OLMA / palazestrant-er-positive-mbc — Palazestrant (OP-1250) for estrogen-receptor-positive, HER2-negative advanced or metastatic breast cancer, second and third line

Program analysis · bpiq_drug_id 15137 · prepared 2026-08 (refresh of 2026-08-04) · USD · framework v5.6.1 · Coverage: CLEARED

Company context, financials, ownership, options and the full pipeline: ../company.md.

How to read this. Written for a reader who has not studied pharmacology or finance. Terms specific to this program are in the Glossary below; general terms are defined once in framework/04-glossary.md. Tags: [VERIFIED — source] / [UNVERIFIED] / [WEB ESTIMATE — source, date].

Glossary

Executive summary

• What it is (one sentence): A once-daily tablet that blocks the estrogen receptor at both of its two switches and destroys it, designed to keep working in breast cancers that have stopped responding to standard hormone treatment — whether or not they carry the ESR1 mutation that usually explains that resistance.
• The event and when (as disclosed): Topline results from OPERA-01, the 510-patient Phase 3 trial of palazestrant alone against standard endocrine therapy. On 2026-08-10 Olema completed enrollment and moved the topline guidance from “fall 2026” to “the first quarter of 2027” [VERIFIED — Q2 2026 results release]. That is a quarter, not a day; BPIQ stores 2027-03-31 as a placeholder. The shares fell 12.6% on the news.
• The main reason it could work: The drug blocks a second switch on the receptor (AF1) that every approved competitor leaves alone — the exact mechanistic argument for working in the ESR1 wild-type patients where the competitors do not. Its own Phase 1/2 data are consistent: in the 49 second-/third-line patients, median progression-free survival was 7.2 months overall and 7.3 months in ESR1-mutant patients, so the wild-type patients did about as well [VERIFIED — PubMed, Future Oncol 2026, DOI 10.1080/14796694.2025.2608863].
• The main risk: OPERA-01 has two primary endpoints and both must be met, one per ESR1 stratum. The mutant half has four precedents in its favour. The wild-type half has none — every oral single-agent endocrine drug tested there has failed, which is why elacestrant, imlunestrant and vepdegestrant all carry mutant-only labels. The thesis rests on ~26 wild-type patients in an uncontrolled Phase 1/2 subset, and the registrational dose (90 mg) sits below the dose (120 mg) that generated every published efficacy figure.
• What it means for the stock: At $10.43 the shares sit 17.6% up a 52-week range of $4.91–$36.26, having now fallen 71% from a high printed on a competitor’s data and another 12.6% on their own one-quarter delay [VERIFIED — ../company.md C.4, C.7]. The de-rated setup of the 2026-08-04 analysis is intact and slightly deeper: no dilution mechanism fires on the event, short interest has risen five consecutive settlements to 18.2% of shares outstanding, and the modelled expected value sits above spot across the whole probability band. The costs of the slip are five to eight months of waiting, a first crack in a previously clean execution record, and two Roche FDA decisions landing in the middle of the run-up window.

0. Program-tier coverage — CLEARED

Company-tier coverage is in ../company.md C.0.

A. Scientific & clinical assessment

A.1 Mechanism of action and scientific rationale

About 70% of breast cancers are ER-positive: their cells carry the estrogen receptor, and when estrogen binds it, it switches on the genes that make the cell grow [VERIFIED — PubMed, Future Oncol 2026, DOI 10.1080/14796694.2025.2608863]. The whole treatment strategy is to interrupt that signal, and three ways exist: aromatase inhibitors stop the body making estrogen; SERMs such as tamoxifen sit in the receptor and block estrogen from binding; SERDs both block the receptor and cause the cell to destroy it.

Each has a specific weakness [VERIFIED — same source]. Aromatase inhibitors drive the tumour to acquire ESR1 mutations, after which the receptor is permanently on and no longer needs estrogen. SERMs partly switch the receptor on in other organs. And the only approved SERD that works as a single agent, fulvestrant, has such poor drug-like properties that it must be injected into muscle, capping the deliverable dose.

