Antitumor activity of crizotinib in lung cancers harboring a MET exon 14 alteration

Abstract

MET exon 14 alterations are oncogenic drivers of non-small-cell lung cancers (NSCLCs)¹. These alterations are associated with increased MET activity and preclinical sensitivity to MET inhibition². Crizotinib is a multikinase inhibitor with potent activity against MET³. The antitumor activity and safety of crizotinib were assessed in 69 patients with advanced NSCLCs harboring MET exon 14 alterations. Objective response rate was 32% (95% confidence interval (CI), 21-45) among 65 response-evaluable patients. Objective responses were observed independent of the molecular heterogeneity that characterizes these cancers and did not vary by splice-site region and mutation type of the MET exon 14 alteration, concurrent increased MET copy number or the detection of a MET exon 14 alteration in circulating tumor DNA. The median duration of response was 9.1 months (95% CI, 6.4-12.7). The median progression-free survival was 7.3 months (95% CI, 5.4-9.1). MET exon 14 alteration defines a molecular subgroup of NSCLCs for which MET inhibition with crizotinib is active. These results address an unmet need for targeted therapy in people with lung cancers with MET exon 14 alterations and adds to an expanding list of genomically driven therapies for oncogenic subsets of NSCLC.

Extended Data Fig. 1. Patient Evaluation Groups.

Shown is a flow diagram summarizing study enrollment and patient evaluation groups in the MET exon 14-altered advanced NSCLC expansion cohort of the ongoing PROFILE 1001 study as of January 31, 2018. Patients who received ≥1 dose of crizotinib were included in the safety population and analyses of PFS and OS. OS, overall survival; PFS, progression-free survival.

Extended Data Fig. 2. Best Percent Change in Target Lesions from Baseline in MET Exon 14-Altered NSCLCs and MET Exon 14 Alterations (Splice Site Region and Type) by Central and Local Testing.

A plot of the best response to crizotinib in 52 patients with MET exon 14-altered NSCLCs is shown. The bars indicate the best percentage change in the sum of target tumor measurements from baseline. Below the plot, retrospective central testing results for the MET exon 14 alteration splice site region (A) and mutation type (C) are depicted in relation to best response. For comparison, prospective local testing results for region (B) and type (D) are also included. In rows A and B, the splice acceptor region includes alterations in the splice acceptor region, polypyrimidine tract, and branching point. Cases classified as unknown include MET exon 14 alterations for which DNA coding region information was not available, such as alterations detected using an RNA-based assay. White space indicates that no results were reported by the central (rows A, C) or local (rows B, D) assay, or that the reported results could not be analyzed for the biomarker of interest. Retrospective central testing confirmed the presence of a MET exon 14 alteration in 88% of the 40 patients with tumor tissue analyzed. Central testing confirmation was not obtained for five patients with MET exon-14-altered NSCLC as determined by local testing methods. Of these five central testing-negative patients, four did not pass full quality control metrics (mostly attributed to low tumor purity or tumor input issues) but were reportable. One patient was determined to have MET exon 14-altered NSCLC by local testing and ROS1 rearranged NSCLC by central testing (as indicated by an asterisk).

Extended Data Fig. 3. Concurrent Alterations in Tumor

Shown are concurrent alterations observed by retrospective molecular profiling of archival tumor tissue (FoundationOne CDx). Each patient with available data is represented by a column. The colored rectangles above each column represent the best objective response to crizotinib. Within a column, each gene of interest for which a concurrent alteration is present is represented by a colored rectangle corresponding to the alteration type. Concurrent genomic alterations (average number of alterations per patient was 4.3 [range: 0 to 12]) were identified in tumor tissue from 35 of 40 (88%) patients with analyzable samples. MDM2 amplification was detected in non-responders, but not observed in responders. No notable response differences were seen in relation to absence or presence of TP53 mutation.

Extended Data Fig. 4. Progression-Free Survival by MET Exon 14 Alteration Detection in ctDNA.

