Adjuvant Immunotherapy Combination Shows Activity in AXL-Positive NSCLC

¹

“The clinical benefit seen with the drug combination in AXL-positive patients is impressive and provides a potential new treatment approach for patients with low or negative PD-L1 status,” said presenting author Matthew G. Krebs, MBChB, PhD. Krebs is a clinical senior lecturer in Experimental Cancer Medicine at the University of Manchester.

AXL kinase is a cell membrane receptor widely expressed in human cancers, can cause cancer treatment failure, and is regarded as a negative prognostic factor for many cancers including NSCLC.²

“[AXL] is associated with resistance to immunotherapy and [AXL] presents itself as a nice potential target to therapeutically tackle,” added Krebs.

Study Design

The phase II trial (NCT03184571) of patients with NSCLC comprises 3 cohorts of patients with NSCLC previously treated with one line of therapy. The results of cohort A were presented at SITC 2019. Investigators enrolled patients with platinum-refractory NSCLC and no prior exposure to a PD-1/PD-L1 inhibitor to receive a 400 mg loading dose of bemcentinib daily for 1 to 3 days, followed by 200 mg once daily thereafter. Patients also received 200 mg of pembrolizumab as an infusion once every 3 weeks.

The primary endpoint is overall response. Secondary endpoints include disease control rate (DCR), duration of response, progression-free survival (PFS), survival at 12 months and response by biomarker status.

A total of 50 patients received at least 1 dose of the combination. The median age was 65 years (range, 39-82) and a majority of patients were either ex-smokers (58%) or current smokers (20%).

A 50-50 split was observed in 30 patients for whom AXL composite score was available. “The AXL composite score is an assessment of the AXL expression in both the tumor and immune cell by immunohistochemistry,” explained Krebs. “This is different from just looking at AXL [expression] in the tumor itself, which is done historically and what we did initially in our analysis. It’s really important that we focus on AXL expression in both tumor and immune cells.”

Additionally, a majority of patients (n = 37) were either negative (55%) or low-expression of PD-L1 (24%). Only 10% of patients reported strong positive PD-L1 expression status.

Combination Promising for AXL-Positive Expression

The overall response rate (ORR) for the 44 responders was 25%, meeting the primary endpoint of the study (TABLE).¹Patients who were AXL-positive (n = 15) demonstrated a greater response versus patients who were AXL-negative (n = 15) with ORRs of 33% versus 7%, respectively. The DCR for the overall population, AXL-positive, and AXL-negative subgroups was 57%, 73%, and 40% respectively.

Median PFS for all patients (N = 50) was 4.1 months. When separated by AXL expression, a clear difference in benefit was observed in AXL-positive patients with median PFS of 8.4 months compared with 2.9 months in AXL-negative patients (P= .029).

“What we show here is [that the] majority of patients who responded were AXL-positive,” said Krebs. “The response isn’t being driven by PD-L1 status, the majority of these patients were either negative or low-PD-L1 expression…[patients who are] AXL-positive are known to have a poor prognosis, and they are having a particularly long PFS in this combination. It could be that AXL could be a predictive factor for selecting patients for treatment with this combination.”

Overall survival rates are still maturing for this cohort.

Regarding safety, the combination was well tolerated in the chemotherapy-refractory patient population. The most common all-grade adverse events (AEs) were increased transaminases (38%), asthenia/fatigue (30%), and diarrhea (24%). Reported grade 3 AEs were minimal with 7 patients (14%) reporting elevated transaminases, 4 (8%) reporting asthenia/fatigue, and 1 (2%) reporting anemia.

AEs leading to discontinuation of treatment included 1 case of Grade 4 pneumonia, and 2 cases of Grade 3 transaminases.

“Treatment related adverse events were considered to be less severe and better tolerated than for other TKI [tyrosine kinase inhibitors] combinations in NSCLC,” said Krebs.

In additional translational analysis, investigators conducted a differential gene analysis on 3 patients who showed the most benefit. The analysis showed that responders had increased expression ofAXL,TGFB1, myeloid activation, and EMT signatures.

“The most striking things about this looking at the signatures that were coming out and comparing them to the literature is that these are signatures associated with very poor prognosis and also signatures associated with lack of response to immunotherapy,” said Krebs. “Here we are showing that with our combination of an AXL-inhibitor and immunotherapy with pembrolizumab we are getting better responses.”

Krebs also noted that PD-L1 and IFNγ expression were not predictive of patient response.

Recruitment is ongoing in the second and third cohorts of the study. Cohort B will evaluate bemcentinib and pembrolizumab in patients with PD-1/PD-L1 inhibitor—refractory NSCLC and cohort C will evaluate patients who have received chemoimmunotherapy as first-line treatment.

TABLE. Efficacy Data by Biomarker Response¹

ORR

DCR

Overall population (N = 44)

25%

57%

AXL-positive (n = 15)

33%

73%

AXL-negative (n = 15)

7%

40%

PD-L1 strong positive; TPS ≥50% (n = 4)

25%

25%

PD-L1 positive; TPS 1%-49% (n = 13)

23%

62%

PD-L1 negative; TPS <1% (n = 20)

20%

50%

DCR indicates disease control rate; ORR, overall response rate; and TPS, tumor proportion score.

References

1. Krebs MG. A phase II study of bemcentinib (BGB324), a first-in-class selective AXL inhibitor, in combination with pembrolizumab in patients with advanced NSCLC: updated analysis. Presented at: 34th Annual Meeting of the Society for Immunotherapy of Cancer. November 6-10, 2019: National Harbor, Md.
2. Terry S, Abdou A, Engelsen AST, et al. AXL targeting overcomes human lung cancer cell resistance to nk- and ctl-mediated cytotoxicity.Cancer Immunol Res. 2019;7(11):1789-1802. doi: 10.1158/2326-6066.CIR-18-0903.