Bemcentinib Plus Pembrolizumab Combination Holds Promise in Clinical Trial of NSCLC

Matthew Krebs, MBChB, PhD

Matthew Krebs, MBChB, PhD

The first-in-class AXL inhibitor bemcentinib (BGB324) plus checkpoint inhibitor pembrolizumab (Keytruda) induced promising activity in patients with non—small cell lung cancer (NSCLC) regardless of PD-L1 status, according to a phase II clinical trial. The data also suggest a greater benefit with the combination in patients who are AXL-positive.

The primary end point, overall response rate (ORR), was met infindings from the first cohort of patients, which were presented at the Society for Immunotherapy of Cancer’s 34^thAnnual Meeting. In cohort A, patients had received prior platinum-based chemotherapy in the frontline setting and were immunotherapy-naïve. Of the 44 patients included in this analysis, the ORR was 25%. However, patients who were AXL-positive had an ORR of 33% versus 7% in those who were AXL-negative.

The disease control rate overall was 57% versus 73% in the AXL-positive group and 40% in the AXL-negative group. The median progression-free survival (PFS) was 4.1 months in the overall cohort compared with 8.4 months in the AXL-positive group and 2.9 months in the AXL-negative group. Overall survival data were still maturing at the time of data cutoff.

The most common any grade adverse events included transaminases (38%), asthenia/fatigue (30%), and diarrhea (24%). The regimen was generally well-tolerated among patients, and grade 3 toxicities were minimal, which included 7 patients (14%) with elevated transaminases, 4 (8%) with asthenia/fatigue, and 1 (2%) with anemia.

In an interview withTargeted Oncology, Matthew Krebs, MBChB, PhD, a clinical senior lecturer in Experimental Cancer Medicine at the University of Manchester, discussed the findings for the clinical trial evaluating the AXL inhibitor plus pembrolizumab in patients with NSCLC. He highlighted the preclinical data that provided rationale for this trial.

TARGETED ONCOLOGY: Could you discuss the preclinical data for the AXL inhibitor?

Krebs:AXL drives EMT and resistance to cytotoxic T-cell mediated cell death, and if we give an AXL inhibitor preclinically, we can reduce the AXL, which makes the cells more sensitive to T-cell mediated killing. Therefore, it makes it an ideal partner for checkpoint inhibitors. We did in vivo models looking at bemcentinib with checkpoint inhibitors, and we see very good synergistic effects preclinically in lung cancer as well as other cancer types, which was the rationale for going on to complete this clinical trial.

TARGETED ONCOLOGY: What was the rationale for combiningbemcentinib with pembrolizumab specifically?

Krebs:We have a lot of good data from the KEYNOTE studies regarding pembrolizumab, particularly for patients with high PD-L1 status over 50%. We know they can have great responses and outcomes, and recent data of immunotherapy plus chemotherapy in the frontline has very beneficial effects across any PD-L1 status.

TARGETED ONCOLOGY: Could you shed some light on the background of your study?

Krebs:The preclinical supported going ahead with this combination, so we wanted to see what the effectiveness of the combination was in patients with NSCLC. We chose to go for a second-line setting, so patients who had initially failed frontline treatment with chemotherapy or were immunotherapy-naïve. We didn’t mind whether they were PD-L1-negative, -positive, -high, or -low. The majority of our patients in this study were PD-L1-low or -negative; typically, these patients don’t tend to respond well to pembrolizumab monotherapy.

The other key point is that AXL positivity is a poor prognostic biomarker, so it is important patient group to try and target with new treatment.

TARGETED ONCOLOGY: What were the methods of design?

Krebs:The study was a single-arm phase II study for patients with NSCLC who failed previous platinum-based chemotherapy. All patients received the combination regardless of their PD-L1 status. The study was performed in 2 stages. In the initial stage, we wanted to see if there were any responses, and if we saw a certain number of responses, we expanded into a total number of 50 patients. Ultimately, we met the end point for the study.

I should note we only presented data for patients from cohort A, which is a group of patients who received only platinum-based chemotherapy in the first-line setting and were immunotherapy-naïve. This study had a cohort B and C that are still ongoing. In cohort B, patients received checkpoint inhibitors as monotherapy in the first-line setting and subsequently failed; we gave the combination to see if we could reverse resistance. In cohort C, we had patients who had received chemotherapy and immunotherapy combinations in the first-line setting.

TARGETED ONCOLOGY: What were the findings from cohort A?

Krebs:For the cohort A patients, the ORR was 25% for the whole population, but we wanted to look for any differentiated response based on AXL expression. We looked at the AXL composite score, which is an assessment of the degree of AXL expression by immunohistochemistry on both the tumor and immune cells. Through the propriety scoring method, we could decide where patients were AXL-positive or -negative, and that showed a difference in terms of clinical benefit for those who were positive and those were negative.

The ORR was 25% for the whole population, but if you divide that between patients who were AXL-positive and -negative, the ORR was 33% for AXL-high and 7% for AXL-negative, so there was a 5-fold difference in the response rate. That was also brought out in PFS; the PFS for the whole population was 4.1 months, but if we divided that by the composite AXL score, those who were positive had a PFS of 8.4 months compared with 2.9 months for those who are AXC-negative. That is interesting because usually, those patients who are AXL-positive would be poor prognosis because the AXL is driving the EMT. We believe that with bemcentinib, we are reversing that process and making the tumors much more sensitive to the immunotherapy as we saw in the preclinical models.

One further thing we did was look for differential gene expression by RNA sequencing between those who are responding and those who are not. The signatures we came up with were fascinating. Those who were responding had these signatures of poor prognostic patients and those who typically would not respond to immunotherapy. Similar data have been reported by other groups as well. The fact that these patients are having such good benefit to our combination again gives credence to the idea that we are reversing that EMT process of making the immune environment more susceptible to the inhibitors.

TARGETED ONCOLOGY: What are the next steps for this combination?

Krebs:We want to see whether we see similar results in cohort B and cohort C. With those signatures that we picked up in the exploratory analysis that might be a predictive biomarker for response, we clearly want to validate that and see whether that might also be applicable in cohorts B and C. If that does pan out, we want to take this forward into larger studies potentially in the randomized setting and potentially even in the frontline setting as well.