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

156

Somatic loss of the tumour suppressor RB1 is a common event in tubo-ovarian 157 high-grade serous carcinoma (HGSC), which frequently co -occurs with alterations in 158 homologous recombination DNA repair genes including BRCA1 and BRCA2 (BRCA). We 159 examined whether tumour expression of RB1 was associated with survival across ovarian 160 cancer histotypes (HGSC, endometrioid (ENOC), clear cell (CCOC), mucinous (MOC), low -161 grade serous carcinoma (LGSC)), and how co-occurrence of germline BRCA pathogenic 162 variants and RB1 loss influences long-term survival in a large series of HGSC. 163 Patients and m ethods: RB1 protein expression patterns were classified by 164 immunohistochemistry in epithelial ovarian carcinomas of 7436 patients from 20 studies 165 participating in the Ovarian Tumor Tissue Analysis consortium and assessed for associations 166 with overall survival (OS), accounting for patient age at diagnosis and FIGO stage. We 167 examined RB1 expression and germline BRCA status in a subset of 1 134 HGSC, and related 168 genotype to survival, tumour infiltrating CD8+ lymphocyte counts and transcriptomic 169 subtypes. Using CRISPR-Cas9, we deleted RB1 in HGSC cell lines with and without BRCA1 170 mutations to model co-loss with treatment response. We also performed genomic analyses on 171 126 primary HGSC to explore the molecular characteristics of concurrent homologous 172 recombination deficiency and RB1 loss. 173 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 8

RB1 protein loss was most frequent in HGSC (16.4%) and was highly correlated with 174 RB1 mRNA expression. RB1 loss was associated with longer OS in HGSC (hazard ratio [HR] 175 0.74, 95% confidence interval [ CI] 0.66-0.83, P = 6.8 x10-7), but with poorer prognosis in 176 ENOC (HR 2.17, 95% CI 1.17-4.03, P = 0.0140). Germline BRCA mutations and RB1 loss co-177 occurred in HGSC (P < 0.0001). Patients with both RB1 loss and germline BRCA mutations 178 had a superior OS (HR 0.38, 95% CI 0.25-0.58, P = 5.2 x10-6) compared to patients with either 179 alteration alone, and their median OS was three times longer than non-carriers whose tumours 180 retained RB1 expression (9.3 years vs. 3.1 years). Enhanced sensitivity to cisplatin (P < 0.01) 181 and paclitaxel (P < 0.05) was seen in BRCA1 mutated cell lines with RB1 knockout. Among 182 126 patients with whole-genome and transcriptome sequence data , combined RB1 loss and 183 genomic evidence of homologous recombination deficiency was correlated with transcriptional 184 markers of enhanced interferon response, cell cycle deregulation, and reduced epithelial -185 mesenchymal transition in primary HGSC. CD8+ lymphocytes were most prevalent in BRCA-186 deficient HGSC with co-loss of RB1. 187

Co-occurrence of RB1 loss and BRCA mutation was associated with 188 exceptionally long survival in patients with HGSC, potentially due to better treatment response 189 and immune stimulation. 190 191

192 Despite a high response rate to primary treatment, the progressive development of acquired 193 drug resistance is common in tubo-ovarian high-grade serous carcinoma (HGSC), a histotype 194 that is associated with approximately 70% of ovarian cancer deaths1. The frequent acquisition 195 of resistance-conferring alterations in HGSC2-4 suggests that the development of drug 196 resistance may be inevitable when curative surgery is not achieved in these patients. Countering 197 that view, however, is the observation that a small subset of patients with HGSC advanced 198 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 9 disease experience an exceptional response to treatment, survive well beyond a median of 3 .4 199 years5, and in some cases, remain disease free 6,7. Interest in studying long-term cancer 200 survivors is growing as they may assist the discovery of prognostic biomarkers, novel 201 treatments, and approaches to limit the development of resistance8. 202 Several clinical and molecular factors that influence treatment response and overall 203 survival (OS) in HGSC have been described. Complete surgical debulking is associated with a 204 more favourable outcome compared to patients left with residual disease 9-11. Molecular 205 subtypes defined by distinct gene expression patterns in primary HGSC are associated with 206 different outcomes12, including the poor survival C1/mesenchymal subtype that is more often 207 seen in patients where complete surgical tumour resection cannot be achieved13-15. By contrast, 208 the C2/immunoreactive subtype is typified by extensive infiltration of intraepithelial T cells12, 209 a feature known to be strongly associated with improved survival 16,17. Tumours arising in 210 individuals with germline or somatic alterations in BRCA1 or BRCA2 genes are typically more 211 responsive to conventional chemotherapy and poly(ADP-ribose) polymerase ( PARP) 212 inhibitors, whereas those tumours with intact homologous recombination (HR) DNA repair are 213 more often resistant to treatment 18-20. Patients with germline BRCA1 or BRCA2 pathogenic 214 variants show more favourable survival at five years post-diagnosis compared to non-carriers, 215 with BRCA2 mutation carriers retain ing a long -term (>10 year) survival advantage 21-23. 