Early Results Of Verastem's VS-7375 In Pancreatic Cancer, Or How To Read A Reduction In CA19-9

Disclosure: The author holds a beneficial long position in Verastem Oncology, Inc. (NASDAQ: VSTM). The author holds no position, long or short, in Revolution Medicines, Inc. (NASDAQ: RVMD) or GenFleet Therapeutics (Shanghai) Inc. (HKEX: 2595). This article is provided for informational and educational purposes only and is not financial advice. Although the author is a physician, the discussion of CA19-9 and the interpretation of the TARGET-D 101 data below represent a personal analytical and clinical perspective on the underlying science and do not constitute medical advice, a diagnosis, or a treatment recommendation, nor do they establish a physician-patient relationship. The author receives no direct compensation for this article and has no business relationship with any of the companies referenced. This article has not been independently peer-reviewed. Please see the full "Legal Information and Disclosures" section below.

In April I simulated four non-covalent KRAS G12D inhibitors, and my computational model suggested Verastem's investigational oral KRAS G12D (ON/OFF) inhibitor VS-7375 may reach the deepest signaling suppression of the group, if the published picomolar binding holds up. Verastem licensed VS-7375 from the Chinese biotech GenFleet in January 2025. GenFleet develops the molecule in China as GFH375, and reported a ~41% response rate at 600 mg in 59 heavily pretreated patients with advanced pancreatic ductal adenocarcinoma in a phase 1/2 monotherapy study. On June 23, 2026, Verastem released the first efficacy update from its own US trial in metastatic pancreatic ductal adenocarcinoma (mPDAC), TARGET-D 101. In this article I will only focus on the mPDAC cohort, although Verastem also reported promising results of VS-7375 in metastatic colorectal cancer (mCRC) and advanced non-small cell lung cancer (NSCLC).

Verastem enrolled more than 20 previously treated mPDAC patients at each of the 600 mg and 900 mg doses, but only 7 at 600 mg and 1 at 900 mg had completed at least six months of follow-up. Dose escalation continues at 1200 mg. With follow-up this short, Verastem could not report a meaningful objective response rate or disease control rate.

At 900 mg, Verastem reported a CA19-9 reduction of more than 50% in 13 of 14 patients, all of whom had a baseline above 37 U/mL and at least one scheduled on-treatment CA19-9 assessment, and all of whom remain on treatment. Verastem did not disclose how many 900 mg monotherapy patients those two criteria excluded, and the release gives a data cutoff of June 12, 2026 for the safety analysis but provides no time point at which CA19-9 was measured over the course of treatment. Verastem also reported dose-dependent activity between 600 and 900 mg, with 900 mg reaching target plasma exposure and separating clearly from 600 mg, and low-grade gastrointestinal toxicity at both doses that eases after the first cycle. So the only quantified efficacy readout in the monotherapy mPDAC cohort from TARGET-D 101 is a reduction in CA19-9. What does this tell us?

CA19-9 is a sialylated Lewis blood-group antigen produced by pancreatic and biliary epithelium and a sensitive and specific serum marker in PDAC. In my clinical practice it is part of the standard laboratory panel that patients with resectable PDAC receive before surgery and at every follow-up. Nevertheless, reading it is not trivial, and it is not a validated surrogate marker.

The FDA-NIH BEST glossary defines a surrogate endpoint as "a substitute for a direct measure of how a patient feels, functions, or survives," and grades surrogates by their level of clinical validation as validated, reasonably likely, or candidate. CA19-9 in PDAC is only a candidate surrogate, because validation would require randomized evidence that a drug's effect on the marker predicts its effect on survival. That evidence has not been produced and may never be.

In May 2026 Yeh et al. reported that among 615 PDAC patients, median survival was 13.5 months for patients with CA19-9 levels of 7 U/mL or below, 23.2 months from 7 to 37 U/mL, 22 months from 37 to 200 U/mL, and 12.8 months above 200 U/mL, so the relationship between the absolute value and survival is not monotonic. A value below 7 U/mL need not mean little disease, since it can also reflect an inability to produce CA19-9 in Lewis-negative patients, who lack a functional fucosyltransferase. The prognosis of these nonproducers was as poor as that of patients with the highest CA19-9 levels, and 44% (51 of 116) of those reading at or below 37 U/mL were nonproducers of this kind.