Palazestrant is a once-daily tablet that is both a SERD and what Olema calls a complete estrogen receptor antagonist (CERAN). The estrogen receptor has two separate transcription switches, AF1 and AF2. Existing SERDs block AF2 and degrade some receptor, but leave AF1 able to work, and laboratory work shows residual receptor protein survives their treatment [VERIFIED — same source]. Palazestrant is designed to block both AF1 and AF2 as well as degrading the receptor, so whatever receptor survives is also silenced — complete shutdown of ER-driven transcription regardless of ESR1 status.

How well the target is validated. As well as any target in oncology. The estrogen receptor has been drugged successfully since the 1970s, with four oral agents approved or filed between January 2023 and June 2026 [VERIFIED — FDA approval announcements; Roche NDA acceptances]. This sweep adds the independent genetics check the last one could not run: Open Targets resolves ESR1 (ENSG00000091831) and the ER+ breast-cancer disease entities cleanly [VERIFIED — Open Targets search_entities 2026-08-12]. Nobody doubts the target; what is doubted is the population claim.

The exact scientific step the next readout must prove. Not that blocking the estrogen receptor helps — that is settled. OPERA-01 must prove that the AF1-plus-AF2 mechanism converts into a measured progression-free survival benefit in patients whose tumours have no ESR1 mutation, on top of a benefit in mutant patients. Both are separate primary endpoints and both must be met.

That is precisely where the class has failed. Imlunestrant’s EMBER-3 produced benefit confined to ESR1-mutant patients (median PFS 5.5 vs 3.8 months, hazard ratio 0.62, 95% CI 0.46–0.82) and was approved for them alone [WEB ESTIMATE — EMBER-3 as reported at SABCS 2024/2025]. Vepdegestrant’s VERITAC-2 likewise; approved 2026-05-01 for mutant patients only [VERIFIED — FDA]. Elacestrant’s label has been mutant-only since January 2023 [VERIFIED — FDA]. The OPERA-01 design paper says it outright: “recent reports indicate that the clinical benefit in ER+, HER2– ABC may be limited to patients with ESR1-mutated tumors, rather than the entire patient population” [VERIFIED — PubMed, DOI 10.1080/14796694.2025.2608863].

The honest scientific risk. The mechanistic argument is good but preclinical, and the clinical evidence for the wild-type claim is a subgroup of roughly 26 patients inside a single-arm study. In the 49 second-/third-line patients, median PFS was 7.2 months overall and 7.3 months in the 23 ESR1-mutant patients [VERIFIED — same source] — nearly identical figures, so the wild-type patients must have done about as well. That is the single most encouraging number in this analysis, and also the thinnest. Nothing published since 2026-08-04 has added to it. A further caution from the company’s own combination data: in the ribociclib arm’s prior-CDK4/6 group, wild-type patients did worse than mutant ones (9.2 vs 13.8 months median PFS) [VERIFIED — same source].

A.2 Clinical development plan, timeline, feasibility, resourcing

Trials

Dates: first and last patient in, database lock, topline. Enrolment began 2023-11-16 [VERIFIED — CT.gov]. Enrollment completion was announced 2026-08-10 — the first hard gate on the trial’s arithmetic ever disclosed — but the completion date itself was not stated, and database lock has still never been disclosed [VERIFIED — Q2 2026 release; absent from CT.gov and all four PubMed records]. The topline is now guided “first quarter of 2027”.