216 Although deleterious mutations in BRCA1, BRCA2 and other genes involved in HR DNA repair 217 are associated with a favourable response to treatment, these are not sufficient alone to confer 218 long-term survival and a large proportion of such patients experience a typical disease 219 trajectory. A differential outcome in mutation carriers can in part be ascribed to alternative 220 splicing24 or retention of the wild-type BRCA allele in tumours25, both of which appear to limit 221 the effectiveness of chemotherapy. 222 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 10 We previously characterised a small series of HGSC exceptional survivors and found 223 that co-occurring loss of function alterations in both BRCA and RB1 were associated with 224 unusually favourable survival7,26. Disruption of the RB pathway is found in many cancer types 225 but with variable impacts on patient outcome. For example , co -loss of RB1 and BRCA is 226 associated with shorter survival in breast and prostate cancer, possibly due to lineage switching 227 and resistance to hormonal therapy 27-29. A transcriptomic signature of RB1 loss was recently 228 described to be associated with poor outcomes across cancer types30. We have previously found 229 that chromosomal breakage is the most common mechanism of RB1 inactivation in HGSC 3, 230 accounting for approximately 80% of all RB1 alterations. In addition to its crucial role in cell 231 cycle regulation, RB1 is involved in non-canonical functions in a context- and tissue-dependent 232 manner31-33, including HR mediated DNA repair. Loss of RB1 expression in HGSC has been 233 associated with a survival benefit 34, including in the context of abnormal block -like p16 234 staining35. 235 Factors underlying the association of RB1 loss with improved outcome in HGSC are 236 unknown. Here, we contrast the pattern and clinical consequences of RB1 loss in HGSC with 237 other epithelial ovarian cancer subtypes, investigate the relevance of co-occurring BRCA1 or 238 BRCA2 mutations and RB1 loss in HGSC patients, and explore the functional effects of 239 combined BRCA and RB1 impairment in HGSC cell lines. 240 241 PATIENTS AND METHODS 242 Patient cohorts 243 The study population consisted of 7436 patients diagnosed with invasive epithelial ovarian, 244 peritoneal or fallopian tube cancer from 20 studies or biobanks participating in the Ovarian 245 Tumor Tissue Analysis (OTTA) consortium36 (Supplementary Fig. S1) . Written informed 246 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 11 consent or IRB approved waiver of consent was obtained at each site for patient recruitmen t, 247 sample collection, and study protocols (Supplementary Table S1). 248 Whole-genome sequence and matched transcriptome sequence data of primary HGSC 249 tumours were available from 126 patients from the Multidisciplinary Ovarian Cancer 250 Outcomes Group (MOCOG) study 26 (Supplementary Fig. S1) . This cohort consisted of 34 251 short-term survivors (OS <2 years), 32 moderate-term survivors (OS ≥2 and <10 years) and 60 252 long-term survivors ( OS ≥10 years) with advanced stage ( IIIC/IV) disease, enrolled in the 253 Australian Ovarian Cancer Study (AOCS), the Gynaecological Oncology Biobank at 254 Westmead Hospital (Sydney) or the Mayo Clinic Study. 255 256 Molecular analyses 257 RB1 protein expression was determined by immunohistochemistry (IHC) staining and scoring 258 of tissue microarrays (TMAs) from formalin-fixed paraffin-embedded (FFPE) tumour samples, 259 using our previously described protocol 7 (RB1 antibody clone 13A10, Leica Biosystems; 260 Supplementary Material). Subsets of HGSC patients had additional molecular or immune data 261 available (Supplementary Fig. S1), including tumour p53 protein expression status previously 262 classified37 as normal (wild -type) or abnormal (overexpression, complete absence, and 263 cytoplasmic), germline BRCA1 and BRCA2 pathogenic variant status obtained from OTTA , 264 RB1 mRNA tumour expression obtained using Nano String (ref34 and unpublished data) , 265 transcriptional subtypes of tumours using NanoString 38 and CD8+ tumour infiltrating 266 lymphocyte (TIL) density was previously classified39 based on the number of CD8+ TILs per 267 high-powered field: negative (no TILs), low (<3 TILs), moderate (3-19 TILs) or high (≥20 268 TILs). 