Longitudinal mutational analysis of TP53 in plasma circulating tumor DNA (ctDNA) in patients (pts) with solid tumors in a phase I study of BI 907828, an MDM2–p53 antagonist.

Abstract

3036 Background: Inactivation of p53 can occur due to TP53 mutations or downregulation of wild-type p53 by its primary negative regulator, MDM2. Targeting the MDM2–p53 interaction may therefore restore p53 function. BI 907828, a highly potent MDM2–p53 antagonist, is being evaluated in a phase Ia/b study in pts with advanced solid tumors (NCT03449381). During dose escalation (phase Ia), BI 907828 demonstrated a manageable safety profile and early signs of efficacy. Here we present data from a longitudinal mutational analysis of TP53 using ctDNA. The objective of the analysis was to identify if mutations in TP53 can be associated with possible acquired resistance to BI 907828. Methods: Collection of blood samples for ctDNA analyses was optional and sampled longitudinally from pts at baseline and every cycle until end of treatment (EoT). ctDNA was purified from plasma and analyzed using tumor-specific next generation sequencing (NGS; custom-made 8-gene panel using KAPA HyperCap technology, including TP53) to identify tumor-derived somatic mutations that may be relevant to understanding resistance mechanisms to BI 907828. The limit of detection was determined to be 0.5% mutant allele frequency; common polymorphisms were filtered out. Results: Plasma samples for ctDNA analysis were available from 44/54 (81%) pts enrolled to phase Ia. Baseline plasma samples from 26/54 (48%) pts and EoT samples from 41/54 (76%) pts were analyzed. Mutations were found in 6/26 (23%) samples at baseline. In 2 pt samples, mutations were found in both tumor and ctDNA, but concordant results were limited to 1 pt. At EoT, no (0) mutations were found in 16/41 (39%) samples, 1 mutation in 11/41 (27%), 2 mutations in 6/41 (15%), 3 mutations in 2/41 (5%) and 4 mutations in 6/41 (15%). No sample had more than 4 mutations. The most frequent mutation at EoT (in 9/41 [22%] samples) was R175H, a known p53 loss of function (LOF) mutation, followed by R248Q/W LOF (in 6/41 [15%) samples). Mutations in non- TP53 genes were detectable independent of TP53 mutation status, suggesting ctDNA was present in all samples. No clear association of induction of mutations could be identified for any BI 907828 dose or dosing schedule, treatment duration, or tumor type. 9/20 (45%) pts with no detectable mutations at baseline maintained wild-type TP53 status at EoT. Signs of efficacy were observed in pts with acquired TP53 LOF mutations, suggesting that these mutations have no effect on efficacy. Analysis of TP53 mutational status at EoT in conjunction with correlative clinical efficacy will be presented. Conclusions: This mutation analysis represents one of the most comprehensive assessments of longitudinal ctDNA by NGS for TP53 from a clinical trial of an MDM2–p53 antagonist. Preliminary data suggest that BI 907828 does not systematically lead to broad acquisition of resistance by inducing alterations in TP53. Clinical trial information: NCT03449381 .