Date cross-check (rule 24). Three sources, reported as three:

Date slippage. The guidance tightened twice, was reiterated once, and has now slipped once — the first slip on record: “Topline data in 2026” (2025-05-13) → “on track for OPERA-01 topline data in H2 2026” (2025-11-10) → “on track for topline data in fall 2026” (2026-03-16) → reiterated (2026-05-12) → “expected in the first quarter of 2027” (2026-08-10). The first four entries are carried from the BPIQ note field as read on 2026-08-04; the feed has since replaced that history with only the newest entry (../company.md C.8) [VERIFIED — BPIQ note field read 2026-08-04 and 2026-08-12; Q2 2026 release]. One slip in fifteen months is still a good record, but the 2026-08-04 analysis leaned on “never slipped” as an execution signal, and that sentence is no longer true.

Feasibility matrix

Resourcing sufficiency. Comfortable through the readout. $461.1M at 2026-06-30 against $14.7–22.4M a month on the three burn readings gives 22–31 months of runway [VERIFIED and modelled — ../company.md C.3]; the Q1 2027 topline needs no new money. Burn is still rising (Q2 R&D $57.8M vs $43.9M a year earlier) with OPERA-02 mid-enrolment. Expect a discretionary raise into strength — the November 2025 offering at $19.00 and the newly surfaced H1 2026 ATM sales at ~$24.5 average are the twice-demonstrated pattern [VERIFIED — ../company.md C.3].

A.3 Target label, target product profile, strategic questions, designations

A.3a Target label. Palazestrant as a single agent, once daily by mouth, for adults with ER+/HER2− locally advanced or metastatic breast cancer progressed after one or two prior lines of endocrine therapy for advanced disease including a CDK4/6 inhibitor — regardless of ESR1 mutation status. That last clause is the entire commercial argument.

A.3b Target product profile

A.3c Strategic Go/No-Go questions. The next decision is registration, so the pre-Phase-III / registration set applies.

A.3d Regulatory designations.

• FDA Fast Track designation for palazestrant in ER+/HER2− metastatic breast cancer progressed after ≥1 endocrine line including a CDK4/6 inhibitor; granted July 2022, restated April 2026 [VERIFIED — press feed; BPIQ note as read 2026-08-04]. What it grants: more frequent FDA contact and eligibility for rolling review; it does not shorten the review clock or lower the evidence bar.
• No Breakthrough Therapy, no Orphan Drug, no Priority Review voucher found anywhere in this sweep [VERIFIED — absent from the 287-item press feed, all four PubMed records, all six CT.gov records].

A.4 Target product profile valuation matrix

Calibration examples, not asset claims: in oncology each incremental month of overall survival is worth roughly $150–300M in peak-sales potential; oral formulations command roughly 15–25% price premiums over injectables.

A.5 Evidence quality and endpoints

Endpoints

A.5b Key opinion leaders.

Panel as of. 2026-08-12.

CT.gov names no investigators for OPERA-01: search_investigators returns only anonymous “Research Coordinator” contacts on the combination study, and the OPERA-01 record itself lists every one of its 233 sites as an unnamed “Clinical Trial Site” with no overall officials [VERIFIED — CT.gov search_investigators and get_trial_details 2026-08-12]. The investigator panel below is therefore the design paper’s academic authorship — the trial’s steering investigators — which is a sponsor-selected subset by construction. No named independent voice on this program’s endpoint was found this sweep (see below), so the independent-voices table is empty and the judgement is written accordingly.

Investigators

(Two further authors, Lianqing Zheng and Elisabeth de Kermadec, are Olema employees and are recorded here as sponsor-side rather than investigators.)

Independent voices

No named independent voice found. The searches made: PubMed (all four records carry Olema co-authors, so none yields an independent commentator), a targeted web search for named commentary on the OPERA-01 wild-type endpoint (returned only sponsor-aligned trade-press quotes from design authors), and the trade press around the 2026-08-11 slip (analyst commentary, not clinical). The nearest thing to independent sentiment remains a survey of 20 breast-cancer research leaders reported at SABCS treating the approved oral SERDs as broadly interchangeable — unattributable to any named person, so it cannot be a row here.