269 The MOCOG whole -genome and transcriptome sequencing dataset of 126 short -, 270 moderate- and long-term survivors was uniformly processed as previously described26, and 271 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 12 included detailed characterisation of each tumour sample for inactivating alterations in RB1 272 and HR pathway genes, including germline and/or somatic mutations in BRCA1, BRCA2, 273 BRIP1, PALB2, RAD51C and RAD51D, or promoter methylation of BRCA1 and RAD51C. 274 Homologous recombination deficiency (HRD) status was assessed using the CHORD 275 (Classifier of Homologous Recombination Deficiency) method40, which uses specific base 276 substitution, indel and structural rearrangement signatures detected in tumo ur genomes to 277 generate BRCA1-type and BRCA2-type HRD scores. Primary tumours were classified as either 278 BRCA1-HRD & RB1 altered; BRCA1-HRD & RB1 wild-type; BRCA2-HRD & RB1 altered; 279 BRCA2-HRD & RB1 wild-type; homologous recombination proficient (HRP) & RB1 altered, 280 or HRP & RB1 wild-type. For details on differential gene expression analyses, see 281 Supplementary Material. 282 283 Cell culture 284 The AOCS patient-derived cell lines (AOCS1, AOCS3, AOCS7.2 AOCS9, AOCS11.2, 285 AOCS14, AOCS16, AOCS22, AOCS30) were established from ascites drained from patients 286 with HGSC, as previously described4. All AOCS cell lines were authenticated against matched 287 patient germline DNA using short tandem repeat markers (STR, GenePrint10 System, 288 Promega). Commercial cell lines OAW28 and CAOV3, categori sed as likely HGSC41, were 289 purchased from the American Type Culture Collection (ATCC) , and JHOS2 and OVCAR4 290 were obtained from the National Cancer Institute Repository . Commercial lines were 291 authenticated by comparing STR profiles (GenePrint10 System, Promega) to those published 292 by online repositories (Cancer Cell Line Encyclopaedia, The Cancer Genome Atlas) before use 293 in experiments. Cell lines were confirmed to be free of Mycoplasma by PCR at each revival 294 and after finishing experiments. For details on cell growth conditions, CRISPR-mediated gene 295 knockout, and molecular and functional cell line characterisation, see Supplementary Material. 296 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 13 297 Statistical analyses 298 Cox proportional hazards models were used to estimate hazard ratio s (HRs) with 95% 299 confidence intervals (CIs) using the ‘coxph’ function of the R package survival (v3.2-7). Final 300 models were fitted using Cox regression adjusted for age at diagnosis and FIGO stage. A spline 301 function was used for age at diagnosis with degree of freedom (df) 5 to account for the non-302 linear effect of the continuous variable. Regression models were fitted separately by histotype. 303 The HGSC regression models were also stratified by site of participant recruitment, and sites 304 with fewer than 10 events within the study period were excluded. The ENOC regression model 305 was not stratified by site due to the limited number of overall patients per site. The OTTA 306 survival dataset was right censored at 10 years from diagnosis to reduce the number of non -307 ovarian cancer related deaths. In the final Cox regression model, there was evidence for 308 deviation from the proportional hazard assumption, but the degree of deviation was not 309 substantial when considered alongside the large sample size and Schoenfeld residuals. The 310 Kaplan–Meier method was used to estimate and plot progression -free and overall survival 311 probabilities, and the log -rank (Mantel –Cox) test used to compare the survival duration 312 between subgroups. In the Kaplan-Meier curves, the number of patients at risk on the date of 313 diagnosis (time = 0) may be fewer than subsequent time intervals, owing to left truncation of 314 follow-up resulting from delayed study enrolment at some OTTA sites . Differences in 315 proportions of categorical features were assessed by either the chi-square or Fisher’s exact test 316 as indicated. Differences in continuous variables were assessed by either a Wilcoxon Rank 317 Sum Test or a Kruskal -Wallis test. All in vitro assays were performed across at least three 318 independent experiments, and data are expressed as mean ± standard error of the mean (SEM) 319 as indicated, from a minimum of three independent measurements. All statistical tests were 320 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 14 two-sided and considered significant when P < 0.05. Statistical analyses were performed using 321 either Prism (v9.3.1) or R (v3.6.3). 322 323

