In Forbes, Dr. Robert Pearl, former CEO of The Permanente Medical Group, argues that insurers and employers will not invest more in primary care until they see evidence that it pays off. We agree with much of his diagnosis but respectfully disagree with its premise, and we offer the published literature, our own blood pressure control data and our Cigna Tier 1 designation as exhibits.

Two sections turn to Precision Medicine. A new study shows that Alzheimer’s risk prediction sharpens considerably when symptoms guide blood biomarker testing, which supports the layered approach of genetics, symptoms, blood testing and neurology evaluation that we already use. And a new viewpoint from authors of the 2026 ACC/AHA lipid guideline argues that ApoB testing should be selective rather than universal. We explain what ApoB measures, using an analogy we all know too well (the Los Angeles freeway system), and why we check it at every annual exam anyway.

Finally, a Lancet Oncology analysis finds that one in eight cancers worldwide is caused by an infection, with HPV one of the two leading culprits. We review where HPV vaccine policy stands after a year of upheaval in Washington, and what a new national survey reveals about how confused American adults have become about which vaccines they need.

Subscribe now

1. Flu Season Arrives Early: Which Shot, and When

The virus has changed, and it arrived early. Last season (2025–26) was dominated by H3N2, which made up 83.3% of subtyped influenza A viruses. This season (2026–27), H1N1 accounts for 98.4%. Weekly lab-confirmed flu hospitalizations nearly tripled between early September and early October, from 1,133 admissions in the week ending September 5 to 3,334 in the week ending October 3, according to CDC data. The national admission rate rose from 0.3 to 1.0 per 100,000 people.

California leads the country, and the scale is striking. Last week, West Coast test positivity reached 10.7%, against 0.5% a year ago. National positivity was 5%, a 10-year high for this point in the season. Eight Southern states, including Florida and South Carolina, are also elevated, so the West isn’t alone. It is just well ahead. The CDC still describes national activity as within expected seasonal variation.

Adults 65 and older are being hospitalized at the highest rate, 3.3 per 100,000, more than three times the rate in adults aged 50 to 64. This year’s vaccine carries an updated H1N1 component, A/Missouri/11/2025, selected in March, and should be well matched to the circulating strain.

Why the western United States is leading this year is unclear. Flu seasons here usually begin in the Southeast and spread north and west. This season’s H1N1 carries a new mutation, G155E, and small studies suggest antibodies from last season’s infections recognize it poorly, particularly in teenagers and young adults. Population immunity to H1N1 is also less recent. Whether the new mutation affects vaccine protection is not yet known. Vaccination still reduces severe illness and hospitalization, and the case for getting it early is stronger, not weaker.

Three vaccines, three populations

Standard-dose vaccine. This is the conventional egg- or cell-based trivalent shot. It is recommended for everyone 6 months and older, and it remains the right choice for most adults under 65.

High-dose vaccine. This shot contains four times the antigen of the standard dose, which offsets the weaker immune response that comes with age. In its pivotal randomized trial in adults 65 and older, laboratory-confirmed influenza occurred in 1.4% of high-dose recipients versus 1.9% with the standard dose. That is a relative efficacy of about 24%. It remains the standard of care for patients 65 and older.

mRNA vaccine (mFlusiva). After the regulatory saga we covered in our March 2026 newsletter, the FDA approved Moderna’s mRNA-1010 on August 5 for all adults 50 and older. It was the first mRNA influenza vaccine licensed in the United States, following unanimous votes by the FDA’s advisory committee. In the Phase 3 trial, it showed 26.6% greater relative efficacy than the standard-dose shot. It has not been compared head-to-head with the high-dose vaccine in adults 65 and older. Until that comparison exists, high-dose remains our default for that age group. Moderna is distributing the vaccine through select retail pharmacies.

On the horizon. A Phase 3 trial in The Lancet Infectious Diseases found that an adjuvanted, cell-based, higher-dose vaccine (aQIVc) produced stronger antibody responses in adults 50 and older than an adjuvanted egg-based vaccine, including against H1N1. These are antibody results, not proof of fewer infections, and the vaccine is not yet available in the United States.

Figure 1. Seasonal influenza vaccines for adults, 2026–27. Sources: FDA; DiazGranados CA, et al. N Engl J Med 2014 (high-dose trial); Moderna mRNA-1010 Phase 3 data; Essink BJ, et al. Lancet Infect Dis 2026 (aQIVc).

Getting vaccinated at Santa Monica Primary Care

We stock both the high-dose and standard-dose vaccines. You do not need a physician appointment. Call us at (310) 828-4411 to book a nurse visit at any time during office hours, 8:30 am to 5:30 pm. The CDC recommends vaccination by the end of October. With California’s season already well underway, sooner is better: protective antibody levels take about two weeks to develop after the shot.

Reference: Essink BJ, et al. MF59-adjuvanted cell-derived higher-dose quadrivalent influenza vaccine in adults aged 50 years and older: a phase 3 randomised controlled trial. Lancet Infect Dis. 2026.

2. The Return on Primary Care: A Response to Forbes

Dr. Robert Pearl led The Permanente Medical Group, Kaiser’s Northern California physician group, for nearly two decades. In a recent Forbes column, he asks why America doesn’t invest more in primary care. His answer: insurers, employers and Medicare are not convinced that spending more will improve quality and lower costs, because they lack the evidence that it will. Until primary care physicians accept financial risk and prove a return, he argues, the money will not come. The pay gap he cites is familiar to anyone in the field: many specialists earn at least twice what primary care physicians do, if not more.

Much of Pearl’s diagnosis is sound. Fee-for-service rewards volume over outcomes. Pay-for-performance has a poor track record: a 2025 BMJ review found incentivized measures improved at first, faded within three years, and unincentivized care deteriorated along the way. His prescription is capitation, with physicians pooling into groups large enough to absorb financial risk. That is a reasonable direction.

Opinion: Where we differ with Dr. Pearl is on the premise that the evidence is lacking. The case that primary care improves health and lowers costs is substantial and has been building for decades. Barbara Starfield’s landmark 2005 review in the Milbank Quarterly assembled decades of data linking primary care supply to lower mortality, fewer avoidable hospitalizations, and lower total spending, both within the United States and across nations. In 2019, Sanjay Basu and colleagues at Stanford analyzed 3,142 U.S. counties in JAMA Internal Medicine. Each additional 10 primary care physicians per 100,000 people was associated with a 51.5-day gain in life expectancy, compared with 19.2 days for the same increase in specialists. The National Academies’ 2021 report, Implementing High-Quality Primary Care, concluded that primary care is the only component of American health care where greater supply is associated with better population health and more equitable outcomes.

The cost data are no weaker. Our practice participates in the Medicare Shared Savings Program through Aledade, the largest network of independent primary care practices in the country. In 2024, Aledade’s ACOs saved Medicare more than $1 billion. They cared for roughly 11% of program beneficiaries while generating nearly 20% of its shared savings. Of those ACOs, 93% earned shared savings, compared with 73% of ACOs outside the network. The savings are calculated and paid by CMS.

That model also answers Pearl’s concern about risk. He argues that solo and small practices cannot absorb the financial risk of capitation alone. He is right. That is why independent practices like ours pool risk through Aledade, without surrendering clinical autonomy to a hospital system.

Pearl’s own example illustrates the point. He cites CDC data showing that roughly 124 million American adults have hypertension and only 21% have it controlled. He notes that leading health systems reach 80% control, and that achieving this nationally could save as much as $262 billion. The best-documented case is Pearl’s former organization. Kaiser Permanente Northern California raised control among its hypertensive members, more than 650,000 people by 2009, from 44% in 2001 to 80% by 2009. It did so with a patient registry, shared performance data, medical assistant blood pressure visits and single-pill combination medications. Every one of those components is primary care work, and the very work fee-for-service undervalues. Kaiser’s 80% does not necessarily translate to other health systems. UCLA Health, Cedars-Sinai and Providence, for example, do not publish their hypertension control rates.

Our own numbers make the same case. According to our Aledade quality data, 96% of our patients with hypertension have their blood pressure controlled below 140/90 mmHg, by the standard CMS quality measure. That compares with roughly 21% of American adults with hypertension nationally and the 80% Kaiser achieved. The comparison is not like for like: the national figure includes people who are undiagnosed, untreated, or without a regular physician. That is the point. Blood pressure gets controlled when patients are engaged in primary care. Through our ACO collaboration with Aledade, we track and publish our quality metrics, and hypertension control is one of them. Over the next year, that list will expand to include colorectal and breast cancer screening, among others.

Figure 2. Hypertension control (blood pressure below 140/90 mmHg). Sources: Aledade practice quality data, CMS quality measure 165, 2026; Jaffe MG, et al. JAMA 2013;310:699–705 (Kaiser Permanente Northern California, 2009); CDC (U.S. adults with hypertension).

Payers do measure value, practice by practice. Cigna’s tiered benefit design steers members toward providers it rates highly on quality and cost efficiency. Each provider is compared against peers in the same specialty and market. Members who choose a Tier 1 provider pay a lower copay or coinsurance. Santa Monica Primary Care has been designated a Cigna Tier 1 provider for 2026. The open question is whether they will invest in it at scale.

Pearl is right that primary care must demonstrate its value to those who pay for it; much of that demonstration already exists.

3. Alzheimer’s Risk: Symptoms First, Then Blood

Blood tests for Alzheimer’s disease have become accurate enough that the debate has shifted from whether they work to whom to test. A new study in Dementia and Geriatric Cognitive Disorders by Matthew Howe and colleagues, a Brown University group focused on plasma biomarkers, makes a practical case: start with symptoms.

The investigators used data from the Alzheimer’s Disease Neuroimaging Initiative (ADNI). They built their model in ADNI-2 (603 participants) and tested it in an independent ADNI-3 cohort (378 participants). The outcome was progression to mild cognitive impairment or dementia with confirmed amyloid pathology over four years. A machine learning model converted participants’ subjective ratings of their own cognition into a baseline symptom-burden score.

Symptoms alone predicted progression reasonably well, with an area under the curve (AUC) of 0.79. Adding blood biomarkers raised the AUC to 0.90. Participants with both elevated symptom scores and abnormal biomarkers had a four-year progression incidence of 72.1%, a 19-fold higher risk than their peers. The findings held up in the validation cohort.

ADNI participants are research volunteers who skew highly educated, and its assays differ from commercial tests. Still, the design mirrors real clinical care and aligns with the Alzheimer’s Association’s first blood-biomarker guideline (2025), which positions these tests for patients with objective cognitive impairment, not population screening.

How this fits our approach

The study supports the layered approach we already use, in which each tool answers a different question (Figure 3). Genetics (ApoE genotyping and an Alzheimer’s polygenic risk score through our collaboration with Biography Health) identifies inherited risk but not disease; many ApoE4 carriers never develop dementia. Symptoms, new memory complaints, especially when family notices them too, are the trigger, the role this study gives them. Blood biomarkers, the Labcorp ATN Profile we reviewed in our September 2026 newsletter, tell us whether pathology is present. Neurology confirms with neuropsychological testing and amyloid PET or spinal fluid analysis, which is still required before disease-modifying antibody therapy.

Figure 3. A layered approach to Alzheimer’s risk at Santa Monica Primary Care, with predictive performance of symptoms and blood biomarkers. Source: Howe MD, et al. Dement Geriatr Cogn Disord 2026.

A positive blood test in someone with no symptoms answers a question nobody asked and raises several that no one can yet answer. A positive test in someone with symptoms changes management.

Reference: Howe MD, Zide BS, Hayes K, Ross D, Barker MS, Kunicki ZJ, Rudolph J, Huey E. Symptom-Guided Blood Biomarker Testing for Alzheimer’s Disease. Dement Geriatr Cogn Disord. 2026 Oct 12 [Epub ahead of print].

4. ApoB: Why Your LDL Number Isn’t the Whole Story

For decades, LDL cholesterol has been the number patients know and physicians treat. It appears on every standard lipid panel, it anchors every guideline, and lowering it saves lives. But LDL cholesterol measures how much cholesterol is being carried, not how many particles are carrying it, and it is the particles that drive atherosclerosis. Apolipoprotein B (ApoB) counts those particles directly. A new viewpoint from authors of the 2026 ACC/AHA lipid guideline argues that most patients do not need it. We check it in every patient at their annual exam. Both positions have a logic, and the difference comes down to who is asking.

Think of your bloodstream as the 405 at rush hour. The vehicles are lipoprotein particles: LDL, VLDL, IDL and Lp(a). Each one carries cholesterol as cargo. Atherosclerosis is what happens when vehicles crash through the guardrail and lodge in the artery wall. The more vehicles on the road, the more crashes.

The three lipid measures we discuss with patients count different things on that freeway.

• LDL-C (LDL cholesterol) measures the total cargo carried by LDL vehicles. It is the number on a standard lipid panel, and it is usually calculated rather than measured directly.

• LDL-P (LDL particle number) counts the LDL vehicles themselves, regardless of how much each one carries.

• ApoB (apolipoprotein B) counts every atherogenic vehicle on the road, LDL and the others alike. Each of these particles carries exactly one ApoB molecule, like a license plate, so measuring ApoB is a census of all the traffic capable of causing a crash. It is the closest thing we have to a single aggregate measure of “bad cholesterol.”

Usually these measures agree. Sometimes they don’t. Insulin resistance, diabetes and high triglycerides tend to produce many small vehicles each carrying little cargo. In that situation, LDL-C can look reassuring while the freeway is gridlocked.

Figure 4. LDL-C, LDL-P and ApoB as cargo and vehicles on the freeway. Two patients with identical LDL-C can carry very different numbers of atherogenic particles. Illustrative schematic; values are not patient data.

The case for selective testing

The viewpoint, from Johns Hopkins, Emory, Mass General Brigham and others and published in JACC: Advances, argues that ApoB should be measured selectively rather than universally. Its senior author is Dr. Roger Blumenthal, who chaired the 2026 ACC/AHA lipid guideline we summarized in our March 2026 newsletter.

Their argument rests on better math. Much of the classic discordance between LDL-C and ApoB was documented with the old Friedewald equation, which underestimates LDL-C at low levels and when triglycerides are high. Friedewald also assumes the patient is fasting. Triglycerides rise after a meal, and because the equation subtracts a triglyceride-based estimate, a non-fasting sample can push the calculated LDL-C artificially low. Newer methods, the Martin-Hopkins and NIH Sampson equations, close most of that gap and hold up better in non-fasting samples. In people without obesity or diabetes, the median gap between ApoB and LDL-C is essentially zero. Among people without severe hypertriglyceridemia who reach a Martin-Hopkins LDL-C below 70 mg/dL, fewer than 10% still have an ApoB of 70 mg/dL or above.

The picture changes in metabolically higher-risk patients. Among those on a statin or with diabetes, metabolic syndrome or high triglycerides, roughly 25% to 50% with an LDL-C below 70 mg/dL still carry an elevated ApoB of 60 to 80 mg/dL. That is the population the 2026 guideline targets:

• Class 2a: measure ApoB in patients already on lipid-lowering therapy, particularly those with established cardiovascular disease, cardiovascular-kidney-metabolic syndrome, type 2 diabetes or elevated triglycerides, to decide whether to intensify treatment once LDL-C goals are met.

• Class 2b: measuring ApoB in patients not on therapy “may be reasonable.”

The authors offer a practical rule: if ApoB exceeds LDL-C by 5 mg/dL or more, intensifying therapy toward a lower LDL-C target is reasonable. For the highest-risk patients, simply aiming for an LDL-C below 50 mg/dL eliminates most of the discordance.

Why we check it every year

The authors’ logic is sound for population-level policy, where a $75 test multiplied across millions of people adds up. For an individual patient, the calculation looks different. A large share of our patients fall into the groups the guideline itself prioritizes: those on statins, those with metabolic risk, those with elevated triglycerides or Lp(a). ApoB is part of every annual exam at Santa Monica Primary Care, alongside our extended lipid panel. For most patients it confirms what LDL-C already shows. For a meaningful minority, it reveals heavy traffic that the cargo count missed, and it changes their treatment.

Reference: Razavi AC, Shapiro MD, Natarajan P, Gianos E, Martin SS, Blumenthal RS. Prioritizing Selective Use of Apolipoprotein-B Testing in the Era of Modern LDL-Cholesterol Calculation. JACC: Advances. 2026;103220.

5. Preventable Cancers, Confusing Policy: HPV and Adult Vaccination

A new worldwide analysis from the International Agency for Research on Cancer (IARC), published in The Lancet Oncology, estimates that 2.3 million new cancer cases in 2024 were caused by infections. That is 12% of all cancers diagnosed worldwide, or one in eight. Five pathogens account for nearly all of them, led by Helicobacter pylori and HPV (Figure 5).

HPV alone causes essentially every cervical cancer worldwide, roughly 600,000 cases a year. The study added 16 infection-cancer links not included in the previous 2018 estimate and is the first global analysis to incorporate HIV.

Figure 5. New cancer cases attributable to the five leading infectious agents worldwide, 2024. Source: Rumgay H, et al. Lancet Oncol 2026;27:1237–1248.

More than three-quarters of these cancers occur in low- and middle-income countries, where, the authors note, proven prevention tools are underused because of “societal, financial, and political barriers.” In the United States, the barrier is increasingly political.

HPV in the United States

HPV causes more than 36,500 cancers in the U.S. each year. Most people think of cervical cancer, but the most common HPV-associated cancer in this country is now oropharyngeal cancer of the back of the throat. HPV causes about 70% of those cases, and men are diagnosed far more often than women. More than four in ten HPV-caused cancers occur in men.

The vaccine works. Since it was introduced, infections with the HPV types that cause most cancers have fallen 88% among teenage girls and 81% among young adult women. Cervical precancers caused by the most dangerous types have dropped 40% among vaccinated women.

Federal policy

Robert F. Kennedy Jr. called Gardasil “the most dangerous vaccine ever invented” in a 2019 video. His financial ties to litigation against its manufacturer drew scrutiny at his confirmation hearings. As HHS Secretary, he fired all 17 ACIP members in June 2025. A new ACIP workgroup was then chartered to re-examine the HPV vaccine’s age, dosing, safety and efficacy, which no reputable expert body considered necessary.

On January 5, 2026, HHS rewrote the childhood schedule without an ACIP vote. HPV went from two doses to one. In fairness, the one-dose science is real. A large NIH-funded trial in Costa Rica, published in the New England Journal of Medicine, found a single dose provided protection against persistent infection that was statistically no worse than two doses, with at least 97% efficacy against HPV-16 and HPV-18. The previous ACIP was weeks away from a formal vote on that very question when it was dissolved. The problem was the process: a schedule change imposed from above, with no FDA labeling to support a single-dose regimen and no formal deliberation behind it.

On March 16, a federal judge in Massachusetts stayed the January schedule, the appointments of the new ACIP members, and every vote they had taken. The government appealed only the ACIP portion, and the First Circuit heard arguments on October 6. Whatever that court decides, the schedule pause remains in place.

For now, the pre-2026 recommendations stand:

• Routine vaccination at ages 11 to 12, which can begin as early as age 9.

• Two doses if the series starts before age 15. Three doses if it starts at 15 or later, or for anyone who is immunocompromised.

• Catch-up vaccination for everyone through age 26.

• Adults 27 to 45: vaccination based on shared decision-making with your physician. Many adults in this group have already been exposed, which reduces the benefit, but some will still gain protection.

The adult picture: support without clarity

The damage from shifting federal guidance shows up most clearly in adults. In an August 2026 NORC survey of 1,206 U.S. adults for the National Foundation for Infectious Diseases, 77% agreed that vaccines are among the best protections against serious infectious disease. Yet only 48% tried to get any of the four major respiratory vaccines in the past year: 44% for flu, 23% for COVID-19, 10% for pneumococcal and 9% for RSV.

The gap was driven less by hesitancy than by confusion. More than a quarter of respondents (28%) either believed the flu vaccine was not recommended for someone like them or did not know. For COVID-19, that figure reached 49%. When a federal schedule is rewritten by memo, frozen by a court, and appealed in turn, it is unsurprising that half of American adults cannot say whether a vaccine applies to them. Most are willing to be vaccinated; they just need clear guidance. More than half (55%) named their healthcare professional as their most trusted source.

What this means for our patients

If you are 26 or younger and not fully vaccinated against HPV, complete the series. If you are 27 to 45 and at risk of new exposure, it is worth a conversation. If you have children or grandchildren approaching age 9, start the discussion with their pediatrician now, since the vaccine works best before any exposure.

For most of our adult patients, the vaccines that matter this fall are the respiratory ones. Flu vaccine is recommended for everyone (see Section 1). Pneumococcal and RSV vaccines depend on age and health conditions. Hepatitis B vaccination is recommended for adults through age 59, and a one-time hepatitis C screening is recommended for every adult. If you are unsure which of these apply to you, ask us. We can review your eligibility and send a prescription to your pharmacy. Most adults, the survey suggests, never ask.

The most effective cancer vaccine ever developed is twenty years old, and its greatest remaining obstacle is not science. The upshot of all the confusion is already visible: U.S. measles cases have passed 4,000 this year, 78% above all of 2025, with deaths roughly doubled.

Reference: Rumgay H, Georges D, Huang Y, et al. Global burden of cancer attributable to infections in 2024: a worldwide incidence analysis. Lancet Oncol. 2026;27:1237–1248.

6. Follow Us On Social Media

Twitter (X): @santa_care

Facebook: https://www.facebook.com/santamonicaprimarycare

Instagram: @santa_monica_primary_care

LinkedIn: https://www.linkedin.com/in/philip-bretsky-mdphd/

BlueSky: @bretskymd.bsky.social

Substack: @bretskymd

Follow our blog posts on: www.drbretsky.com/blog

Archived issues of our newsletter can be found at: https://drbretsky.com/newsletters/

If you know somebody who would like to sign up to receive our newsletter, they can do so at: https://lp.constantcontactpages.com/su/JtF9zzR

This edition starts with LDL, because Lp(a) is best understood as an LDL particle with an attachment. From there we cover how Lp(a) builds soft plaque, what five years of annual testing at our practice has shown us, the imaging that matters, and why a well-run trial can fail.

Subscribe now

1. Starting With LDL

Cholesterol has an undeserved reputation. It is essential. Every cell membrane in the body is a lipid bilayer, and cholesterol sits between its phospholipids, making up roughly a third of the membrane’s lipid molecules and governing how fluid or rigid it is. Cholesterol is the raw material for steroid hormones (estrogen, testosterone, cortisol), vitamin D, and bile acids. It is a major component of myelin, the insulating sheath around nerve fibers, and the brain holds about a quarter of the body’s cholesterol. The brain, however, makes its own behind the blood–brain barrier, which LDL does not cross. That distinction answers a persistent piece of misinformation: lowering LDL in the blood does not starve the brain of cholesterol.

Cholesterol does not dissolve in blood, so it travels inside lipoproteins. Low-density lipoprotein (LDL) delivers cholesterol to tissues, and each LDL particle carries exactly one copy of apolipoprotein B (apoB). HDL, built around apolipoprotein A1 (apoA1), runs largely in reverse, returning cholesterol to the liver. The liver clears LDL through its LDL receptors, and our main therapies all work through that pathway. Statins cut the liver’s own cholesterol production, prompting it to make more receptors. Ezetimibe (Zetia) blocks absorption of cholesterol from the gut. PCSK9 inhibitors (Repatha, Praluent) prevent the destruction of LDL receptors, so each one is recycled more times.