Judgement

Dissent

B. Commercial assessment

B.0 Current treatment algorithm

What a patient with ER+/HER2− advanced breast cancer receives today, and where palazestrant would fit [VERIFIED — DOI 10.1080/14796694.2025.2608863, citing NCCN and ASCO guidelines]:

1. First line. Endocrine therapy — usually an aromatase inhibitor — plus a CDK4/6 inhibitor. Works for a long time; almost everyone eventually progresses.
2. At progression, a blood test. ctDNA is tested for an ESR1 mutation; up to about half of patients on an aromatase inhibitor have acquired one.
3. Second line, ESR1-mutant. Elacestrant, imlunestrant or vepdegestrant, or fulvestrant, or another aromatase inhibitor.
4. Second line, ESR1 wild-type. Fulvestrant or a different AI, possibly with everolimus or (if PIK3CA-mutant) alpelisib. None of the three approved oral drugs is licensed here. This is the gap.
5. Later lines. Chemotherapy or an antibody-drug conjugate — considerably more toxic; steps 3 and 4 exist to delay this point.

Where palazestrant would fit. Steps 3 and 4, aiming at both. In step 3 it would be a fourth entrant with a mechanistic story. In step 4 it would be the first oral single agent with a licence. Step 4 is where the value is, and step 4 is the endpoint that has never been met.

B.1 Competitive landscape and positioning

Where this asset wins, and the single fact the thesis rests on. Unchanged: palazestrant wins in exactly one place — ESR1 wild-type patients in the second and third line, where it would have no oral competition, because every approved oral is mutant-only and giredestrant’s positive second-line trial used a combination. The single fact: in the 49 second-/third-line Phase 1/2 patients, mPFS was 7.2 months overall and 7.3 months in the 23 mutant patients, so the ~26 wild-type patients must have done about as well [VERIFIED — DOI 10.1080/14796694.2025.2608863]. That is the only clinical evidence in existence that this drug behaves differently from its competitors in the population that matters, and it is 26 uncontrolled patients.

The competitive fact that cuts both ways, sharpened. This stock trades as an oral-SERD-class proxy (../company.md C.7), and the class’s next two catalysts belong to Roche, inside this program’s own pre-readout window: giredestrant’s adjuvant decision by 2026-11-30 and its +everolimus decision by 2026-12-18. An approval re-rates the class up (the lidERA readout was worth +130% to OLMA); a surprise rejection de-rates it. Neither says anything about palazestrant’s wild-type endpoint.

Calibration example, not an asset claim: oncology first-movers in novel mechanisms have shown roughly 3.2× higher peak-sales potential but about 1.8× higher development risk.

B.2 Addressable market

Launch markets. The United States first, on a Fast Track designation and an FDA-agreed dose; the 26-country, 233-site footprint is built for multi-region filing [VERIFIED — CT.gov].

B.3 Value and feasibility

B.3a Expected peak sales. United States only, conditional on approval, before ex-US revenue and before the larger first-line opportunity OPERA-02 addresses. Inputs unchanged from 2026-08-04 — nothing in this sweep moved them.

Which inputs are not defensible, named individually. Unchanged: patients (the US incidence figure under the verified percentage chain was not verified) and price (the only anchor, elacestrant’s US list price, still shows a more-than-tenfold spread between sources — ~$26,105 per 30-day supply vs ~$2,500 per cycle — neither a net price) [WEB ESTIMATE — retail pricing sources; conflict unresolved]. Share is modelled from B.1.

Cross-check against published third-party figures. GlobalData’s model now circulates as “$1.3bn in 2031” [WEB ESTIMATE — clinicaltrialsarena/GlobalData, 2026] alongside the earlier $686M-by-2036 page the 2026-08-04 analysis used; Stifel initiated in February 2026 calling peak sales “exceeding $3 billion” [WEB ESTIMATE — Stifel initiation coverage, 2026-02]. The base case above sits below all of these; the spread between them (2×–4×) is a measure of how unpriced this asset’s label breadth is, not a reason to raise the base.