324 Loss of RB1 expression is most frequent in HGSC 325 RB1 protein expression was assessed by IHC in tumour samples from 7436 ovarian cancer 326 patients using TMAs from 20 centres participating in the OTTA consortium (Supplementary 327 Tables S1 and S2). RB1 tumour expression was classified as either retained or lost in 6564 328 samples, with 872 samples excluded that had either subclonal loss (n = 66), cytoplasmic (n = 329 17), or uninterpretable results (n = 789) due to either sample drop out or the absence of an 330 internal positive control (Fig. 1A, Supplementary Material). 331 RB1 loss was most frequent in HGSC (16.4%), followed by endometrioid ovarian 332 cancer (ENOC; 4.1%, Chi-square P < 0.0001, Fig. 1B). Loss of RB1 expression was less 333 frequent in all other histotypes (1.8% to 2.8%). RB1 mRNA expression was also assessed by 334 NanoString in a subset of HGSC tumours (n = 2552) and was significantly associated with RB1 335 protein expression (Fig. 1C, P < 0.0001). 336 337 RB1 loss is associated with longer survival in HGSC 338 Loss of RB1 protein expression was associated with longer OS in patients with HGSC (HR 339 0.74, 95% CI 0.66 -0.83, P = 6.8x10-7; Table 1) following multivariate analysis adjusting for 340 stage and age at diagnosis and stratified by study. Patients with HGSC were comparable in 341 terms of stage regardless of RB1 loss or retained expression (P = 0.9246), however those with 342 RB1 loss had a younger age at diagnosis (median 59 years versus 61 years, P = 0.0003; 343 Supplementary Table S3). Median OS was 4.7 years for patients with RB1 loss compared to 344 3.6 years for those with retained RB1 expression (Fig. 1D). 345 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 15 In contrast to HGSC, loss of RB1 expression in tumours from patients with ENOC was 346 associated with advanced stage (P = 0.0003) and poorer survival (HR 2.17, 95% CI 1.17-4.03, 347 P = 0.0140; Table 1, Fig. 1E, Supplementary Table S4). RB1 loss and abnormal p53 protein 348 expression, which is highly predictive of TP53 mutation42, were strongly correlated (chi-square 349 P < 0.0001; Supplementary Fig. 2A). While TP53 mutation is known to be associated with 350 inferior survival in patients with ENOC 37,43, we note that combined RB1 loss and abnormal 351 p53 expression w ere associated with the shortest patient survival (median OS 3. 0 years; 352 Supplementary Fig. 2B), suggesting that loss of RB1 and TP53 mutation have a compounding 353 negative impact on survival in patients with ENOC. 354 355 Combined RB1 loss and germline BRCA mutation is associated with exceptional ly good 356 survival 357 We previously observed that co-occurrence of somatic RB1 protein loss and BRCA1 or BRCA2 358 alteration (somatic or germline) was associated with longer progression-free survival (PFS) 359 and OS in HGSC7. Here, germline BRCA1 and BRCA2 status was available for 1134 HGSC 360 patients for which we had RB1 IHC data (Supplementary Fig. S1). Consistent with having a 361 younger age of diagnosis, patients with RB1 loss were more likely to have concurrent germline 362 BRCA1 or BRCA2 mutations than those with retained RB1 expression (Fig. 1F, Chi-square P 363 < 0.0001). Patients with both RB1 loss and a germline BRCA mutation had a 62% reduced risk 364 of death compared with non -carriers with retained RB1 (HR 0.38, 95% CI 0.25 -0.58, P = 365 5.2x10-6; Table 1). The median OS of BRCA germline carriers with RB1 loss was three times 366 longer than non-carriers with RB1 retained tumours (median OS 9.3 years vs. 3. 1 years, 367 respectively), while median OS was 5.2 years for BRCA carriers with retained RB1 expression 368 and 4.5 years for non-carriers with RB1 loss (Fig. 1G; Supplementary Table S5). 369 370 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 16 Enhanced response to chemotherapy in cells with impaired BRCA and RB1 function 371 To investigate whether co -occurrence of RB1 and BRCA alterations enhances sensitivity to 372 standard-of-care ovarian cancer drugs, nine patient -derived HGSC cell lines with confirmed 373 pathogenic TP53 mutation and known RB1 and BRCA status were treated with cisplatin, 374 paclitaxel and olaparib (Supplementary Fig. S 3A-C). AOCS14, the only cell line with a 375 germline BRCA1 mutation and concomitant loss of RB1 expression, showed the best response 376 to cisplatin and olaparib, and was the second most sensitive cell line to paclitaxel. In contrast 377 AOCS11.2, a line with BRCA1 promoter methylation and loss of RB1 expression, was 378 relatively resistant to paclitaxel and olaparib. A mong cell lines with intact RB1 protein 379 expression and BRCA wildtype background, AOCS3 was resistant to cisplatin, paclitaxel and 380 olaparib. 