For a drug that came out of forty years of unfinished neuroscience, the news from ENDO 2026 should have been straightforward. On Sunday in Chicago, the University of Illinois Chicago’s Pauline Maki stood in front of the Endocrine Society’s annual meeting and presented the first real-world numbers for Veozah, Astellas’s non-hormonal menopause pill. The 656 women in the analysis, she said, reported significant improvements in hot flashes, depression, and anxiety, with mood gains turning up as early as four weeks. The benefits looked, in her phrase, “similar to what was seen in clinical trials.” The conference write-up carried the line and the upbeat framing, and stopped there. What it did not mention, and what Astellas’s own safety boilerplate carefully avoided naming, was the boxed warning the FDA stamped on the same drug’s label in December 2024, the most serious alert the agency issues, after a patient on the drug developed serious liver injury within forty days of starting it.

That gap, between Veozah on the slides and Veozah on the label, is what ENDO 2026 was actually reporting on. The upbeat write-ups just left the second half out.

Some of why Astellas wants the upbeat read to stick has to do with how hard it was to get a non-hormonal menopause drug this far at all. For most of the last twenty years, the menopause-medicine conversation looped through the same two options. Take estrogen, and accept the cancer and clot anxieties left behind by the Women’s Health Initiative scare at the turn of the millennium. Or take an SSRI or SNRI repurposed off its depression label, and accept that the data for hot flashes was modest and that a meaningful share of women never got much relief. The biology of the hot flash itself, the specific reason a fifty-three-year-old wakes at 3 a.m. drenched and furious, stayed largely undisturbed in the background.

That changed when endocrinologists zeroed in on a small cluster of neurons in the hypothalamus called KNDy neurons. As estrogen falls, those neurons go hyperactive and signal the brain’s thermostat to misfire. Block the right receptor on that circuit, the neurokinin-3 receptor, and the misfiring quiets. Fezolinetant, marketed as Veozah, was the first molecule to translate that mechanism into a pill, and the FDA approved it in May 2023.

OPTION-VMS, the study Maki presented, is not the trial that backed up that approval. According to Astellas’s October 2025 description of the Phase IV protocol, it is observational. Women in this preliminary cut came in already on a non-hormonal therapy their own clinicians had prescribed: 201 on fezolinetant, 329 on an SSRI or SNRI, 126 on other options like gabapentin and oxybutynin. No placebo arm. No blinding. Patients knew which pill they were taking, and so did their doctors. That design is useful for tracking what happens once a drug reaches actual clinics, but it cannot separate the effect of the molecule from the effect of being in a study and telling a researcher every few weeks that you feel better.

It is also an Astellas-sponsored design. The company’s October release on the same study credits Astellas medical-affairs staff and academic collaborators including UNC’s Genevieve Neal-Perry. The press release puts symptom-improvement language up top, rolls outcomes into MENQOL questionnaire domain scores rather than absolute hot-flash counts, and routes the safety conversation into a “U.S. Important Safety Information” block that walks readers through monitoring requirements without ever using the FDA’s phrase for them. There is a difference between disclosing a fact and naming it; the release banks on the difference.

The placebo-controlled evidence behind the approval is worth holding next to the upbeat real-world summary. SKYLIGHT 1, the pivotal Phase 3 trial in The Lancet, reported a 61 percent reduction in moderate-to-severe hot-flash frequency on fezolinetant 45 milligrams at week 12, against a 35 percent reduction on placebo. The placebo-adjusted gap, about 26 percentage points, is meaningful and worth noting, and it is also not dramatic. Most women in those trials finished twelve weeks still having hot flashes, just fewer of them.

The boxed warning is where the picture sharpens. The FDA opened a safety review in September 2024 after a postmarketing case in which a patient on fezolinetant developed jaundice, dark urine, fatigue, and abnormal liver enzymes within roughly forty days of starting treatment, then recovered after stopping the drug. In December 2024, the agency upgraded the language to a boxed warning. The updated label now calls for liver-function bloodwork before the first pill, monthly tests through month three, and additional tests at month six and month nine. Six liver panels in the first nine months on a hot-flash drug. The Astellas ENDO summary lists “liver problems” under a single line of safety boilerplate and never uses the phrase the FDA chose.

The price keeps the contrast honest. Veozah launched at roughly $550 a month in 2023 and, per GoodRx in 2026, now runs from about $550 to roughly $765 a month at retail pharmacies, with no generic on the way. A year of the drug alone at the top of that range clears $9,000 cash. Layer the FDA-mandated lab monitoring on top, and the policy question hiding inside the ENDO slides comes into view: who pays $9,000 plus standing labs for a 26-percentage-point improvement on a class of symptoms that, for many women, will fade on their own inside a few years.

The field is about to stop being a one-drug field, which is the other reason this ENDO presentation arrived when it did. Bayer’s elinzanetant, a dual NK1/NK3 receptor antagonist marketed as Lynkuet, picked up FDA approval on October 24, 2025 as the second non-hormonal pill in this space and the first to hit both receptors at once. The data point worth watching is head-to-head trial work refereed by independent researchers rather than by company medical affairs: who benefits, at what cost, under what monitoring burden, with what liver risk. Until those numbers land, the right way to read Astellas’s ENDO slides is to keep the order honest: the neuroscience under fezolinetant is solid and a long time coming, the molecule does something for some women, and the distance between those two sentences and the picture on the conference slides is where the boxed warning, the price tag, and the questionnaire-scale framing all live.

Sources

  1. News-Medical, ENDO 2026 coverage of the OPTION-VMS preliminary real-world analysis (June 14, 2026)
  2. Astellas, OPTION-VMS Phase IV observational study release (October 22, 2025)
  3. FDA Drug Safety Communication on serious liver injury with Veozah (fezolinetant)
  4. The Lancet, SKYLIGHT 1 Phase 3 randomised controlled trial of fezolinetant (2023)
  5. Pharmacy Times, FDA approval of elinzanetant (Bayer’s Lynkuet), October 2025
  6. GoodRx, Veozah 2026 cash pricing