PolyPEPI1018

PolyPEPI1018 is an off-the-shelf therapeutic cancer vaccine from Treos Bio: six long (30-mer) peptides encoding 12 epitopes drawn from seven colorectal-cancer-associated antigens, formulated with Montanide ISA51VG adjuvant [1]. It is being tested in combination with Roche's PD-L1 inhibitor atezolizumab in microsatellite-stable metastatic colorectal cancer (MSS mCRC), the roughly 95% of colorectal cancers that don't respond to checkpoint inhibitors alone [1]. The Phase 2 trial (NCT05243862, n=18) recently transitioned from active-not-recruiting to completed on ClinicalTrials.gov, so a readout should be imminent or already at hand [2]. The commercial pitch: convert cold tumors to hot ones in a large indication where every prior immunotherapy attempt has failed.

Novel investigational biologic, no approvals anywhere. The lead Phase 2 study (NCT05243862, OBERTO-301) evaluates PolyPEPI1018 plus atezolizumab in previously treated MSS mCRC and has now completed at a target enrollment of 18 patients [2]. Two supporting Phase 1 studies wrapped earlier: OBERTO-101 (NCT03391232), an 11-patient monotherapy-plus-maintenance study run by Treos Bio, and OBERTO-201 (NCT05130060), a 15-patient investigator-sponsored combination with TAS-102 (trifluridine/tipiracil, an oral chemotherapy used in refractory CRC) at Mayo Clinic [3][4]. Published Phase Ib data from Hubbard et al. in Clinical Cancer Research 2022 (OBERTO-101, n=11) documented vaccine-specific CD8 T cell responses in 9 of 10 evaluable patients (90%) by IVS ELISpot, and reported encouraging survival signals when the vaccine was added to standard maintenance therapy [1]. No breakthrough, fast track, orphan, or RMAT designations are publicly disclosed. Treos Bio is a small privately held company (UK/Hungary), so investor visibility depends on conference presentations rather than earnings calls. The next catalyst is a Phase 2 data drop, plausibly at ESMO GI, ASCO GI, or AACR in the coming quarters. Expected readout for the combination arm is Q4 2026 or early 2027 based on trial completion timing, but no scheduled presentation has been announced.

The immune system normally hunts cells displaying abnormal proteins. Cancer cells often overproduce tumor-associated antigens (TAAs), proteins that healthy cells barely make. A therapeutic cancer vaccine injects short fragments (peptides) of these antigens together with an immune adjuvant, training killer T cells to recognize and destroy any cell displaying them. PolyPEPI1018 is built as six 30-mer peptides that together encode 12 epitopes drawn from seven TAAs frequently expressed in colorectal cancer: EPCAM, SURVIVIN, TSP50, FBXO39, SPAG9, CAGE1, and MAGE-A8 [1]. The peptides are emulsified with Montanide ISA51VG, a water-in-oil adjuvant that creates a depot at the injection site and recruits antigen-presenting cells [1]. The central scientific tension for any off-the-shelf peptide vaccine is HLA restriction. T cells recognize peptides only when those peptides are presented on the patient's own HLA (human leukocyte antigen) molecules, and HLA type varies substantially across individuals. A vaccine peptide that binds one patient's HLA may be invisible to another patient's immune system. Treos Bio addresses this by choosing long 30-mer peptides that internally contain multiple short epitopes predicted to bind across many HLA alleles. In the in-silico cohort reported in Hubbard et al., 98% of subjects were predicted to have at least one vaccine-specific HLA class I binding epitope (PEPI) and 91% were predicted to have two or more [1]. Whether that predicted coverage translates to clinical benefit across HLA backgrounds is what the Phase 2 is testing. The bet in MSS mCRC is specific: these tumors have few mutations, few neoantigens, and few tumor-infiltrating lymphocytes, so the immune system never mounts a response in the first place. That is why anti-PD-1 monotherapy (pembrolizumab, nivolumab) works in the ~5% of CRC that is microsatellite-instability-high but does essentially nothing in the MSS majority. A vaccine can, in principle, manufacture the T cell response the tumor failed to provoke. Adding atezolizumab, which blocks PD-L1 and takes the brake off T cells inside the tumor, is meant to keep those vaccine-primed T cells functional once they arrive [1]. Mechanism validation is thin. Sipuleucel-T (prostate) is the only approved therapeutic cancer vaccine for a solid tumor and was a commercial disappointment. The peptide vaccine graveyard includes tecemotide, GVAX, and PROSTVAC. The Phase Ib CD8 response rate and survival signal in Hubbard et al. are the strongest positive datapoints for this specific asset, but they came from a small, single-arm study [1].