The problem is excess. Surplus apoB particles enter the artery wall and stay there. Across 26 statin trials, every 39 mg/dL (1 mmol/L) drop in LDL cholesterol produced a 22% relative reduction in major vascular events. That number is the benchmark against which every new lipid drug is measured.

The standard lipid panel misses important detail. Two people with identical LDL cholesterol can carry very different numbers of particles. Someone whose LDL comes packed into many small, dense particles has more apoB particles available to enter the artery wall. The 2026 ACC/AHA guideline acknowledges this. For the first time, it brings apoB into a U.S. guideline as a recommended measurement, particularly in patients with diabetes, high triglycerides, or very low LDL cholesterol, in whom the standard LDL estimate can understate true risk. It also restores numeric treatment goals, with apoB goals matched to LDL goals by risk category:

• Very high risk: LDL cholesterol <55 mg/dL, apoB <55 mg/dL

• High risk: LDL cholesterol <70 mg/dL, apoB <70 mg/dL

• Borderline to intermediate risk: LDL cholesterol <100 mg/dL, apoB <90 mg/dL

When LDL and non-HDL goals are met but apoB remains elevated, the guideline recommends intensifying treatment further.

At SMPC we have measured beyond the standard panel for years. Our lipid cascade includes LDL particle number, particle subtype and size, apoB, apoA1, and Lp(a). ApoB counts every atherogenic particle, apoA1 reflects the protective HDL side, and particle size and number reveal discordance that a standard LDL cholesterol result conceals.

2. Lp(a): The Basic Science

Lp(a) is an LDL particle with a second protein attached. The core is standard: cholesterol inside, one apoB on the surface. Bolted onto the apoB by a single disulfide bond is apolipoprotein(a), a long chain of looped protein segments called “kringles,” copied almost directly from plasminogen, the precursor of the body’s clot-dissolving enzyme. The attachment gives Lp(a) three properties LDL lacks.

• Inflammation. Lp(a) is the preferred carrier of oxidized phospholipids (OxPL), chemically damaged fats that activate the vessel lining and recruit inflammatory cells.

• Clotting. Apo(a) resembles plasminogen but has no clot-dissolving activity. It competes with plasminogen, so a clot that forms on a ruptured Lp(a)-rich plaque is harder to break down.

• Retention. Apo(a) grips fibrin and the connective tissue of the artery wall, giving Lp(a) a second anchor once it enters.

Lp(a) levels are set almost entirely by the LPA gene, specifically by the number of repeating kringle IV type 2 segments, which ranges from one to more than 40. Fewer repeats produce a smaller apo(a) that the liver secretes more efficiently, and therefore a higher blood level. Plasma Lp(a) is 70–90% heritable, largely unresponsive to lifestyle, and poorly controlled with statins and PCSK9 inhibitors. The reason is mechanical: our drugs accelerate clearance through the LDL receptor, while Lp(a) levels are governed mainly by production. Statins may even raise Lp(a) modestly.

Particle for particle, Lp(a) is more dangerous than LDL. Genetic analyses anchored to apoB estimate that each Lp(a) particle carries roughly six times the risk of an LDL particle. But Lp(a) particles are far fewer in number, a point that becomes central when we reach the trial.

3. Where Lp(a) Goes, and How It Becomes Soft Plaque

Plaque does not form evenly along the coronary arteries. It concentrates where blood flow is slow, reversing, or turbulent (low or oscillating shear stress): the outer walls of branch points, the inner side of curves, and proximal segments. In practice, that means the left main bifurcation, the first few centimeters of the left anterior descending (LAD) artery and its diagonal branches, the circumflex–marginal junction, and the proximal right coronary artery. At these sites the endothelium, the vessel’s inner lining, is more permeable, and apoB particles small enough to cross it, LDL and Lp(a) alike, slip into the wall.

Once inside, apoB binds to proteoglycans, the sugar-coated proteins of the wall, and apo(a) adds its second anchor. The OxPL cargo activates the overlying endothelium. Circulating monocytes attach, migrate into the wall, and become macrophages, which engulf the trapped, oxidized particles until they become lipid-laden foam cells. When foam cells die, they spill their contents, forming a lipid-rich necrotic core beneath a thin fibrous cap.

This is soft plaque: noncalcified, lipid-rich, and the type most prone to rupture. It grows outward rather than inward (positive remodeling), so the channel for blood flow can look normal on a stress test or even an angiogram. Calcium is laid down years later. Calcification tends to stabilize a plaque, and it is the only component a calcium score detects.

Imaging shows Lp(a) favoring this soft phenotype. In a CT angiography substudy of the DIAMOND trial (Kaiser et al., JACC 2022), patients with advanced multivessel disease and Lp(a) of 70 mg/dL or higher showed accelerated progression of low-attenuation plaque. In one example patient with Lp(a) of 82 mg/dL, low-attenuation plaque volume rose from 81.4 to 132.9 mm³ in a single year. Low-attenuation plaque is the CT signature of a necrotic core, and in SCOT-HEART it was the strongest predictor of heart attack, outperforming risk scores, stenosis severity, and the calcium score.

4. How Much Heart Disease Is Lp(a)?

About 1 in 5 people worldwide have elevated Lp(a), and nearly a third of people who develop heart disease early in life do. At the individual level, people with Lp(a) above 50 mg/dL had a 35% higher risk of cardiovascular events than those below 15 mg/dL in a meta-analysis of statin trials. Risk concentrates at the extremes: in the Copenhagen studies, excess risk sat primarily in the top 5%, above roughly 90 mg/dL.

At the population level, the picture is more modest. The population attributable fraction, the share of all cardiovascular disease that would disappear if elevated Lp(a) did not exist, was 5.8% for Lp(a) above 100 nmol/L in UK Biobank. It varies by ancestry: in INTERHEART, South Asians had both the highest odds ratio for heart attack and the highest attributable risk, at 10%. By comparison, the apoB/apoA1 ratio in INTERHEART accounted for roughly half of heart attacks. Lp(a) is a major risk factor for a minority of individuals and a minor one for the population. An Lp(a) drug was never going to be the next statin for everyone; the hope was a missing statin for one fifth of patients.

5. What We Have Seen at SMPC: Lp(a) Moves

Guidelines recommend measuring Lp(a) once in a lifetime, on the premise that it is genetically fixed. We have included Lp(a) in routine annual labs at Santa Monica Primary Care for about five years. In most patients the convention holds: levels are highly genetic and generally stable. In others, we have seen wide swings without a clear explanation.

The literature is catching up. In a Nashville cohort of 609 people tested twice on the same assay, Lp(a) changed by 10 mg/dL or more in 38% of individuals, with larger changes in women and in Black individuals. The menopausal transition is one identifiable driver. In the Copenhagen General Population Study, which followed more than 37,000 women and 32,000 men, Lp(a) rose with age in both sexes, with an additional increase in women around age 50. Levels were 27% higher after menopause and 12% lower during hormone replacement therapy. Other contributors include thyroid and kidney disease, acute inflammation (Lp(a) behaves partly as an acute-phase reactant), and the lab itself: assays differ in how they handle apo(a) size variation, so a change of laboratory can masquerade as a change in the patient.

A single lifetime value is adequate for most people. It is not adequate for a woman in her 40s or 50s, a patient near a treatment threshold, or anyone whose number has moved. Repeat testing costs little and occasionally changes a patient’s risk category.

On treatment, our experience matches the trial data: patients on PCSK9 inhibitors see roughly 20–30% reductions in Lp(a), in line with the median 27% reduction in FOURIER and ODYSSEY Outcomes. The effect is real but modest, and it was smaller in FOURIER patients with the highest Lp(a).

6. Imaging: Why Lp(a) Calls for CT Angiography, and What It Costs

In our March 2026 edition we discussed the joint risk of Lp(a) and coronary calcium. A calcium score (CAC) remains a useful, inexpensive tool. But a calcium score can detect, well, only calcium. Lp(a)’s characteristic lesion is soft, noncalcified plaque. A score of zero rules out calcified plaque; it does not rule out a lipid-rich plaque that has not yet calcified, the stage at which Lp(a) appears to do its damage.

Coronary CT angiography (CCTA) with AI-quantified plaque analysis, which we use through Cleerly, measures noncalcified and low-attenuation plaque volume directly and establishes a baseline against which progression can be tracked.

The downside is cost. A calcium score runs about $150; a CCTA with Cleerly analysis, about $1,500. CCTA also requires intravenous contrast and a modestly higher radiation dose.

For patients without symptoms, insurance generally won’t pay for either scan. Medicare’s position is explicit: a calcium score done on its own is a “screening service” and denied as “not medically necessary.” Private insurers largely follow the U.S. Preventive Services Task Force, which has not endorsed calcium scoring (its last review dates to 2018), so the Affordable Care Act requires no coverage. This is despite the 2026 ACC/AHA guideline giving CAC scoring its strongest (Class 1) recommendation for men 40 and older and women 45 and older. Once chest pain starts, CCTA is covered. Before chest pain, when the scan might have prevented it, it is not.

A few states have stepped in. Texas has required since 2009 that insurers pay up to $200 every five years for a calcium score or carotid ultrasound in men 45–76 and women 55–76, plus anyone with diabetes or at intermediate or higher risk. New Mexico requires coverage of a calcium scan every five years for adults 45–65 at intermediate risk. California has no such requirement. Our health system will pay generously for a heart attack and grudgingly, if at all, for the scan that might have prevented one.

For most patients, a calcium score is a sensible first look. For patients with elevated Lp(a), a strong family history of early heart disease, or a calcium score of zero that does not fit the rest of their risk profile, the additional information justifies the additional cost.

7. Lp(a)HORIZON: Design, Result, and Why It May Have Missed

The trial. Lp(a)HORIZON randomized 8,323 patients with established cardiovascular disease and Lp(a) of at least 70 mg/dL to pelacarsen or placebo. The primary endpoint was a composite of cardiovascular death, heart attack, stroke, and urgent coronary revascularization, with a prespecified higher-risk subgroup at Lp(a) of 90 mg/dL or above. Participants injected themselves monthly, with follow-up extending to seven years.

The result. Lp(a) fell. Events did not. Only topline results have been released; event rates, hazard ratios, the 90 mg/dL subgroup, and the achieved Lp(a) reduction are pending, with full results expected at a major cardiology meeting in November.

The New York Times coverage. Gina Kolata covered the story twice, including a September 11 follow-up titled “A Promising Heart Drug Failed. What Went So Wrong?” Cardiologists had hoped the drug could be as important as a statin, and when Novartis released its results late on a Friday afternoon, after the markets closed, they were shocked. Duke cardiologist Dr. John Alexander put it simply: “There’s something we don’t understand here.” Yale’s Dr. Harlan Krumholz told the Times that even strong evidence of two things occurring together is never proof that one causes the other.

Lp(a)’s case does rest on more than correlation. Mendelian randomization studies, which use inherited gene variants as a natural randomized trial, support causation. But they show that a lifetime of high Lp(a) causes disease, not that lowering it late in life undoes the damage. Alzheimer’s disease offers a close parallel. Amyloid is causally tied to the disease; inherited mutations in the amyloid pathway cause early-onset Alzheimer’s. Lecanemab (Leqembi) clears amyloid from the brain effectively. Yet in its pivotal trial it slowed decline by 0.45 points on an 18-point dementia scale over 18 months, a 27% relative slowing that many patients and families would struggle to notice, and it carries a risk of brain swelling and microbleeds. Its predecessor, aducanumab, won accelerated approval on amyloid reduction alone and was withdrawn in 2024. A causal molecule, effectively cleared, produced only a modest clinical benefit. Removing a lifelong exposure late in the disease may simply come too late.

What lipid drugs actually deliver.

• Statins: 22% relative risk reduction per 39 mg/dL (1 mmol/L) of LDL lowering, across 26 trials.

• Ezetimibe (IMPROVE-IT): LDL fell from 69.5 to 53.7 mg/dL; events at seven years were 32.7% vs 34.7%, a 6.4% relative reduction. Demonstrating it required about 18,000 patients followed for a median of six years.

• PCSK9 inhibition (FOURIER): LDL fell to a mean of 30 mg/dL; the primary endpoint fell by 15%, and cardiovascular death, heart attack, or stroke by 20%, over a median of 2.2 years.

Lp(a)HORIZON assumed that an 80% cut in Lp(a) would produce a 20% relative risk reduction. That is more than a PCSK9 inhibitor delivered, from a trial less than a third the size of FOURIER.

The cost of immodest design. Our March 2026 edition made the same point about the Galleri multi-cancer blood test. Its NHS trial assumed a 20% reduction in late-stage cancers over roughly two years, then used a 3:1 randomization that further reduced statistical power. If the true effect had been 10%, the trial had about a 20% chance of detecting it. The failure mode is the same in both cases. Phase 2 results tend to flatter: early studies are small, and the drugs or tests that advance are the ones that happened to look best (the “winner’s curse”). Investigators, understandably confident, size the definitive trial around an optimistic effect, which keeps it smaller, faster, and cheaper. If the real effect is half that size, the trial is underpowered and a modest but real benefit is recorded as a null result.

A negative trial of this kind does not prove that a drug or test is ineffective. It proves that it is not as good as its sponsors believed. Those are very different conclusions, and the full dataset, particularly the confidence interval around the hazard ratio and the 90 mg/dL subgroup, will tell us which applies here.

Other design features that may have diluted the signal:

• Absolute, not relative, lowering. Lp(a) particles are far fewer than LDL particles, so the absolute drop matters. Genetic estimates suggest roughly a 100 mg/dL fall is needed to match lowering LDL by 39 mg/dL. An 80% cut from the 70 mg/dL entry threshold is only 56 mg/dL.

• Late intervention. Every participant already had established vascular disease.

• Excellent background care. Mean LDL was around 66 mg/dL. The better the baseline treatment, the fewer events remain for any new drug to prevent.

• A dilute endpoint. Urgent revascularization and stroke may be less sensitive to Lp(a) than heart attack.

The surrogate endpoint debate. In Outlive, Dr. Peter Attia argues that because atherosclerosis develops over decades while outcome trials capture only a few years, we should act on biomarkers such as apoB rather than wait for events. This is a point on which Dr. Attia and I disagree. The time argument has some merit: in FOURIER, the reduction in cardiovascular death, heart attack, or stroke grew from 16% in the first year to 25% beyond it. But Lp(a)HORIZON ran for up to seven years, and the biomarker still did not translate.

My objection is an epidemiologist’s. A biomarker tells you what happened to a molecule. Event reduction tells you what happened to people, and at the population level that is what matters: heart attacks prevented, the number needed to treat, the harms, and the cost of treating millions of people for decades. Medicine’s history of surrogate failures is long:

• CAST: antiarrhythmic drugs suppressed abnormal heartbeats on monitoring and increased deaths.

• Rosiglitazone (Avandia): approved on the strength of lowering blood sugar, then linked by a 2007 Cleveland Clinic meta-analysis of 42 trials (Nissen and Wolski) to 43% higher odds of heart attack.

• Niacin: improved nearly every lipid number, including Lp(a), then failed two outcome trials.

• Torcetrapib: raised HDL and increased mortality.

• Aducanumab: cleared amyloid and was withdrawn.

LDL and apoB have earned surrogate status, because dozens of trials across multiple mechanisms show that lowering them reduces events in proportion. Lp(a) has not yet earned it. Plaque imaging offers a middle path: Eli Lilly began a trial this June testing whether lepodisiran reduces coronary plaque in patients with elevated Lp(a).

What remains in the pipeline. Two large outcome trials continue: OCEAN(a) with olpasiran (Amgen) and ACCLAIM-Lp(a) with lepodisiran (Eli Lilly). Both lower Lp(a) more deeply than pelacarsen, by more than 95% and 93.9% respectively. If the problem was insufficient lowering, these trials will show it.

8. How We Approach Lp(a) and Cholesterol at SMPC

A failed drug does not erase a risk factor. Elevated Lp(a) still identifies people at higher lifetime risk; what it changes is how hard and how early we manage everything else. Our approach combines four layers.

Advanced biomarkers, measured over time. Our annual labs include the full lipid cascade described in Section 1. We repeat Lp(a) every year rather than once in a lifetime, because we have seen it move, particularly around menopause.

Genetics. Through our collaboration with Biography Health (www.biographyhealth.com), polygenic risk scores add an inherited dimension that a single blood test cannot capture. Lp(a) is one inherited risk; polygenic scores capture many others.

Imaging matched to risk. We start with a calcium score for most patients. For those with elevated Lp(a), a strong family history, or results that do not add up, we use CCTA with Cleerly to measure soft plaque directly and track it over time.

Personalized risk reduction. When Lp(a) is elevated, we set lower apoB and LDL targets than the guideline defaults and reach them with statins, ezetimibe, PCSK9 inhibitors, or bempedoic acid as needed. Blood pressure, glucose, and lifestyle get the same intensity. Where we can see plaque, treatment is guided by the plaque rather than by a risk calculator alone.

We cannot yet lower Lp(a) enough to matter. We can find out what it has done to a patient’s arteries and treat the rest of the risk profile accordingly.

9. Follow Us On Social Media

Twitter (X): @santa_care

Facebook: https://www.facebook.com/santamonicaprimarycare

Instagram: @santa_monica_primary_care

LinkedIn: https://www.linkedin.com/in/philip-bretsky-mdphd/

BlueSky: @bretskymd.bsky.social

Substack: @bretskymd

Follow our blog posts on: www.drbretsky.com/blog

Archived issues of our newsletter can be found at: https://drbretsky.com/newsletters/

If you know somebody who would like to sign up to receive our newsletter, they can do so at: https://lp.constantcontactpages.com/su/JtF9zzR

Literature Cited

News and trial announcements

1.Novartis. Lp(a)HORIZON Phase III topline results for pelacarsen. Press release, September 4, 2026.

2.Kolata G. Heart drug fails in Novartis trial. New York Times, September 4, 2026.

3.Kolata G. A promising heart drug failed. What went so wrong? New York Times, September 11, 2026.

4.Cox CE. Lp(a)HORIZON: top-line results dash hopes for pelacarsen. TCTMD, September 8, 2026.

5.Design and rationale of the Lp(a)HORIZON trial. PubMed 40185318.

LDL, apoB and guidelines

6.Cholesterol Treatment Trialists’ Collaboration. Efficacy and safety of more intensive lowering of LDL cholesterol. Lancet 2010;376:1670–81.

7.Blumenthal RS, Morris PB, et al. 2026 ACC/AHA/Multisociety Guideline on the Management of Dyslipidemia. Circulation 2026.

Lp(a) biology and risk

8.Kronenberg F, et al. Lipoprotein(a) in atherosclerotic cardiovascular disease and aortic stenosis: EAS consensus statement. Eur Heart J 2022;43:3925–46.

9.Björnson E, et al. Lipoprotein(a) is markedly more atherogenic than LDL. J Am Coll Cardiol 2024.

10.Tsimikas S, et al. Statin therapy increases lipoprotein(a) levels. Eur Heart J 2020.

11.Considerations for routinely testing for high lipoprotein(a) (PCSK9 inhibitors and Lp(a)). 2023.

12.Willeit P, et al. Baseline and on-statin Lp(a) and cardiovascular risk: individual patient-data meta-analysis. Lancet 2018;392:1311–20.

13.Welsh P, et al. Lipoprotein(a) and cardiovascular disease: prediction, attributable risk fraction, and estimating benefits from novel interventions. Eur J Prev Cardiol 2021;28:1991–2000.

14.Paré G, et al. Lipoprotein(a) levels and the risk of myocardial infarction among 7 ethnic groups (INTERHEART). Circulation 2019;139:1472–82.

15.Yusuf S, et al. Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (INTERHEART). Lancet 2004;364:937–52.

16.Burgess S, et al. Association of LPA variants with risk of coronary disease and the implications for Lp(a)-lowering therapies. JAMA Cardiol 2018;3:619–27.

Lp(a) variability

17.Harb T, et al. Intra-individual variability in lipoprotein(a). Eur Heart J Open 2024;4:oeae064.

18.Simony SB, et al. Sex differences of lipoprotein(a) levels and associated risk by age: the Copenhagen General Population Study. Atherosclerosis 2022.

Imaging and coverage

19.Kaiser Y, et al. Association of lipoprotein(a) with atherosclerotic plaque progression. J Am Coll Cardiol 2022;79:223–33.

20.Williams MC, et al. Low-attenuation noncalcified plaque on coronary CT angiography predicts myocardial infarction (SCOT-HEART). Circulation 2020;141:1452–62.

21.Bhatia HS, et al. Joint association of Lp(a) and coronary artery calcium. J Am Coll Cardiol 2026.

22.Centers for Medicare & Medicaid Services. Local Coverage Determination L35121: Coronary CT Angiography.

23.Texas Legislature. HB 1290 bill analysis (Texas Heart Attack Prevention Bill). 2009.

24.Blaha MJ, et al. Insurance payers should cover selective coronary artery calcium testing in intermediate-risk primary prevention patients. Circulation.

Lipid outcome trials and trial design

25.Cannon CP, et al. Ezetimibe added to statin therapy after acute coronary syndromes (IMPROVE-IT). N Engl J Med 2015;372:2387–97.

26.Sabatine MS, et al. Evolocumab and clinical outcomes in patients with cardiovascular disease (FOURIER). N Engl J Med 2017;376:1713–22.

27.Schwartz GG, et al. Alirocumab and cardiovascular outcomes after acute coronary syndrome (ODYSSEY Outcomes). N Engl J Med 2018;379:2097–107.

28.Clinical trial design for lipoprotein(a)-lowering therapies: JACC Focus Seminar 2/3. J Am Coll Cardiol 2023.

Surrogate endpoints

29.Echt DS, et al. Mortality and morbidity in patients receiving encainide, flecainide, or placebo (CAST). N Engl J Med 1991;324:781–8.

30.Nissen SE, Wolski K. Effect of rosiglitazone on the risk of myocardial infarction and death from cardiovascular causes. N Engl J Med 2007;356:2457–71.

31.AIM-HIGH Investigators. Niacin in patients with low HDL cholesterol levels receiving intensive statin therapy. N Engl J Med 2011;365:2255–67.

32.HPS2-THRIVE Collaborative Group. Effects of extended-release niacin with laropiprant in high-risk patients. N Engl J Med 2014;371:203–12.

33.Barter PJ, et al. Effects of torcetrapib in patients at high risk for coronary events (ILLUMINATE). N Engl J Med 2007;357:2109–22.

34.van Dyck CH, et al. Lecanemab in early Alzheimer’s disease (CLARITY AD). N Engl J Med 2023;388:9–21.

35.Attia P, with Gifford B. Outlive: The Science and Art of Longevity. Harmony Books; 2023.

Pipeline

36.Novel lipoprotein(a) therapies: a comprehensive review. 2026.

37.OCEAN(a)-Outcomes, olpasiran (NCT05581303).

38.ACCLAIM-Lp(a), lepodisiran (NCT06292013).

39.Lepodisiran coronary plaque trial (NCT07613294).

Practice collaboration

40.Biography Health. www.biographyhealth.com.

Prior SMPC newsletters

41.March 2026: The Clinical Trials Edition (Galleri, Moderna, lipid guidelines).

A new study also puts one of our oldest tools, coronary calcium scoring, to an actual test — confirming that it still needs to be pointed at the right patient to matter, and that it isn’t the whole picture on its own.

