Genomic Portraits: Reflections into a Tumor's Response to Therapy

In this issue of Clinical Cancer Research, Menssouri and colleagues (1) provide updates on the prospective MATCH-R trial that studied the evolution of clonal architecture of tumors from patients with metastatic castration-resistant prostate cancer (mCRPC) who had undergone treatment with molecular targeted agents such as the androgen receptor signaling inhibitors (ARSI), enzalutamide and abiraterone acetate. This ongoing trial planned to accrue 1,500 patients with the primary clinical endpoints studying the molecular and phenotypic type characteristics of tumors from patients who became resistant to ARSIs as well as assessing the frequency of molecular alterations when compared with pretreatment samples. Patients must have had evidence of radiographic disease progression after 6 months of therapy followed by tissue biopsy; PSA progression was insufficient. Molecular profiling of tumors consisted of whole-exome sequencing (WES), RNA sequencing (RNA-seq), and panel sequencing. A previous report from 2020 from this group (2) accrued 303 cases (91%). Gleaned from these biopsies, were 278 (83%) samples that had sufficient quality for analysis by high-throughput next-generation sequencing (NGS). All 278 samples underwent targeted NGS, 215 (70.9%) RNA-seq, and 222 (73.2%) WES. Samples were of sufficient quality to enable the generation of patient-derived xenografts as well as transfer of tissue implants into nude mice or NSG mice for translational evaluation (Fig. 1; ref. 2).

While numerous studies have suggested various mechanisms of resistance to these agents (3), reports have been retrospective and based on tissue samples that may have been suboptimally preserved or too old for reproducible analyses. Among the reports, mechanisms of resistance to abiraterone have included overexpression of CYP17A1 (4), activation of the 5α-dione pathway (5, 6), the androgen splice variant (ARV7; ref. 7), as well as genomic alternations such as ATM, AR, BRCA2, and TP53 (8), and increased ERbB2 signaling (9). Similarly, investigation of resistance of enzalutamide was equally challenging although more varied and included activation of the Wnt pathway (10), phenotypic switch (11), and interestingly, NOTCH and HMGCR activation (12) in addition to the androgen receptor (AR) mutations and other splice variants seen with abiraterone and similar agents (13, 14).

In the current article (1), the authors conducted a prospective trial in patients with mCRPC for whom pretreatment tissue biopsies were feasible. Although there may be mechanistic differences between primary and acquired resistance, the authors’ main hypothesis was that interrogation of the genomic landscape may lead to determinants of primary and acquired resistance. Paired biopsies at the time of resistance were paramount to addressing these questions. The tumor specimens were meticulously controlled for analysis by significant annotation and histologic review that allowed for quality control. WES of both tumor and germline DNA were determined via the Illumina NextSeq500 and hiSwq2000 systems with RNA-seq performed on the NextSeq500 platform. Transcriptomic analyses were also performed with special focus on the AR and neuroendocrine transformations. The current tissue analyses at baseline pretreatment with ARSIs with enzalutamide used by the majority of the patients and included 44.1% prostate, 40.7% nodal disease with 5% were bone, lung and liver, respectively. Overall, the frequency of genomic alterations was similar to those reported in other studies (15, 16). For those patients with primary resistance, the number of clonal or subclonal mutations did not differ between responders and non-responders; driver genes such as AR or TP53 did not have statistical relevance. AR-V7 expression did not appear to be associated with primary resistance. Interestingly, there was increased AR and NOTCH pathway activity in patients who responded to these agents, whereas non-responders developed activation in the Hedgehog pathway. While immune hallmark pathways were activated in non-responders, the epithelial-to-mesenchymal (EMT) transition and IL6/JAK/STAT pathways appeared highly active. Data suggested that low AR activity and EMT-like phenotype were highly associated with ARSI resistance with confirmation of low AR activity associated with EMT and stemness appeared consistent with de novo enzalutamide resistance. Failure of responses in a patient with germline BRCA2 to multiple ARSIs including PARP inhibitors indicated the emergence of different clones including one with a BRCA2 reversion mutation and one leading to the development of NEPC.

How does these data change how we decide on the right treatment for the right genomic profile of the disease and can it reliably be used in painting a “genomic portrait” for every patient to guide therapy? The authors have been able to prospectively perform genomic analyses across a spectrum of analytic platforms that have been able to provide some insight into the changes in the mutational and behavioral landscape of mCRPC. While there are limitations to most studies, the development of metastatic disease is not always to a single-site; the multi-clonality of the disease makes it difficult for single-agent therapy to address all the genomic and pathway alternations described. The authors have confirmed the works of others regarding the observation of low AR activity as being indicative of low responders to ARSIs. However, the emergence of the sonic Hedgehog (SHH) signaling pathway as a means of resistance is of interest. For example, clonal stem cells as noted by cell makers CD133 and CD44 are glycoproteins that are associated with stemness properties and are under SHH signaling for their regulation. As such, aberrant SHH pathway activation may lead to the development of specific stem cell lineages with their ultimate transformation into a more aggressively behaving phenotype. The authors indicate a profile for treatment citing that another pathway, NOTCH may have implications for disease response. Patients with low SHH, high AR, and high NOTCH were associated with clinical benefit. The converse, those with increased mesenchymal activity, high SHH, low NOTCH, and low AR may derive benefit from SHH targeted therapy albeit early trials with such agents were unsuccessful. However, given the increasing number of agents to which patients become resistant, it may not be unreasonable to investigate second- or third-generation SHH or NOTCH inhibitors. While there is no one specific genomic/pathway profile that can definitively point to a therapeutic treatment, nevertheless, this study provides high-quality annotated material on which some measure of treatment insight may be based. This study, in addition to others, provides real-world evidence for the importance of obtaining tissue samples throughout the various clinical states of disease transition, and the need to continue these efforts to ultimately generate a dynamic patient profile for treatments throughout the course of the disease. One caveat is that not all clones are created equal and therefore how to reconcile the different metastatic disease behavior in which to garner the appropriate treatment that addresses all sites remains unclear. This is where the benefit of combination intensification therapies may prove to be a better approach not only in the early de novo metastatic hormone-sensitive state but clearly in the metastatic castration-resistant state as well.

No disclosures were reported.