A change from an elevated CA19-9 baseline appears to carry more information than any absolute value, and this is the setting of the 14 patients Verastem reported.

In the phase III MPACT trial of nab-paclitaxel and gemcitabine in mPDAC, patients in the combination arm with any decrease in CA19-9 by week eight had a median survival of 13.2 months against 8.3 for those without. MPACT also puts the more-than-50% reduction Verastem reported in 13 of 14 patients in context: at week eight in the combination arm, 82% of patients had some decrease in CA19-9, 78% at least 20%, and 58% at least 60%.

A retrospective analysis of ACCORD 11/PRODIGE 4 pointed the same way: in the pooled population a week eight decrease of at least 20% was associated with better overall and progression-free survival (both p < 0.001), but broken down by arm the association reached significance only in the gemcitabine arm (OS p = 0.030, PFS p < 0.001), not the FOLFIRINOX arm (OS p = 0.063, PFS p = 0.693).

Both are associations observed under "conventional" cytotoxic chemotherapy, and it does not formally follow that they transfer to an investigational KRAS-targeted therapy.

MPACT and ACCORD 11/PRODIGE 4 both measured CA19-9 at a fixed eight weeks, whereas Verastem did not report the time point at which it measured the reductions it reported. In August 2021 Tomishima et al. published a retrospective series of 79 patients with unresectable locally advanced PDAC that asked how long a reduction has to last to matter. A reduction of more than 44% from baseline separated long-term survivors, but only when it persisted: a reduction sustained beyond two months did not predict survival significantly (p = 0.1), and only a reduction sustained beyond three months reached significance, at p = 0.04. Verastem gave neither a time point nor a duration for its CA19-9 reduction, so it could rest on a single early reading of the kind Tomishima et al. found uninformative. However, it should be noted that Tomishima's retrospective cohort consisted of patients with unresectable locally advanced PDAC, rather than mPDAC.

A tumor in the pancreatic head may compress or invade the distal common bile duct, and where it does, the resulting cholestasis can raise CA19-9 independently of tumor burden. In my own clinical experience cholestasis alone can drive the marker to high levels. Obstruction is then usually relieved by ERCP with stent placement, and decompression can lower the marker substantially whether or not the tumor has responded. Among 128 patients admitted with obstructive jaundice, CA19-9 fell after endoscopic drainage in 19 of 38 malignant cases, with no antitumor drug involved. Verastem reported nothing about biliary obstruction or intervention in this cohort. However, it is unlikely, in my opinion, that a reduction this uniform across 14 patients rests mainly on stenting. Most patients in this second-to-fourth-line setting would probably have been stented earlier in their course if necessary, but we cannot know, because Verastem did not report it.

Revolution Medicines reported a different measure for zoldonrasib, its KRAS G12D (ON) inhibitor. In a phase 1 study, 40 PDAC patients received zoldonrasib at 1200 mg per day, 20 as a single daily dose and 20 as 600 mg twice daily, with a 30% response rate and 80% disease control. Among the 28 with KRAS G12D detectable in ctDNA at baseline and evaluable on treatment, the allele fraction fell by more than 50% in 86% and cleared entirely in 39%. Allele fraction in ctDNA tracks the mutant clone, but it is no more validated as a surrogate than CA19-9 and is not standard of care. GenFleet's own ESMO data found detectable plasma KRAS G12D in only 71% of the patients with baseline ctDNA available (44 of 62), whereas CA19-9 is inexpensive and drawn anyway. The allele fraction might also miss subclones that have already escaped a G12D-targeted therapy.

Revolution Medicines' multi-selective RAS(ON) inhibitor daraxonrasib is further along. Its phase 3 RASolute 302 trial, which I covered on the topline in April and which has since been presented in the ASCO plenary on May 31 and published in the New England Journal of Medicine, raised median overall survival in the intention-to-treat population to 13.2 months against 6.7 on chemotherapy, a hazard ratio of 0.40.