These figures are conditional on approval and exclude: ex-US revenue; the OPERA-02 first-line opportunity; OP-3136; partnership economics.

B.3b Feasibility

B.4 Product-development risk

Framing questions.

1. Does the development plan support the target product profile claims? Yes, precisely: two co-primary endpoints, one per ESR1 stratum, randomisation stratified on the same variable, tested centrally on blood. The design does not hedge — if only the mutant endpoint hits, the company’s own trial refutes the profile’s central claim.
2. Will the identified risks affect the target product profile? Two will: the 90 mg registrational dose attacks the efficacy claim (no published efficacy at that dose), and 62.8% nausea attacks the tolerability claim in a chronic daily tablet.
3. If a risk cannot be mitigated, is the asset still differentiated? Only in one scenario. A wild-type failure leaves a fourth mutant-only oral, three-plus years late, with no head-to-head data. There is no fallback differentiation; the mechanism claim and the commercial case are the same claim.

Readout

Sources

The modelled estimate. Registry primary completion (2026-06-30) plus the rule 34 default of two to four months to database lock and analysis gives late August to late October 2026 — a range the company’s own fresh Q1 2027 guidance has already overtaken. The arithmetic is recorded so it can be argued with; its input is stale, which is the argument against it.

Window

Basis. Earliest is the guided quarter’s open: the guidance is two days old, issued alongside the enrollment-completion announcement, and a company that has just pushed a date out does not then beat it by weeks. Likeliest is mid-quarter, because an event-driven dual-primary PFS analysis plus lock and cleaning tends to consume most of a guided quarter, and this management has shown it guides to windows it only just holds. Latest absorbs one further quarter of slip, because the record now contains exactly one slip of exactly one quarter and nothing rules out a second. Precision is PERIOD — a quarter is named, no month, no day. Confidence is MEDIUM: the guidance is fresh and anchored to a completed gating event (enrollment), but the same company’s previous window was reiterated four times and still slipped, and the registry offers no independent check.

Disagreement. UNRESOLVED — but asymmetrically. The registry (primary completion 2026-06-30, still RECRUITING) and the modelled estimate built on it point earlier than, and contradict, the company’s Q1 2027 guidance. The registry record is demonstrably unmaintained: its status field contradicts the sponsor’s own enrollment-completion announcement of 2026-08-10, which is the stronger evidence. Neither is averaged; both are recorded; the window is built on the company source and says so (rule 24).

Date slippage.

Five statements, four transitions: two tightenings, one reiteration, one slip (count: 1). The first four entries are carried from the BPIQ note field as read at the 2026-08-04 sweep; the feed has since replaced that history (../company.md C.8).

Attribution

Status. CLEAN — as computed by lib/clustering.mjs’s attributionFor over this ticker’s pipeline for bpiq_drug_id 15137, run 2026-08-12.

Conflicts

Row 15137 is the only has_catalyst: true row of the eight, so the conflict set is empty by construction — nothing else on Olema’s own pipeline can land inside the window. The two Roche FDA decisions of 2026-11-30 and 2026-12-18 are not attribution conflicts (they are another company’s catalysts, outside the computation’s scope, which spans this ticker’s pipeline only), but they are named in the stock call and the run-up basis because this equity demonstrably trades on them (../company.md C.7).

Note. (blank — status is CLEAN)

