Open in interactive viewer → charts, metric popovers & call review

ProMIS Is Asking the Market to Believe the Alzheimer's Failure Was Someone Else's

A $29 million biotech, $44 million of cash, and a Q1 2027 readout that management is already framing as a referendum on selectivity.
PMN · Earnings Call · 2026-11-12

ProMIS Is Asking the Market to Believe the Alzheimer's Failure Was Someone Else's

A quarter that was really a pre-bunking exercise

ProMIS Neurosciences did not report a quarter so much as a campaign speech. There was no revenue, no partnership, no new clinical dataset on the Q3 call. The financials were almost an afterthought: $44.1 million of cash, R&D down to $6.7 million from $9.8 million, G&A flat at $1.9 million, and a net loss of $8.2 million. The real event had happened two weeks earlier, when another company's oligomer-selective Alzheimer's antibody failed its Phase 2. Management's job on this call was to make sure investors did not tar PMN310 with that result. The argument is precise, and it is the whole investment case. The competitor's antibody, in ProMIS's reading, reduced amyloid plaque on PET and caused ARIA-E in roughly 10% of treated patients. That is not the behavior of a molecule that selectively neutralizes toxic oligomers; it is the behavior of a plaque-directed approved antibody with a familiar safety burden.

a molecule that clears plaque and causes aria is behaving much like the existing plaque directed drugs. So in our view, that trial did not test the oligomer hypothesis.

Neil K. Warma, Chief Executive Officer · 2026-11-12
Whether or not one buys the defensive logic, it is a clever framing: the failure becomes evidence that true selectivity has not yet been clinically tested—and PMN310 is positioned as that test.

Selectivity is the product

PMN310 is built on an IgG1 backbone to bind and clear oligomers while avoiding monomers, fibrils, and plaque. The company's 6-month interim from PRECISE AD—blinded, pooled, 136 safety-evaluable patients—showed no ARIA-E in any APOE genotype, ARIA-H at 4.4% (mild, asymptomatic, in line with placebo), no drug-related discontinuations, and no treatment-related serious adverse events. On the biomarker side, plasma p-tau217 fell about 15% and CSF MTBR-tau243 about 13% from baseline. Management is careful to call these blinded, pooled observations rather than a determination of efficacy, but the call's tone treats them as a down payment on the thesis. The scientific differentiator is the assay. ProMIS claims it can measure absolute concentrations of oligomers in CSF, not just antibody saturation. “our assay is quite a bit different in that it measures the absolute concentrations of A-beta oligomers themselves” — Larry Douglas Altstiel, Executive · 2026-11-12 If the Q1 2027 top line can show a direct drop in oligomers alongside clinical separation and a clean ARIA E profile, management believes it will have connected mechanism to outcome in a way no competitor has. That is the bull case in one sentence.

What a win means, and what it doesn't

The company has tried to pre-define success so the readout cannot be reframed later. On efficacy, management is looking for roughly 15-20% slowing of decline on either CDR Sum of Boxes or iADRS at 12 months, with directional trends in key biomarkers. “if we saw an efficacy rate kind of in the range of you know, roughly 15 to 20 percent, similar efficacy to what we are seeing with marketed drugs with a significantly better safety profile that we feel would be a significant win” — Neil K. Warma, Chief Executive Officer · 2026-11-12 The comparison is to the 22-27% range seen at 18 months in the larger competitor study, annualized down to 12 months—an interpolation management acknowledges is imperfect. The bigger claim is on safety. If ARIA-E comes in at or near placebo, ProMIS argues it can dramatically reduce the MRI monitoring burden that makes current anti-amyloid therapy so cumbersome. That is not just a clinical nicety; it is a commercial wedge. The company has already hired a chief commercial officer from Biogen and Roche, is developing a sub Q arm for the registrational study, and intends to meet the FDA after the full dataset. “We are not waiting--not sitting and waiting for top line data to come to plan the next phase.” — Neil K. Warma, Chief Executive Officer · 2026-11-12 That is the right posture for a binary story, but it also means spending is starting to build before proof.

The balance sheet is the other thesis

Here is the number that makes this more than another micro-cap Alzheimer's call: the equity trades at a market value below the cash on the balance sheet. With a market cap around $29.5 million and $44.1 million of cash, the public market is assigning a negative enterprise value to PMN310. That is either a profound mispricing or a clear verdict that the probability-weighted value of the program, after burn and financing risk, is less than zero to current shareholders. The truth will not be known until the readout. The cost trajectory is at least moving in the right direction. R&D expense fell 32% year over year in Q3 as direct PMN310 costs dropped by $3.6 million; the last filed quarter showed $10 million of R&D and a $12 million net loss, so the Q3 improvement is real. Net income improved to an $8.2 million loss from $11.6 million a year ago. And cash runway has been volatile through repeated raises, but management now says the $44.1 million carries the company through 2027, past the PRECISE AD readout. That runway does not obviously fund a registrational study. Management repeatedly used the phrase registration study but declined to give size, cost, or start date beyond towards the end of 2027, subject to FDA feedback. A single Phase 3-style trial plus a sub-Q bridging program is not cheap. If the data are good, the company will likely need to finance or partner; if the data are bad, the cash is the only support. The negative enterprise value suggests the market is already pricing the latter.

The real catalyst is still in front of us

The CTAD conference next week will not bring new data—management was explicit that the poster simply repeats the July interim. The next real event is unblinded 12-month top-line data in Q1 2027. Everything on this call is an attempt to own the narrative before that binary moment: the competitor's failure is not a referendum on Alzheimer's disease antibodies, it is a referendum on selectivity; the interim safety is not efficacy, but it is encouraging; the cash is not a strategic moat, but it buys time. “PMN310 was designed for that selectivity, and our interim data to date in the lab and in patients fully support our hypothesis.” — Neil K. Warma, Chief Executive Officer · 2026-11-12 That is a coherent story. It is not yet a proven one. For a company this small, the stock is less an earnings vehicle than a call option on a mechanistic argument—and the market is currently pricing that option below the cash backing it. The Q1 readout will tell us whether the option is mispriced or merely cheap for a reason.