A reduction of more than 50% in 13 of 14 patients with initially elevated CA19-9 levels is an encouraging early signal, but it does not yet reliably suggest a survival benefit, since the time point is unstated, the marker is not validated as a surrogate, its behavior under KRAS inhibition rather than cytotoxic chemotherapy is not sufficiently tested, and biliary decompression may lower it on its own.

Verastem has meanwhile dosed the first patient in the registration-directed Phase 2 trial TARGET-D 201 on June 16, expects to complete enrollment across three Phase 2 trials by year end, and plans to begin Phase 3 trials in the first half of 2027. So far the company has committed only to a TARGET-D 101 update in the second half of 2026. With longer follow-up, that update may bring the first confirmed RECIST responses and an early read on how durable they are.

Follow me on X for frequent updates (@chaotropy).

Legal Information and Disclosures

General Disclaimer & No Financial Advice: The content of this article is for informational and educational purposes only. It represents the personal opinions of the author as of the date of publication and may change without notice. The author is not a registered investment advisor or financial analyst. This content is not intended to be, and shall not be construed as, financial, legal, tax, or investment advice. It does not constitute a personal recommendation or an assessment of suitability for any specific investor. This article does not constitute an offer to sell, a solicitation of an offer to buy, or a recommendation of any security. Readers should conduct their own independent due diligence and consult with a certified financial professional before making any investment decisions. This article is not a securities research report and has not been prepared in accordance with legal requirements designed to promote the independence of investment research.

Medical Disclaimer: Although the author possesses a medical background, the information presented here regarding clinical trials, biomarkers, or pharmaceutical mechanisms is strictly for the purpose of educational discussion and general commentary regarding the underlying science. It does not constitute medical advice, a diagnosis, or a treatment recommendation, nor does it establish a physician-patient relationship. The author's references to personal clinical experience are anecdotal, are offered only to frame the interpretation of published data, and are not evidence and not generalizable to any individual patient. Readers should never disregard professional medical advice or delay in seeking it because of something read in this article. Always consult a qualified healthcare provider regarding any medical condition.

Accuracy and Third-Party Data: Clinical trial data, press releases, regulatory filings, and published literature referenced in this article are sourced from ClinicalTrials.gov, company press releases, peer-reviewed journals, and conference abstracts and presentations. While the author believes these sources to be reliable, the completeness, timeliness, or correctness of this data cannot be guaranteed. The author assumes no liability for errors, omissions, or the results obtained from the use of this information.

Disclosure of Interest: The author holds a beneficial long position in Verastem Oncology, Inc. (NASDAQ: VSTM). The author holds no position, long or short, in Revolution Medicines, Inc. (NASDAQ: RVMD) or GenFleet Therapeutics (Shanghai) Inc. (HKEX: 2595). The author reserves the right to buy or sell these securities at any time without further notice. The author holds no position in any other company referenced in this article. The author receives no direct compensation for the production of this content and maintains no business relationship with any of the companies mentioned. The author has no access to non-public preclinical data, clinical trial data, unblinded interim results, or internal company information. Readers should be aware that the author's long position in Verastem, whose investigational drug is the subject of this article and is compared here against candidates from Revolution Medicines in which the author holds no position, creates a potential for confirmation bias in framing, selection of comparators, and interpretation of the data, despite efforts to mitigate this through hedged language, transparent presentation of what was and was not disclosed, and citation of primary sources.

Forward-Looking Statements & Risk: This article contains forward-looking statements regarding clinical trial outcomes, regulatory pathways, and the potential clinical value of investigational candidates. These statements are based on current publicly available information and are subject to significant risks and uncertainties. Actual results may differ materially. Preliminary, interim, and biomarker-based data may not be predictive of the results of ongoing or later clinical trials. Investing in biotechnology and pharmaceutical securities involves a high degree of risk, including the potential for total loss of principal. Past performance is not indicative of future results. No reader should rely on this article as a basis for any investment decision. The author has no obligation to update this article if new information becomes available or if the author's position changes.

Copyright: All original content, including text, is the property of the author and may not be copied, reproduced, or published, in whole or in part, without prior written consent, except as permitted by applicable law or the terms of the platform on which it is published. Use of this content for training machine learning or AI models is not permitted without explicit authorization. Third-party or public domain images remain subject to their respective rights and are not claimed as the author's property.