The Kaplan-Meier curves for progression-free survival in patients treated with crizotinib are shown according to detection of MET Exon 14 alterations in ctDNA by plasma profiling. Progression-free survival was defined as the time from the date of the first dose of crizotinib to objective disease progression or death from any cause. Hash marks on the survival curve indicate censoring of data. *P-value from 2-sided log-rank test comparing survival distributions among ctDNA-positive versus ctDNA-negative patients. ^†HR from Cox proportional hazards regression – assuming proportional hazards, an HR >1 indicates a greater risk of disease progression or death among ctDNA-positive versus ctDNA-negative patients. ctDNA, circulating tumor DNA; HR, hazard ratio; NE, not estimable; PFS, progression-free survival.

Extended Data Fig. 5. Concurrent Alterations in ctDNA.

Shown are concurrent alterations observed by retrospective molecular profiling of baseline plasma samples (PlasmaSELECTR64). Each patient with available data is represented by a column. The colored rectangles above each column represent the best objective response to crizotinib. Within a column, each gene of interest for which a concurrent alteration is present is represented by a colored rectangle corresponding to the alteration type. Concurrent genomic alterations (average number of alterations per patient was 2.36 [range: 0 to 8]) were identified in ctDNA from 25 of 36 (69%) patients with analyzable samples. ctDNA, circulating tumor DNA.

Figure 1.. Best Percent Change in Target Lesions from Baseline in MET Exon 14-Altered NSCLCs.

A plot of the best response to crizotinib in 52 patients with MET exon 14-altered NSCLCs is shown. Cases not evaluable for response, with early death, and without measurable disease in the response-evaluable population of 65 patients were excluded from this plot. The bars indicate the best percentage change in the sum of target tumor measurements from baseline. As indicated by the asterisk, one patient was determined to have MET exon 14-altered NSCLC by local testing and ROS1-rearranged (and MET wild type) NSCLC by central testing. Below the plot, local testing results for the MET exon 14 alteration splice site region (A) mutation type (B), and the presence of concurrent increased (↑) MET copy number (C), as well as the presence of a MET exon 14 alteration in ctDNA, by central testing (D) are depicted in relation to best response. Note that while the two cases with increased MET copy number were called MET-amplified by local testing, none of these tumors had high-level amplification. In row A, the splice acceptor region includes alterations in the splice acceptor region, polypyrimidine tract, and branch point. Cases classified as unknown include MET exon 14 alterations for which DNA coding region information was not available, such as alterations detected using an RNA-based assay. In rows B and C, a white space indicates that no results were reported by the local assay, or that the reported results could not be analyzed for the biomarker of interest. In row D, a white space indicates that a plasma sample was not available for ctDNA testing, such as in patients who were treated prior to an amendment that included plasma ctDNA collections.

Figure 2.. Duration of Crizotinib Therapy in MET Exon 14-Altered NSCLCs.

A plot of the duration of crizotinib therapy in 65 response-evaluable patients with MET exon 14-altered NSCLCs is shown. To the left of the plot, local testing results for the MET exon 14 alteration splice site region (A) mutation type (B) and the presence of concurrent increased (↑) MET copy number (C), as well as the presence of a MET exon 14 alteration in ctDNA by central testing (D) are depicted in relation to the duration of therapy and best response. While the cases with increased MET copy number were called MET-amplified by local testing, none of these tumors had high-level amplification. Arrows indicate patients for whom crizotinib therapy was ongoing at the time of data cutoff. As indicated by the asterisk, one patient was determined to have MET exon 14-altered NSCLC by local testing and ROS1-rearranged NSCLC by central testing. In column A, the splice acceptor region includes alterations in the splice acceptor region, polypyrimidine tract, and branching point. Cases classified as unknown include MET exon 14 alterations for which DNA coding region information was not available, such as alterations detected using an RNA-based assay. In columns B and C, a white space indicates that no results were reported by the local assay, or that the reported results could not be analyzed for the biomarker of interest. In column D, a white space indicates that a plasma sample was not available for ctDNA testing, such as in patients who were treated prior to an amendment that included plasma ctDNA collections.

Figure 3.. Progression-Free Survival.

The Kaplan-Meier curve for progression-free survival in patients with MET exon 14-altered NSCLCs treated with crizotinib is shown. Progression-free survival was defined as the time from the date of the first dose of crizotinib to objective disease progression or death from any cause. The shaded area represents the 95% Hall–Wellner confidence bands. Hash marks on the survival curve indicate censoring of data. PFS, progression-free survival.