The phosphatidylserine targeting antibody bavituximab plus pembrolizumab in unresectable hepatocellular carcinoma: a phase 2 trial

Abstract

Immune checkpoint inhibitors targeting PD-1/L1 have modest efficacy in hepatocellular carcinoma as single agents. Targeting membranous phosphatidylserine may induce pro-inflammatory and -immune stimulating effects that enhance immunotherapy activity. This hypothesis was tested in a single-arm phase 2 trial evaluating frontline bavituximab, a phosphatidylserine targeting antibody, plus pembrolizumab (anti-PD-1) in patients with unresectable hepatocellular carcinoma (NCT03519997). The primary endpoint was investigator-assessed objective response rate among evaluable patients, and secondary end points included progression-free survival, incidence of adverse events, overall survival, and duration of response. Among 28 evaluable patients, the confirmed response rate was 32.1%, which met the pre-specified endpoint, and the median progression-free survival was 6.3 months (95% CI, 1.3–11.3 months). Treatment related-adverse events of any grade occurred in 45.7% of patients, with grade 3 or greater adverse events in 14.3% of patients. Adverse events of any cause were observed in 33 patients (94.3%), with grade 3 or greater adverse events in 11 patients (31.4%). Prespecified exploratory analyses of baseline tumor specimens showed that a depletion of B cells, and the presence of fibrotic tissue and expression of immune checkpoints in stroma was associated with tumor response. These results suggest that targeting phosphatidylserine may lead to synergistic effects with PD-1 blockade without increasing toxicity rates, and future studies on this therapeutic strategy may be guided by biomarkers characterizing the pre-treatment tumor microenvironment.

Introduction

Hepatocellular carcinoma (HCC) is the most common liver cancer and the fourth most frequent cause of cancer-related deaths worldwide¹. Given a high recurrence rate among patients who undergo resection and a high frequency of advanced stage disease at the time of diagnosis, effective systemic treatments for HCC are needed to combat its growing burden of disease. While tyrosine kinase inhibitors targeting VEGF signaling were previously the mainstay of HCC therapies, immune checkpoint inhibitor combinations have emerged as the preferred frontline treatments for advanced HCC based on superior anti-tumor effects and meaningful survival benefits. In the phase 3 IMbrave150 study, atezolizumab (anti-PD-L1) plus bevacizumab in the first line setting demonstrated an unprecedented objective response rate (ORR) of 30% and overall survival (OS) benefit versus sorafenib². Similarly, the phase 3 HIMALAYA trial showed that durvalumab (anti-PD-L1) plus tremelimumab (anti-CTLA4) was superior to sorafenib in both OS benefit and ORR (20.1% versus 5.1%)³. These studies demonstrate the clinical benefit of rationally designed combination immunotherapy regimens in a subset of patients with HCC. However, an increased risk of specific toxicities associated with each of these regimens limits their utility. Specifically, atezolizumab plus bevacizumab is associated with a greater risk of variceal bleeding, cardiovascular events, and clinically significant proteinuria which is attributed to the on-target effects of bevacizumab². Tremelimumab plus durvalumab is associated with a greater risk of serious treatment-related toxicities compared to durvalumab alone (17.5% versus 8.2%)³. Thus, a treatment regimen with similar efficacy but an improved safety profile compared to approved immune checkpoint inhibitor combinations may have greater utility in patients with HCC who frequently have concurrent morbidities that may limit tolerance to adverse events.

Phosphatidylserine is a cytoplasmic-facing anionic phospholipid which has multiple essential functions in cellular membranes^4,5. Externalization of phosphatidylserine occurs as a result of apoptosis or cellular stress, but is also frequently observed in cancer cells and within stromal cells⁴. Exposed phosphatidylserine and phosphatidylserine dependent-receptors include the Tyro3, Axl, and Mer (TAM) family of negative immune regulators may facilitate cancer immune escape⁴. Preclinical evidence indicates that antibody-based targeting of phosphatidylserine promotes pro-inflammatory pathways, recruitment of tumoricidal macrophages, dendritic cell maturation, and potent T-cell immunity in multiple cancer models⁶. Bavituximab is a genetically engineered immunoglobulin gamma 1 (IgG1) chimeric (human/mouse) antibody that targets phosphatidylserine by binding to the phosphatidylserine-binding protein β2-glycoprotein 1 (β2GP1)⁶. While bavituximab has been demonstrated to be well tolerated as a single agent and in combination with chemotherapy, its use as a single agent in a phase I study of refractory advanced solid cancers and in combination with docetaxel in a phase 3 trial for advanced non-small-cell lung cancer did not show improved clinical benefit^7,8,9. In the frontline setting for HCC, bavituximab plus sorafenib was also not associated with a meaningful increase in ORR (5.3%)¹⁰. Importantly, prior use of bavituximab was not guided by its recently discovered immunogenic mechanism of action. The pleiotropic pro-inflammatory and -immune stimulating effects of bavituximab on tumor cells, immune cells, and stroma suggests that it may enhance the activity of immune checkpoint inhibitors. This is supported by preclinical evidence that targeting phosphatidylserine enhances either PD-1 or CTLA-4 blockade in multiple cancer models^11,12,13.