Market and timing for this event

• Plain takeaway. The de-rated setup of the 2026-08-04 analysis got more extreme: the shares now sit at $10.43 — 17.6% of a 52-week range running $4.91 to $36.26 — after falling 71% from a competitor-made high and then another 12.6% on their own one-quarter slip [VERIFIED — ../company.md C.4, C.7]. No dilution mechanism fires on the event, short interest has risen five consecutive settlements to 18.2% of shares outstanding, and the event itself moved five months further away. What is priced in is scepticism about the wild-type endpoint plus, now, a timing discount; what is not priced in is either side of the readout.
• Months to this catalyst. About 4.7 months to readout.window.earliest (2027-01-01), seven to the likeliest mid-quarter landing. Precision is PERIOD — a quarter is named, no month, no day — so every timing statement here inherits that coarseness.
• Expected move around this event. A bracket of roughly 35% to 75%, not a point. ../company.md C.6: the January 2027 $10 straddle prices ~51% at the mid on readable IV (~0.96–1.20), but at-the-money depth is negligible, spreads exceed the mid, and — structurally — no listed expiry spans the guided window, so the chain cannot cleanly price the event at all.
• Nearest comparable past reaction. Still none that fits. The 2025-11-18 giredestrant class re-rating (+130% to +197%) and the 2026-03-09 persevERA de-rating (−33.4% over four weeks) are the plausible-magnitude brackets, both made of Roche’s data. The newest calibration point is Olema’s own: −12.6% on a one-quarter timing slip with no efficacy content (2026-08-11). Olema has still never reported a registrational efficacy result [VERIFIED — ../company.md C.7].
• Materiality. Dominant, per ../company.md C.2 — the only pipeline row with a disclosed catalyst, the sole registrational readout, and a 12.6% one-day repricing on a pure timing change proves the equity hangs on it. The stock call below is consistent with that (rule 27). OPERA-02 (topline ~2028) and OP-3136 stop it being fully binary.
• Date slippage. One slip (of one quarter) after two tightenings and a reiteration — see Readout. The 2026-08-04 analysis’s “fifteen months of unmoved guidance” argument is retired.

Spot. $10.43, the 2026-08-11 close, read 2026-08-12 from BPIQ (../company.md C.4). The committed price cache’s own bar for 2026-08-11 carries a null close (C.8), so the cache’s last non-null close is the pre-slip $11.94 of 2026-08-10; every figure below uses $10.43 and says so.

Scenario prices

Excluded anchors, named rather than omitted. A dilution mechanism that fires on the event: none exists (miss row). This ticker’s own past readout on its own registrational data: none exists; the largest own-data move in the catalyst feed remains +11.57% on an IND clearance, and the newest own-news print (−12.6%) was a timing slip, not data.

Expected value. 40% applied to the midpoint of each range: 0.40 × $23.75 + 0.60 × $6.50 = $13.40, which is +28.5% against the $10.43 spot. Arithmetic, not advice, not a price target. The sign does not flip inside the band: at 30% the expected value is $11.68 (+11.9%); at 52% it is $15.47 (+48.3%). As at the prior sweep, that stability is a consequence of how far the shares have fallen, not of confidence in the trial — and it survived a 12.6% further fall with the probability unchanged.

Run-up

Entry and exit

Predicted move and peak

Priority score drivers

Priority score 13 · formula_version 1.0.0

(Computed by lib/runup.mjs priorityScore. The PERIOD-precision date gates the strong underlying setup down hard — the six other drivers combine to a base of ~65 before the date-confidence gate multiplies it by 0.2 and the sub-threshold ceiling applies. That is the formula doing its job: a quarter is not a tradeable date.)

Settlement. Left null at lock time, per 05-prediction-protocol.md § Run-up settlement. The exit rule cannot yet resolve to a date: the committed cache ends 2026-08-11, before the window opens (lib/runup.mjs resolveExit returns null).

Verdict

What I would do. Buy, small — same call as 2026-08-04, at a 9% better price and a five-month longer wait, sized for a real chance of losing half.

Why. Nothing about the science moved: the AF1-plus-AF2 differentiation, the unhedged two-endpoint design, and the 26-uncontrolled-patient thinness of the wild-type evidence are all exactly where they were. What moved is the price (−12.6% on a timing slip with zero efficacy content), the calendar (topline Q1 2027, enrollment now complete — the largest operational uncertainty resolved in the trial’s favour), and the execution record (first slip after fifteen clean months — a real cost, priced the day it happened). At $10.43 the market pays ~$620M of enterprise value for an asset whose US-only base-case peak is ~$715M a year conditional on approval, with no dilution trigger on the event, short interest rising five straight settlements to 18.2% of shares outstanding, and a modelled expected value above spot at every point in the probability band. The slip also created a specific new risk this document did not carry before: five months of dead money through two Roche FDA decisions that can re-rate the class in either direction before Olema says a word.