We update our own six-year COVID-19 case log, this issue which shows the same late-summer pattern reasserting itself for the sixth year running. We also cover a newly available option for preventing infection after a known exposure, for anyone with travel or gatherings ahead this fall — along with what it actually costs.

We’re highlighting a fuller “Conditions We Treat” library on our website this month, built around real evidence and, in at least one case, this practice’s own research history rather than generic condition summaries.

And we close with the increasingly tangled state of vaccine policy in this country, where recommendations, access, and even the disease counts themselves now depend on who you ask — along with some welcome news on the mRNA flu vaccine front, and a few changes we’ve made in-house to close some of the access gaps ourselves.

1. A New Blood Test for Alzheimer’s

The FDA cleared a new blood test for Alzheimer’s disease in late August: PrecivityAD2, made by C2N Diagnostics, a spinout of Washington University School of Medicine in St. Louis. It’s the third FDA-cleared Alzheimer’s blood test, but the first cleared down to age 40. Using mass spectrometry, it measures two plasma biomarkers — Aβ42/40 and p-tau217/np-tau217 — and reports a composite result: negative, likely positive, or positive for amyloid pathology. In a validation cohort of 1,142 symptomatic adults, it correctly identified 97.6% of positive biomarkers, with performance holding up even in patients with chronic kidney disease or diabetes (patients amongst whom amyloid detection has previously been variable / invalid). It’s intended for patients already showing cognitive symptoms, not as a general screening tool, and it replaces what previously required PET imaging or a lumbar puncture (spinal tap).

Our own workup for neurodegenerative risk hasn’t changed with this new test — it’s added a tool, not replaced the sequence. We still genotype for ApoE4 status before ordering any amyloid biomarker test, since

it remains the strongest predictor of ARIA (amyloid-related brain swelling or microhemorrhage) risk on anti amyloid antibody therapy — a risk that rises sharply in ApoE4 homozygotes (the E4/E4 genotype) and is high enough that the UK’s regulator declined to approve lecanemab in that group altogether, even though it remains available to them here in the US.

How PrecivityAD2 differs from the ATN Profile we already run through Labcorp

Labcorp’s ATN Profile measures three markers — amyloid, tau, and neurofilament light chain (NfL) for neurodegeneration — while PrecivityAD2 measures only two, amyloid and tau, combined into one mass spectrometry-based score. It’s an “AT” test, not a full “ATN” one, and says nothing about disease severity. We reserve PrecivityAD2 for two situations: patients under 55, where Labcorp’s FDA-cleared components aren’t yet labeled, and as a confirmatory check when an immunoassay-based amyloid result sits in a gray zone rather than resolving cleanly.

Biography Health’s polygenic risk scoring, and why it’s not the same question as ApoE4 Biography Health’s DNAbled Blueprint runs polygenic risk scores (PRS) through two platforms — DynamicDNA and Galatea Bio — which use different reference populations and often produce different percentile estimates for the same condition; that’s expected, not an error, and where they diverge, we weigh both against the patient’s actual clinical picture. For Alzheimer’s, PRS and ApoE genotypes answer different questions: ApoE4 is one gene with a large individual effect, while PRS aggregates many small-effect variants elsewhere in the genome. A patient can carry one ApoE4 allele while their Alzheimer’s PRS sits at an unremarkable percentile — both pieces are needed, not an average of the two. Our neurodegenerative workup layers all three: ApoE genotype, polygenic background risk, and a blood biomarker test measuring current disease activity. More on the collaboration is on our Genomic Analysis and Biography Health pages, on our newly revamped drbretsky.com website.

2. Coronary Calcium, Cleerly, and What a New Study Actually Tells Us

The 2026 ACC/AHA/Multisociety dyslipidemia guideline we covered in the March issue newly recommended selective use of coronary artery calcium (CAC) scoring in adults whose 10-year PREVENT ASCVD risk falls between 3% and 10% — the borderline-to-intermediate range where the decision to start a statin is genuinely uncertain. A study out this month in Annals of Internal Medicine, drawing on the Multi Ethnic Study of Atherosclerosis (MESA) cohort across six U.S. sites, put that recommendation to an actual test: does adding a CAC score to the PREVENT-ASCVD equation meaningfully change risk prediction? The answer, across the full study population, was disappointingly modest — CAC added little discriminative value overall. But when the analysis was restricted to patients already sitting in that borderline risk band, CAC scoring earned its place, meaningfully reclassifying risk in the group the guideline actually intended it for. In other words: don’t order a CAC scan on everyone, but for the patient sitting at the edge of a statin decision, it still does real work.

We’ve been scanning exactly this population for years, and the new data validates rather than changes that practice. But CAC scoring has a blind spot worth naming plainly: an Agatston score only counts calcified plaque. It says nothing about the soft, non-calcified plaque that is often the more dangerous kind — the plaque most likely to rupture and cause an acute event — and a CAC of zero in a younger patient, or one already on a statin (which can paradoxically increase visible calcification while stabilizing plaque), doesn’t mean the coronary arteries are clean.

Cleerly, the AI-driven coronary CT angiography platform we’ve incorporated into the practice, is where that blind spot actually gets addressed. Where CAC gives you one number, Cleerly analyzes the full coronary CT angiogram vessel by vessel, quantifying both calcified and non-calcified plaque burden and grading stenosis severity directly rather than inferring it from a calcium proxy. In practice, we reserve full CCTA-with Cleerly analysis for patients where the clinical picture doesn’t fully resolve with CAC alone — a borderline PREVENT score with a strong family history or an elevated Lp(a), for instance, or a CAC of zero in a patient whose risk profile makes that reassuring result feel incomplete. The MESA data is a useful reminder that even guideline-endorsed tools have a fairly narrow lane where they change management; the point of layering CAC and Cleerly is knowing which patients actually sit in that band.

3. COVID in the Practice: A Late-Summer Pattern We’ve Now Tracked for Six Years

We’ve kept a case log for this practice since COVID’s earliest weeks in Los Angeles in March 2020, and that six years of data lets us say something with more confidence than a single season’s numbers would allow: the late-summer uptick we’re seeing right now is not new. It shows up, in some form, in every one of the six years we’ve tracked — smaller each year as population immunity has built, but never fully absent. 2026 stands out mainly for its silence beforehand: March through July produced not a single documented case in the practice, the quietest and longest gap in our entire six-year log. The cluster that followed in late August and continued into September, by contrast, looks almost exactly like the same months in 2024 and 2025 — smaller in absolute terms each time, but recognizably the same shape.

That last point is worth sitting with. Two very different eras of undercounting sit behind these numbers. In 2020, cases were missed because testing capacity didn’t exist yet at the scale needed — a number of infections in our earliest log were confirmed only in retrospect, sometimes years later, via antibody testing done for unrelated reasons. In 2026, testing capacity is not the constraint; a home antigen test is a ten-dollar drugstore item. The constraint now is reporting — a positive result on that test never reaches any public health system unless a patient picks up the phone. Both eras produce an undercount. They just get there by opposite mechanisms, which is exactly why we think the “case count” headline means less with each passing year, and why a practice’s own longitudinal log, however small, still tells you something official numbers increasingly can’t.

Post-Exposure Prophylaxis: Ensitrelvir (Xocova)

A question we’ve been fielding more often as travel picks back up: someone in the traveling party or household tests positive — is it worth taking something before you get sick too? Until this year, the honest answer was that there wasn’t much to offer beyond watching for symptoms. That changed on June 1, 2026, when the FDA approved ensitrelvir (Xocova, Shionogi) as the first oral option for post-exposure prophylaxis (PEP) of COVID-19, for anyone 12 and older following known exposure to an infected household member. Like Paxlovid, ensitrelvir works by blocking the SARS-CoV-2 main protease, but the indication is different: it’s taken by someone who is not yet infected, not someone already symptomatic. The approval rested on the SCORPIO-PEP trial, a placebo-controlled study of over 2,000 household contacts of a symptomatic case, and the regimen itself is straightforward — three tablets on day one, then one tablet daily for four more days.

The rationale for taking it seriously extends past the days spent feeling sick. Close to half of household contacts of someone with COVID-19 go on to develop infection themselves, and a meaningful share of those infections carry downstream consequences well past the acute illness — new or worsened neurologic, cardiovascular, respiratory, and renal issues in the following year, and for a subset of patients, Long COVID. Preventing the infection in the first place is, by definition, the only way to prevent all of that.

Cost is the practical catch. Xocova has no generic yet and runs around $1,375 cash for the five-day, seven tablet course. Shionogi’s manufacturer copay program can bring that to $0 for commercially insured patients, but the fine print excludes anyone on Medicare, Medicaid, VA benefits, TRICARE, or Indian Health Service coverage — which means many of the older, higher-risk patients this drug is arguably most useful for get no discount from the manufacturer at all, and are left to whatever their Part D plan decides to cover for a newly created indication. A separate assistance program for uninsured patients exists but is new enough that we’d encourage checking current terms before assuming it applies. For a patient asking whether it’s worth getting a prescription on hand before a trip, particularly one involving close quarters with family or a group, our answer is generally yes if cost isn’t the deciding factor — especially layered on top of a current COVID and flu vaccine, which is its own separate and still-worthwhile piece of protection heading into fall.

4. Website Updates: A Fuller “Conditions We Treat” on the Santa Monica Primary Care Website

We’ve built out a proper conditions library on the Santa Monica Primary Care website, organized into thirteen categories spanning cardiovascular, metabolic, digestive, respiratory, musculoskeletal, neurologic, and several other domains — from hypertension and atrial fibrillation to SIBO, sleep apnea, osteoporosis, and chronic kidney disease. The organizing principle wasn’t to produce another generic list of condition definitions; every page is built around the actual evidence base for that condition and, where it’s actually relevant, the specific research and clinical approach this practice brings to it.

The Alzheimer’s page is a good example of what we mean by that. It walks through the amyloid/tau pathophysiology, the 1993 discovery that established APOE-epsilon 4 as the first genetic risk factor identified for late-onset Alzheimer’s, and then something more specific: the finding, from a 1999 paper Dr. Bretsky co-authored, that epsilon 4’s risk is not sex-neutral — female carriers show meaningfully greater Alzheimer’s-related pathology and faster cognitive decline than male carriers with the identical genotype, a result extended in a 2003 follow-up in Neurology on longitudinal cognitive decline specifically. That’s not incidental detail. It’s the same genetic marker discussed in this issue’s lead section, decades earlier in this practice’s own research history, and it’s now a required part of the safety workup before starting lecanemab or donanemab given the ARIA risk in epsilon 4 carriers we described above. The page closes with how this actually plays out clinically here — APOE genotyping and PRS through Biography Health, ATN biomarker testing through Labcorp, NeuroQuant volumetric MRI when indicated, and coordination with USC’s Alzheimer’s Disease Research Center for diagnostic workup and treatment access.

The rest of the library follows the same approach: real evidence, real citations, and where applicable, the specific tools and referral relationships this practice actually uses. Browse the full set at santamonicaprimarycare.com/conditions.html.

5. Vaccine Policy Chaos: Recommendations, Access, and Disease Counts Now Have Three Different Answers

For most of our lifetimes, three separate questions — what vaccines are recommended, whether you can walk in and get one, and what’s actually happening with disease in this country — had a single answer, and that answer was the CDC. This fall, those three questions have three different answers, and none of them are the CDC.

The recommendations are now coming from four of the country’s largest medical societies — the American Academy of Pediatrics, American College of Obstetricians and Gynecologists, American Academy of Family Physicians, and Infectious Diseases Society of America — who issued joint 2026-27 respiratory season guidance directly, posted to a new American Medical Association site, SpreadTheFacts.org, with the underlying scientific reasoning published in JAMA. This work used to run through the CDC’s Advisory Committee on Immunization Practices (ACIP); that collaboration ended in June 2025 when the committee’s membership was wiped, and both committee actions and membership remain in legal limbo. In its absence, the societies commissioned their own evidence review through the University of Minnesota’s Vaccine Integrity Project, part of CIDRAP, working alongside the AMA. The resulting recommendations: flu vaccine for everyone six months and older, with an enhanced formulation (high-dose, adjuvanted, recombinant, or mRNA) preferred for adults 65 and up; an updated COVID vaccine for adults 19 and older, with a second dose six months later for those 65 and older, recommended during pregnancy in any trimester, and children 6 through 23 months eligible; and a single RSV dose for adults 75 and older and higher-risk adults 50–74, plus pregnant women in a window now extended through March 1 (previously ending January 31), reflecting RSV seasons trending later and longer.

Access is a separate, messier question, because most Americans get vaccinated at a pharmacy rather than their own physician’s office — often for the mundane reason that vaccines are typically billed through the pharmacy benefit rather than the medical benefit, which can leave a fully in-network primary care practice technically “out of network” for the vaccine itself, even though it’s in-network for everything else that practice does. Pharmacists then operate under state law, some of which explicitly ties their scope of practice to a CDC recommendation rather than a professional society’s. The FDA approved the updated 2026-27 COVID vaccines on August 27 — so yes, they are available now — but only for adults 65 and older and younger adults with a qualifying condition, a list broad enough that FDA officials have previously estimated it covers 30–60% of American adults. A younger, healthy patient without a risk factor on that list falls outside the current approval and would need an individualized, off-label discussion with us directly. Walgreens will let patients self-attest to meeting that criteria everywhere except Arizona and Oregon; CVS is using “shared clinical decision-making” without clarifying whether self-attestation counts; Walmart will require a prescription in some states. Whether private insurance covers a vaccine administered outside a CDC recommendation is genuinely murky — several major insurers say they’ll pay anyway, but nothing obligates them to.

One unambiguous bit of good news buried in all this: the Moderna mRNA flu vaccine saga we covered in the March issue has actually resolved, and more favorably than the FDA’s initial Refusal-to-File letter suggested it would. mFlusiva (mRNA-1010) received full FDA approval on August 5 for adults 50 to 64, with accelerated approval for adults 65 and older pending a postmarketing study, making it the first mRNA based flu vaccine to reach the US market. The theoretical case for an mRNA platform over an adenoviral vector one is that mRNA sidesteps anti-vector immunity — repeated dosing with an adenoviral-vector vaccine can blunt the immune response to the vector itself over successive doses, a problem mRNA vaccines don’t have — and its antigen sequence can be swapped each season faster, and in principle with tighter strain-matching, than egg-based or recombinant protein platforms allow. mFlusiva’s 26.6% relative efficacy edge over a standard-dose comparator is an early real-world data point in that direction, though it’s not yet clear how much of that gain is platform-specific versus simply a higher effective antigen dose. Moderna expects it available at select retailers before this flu season is over — a real option this fall, not a hypothetical one.

Given all of this, our own practice follows EU and Australian vaccination guidelines rather than waiting on domestic clarity that hasn’t come — if nothing else, “EU and Australia” is a considerably easier phrase to type than the AAP/ACOG/AAFP/IDSA alphabet soup above, though the underlying guidance is the actual reason, not the acronym count. One important caveat: Australia’s seasonality is the reverse of ours, so their “October” guidance is written for their spring, not ours, and has to be translated into our calendar rather than copied onto it directly. We continue recommending twice-annual COVID vaccination for patients over 75, annual for those over 65 or with chronic cardiopulmonary disease or immune compromise, and individualized decisions for younger, healthy patients based on exposure risk and infection history — largely consistent, as it happens, with where the AAP/ACOG/AAFP/IDSA guidance above has now landed.

To close some of the access gaps described above ourselves, we now stock both the high-dose flu vaccine for patients 65 and older and the standard-dose formulation in-house, along with MMR — specifically so patients aren’t left dependent on a pharmacy’s willingness to administer either. That MMR stock has already mattered: we’ve had more than one patient referred for MMR after a documented sub-protective measles antibody level, only to be turned away at the pharmacy because the pharmacist judged the vaccine “not indicated.” Whether a vaccine is indicated for a specific patient, based on a specific lab result, is a clinical decision made by the physician who ordered the antibody test and reviewed it — not a determination a pharmacist is positioned to make, and not one that should leave an under-immunized patient without protection. That stakes feel especially real right now given RFK Jr.’s continued refusal to formally acknowledge measles — the most infectious of all infectious diseases — as a public health emergency, despite case counts climbing and two pediatric deaths already reported in Pennsylvania. (If you were hoping to get through this newsletter without an RFK Jr. jab: sorry to “disappoint.”)

As many of you already know, COVID antibody testing (spike and nucleocapsid) has long been a standard part of our Annual Wellness panel, not something ordered only on request. Blended immunity — vaccine derived protection plus infection-derived protection — remains useful for individualizing vaccination timing, and having it as a standard component means we have that information on every patient, not only the ones who happen to ask for it.

The recommendations exist. The infrastructure to deliver them consistently and count the results honestly is what’s currently in question.

6. Follow Us On Social Media

Twitter (X): @santa_care

Facebook:https://www.facebook.com/santamonicaprimarycare

Instagram:@santa_monica_primary_care

LinkedIn: https://www.linkedin.com/in/philip-bretsky-mdphd/

BlueSky: @bretskymd.bsky.social

Substack: @bretskymd

Follow our blog posts on: www.drbretsky.com/blog

Archived issues of our newsletter can be found at: https://drbretsky.com/newsletters/

If you know somebody who would like to sign up to receive our newsletter, they can do so at: https:// lp.constantcontactpages.com/su/JtF9zzR

On a more constructive note, we also cover a new clinically available polygenic risk score covering eight cardiovascular conditions from a single blood or saliva sample — a natural companion to the BARCODE1 prostate cancer score we covered last edition, and to our ongoing Biography Health collaboration.

From there we turn to a retraction that should have happened two years ago: BMJ Public Health has formally withdrawn the Mostert et al. paper that anti-vaccine advocates — including a witness at a recent U.S. Senate hearing — have leaned on to argue that COVID-19 vaccines drove excess mortality. We go through what the paper actually claimed, and why it was wrong.

On the respiratory front, we look at two recent papers on influenza management: the ASTRA trial, the first head-to-head ranking of baloxavir, favipiravir, and oseltamivir in non-severe flu, and a Japanese VENUS-study target trial emulation on the timing of flu vaccination in older adults.

We close with two items from the summer’s general medical press: the ongoing multistate Cyclospora outbreak, first flagged here in July and since ballooned past 24,000 cases, and new CDC data showing the U.S. death rate fell to its lowest level ever recorded in 2025 — set against adult flu vaccination coverage, which is falling in the opposite direction.

1. The MMR Executive Order: A Solution in Search of a Problem

On August 10th, President Trump signed an executive order titled the “Gold Standard Childhood Vaccine Recommendations,” which does two things of note. First, it narrows the list of universally recommended childhood vaccines from 17 to 11, shifting shots like RSV and hepatitis A/B into a “high-risk” or shared-decision-making category. Second, and more consequentially for this newsletter, it directs HHS to draft, within 90 days, recommendations for administering the combined measles-mumps-rubella (MMR) vaccine as three separate single-antigen shots, “once such products are domestically available.”

Those products are not domestically available. There is no FDA-licensed monovalent measles vaccine, monovalent mumps vaccine, or monovalent rubella vaccine currently on the U.S. market. The combined MMR vaccine — first licensed in this form in 1971, built on a measles component that has been in continuous use since 1963 — is what exists. Splitting it into three would require manufacturers to develop, trial, and license entirely new single-antigen products, a process that takes years, not the 90 days the order allots for a recommendation. In the meantime, the practical effect of the order is rhetorical: it tells families that the current, extensively studied combination vaccine is somehow suspect, without offering an alternative that exists anywhere outside a press release.

The evidentiary basis for splitting MMR is, to be blunt, nonexistent. The autism hypothesis traces to Andrew Wakefield’s 1998 Lancet paper, retracted in 2010 after evidence of data manipulation and loss of Wakefield’s medical license. It has since been contradicted by a large and consistent body of subsequent research:

• Taylor et al., The Lancet, 1999 — UK population data found no step-change in autism diagnoses coinciding with MMR’s introduction or any clustering of diagnoses around the time of vaccination.

• Madsen et al., NEJM, 2002 — a Danish cohort of 537,303 children found no association between MMR vaccination and autism, and no association with age at vaccination, time since vaccination, or date of vaccination.

• DeStefano et al., Pediatrics, 2004 — a U.S. case-control study similarly found no relationship between MMR timing and autism diagnosis.

• Institute of Medicine, 2004 and 2011 reviews — two separate National Academy of Medicine committees reviewed the full body of evidence and concluded it favored rejecting a causal relationship between MMR and autism.

• Hviid et al., Annals of Internal Medicine, 2019 — a second, larger Danish cohort of more than 650,000 children found no increased autism risk associated with MMR vaccination, including in siblings of autistic children, the subgroup where a hidden effect would most likely surface.

• Jain et al., JAMA, 2015 — separately examined children with older autistic siblings and reached the same conclusion: no increased risk associated with MMR vaccination.

• Taylor, Swerdfeger & Eslick, Vaccine, 2014 — a meta-analysis spanning more than 1.2 million children across multiple study designs and countries found no association between vaccination and autism.

No single one of these settles the question by itself. Taken together, across different populations and methods over more than two decades, they’re about as close to settled as clinical epidemiology gets.

Why does this matter to our adult patients? Two reasons. First, measles is the most contagious pathogen in clinical infectious disease, with an R₀ in the range of 12 to 18, meaning a single case in an unvaccinated population can seed over a dozen secondary infections — see the figure below for how it compares to other pathogens we discuss regularly in this newsletter. U.S. measles cases are already at a 35-year high. Any policy that adds friction to the vaccination schedule — more visits, more copays, more chances for a family to fall through the cracks between doses — widens the susceptible population, and susceptible populations are how outbreaks reach adults who assume their own childhood immunity or a single adult booster is sufficient. Second, this order is not happening in isolation. It follows a CDC advisory panel push for the same monovalent-MMR concept last fall, which a federal judge blocked in March on procedural grounds. This is the same policy, recycled, now delivered by executive order rather than through ACIP’s normal evidentiary review — a workaround, not a scientific advance.

This isn’t the first time this newsletter has tracked institutional dysfunction at the intersection of vaccine policy and this administration. In an interesting bit of timing, the FDA approved Moderna’s mRNA seasonal flu vaccine, mFLUSIVA, on August 5th — closing the loop on the Dr. Prasad saga we covered in our March newsletter. Two administrations’ worth of vaccine whiplash, one month apart.

2. Precision Medicine: A Multi-Disease Cardiovascular Risk Score

Last edition covered BARCODE1, the saliva-based polygenic risk score (PRS) for prostate cancer. A parallel effort in cardiovascular medicine reached the same stage this spring: a clinically orderable, multi-condition cardiovascular PRS, developed by Anika Misra and colleagues at Massachusetts General Hospital and published online in JACC in March, with validation data reported at TCTMD in May.

The score was built from genetic and clinical data on 245,394 participants in the NIH’s All of Us Research Program, then validated in 53,306 participants from the Mass General Brigham Biobank. Rather than targeting a single disease, it reports risk across eight cardiovascular conditions from one blood or saliva sample: coronary artery disease (OR 3.7 comparing high to average genetic risk), type 2 diabetes (OR 3.1), atrial fibrillation (OR 3.0), hypercholesterolemia (OR 4.1), hypertension (OR 2.1), venous thromboembolism (OR 1.9), thoracic aortic aneurysm (OR 1.7), and elevated lipoprotein(a) (OR 41.0). That last figure is not a typo — Lp(a) is highly heritable, and a PRS captures that genetic architecture more directly than almost any other cardiovascular trait we measure, which is part of why we’ve incorporated Lp(a) testing into our own practice for years.

