Leaked screenshots sent Moderna's shares down 17.6% before the companies released the data. The official figures show a 42% overall cut in recurrence risk, with no effect in the first six months and a 58% reduction afterwards.
On Thursday afternoon Moderna lost about $17 billion of market value on screenshots posted to social media. The companies did not publish their data until that evening.
The leak brought forward a repricing that Saturday's data would likely have caused. Investors reacted to the overall 42% figure. The sell off says more about a valuation that had nearly quadrupled than about failed efficacy. The release's own split at month six shows why 42% understates the effect once the regimen has been given.
How the data leaked
Results for the melanoma vaccine intismeran began to appear on X on Thursday afternoon. The conference organiser ESMO then wrote on X that the embargo on the trial abstract "has been breached". It lifted the embargo on the full dataset.
No one has said who leaked the data, and no wrongdoing is alleged against Moderna or Merck. The data were due at ESMO's first Presidential Symposium on Saturday 24 October.
Moderna fell as much as 26% in the session. According to Yahoo Finance price history, it closed at $202.63 against $245.81 on Wednesday, a drop of 17.6%. On roughly 399 million shares, that is $43 a share and about $17 billion.
What the trial found
The full results arrived that evening, when Merck and Moderna published them. The trial enrolled 1,137 patients with resected melanoma at stages IIB to IV, randomised two to one.
Adding intismeran to Keytruda cut the risk of recurrence or death by 42% over the whole trial, a hazard ratio of 0.58. Distant metastasis or death fell by 38%. Median follow up was 24.6 months.
That 42% sat at the low end of what investors wanted. BioPharma Dive reported that Wall Street expected a 40% to 50% cut, and that Moderna had called 35% clinically significant. An analyst at RBC put the result at the low end of expectations. One at William Blair called the fall a "sell-the-news reaction".
The price had run far ahead of that band. On 19 August the companies said only that the trial had met its endpoints, with no numbers, and the shares jumped 177% in a day. By Wednesday they had risen about 3.9 times from the old close.
Some investors had priced in more than melanoma, according to Citi analysts quoted by Clinical Trials Arena. They said some were "assigning value to an effective-cure narrative" across further solid tumours. A clinically meaningful result can still trigger a valuation cut if investors expected more.
Why the six-month split matters
The release also shows how the effect changed over time. For the first six months, while patients were still being immunised, results were "comparable with KEYTRUDA alone". The hazard ratio was 0.99. After month six it was 0.42, a 58% cut.
The companies call the later figure a pre specified sensitivity analysis, so it is not the primary result. A trial that starts flat and then separates will always report an average below its later effect. Keytruda alone cut recurrence 43% against placebo in earlier work, so the bar for an add on was high.
Patients who received at least one planned dose did better. Their cut was 49% on recurrence and 45% on distant metastasis. The 49% matches the five year Phase 2b figure from June.
The vaccine arm also kept more patients disease free at two years. In it, 82.2% of vaccine patients were free of recurrence against 70.0% on Keytruda alone, a gap of 12.2 points. Distant metastasis free survival was 88.5% against 80.2%, a gap of 8.3 points.
Safety looked unremarkable. Grade 3 to 5 treatment related events were 14.2% in the vaccine arm and 13.4% in the control arm. The companies reported no new safety signals and said they are engaging with regulatory authorities.
The Phase 2b trial covered stage III and IV patients. This one reaches stage IIB, where fewer patients relapse, so the same hazard ratio takes longer to show. The two percentages are not directly comparable, though the release says results were generally consistent across pre-specified patient subgroups.
The unanswered question is how long the benefit lasts. Overall survival data are not mature, though the release reports a favourable trend. Nobody has yet seen how long the 58% after month six holds.
The debt and share-price test
Moderna's convertible debt puts a price line just above Thursday's close. On 28 August the company sold $2.6 billion of zero coupon convertible notes due 2032. The conversion price is about $210.58, a 47.5% premium to the $142.77 share price the day before.
Thursday's close of $202.63 is about 4% below that conversion price. Holders therefore own a bond that is not yet in the money, though the shares sit close. Moderna said the proceeds could fund growth in oncology or repay debt.
Moderna also paid about $285 million for a capped call, a hedge against dilution. Its cap price is $392.62 a share, a 175% premium to the pre-pricing close.
The shares largely recovered from a further fall on Friday. With the full data out, Moderna opened at $196.69 and fell to $183.00, 9.7% below Thursday's close. By 11:36 am Eastern it was back at $201.95 intraday, 0.3% below that close, as buyers returned after the further decline.
Merck, the vaccine partner, fell less than Moderna. It dropped as much as 3.9% on Thursday and closed down 1.8%. At 11:36 am Eastern on Friday it was down 1.8% intraday at $144.64, against Moderna's 0.3% shortfall at that time.
The next data and results dates
On Saturday ESMO's Presidential Symposium will show the fuller curves, including whether the line after month six keeps separating. That decides whether the later 58% cut is durable, which the 42% average cannot show. The paper is accepted at the New England Journal of Medicine.
The price test is the $210.58 conversion price, about 4% above Thursday's close. Above it, holders can convert into stock worth more than the notes' face value. Below it, the notes behave like zero coupon debt.
The next hard date is Friday 6 November, when Moderna reports third quarter results. It is the next filing that shows what the company has spent.