NCT05243862 (OBERTO-301) is a Phase 2, single-arm, open-label study of PolyPEPI1018 plus atezolizumab in patients with MSS mCRC who progressed on standard chemotherapy [2]. Enrollment target was 18 patients, a tell that this is a signal-seeking screen rather than a registrational study. The primary endpoint is the incidence and severity of treatment-related adverse events, with clinical activity (objective response rate or ORR, the percentage of patients whose tumor shrinks measurably by RECIST criteria; progression-free survival or PFS, the time until disease worsens or death; and overall survival) and immune correlates as secondary endpoints [2]. Sponsor is Treos Bio Limited. The trial has transitioned to completed status per ClinicalTrials.gov, so top-line safety and preliminary efficacy should surface soon [2]. Design caveats are honest. A single-arm 18-patient study can only produce hypothesis-level efficacy data. There is no comparator, so any response numbers will be judged against historical controls, which for third-line MSS mCRC are brutal: ORR to checkpoint monotherapy is near zero, and median overall survival is roughly 6 to 8 months on regorafenib or trifluridine/tipiracil (both oral small-molecule drugs approved for late-line CRC). That low bar cuts both ways. A durable ORR anywhere in the 15% to 20% range would be genuinely interesting for a vaccine in this population. Anything less than that, or responses that fizzle within a few months, will fit the historical pattern of cold-tumor vaccine attempts and be hard to distinguish from noise given the sample size.

Our model estimates a 6% chance this drug is eventually approved. It starts from the historical base rate for Phase 2 drugs in this area (about 13%), then adjusts using ten facts about the trial and sponsor. What moves the number most: it is helped by a non-randomized design; it is held back by the sponsor's thin or weak approval record, smaller-than-typical enrollment for this phase, and weak or limited earlier-phase results. The other facts land near average for this stage, so they leave the estimate roughly where the base rate put it.

Efficacy risk dominates. The central question in MSS mCRC is whether any vaccine plus checkpoint combination can turn cold tumors hot durably. Bintrafusp alfa (a TGF-beta-trap/PD-L1 fusion, program discontinued by Merck KGaA and GSK in 2021 after Phase 3 failures in NSCLC and biliary tract cancer [6]), multiple neoantigen vaccines, and various TLR agonists (Toll-like-receptor agonists, a class of injected immune-stimulating molecules that mimic bacterial or viral signals to prime innate immunity) have tried and mostly failed. Even a positive Phase 2 signal at n=18 is fragile and would need replication in a controlled Phase 3 against regorafenib, trifluridine/tipiracil, or fruquintinib (all oral small-molecule drugs approved for late-line CRC) as the standard-of-care comparator. Safety risk is modest. Peptide vaccines are generally well-tolerated on their own; the main local reactions come from the Montanide ISA51VG adjuvant depot. The atezolizumab combination adds standard checkpoint inhibitor toxicity: immune-mediated colitis (inflammation of the colon), pneumonitis (inflammation of the lungs), hepatitis (inflammation of the liver), and endocrinopathies (thyroid, adrenal, and pituitary dysfunction) at low single-digit percentages, all well-characterized from the PD-L1 label. No mechanism-based cardiac or neurologic flags for this class. Execution risk is real and probably the dominant risk. Treos Bio is a small privately held biotech: Tracxn and Crunchbase list total funding of roughly $47M across a Series A (~$8M) and Series B (~$14M), with named investors including Luminous Ventures, Isabel Fox, Future Fund (UK), Outsized Ventures, and Proxenia Venture Partners, plus non-dilutive grant and clinical-collaboration support [7]. Current runway is not publicly disclosed. That capital base is nowhere near what is needed to run a Phase 3 in MSS mCRC solo, so a registrational program almost certainly requires a partnership or acquisition, and deal terms for an unpartnered small-cap vaccine developer in a graveyard indication are typically modest. Commercial risk is meaningful even on approval. MSS mCRC late-line care runs on cheap generics and oral small molecules. A vaccine plus checkpoint regimen would need to demonstrate clear overall survival benefit against a defined comparator to justify combination pricing, and payer pushback on cancer immunotherapy combinations is already a known headwind.

Worth watching, cheaply. The specific signal that would matter: a confirmed objective response rate above roughly 15% in the Phase 2 readout, or clear PFS/OS separation from historical MSS mCRC third-line controls, paired with translational data showing vaccine-induced T cell infiltration into tumors [1]. Either result would stand out in a cancer vaccine space defined by disappointments. Check back for abstract titles at ESMO 2026, ASCO GI 2027, and AACR 2027 with Treos Bio or Hubbard listed as lead. The Mayo Clinic Phase 1b OBERTO-201 with TAS-102 (NCT05130060) is a secondary datapoint that could arrive on a similar timeline and would test whether the vaccine adds anything to a chemo backbone rather than a checkpoint backbone [3]. If the Phase 2 data is underwhelming, this drops into the same bucket as PROSTVAC and tecemotide and a big-pharma partnership becomes unlikely. If it works, the natural home is Roche, given the atezolizumab combination, though Merck (KEYTRUDA franchise defense in CRC) or BMS (nivolumab) could compete for the asset. Watch also for how Treos Bio talks about funding runway at any presentation; a small private company with roughly $47M raised needs a partnership or a bridge before a Phase 3 conversation is credible. Base case: incremental data, no partnership, extended runway questions for Treos Bio. Upside case: a durable response signal that reframes the cold-tumor vaccine thesis and triggers a licensing deal.