Bavituximab

Bavituximab is a chimeric antibody (built from a mix of mouse and human protein components so it targets like a mouse antibody but is less likely to be rejected by the human immune system) that targets phosphatidylserine on tumor blood vessels and stressed cancer cells. It was originally developed by Peregrine Pharmaceuticals (which reincorporated as the CDMO Avid Bioservices in 2018) and now has active and completed Phase 2 combinations with pembrolizumab across head and neck, hepatocellular, and gastric cancers [1][2][3]. The head and neck trial (NCT04150900) is an investigator-initiated study at the University of Maryland that stalled enrollment at 7 patients, well short of a decision-grade signal [4]. Commercial rights were transferred from Peregrine/Avid to OncXerna Therapeutics in February 2018 [5]; any subsequent transfer to another rights holder is not confirmed in public filings.

Bavituximab is investigational everywhere; it has never been approved in any indication despite roughly two decades of clinical work that began with early Phase 1 and Phase 2 antiviral studies in hepatitis C and HCV/HIV coinfection in the late 2000s and early 2010s [6] before pivoting to oncology. It carries no active FDA breakthrough, fast track, orphan, or accelerated approval designation. The program's high-water mark and inflection was the SUNRISE Phase 3 in second-line non-squamous NSCLC (582 patients, bavituximab plus docetaxel vs. docetaxel plus placebo), which was halted for futility at a pre-specified interim analysis in February 2016 after the control arm dramatically outperformed overall survival expectations [7]. That failure effectively ended Peregrine's oncology ambitions. Peregrine reincorporated as Avid Bioservices in 2018 and exited drug development to focus on contract manufacturing; commercial rights to bavituximab were assigned to OncXerna Therapeutics in February 2018 under an asset purchase agreement (up to $95M in milestones plus mid-teens royalties to Avid) [5]. The current clinical footprint is a set of investigator-initiated Phase 2 combinations with pembrolizumab: NCT04099641 in gastric and gastroesophageal junction cancer (n=80, completed, biomarker-correlated results published in Curr Oncol 2026) [1], NCT03519997 in hepatocellular carcinoma (n=35, completed, Nat Commun 2024) [2], NCT03139916 in newly diagnosed glioblastoma (n=36, completed, Clin Cancer Res 2023) [3], and the head and neck study NCT04150900 (Active-Not-Recruiting at 7 patients) [4]. There is no announced Phase 3, no BLA activity, and no publicly disclosed pivotal readout timeline. The realistic near-term catalyst is licensing or a randomized Phase 2 restart in a biomarker-defined population, not a regulatory event.

Phosphatidylserine (PS) is a lipid that normally sits on the inner face of a cell's membrane. When cells are stressed, dying, or building new blood vessels the way tumors do, PS flips to the outside. Immune cells read exposed PS as an 'eat me quietly, don't inflame' signal, and tumors exploit that to keep the surrounding immune environment sedated [8]. Because PS is a lipid, antibodies cannot target it directly the way they target proteins; β2-glycoprotein I is a blood protein that binds spontaneously to exposed PS and creates a protein coat that bavituximab can actually grip. That indirect binding is a fundamental constraint of the mechanism, not a quirky detail, because the drug's activity depends on β2GPI being available and correctly oriented on the PS surface. Once bound, bavituximab rewrites the message from 'ignore' to 'attack' by engaging immune-cell Fc receptors and by depleting myeloid-derived suppressor cells (MDSCs), a class of immune cells that tumors recruit to shut down T-cell responses. In newly diagnosed glioblastoma patients treated with bavituximab plus radiation and temozolomide, Ly et al. showed measurable MDSC reduction in blood, the cleanest pharmacodynamic readout the drug has produced [3]. The mechanism linking PS coating to MDSC reduction is not fully characterized, however, and may be indirect (for example, mediated by vascular remodeling or Fc-receptor engagement on other myeloid cells) rather than direct depletion of PS-exposing MDSCs. The mechanistic thesis, that reversing PS-mediated immunosuppression should synergize with PD-1 blockade, is biologically coherent and supported by preclinical work. What the program still lacks after two decades is a validated patient-selection biomarker. Every drug that tunes an immunosuppressive knob in the tumor microenvironment has had to solve the 'helps a subset, blurs the average' problem, and bavituximab has not solved it in a randomized setting.