The more clinically useful finding sits in how the score performs among patients who are hardest to counsel: those in the gray zone of conventional risk calculators. Among people originally classified as borderline or intermediate risk by the pooled cohort equations, adding the PRS meaningfully reclassified risk in both directions — some patients moved up, some moved down — and the same held true when the score was layered onto the newer PREVENT equations we discussed in March, and onto CHARGE-AF for atrial fibrillation risk specifically. In prospective follow-up (median 7.6 years) among participants healthy at baseline, a high score was associated with roughly double the risk of incident coronary artery disease and atrial fibrillation, and comparable increases for type 2 diabetes and venous thromboembolism.

This is precisely the kind of tool our Biography Health partnership is built around: not a single-disease test, but a genomic layer added to conventional risk assessment that helps sharpen a decision that’s otherwise a coin flip — the patient with borderline LDL and a strong family history who isn’t sure whether starting a statin at 42 makes sense, or the patient wondering whether an elevated Lp(a) runs in the family. A PRS doesn’t override clinical judgment or replace the PREVENT calculation; it adds a layer that’s fixed at birth and doesn’t shift with this year’s cholesterol panel, which is exactly what makes it useful for the small number of decisions where conventional risk factors leave real ambiguity.

3. A COVID-Era Retraction, Two Years Late

On August 11th, BMJ Public Health formally retracted “Excess mortality across countries in the Western World since the COVID-19 pandemic: ‘Our World in Data’ estimates of January 2020 to December 2022,” a paper published in May 2024 by Saskia Mostert and colleagues at the Princess Máxima Center for Pediatric Oncology in the Netherlands.

What the paper claimed. Mostert and colleagues used two publicly available mortality datasets — Our World in Data and the World Mortality Dataset — to calculate excess mortality (deaths above a pre-pandemic expected baseline) across roughly 47 Western countries from January 2020 through December 2022. Their headline finding was that excess mortality remained elevated through 2022 — two years into the pandemic, after COVID-19 vaccines had been widely deployed — despite the implementation of both containment measures and vaccination campaigns. The paper then moved from that descriptive finding into a lengthy discussion section arguing that COVID-19 vaccination itself may have contributed meaningfully to this continued excess mortality — citing reports of vaccine-associated myocarditis, pericarditis, and other adverse events as a plausible explanatory mechanism, without actually testing that hypothesis against the data the paper had assembled. The paper did not run a country-level analysis correlating vaccination rates with excess mortality; it simply observed that both vaccines and continued excess deaths were present in the same time period, and devoted substantial space to the argument that vaccines were a contributing cause.

Why it was wrong. That is a textbook conflation of correlation and causation, and critics identified the problem within days of publication. The paper’s own numbers, properly interrogated, point the opposite direction: Stuart McDonald, an actuary who audits pandemic mortality claims professionally, found that countries with higher vaccination rates tended to have lower excess deaths in the paper’s own dataset — the association ran opposite to what the discussion section insinuated. Separately, Ariel Karlinsky, the Hebrew University economist who co-created the World Mortality Dataset the paper relied on, contacted the journal starting in June 2024, arguing the analysis misused his data and had duplicated substantial blocks of text and mathematical notation from his own previously published work without proper attribution — a plagiarism concern layered on top of the causal-inference problem. An expression of concern went up within a month of publication, and BMJ Public Health opened a formal institutional investigation through the Princess Máxima Center shortly after. The final retraction notice cites “misinformation in the discussion regarding the possible causes of excess mortality” alongside “concerns about quality, originality and messaging” — language that maps directly onto both problems: a discussion section that argued for a causal link the data didn’t support, built partly on text that wasn’t the authors’ own.

It took two more years, and a starring role in a U.S. Senate subcommittee hearing where a witness cited the paper to argue for a vaccine-cancer link, for the retraction to actually happen. The original study never explicitly stated that vaccines caused the excess deaths it measured; the ambiguity was the point of leverage. A paper doesn’t need to state a false conclusion outright to do damage — it just needs to leave the door open, and the door was walked through, repeatedly, by people with an existing narrative to confirm. Retraction, two years on, corrects the record. It does not undo the two years of amplification that preceded it.

4. Influenza: Ranking the Antivirals, Timing the Shot

Two papers, both essentially fresh off the press, sharpen how we should think about seasonal flu management in two different populations — young low-risk adults choosing an antiviral, and older adults deciding when to get a flu shot.

Antiviral head-to-head: the ASTRA trial. Published in The Lancet Infectious Diseases, ASTRA is the first trial to directly rank baloxavir, favipiravir, and oseltamivir against one another rather than against placebo alone. Led by Podjanee Jittamala at Mahidol University (Bangkok, Thailand), the trial enrolled 729 adults aged 18 to 60 (modified intention-to-treat population; median age 31; 63% female; 76% influenza A) across sites in Thailand, Laos, Nepal, and Brazil between February 2023 and December 2025, all with PCR-confirmed influenza, symptomatic for four days or less, and at low risk of progressing to severe disease. Patients were randomized to single-dose baloxavir (weight-based, 40 or 80 mg), a five-day favipiravir regimen, five days of standard-dose oseltamivir, or no study drug.

All three antivirals significantly outperformed no treatment on the primary endpoint, rate of oropharyngeal viral RNA clearance, but not equally. Baloxavir produced the largest acceleration (86% faster than no drug), followed by favipiravir (66%) and oseltamivir (49%), each with a posterior probability of 1.0. Head-to-head against baloxavir, oseltamivir was clearly inferior (20% slower clearance), while favipiravir’s non-inferiority claim narrowly missed (10% slower, confidence interval crossing zero).

All three drugs shortened fever duration modestly, by 0.5 to 0.9 days versus no treatment, but none significantly reduced time to complete symptom resolution — the viral-clearance/symptom-resolution dissociation this section has flagged before, now with real numbers behind it. Patients taking these drugs clear detectable virus faster and run a fever slightly less long, but “feeling sick” overall doesn’t reliably resolve any sooner. That is a useful thing to tell a patient asking whether Tamiflu will get them back to work by Thursday: probably not, though it may shorten how long they’re shedding virus onto coworkers. The authors frame baloxavir as the most potent oral option in this class, with favipiravir a reasonable second-line choice where regionally available and both worth further study for pandemic stockpiles.

Vaccine timing in older adults: the VENUS target trial emulation. The second paper, published ahead of print in Clinical Microbiology and Infection (Mimura et al.), takes a clever approach to the timing question — a target trial emulation using cloning, censoring, and weighting to simulate seven vaccination strategies across four Japanese influenza seasons, 2016/17 through 2019/20. Each season’s cohort ran 131,000 to 138,000 adults aged 65 and older (median age 76–78).

The result is more interesting than a simple “vaccinate early” conclusion. Delayed vaccination was actually associated with the lowest influenza incidence in two of the four seasons studied (2017/18 and 2018/19), and hospitalization followed the same season-dependent pattern. All-cause mortality told a more consistent story: early vaccination reduced all-cause mortality relative to no vaccination in every season studied, with risk ratios ranging roughly from 0.80 to 1.00 depending on season and strategy, and the absolute risk difference was consistently larger for mortality than for the flu-specific outcomes.

This dissociation could reflect biology — waning antibody titers leaving late-vaccinated patients under-protected precisely as a season’s incidence curve peaks, a pattern that would show up more clearly in a broad, noisy outcome like all-cause mortality than in influenza-specific case counts. It could also reflect residual confounding a target trial emulation can’t fully remove: healthier, more health-engaged older adults are more likely to get vaccinated in October than in November or December, and that same engagement plausibly lowers all-cause mortality independent of the vaccine itself. The authors don’t overclaim either reading. What supports the mortality signal over the confounding explanation is consistency — the same direction of effect across four seasons with different dominant strains and different epidemic curves is a reasonably high bar for confounding alone to clear. It supports what we already recommend: get the flu shot in October, not because it’s guaranteed to be the best month for avoiding influenza specifically, but because the mortality data doesn’t move.

5. Cyclospora: An Update

We first covered this outbreak in July, when case counts were still in the low thousands. As of the most recent CDC data, more than 24,000 cyclosporiasis infections have been reported nationally, including at least 13,895 laboratory-confirmed cases and 740 hospitalizations. Two deaths have now been attributed to the outbreak, both in patients with significant underlying illness. The rate of new case reporting has slowed but not stopped. The source remains iceberg lettuce from Taylor Farms de Mexico, distributed widely through fast-casual restaurant chains including Taco Bell, and the recall, now weeks old, has not fully arrested new infections given the parasite’s incubation period and the lag in case reporting.

CDC’s foodborne surveillance network made cyclospora testing optional in July 2025, a change some infectious disease experts believe is complicating the current response, though federal health officials have publicly disputed that the surveillance change is affecting the investigation. Cyclosporiasis remains a largely self-limited illness in immunocompetent patients — watery diarrhea, bloating, fatigue, low-grade fever in a minority — but the standard treatment (trimethoprim-sulfamethoxazole) is worth knowing for patients with a sulfa allergy, for whom options are more limited. If you or a family member have had prolonged watery diarrhea after eating shredded lettuce or a fast-casual salad since late June, call the office rather than waiting it out.

6. The Good News: U.S. Death Rate Hit a Record Low in 2025

Amid a newsletter otherwise devoted to policy dysfunction and outbreaks, here’s something that actually went right. The U.S. death rate fell to 689.2 per 100,000 people in 2025, a 4.6% drop from 2024 and the lowest rate on record, according to provisional CDC data released July 2nd. The decline was seen across all age groups and both sexes, and demographers expect the corresponding life expectancy figure, not yet finalized, to set a new record when it’s calculated.

The sharpest driver was a continued decline in drug overdose deaths, particularly among adults aged 25 to 34, whose death rate fell from 124.5 to 111.1 per 100,000 year over year — a decade of opioid-driven losses in that age group finally bending the other way. Heart disease and cancer remain, as they have for decades, the two leading causes of death nationally. One line item stands out against that backdrop, and it’s worth reading alongside the second panel of the figure above: influenza moved up the leading-causes-of-death list this year, from 11th to 8th, at the same time adult flu vaccination coverage has fallen to its lowest level in roughly seven seasons — 41.9% in 2024–25, down from 52.2% just four seasons earlier. Even in a year when nearly every other mortality trend improved, the one respiratory illness with a decades-old, well-studied vaccine moved the national numbers in the wrong direction.

7. Follow Us On Social Media

Twitter (X): @santa_care

Facebook: https://www.facebook.com/santamonicaprimarycare

Instagram: @santa_monica_primary_care

LinkedIn: https://www.linkedin.com/in/philip-bretsky-mdphd/

BlueSky: @bretskymd.bsky.social

Substack: @bretskymd

Archived issues of our newsletter can be found at: https://drbretsky.com/newsletters/

If you know somebody who would like to sign up to receive our newsletter, they can do so at: https://lp.constantcontactpages.com/su/JtF9zzR

Elsewhere the news is more encouraging, if more complicated. An oral PCSK9 inhibitor reported Phase 3 results strong enough to compete with the injectable antibody drugs, arriving in the same month that a new JAMA analysis quantified just how many additional Americans the 2026 dyslipidemia guideline has made eligible for statin therapy—21.5 million newly eligible, the majority of them at meaningfully lower cardiovascular risk than the population statins were originally studied in. A UK trial published in the New England Journal of Medicine demonstrated that a saliva-based polygenic risk score can identify clinically significant prostate cancers that PSA testing misses entirely, work we’re now folding into our genomic screening partnership with Biography Health.

And because no topic generates more questions in this office than GLP-1 receptor agonists, this month’s newsletter closes with a longer, more clinical treatment of what the evidence actually supports for semaglutide, tirzepatide, and the new oral formulations—where the data are strong, where we are still extrapolating from a narrower trial population than the marketing implies, and what happens, physiologically and not just anecdotally, when patients stop.

Subscribe now

1. The FDA’s Peptide Panel: Strong Animal Data, an Unclosed Human Evidence Gap, and a Compounding Pathway That May Guarantee It Stays That Way

On July 23 and 24, the FDA’s Pharmacy Compounding Advisory Committee (PCAC) met to evaluate seven peptides—BPC-157, KPV, TB-500, MOTS-c, emideltide (DSIP), epitalon, and Semax—for inclusion on the Section 503A Bulk Drug Substances List, the mechanism by which compounding pharmacies are permitted to dispense a substance without full FDA drug approval. The committee’s own briefing documents, prepared by FDA career scientists applying the four-factor statutory framework under 21 CFR 216.23(c)—physical and chemical characterization, historical use in compounding, evidence of effectiveness, and evidence of safety—recommended against all seven. On the first day of the meeting, the committee voted 8–6, with one abstention, to recommend BPC-157 for inclusion despite that recommendation; KPV, TB-500, and MOTS-c followed the same pattern. The votes are non-binding and still require formal FDA rulemaking before taking effect, but they represent a meaningful signal of where this is headed (CNN).

It is worth being precise about what “the evidence isn’t there” actually means for these compounds, because the honest picture is more nuanced than either side of the current debate tends to present it. For BPC-157 specifically, the preclinical literature is genuinely substantial—a body of rodent studies spanning gut mucosal healing, tendon and ligament repair, and anti-inflammatory signaling that has accumulated over roughly two decades (Józwiak M, et al. Multifunctionality and Possible Medical Application of the BPC 157 Peptide—Literature and Patent Review. Pharmaceuticals. 2025). What is largely absent is controlled human data: dose-response relationships, pharmacokinetics in people, and outcome data from anything resembling a randomized trial—the “Gold Standard Science” that RFK Jr. repeatedly references. That is a different problem than “no evidence,” and it deserves a different response than either uncritical adoption or blanket dismissal. Strong preclinical signal combined with an unclosed human evidence gap is, properly understood, a reasonable basis for continued research interest. It is not evidence for compounding-pharmacy distribution to the general public.

The panel’s composition has drawn substantial press attention, and for good reason—several new appointees run peptide clinics, sell the compounds directly, or maintain substantial commercial followings promoting them, a marked departure from prior PCAC panels, which drew primarily from academic pharmacology and toxicology departments. But I want to redirect attention toward a structural problem that matters more than any individual conflict of interest. FDA drug approval comes bundled with an evidence-generating apparatus: defined dosing, controlled trials, and mandatory post-marketing surveillance. Compounding-pharmacy access carries none of that scaffolding. Once a peptide is legally available through a local compounding pharmacy, the prescriber’s incentive is simply to prescribe it; there is no registry requirement, no mandated follow-up laboratory work, and no standardized outcome reporting attached to the pathway. Multiply that across the thousands of patients who will likely seek these compounds once they’re more widely available, and the result is years of real-world use that generates essentially no usable evidence—not because the compounds are necessarily dangerous, but because the distribution model we’ve chosen forecloses the possibility of ever finding out.

That is where I think the responsibility now falls on physicians rather than regulators. If we are going to prescribe these compounds—and many of us will, given patient demand—the burden is on us to build the rigor that the compounding pathway does not require: baseline laboratory work, clearly defined endpoints, structured follow-up intervals, and honest documentation of outcomes, positive and negative. Absent that discipline, we are not practicing evidence-based medicine on a new frontier; we are simply adding anecdotes to a pile nobody can interpret. For patients asking about these compounds, my advice is to treat “peptide” as a category rather than a guarantee—your body already produces many legitimate, well-studied peptide hormones, insulin foremost among them—and to ask any prescriber what specifically they intend to measure before and after treatment, not simply whether they’re willing to write the prescription.

For a longer, clinician-level discussion of these specific data gaps, see Dr. Michael Albert’s overview, available on VuMedi: Peptides: Concerns and Challenges.

2. “Diarrhea Lettuce”: A Cyclospora Outbreak, and the Surveillance System We Chose to Stop Funding

Regular readers of this newsletter know I try to avoid leading with internet ephemera, but the meme circulating this month captures something real: a person commits to eating healthily, the lettuce responds with a bout of explosive diarrhea, and the Oreos in the pantry point out that they, at least, would never do that to you. It is funny. The outbreak behind it—caused by a parasite, not the bacteria that usually headline these stories—is not.

The FDA and CDC are jointly investigating a multistate outbreak of cyclosporiasis linked to shredded iceberg lettuce, sourced from a single Mexican supplier, and served at Taco Bell locations across Indiana, Kentucky, Michigan, Ohio, and West Virginia. Michigan has borne the brunt of it, with more than 3,700 confirmed cases as of mid-July. Cyclospora cayetanensis is a protozoan parasite that causes prolonged, watery diarrhea, often accompanied by fatigue, cramping, and low-grade fever, and—unlike most of the foodborne pathogens we discuss in this newsletter—it does not reliably resolve on its own. Untreated infection can persist for weeks to months, with symptoms waxing and waning, and the standard treatment is a specific antibiotic, trimethoprim-sulfamethoxazole, rather than supportive care alone. Washing produce reduces but does not reliably eliminate risk, since the parasite’s oocysts adhere tenaciously to leaf surfaces and are resistant to simple rinsing.

Some seasonality is expected here and should not itself be read as alarming: the CDC designates May 1 through August 31 as cyclosporiasis season in the US, because the parasite’s oocysts require one to two weeks in warm environmental conditions to mature into an infectious form after being shed in feces, and because the growing season for the imported fresh produce most often implicated—basil, cilantro, raspberries, and leafy greens—coincides with those same months. A modest summer uptick is the baseline. What is not baseline is an outbreak of this magnitude: more than 11,000 confirmed and probable cases reported nationally as of late July, against roughly 2,700 for all of 2025.

The epidemiologically important part of this story is not the pathogen but the infrastructure gap that let it grow to this scale before detection. FoodNet, the CDC’s active foodborne-illness surveillance network operating across ten sentinel states since 1996, has historically worked by contacting clinical laboratories directly to identify cases in near-real time, rather than waiting for the slower, passive chain of case reports to accumulate through the usual public health reporting structure. That distinction—active surveillance versus passive reporting—is the entire reason FoodNet exists, and it is the reason outbreaks tracked by the network have historically been identified and contained faster than those relying on standard reporting alone. As of mid-2025, budget reductions cut FoodNet’s mandatory pathogen list from eight organisms to two: Salmonella and Shiga toxin–producing E. coli. Cyclospora surveillance, along with six other pathogens, was downgraded to optional reporting. The practical consequence is exactly what we are watching unfold: an outbreak that likely grew substantially larger, and was identified substantially later, than it would have been under the surveillance regime that existed eighteen months ago.

This was self-inflicted, and it was foreseeable. You cannot manage a public health threat you have deliberately stopped measuring. Defunding active surveillance does not make outbreaks smaller or rarer; it makes them invisible for longer, and the bill for that invisibility is paid by patients, not by the budget line item that generated the savings. We watched a version of this same pattern play out earlier this year with the erosion of SARS-CoV-2 sequencing capacity ahead of the 2026–27 vaccine strain selection process. This month it is a parasite in lettuce rather than a respiratory virus, but the underlying mechanism of institutional failure is identical.

For patients: if you have experienced diarrhea lasting more than several days, particularly if you recall eating at one of the affected locations in the relevant states, please contact our office rather than waiting it out—this infection generally requires a specific antibiotic rather than time alone.

Source: FDA, Investigation of 5-State Outbreak of Cyclospora Illnesses: Iceberg Lettuce (July 2026), and CDC FoodNet Fast surveillance data. fda.gov/food/outbreaks-foodborne-illness

3. An Oral PCSK9 Inhibitor Arrives the Same Month the Statin-Eligible Population Grew by 21.5 Million

Two pieces of lipid-management news landed within days of each other in July, and I think they are best read together rather than separately.

The first is a JAMA analysis published July 20 quantifying the population-level effect of the 2026 AHA/ACC/multisociety dyslipidemia guideline, which we first covered in our March newsletter when it replaced the 2018 cholesterol guideline. Using nationally representative NHANES data and the newer PREVENT-ASCVD risk equations, the authors estimate that 87.5 million American adults—56.6% of the nonpregnant population aged 30 to 79—now meet criteria for primary prevention statin therapy, including 21.5 million individuals who were not eligible under the prior guideline.

The second piece of news is more straightforwardly good: enlicitide, an investigational once-daily oral PCSK9 inhibitor, reported Phase 3 results this year showing LDL-C reductions in the range of 56% to 65% on top of background lipid-lowering therapy, across the CORALreef Lipids, CORALreef heFH, and CORALreef AddOn trials, with a safety profile comparable to placebo. That efficacy is in the same range as the injectable PCSK9 monoclonal antibodies, evolocumab (Repatha) and alirocumab (Praluent), but delivered as a daily pill rather than an every-other-week injection—a meaningful access improvement for patients who have declined injectable therapy on convenience or needle-aversion grounds, and for those who have not reached goal LDL on statin plus ezetimibe alone. Enlicitide is not yet FDA approved.

Read together, these two developments cut in slightly different directions. An effective, well-tolerated oral non-statin option is genuinely useful for higher-risk patients who need more aggressive LDL lowering than statins alone provide. But its arrival coincides with a guideline expansion that is pulling millions of comparatively low-risk, often younger patients into pharmacotherapy discussions where the absolute benefit is smaller and the multi-decade commitment—financial, and in terms of side-effect exposure—is proportionally larger relative to that benefit. Easier access to an effective drug is not, by itself, a reason to prescribe it reflexively. If your LDL-C, Lp(a), or ApoB has newly placed you in an eligibility category under the 2026 guideline, I would encourage a visit to review your specific numbers against your actual risk profile rather than treating guideline eligibility as a standing prescription.

References: Anderson TS, Wilson LM, Sussman JB. Implications of the 2026 Dyslipidemia Guideline for Primary Prevention Statin Therapy. JAMA. Published online July 20, 2026. doi:10.1001/jama.2026.11246. Navar AM, et al. Phase 3 study results for oral PCSK9 inhibition with enlicitide. J Clin Lipidol. 2026.

4. Reframing Prostate Cancer Risk via Polygenic Risk Scoring Risk

Prostate cancer screening has long been constrained by the limitations of prostate-specific antigen (PSA) testing, which is elevated by benign prostatic hyperplasia, prostatitis, and normal aging in addition to malignancy, producing a well-documented high false-positive rate and a parallel problem of biopsies detecting indolent cancers that would never have become clinically significant. The BARCODE1 trial, published in the New England Journal of Medicine, offers a different approach. Investigators recruited men aged 55 to 69 from primary care centers in the United Kingdom and calculated a polygenic risk score (PRS) from 130 germline genetic variants known to be associated with prostate cancer risk, using DNA extracted from a mailed saliva kit. Of 40,292 men invited to participate, 6,393 completed genotyping; the 745 men whose PRS fell at or above the 90th percentile were invited for MRI and transperineal biopsy irrespective of their PSA level. Prostate cancer was detected in 40.0% of that high-PRS group, and of the cancers requiring clinical management by NCCN criteria, 70.9% would have been missed entirely by the standard UK PSA-driven diagnostic pathway.

The reason a genetic risk score outperforms PSA in this context is straightforward: PSA is a biomarker that fluctuates with inflammation, infection, and benign glandular enlargement, none of which have any bearing on cancer risk. A polygenic risk score measures inherited susceptibility that is fixed at conception and can therefore be assessed decades before any symptom, biomarker abnormality, or imaging finding would prompt an evaluation under the conventional pathway.

