Elraglusib

Elraglusib is Actuate Therapeutics' lead asset - a small molecule originally developed as a GSK-3β inhibitor but which a 2024 mechanistic paper argues works primarily by destabilizing microtubules, opposite in direction but at the same cellular target as taxanes [1]. The headline event: a randomized Phase 2 (the 1801 study) combining elraglusib with gemcitabine/nab-paclitaxel (GnP) beat GnP alone in metastatic pancreatic ductal adenocarcinoma (PDAC), with median overall survival of 10.1 vs 7.2 months, hazard ratio 0.63 (log-rank p=0.01), and roughly doubled 12-month survival (44.4% vs 22.3%); follow-up at ASCO GI 2026 also reported a ~5x increase in 2-year survival (12.9% vs 2.6%) [2,11]. Full results were published in Nature Medicine in 2026 [2]. The pipeline node referenced here is a separate investigator-sponsored Phase 2 at Mass General (NCT05077800) testing elraglusib on top of FOLFIRINOX (a four-drug regimen of folinic acid, fluorouracil, irinotecan, and oxaliplatin) with losartan - a different combination, different patient flow, run by Colin Weekes rather than Actuate directly [3]. For Actuate (ticker: ACTU, IPO'd August 2024), elraglusib is essentially the entire company. The stock moves as a binary bet on PDAC registration. With 1801 positive, the question shifts from 'does it work?' to 'will FDA accept a randomized Phase 2 as a registration trial, or do they need a Phase 3 with limited cash?'

Novel compound, never approved anywhere. Elraglusib has Orphan Drug and Fast Track designations from FDA for pancreatic cancer, and Actuate has stated it plans to pursue an accelerated approval pathway based on the 1801 Phase 2 data [4]. Actuate disclosed positive top-line results from the randomized Phase 2 1801 study (elraglusib + GnP vs. GnP alone in untreated metastatic PDAC, n=233 in the randomized portion) in late 2024, full publication in Nature Medicine in 2026 [2], with extended follow-up presented at ASCO GI on January 12, 2026 [11]. Earlier Phase 2 single-arm data appeared in ESMO Open in 2025 [5]. The first-in-human Phase 1 was published in Clin Cancer Res in 2024 - a 162-patient dose-escalation across solid tumors showing manageable safety [6]. Beyond pancreatic cancer, Actuate has signals in pediatric Ewing sarcoma and salivary gland carcinoma, the latter recently reported in Clin Cancer Res [7]. The investigator-sponsored MGH trial (NCT05077800, n=70) is listed as active, not recruiting as of mid-2026, primary endpoint progression-free survival on the FOLFIRINOX + losartan + elraglusib regimen [3]. A second NCI-sponsored trial (NCT06896188, RiLEY) is enrolling 12 patients with elraglusib + retifanlimab (anti-PD-1) + modified FOLFIRINOX [8]. Actuate's most recent 8-K (May 2026) signals continued regulatory engagement around the registration pathway [9].

Here's the awkward part. Elraglusib was marketed for a decade as a selective inhibitor of GSK-3β - a kinase (an enzyme that adds phosphate groups to other proteins) that sits at the crossroads of Wnt signaling, glucose metabolism, and inflammation. The pitch was that cancer cells, especially pancreatic ones, lean on GSK-3β to survive chemo, so blocking it sensitizes tumors. A 2024 paper from Coats et al. (Saurin lab, Dundee) in Cancer Res Commun blew that up: they showed elraglusib's cytotoxicity in cancer cells comes from directly destabilizing microtubules - the same protein scaffolds that taxanes and vincas bind - and the effect is independent of GSK-3 inhibition [1]. In other words, elraglusib is probably a tubulin-binder that also inhibits GSK-3β at higher concentrations. Two implications. First, the combination with nab-paclitaxel is mechanistically mixed: nab-paclitaxel stabilizes microtubules while elraglusib destabilizes them - opposite directional effects on the same structural target rather than identical mechanisms, so the concern is convergent on-target toxicity (neuropathy, myelosuppression) rather than redundant pharmacology. Second, the validated drug class is no longer 'GSK-3 inhibitor' (a graveyard of failures across CNS and oncology - see Walz et al. 2017 review [10]) but 'microtubule-active agent' - a well-trodden but crowded space. Complicating both stories: 1801 correlative data reported 7-40x increases in tumor-infiltrating cytotoxic immune cells in the elraglusib arm, suggesting a clinically relevant immunomodulatory signal that neither the GSK-3 nor the microtubule frame fully explains [2]. The positive 1801 readout means the drug does something useful; the mechanism story Actuate has been telling needs revision.