NCT04150900 is a single-arm, open-label Phase 2 combining bavituximab with pembrolizumab in recurrent or metastatic head and neck squamous cell carcinoma, sponsored by the University of Maryland Baltimore under protocol 1922GCCC [4]. Primary endpoint is investigator-assessed complete plus partial response rate. The trial is Active-Not-Recruiting at 7 enrolled patients, an order of magnitude below the typical Simon two-stage HNSCC combo target of 30 to 45 patients. (Simon two-stage is a standard Phase 2 design that checks for an early efficacy signal in a small first cohort before enrolling the full study population, letting sponsors abandon obvious failures quickly.) There is no comparator arm and no disclosed PD-L1 CPS enrichment strategy. CPS (Combined Positive Score) is a lab measurement of PD-L1 protein levels across both tumor cells and surrounding immune cells, used to predict which patients are likely to respond to checkpoint inhibitors. That absence is a problem for interpretability: pembrolizumab monotherapy already produces roughly 15 to 20 percent ORR in CPS-positive recurrent HNSCC, so any single-arm combo signal below about 30 percent is impossible to attribute to bavituximab. With n=7, the study cannot generate that resolution, and there is no BLA (Biologics License Application, the FDA submission required to approve a biologic drug) pathway visible from this trial. The more informative trials in the broader program are the completed gastric/GEJ study NCT04099641 (n=80, biomarker-correlated outcomes reported in Curr Oncol 2026) [1] and the HCC study NCT03519997 (n=35, ORR ~32 percent with a favorable signal in HBV-positive patients, Nat Commun 2024) [2]. The HBV-positive signal is biologically plausible because chronic HBV infection maintains a smoldering inflammatory environment in the liver that includes elevated PS exposure on stressed hepatocytes and infected cells; a drug that flips PS-mediated tolerance into activation has more substrate to work on in that context. This hypothesis needs prospective validation. Those two trials provide the real evidentiary basis for the PS-plus-PD-1 combination thesis; the HNSCC study anchored to this node does not.

Our model estimates a 5% 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 smaller-than-typical enrollment for this phase, the sponsor's thin or weak approval record, 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 largest randomized bavituximab dataset, SUNRISE in second-line non-squamous NSCLC, was halted for futility in February 2016 (582 patients, bavituximab plus docetaxel vs. docetaxel; the docetaxel-alone arm outperformed OS expectations), and no biomarker has since emerged to explain a responder subset [7]. Biomarker selection is absent from most current combo trials; the gastric Phase 2 published biomarker-correlated outcomes in Curr Oncol 2026, which is a directional positive but needs replication in a randomized design [1]. Execution risk is severe for NCT04150900 specifically: 7 enrolled against any reasonable Simon two-stage target means either the sponsor lost interest, funding dried up, or accrual competition from the crowded pembrolizumab combo shelf killed it. Competitive risk is severe and, from an investor's perspective, more important than intra-class rivalry (there are no other PS-targeting biologics in late-stage development). The real competition is the established IO-combination shelf: in HCC, lenvatinib plus pembrolizumab is FDA-approved on the KEYNOTE-524 label and atezolizumab plus bevacizumab (IMbrave150) is the first-line standard; in HNSCC, pembrolizumab monotherapy or pembro plus chemotherapy is the frontline standard from KEYNOTE-048 and nivolumab is approved in the second-line setting; in gastric, nivolumab plus chemotherapy is approved first-line (CheckMate-649) and ramucirumab plus pembrolizumab has active data. Bavituximab would need to show a clearly differentiated ORR or OS signal in a biomarker-selected population to earn a place on that shelf. Safety risk is comparatively low. Bavituximab has thousands of patient-years of exposure across prior trials with a manageable profile of infusion reactions and low-rate thromboembolic events; PS is exposed on activated platelets, which is the on-mechanism concern, but it has not produced a black-box safety signal. Commercial risk is existential. Avid Bioservices is now a CDMO, OncXerna is a small private company, and there is no partner with checkpoint-inhibitor commercial infrastructure. Even a positive readout would require a well-capitalized licensee to run a pivotal trial, and payer coverage for a novel biologic layered on top of pembrolizumab requires clear differentiation that the current data package does not deliver.

Noise, with a narrow caveat. The head and neck trial anchored to this node is functionally dormant at 7 patients and will not produce a decision-grade signal. What is worth tracking in the broader bavituximab program is whether the current rights holder (OncXerna as of the 2018 Avid asset purchase, or any subsequent licensee) takes the gastric/GEJ biomarker-correlated data [1] or the HCC HBV-positive signal [2] into a randomized Phase 2 or Phase 3 with prospective biomarker enrichment. That would be the first genuinely modern development plan the asset has produced in two decades. Check back if any of three things happen: a licensing or partnership announcement pairing the rights holder with a checkpoint-inhibitor commercial player; initiation of a randomized bavituximab-plus-pembrolizumab study in gastric, HCC, or HNSCC with biomarker enrichment as a primary hypothesis; or an FDA fast track or breakthrough designation. Absent any of those, this is a long-tail asset kept alive by academic investigators rather than a corporate program. The 2018 Peregrine-to-Avid pivot, in which the original developer effectively walked away from bavituximab and turned itself into a contract manufacturer, is the most honest signal about where private capital thought the drug was going. For portfolio purposes, treat pipeline-bavituximab as an option with very low delta, worth a quarterly glance and no active tracking.