381 Except for the chemo-naïve cell line s AOCS30 and AOCS14 , all other lines were 382 derived from patients previously treated with chemotherapy. Since the evaluation of HGSC 383 cell lines with existing RB1 mutations may have been confounded by their prior, differential 384 exposure to chemotherapy we therefore characterised responses in isogenically matched lines 385 deleted of RB1 and/or BRCA1. We first inactivated RB1 in two BRCA1-mutant (AOCS7.2, 386 AOCS16) and one wild-type line (AOCS1) using CRISPR-Cas9 (Fig. 2A, Supplementary Fig. 387 S4A). RB1 knockout clones of the BRCA1-mutant cell line AOCS7.2 had enhanced sensitivity 388 to cisplatin and paclitaxel compared to RB1 wild-type clones, which was observed both in 389 short-term drug assays (72 hours, Fig. 2 B) and long er-term clonogenic survival assays (12 390 days, Fig. 2C). In this cell line , sensitivity to paclitaxel and olaparib was increased after RB1 391 knockout (paclitaxel IC50 92.0 nM versus 11.8 nM, P < 0.0001; olaparib IC50 6.1 versus 1.1 392 nM, P < 0.0001). Further, significantly fewer colonies grew in this BRCA1-mutant cell line 393 after RB1 knockout upon treatment with cisplatin ( P = 0.01), paclitaxel ( P = 0.02) or a 394 combination of both drugs (P = 0.067) in a clonogenic survival assay (n = 3). This effect was 395 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 17 not apparent in the BRCA-wild-type line (AOCS1) or the other BRCA1-mutant line (AOCS16). 396 Western blot and IHC analysis ( Supplementary Fig. S 4A) found that AOCS16 lacked 397 expression of p16, which may functionally disrupt the RB1 pathway irrespective of an RB1 398 knockout44. 399 Given that RB1 plays a central role in the negative control of the cell cycle 44,45, we 400 tested whether the enhanced chemosensitivity of RB1 knockout AOCS 7.2 cells was associated 401 with increased cell division. Live cell imaging showed similar growth rates of RB1 wildtype 402 and knockout clones of all three isogenically matched HGSC cell lines ( Supplementary Fig. 403 S4B). In both BRCA wild-type and BRCA1 mutant cell lines, RB1 knockout did not alter cell 404 cycle distribution at baseline or after 24 hours of cisplatin treatment (Supplementary Fig. S4C). 405 Paclitaxel treatment resulted in a larger proportion of cells with a tetraploid DNA content in 406 RB1 knockout cells compared to RB1 wild-type cells, indicating arrest in the G2 or M phase of 407 the cell cycle. This effect was observed in all cell lines independent of BRCA or p16 status, 408 however the arrest was more profound in the AOCS7.2 cell line (AOCS1, G2/M difference 409 8.59% ± 4.73%, P = 0.144; AOCS16, G2/M difference 8.13% ± 4.45%, P = 0.142; AOCS7.2: 410 G2/M difference 14.49% ± 3.99%, P = 0.022; Supplementary Fig. S4C). 411 We extended our analysis of isogenically matched pairs by inactivating BRCA1 and/or 412 RB1 in the chemo-naïve cell line AOCS30. While we were readily able to establish RB1 413 knockout lines, all BRCA1 targeted clones were hemizygous for BRCA1 deletion and retained 414 BRCA1 expression (Supplementary Table S6), suggesting that engineered homozygous loss of 415 BRCA1 was cell lethal, even in a tumour type where BRCA1 loss is frequently observed46. 416 417 Genomic and transcriptional landscape of HGSC with combined inactivation of BRCA and 418 RB1 419 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 18 To further understand how RB1 loss may impact the biology of HGSC with co-loss of BRCA1 420 or BRCA2, we explored matched whole -genome and transcriptome data of primary HGSC 421 tumours in the MOCOG cohort26 of 126 short- (OS <2 years), moderate- (OS ≥2 to <10 years) 422 and long-term (OS ≥10 years) survivor patients (Supplementary Fig. S1). Each tumour genome 423 was classified according to their HRD and RB1 status, resulting in 6 groups: BRCA1-HRD & 424 RB1 altered (n = 13); BRCA1-HRD & RB1 wild-type (n = 36); BRCA2-HRD & RB1 altered (n 425 = 8); BRCA2-HRD & RB1 wild-type (n = 20); HRP & RB1 altered (n = 4), or HR P & RB1 426 wild-type (n = 45; Fig. 3A). 427 The cohort had been selected for a long-term survivor study26 and hence was enriched 428 for patients with very long survival. Among BRCA2-HRD patients, those with RB1 alterations 429 had longer OS (median OS 17.0 years) compared with those without RB1 alterations (median 430 OS 11.7 years, P = 0.0004; Fig. 3B). Similarly, BRCA1-HRD patients with RB1 alterations 431 survived longer (median OS 10.4 years) than those with an intact RB1 gene (median OS 7.1 432 years). There were few HRP tumours with RB1 alterations, however these patients had a worse 433 survival (median OS 1.4 years) compared to the HRP group with no RB1 alteration (median 434 OS 2.4 years). 