Two trials matter for this node. NCT05077800 (the trial this node was ingested from) is an investigator-initiated Phase 2 at MGH, n=70, single-arm, testing modified FOLFIRINOX + losartan + elraglusib in metastatic PDAC. Primary endpoint is progression-free survival. Status: active, not recruiting [3]. The losartan piece is interesting - Boston groups (Jain lab) have argued the angiotensin blocker remodels pancreatic tumor stroma to improve drug delivery, so this trial is layering two stromal/sensitizer hypotheses on top of chemo. The trial does not have a control arm, which limits interpretability - single-arm Phase 2 in PDAC has a long history of looking good and then failing Phase 3. The key commercial dataset is Actuate's own 1801 study: randomized, open-label Phase 2, 233 evaluable patients with previously untreated metastatic PDAC, randomized to GnP ± elraglusib. Primary endpoint was overall survival. Headline numbers: median OS 10.1 vs 7.2 months, HR 0.63 (log-rank p=0.01), 12-month OS 44.4% vs 22.3% (a doubling), 24-month OS 12.9% vs 2.6% at ASCO GI 2026 follow-up [2,11]. Published in Nat Med 2026 [2]. The randomization is what makes this credible. The open-label design and modest size are what make FDA's willingness to grant accelerated approval uncertain - pancreatic cancer regulators have historically wanted a confirmatory Phase 3, particularly when the experimental arm has more toxicity.

Our model gives this drug a 5% chance of eventually being approved. That starts from a 13% historical approval rate for Phase 2 drugs in this area, then gets adjusted based on ten facts about the trial and sponsor. The estimate is helped by the trial's light or open-label blinding, and held back by the sponsor's thin approval record, weak earlier-phase results, and a randomized design. Most other factors landed near average, so they left the estimate close to where the base rate put it.

Efficacy risk: 1801 was open-label, n=233 in the randomized portion. PDAC Phase 3 confirmatory trials have a brutal record of erasing Phase 2 OS signals (see the long list of post-MPACT failures - masitinib, evofosfamide, ruxolitinib all looked promising in Phase 2 and missed in Phase 3). Without a predictive biomarker stratifying responders, elraglusib could be a modest-benefit drug that gets washed out at scale, even with the HR 0.63 starting point. Safety risk: stacking a microtubule-destabilizing agent on top of nab-paclitaxel (a microtubule stabilizer) raises real questions about convergent on-target toxicity - peripheral neuropathy and myelosuppression - that are hard to fully evaluate without the complete 1801 safety tables. Mechanism risk: the 2024 Coats et al. finding means the company's biomarker strategy (if tied to GSK-3 pathway) is probably wrong; any rational predictive biomarker work needs to restart [1]. The immunomodulatory signal at 1801 [2] also reopens the question of which patient subset benefits most. Execution risk: Actuate's cash runway. The 2024 10-K (filed March 2025) describes a small-cap balance sheet that, based on disclosed operating expenses, likely supports only a few quarters of runway without dilution or partnership [4] - running a confirmatory Phase 3 in PDAC (~$80-150M) without a partner is hard. Commercial risk: NALIRIFOX (liposomal irinotecan + 5-FU + oxaliplatin) became first-line standard of care (SOC) for metastatic PDAC in 2024 based on NAPOLI-3, eroding GnP's first-line share. Elraglusib's data is on the GnP backbone; a label limited to GnP may capture a shrinking slice unless Actuate runs a NALIRIFOX combination study. Regulatory risk: FDA may demand a Phase 3 before accelerated approval.

Worth watching. This is one of the few PDAC programs with a positive randomized Phase 2 OS readout, and follow-up at ASCO GI 2026 (January 12, 2026) extended that signal into the 2-year survival tail - a 5x improvement that's hard to dismiss as noise [2,11]. The thing to watch now is FDA's response to Actuate's pre-NDA or Type B meeting outcomes, which should surface in 8-K filings over the next few quarters [9]. Two specific signals would change the call: (1) FDA agrees to accept the 1801 study for accelerated approval with a confirmatory Phase 3 - that's a multi-bagger setup for ACTU equity; (2) a partnership or licensing deal with a larger oncology player to fund Phase 3 - Actuate doesn't have the balance sheet to go it alone, and a partner validates the data externally. Conversely, if FDA demands a Phase 3 with no AA pathway and no partner emerges, the asset is functionally stranded at a company that can't develop it. Next data catalysts: ASCO 2026 in June for any updated subgroup or biomarker analyses, then ASCO GI 2027 in January for the next major OS update. The microtubule-versus-GSK-3 mechanism story is also worth tracking - if independent groups confirm the Coats/Saurin finding, the entire scientific narrative around this drug shifts, which affects how rational the combinations look [1].