At Santa Monica Primary Care, we have incorporated this kind of genomic risk stratification into our ongoing collaboration with Biography Health. Their whole genome sequencing panel—collected via cheek swab or saliva sample, priced at approximately $1,200—extends beyond a single-cancer PRS to assess inherited risk across cardiovascular, metabolic, autoimmune, and multiple cancer domains simultaneously, prostate cancer included. This is not a replacement for PSA testing, nor is it diagnostic on its own; its value is in risk stratification well in advance of any conventional finding, allowing us to know, for a given patient, whether a borderline PSA result next year warrants an immediate MRI referral or a routine repeat test in six months. That is Precision Medicine functioning as intended: not indiscriminately more testing, but calibrated testing intensity matched to actual inherited risk.

Reference: Eeles RA, et al. Assessment of a Polygenic Risk Score in Screening for Prostate Cancer (BARCODE1). N Engl J Med. 2026.

5. GLP-1 Receptor Agonists: A Fuller Clinical Picture for the Question I’m Asked Most Often

No topic in this practice generates more patient questions right now than GLP-1 receptor agonists, and the public conversation around them—equal parts enthusiasm and social-media alarmism—rarely reflects the actual state of the evidence. I want to walk through it in more detail than usual.

The evidence we have is genuinely strong. In the SELECT trial, semaglutide reduced major adverse cardiovascular events by 20% in adults with obesity and established cardiovascular disease who did not have diabetes—a hard clinical outcome, not a surrogate marker, in a well-defined and appropriately high-risk population. For patients who resemble that trial population, the evidence for benefit is about as good as cardiovascular pharmacotherapy evidence gets.

That same strength of evidence creates an observation bias worth naming explicitly. The outcomes we are currently seeing in clinical practice are disproportionately drawn from the sickest, highest-risk patients, because that is who has been enrolled in the pivotal trials and who insurers have prioritized for coverage. That population has the most to gain in absolute terms, which is exactly what the data show. It does not follow automatically that a healthy 35-year-old with a BMI of 31 and no additional cardiometabolic risk factors will see benefits of the same magnitude—the general principle, familiar from decades of statin and antihypertensive research, is that absolute risk reduction scales with baseline risk. This does not mean lower-risk patients derive no benefit from these medications. It means the benefit is smaller, and the cost, side-effect burden, and multi-year commitment need to be weighed against a correspondingly smaller absolute gain. We will not have clean outcome data in this lower-risk population for years, because the trials generating our current evidence base were not designed to answer that question.

Oral formulations represent a real trade-off, not simply a downgrade. The FDA approved orforglipron (Foundayo) this month, the first oral GLP-1 receptor agonist that can be taken without food or water restrictions—a genuine convenience advantage over oral semaglutide, which requires dosing on an empty stomach with a narrow water-only window. The efficacy trade-off is real, however: orforglipron produced approximately 11% mean weight loss at 72 weeks in its pivotal trial (ATTAIN-1), oral semaglutide 25 mg approximately 13.6% at 64 weeks (OASIS-4), against 15% to 20% for the injectable formulations. Gastrointestinal side effects are broadly similar across formulations. For patients unwilling to tolerate injections, that trade may be entirely reasonable. For patients pursuing maximal weight loss, the oral formulations are not simply “the same drug in pill form”—they occupy a meaningfully different efficacy tier.

The musculoskeletal costs of rapid pharmacologic weight loss are real and underdiscussed. Roughly 20% to 25% of the weight lost on these medications is lean mass rather than fat—a proportion comparable to what is seen with aggressive caloric restriction generally, though it occurs more rapidly and often in patients who are not concurrently engaged in resistance training. The SURMOUNT-1 DXA substudy found a 10.9% reduction in total lean mass with tirzepatide over 72 weeks. Bone density is affected as well: a 52-week randomized trial of semaglutide specifically designed to assess skeletal outcomes found a 2.6% reduction in total hip bone mineral density and a 2.1% reduction at the lumbar spine, accompanied by increased bone resorption markers without a compensatory increase in bone formation. None of this is a reason to avoid these medications in appropriate candidates. It is a reason to prescribe them alongside a structured resistance-training and adequate-protein plan from the outset, and to obtain a baseline DEXA scan in patients who are older, postmenopausal, or otherwise at elevated fracture risk before initiating therapy.

The question patients almost never ask before starting is what happens if they stop. In the STEP 1 extension trial, patients who discontinued semaglutide after 68 weeks of treatment regained two-thirds of their lost weight within the following year—11.6 of the 17.3 percentage points of body weight they had lost, versus 1.9 percentage points of gain in the placebo group over the same interval. SURMOUNT-4 demonstrated a comparable pattern with tirzepatide discontinuation. This regain is not a failure of willpower; appetite regulation and metabolic set point drift back toward baseline once the drug’s pharmacologic effect is withdrawn, in essentially the same way blood pressure rises again after an antihypertensive is stopped. The practical implication is that the conversation about long-term intent belongs at the start of treatment, not after it: is this being initiated as indefinite chronic therapy, analogous to a statin, or as a bridge toward some other endpoint? If the latter, what specifically is the plan for maintaining results once the medication is withdrawn?

A few claims circulating on social media are worth addressing directly. Compounded semaglutide is not necessarily equivalent to the FDA-approved product; compounded formulations produced during the recent supply shortage have not undergone the same purity and dosing verification, and quality has been documented as inconsistent across compounding sources. The framing of these medications as “the easy way out” does not withstand scrutiny against the STEP trial regain data, which is itself among the strongest evidence that obesity behaves as a chronic, biologically regulated condition rather than a simple deficit of willpower. And the phenomenon colloquially termed “Ozempic face”—facial volume loss from rapid fat reduction—is a real cosmetic effect but not a safety signal; the adverse effects that genuinely warrant clinical attention are gastrointestinal intolerance, gallstone formation, and the rare but documented risk of pancreatitis, which is a separate conversation from cosmetic appearance.

Taken together, these are excellent medications for the right patient, deployed with an actual plan—for muscle preservation, for bone protection, and for what happens at the end of treatment, whenever that end arrives.

References: Lincoff AM, et al. Semaglutide and Cardiovascular Outcomes in Obesity without Diabetes (SELECT). N Engl J Med. 2023. Wilding JPH, et al. Weight regain and cardiometabolic effects after withdrawal of semaglutide: the STEP 1 trial extension. Diabetes Obes Metab. 2022. Look M, et al. SURMOUNT-1 DXA substudy. Diabetes Obes Metab. 2025.

6. Follow Us on Social Media

Twitter (X): @santa_care

Facebook: https://www.facebook.com/santamonicaprimarycare

Instagram: @santa_monica_primary_care

LinkedIn: https://www.linkedin.com/in/philip-bretsky-mdphd/

BlueSky: @bretskymd.bsky.social

Substack: @bretskymd

Follow our blog posts on: www.drbretsky.com/blog

Archived issues of our newsletter can be found at: https://drbretsky.com/newsletters/

If you know somebody who would like to sign up to receive our newsletter, they can do so at:

https://lp.constantcontactpages.com/su/JtF9zzR

Subscribe now

On other fronts, the FDA has announced a seismic shift in how clinical trials are conducted—real-time data monitoring that collapses the months-long hiatuses between trial phases, potentially accelerating drug development by 1-2 years. For patients waiting for lifesaving therapies, this modernization matters.

On the respiratory front, as we transition into summer, spring’s relative lull in infection is giving way to a parainfluenza uptick—a timely reminder that seasonal epidemiology has rhythms that merit our attention, even in the post-pandemic era.

For practitioners engaged in Precision Medicine, June also brings advances in early menopause genetics. We highlight a case report that illustrates new knowledge convergence in polygenic risk scores, whole genome sequencing, and early menopause—tools that can better frame, focus, and identify women in perimenopause and inform discussions about hormone therapy, a conversation that has been unfairly shadowed by Women’s Health Initiative data from decades past.

Finally, both COVID and influenza vaccine composition for 2026-2027 are now set. Monovalent XFG-targeted COVID vaccines and updated flu strains reflect sound epidemiological science, yet the decisions reveal institutional dysfunction that nearly derailed the process.

1. Real-Time Clinical Trials: Closing the Gaps Between Phases

For decades, drug development has followed a sequential, phase-based architecture: Phase 1 (safety/dosage in healthy volunteers) → pause → Phase 2 (efficacy in diseased patients) → pause → Phase 3 (large pivotal trial). Between each phase, there are lengthy hiatuses for data analysis, regulatory review, and protocol revisions. The FDA reviews data after each phase concludes (Figure 1 below).

This made sense in the 1970s. Computers were cumbersome. Data collection meant paper forms. The idea of real-time analytics was science fiction.

Today, it is neither. Real-Time Clinical Trials (RTCT) allow the FDA to monitor safety signals and efficacy endpoints continuously as trials progress, eliminating (or drastically reducing) inter-phase delays. A Phase 2 trial that might have taken 3-4 years to complete (12 months enrollment + 24 months follow-up + 9 months hiatus before Phase 3) could instead transition to Phase 3 within weeks of meeting pre-specified efficacy criteria. The FDA has already launched two proof-of-concept RTCT trials and will be accepting applications for a pilot program launching this summer.

Why be cautiously optimistic but skeptical? RTCT requires robust data infrastructure—something smaller biotech and academic centers may lack. Real-time monitoring also introduces bias: analysts exposed to emerging treatment differences may be tempted to cherry-pick favorable findings. The FDA’s historical track record on novel methodologies is mixed (see our March 2026 newsletter on Dr. Prasad’s Moderna flu vaccine decision: https://drbretsky.com/newsletters.html). But the potential is real: for patients with pancreatic cancer, ALS, or aggressive leukemias, a 12-24 month reduction in approval timelines could be transformative.

2. Spring Respiratory Trends: Parainfluenza’s Summer Rise

Spring represents a relative lull in the respiratory infection calendar. Influenza season formally concludes on April 1st, and COVID-19 positivity rates remain traditionally low as we transition into warmer months.

However, a new player is stepping into the spotlight. According to the HealthTrackRx Respiratory Trends Report for May 2026, while overall respiratory pathogen positivity remains suppressed, Parainfluenza (Types 1-4) has risen slightly in positivity rates starting the week of May 10, aligning with Parainfluenza Type 3’s typical prevalence in spring-summer months.

This is not a significant outbreak but warrants attention. Parainfluenza infections often manifest as cold-like symptoms, croup, or bronchitis, and spread easily at summer camps, daycares, and pools. Unlike influenza (which has vaccines) or COVID (which has antivirals), there is no vaccine or specific antiviral medication for parainfluenza—prevention relies on core respiratory hygiene (Figure 2 below).

For the full monthly respiratory trends report and tracking data, visit: https://www.healthtrackrx.com/top-respiratory-infections/

3. The Genetics of Early Menopause: A Multi-Center Study and New Risk Stratification

In February 2026, a landmark multi-center study was published evaluating the utility of polygenic risk scores (PRS) in stratifying women’s genetic risk for early menopause. The study, “Genetic risk stratification and risk factors of early menopause in women: a multi-center study utilizing polygenic risk scores,” analyzed genotyping data from over 1,000 women and established a validated PRS model to predict early menopause risk. The findings underscore that early menopause—occurring before age 45, with onset of perimenopause in the 40s or earlier—is a genetically stratified condition that can be predicted and managed with precision.

The menopausal transition, or perimenopause, typically unfolds in the early 50s, marked by irregular cycles, variable hormone levels, and the gradual waning of ovarian function. For women with genetically determined early menopause, this transition compresses the reproductive lifespan and brings an earlier onset of menopausal symptoms—hot flushes, night sweats, insomnia, vaginal dryness, mood changes, and cognitive shifts. The etiology of early menopause is heterogeneous—autoimmune, iatrogenic (chemotherapy, radiation), chromosomal, or idiopathic. The recent multi-center study identified high-risk and intermediate-risk groups based on PRS percentiles, demonstrating that genetic stratification can meaningfully differentiate women who will experience early menopause from those who will not.

At Santa Monica Primary Care, we collaborate with Biography Health to integrate whole genome sequencing and AI-powered genomic insights into our understanding of reproductive endocrinology. For a woman presenting with early menopause—whether through symptoms, biochemical evidence, or other findings—genomic data can clarify whether her trajectory reflects a genetically determined earlier transition versus an acquired process (thyroid autoimmunity, ovarian antibodies, metabolic stress).

A case example: Consider a 45-year-old woman with an elevated early menopause PRS, now presenting with biochemical confirmation of menopause: FSH 71.3 mIU/mL, LH 51.5 mIU/mL, and estradiol 35.4 pg/mL—all in the postmenopausal range. Concurrent hormonal assessment reveals low testosterone (<3 ng/dL) and low DHEA-S (44.5 μg/dL), consistent with a comprehensive decline in ovarian and adrenal steroidogenesis. The genomic PRS validates that this trajectory is not incidental or artifact—her genetic architecture predisposes her to an earlier menopausal transition.

Hormone replacement therapy: Moving beyond the Women’s Health Initiative. For decades, Women’s Health Initiative (WHI) data hung over HRT discussions like a dark cloud. Published in 2002, WHI reported increased breast cancer and cardiovascular risk in postmenopausal women using conjugated estrogen plus medroxyprogesterone. The fear was warranted—but the nuance was often lost. WHI enrolled women with an average age of 63, many years into menopause, many with metabolic risk factors. The regimen (conjugated equine estrogen + MPA) was one of many formulations. The trial was not designed to address symptomatic women in the transition or women with early menopause seeking bone and cardiovascular protection.

Subsequent research and meta-analyses have reframed HRT risk-benefit, particularly for women initiating therapy in their 40s or early 50s. The cardiovascular risk appears minimal in younger women, the breast cancer risk is modest and reversible upon cessation, and the bone protection is substantial. For a woman with genetic predisposition to early menopause, the discussion shifts further: earlier estrogen decline accelerates bone loss and lipid perturbation. Baseline bone mineral density assessment (DEXA) is now recommended to establish a starting point, independent of osteoporosis PRS—her genetic risk for early menopause is itself a risk factor for accelerated bone loss.

Cardiovascular considerations: Early menopause can adversely affect the lipid trajectory and vascular function. This woman’s elevated LDL-C and apoB, now viewed through the lens of an estrogen-depleted state, take on additional urgency. Estrogen’s anti-atherogenic effects—improved lipid metabolism, improved endothelial function, reduced inflammation—are forfeited earlier in women with genetic early menopause. These factors should inform her overall cardiovascular risk reduction strategy alongside pharmacotherapy.

Clinical decision-making: Decisions regarding hormone therapy should be made in consultation with gynecology or reproductive endocrinology, weighing symptom burden (hot flushes, night sweats, vaginal dryness, mood changes, insomnia), bone protection (DEXA baseline, fracture risk), and cardiovascular protection (lipid trajectory, vascular risk markers, family history). For young women with genetically determined early menopause, the benefits of timely HRT often outweigh the risks. This is Precision Medicine in action: genomic data, biochemical markers, and clinical judgment converging to personalize therapy.

Reference: Zhong W, et al. Genetic risk stratification and risk factors of early menopause in women: a multi-center study utilizing polygenic risk scores. Frontiers in Endocrinology. February 2026. https://www.frontiersin.org/journals/endocrinology/articles/10.3389/fendo.2024.1518288/full

4. The 2026 Ebola Outbreak: Declining Mortality, Misguided Policy, and the Specter of Hemorrhagic Fever

As of June 22, the Democratic Republic of Congo has confirmed over 1,000 Ebola cases—the second largest outbreak on record and the 17th in DRC. Cases are rising faster than in any prior outbreak. Bundibugyo virus drives the epidemic; no vaccine or antiviral exists.

Historical context: Ebola emerged in 1976 in two simultaneous outbreaks: Sudan and Zaire (now DRC). A filovirus—belonging to the family that includes Marburg—Ebola causes hemorrhagic fever: fever, myalgia, and catastrophic vascular leak leading to shock and multi-organ failure. Early outbreaks showed mortality rates up to 90%.

Yet Ebola differs fundamentally from other viruses we have highlighted in prior newsletters—measles, COVID-19, RSV. It is not airborne. Transmission requires direct contact with blood or body fluids, or contaminated needles and medical equipment. There is no animal reservoir; humans are the only host. This biological fact should limit spread in modern healthcare settings—yet in 1976 Zaire, the virus found its way into a hospital lacking basic infection control. Healthcare workers died. Patients died. Families caring for the sick died. The disease turned medicine itself into a vector. That terror shaped how the world understood Ebola: not as a public health inevitability like measles, but as a catastrophic breach of medical safety.

Returning to 2026: Modern supportive care and experimental treatments have transformed Ebola’s trajectory. Without treatment, 90% of cases are fatal. With early intervention and supportive care, the current outbreak shows 55-60% mortality—a meaningful decline from the pre-treatment era. This reflects decades of epidemiological study, clinical experience, and therapeutic innovation. Every percentage point reduction represents lives saved.

Yet institutional fear and political calculation can override evidence-based practice, and nobody does that better than the RFK Jr.-led HHS. In 2014, when Americans were infected with Ebola, Dr. Anthony Fauci donned a Positive Pressure Personnel Suit and provided direct clinical care (https://www.science.org/content/article/why-nihs-anthony-fauci-treating-ebola-patients-himself). In 2026, RFK Jr. prefers to quarantine American patients in Kenya—an unwilling neighbor with no current cases—generating violent protests in Nanyuki and undermining the very public health principles that enabled mortality decline. One approach conveys scientific understanding—the other epidemiologic nonsense.

5. Respiratory Vaccine Composition for 2026-2027: COVID-19 and Influenza Strain Selection

As fall approaches, both COVID-19 and influenza vaccines for 2026-2027 are finalized. The FDA’s strain recommendations reflect sound epidemiological science but were complicated by eroding surveillance capacity.

COVID-19: Monovalent XFG-Targeted

The FDA recommended monovalent JN.1-lineage XFG vaccines for 2026-2027, shifting from 2025-26’s LP.8.1. XFG emerged as the dominant global variant for six months, favoring a monovalent approach that maximizes immunological focus and manufacturing efficiency.

The decision process revealed surveillance gaps: CDC data at the VRBPAC meeting was a month old—nearly useless for tracking RNA viruses. Sequencing volumes and state data sharing have declined, reflecting losses in government surveillance staff and academic funding.

Given the lack of meaningful vaccine advice from HHS/CDC, we follow EU and Australian guidelines. These recommend twice-annual COVID vaccination for adults over 75, and once-annual vaccination for those over 65. Annual vaccination also remains reasonable for immunocompromised individuals and those with chronic pulmonary or cardiovascular disease. For younger, healthy individuals without risk factors, vaccination decisions should reflect individual exposure risk and prior infection history.

At Santa Monica Primary Care, we measure COVID antibodies (spike protein and nucleocapsid) to gauge blended immunity—the combination of vaccine series protection and community-acquired immunity. This personalized approach informs vaccination timing and frequency for individual patients.

Influenza: All Three Components Updated

On March 13, 2026, the FDA updated all three flu vaccine components: A/Missouri/11/2025 (H1N1)pdm09-like, A/Darwin/1454/2025 (H3N2)-like, and B/Tokyo/EIS13-175/2025 (B/Victoria lineage)-like. The H3N2 update targets subclade K, which circulated widely in 2025-2026.

Clinical Guidance

For influenza: All individuals 6 months and older should receive the seasonal flu shot. Vaccination reduces infection risk and, if breakthrough infection occurs, significantly reduces severity. Aim for September or October before peak transmission.

6. Follow Us on Social Media

Twitter (X): @santa_care
Facebook: https://www.facebook.com/santamonicaprimarycare
Instagram: @santa_monica_primary_care
LinkedIn: https://www.linkedin.com/in/philip-bretsky-mdphd/
BlueSky: @bretskymd.bsky.social
Substack: @bretskymd

Follow our blog posts on: www.drbretsky.com/blog

Archived issues of our newsletter can be found at: https://drbretsky.com/newsletters/

If you know somebody who would like to sign up to receive our newsletter, they can do so at:
https://lp.constantcontactpages.com/su/JtF9zzR

From The Department of Current Infectious Disease Outbreaks in the United States, some welcome news has arrived here at the end of April, namely the conclusion of the South Carolina measles outbreak. As case numbers remained steady at 997 for 42 days (two full measles incubation periods), public health officials officially declared the outbreak over on April 27th.

We then move into a discussion of the Coronary Artery Calcium (CAC) can. As many of our patients already know, this is a quick, non-invasive heart imaging study using a CT scanner. It looks for evidence of calcium deposits in the arteries that supply blood to the heart muscle. In our clinic, we use CAC scores to improve our individualized cardiovascular risk prediction and to set lifestyle or pharmacologic treatment goals.

With more patients asking about Peptides, we then cover some basic contours of these compounds, how they are regulated (spoiler alert, they aren’t) and what steps will need to be taken to show their safety and efficacy. Lastly, we finish with data on another, more famous, peptide, namely the GLP-1 receptor agonists, and what happens when these medications are stopped.

Our Social Media links are also listed at the bottom of this Newsletter. You can follow us there. Copies of this Newsletter are now newly available on Substack.

1. Measles in South Carolina

On April 27th, the South Carolina Department of Public Health (DPH) recently declared an end to a major measles outbreak which began in October 2025. With a total of 997 confirmed cases and 21 hospitalizations in the six months of the outbreak, it was the largest seen in the United States since 1991.

Previous newsletters have covered the epidemiology and transmission patterns of measles. But as a brief recap, measles in 1948 was described by Kenneth Maxcy (Footnote #1) as the “simplest of all infectious diseases.” Why would he say such a thing? Well it is simple insofar as it is a nearly universal disease and is spread only person-to-person via aerosol (primarily sneezing). There is no animal vector or reservoir. Because measles is so infectious, quarantine is ineffective and used only as a temporary measure in small, isolated populations to permit vaccination efforts. Measles incidence is widely considered one of the most sensitive and “honest” indicators of a public health system’s overall strength and infrastructure.

A two dose series of the measles vaccine (MMR or MMRV) is 97% effective and, for the vast majority of people, provides lifelong immunity. The vaccine administered within 72 hours of exposure to the virus can still prevent disease or significantly reduce the severity of symptoms.

As of April 17th, South Carolina reported a stable count of 997 measles cases, 95% of which were concentrated in Spartanburg County. This appears to be the conclusion of one of the most significant public health events in recent US history. This outbreak began in early October 2025 with the South Carolina Department of Public Health confirming 8 cases in the Upstate region of the state (link: https://dph.sc.gov/news/dph-confirms-measles-outbreak-upstate-region). Two of these cases had no history of travel, therefore implicating previously “unrecognized community spread.” Two schools subsequently were identified as locations with confirmed measles cases. Only 21% of students at one school and 82% at the other were up-to-date on school vaccinations, both far below the threshold 95% required to keep measles at bay.

By mid-October South Carolina health officials were optimistic that they had controlled the spread of measles stating:

“The quarantine of students from Global Academy and Fairforest Elementary appears to be working. The recent confirmed cases in those exposure groups did not lead to additional case investigations because they were already in quarantine.”

Those of you who read the opening paragraph of this section will already know that quarantine is essentially ineffective so what happens next will not be a surprise. But for completeness here is what happened (Figure 1 below)

Figure 1: New and cumulative measles outbreak cases reported by the South Carolina Department of Public Health.

Those numbers continued to climb throughout April, leveled off at 997 and remained steady therefore permitting officials to formally declare an end to the outbreak. Unsurprisingly, unvaccinated individuals accounted for 95% of the cases, partially vaccinated (one dose of the two series MMR) were 2% of the cases and 2.5% of cases were amongst those considered to be fully vaccinated. Children and adolescents were disproportionately affected with those 5-17 years old making up 60% of cases and those under 5 accounting for 26% of cases. Adults 18+ were only 10% of the total case count.

