Revisiting clinical trials using EGFR inhibitor-based regimens in patients with advanced non-small cell lung cancer: a retrospective analysis of an MD Anderson Cancer Center phase I population

Abstract

Purpose: Single-agent EGFR inhibitor therapy is effective mainly in patients with lung cancer and EGFR mutations. Treating patients who develop resistance, or who are insensitive from the outset, often because of resistant mutations, other aberrations or the lack of an EGFR mutation, probably requires rational combinations. We therefore investigated the outcome of EGFR inhibitor-based combination regimens in patients with heavily-pretreated non-small cell lung cancer (NSCLC) referred to a Phase I Clinic.

Methods: We reviewed the electronic records of patients with NSCLC treated with an EGFR inhibitor-based combination regimen: erlotinib and cetuximab; erlotinib, cetuximab and bevacizumab; erlotinib and dasatinib; erlotinib and bortezomib; or cetuximab and sirolimus.

Results: EGFR mutations were detected in 16% of patients (21/131). EGFR inhibitor-based combination regimens were administered to 15 patients with EGFR-mutant NSCLC and 24 with EGFR wild-type disease. Stable disease (SD) ≥6 months/partial remission (PR) was attained in 20% of EGFR-mutant patients (3/15; two with sensitive mutations and secondary resistance to prior erlotinib, and one with a resistant mutation), as well as 26% of evaluable patients (5/19) with wild-type disease. One of three evaluable patients with squamous cell histology achieved SD for 26.5 months (EGFR wild-type, TP53-mutant, regimen=erlotinib, cetuximab and bevacizumab).

Conclusions: Eight of 34 evaluable patients (24%) with advanced, refractory NSCLC evaluable for response achieved SD ≥6 months/PR (PR=3; SD ≥6 months=5) on EGFR inhibitor-based combination regimens (erlotinib, cetuximab; erlotinib, cetuximab and bevacizumab; and, erlotinib, bortezomib), including patients with secondary resistance to single-agent EGFR inhibitors, resistant mutations, wild-type disease, and, squamous histology.

Figure 1. 3-D Waterfall plot

Best response by RECIST, of 15 NSCLC patients with EGFR positive-mutations treated with an EGFR inhibitor-based regimen. Patients with clinical progression or with new metastases were graphed as 20% progression. Time to treatment failure in months is represented by solid lines and the arrow indicates that the patient was still on study when the data was censored. Patients with PIK3CA mutations in addition to EGFR mutation and patients who received prior EGFR inhibitor therapy are designated as such. Dotted horizontal line at -30% indicates border for partial response.

Figure 2. Computed tomography (CT) scans of a NSCLC patient (case #5, Table 2) with two sensitive EGFR mutations (L858R in exon 21 and G873E in exon 21) and a PIK3CA mutation (E542K in exon 9)

a) CT at baseline, and b) CT taken 5 months after treatment initiation with erlotinib/cetuximab/bevacizumab demonstrating a PR (-55%). Duration of response = 9+ months. Patient had received prior erlotinib for 14.3 months.

Figure 3. Kaplan-Meier estimates of time to treatment failure in 24 NSCLC patients with EGFR wild-type disease treated with EGFR inhibitor-based combination therapies

in phase I clinical trial program (3.3 months) vs. time to treatment failure on their last standard therapy (2.3 months; p=0.045; log-rank test)