Ripretinib versus sunitinib in gastrointestinal stromal tumor: ctDNA biomarker analysis of the phase 3 INTRIGUE trial

Abstract

INTRIGUE was an open-label, phase 3 study in adult patients with advanced gastrointestinal stromal tumor who had disease progression on or intolerance to imatinib and who were randomized to once-daily ripretinib 150 mg or sunitinib 50 mg. In the primary analysis, progression-free survival (PFS) with ripretinib was not superior to sunitinib. In clinical and nonclinical studies, ripretinib and sunitinib have demonstrated differential activity based on the exon location of KIT mutations. Therefore, we hypothesized that mutational analysis using circulating tumor DNA (ctDNA) might provide further insight. In this exploratory analysis (N = 362), baseline peripheral whole blood was analyzed by a 74-gene ctDNA next-generation sequencing-based assay. ctDNA was detected in 280/362 (77%) samples with KIT mutations in 213/362 patients (59%). Imatinib-resistant mutations were found in the KIT ATP-binding pocket (exons 13/14) and activation loop (exons 17/18). Mutational subgroup assessment showed 2 mutually exclusive populations with differential treatment effects. Patients with only KIT exon 11 + 13/14 mutations (ripretinib, n = 21; sunitinib, n = 20) had better PFS with sunitinib versus ripretinib (median, 15.0 versus 4.0 months). Patients with only KIT exon 11 + 17/18 mutations (ripretinib, n = 27; sunitinib, n = 25) had better PFS with ripretinib versus sunitinib (median, 14.2 versus 1.5 months). The results of this exploratory analysis suggest ctDNA sequencing may improve the prediction of the efficacy of single-drug therapies and support further evaluation of ripretinib in patients with KIT exon 11 + 17/18 mutations. ClinicalTrials.gov identifier: NCT03673501.

Fig. 1. ctDNA analysis and detection.

Patients are included in multiple groups if they had more than one mutation; patients can have multiple mutations in the same exon. Groups under each of the categories (KIT exon 9, 11, 13/14 or 17/18) are not mutually exclusive, and patients may appear in more than one box. Bold indicates patients who were included in the analysis populations for the current manuscript. CNV, copy number variation; QC, quality control; R, ripretinib; S, sunitinib. ^actDNA detected only for SNV/INDEL; two patients had CNV-only mutations and were categorized as ctDNA not detected.

Fig. 2. Heterogeneity of ctDNA mutations in the KIT ATP-binding pocket (exons 13/14) and activation loop (exons 17/18).

This plot illustrates the number of mutations; each patient could have multiple mutations. The letters in the bubbles and in front of each listed codon represent amino acids. A, alanine; ATP, adenosine triphosphate; C, cysteine; D, aspartic acid; E, glutamic acid; F, phenylalanine; G, glycine; H, histidine; I, isoleucine; K, lysine; L, leucine; M, methionine; N, asparagine; P, proline; R, arginine; S, serine; T, threonine; V, valine; Y, tyrosine. ^aE640_L641delinsD. ^bRipretinib: R815_D816delinsN; sunitinib: R815_D816delinsK.

Fig. 3. Forest plot of PFS by mutational subgroup.

Data are represented as hazard ratio (HR) ± 95% confidence interval (CI). PFS was summarized using the Kaplan-Meier method with associated two-sided 95% CIs calculated using the Brookmeyer and Crowley method. HRs were obtained from the unstratified Cox proportional hazard model. Nine patients were included in multiple groups, including one patient with mutations in KIT exon 9 and KIT exon 11 and eight patients with mutations in KIT exon 11 and PDGFRA; the exon 11 + 13/14-only group excludes patients with mutations in KIT exons 9, 17 and 18; the exon 11 + 17/18-only group excludes patients with mutations in KIT exons 9, 13 and 14. Data cutoff: 1 September 2021.

Fig. 4. Kaplan-Meier analysis of PFS for patients treated with ripretinib or sunitinib in the KIT exon 11 + 13/14 (a) and KIT exon 11 + 17/18 (b) populations.

PFS was summarized using the Kaplan-Meier method with associated two-sided 95% CIs calculated using the Brookmeyer and Crowley method. HRs and P values were obtained from the unstratified Cox proportional hazard model and two-sided unstratified log-rank tests, respectively. Data cutoff: 1 September 2021. P values are nominal. NE, not estimable.

Fig. 5. Best percent change from baseline in the sum of target lesion diameters in patients treated with ripretinib or sunitinib.

a,b, Patients with KIT exon 11 + 13/14 mutations. c,d, Patients with KIT exon 11 + 17/18 mutations. Data cutoff: 1 September 2021. Dotted line at 20% represents the threshold for PD; dotted line at −30% represents threshold for PR. DOR, duration of response; NE, not evaluable; PD, progressive disease; PR, partial response; SD, stable disease.

Extended Data Fig. 1. Frequency of KIT mutations.

Includes both primary and secondary mutations. Only somatic KIT mutations considered; sunitinib arm includes 2 patients not treated. Total patients, N = 213. ctDNA, circulating tumor DNA.

Extended Data Fig. 2. Number of patients with KIT primary (A) and secondary resistance (B) mutations.

Data indicate number of patients with mutations at all affected codons; each patient can have multiple mutations. AA, amino acid.

Extended Data Fig. 3. Kaplan-Meier analysis of OS for patients treated with ripretinib or sunitinib in the KIT exon 11 + 13/14 (A) and KIT exon 11 + 17/18 (B) populations.

OS was summarized using the Kaplan-Meier method with associated 2-sided 95% CIs calculated using the Brookmeyer and Crowley method. HRs and P-values were obtained from the unstratified Cox proportional hazard model and 2-sided unstratified log-rank tests, respectively. Data cutoff: 1 September 2022. P-values are nominal. CI, confidence interval; HR, hazard ratio; NE, not estimable; OS, overall survival.