What would change this. To watch: a second timing slip in the Q3 2026 results (~November), a CT.gov update pushing primary completion past Q1 2027, or a run back above roughly $18 before the readout (which would remove the asymmetry). To a larger position: pre-topline evidence that 90 mg reproduces 120 mg exposure and activity, a commercial partnership for the second-line launch, or a giredestrant adjuvant approval by 2026-11-30 that re-rates the class while OLMA still trades under ~$13.

What to watch.

• 2026-08-12 to ~2026-11: whether CT.gov NCT06016738 is finally updated (status to ACTIVE_NOT_RECRUITING, a new primary completion date). The record has been stale for months; a fresh primary-completion date would be the first independent check on the Q1 2027 guidance.
• By 2026-11-30: the FDA decision on giredestrant adjuvant (Priority Review) — the largest class event before topline; this stock moved +130% on the same drug’s adjuvant data.
• By 2026-12-18: the FDA decision on giredestrant + everolimus in ESR1-mutant advanced disease — a second-line-adjacent class event.
• ~2026-11 (Q3 2026 results): whether “first quarter of 2027” survives its first reiteration. One slip is a record; two is a pattern.
• Any time: a first disclosure of database lock, which would let the Q1 2027 guidance be checked arithmetically; and any ATM activity visible in the Q3 10-Q — Olema sold $41.9M of stock quietly in H1 2026 and would plausibly do it again into any pre-readout strength.
• At the announcement, in order: (1) did the ESR1 wild-type co-primary reach significance — the whole thesis; (2) hazard ratios per stratum against imlunestrant’s 0.62 mutant benchmark; (3) discontinuation and nausea at 90 mg against the 62.8% seen at 120 mg.
• Within days of a positive announcement: an equity raise off the March 2026 shelf or the ATM.

Locked prediction

• Outcome-direction (will the readout succeed on its primary endpoint?): uncertain, leaning miss — probability 40%, band 30–52% [UNVERIFIED — modelled]. The probability that both co-primary endpoints reach statistical significance, unchanged from the superseded 2026-08-04 call because nothing in this sweep bears on trial odds: no new clinical data published, the slip is a timing event, and enrollment completion (mildly positive: the trial is fully powered as designed, with no futility action reported) roughly offsets the negative signal-value of guidance slipping at the first hard gate. It still decomposes into ~75–80% for the mutant endpoint (four precedents) and ~45% for the wild-type endpoint (no precedent, ~26 uncontrolled supporting patients), positively correlated through the shared effect size; 90 mg dose risk and the open-label design are reflected in it. No third party publishes a probability on this event; post-slip analyst targets of ~$36–59 against a $10.43 spot imply materially more optimism than 40% and are named rather than adopted.
• Stock-direction (which way do the shares move?): up — confidence low — window 2026-08-12 to 2027-05-15, basis: the guided Q1 2027 topline window (to 2027-03-31) plus about six weeks for the reaction to settle; no listed option expiry spans the window (C.6). Materiality is dominant per ../company.md C.2 and the call is consistent with it. Attribution is CLEAN within this ticker’s own pipeline, but the window contains two giredestrant FDA decisions (2026-11-30, 2026-12-18) on which this class-proxy equity demonstrably moves; the direction call owns that exposure rather than assuming a quiet window. Called up despite a below-even outcome probability because the asymmetry sits in the price: 17.6% of the 52-week range, a fresh −12.6% timing discount, no dilution trigger, 18.2% of shares short and rising, and an expected value above spot across the whole band. Confidence is low because a single unprecedented endpoint decides it, the company has never reported a registrational result, and the execution record just took its first crack.
• Scenario prices: positive $15.50–$32.00 · miss $4.50–$8.50
• Expected value: $13.40, +28.5% against spot $10.43 (arithmetic, not advice)
• Run-up: entry $10.43 on 2026-08-12, exit rule T-5 trading days before readout.window.earliest — predicted move −10% to +30%, predicted peak +10% to +40% around 2026-12/2027-01 — priority score 13, formula_version 1.0.0 (the PERIOD-precision date gates an otherwise strong setup; see Run-up above)
• Settles on the first public disclosure of OPERA-01 (NCT06016738) topline results. Positive is defined as: blinded-independent-committee-review-assessed progression-free survival showing a statistically significant improvement for palazestrant over standard-of-care endocrine therapy in both pre-specified co-primary populations — patients with a detected ESR1 mutation and patients with no detected ESR1 mutation — each at conventional two-sided significance as pre-specified in the trial’s own analysis plan. Anything else is a miss, including the case where only the ESR1-mutation-detected endpoint is met. Settled from Olema’s own press release or a peer-reviewed or conference presentation of the topline data, cross-checked against the ClinicalTrials.gov record.
• Locked: yes · Settled: no