Ultimately, if we accept Maxcy’s premise that measles serves as the most “honest” indicator of a public health system’s integrity, then the Spartanburg outbreak stands as a scathing indictment of the current US policy trajectory spearheaded by RFK Jr.. For decades, the 95% vaccination threshold was treated as a non-negotiable floor for community safety; however, under the “medical freedom” framework favored by Health and Human Services, that floor has been effectively dismantled in favor of individual autonomy over collective wellness. This 997-case surge was not a failure of the vaccine’s efficacy—that remained expectedly stable at 97% (because the vaccine works!)—but a failure of a policy environment that prioritizes vaccine hesitancy over evidence-based practice. When schools are permitted to operate with uptake rates as low as 21%, we are no longer practicing public health; we are conducting a high-stakes epidemiological experiment on our children. As this outbreak concludes, the lesson is clear: when policy undermines the infrastructure required to keep the “simplest of all infectious diseases” at bay, the resulting morbidity is not just a clinical inevitability—it is a choice.

Footnote #1: Dr. Kenneth Maxcy (July 27, 1889 - December 12, 1966) was a professor of epidemiology at Johns Hopkins School of Public Health and was an authority on rickettsial disease and typhus specifically. He was a member of the Department of Defence’s (now Department of War) Committee on Biological Warfare. He also chaired the committee that recommended fluoridation of drinking water in 1952.

2. Coronary Calcium Scoring, Limits and Expectations

As many of our patients already know, we have begun to recommend the Coronary Calcium Scan more widely as a way to delineate the presence of coronary artery disease or understand their individual risk of developing such. Traditionally, we have relied on population-based risk calculators such as the ASCVD (Atherosclerotic Cardiovascular Disease) risk tool which calculates the “statistical” chance of a heart attack, stroke, or peripheral artery disease based on age, gender, hypertension, diabetes, and cholesterol measures. While these risk estimator tools are useful for public health policy, they lack the specificity for the individual patient sitting in my office.

\Instead, we have increasingly implemented Coronary Artery Calcium (CAC) scanning as a standard clinical tie-breaker. We widely recommend it even for those at low or average risk, typically beginning in the 40s for men and 50s for women. This decade-long gap reflects the well-documented epidemiological lag in cardiovascular disease (CVD) between the men and women, ensuring we catch the “biological” start of the disease rather than just waiting for an event to occur before intervening (aside: for those that are interested, this is the difference between “primary” and “secondary” prevention. For more on that see: https://www.ncbi.nlm.nih.gov/books/NBK537222/ which is still available at the NIH Library, surprisingly it hasn’t been banned yet).

The tool we use to quantify this risk is the Agatston Score, named after Dr. Arthur Agatston more famously known for the South Beach Diet craze in the mid-90s—a plan that prioritized “good fats” and lean proteins over the aggressive carb-cutting of Atkins. While Agatston’s nutritional approach was lacking (2 year follow-up shows that weight lost during the intensive phases is frequently regained), his more enduring contribution to cardiology is development of this density-weighted CT scoring system. The Agatston score isn’t just a measure of how much calcium is present; it accounts for the density of the plaque as well. Because calcium is a late-stage marker of the atherosclerotic process, a score above zero confirms the transition from early cholesterol deposition to more established plaquing.

From a practical standpoint, the CAC scan is one of the most cost-effective metrics in modern medicine. Because insurance remains stubbornly behind the clinical curve, this is almost always an out-of-pocket expense, typically ranging from $100 to $150. Unfortunately, it is not for everybody as those patients with a smoking history, diabetes or elevated Lp(a) are more appropriate candidates for the Coronary CT Angiogram (CCTA), which can cost upwards of $2,500. The CCTA is the superior tool for identifying the “soft,” non-calcified plaque that often precedes a major event, something more likely to occur in our highest-risk patients.

We use these scores to guide the initiation and intensity of therapy using well validated population data, specifically that from the MESA (Multi-Ethnic Study of Atherosclerosis) Trial to predict 10-year risk:

• CAC 0: A powerful negative predictor indicating an extremely low risk (less than 1 in 200 chance) of a major cardiac event over the next 5 years

• CAC 1-99: Mild but some burden of atherosclerotic plaque. Can consider pharmacologic management but stress achieving an optimal BMI, rigorous blood pressure control (targeting <120/80), and a Mediterranean dietary approach.

• CAC 100–300: This is often our trigger to initiate statin therapy, aiming for a target LDL < 100 mg/dL.

• CAC > 300: At this level, your risk is statistically identical to that of a patient in “secondary prevention”—meaning someone who has already had a heart attack. For these individuals, we drive the LDL to < 55 mg/dL and aggressively manage blood pressure toward a goal of 120 mmHg.

Ultimately, the most important distinction for both patients (and physicians, many of whom do not understand this nuance) is that a high CAC score is a risk-stratification tool, not a functional diagnostic for a blockage. A score of 400 identifies a high-risk patient, not necessarily a high-risk lesion. Identifying a high-risk lesion requires functional stress testing, something that often goes hand in hand with a CAC amongst my higher risk patients. The goal is to treat the systemic environment that allowed the plaque to form in the first place. We treat the patient’s vascular milieu, not the specific “speck” on the scan. As I tell my patients: the scan exists to tell us how hard we need to work on your prevention strategy, not to go looking for trouble where none yet exists.

There are a number of great resources for those interested in learning more about the Coronary Calcium Scan. A curated list below is a useful starting point:

Using CAC score to stratify the risk of heart disease: https://pmc.ncbi.nlm.nih.gov/articles/PMC11198403/

Applying CAC Scores for Primary Prevention of ASCVD:

https://www.jacc.org/doi/full/10.1016/j.jcmg.2024.12.008

The Power of a Zero Calcium Score:

https://www.ahajournals.org/doi/10.1161/CIRCIMAGING.124.017714

Use of Coronary Artery Calcium Scoring in Individuals With Elevated Lipoprotein(a)

https://pubmed.ncbi.nlm.nih.gov/41837904/

3. Peptides: All You Wanted to Know But Were Afraid to Ask

In recent months, the American health consumer has entered a period of unprecedented fascination with peptide therapy. What was once a niche protocol whispered about in biohacking circles and bodybuilding forums has exploded into the mainstream, fueled by a surge in social media interest and high-profile endorsements from figures like Joe Rogan and Health Secretary RFK Jr. In fact, the FDA recently announced a summer meeting to ease limits on compounding for seven peptides including BPC-157 (which we discuss in more detail below). As it stands now, peptides are accessible via a “For Research Use Only” label, which is a legal loophole that allows companies to sell chemicals not intended for human consumption, bypassing all FDA safety, purity, and manufacturing standards.

To ground this conversation, let’s first start with some Chemistry. Peptides are simply short chains of amino acids—typically fewer than 50—linked together. They function as the structural middle ground between a single amino acid and a full-sized protein. In a strict clinical sense, we have been using peptides as medical treatments for a century; insulin and the transformative GLP-1 receptor agonists (like Ozempic and Mounjaro) are, point in fact, peptides. However, it is not insulin that we are talking about when referring to “peptides.” Instead, we are referring to a subset of as yet unapproved Body Protection Compounds and growth hormone secretagogues that currently exist in a regulatory grey area.

The most common peptide compounds discussed and used include:

• BPC-157: A synthetic sequence derived from human gastric juice, widely touted for tendon and ligament repair.

• TB-500 (Thymosin Beta-4): Often paired with BPC-157 to promote angiogenesis and wound healing.

• CJC-1295 / Ipamorelin: Peptides designed to stimulate the pituitary gland to release more growth hormone.

The preclinical data (i.e. animal studies or in vitro cell cultures) for these molecules are undeniably intriguing. For instance, rats with transected Achilles tendons, BPC-157 treatment led to significantly faster functional recovery and improved collagen organization compared to controls. When looking at cellular migration patterns in the lab, TB-500 facilitates the migration of repair cells to the site of an injury. This has clear potential application for heart tissue repair after a heart attack, for instance, as well as athletic injuries. CJC-1295 and Ipamorelin work by mimicking GHRH (Growth Hormone Releasing Hormone) to trigger the pituitary gland. Laboratory data suggests these can improve lean mass and sleep quality.

However, evidence of efficacy in humans remains vanishingly thin. For BPC-157, the total number of humans studied in registered safety trials is 14 people. To put that in perspective: statin therapy for high cholesterol has been studied in meta-analyses encompassing over 170,000 participants. If we are to take these biological tools seriously—and I believe we should—they must be subjected to the same “honest” scrutiny as any other intervention. We cannot fast-track safety. To move these from the “gray market” to the clinic, we require:

1. Phase II and III Double-Blind RCTs: We need to move beyond 14-person pilot studies and establish real-world efficacy and long-term safety profiles.

2. Standardized Dosing & Purity Controls: Currently, “research-only” peptides arrive in the mail with no guarantee of concentration or the absence of contaminants.

3. Vigilant Pharmacovigilance: We must be able to track adverse events, particularly since some animal studies suggest that peptides like BPC-157, which promote blood vessel growth, could theoretically accelerate the growth of undiagnosed cancerous cells.

Ironically, the FDA recently fired the one person who consistently championed fundamental questions of clinical integrity (albeit rather haphazardly), Dr. Vinay Prasad. He has argued that when we allow drugs to reach the market based on surrogate endpoints, single-arm trials, or, in the case of BPC-157, essentially no human data at all, we aren’t just “innovating”—we are compromising the safety of the public for the sake of speed. While his primary objections were “non-inferior” drugs brought to market in the field of Oncology, there is clear applicability for his approach when considering peptides. We need rigorous, placebo-controlled evidence that proves these compounds help people live longer or better, rather than just “healing” a rat’s tendon. Until we have that, we must remain healthily skeptical.

4. GLP-1 Receptor Agonists: What Happens When You Stop?

The landscape of metabolic (especially but not limited to diabetes and obesity) health has been fundamentally reshaped by GLP-1 receptor agonists (GLP-1s) like semaglutide (Ozempic, Wegovy) and the dual-agonist tirzepatide (Mounjaro, Zepbound). Originally developed to manage Type 2 Diabetes (T2DM) by enhancing insulin secretion, increasing insulin sensitivity at the cellular level and slowing gastric emptying, their efficacy in inducing significant weight loss—often 15% to 20% of total body weight—has led to a surge in use for obesity management. While these medications are transformative during active treatment, a critical question is: What happens when the weekly injections stop?

It will probably come as no surprise that upon stopping these medications, both the weight and lack of glycemic control return. Two new studies underscore that obesity is a chronic, relapsing condition rather than a temporary hurdle. A large-scale meta-analysis reveals that after stopping GLP-1 RAs, patients regain weight at an average rate of 0.4 kg (nearly 1 lb) per month. On average, individuals return to their baseline weight within just 1.7 years of cessation (link: https://jamanetwork.com/journals/jama/fullarticle/2812936). Of interest, the rate of weight regain is faster than we see when compared to those stopping behavioral weight management programs (such as Weight Watchers) alone.

The concerns after stopping aren’t related to weight alone, but extend to blood pressure, cholesterol and glycemic control. During treatment, we typically see marked improvements in these parameters. But after stopping, blood pressure and triglycerides tend to return to baseline levels within approximately one year of stopping the medication. Improvements in HgA1c (a three month average of glucose levels) and fasting glucose levels reverse as early as 4 months after the last dose. As such, stopping GLP-1 RAs may reverse the protective cardiovascular benefits—such as reduced risk of major heart events—that were achieved during active therapy (link: https://www.bmj.com/content/392/bmj-2025-085304).

A post hoc analysis of the SURMOUNT-4 trial (referenced above) found that 82.5% of participants who stopped tirzepatide regained at least 25% of their lost weight within a year. Notably, those who managed to limit their gain to less than 25% were the only ones who successfully maintained their improvements in waist circumference and insulin sensitivity.

The takeaway message is clear: these medications are highly effective tools, but they are generally not “short-term fixes.” For many, they require a long-term commitment to either maintenance dosing or an incredibly rigorous transition to lifestyle-based weight maintenance to prevent the reversal of metabolic progress. For anybody considering a change in their GLP-1 regimen, a structured taper combined with a reinforced behavioral plan to protect any health gains will be paramount.

5. Follow Us On Social Media

Twitter (X): @santa_care

Facebook: https://www.facebook.com/santamonicaprimarycare

Instagram: @santa_monica_primary_care

LinkedIn: https://www.linkedin.com/in/philip-bretsky-mdphd/

BlueSky: @bretskymd.bsky.social

Substack: @bretskymd

Follow our blog posts on: www.drbretsky.com/blog

Archived issues of our newsletter can be found at: https://drbretsky.com/newsletters/

If you know somebody who would like to sign up to receive our newsletter, they can do so at: https://lp.constantcontactpages.com/su/JtF9zzR

Welcome to the March 2026 Newsletter for Santa Monica Primary Care.

The Clinical Trials Edition

This past month was a big one for results coming out of two highly anticipated clinical trials, each one resulting in some degree of controversy. One of these trials involved a new influenza vaccine, and the fallout involving such ultimately led to the ouster of FDA official (and COVID contrarian) Dr. Vinay Prasad. Ironically, the second time Dr. Prasad has been fired from the FDA within a year.

The second trial was the anticipated result of Galleri’s Multi-Cancer Early Detection (MCED) which is a screening blood test seeking to identify any of more than 50 cancers. This is a blood test that we have offered in our clinic for the past several years. The Galleri trial was a three year investigation involving 140,000 National Health Service (NHS) participants in the United Kingdom. It did not perform as well as researchers expected.

We start out the newsletter with some background on Clinical Trials - why they are done, how they are constructed and some nomenclature to familiarize readers with them. Building on that framework, we will look at the results of both of these clinical investigations. For the Moderna flu-vaccine study, we cover the FDA’s initial response (largely framed by Dr. Prasad), and its subsequent reversal (presumably framed by somebody other than Dr. Prasad).

Lastly, we finish with some news from the American College of Cardiology who “generously” updated their lipid (cholesterol) guidelines in a 132 page document. We have distilled that into a few, hopefully more readable, paragraphs with one great graphic.

Some of our Social Media posts this past month include a summary of seasonal influenza activity, announcing Dr. Bretsky’s 10th consecutive year as a Southern California “Super Doctor” and a video link of his participation in USC’s Keck School of Medicine roundtable discussion of physician career paths in “Beyond the White Coat, MDs Reimagined”.

1. Background on Clinical Trials: Purpose, Design and Endpoints

Clinical trials are structured experiments in humans, designed to answer a specific medical question—namely, does an intervention or diagnostic work, how well, and how safely? To answer these questions, a representative group of individuals are “randomized” into (typically) one of two groups: an intervention group and a control (which may be placebo or no treatment) group. As we discuss further below, the choice of control group anchors the degree to which investigators can be confident about their results and is of particular interest to Dr. Vinay Prasad - formerly the Director of the Center for Biologics Evaluation and Research (CBER) and the Chief Medical and Scientific Officer at FDA. He was tasked with overseeing vaccine, cell/gene therapy, and blood product regulation, with the goal of playing a major role in reshaping FDA approval standards. For his efforts, he was fired - twice. We will discuss how control group selection was problematic in Moderna’s study design.

Endpoints are another critical study design choice and can be event-based (mortality, disease / illness) or a surrogate endpoint (a biomarker for example like cholesterol level). A surrogate endpoint trades speed for accuracy. We will see in reviewing data from Grail’s Galleri multi-cancer detection test that while primary endpoints may not show efficacy, secondary endpoints add nuance to those findings. So while a test may not have “worked” as anticipated in the general population, there would still be pockets of applicability in a more selected patient group.

Because human subjects are involved, clinical trials always require informed consent, prespecified protocols, and monitoring by Review Boards. In the United States., regulatory pathways are run through the FDA, with public registration and results reporting encouraged via ClinicalTrials.gov (maintained by the National Institutes of Health). The FDA’s response to the Moderna mRNA flu vaccine application highlights the chaos of these storied institutions emblematic of our current administration’s oversight and appointing individuals into positions for which they lack qualification, expertise or experience.

2. Moderna’s mRNA influenza vaccine results

It has been no secret that the current administration has no interest in mRNA derived vaccines or therapeutics. This, despite the fact, that mRNA research (which has applications both in communicable disease and cancer) has been funded by the National Institutes of Health (NIH) for decades - dating as far back to the 1960s. However, in 2025, the US Department of Health and Human Services (HHS) elected to wind down late-stage mRNA vaccine development through its Biomedical Advanced Research and Development Authority (BARDA) program. The rationale for such, as stated by Robert F. Kennedy Jr. (our well-documented vaccine skeptic and shirtless, jeans-wearing cold plunge video maker in residence) was to shift towards “safer, broader vaccine platforms that remain effective even as viruses mutate.” This assertion, like many others that he makes, has no evidence supporting such a generalization. Point in fact, no vaccine remains effective as viruses mutate which is why, for instance, we have new, seasonal influenza vaccines each year. In terms of safety, mRNA vaccines compare favorably to other vaccine platforms (e.g. adenoviral vector, inactivated, protein subunit, and live-attenuated) although do have more short-term, treatable side effects with very low rates of serious adverse events.

It was against these headwinds that Moderna’s investigational mRNA-1010 seasonal influenza vaccine (an mRNA-based vaccine designed for adults 50 years of age and older) presented efficacy data to the FDA this past February. It underwent a Phase 3 trial (this is a large-scale clinical trial design intended to be definitive in terms of result accuracy) involving over 40,000 participants during the 2022/2023 and 2023/2024 influenza seasons. Initial results (published March 2025) were encouraging, showing about ~26.6 % superior relative vaccine efficacy compared with standard-dose flu shot.

The “standard-dose” flu shot comparison group, however, became a core issue for the FDA (overlooking the fact that prior reviews with the FDA had failed to identify this as a design flaw) as those over 65 years of age are recommended to have a high-dose influenza vaccine which contains about 4-times the antigen of standard vaccines. While the higher dose vaccine does lead to more side effects, a pivotal trial showed a modest but clinically meaningful improvement in influenza rates - which was 1.4% amongst those receiving a high-dose vaccine and 1.9% amongst those receiving a standard-dose vaccine. As such, the current standard of care recommends a high-dose flu vaccine for those 65 years of age or older, and standard dose flu vaccine for those under the age of 65.

In early February, the FDA issued a rare “Refusal-to-File” letter, signed by Dr. Prasad, declining to begin formal review of Moderna’s biologics license application for mRNA-1010. The stated reason was procedural: the Phase 3 trial compared mRNA-1010 only against a standard-dose influenza vaccine rather than a high-dose version recommended for older adults. As such, the 65+ control group did not reflect the “best-available standard of care” which is a very reasonable critique and should have been addressed in the design phase before recruiting the over 40,000 clinical trial participants.

Of the 40.000 participants, approximately 30,000 were under the age of 65. And, for this group, a standard dose flu-vaccine as a comparison would be appropriate. But the decision to move forward with a formal review was congruent with the stated positions of Dr. Prasad, who has long held that new therapeutics (primarily in Oncology which is his area of expertise), should demonstrate meaningful advantage over current standards before approval. This philosophy is rather at odds with how some clinical trials (most often vaccines, antibiotics, anticoagulants, and devices) have traditionally been structured which is termed ‘non-inferiority’ studies. In such, a therapeutic may reasonably be approved if not worse than an established effective treatment by more than a pre-specified margin. A treatment may be even acceptable even if slightly less effective if it offers other advantages such as improved safety, fewer side effects, lower cost, or easier dosing.

Understandably upset with the decision, Moderna released the refusal-to-file letter publicly and a week later the FDA reversed its decision and agreed to accept the application for review. Soon after Dr. Prasad announced he would step down as CBER director, electing to return to his academic position at UCSF.

All told, the saga of Moderna’s mRNA flu vaccine illustrates the capriciousness of the FDA in this current administration. Despite a well-designed clinical trial with meaningfully positive results, the agency’s initial decision was to sink the entire project. Of course, there was a very reasonable critique of the inadequate control group amongst those 65 years of age and older. However, that was only 25% of the study. Further, regulators showed little practical or real-world knowledge of how therapeutics come to market - unsurprising given that Dr. Prasad is an academic Oncologist with no actual regulatory experience. Students of history would know that the current high-dose influenza vaccine for seniors was on the market for 4 years before randomized clinical efficacy data demonstrating that it reduced influenza infection were published. Why did this happen? Well, the vaccine was not approved “without evidence” — it had well-demonstrated higher immunogenicity (it generated higher antibody levels), an established safety profile and biologic plausibility (4× antigen dose). But clinical efficacy against influenza illness itself was confirmed later.

In all likelihood, a similar scenario with Moderna’s mRNA flu vaccine will occur wherein effectiveness amongst those 65 years and older will be demonstrated from post-marketing studies. But Dr. Prasad, like the Sith, wanted to deal in absolutes which - in the real world - don’t really exist. As it stands, the FDA has now accepted Moderna’s revised submission and set a target date (August 2026) for completing review that could affect availability for the 2026–27 flu season.

3. Galleri’s Multi-Cancer Detection Test, NHS Results

The next clinical trial we will look at is the first key trial for Galleri’s multi-cancer early detection (MCED) blood. While there are several on the market (Cancerguard, CancerSEEK for example) Galleri’s Multi-Cancer Early Detection (MCED) test is probably the best known of them. It screens for more than 50 cancers (with varying degrees of accuracy) via a blood test and is something we have been offering to patients in our office for several years. As it does not yet have FDA approval, it is not covered by insurance. These tests typically cost somewhere between $650 and $950.

Galleri elected to conduct their clinical trial with the UK’s National Health Service (NHS), most likely because of a unified follow-up schema that would identify new cancer cases, unlike the US healthcare system which is fractionated. The population studied included 142,000 healthy (no active symptoms) adults between the ages of 50 and 77. None could have a cancer diagnosis within the previous three years.

Galleri employed an unusual 3:1 case to control design with 104,000 assigned to essentially annual Galleri blood testing (three blood tests over about 2 years of follow-up) and the remaining 38,000 to usual care (no Galleri testing). The primary outcome of interest was to see if those in the Galleri testing group had lower rates of Stage 3 and 4 cancer diagnoses compared to usual care. In other words, was Galleri picking cancer up at earlier and more treatable stages?

In short, the answer is no. The clinical trial failed to demonstrate a statistically significant reduction in Stage 3 and Stage 4 cancers amongst those who received Galleri testing (note that only “top line” results have been reported but an actual peer-reviewed paper has not yet been published). They did report, however, a more than 20% reduction in Stage 4 diagnoses amongst a prespecified group of 12 high-mortality cancers (bladder, esophageal, stomach, head/neck, liver/bile duct, lung, ovarian, pancreatic, sarcoma, small intestine, uterine, and other gastrointestinal cancers).

Notably, the clinical trial used a 3:1 design which serves a practical purpose (primarily the NHS wanted more participants to potentially benefit from screening) but reduces the ability (termed statistical “power”) to detect between group differences. For instance, the Moderna mRNA influenza vaccine trial discussed in Section 2 employed a 1:1 intervention to control design which maximizes statistical power. With a 3:1 design and a rare outcome of interest (late-stage cancers are relatively uncommon in populations with regular screening as would be expected amongst NHS participants), the Galleri test could potentially shift stage distribution rather than the overall effect. In other words, Galleri testing could convert Stage 4 to Stage 3 and Stage 3 to Stage 2.. As such, Stage 4 decreases while Stage 3 increases. But because the outcome of interest is Stage 3 and 4 cancers, the net effect results in a null finding.