435 Examination of genomic features revealed relatively similar patterns within BRCA1-436 HRD and BRCA2-HRD groups, although there were a few discriminatory features identified 437 between those with and without RB1 alterations (Supplementary Figs. S5 and S6). For example, 438 the BRCA1-associated rearrangement signature Ovary_G47 was more enriched in BRCA1-HRD 439 tumours with RB1 alterations compared to those without ( P = 0.039). Among BRCA2-HRD 440 tumours, the mutational signatures DBS6 ( unknown etiology) and SBS3 (associated with 441 HRD)48 were higher in RB1-altered tumours compared to non-altered tumours, although this 442 was not significant ( P = 0.082 and P = 0.1 respectively). Concordantly, the average BRCA1-443 type and BRCA2-type CHORD scores40 were highest in BRCA1- and BRCA2-HRD tumours 444 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 19 with RB1 alterations respectively, indicating a higher probability of HRD . As described 445 previously49, CCNE1 gene amplifications were absent in tumours with both HRD and RB1 446 alterations (P = 0.0006; Supplementary Fig. S7). 447 We hypothesised that tumours with combined HRD and RB1 loss may have unique 448 transcriptional profiles. To explore this, we compared gene expression profiles between each 449 HRD/RB1 group and the reference set of tumours that were HR P and RB1 wild-type 450 (Supplementary Table S7, Supplementary Fig. S8 ). There was significant enrichment of 451 MSigDB hallmark gene sets among genes differentially expressed in BRCA1-HRD tumours 452 with RB1 alterations, the most prominent being interferon gamma response (up), interferon 453 alpha response (up), oxidative phosphorylation (up), and E2F targets (up; adjusted P < 0.0001; 454 Fig. 4A). The differentially expressed genes identified between BRCA2-HRD / RB1 altered 455 tumours and the reference set were significantly enriched for the MSigDB hallmark gene sets: 456 E2F targets (up), epithelial mesenchymal transition (down), G2M checkpoint (up), and TNF 457 alpha signalling via NF-kB (up; adjusted P < 0.0001). 458 Since enhanced tumour cell proliferation has been associated with long -term survival 459 in HGSC7,26, and loss of RB1 might accelerate proliferation31, we evaluated the expression of 460 proliferation markers across the RB1 and BRCA subgroups. BRCA1-HRD tumours with RB1 461 alterations had significantly higher mRNA levels of the cell proliferation related genes PCNA 462 (proliferating cell nuclear antigen) and MCM3 (minichromosome maintenance complex 463 component 3) compared to BRCA1-HRD tumours without RB1 alterations ( P < 0.0001, 464 Supplementary Fig. S 6). However, there were no significant differences in the proportion of 465 Ki-67 positive cancer cell nuclei (P = 0.3297) across the subgroups (Supplementary Fig. S 6), 466 which was previously quantified by immunohistochemistry7 in a subset of primary tumours (n 467 = 59). 468 469 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 20 Germline BRCA mutation carriers with somatic loss of RB1 tumour expression show 470 elevated immune activity 471 Having observed that HGSC with combined RB1 loss and HRD have enrichment of 472 transcriptional signatures associated with an enhanced immune response, we accessed existing 473 immunohistochemical data39 to determine the prevalence of CD8+ TILs in HGSC samples that 474 also had RB1 protein expression and BRCA germline mutation status (n = 868). BRCA carriers 475 with RB1 loss had a significantly higher proportion of tumours ( 79.6%) with moderate and 476 high densities of CD8+ TILs, compared to BRCA carriers with retained RB1 (6 4.9%), non-477 carriers with RB1 loss (72. 4%) and non-carriers with retained RB1 (63. 6%, P = 0.0264; Fig. 478 4B). Tumours with complete absence of CD8+ TILs were the least frequent in BRCA carriers 479 with RB1 loss (4. 1%) compared to the other groups (1 3.8 % of BRCA carriers with retained 480 RB1 tumour expression, 14. 6% of non-carriers with RB1 tumour loss, 18. 8% of non-carriers 481 with retained RB1 tumour expression). 482 Gene expression-based molecular subtypes12,38 also differed by RB1 and BRCA status 483 (P = 0.0271, n = 601; Fig. 4C). As expected, there was enrichment for the C2/immunoreactive 484 subtype, a subtype characterised by the presence of intratumoural CD8+ T cells and good 485 survival, in germline BRCA carriers with RB1 loss ( 32.4%) compared to the other subgroups 486 (between 19.8% and 23.4%). Additionally, tumours with RB1 loss were enriched for the 487 C4/differentiated molecular subtype, a subtype characterised by cytokine expression and good 488 survival, regardless of BRCA status (45.9% in BRCA carriers with RB1 loss, 50.0% in non -489 carriers with RB1 loss, 39.5% in BRCA carriers with retained RB1, 32.1% of non-carriers with 490 retained RB1). BRCA carriers with RB1 loss also had the lowest proportion of the 491 C5/proliferative molecular subtype (2. 7% versus 1 7.2% to 20.3% in the other groups), a 492 subtype associated with diminished immune cell infiltration and poor survival12,19. 493 494 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 21