Pembrolizumab, an anti-PD-1 antibody, has modest anti-tumor activity in HCC. In the non-randomized phase 2 KEYNOTE-224 study, the ORR was 16% in patients previously treated with sorafenib (Cohort 1) and 16% in patients with no prior systemic therapy (Cohort 2)^14,15. In the randomized, double-blind, phase 3 KEYNOTE-240 study testing pembrolizumab versus placebo in the second line setting, OS and PFS did not reach statistical significance per specified criteria, but there was an improvement in ORR (18.3% versus 4.4%)¹⁶. In the randomized, double-blind, phase 3 KEYNOTE-394 study testing pembrolizumab plus best supportive care versus best supportive care in Asia, pembrolizumab was associated with an improved OS, PFS, and ORR (13.7% versus 1.3%)¹⁷. Whether the efficacy of PD-1 blockade can be augmented by targeting novel regulators of the innate immune system in HCC beyond inhibitors of VEGF or CTLA-4 remains unclear.

Here, we investigate whether bavituximab may enhance the efficacy of pembrolizumab in HCC in a multicenter single-arm two-stage phase 2 trial of bavituximab plus pembrolizumab for unresectable HCC in patients who had not received prior systemic therapies. The primary endpoint was investigator determined confirmed ORR among evaluable patients. Treatment with bavituximab plus pembrolizumab in this study met its prespecified endpoint, demonstrating a confirmed ORR of 32.1% (9 out of 28 evaluable patients). Treatment-related adverse events of any grade occurred in 45.7% of patients, with grade 3 or greater adverse events in 14.3% of patients. Depletion of B cells, intratumoral fibrosis, stromal immune checkpoint expression, and immune-rich inflammatory tumors were associated with tumor response.

Results

Patient enrollment and characteristics

From 25 June 2018 to 2 March 2022, 42 patients were screened for eligibility at a National Cancer Institute-designated cancer center (UT Southwestern Harold C. Simmons Comprehensive Cancer Center) and a county safety-net hospital system (Parkland Health), out of which 35 patients were enrolled (Supplementary Fig. 1). Enrolled patients who were not evaluable because they did not complete radiographic evaluation after starting treatment included 2 patients who withdrew consent, 1 patient who withdrew from the study after receiving radiation to a target lesion during their first cycle, 1 patient withdrawn following complications of an elective procedure, and 3 patients removed per investigator for worsening concomitant illness (1 patient who developed suspected immune-related colitis, 1 patient who developed hepatic decompensation during their first cycle, and 1 patient who developed non-islet cell hypoglycemia).

Based on a minimax two-stage design, 15 evaluable patients were enrolled in the first stage where the threshold to meet 3 or more confirmed objective responses was met allowing for the enrollment of an additional 13 evaluable patients in the second stage. Patients included in the primary analysis had a median age of 64 (IQR: 60–67), were primarily male (85.7%), frequently associated with HCV cirrhosis (71.4%), and frequently received prior locoregional therapy (60.7%) (Supplementary Table 1). These demographics are consistent with characteristics of patients with HCC in the broader US population^18,19. Due to the enrollment of patients across clinical sites serving diverse demographics, 50% of the evaluable patients were Black, and 7.1% were Hispanic.

Efficacy

Using RECIST 1.1 criteria, the investigator determined ORR among the 28 evaluable patients was 32.1%, including 2 complete response (CR) and 7 partial responses (PR), which met the prespecified primary endpoint of 8 or greater confirmed objective responses (Fig. 1a). The disease control rate including CRs, PRs, and stable disease (SD) was 64.3%. The median time to response was 2.1 months (range, 1.9–12.7 months), and the median duration of response was 13.3 months, with 4 responses ongoing at the time of data cutoff. The 6-month progression-free survival (PFS) rate was 57.1% (95% CI 38.8–75.4) and the median PFS was 6.3 months (95% CI, 1.3–11.3 months) (Fig. 1b). The median follow-up time was 28.5 months using the reverse Kaplan-Meier method. Noting that statistical power is limited given small sample sizes, subgroup analyses showed that objective tumor responses were not associated with clinical characteristics including age, race, sex, AFP levels, history of prior locoregional therapy, extrahepatic disease, or macrovascular invasion (Supplementary Fig. 2).

Safety

Patients who received at least one dose of drug were included in the safety analysis as prespecified in the trial protocol. Among 35 patients included in the safety analysis, adverse events of any cause were observed in 33 patients (94.3%) and serious adverse events of any cause were observed in 7 patients (20%). Adverse events of any cause that were grade 3 or greater occurred in 11 patients (31.4%). Treatment-related adverse events occurred in 16 patients (45.7%) and serious adverse events attributed to study treatment occurred in 4 patients (11.4%). Treatment-related adverse events that were grade 3 or greater occurred in 5 patients (14.3%). The most common treatment-related adverse event of any grade was diarrhea (17.1%), followed by rash (17.1%), alanine aminotransferase increase (11.4%), aspartate aminotransferase increase (11.4%), and chills (8.6%) (Table 1). Grade 3 and 4 treatment-related adverse events included diarrhea (5.7%), alanine aminotransferase increase (2.9%), aspartate aminotransferase increase (2.9%), and colitis (2.9%) (Table 1). Treatment interruptions due to adverse events occurred in 6 patients, and 1 patient discontinued therapy permanently for grade 4 diarrhea per the trial protocol for suspected high-grade immune-related toxicity.