Further, the company assumed a 20% reduction in late-stage cancer incidence amongst those that were screened, which is a very aggressive target in a short (2 year) follow-up. With a 1:1 design, the clinical trial would be more than 90% likely to detect a difference between the two groups. However, with a 3:1 design, that value dropped to about 75%. If the true reduction were smaller (say 10% of 15%), the likelihood of detecting a difference drops even more precipitously - to 20% if the true reduction is 10%, and 50% at 15%.

The company attempts to correct for this design flaw by reporting the analysis of Stage 4 diagnoses among 12 high-mortality cancers, which are relatively rare. But this doesn’t do much to fix the headline finding, namely that the Galleri test failed to show an effect. Galleri would have been better served by being more conservative in their design assumptions and employing a 1:1 intervention to control participant distribution.

So do these results make Galleri a useless test? Actually it does not but, for now, it moves its clinical utility from asymptomatic individuals (i.e. the general population) to higher risk populations. MCED performance improves as the likelihood of cancer incidence rises and, as such, those with known environmental cancer risk factors (occupational exposure, heavy cigarette or alcohol intake), strong family history, chronic inflammatory disease and cancer survivors would be reasonable individuals amongst whom to employ the test. Further, MCED testing has an advantage in cancers with no effective existing screening such as pancreatic, ovarian, and bile duct cancers. Note that this list overlaps with the 12 high-mortality cancers identified by Galleri in their sub-analysis. But this does move MCEDs away from universal testing and into a more appropriate risk -stratified screening tool.

4. New Lipid Guidelines

This past week, new guidelines for cholesterol screening and management were published by the American College of Cardiology (ACC) and American Heart Association (AHA). We have distilled the 132 page document down into some usable, digestible information.

The primary change endorsed by this working committee is to employ a risk estimate (namely the PREVENT-ASVD equation) in patients as young as 30 years of age. This calculation estimates the 10 year cardiovascular risk (including heart attack, stroke, heart failure and cardiovascular death) using a patient’s age, gender, blood pressure, cholesterol, body mass index, kidney function, anti-hypertensive medication use and presence or absence of diabetes to “begin consideration” of cholesterol-lowering medication. This practice is termed “primary prevention” as it seeks to prevent an adverse health event before it occurs (as distinct from “secondary prevention” where the goal is to prevent a recurrence of an outcome). The guideline panel selected a 10 year estimated risk of 3% or above to begin to consider lipid-lowering therapy.

This earlier consideration for medical therapy (which should be considered carefully, especially amongst women of child-bearing age as statins are contraindicated in pregnancy), was modified by a lifestyle-first approach to high cholesterol. This is nicely summed up in a statement by Dr. Roger Blumenthal who chaired the guideline committee:

“While we want to try to optimize healthy lifestyle habits as the first step to lower cholesterol, we realize that if lipid numbers aren’t within the desirable range after a period of lifestyle optimization, we should consider adding lipid-lowering medication earlier than we would have considered 10 years ago. And lower LDL cholesterol for longer, just like lower blood pressure for longer, results in much greater protection against future heart attack and stroke risk.”

In addition to earlier screening and (potentially) intervention, the committee newly encouraged the addition of advanced biomarker screening outside of the traditional LDL and HDL cholesterol measurements. This includes Lp(a) and ApoB, measurements which we have been running in our office for several years now. Further, they encouraged the use of coronary calcium scans (CAC) to be employed if the decision to initiate lipid-lowering therapy is uncertain and for adults at intermediate (and some at borderline) risk with no prior history of coronary artery disease. Again, we have been utilizing coronary calcium scans in this manner as well as in estimating the 10 year risk of clinically significant coronary artery disease.

Recent data suggests that both Lp(a) and elevated CAC share independent associations with an increased risk of cardiovascular disease. In this study (link: https://www.jacc.org/doi/full/10.1016/j.jacc.2026.02.5067), the greatest risk was noted when both measures were elevated (defined as Lp(a) above 50 and CAC over 300, Figure below).

The authors make the following observation to integrate these findings:

“Lp(a) is a genetically determined exposure with implications for long-term ASCVD risk, while CAC reflects a lifetime of accumulated burden of disease. Thus, Lp(a) measurement may be particularly useful earlier in life to inform a healthy lifestyle, and CAC may be helpful as individuals get older to personalize risk assessment and identify individuals who may benefit from more intensive risk factor modification.”

Lastly, the writing committee endorsed equally the adjunctive use of three non-statin treatments to achieve goal LDL if such can not be attained by statin treatment alone. These include Ezetimibe (Zetia), a PCSK9 monoclonal antibody (e.g. Repatha, Praulent), and bempedoic acid (Nexletol). The PCSK9 monoclonal antibodies are injections every other week (which can be a barrier to initiating therapy) whereas Zetia and Nexletol are once daily pills. All can be added to any ongoing statin treatment.

5. Follow Us On Social Media

Twitter (X): @santa_care

Facebook: https://www.facebook.com/santamonicaprimarycare

Instagram: @santa_monica_primary_care

LinkedIn: https://www.linkedin.com/in/philip-bretsky-mdphd/

BlueSky: @bretskymd.bsky.social

Substack: @bretskymd

Follow our blog posts on: www.drbretsky.com/blog

Archived issues of our newsletter can be found at: https://drbretsky.com/newsletters/

If you know somebody who would like to sign up to receive our newsletter, they can do so at: https://lp.constantcontactpages.com/su/JtF9zzR

Welcome to the February 2026 Newsletter for Santa Monica Primary Care.

The CDC Mucks Up Vaccines Edition

The CDC over the past several months has been laying the groundwork for a wholesale change to vaccine policy in the United States. This started in June 2025 with RFK Jr. firing the entire 17 member Advisory Committee on Immunization Practices (ACIP). He then installed his own replacements. Included amongst these are Dr. Retsef Levi and Dr. Robert Malone who have well documented objections to mRNA vaccines as well as Dr. VIcky Pebsworth who sits on the board of National Vaccine “Information” Center (NVIC) which is a prominent vaccine-critical organization.

The next step that the CDC undertook was to change their “Autism and Vaccines” webpage to state that a link between vaccines and autism “cannot be ruled out” (link: https://www.cdc.gov/vaccine-safety/about/autism.html). This misinformation was deposited despite decades of epidemiologic, basic science and genetic research that definitively demonstrates the lack of causality of vaccination in autism spectrum conditions (side note: an actual coherent discussion of this strong evidence is in Peter Hotez’s excellent and deeply personal book “Vaccines Did Not Cause Rachel’s Autism”). Ironically, some of the best evidence that definitively disassociates autism from vaccination comes from Denmark - a country which Make America Healthy Advocates seek to emulate in terms of vaccine policy.

All of this groundwork led the CDC on January 5th to cut the number of vaccines it recommends for every child, undermining protections against several diseases. No longer recommended is protection against flu, rotavirus, hepatitis A, hepatitis B, forms of meningitis and RSV. RFK Jr. insists that the overhaul won’t result in families who want access to the vaccines losing insurance coverage - a demonstrably false claim as CDC/ACIP recommendations are inexorably tied to insurance reimbursement.

In this Newsletter edition we will begin to chip away at the accumulated misinformation set forth by the CDC. Make no mistake, the MAHA movement and revamped CDC is not working in the best interest of the US population. As we cover below, we are already seeing large scale measles outbreaks - this will only worsen. We will also cover the Hepatitis B vaccine and why the CDC/ACIP’s update guidance will inevitably lead to increases in otherwise preventable cirrhosis and liver cancer. All by restricting access to a demonstrably safe vaccine. As an adult medicine physician, I will highlight the downstream effects on adults from two perspectives. One, what does this mean for those who are currently over 18 and, two, what ripple effects will these childhood vaccination oversights have for them as they reach adulthood.

1. Hepatitis B Vaccine

The first vaccine to be dropped from the childhood immunization schedule was Hepatitis B (abbreviated HBV: Hepatitis B Virus below), a move that occurred in December 2025. Some of those reading this Newsletter (unless you are in healthcare, have specific risk factors or have traveled to a high risk region) will not have gotten a Hepatitis B vaccine as it was introduced to the childhood schedule in 1991. I did not have a Hepatitis B vaccine until I entered medical school in 1997 because of the risk of HBV transmission through accidental contamination of needles and other injected products. HBV is also transmitted sexually. Because HBV’s transmission pattern includes these pathways, it was an attractive target to the ACIP (i.e. “babies don’t share needles, why do they need a Hepatitis B vaccine?”).

Another issue that the ACIP likely considered is that the prevalence (rate) of Hepatitis B infection varies greatly across the world. In fact, chronic Hepatitis B infection is not something we commonly see in the United States, this as a direct consequence of ‘layered’ prevention which (until now) includes prenatal screening and infant vaccination. However, in other regions (China, Southeast Asia, sub-Saharan Africa, parts of the Middle East and South America), the prevalence of Hepatitis B antigen (we are all now familiar with antigens, or viral proteins, from the COVID rapid antigen testing) can be 5-15% of the population.

In these regions, the primary route of acquisition of Hepatitis B amongst infants is from their mother during birth (it is not transmitted across the placenta). Infants born to an HBV chronic carrier mother who expresses the envelope antigen (HBeAg positive) have a 70% (!) chance of becoming infected. For those who do not express the envelope antigen (a measurement of viral replication), that rate drops to 5-10%. Furthermore, the virus can be transmitted from infected household members or through casual contact in childcare settings as the virus is quite stable. This risk extends from birth and through the first 5 years of life but drops sharply thereafter. Risk of chronic HBV infection (i.e., failure to clear) is highest in infancy and falls with age such that:

• Infants: ~90% become chronically infected (~10% clear).

• Children 1–5 years: about 30% chronic (~70% clear).

• Adults: about <5% chronic (~95% clear).

Before universal Hepatitis B vaccination for infants, about 18,000 children in the United States contracted the virus annually. The 1991 introduction of universal Hepatitis B vaccination, resulted in a 90% drop in incidence rates of Hepatitis B amongst children and adolescents, nearly eliminating this chronic infection that can lead to cirrhosis, liver cancer and liver failure (Table below).

Returning to our pre-1991 policy will lead to about 16,000 children annually who become chronically infected (in 1991 there were 4.1 million US births and in 2024 about 3.6 million so this is why there are fewer total annual incident cases). Of those, 1 in 4 will eventually die from liver failure or liver cancer (hepatocellular carcinoma) and 1 in 3 will develop cirrhosis (these two are not mutually exclusive). Running this out over 20 years of misguided infant vaccine policy, we are left with 324,000 needlessly but chronically infected children translating into 81,000 deaths from an entirely preventable cancer.

But wait, you ask, how does the Hepatitis B vaccine work if the baby is already exposed to the virus during childbirth? Well, that’s an excellent question and, again, reveals how little the ACIP understands about basic viral mechanics. After exposure (either during childbirth or via a needlestick), HBV doesn’t instantly bind to the liver. Like many other viruses, there is a ‘window period’ during which HBV only begins to seed liver cells. By inducing enough antibodies, it is possible to prevent the establishment of infection (as a caveat for completeness the standard of care can also include passive neutralizing antibodies in high-risk exposures during childbirth).

So why universal vaccination for infants (if the above were not convincing enough) when we can just test mothers and vaccinate infants accordingly? This is what FDA Chief Dr. Marty Makary thinks we should be doing, namely bringing US vaccine policy more in line with countries like Denmark (source: Welcome to the February 2026 Newsletter for Santa Monica Primary Care.

The CDC Mucks Up Vaccines Edition

The CDC over the past several months has been laying the groundwork for a wholesale change to vaccine policy in the United States. This started in June 2025 with RFK Jr. firing the entire 17 member Advisory Committee on Immunization Practices (ACIP). He then installed his own replacements. Included amongst these are Dr. Retsef Levi and Dr. Robert Malone who have well documented objections to mRNA vaccines as well as Dr. VIcky Pebsworth who sits on the board of National Vaccine “Information” Center (NVIC) which is a prominent vaccine-critical organization.

The next step that the CDC undertook was to change their “Autism and Vaccines” webpage to state that a link between vaccines and autism “cannot be ruled out” (link: https://www.cdc.gov/vaccine-safety/about/autism.html). This misinformation was deposited despite decades of epidemiologic, basic science and genetic research that definitively demonstrates the lack of causality of vaccination in autism spectrum conditions (side note: an actual coherent discussion of this strong evidence is in Peter Hotez’s excellent and deeply personal book “Vaccines Did Not Cause Rachel’s Autism”). Ironically, some of the best evidence that definitively disassociates autism from vaccination comes from Denmark - a country which Make America Healthy Advocates seek to emulate in terms of vaccine policy.

All of this groundwork led the CDC on January 5th to cut the number of vaccines it recommends for every child, undermining protections against several diseases. No longer recommended is protection against flu, rotavirus, hepatitis A, hepatitis B, forms of meningitis and RSV. RFK Jr. insists that the overhaul won’t result in families who want access to the vaccines losing insurance coverage - a demonstrably false claim as CDC/ACIP recommendations are inexorably tied to insurance reimbursement.

In this Newsletter edition we will begin to chip away at the accumulated misinformation set forth by the CDC. Make no mistake, the MAHA movement and revamped CDC is not working in the best interest of the US population. As we cover below, we are already seeing large scale measles outbreaks - this will only worsen. We will also cover the Hepatitis B vaccine and why the CDC/ACIP’s update guidance will inevitably lead to increases in otherwise preventable cirrhosis and liver cancer. All by restricting access to a demonstrably safe vaccine. As an adult medicine physician, I will highlight the downstream effects on adults from two perspectives. One, what does this mean for those who are currently over 18 and, two, what ripple effects will these childhood vaccination oversights have for them as they reach adulthood.

1. Hepatitis B Vaccine

The first vaccine to be dropped from the childhood immunization schedule was Hepatitis B (abbreviated HBV: Hepatitis B Virus below), a move that occurred in December 2025. Some of those reading this Newsletter (unless you are in healthcare, have specific risk factors or have traveled to a high risk region) will not have gotten a Hepatitis B vaccine as it was introduced to the childhood schedule in 1991. I did not have a Hepatitis B vaccine until I entered medical school in 1997 because of the risk of HBV transmission through accidental contamination of needles and other injected products. HBV is also transmitted sexually. Because HBV’s transmission pattern includes these pathways, it was an attractive target to the ACIP (i.e. “babies don’t share needles, why do they need a Hepatitis B vaccine?”).

Another issue that the ACIP likely considered is that the prevalence (rate) of Hepatitis B infection varies greatly across the world. In fact, chronic Hepatitis B infection is not something we commonly see in the United States, this as a direct consequence of ‘layered’ prevention which (until now) includes prenatal screening and infant vaccination. However, in other regions (China, Southeast Asia, sub-Saharan Africa, parts of the Middle East and South America), the prevalence of Hepatitis B antigen (we are all now familiar with antigens, or viral proteins, from the COVID rapid antigen testing) can be 5-15% of the population.

In these regions, the primary route of acquisition of Hepatitis B amongst infants is from their mother during birth (it is not transmitted across the placenta). Infants born to an HBV chronic carrier mother who expresses the envelope antigen (HBeAg positive) have a 70% (!) chance of becoming infected. For those who do not express the envelope antigen (a measurement of viral replication), that rate drops to 5-10%. Furthermore, the virus can be transmitted from infected household members or through casual contact in childcare settings as the virus is quite stable. This risk extends from birth and through the first 5 years of life but drops sharply thereafter. Risk of chronic HBV infection (i.e., failure to clear) is highest in infancy and falls with age such that:

• Infants: ~90% become chronically infected (~10% clear).

• Children 1–5 years: about 30% chronic (~70% clear).

• Adults: about <5% chronic (~95% clear).

Before universal Hepatitis B vaccination for infants, about 18,000 children in the United States contracted the virus annually. The 1991 introduction of universal Hepatitis B vaccination, resulted in a 90% drop in incidence rates of Hepatitis B amongst children and adolescents, nearly eliminating this chronic infection that can lead to cirrhosis, liver cancer and liver failure (Table below).

Returning to our pre-1991 policy will lead to about 16,000 children annually who become chronically infected (in 1991 there were 4.1 million US births and in 2024 about 3.6 million so this is why there are fewer total annual incident cases). Of those, 1 in 4 will eventually die from liver failure or liver cancer (hepatocellular carcinoma) and 1 in 3 will develop cirrhosis (these two are not mutually exclusive). Running this out over 20 years of misguided infant vaccine policy, we are left with 324,000 needlessly but chronically infected children translating into 81,000 deaths from an entirely preventable cancer.

But wait, you ask, how does the Hepatitis B vaccine work if the baby is already exposed to the virus during childbirth? Well, that’s an excellent question and, again, reveals how little the ACIP understands about basic viral mechanics. After exposure (either during childbirth or via a needlestick), HBV doesn’t instantly bind to the liver. Like many other viruses, there is a ‘window period’ during which HBV only begins to seed liver cells. By inducing enough antibodies, it is possible to prevent the establishment of infection (as a caveat for completeness the standard of care can also include passive neutralizing antibodies in high-risk exposures during childbirth).

So why universal vaccination for infants (if the above were not convincing enough) when we can just test mothers and vaccinate infants accordingly? This is what FDA Chief Dr. Marty Makary thinks we should be doing, namely bringing US vaccine policy more in line with countries like Denmark (source: https://www.youtube.com/watch?v=ltl5H_pCp1I). Unfortunately, Dr. Makary is not recognizing a key difference between the US healthcare and that of Denmark which is the abysmal rate of maternal screening in the US. In Denmark, there is a centralized national pregnancy screening program. 99.9% of mothers were tested for Hepatitis B, HIV and syphilis. By contrast, only 82% of women in the US are screened for Hepatitis B, 66.1% for HIV and 79.6% for syphilis. Danish population health datasets also provide some of the strongest longitudinal data refuting claims of an association between vaccines and autism, but this is a topic for another newsletter.

2. Americas Loses Its Measles Free Status

Over the past months, we’ve seen more and more headlines covering the (expected) resurgence of measles - which is the most infectious of infectious diseases (we cover the infectious characteristics of measles below). Amongst these are:

• “South Carolina Measles Outbreak Now the Largest in Recent US History” (Infectious Disease Advisor, Feb 2026) - note on February 3rd, 2026 the case count rose from 29 to 876

• “US Measles Outbreak Tops 1,800 Cases as Respiratory Illness Surveillance Returns” (CIDRAP, Dec 2025).

• “New Measles Spread Across States Threatens US Elimination Status” (Medical Professionals Reference, Jan 2026).

One in particular illustrates just how problematic a case of measles can be from the NBC 10 Philadelphia article: “Passengers at Philadelphia International Airport may have been exposed to measles.”

Philadelphia’s Department of Public Health issued a notice that a person with measles transited Terminals A and B between 8:50 a.m. and 4:00 p.m. on Sunday, November 9 2025, and anyone in those areas during that window may have been exposed. Officials urged potentially exposed travelers to verify that they are immune and to watch for symptoms.

The advice set forth by public health officials from this Philadelphia exposure is standard and worth restating here:

• You are generally considered protected if you were born before 1957, have had documented measles in the past, or have received two doses of an MMR-containing vaccine.
  

• If you are not protected, you should get an MMR dose promptly and talk with your clinician—especially if you are pregnant, under 12 months of age, or immunocompromised.
  

• Someone who has measles is contagious from 4 days before the rash appears to 4 days after. During that period they can infect people who simply share airspace with them
  

For those who like virology: measles is not just “very contagious” but it is the most contagious infection we deal with in routine clinical practice. The basic reproduction number (R₀) is often estimated at 12–18, meaning a single person with measles can infect 12–18 others in a completely susceptible population. That’s an order of magnitude more than the original strain of COVID-19.

Crucially:

• Humans are the only host. There is no animal or environmental reservoir; measles is maintained only by human-to-human transmission
  

• Because it is so infectious, you need about 94–95% of the population immune to prevent sustained transmission

Once measles finds a pocket where vaccine coverage has dipped below that 95% threshold in any community with 100,000 people, it will circulate freely

But the resurgence of measles isn’t just about one infectious traveler at an airport. In early November, the Pan American Health Organization (PAHO) formally announced that the Region of the Americas has lost its verification as free from endemic measles transmission. This is a sad reversal as measles had been declared eliminated from the Americas in 2016—but sustained outbreaks in Canada, the United States, and Mexico have reversed that progress.

The drivers are depressingly predictable:

• Childhood vaccination coverage has fallen below the 95% threshold needed to prevent measles spread,

• Widespread misinformation and vaccine skepticism, amplified after COVID, have eroded trust in public health and in routine immunizations.

• At the same time, U.S. funding cuts to measles campaigns, justified by unsubstantiated vaccine safety claims from current U.S. Health Secretary Robert F. Kennedy Jr., undermine immunization efforts,

That combination is exactly how you turn an “eliminated” disease back into an endemic (freely circulating one). In the United States, 2025 had the highest number of measles cases since the disease was declared eliminated in 2000. There were 2,267 confirmed measles cases across 40 states and jurisdictions. A few key points are:

• Most cases are in children and adolescents. Children under 5 years account for roughly 27% of cases, and those 5–19 years for about 40%, so roughly two-thirds of all U.S. cases are in people under 20.

• Adults are not spared. About one-third (33%) of cases are in adults aged 20 and older—many of whom assumed they were immune but either never completed the two-dose series or have uncertain records

• Vaccination status is the story. Ninety-two percent of U.S. cases are in people who are either unvaccinated or have unknown vaccination history; only about 4% had received a single MMR dose and 4% had received two doses.

• This is not a benign “natural” illness. About 12% of patients have required hospitalization, with the highest hospitalization rate—around 22%—in children under 5. Measles complications include pneumonia, encephalitis, and long-term immune dysfunction.

• CDC reported three measles-associated deaths in 2025, For a disease we already knew how to prevent in the 1960s, this is not acceptable public health policy.

At Santa Monica Primary Care, we’ve been watching this unfold with a mix of déjà vu and frustration. Two concrete things we are doing:

1. Checking immunity, not guessing: As part of annual wellness labs, we have now added a check of MMR titers in adults—particularly those born between 1957 and the late 1980s, when documentation can be spotty. If titers are low or equivocal, we offer MMR vaccination e. This is our way of quietly countering RFK-style vaccine skepticism with something radical: data.
   

2. Treating suspected measles like a dangerous and incredibly infectious airborne pathogen: For any suspected measles case in the office, we take airborne precautions seriously:
   

   • The patient is brought in through a separate entrance, not the main waiting room.

   • Staff caring for the patient wear fit-tested N95 respirators.

   • After the visit, we close that exam room for at least four hours. Measles can remain airborne and infectious in an enclosed space for up to two hours; we use four as a practical safety buffer when scheduling.
     

This is more disruptive to our schedule than most infections—but that is the point. Measles behaves differently from garden-variety respiratory viruses; our infection-control approach has to match its biology.

What this means for you:

If you take nothing else away from this section, make it this:

• If you don’t know whether you’re immune to measles, it is worth checking—either by verifying your records or by drawing titers.

• If you are not immune, MMR vaccination is an extremely safe, highly effective way to close that gap.

