**The Impact of Whole Genome Data on Therapeutic Decision-Making in Metastatic Prostate Cancer: A Retrospective Analysis**

**Megan Crumbaker 1,2,3,**†**, Eva K. F. Chan 1,2,**†**, Tingting Gong 1,4, Niall Corcoran 5,6,7, Weerachai Jaratlerdsiri 1, Ruth J. Lyons 1, Anne-Maree Haynes 1, Anna A. Kulidjian 8,9, Anton M. F. Kalsbeek 1, Desiree C. Petersen 10, Phillip D. Stricker 11, Christina A. M. Jamieson 12, Peter I. Croucher 1,13, Christopher M. Hovens 5,6,\*, Anthony M. Joshua 1,2,3,\* and Vanessa M. Hayes 1,2,4,\***


Received: 18 March 2020; Accepted: 28 April 2020; Published: 7 May 2020

**Abstract:** *Background*: While critical insights have been gained from evaluating the genomic landscape of metastatic prostate cancer, utilizing this information to inform personalized treatment is in its infancy. We performed a retrospective pilot study to assess the current impact of precision medicine for locally advanced and metastatic prostate adenocarcinoma and evaluate how genomic data could be harnessed to individualize treatment. *Methods*: Deep whole genome-sequencing was performed on 16 tumour-blood pairs from 13 prostate cancer patients; whole genome optical mapping was performed in a subset of 9 patients to further identify large structural variants. Tumour samples were derived from prostate, lymph nodes, bone and brain. *Results*: Most samples had acquired genomic alterations in multiple therapeutically relevant pathways, including DNA damage response (11/13 cases), PI3K (7/13), MAPK (10/13) and Wnt (9/13). Five patients had somatic copy number losses in genes that may indicate sensitivity to immunotherapy (*LRP1B, CDK12, MLH1)* and one patient had germline and somatic *BRCA2* alterations. *Conclusions*: Most cases, whether primary or metastatic, harboured therapeutically relevant alterations, including those associated with PARP

inhibitor sensitivity, immunotherapy sensitivity and resistance to androgen pathway targeting agents. The observed intra-patient heterogeneity and presence of genomic alterations in multiple growth pathways in individual cases suggests that a precision medicine model in prostate cancer needs to simultaneously incorporate multiple pathway-targeting agents. Our whole genome approach allowed for structural variant assessment in addition to the ability to rapidly reassess an individual's molecular landscape as knowledge of relevant biomarkers evolve. This retrospective oncological assessment highlights the genomic complexity of prostate cancer and the potential impact of assessing genomic data for an individual at any stage of the disease.

**Keywords:** prostate cancer; precision medicine; whole genome sequencing; optical mapping; therapy

#### **1. Introduction**

Worldwide, prostate cancer (PCa) is the most commonly diagnosed non-cutaneous cancer in men and a leading cause of cancer-related male deaths [1]. Treatment strategies range from observation alone to multi-modal treatment and vary based on clinical and pathological factors such as tumour stage (T Stage), prostate specific antigen (PSA) level, Gleason or International Society of Urological Pathology (ISUP) score and life expectancy. PCa is a heterogeneous disease, and these clinical factors alone cannot predict outcomes accurately. Early disease is potentially curable whereas eventual treatment resistance is an intractable problem in metastatic disease. In both early and advanced disease, escalation of treatment including combination therapies, such as androgen deprivation therapy (ADT) administered with docetaxel with or without radiotherapy to the primary has resulted in improved outcomes [2–5]. However, the escalated treatment comes at the cost of increased toxicity and only a subset of men garner benefit. As such, predictive biomarkers for optimal treatment selection are needed at all stages of the disease.

PCa progression is driven by genomic alterations and, as such, large sequencing efforts have focused on elucidating the succession of events driving its pathogenesis and progression. These sequencing efforts aim to establish new prognostic and therapeutic targets [6–11]. Thus far, these studies have focused on primary tumours from localized cancers and/or heavily pre-treated disease that has become resistant to ADT, termed castrate-resistant prostate cancer (CRPC). It has been established that intraand inter-patient heterogeneity is high [12–16], though certain critical events may occur early in some patients and propagate. In general, genomic changes are thought to accumulate in response to treatment as the disease progresses and the importance of structural variants (SVs) in advanced prostate cancer is an evolving area of research [10,11].

A goal of previous studies is to advance "precision medicine" in PCa. Robinson et al. found that 89% of CRPC samples harboured a clinically actionable genomic alteration [6]. However, clinical trials utilizing the precision medicine paradigm of selecting a targeted drug based on molecular criteria have yielded mixed results. For example, though phosphoinositide 3-kinase (PI3K)/Protein Kinase B (Akt) pathway activating alterations are commonly reported in PCa, PI3K inhibitors have demonstrated limited efficacy to date [17,18]. However, a randomized phase II study of abiraterone +/−ipatasertib, an Akt inhibitor, in metastatic CRPC did find improved antitumoral activity in the combination arm, particularly in men with *PTEN* loss [19]. Similarly, poly ADP ribose polymerase (PARP) inhibitors have shown promise in selected men with CRPC and homologous recombination deficiency [20–23] but not all mutations in the homologous recombination pathway predict a response [21].

Although these large genomic studies have expanded the knowledge of molecular drivers of treatment-naïve primary and metastatic CRPC, they have generally viewed the data as a cohort without looking at cumulative alterations and their potential therapeutic impact within the individuals. Likewise, hormone sensitive (HSPC) metastatic disease has also been largely neglected. In this study, we performed whole genome sequencing (WGS) on men with confirmed PCa in order to assess the collective genomic

eventsinindividual cases and theirimpact on real-world therapeutic decisions. Recognizing theimportance of SVs in prostate cancer and the limitations of WGS in detecting large genomic rearrangements, we also performed whole genome optical mapping (WGM) on a subset of the samples.