495 Identifying the determinants of long-term patient survival, particularly in cancers with a 496 generally unfavourable prognosis such as HGSC, may reveal novel therapeutic targets and 497 inform personalised treatment strategies8. Improved survival associated with RB1 loss has been 498 described previously in HGSC7,34,35,50 but the underlying factors contributing to this survival 499 benefit have not been studied to date. We assessed tumour samples from a cohort of more than 500 7,000 women with ovarian cancer , including a subset with high resolution genomic data , to 501 understand how RB1 loss may impact on therapeutic response and patient survival. 502 Alteration of the RB1 pathway is a frequent event in tumourigenesis, including loss of 503 regulators such as p16, activation of D - and E -type cyclins and their associated cyclin 504 dependent kinases, and loss of RB1 itself (reviewed in 51). Our study showed that RB1 loss is 505 associated with longer survival in patients with advanced stage HGSC, but by contrast, loss of 506 RB1 in ENOC was associated with a shorter survival , particularly in combination with p53 507 mutation. Similar to ENOC , i n endocrine -driven breast and prostate cancer, RB1 loss is 508 associated with poorer survival : early co -loss of BRCA2 and RB1 is associated with an 509 aggressive, castration-resistant prostate cancer subtype (CRPC) characterised by epithelial-to-510 mesenchymal transition and shorter survival29. RB1 loss facilitates lineage plasticity and, with 511 p53-comutation, leads to an androgen-independent phenotype52,53 and consequently resistance 512 to anti-androgen therapy. In estrogen-receptor (ER) positive breast cancer, CDK4/6 inhibitor 513 resistance is associated with RB1 loss and cyclin E2 activation54,55. 514 Triple negative breast cancer (TNBC) provides an important contrast to the findings for 515 RB1 loss in ER-positive breast cancer. In TNBC, RB1 loss is most common in the basal-like 516 subtype, where BRCA1 mutation and promoter hypermethylation is associated with frequent 517 RB1 gene disruption and RB1 loss 28. RB1 loss alone, as well as co -occurrence with BRCA1 518 promoter hypermethylation , is associated with a favourable chemotherapy response and 519 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 22 outcome27,56-58. Notably, TNBC and HGSC are more similar than the cancers that they are 520 grouped with anatomically, sharing gene expression patterns, genetic drivers including BRCA1 521 and BRCA2, ubiquitous loss of TP53, extensive copy number variation, and susceptibility to 522 platinum-based chemotherapy 59,60. Taken together, the relationship between RB1 loss and 523 patient survival appears to be dependent on cancer type and molecular context61. 524 Some, but not all TNBC and early metastatic prostate cancer s are associated with 525 germline variants in BRCA1, BRCA2 and other genes involved in HR DNA repair. However, 526 previous tumour studies of RB1 expression have not also defined the HRD status of individual 527 samples. A strength of this study was the known BRCA germline status of 1134 of the HGSC 528 patients for which we also had RB1 protein expression, and this revealed the strong association 529 of co-mutation in either BRCA1 or BRCA2 and RB1 with survival. In addition to germline 530 mutations in BRCA1 or BRCA2, germline or somatic mutations, and promoter methylation of 531 other genes involved in HR DNA repair, such as RAD51C, can result in a similar molecular 532 phenotype, characterised by distinct genomic scarring26. Using whole-genome sequence data, 533 we determined the likely tumour HRD status in a subset of 126 tumours using an algorithm 534 that recogni ses genomic scarring associated with HRD (Fig. 3A), rather than simply 535 designating BRCA mutation status, which does not account for all mechanisms of HR repair 536 inactivation. Although the number of samples with RB1 loss and HR proficiency was small, 537 the very poor outcome we observed with this group indicated that for RB1 to impart a survival 538 benefit in HGSC, it must occur in a n HRD background. Validation of this finding in a larger 539 cohort may further inform how RB1 loss could favourably influence survival in certain 540 histological and molecular contexts. 541 We have previously noted that enhanced proliferation in HGSC is associated with long-542 term survival7,26 and it is reasonable to suggest that RB1 loss may be imparting an effect 543 through deregulating the cell cycle. However, data on the effect of RB1 loss on proliferation in 544 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 23 HGSC tumours and cancer cell lines is inconsistent. RB1 knockout in our HGSC cell lines did 545 not cause cell cycle alterations in the absence of treatment, and despite differences in 546 proliferative markers at the mRNA level, there was no significant difference in the proportion 547 of Ki-67 positive nuclei between tumours with or without RB1 protein expression. In a recent 548 OTTA study, Ki-67 expression was not associated with survival in HGSC; however, there was 549 strong correlation between loss of RB1 and the proliferative marker MCM3 62, which may 550 provide a more accurate measure of tumour cell proliferation than Ki-6763. 