• If you develop fever, cough, red eyes, and a rash after a known exposure (like the Philadelphia airport window above), please call before coming to the office so we can bring you in safely.
  

Measles came back not because the virus got smarter, but because our vaccination policy—shaped in part by deliberate misinformation from people now in charge of federal health funding—got dumber. The good news is that individual immunity is still a switch you can flip on to keep yourself and your family protected.

2. ACIP Recommendations in the Midst of Rising Measles Incidence

While measles cases are climbing and the Americas have lost their “measles-free” status, the Advisory Committee on Immunization Practices (ACIP) has chosen this moment to…make it harder for kids to get fully immunized.

At its September 18, 2025 meeting, ACIP voted to no longer recommend the combined measles–mumps–rubella–varicella (MMRV) vaccine as the first dose for children under 4. Instead, they now recommend using two separate injections—MMR plus varicella (MMR+V)—for the 12–15 month visit, reserving MMRV as an option only for the second dose at 4–6 years. The justification: febrile seizures.

Here are the numbers ACIP reviewed:

• With MMR+V split vaccine scenario (i.e. two separate shots at the same visit), the risk of a febrile seizure is about 1 in 3,000–4,000 doses.

• With MMRV as a single vaccine the first dose, the risk is roughly double that (relative risk).

• No increased seizure risk has been seen with the second dose for either schedule.
  

Based on this, the committee voted 8–3 (1 abstention) against recommending MMRV for the first dose. This vote demonstrates the committee’s lack of understanding of Relative Risk versus Absolute Risk. To be clear: a seizure is scary for any parent. But the absolute risk being debated is on the order of one extra febrile seizure per several thousand doses (about 2,500) of MMRV, almost all of which resolve without neurologic sequelae.

A few members pushed back. Cody Meissner pointed out that febrile seizures are common, usually benign, and overwhelmingly not vaccine-related, with excellent prognosis and no evidence of long-term harm in the vast majority of cases. Restricting an effective combination vaccine, he argued, risked lowering coverage and therefore herd immunity—something that has long-term consequences. Joseph Hibbeln made a similar point: if you are going to change a long-standing recommendation in a way that adds barriers and complexity, the evidence had better be overwhelming. Here, it isn’t.

. Meanwhile, the disease we’re trying to prevent:

• Hospitalizes about 12% of all U.S. measles cases, and hospitalizes 22% of children under 5.

• Causes pneumonia, encephalitis, and long-term immune suppression.

• Killed 3 Americans in 2025.
  

And this is the moment to make the most efficient, kid-friendly formulation less available for first doses?

ACIP did at least vote (8–1, with 3 abstentions) to keep access to MMRV through the Vaccines for Children (VFC) program, so combination vaccine supply remains. But the new guidance sends a mixed message to parents at exactly the wrong time.

It is not surprising that major medical societies have decided they’ve had enough. In an updated complaint in American Academy of Pediatrics v. Robert F. Kennedy Jr., the AAP, ACP, and others are now asking the court to force Secretary Kennedy to reconstitute ACIP with entirely new members, chosen for actual expertise and a balanced range of perspectives rather than ideological loyalty. The lawsuit argues that the current committee was not properly constituted, is “inappropriately influenced” by the appointing official, and is making changes that increase confusion, fuel hesitancy, and erect barriers to care—exactly what we are seeing with the MMRV decision.

3. Influenza A Activity

Our September 2025 newsletter highlighted seasonal vaccines for Fall 2025, namely Influenza, COVID and RSV. While we have not seen an unexpected number of cases of COVID in our clinic (more on this below), Influenza A cases have surged. We have seen cases even amongst those who received the 2025/2026 seasonal Influenza vaccine. Historically, in Los Angeles, we have seen the highest rates of influenza during the week between Christmas and New Years

The CDC through its FluSurv-NET surveillance program, reported a peak hospitalization rate for influenza the week of December 21st - December 27th 2025 at 12.8 hospitalizations per 100,000 population. This is the second highest weekly peak since 2010-2011 and the pediatric hospitalization peak was the highest observed since 2010-2011.

Rates have dropped off a bit since late December but, in the last week, there has been a slight resurgence. Laboratories are reporting that 18% of respiratory swab samples submitted are positive for influenza and nearly 5% of all visits to outpatient healthcare facilities were for respiratory illnesses. Both are above the typical baseline and have risen in the past week. This season alone influenza has accounted for more that 20 million illnesses, 270,000 hospitalizations and 11,000 deaths. A total of 130 million doses of influenza vaccine have been given in the United States (38% of the population). This lags the typical percentage of Americans that receive a flu vaccination annually, which tends to be about 45-50%.

There is still time to get vaccinated against flu as the traditional end of flu season is April 1st. Flu vaccination has been shown definitively to reduce the risk of acquiring the infection and, if a breakthrough infection occurs, significantly reduce the severity of illness. We stock influenza vaccines in the office.

Influenza Vaccine Efficacy: As above, seasonal influenza vaccination provides important protection for flu itself and associated potential complications. The longstanding recommendations remain the same this year such that all individuals 6 months and older (with very few exceptions) should be getting the seasonal flu shot. One potentially concerning aspect of this season’s surveillance and reporting is the possibility of a mismatch between the 2025/2026 flu vaccine and the predominant circulating influenza viruses.

While it is too soon to accurately estimate this season’s influenza vaccine efficacy, we have noted in the United States a predominance of Influenza A (H3N2). This differs from what we saw earlier in 2025 from surveillance networks in eight Southern Hemisphere countries, where the predominant subtype was Influenza A (H1N1). It is important to note that this season’s flu vaccine is directed at both H1N1 and H3N2 subtypes of Influenza A (as well as an Influenza B subtype).

Initially estimates of vaccine efficacy from the Southern Hemisphere measuring protection against H1N1 illness (which was the primary circulating virus) for outpatient visits was 50.4% and for hospitalization was 49.7%. This represents a typical-mid range efficacy (40-60% is common when the vaccine/strain match is moderate) and is slightly better than the 2023/2024 season when vaccine efficacy was 39% against hospitalization. A 50% reduction in medically-attended influenza and hospitalization remains a significant public health benefit.

4. COVID-19 Activity, Santa Monica Primary Care

As we have discussed in our prior newsletters, COVID infections display a seasonal distribution with two distinct peaks - the largest being in the Winter months (November/December/January) and a second in mid to late summer (July/August). This pattern is why we have a recommendation in the Fall to have the COVID booster in September or October and in the Spring to time the vaccine in April or May for those most susceptible to severe illness with COVID infection.

The Los Angeles County Health Department stopped reporting daily incidence and mortality rates for COVID-19 in October 2023 and ended weekly press-release updates in February 2024. As such, we have relied on our own data from Santa Monica Primary Care to track COVID seasonality.

But looking at the graph for 2025/2026 below, our experience in the clinic over the past 12 months tracks with the expected dual seasonality. While we had a relatively steady state of cases from February through July (between 2 and 4 cases each month), there was a significant increase in cases in August and September (8 and 6 cases respectively) before declining to only 1 case in October.

As with most respiratory viruses, an increase during the late Fall to Winter months is expected. Again, this is seen as case numbers increased from 1 in October and November 2025 to 2 and 3 in December 2025 and January 2026 respectively. Similar to 2024/2025 it looks as if in 2025/2026 the more significant seasonal peak occurs in the Summer, not the Winter.

5. Follow Us On Social Media

Twitter (X): @santa_care

Facebook: https://www.facebook.com/santamonicaprimarycare

Instagram: @santa_monica_primary_care

LinkedIn: https://www.linkedin.com/in/philip-bretsky-mdphd/

BlueSky: @bretskymd.bsky.social

Substack: @bretskymd

Follow our blog posts on: www.drbretsky.com/blog

Archived issues of our newsletter can be found at: https://drbretsky.com/newsletters/

If you know somebody who would like to sign up to receive our newsletter, they can do so at: https://lp.constantcontactpages.com/su/JtF9zzR

Further, advanced paternal age shows a dose response such that every 10 year increase in paternal age increases ASD risk by about 20%. There is no such dose-response seen with Tylenol. In fact, one could just as easily argue that mothers partnered with older men take Tylenol more frequently than those partners with younger men (this has never been studied so, for now, cannot be disproven).

I add paternal age as a biologically plausible culprit of ASD to reflect the off-base simplicity of the White House’s admonition for pregnant women to not take Tylenol. If the White House were serious, the moderately strong epidemiologic evidence of advanced paternal age and stronger effect on ASD should have led them to warn against having children with men over the age of 50. But while paternal age factors into risk discussion, it exerts a relatively low absolute risk on ASD given its aforementioned multifactorial nature. Using large Swedish meta-analysis datasets (sources: Reichenberg et al., JAMA Psychiatry 2014; Hultman et al, Molecular Psychiatry 2011), you get the following

Just as we would advise prospective parents that paternal age is one of many modest contributors to ASD, Tylenol should be viewed as associated with but not deterministic of ASD. Tylenol remains the safest treatment strategy for pregnant women when experiencing pain and fever. It is important to note that an uncontrolled fever in early pregnancy doubles (relative risk of 1.9 fold) the risk of neural tube defects (link:https://pubmed.ncbi.nlm.nih.gov/15703536/).

Below we review the single study that the White House uses as the keystone for its argument against Tylenol use during pregnancy. In addition to directly contradicting 6 other systematic review studies the article directly references, it uses a curious study design called “Navigation Guide” methodology which has been applied to environmental health research where gold standard randomized clinical trials are unethical.

Tylenol was first developed in 1955 and became widely available for adults in the 1960s, quickly establishing itself as the leading non-aspirin pain reliever. Autism spectrum disorders were first described in the early 1940s, with cases from Johns Hopkins (Kanner, 1943) and Vienna (Asperger, 1944). Asperger’s work described children with typical cognitive profiles and age-appropriate language skills but social difficulties and restricted interests. So The White House has identified a ‘cause’ that didn’t even exist when the condition was first described.

Before concluding, we will do a quick review of seasonal vaccines as well as recent data on their demonstrated efficacy to date. There is still time to get recommended Fall 2025 vaccines.

1. The White House Statement

“The Trump Administration does not believe popping more pills is always the answer for better health. There is mounting evidence finding a connection between acetaminophen use during pregnancy and autism — and that’s why the Administration is courageously issuing this new health guidance. Additionally, the Trump Administration is approving a new treatment option that has been found to improve some autism symptoms [wait - I thought you just said popping pills wasn’t the answer - sorry to interrupt]. President Trump pledged to address America’s skyrocketing rates of autism, and his team is deploying Gold Standard Science [not a thing] to deliver on this pledge. We will not be deterred in these efforts as we know millions across America are grateful.”

— Press Secretary Karoline Leavitt

So where to even start with this statement? While one is tempted to begin with the first sentence - that popping more pills is not the answer to better health - a more constructive place to start is with the “mounting evidence” of a connection between Tylenol and pregnancy.

The bulk of the White House’s thesis comes from a single article by Prada et al. published this August in Environmental Health entitled “Evaluation of the evidence on acetaminophen use and neurodevelopmental disorders using the Navigation Guide methodology.” The first question I had was “what is Navigation Guide methodology?” - having never heard of it before in the context of health research. There is a reason, I hadn’t heard of it, that being it is typically used in environmental health and toxin exposure investigations.

Navigation Guide methodology uses an applied framework in the case where randomized trials are unethical or impossible. This would include exposures to pesticides, carcinogens or other compounds which are known to cause harm and, as such, knowingly exposing a research group to such would be unethical. For instance, the risk to children from pesticides, flame retardants or other household chemicals was examined using Navigation Guide methodology in the CHEERS study of 2004-2005 - exposures to which in a case/control fashion would clearly be unethical.

The authors’ rationale for apply this fringe statistical methodology to Tylenol and autism relies on the general reluctance to include pregnant women in placebo controlled medication trials. This is primarily due to the fact that any potential harm to the fetus (teratogenicity, miscarriage, neurodevelopmental effects) would be irreversible. Additionally, pregnancy alters drug pharmacokinetics (absorption, metabolism, distribution), making dose-response data difficult to interpret and generalize. However, Acetaminophen (Tylenol) has a long safety record making the application of an analytic framework better suited for environmental exposures a curious choice indeed.

The Prada article itself cites 6 previous systematic reviews and meta-analyses which considered the association between prenatal acetaminophen exposure and neurodevelopmental disorders (which the authors consider to be Autism Spectrum Disorder and Attention Deficit / Hyperactivity Disorder). Here are the conclusions of those 6 studies with more traditional and robust analytic designs.

1. Banerjee et al (2007, Acta Paediatrica): “Acetaminophen not yet identified as a candidate exposure.” This article identified maternal smoking, alcohol and toxins as ADHD risks.

2. Khan FY et al. (2022, Cureus): Association possible, but “a mystery to resolve” with evidence that is “insufficient and inconsistent.”

3. Kim JH et al. (2020, Lancet Psychiatry): Acetaminophen exposure across all aggregated studies showed weak evidence of association; higher-quality studies often attenuated the link.

4. Marsarwa et al (2020, Paediatr Perinat Epidemiol): Observed relationships “likely reflect bias and residual confounding rather than true causation.”

5. Patel R et al. (2022, Int J Environ Res Public Health): Available evidence “does not establish a causal link.”

6. Ricci C et al. (2023, Paediatr Perinat Epidemiol): Small association observed but “cannot confirm causality”

Clearly none of these support the Prada conclusion and most certainly do not represent “mounting evidence” of a connection between acetaminophen use during pregnancy and autism.

2. So What’s A Better Way to Approach The “Cause” of Autism?

It has been clear to even a causal observer, the RFK Jr. and HHS were focused on finding a single ‘cause’ of autism. In April of this year, RFK Jr. stated the following:

“We are going to look at vaccines, but we are going to look at everything. Everything is on the table,” Kennedy said. “Our food system, our water, our air, different ways of parenting, all the kinds of changes that may have triggered this epidemic.”

He suggested that by September 2025 they would find the cause. Tylenol was not on my bingo card as the scapegoat (vaccines were). And here on 14 August 2025, The White House has provided the “answer” via the Prada et al. article - except it didn’t. Autism and ASD (Autism Spectrum Disorder) are far more complex and include both genetic and environmental factors. Any suggestion of a single causative factor is simply false and a vast oversimplification of a very complex chain of causality - whether it be Tylenol or vaccines. A coherent discussion of autism and neurodevelopmental disorders at a minimum needs to consider the genetic underpinnings of such. These are robust and have been well described.

When considering genes that have shown a correlation with autism spectrum disorder risk, one can quite easily appreciate their biologic plausibility. These include:

1. SHANK3 - builds the scaffolding that connects neurons, creating a communication framework between cells

2. NRXN1 (Neurexin 1) - a gene encoding for a cell-adhesion molecule that allows neurons to line up for efficient signal transmission

3. CHD8 (one of the most reliably and reproducibly seen ASD-linked genes) which acts as a genome organizer, turning genes on or off during fetal brain development. It helps control how neurons mature

4. MECP2 - a regulatory gene that limits expression of other genes. In cases where it does not work, neurons become overactive and are uncoordinated.

5. TBR1 - a transcription factor guiding brain cell differentiation and maturation

Designing a study that integrates genetic risk (which is reviewed more fully below) as well as the potential influence of environmental exposures is incredibly complex and time consuming. Further, when the environmental exposures which may modify risk occur during critical periods in fetal brain development, this makes the study design all the more complex.

For those interested in an even more full accounting of Autism Spectrum Disorder causes (and personal experience with such) I would recommend Dr. Peter Hotez’s outstanding and highly personal book “Vaccines Did Not Cause Rachel’s Autism.”

3. Genetic Factors in ASD: Up to 80% of Risk

You read that correctly - Autism and Autism Spectrum Disorders (ASD) are amongst the most heritable of neurodevelopmental conditions, with anywhere between 50% to 80% of their risk attributable to genetic factors. It is massive (although deliberate) oversight insofar as the White House’s “discovery” makes no mention of genetic factors in ASD. As a comparison, only 5-10% of breast cancers are attributable to inherited (genetic) factors but any discussion of breast cancer risk factors would be incomplete without including genetic susceptibility inherent in BRCA-1 and BRCA-2 carrier states. Further, twin- and family based studies further indicate that genetic factors play critical roles in ASD, such that the concordance rate among monozygotic (identical) twins is higher (60–90%) than that among dizygotic (fraternal) twins (0–30%).

But similar to breast cancer, ASD is associated with some high risk but low frequency genetic mutations (this is similar to BRCA-1 and BRCA-2 which are rather uncommon in the general population but confer large effect sizes on the risk of breast and ovarian cancer). Familial clustering and genetic overlaps with ADHD, schizophrenia, and language delay suggest additive risk across hundreds of different genes. This is truly complex and does not lend itself to the simple one size fits all answer that the White House has put forth.

4. Using Advanced Paternal Age and ASD Risk, Illustrating Why Tylenol is Safe in Pregnancy

Interestingly, there is a rather strong association between advanced paternal age and Autism Spectrum Disorders, likely due to age-related germline mutations that are then passed along. In fact advanced paternal age shows a higher relative risk (1.55 - oldest versus reference age group) as compared to Tylenol (1.19 at its highest pooled estimate). In a Swedish cohort, fathers aged 50 years and older had a 2.2 fold increased risk of fathering a child with autism compared with fathers under the age of 30 (link: https://pubmed.ncbi.nlm.nih.gov/21116277/).

Further, advanced paternal age shows a dose response such that every 10 year increase in paternal age increases ASD risk by about 20% (see Table below). There is no such dose-response seen with Tylenol. In fact, one could just as easily argue that mothers partnered with older men take Tylenol more frequently than those partners with younger men (this has never been studied so, for now, cannot be disproven).

I add paternal age as a biologically plausible culprit of ASD to reflect the off-base simplicity of the White House’s admonition for pregnant women to not take Tylenol. If the White House were serious, the moderately strong epidemiologic evidence of advanced paternal age and stronger effect on ASD should have led them to warn prospective mothers against having children with men over the age of 50. But while paternal age factors into risk discussion, it exerts a relatively low absolute risk on ASD given its aforementioned multifactorial nature. Using large Swedish meta-analysis datasets (sources: Reichenberg et al., JAMA Psychiatry 2014; Hultman et al, Molecular Psychiatry 2011), you get the following:

Summary Table: Relative Risk (RR) of Escalating Paternal Age and ASD

Just as we would advise prospective parents that paternal age is one of many modest contributors to ASD, Tylenol should be viewed as associated with but not deterministic of ASD. Tylenol remains the safest treatment strategy for pregnant women when experiencing pain and fever. It is important to note that an uncontrolled fever in early pregnancy doubles (relative risk of 1.9 fold) the risk of neural tube defects (link:https://pubmed.ncbi.nlm.nih.gov/15703536/).

As is the case with vaccine guidance, we once again need to look to other global agencies for substantive health advice since our own institutions (HHS, CDC and NIH) have gone horribly awry. So in response to the White House’s warning about Tylenol and pregnancy, the Australian Therapeutic Goods Administration reaffirmed that Acetaminophen remains “Pregnancy Category A” deeming it safe when used as directed with “robust scientific evidence showing no causal link” with autism/ADHD. Similarly, the European Medicines Agency confirmed that acetaminophen remains an important option for pregnant women with no evidence that taking such “during pregnancy causes autism in children.” And, lastly, the WHO’s statement emphasized that “there is currently no conclusive scientific evidence” linking acetaminophen use in pregnancy with autism.

5. Returning to Our More Usual Subjects, Vaccine Efficacy

Our September 2025 newsletter highlighted seasonal vaccines for Fall 2025, namely Influenza, COVID and RSV. As vaccinations began in March for the Southern Hemisphere, we do now have some preliminary data on their efficacy.

Influenza: As we discussed last month, seasonal influenza vaccination provides important protection for flu itself and associated potential complications. The longstanding recommendations remain the same this year such that all individuals 6 months and older (with very few exceptions) should be getting the seasonal flu shot. When considering surveillance networks in eight Southern Hemisphere countries thus far in 2025, the predominant subtype seen was Influenza A H1N1. Vaccine efficacy measuring protection against illness for outpatient visits was 50.4% and for hospitalization was 49.7%. This represents a typical-mid range efficacy (40-60% is common when the vaccine/strain match is moderate) and is slightly better than the 2023/2024 season when vaccine efficacy was 39% against hospitalization. Nevertheless, a 50% reduction in medically-attended influenza and hospitalization remains a significant public health benefit. For those patients still seeking a seasonal flu shot, we do have them available in the office.

COVID: Although there are no data available as yet for the 2025/2026 seasonal COVID vaccination, a recently published (October 8th, 2025 in New England Journal of Medicine) Veterans Health Administration study does give us some information on last year’s COVID vaccine efficacy. When comparing those who received the 2024/2025 COVID-19 and influenza shots on the same day (164,132 participants) to those who received influenza only (131.839 participants) the COVID-19 vaccine demonstrated 29.3% efficacy against Emergency Department visits, 39.2% against hospitalization and 64.0% against COVID-19 associated mortality. Participants were followed for 6 months after vaccination. The average age of those receiving the COVID-19 vaccine was 71.5 years and, given that it was a VA study, overwhelmingly male (92%).

RSV: While it has long been known that RSV can cause significant illness (and death) in a pediatric population, only recently with improved RSV testing, we now understand that RSV is a major cause of hospitalization in older adults as well. As such, the RSV vaccine has been routinely recommended for adults 75 years and older. Recent studies have shown robust protection with a vaccine efficacy of 77% in reducing RSV-associated lower respiratory tract illness (source: https://pubmed.ncbi.nlm.nih.gov/41016728/). Further, there seems to be little difference in terms of serious adverse events when comparing the RSV vaccine to placebo. Current data suggest at least 2 years of protection which reflects both vaccine durability and the relatively slow antigenic drift of RSV (i.e. the same RSV strain circulates for multiple seasons). As such, there is not yet any recommendation for an additional booster amongst older adults.

6. COVID-19 Activity, Santa Monica Primary Care

As we have also discussed previously, COVID infections display a seasonal distribution with two distinct peaks - the largest being in the Winter months (November/December/January) and a second in mid to late summer (July/August). This pattern is why we have a recommendation in the Fall to have the COVID booster in September or October and in the Spring to time the vaccine in April or May.

The Los Angeles County Health Department stopped reporting daily incidence and mortality rates for COVID-19 in October 2023 and ended weekly press-release updates in February 2024. As such, we have relied on our own data from Santa Monica Primary Care to track COVID seasonality.

But looking at the graph for 2025 below, our experience in clinic this calendar year generally tracks with the expected dual seasonality. While we had a relatively steady state of cases from February through July (between 2 and 4 cases each month), there was a significant increase in cases in August and September (8 and 6 cases respectively) before declining to only 1 case in October.

Summary Graph: COVID-19 Cases Santa Monica Primary Care, January - October 2025

7. Follow Us On Social Media

Twitter (X): @santa_care

Facebook: https://www.facebook.com/santamonicaprimarycare

Instagram: @santa_monica_primary_care

LinkedIn: https://www.linkedin.com/in/philip-bretsky-mdphd/

Blue Sky: @bretskymd.bsky.social

Substack: @bretskymd

Follow our blog posts on: www.drbretsky.com/blog

Archived issues of our newsletter can be found at: https://drbretsky.com/newsletters/

If you know somebody who would like to sign up to receive our newsletter, they can do so at our newly redesigned sign-up page: https://lp.constantcontactpages.com/su/JtF9zzR