Concurrent RB1 loss and BRCA-deficiency predicts enhanced immunological response and long-term survival in tubo-ovarian high-grade serous carcinoma

Abstract

Background: Somatic loss of the tumour suppressor RB1 is a common event in tubo-ovarian high-grade serous carcinoma (HGSC), which frequently co-occurs with alterations in homologous recombination DNA repair genes including BRCA1 and BRCA2 (BRCA). We examined whether tumour expression of RB1 was associated with survival across ovarian cancer histotypes (HGSC, endometrioid (ENOC), clear cell (CCOC), mucinous (MOC), low-grade serous carcinoma (LGSC)), and how co-occurrence of germline BRCA pathogenic variants and RB1 loss influences long-term survival in a large series of HGSC.

Patients and methods: RB1 protein expression patterns were classified by immunohistochemistry in epithelial ovarian carcinomas of 7436 patients from 20 studies participating in the Ovarian Tumor Tissue Analysis consortium and assessed for associations with overall survival (OS), accounting for patient age at diagnosis and FIGO stage. We examined RB1 expression and germline BRCA status in a subset of 1134 HGSC, and related genotype to survival, tumour infiltrating CD8+ lymphocyte counts and transcriptomic subtypes. Using CRISPR-Cas9, we deleted RB1 in HGSC cell lines with and without BRCA1 mutations to model co-loss with treatment response. We also performed genomic analyses on 126 primary HGSC to explore the molecular characteristics of concurrent homologous recombination deficiency and RB1 loss.

Results: RB1 protein loss was most frequent in HGSC (16.4%) and was highly correlated with RB1 mRNA expression. RB1 loss was associated with longer OS in HGSC (hazard ratio [HR] 0.74, 95% confidence interval [CI] 0.66-0.83, P = 6.8 ×10^-7), but with poorer prognosis in ENOC (HR 2.17, 95% CI 1.17-4.03, P = 0.0140). Germline BRCA mutations and RB1 loss co-occurred in HGSC (P < 0.0001). Patients with both RB1 loss and germline BRCA mutations had a superior OS (HR 0.38, 95% CI 0.25-0.58, P = 5.2 ×10^-6) compared to patients with either alteration alone, and their median OS was three times longer than non-carriers whose tumours retained RB1 expression (9.3 years vs. 3.1 years). Enhanced sensitivity to cisplatin (P < 0.01) and paclitaxel (P < 0.05) was seen in BRCA1 mutated cell lines with RB1 knockout. Among 126 patients with whole-genome and transcriptome sequence data, combined RB1 loss and genomic evidence of homologous recombination deficiency was correlated with transcriptional markers of enhanced interferon response, cell cycle deregulation, and reduced epithelial-mesenchymal transition in primary HGSC. CD8+ lymphocytes were most prevalent in BRCA-deficient HGSC with co-loss of RB1.

Conclusions: Co-occurrence of RB1 loss and BRCA mutation was associated with exceptionally long survival in patients with HGSC, potentially due to better treatment response and immune stimulation.

Figure 1.. Expression of RB1 and survival associations across ovarian cancer histotypes.

(A) Representative images of immunohistochemical detection of RB1 expression in ovarian carcinoma tissues, showing examples of the three most common expression patterns: retained, lost and subclonal loss. (B) Proportion of patients with loss or retention of RB1 protein expression in tumour samples by ovarian cancer histotypes. Chi-square P value reported for difference in proportions across all histotypes. HGSC, tubo-ovarian high-grade serous carcinoma; LGSC, low-grade serous carcinoma; MOC, mucinous ovarian cancer; ENOC, endometrioid ovarian cancer; CCOC, clear cell ovarian cancer. (C) Boxplots show RB1 mRNA expression (NanoString) by RB1 protein expression status; lines indicate median and whiskers show range (Mann-Whitney test P value reported). Kaplan-Meier analysis of overall survival in patients diagnosed with HGSC (D) and ENOC (E) stratified by tumour RB1 expression. (F) Loss of RB1 tumour expression is more common in germline BRCA1 and BRCA2 mutation carriers than retained RB1 expression. Chi-square P value is reported. (G) Kaplan-Meier estimates of overall survival in HGSC patients by combined germline BRCA and tumour RB1 expression status.

Figure 2.. Sensitivity to therapeutic agents in BRCA1-mutant cell lines with RB1 knockout.

(A) RB1 was knocked out using CRISPR/Cas9 in 3 patient-derived Australian Ovarian Cancer Study (AOCS) HGSC cell lines with either wild-type or mutant BRCA1 background. Representative Western Blots show protein levels of RB1 and phosphorylated RB1 (pRB1) compared to GAPDH loading control in single cell cloned, homozygous RB1 wildtype (WT) and knockout (KO) colonies in comparison to heterogeneous populations with a scramble single guide RNA (sgRNA). Independent blots were used for RB1 and pRB1. (B) Cell viability was compared between RB1 WT and KO clones following treatment with cisplatin (72 hours), paclitaxel (72 hours) or olaparib (120 hours). Nonlinear regression drug curves are shown; P values of a curve fit, extra sum-of squares F test (ns, not significant; ** P < 0.01; **** P < 0.0001; n = 3). Error bars indicate ± SEM; for some values error bars are shorter than the symbols and thus are not visible. (C) Proportion of surviving colonies following 16 days of treatment with cisplatin, paclitaxel or a combination of both (with half of the IC50 determined per drug and cell line respectively) relative to DMF vehicle control (n = 3 replicates). Data are presented as mean ± SEM. Mean values were compared by student’s t-test (ns, not significant; *P < 0.05; **P < 0.01). Representative scans of the fixed cell colonies stained with crystal violet are shown for each condition.

Figure 3.. Genomic landscape of high-grade serous ovarian tumours with co-occurring BRCA and RB1 alterations.

(A) Pathogenic germline and somatic alterations in homologous recombination (HR) and DNA repair genes detected by whole-genome sequencing and DNA methylation analysis of 126 primary HGSC samples are shown, as well as alterations in immune genes and CCNE1. Samples are grouped by HRD and RB1 status (wt, wild-type; mut, mutation). Bars at the top indicate the number of alterations in each listed gene per patient. Patients are annotated with survival group (LTS, long-term survivor, OS >10 years; MTS, mid-term survivor, OS 2–10 years; STS, short-term survivor, OS <2 years), tumour CHORD scores, and the proportion of structural variant (SV) type (DUP, duplication; DEL, deletion; INV, inversion; ITX, intra-chromosomal translocation). (B) Kaplan-Meier estimates of progression-free and overall survival of patients with according to HR status (BRCA1-type HRD, BRCA2-type HRD or homologous recombination proficient tumours) and RB1 status (mut, mutation; wt, wild-type).

Figure 4.. Characterisation of HGSC with co-loss of RB1 and BRCA.

(A) Gene set enrichment analysis indicating up- and downregulated pathways in tumours according to BRCA and RB1 status. HRP, homologous recombination proficient; HRD, homologous recombination deficient; RB1wt, RB1 wild-type; RB1m, RB1 altered. (B) Proportion of tumour infiltrating lymphocytes (TILs) in HGSC tumours grouped by RB1 expression and BRCA germline mutation status (Chi-square P value is indicated). (C) Proportion of tumours classified as each HGSC molecular subtype grouped by RB1 expression and BRCA germline mutation status (Chi-square P value is indicated; C5.PRO, C5/proliferative subtype; C4.DIF, C4/differentiated subtype; C2.IMM, C2/immunoreactive subtype; C1.MES, C1/mesenchymal subtype).