Increased Pathway Complexity Is a Prognostic Biomarker in Metastatic Castration-Resistant Prostate Cancer

Abstract

Metastatic castration-resistant prostate cancer (mCRPC) is a heterogeneous disease, characterized by common and rare driver gene alterations that provide a selective growth advantage for progressing tumour cells. We hypothesized that the number of distinct gene driver alteration-affected pathways or gene classes was associated with poor prognosis in patients initiating androgen receptor signalling inhibitors (ARSi). We performed a post hoc analysis of an amalgamated baseline circulating tumour DNA (ctDNA) mutational landscape dataset of ARSi-treated men with mCRPC (n = 342). We associated the detected hotspot, pathogenic, and/or high impact protein function-affecting perturbations in 39 genes into 13 pathways. Progression-free (PFS) and overall survival (OS) were analysed using Kaplan-Meier curves and multivariate Cox regression models. Driver gene alterations were detected in 192/342 (56.1%) evaluable patients. An increased number of affected pathways, coined pathway complexity index (PCI), resulted in a decremental PFS and OS, and was independently associated with prognosis once ≥3 pathway or gene classes were affected (PFS HR (95%CI): 1.7 (1.02-2.84), p = 0.04, and OS HR (95%CI): 2.5 (1.06-5.71), p = 0.04). Additionally, visceral disease and baseline PSA and plasma ctDNA levels were independently associated with poor prognosis. Elevated PCI is associated with poor ARSi outcome and supports comprehensive genomic profiling to better infer mCRPC prognosis.

Keywords: biomarker; cfDNA; mCRPC.

Figure 1

CONSORT diagram of patients with metastatic castration-resistant prostate cancer (mCRPC) treated with androgen receptor signalling inhibitors (ARSi). A total of 342 patients with mCRPC who underwent cell-free DNA profiling received ARSi (i.e., abiraterone acetate or enzalutamide). Five patients were excluded from downstream analysis as they were positive for the microsatellite instability (MSI) or hypermutator genotype. VANC, Vancouver Prostate Centre cohort; CORE/PROBIO, Centre for Oncological Research and Prostate Biomarker cohort; PCI, pathway complexity index.

Figure 2

Prevalence of genomic alterations in baseline plasma ctDNA samples from patients with metastatic castration-resistant prostate cancer-initiating abiraterone or enzalutamide (n = 342). (A) Gene-level alteration frequencies, colour-filled according to perturbation type. (B) Pathway or gene class-level alteration frequency. Abbreviations: AMP, amplification; DEL, deletion; SNV/INDEL, single nucleotide variation, insertion or deletion; GSR, genomic structural rearrangement.

Figure 3

Increasing pathway complexity is associated with poor outcomes to abiraterone or enzalutamide in metastatic castration-resistant prostate cancer. Kaplan–Meier analysis of progression-free (A) and overall survival (C), stratified according to the pathway complexity index (PCI) at baseline. p-value is calculated via the log-rank test. Multivariate Cox regression analysis (hazard ratio (confidence interval)) of progression-free (B) and overall (D) survival using baseline characteristics and pathway complexity index (using a PCI = 0 as reference). Within the multivariate cox regression model, serum PSA and plasma ctDNA levels were incorporated and modelled as continuous variables using restricted cubic spline models. The correlation between increasing levels of serum PSA or plasma ctDNA and time-to-event hazard ratio’s (for progression-free (PFS) and overall survival (OS)) is graphically presented in Panel B and D (right), using the median PSA (36 ng/mL) and ctDNA fraction (6%) as reference points (i.e., HR = 1). The observed positive correlation was tested for significance using the Wald test (both p < 0.001). Abbreviations: ARSi, androgen receptor signalling inhibitors; CT, chemotherapy; ctDNA, circulating tumour DNA fraction; PSA, prostate-specific antigen. p-values calculated via Wald test.