Program data-quality flags

• catalyst_date 2027-03-31 is a synthesized quarter-end placeholder generated from “Q1 2027” (rule 23). Never a scheduled date; all timing reads readout.window.
• The ClinicalTrials.gov record for NCT06016738 is demonstrably unmaintained. Status RECRUITING and primary completion 2026-06-30 both contradict the sponsor’s own 2026-08-10 enrollment-completion announcement. The 2026-08-04 analysis reported this as a two-way unresolved conflict; the Q2 release resolved which reading was true (both: the registry was stale and the fall topline was at risk). The registry currently offers no independent check on the Q1 2027 guidance, which is itself a finding.
• The BPIQ note field was replaced, not appended: the 15-month guidance history the prior sweep read from it is gone from the feed. The slip sequence above carries its first four entries from the 2026-08-04 record, tagged as such.
• The committed price cache’s 2026-08-11 bar has a null close with real volume — the −12.6% slip-reaction session itself. Spot, the run-up entry and the 52-week position in this document use BPIQ’s $10.43 and say so; cache-derived figures resolve to the pre-slip $11.94 until the bar self-heals on the next fetch.
• Database lock has never been disclosed; enrollment completion was announced without a date. The Q1 2027 guidance still cannot be checked against the trial’s own arithmetic from outside.
• The modelled readout estimate rests on a stale registry input and points before the company’s own guidance; it is recorded with its arithmetic and given no weight in the window (Readout).
• Open Targets was CALLED successfully this sweep — the standing global throttle documented in 02-connectors.md did not fire — closing the one coverage gap the 2026-08-04 analysis carried.
• The registrational dose (90 mg) differs from the dose behind every published efficacy figure (120 mg); the IDMC’s unblinded Part 1 data are not public. Unchanged.
• Every publication on this molecule has Olema employees as co-authors; no independent replication of the AF1 claim. Unchanged — the PubMed hit set did not move between sweeps.
• The elacestrant price anchor remains internally contradictory (>10× spread, neither figure a net price); B.3a’s $100,000–$180,000 is a judgement bracket, not a traced figure.
• The US patient count under B.3a remains unverified; the 18,000–30,000 range is modelled and all three peak-sales scenarios inherit it.
• No patent number, expiry, licence or royalty for palazestrant was found; the Aurigene royalty attaches to OP-3136. Market exclusivity stays recorded as unassessable.
• The BPIQ historical-catalyst feed omits the events that moved this stock most — both Roche moves and now the 2026-08-11 slip reaction. Any timing read built on it alone would be wrong.
• GlobeNewswire timed out on three fetch attempts for the Q2 release; it was read in full from a BioSpace mirror and cross-checked against the 10-Q, which agrees on every overlapping figure.