551 In addition to its role in driving progression through the G1 stage of the cell cycle, RB1 552 has non-canonical functions. RB1 has been shown to participate in HR DNA repair through 553 interactions with BRG1 and ATM64. A recent pan-cancer study65 found that combined loss of 554 TP53 and RB1 was associated with a particularly high genome -wide loss-of-heterozygosity 555 score, one of the key elements of genomic scarring associated with HRD. In our whole-genome 556 analysis, HGSC tumours with dual loss of HRD and RB1 did not exhibit overall higher 557 mutation burden; however, we did observe elevated levels of mutational signatures associated 558 with HRD, which may be evidence of compounding DNA repair defects . It remains possible 559 that the combined inactivation of RB1 and HR genes contribute to enhanced chemotherapy 560 response and/or an impaired ability for tumour cells to develop therapy resistance. 561 When we evaluated a set of patient derived HGSC lines, those with germline BRCA1 562 mutation and RB1 alteration were most sensitive to cisplatin and olaparib. Knockout of RB1 in 563 the AOCS 7.2 cell line which had a pre-existing BRCA1 mutation, resulted in an increase in 564 chemosensitivity, consistent with the notion that co -mutation enhances chemotherapy 565 response7. Unfortunately , despite considerable effort s, we were unable to generate a larger 566 series of isogenically matched cell lines with combinations of conditional knockout s of RB1 567 and BRCA1 as all surviving clones retained at least one BRCA1 allele. BRCA1 loss is embryonic 568 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 24 lethal and engineered loss in cell lines has been reported as lethal elsewhere including in the 569 human haploid cell line, HAP146. 570 Our data provides evidence of an enhanced immunogenicity in HGSC with RB1 loss, 571 with higher CD8+ TIL counts and upregulated expression of IFN -γ signalling pathways. RB1 572 has been shown to inhibit innate IFN -β production in immunocompetent mice 66 and RB1 573 deficiency triggered an increased IFN -β and IFN-α secretion. Co-mutation of RB1 and TP53 574 was recently found to be associated with an enhanced response to the immune checkpoint 575 inhibitor atezolizumab in metastatic urothelial bladder cancer 67. Similarly, a case report 576 described a complete response to atezolizumab in heavily pre-treated, RB1-negative TNBC68. 577 This generates the hypothesis that RB1 loss could predict response to such therapies in HGSC, 578 since this tumour type ubiquitously harbours TP53 mutations69. However, a recent biomarker 579 study in ovarian cancer patients treated with atezolizumab or placebo and standard 580 chemotherapy found that deleterious mutations in RB1 were prognostic for a better PFS, 581 regardless of the addition of atezolizumab 70. While it appears RB1 loss alone may not be 582 predictive of response to the PD -L1 inhibitor atezolizumab, response rates to PD -1/PD-L1 583 pathway checkpoint inhibitors are generally quite low in HGSC, with the best objective 584 response rates between 8% and 15% 71. Our study has identified a subset of patients with 585 combined RB1 and BRCA inactivation who demonstrate exceptional immune responses and 586 may provide clues for the development of new immunotherapeutic strategies for HGSC that 587 extend beyond targeting PD-L1/PD-1. 588 Our work highlights the importance of RB1 loss to treatment response and survival and 589 focuses attention on other therapeutic opportunities in this subset of HGSC patients. 590 Approximately 20 percent of HGSC patients have somatic loss of RB1 assessed using genomic 591 data3,26, a figure that is consistent with the immunohistochemical results obtained in the large 592 patient cohort described here. Both approaches indicate that RB1 loss is generally clonal, 593 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 25 enhancing its value as a therapeutic target if selective inhibitors can be identified. Casein kinase 594 2 (CK2) inhibitors have been reported to enhance the sensitivity of RB1-deficient TNBC and 595 HGSC cells to carboplatin and niraparib 72. In addition, Aurora kinase A and B inhibition is 596 synthetically lethal in combination with RB1 loss in breast and lung cancer cells 73-75. 597 Irrespective of HRD status, RB1 mutations correlate with sensitivity to WEE1 inhibition in 598 TP53 mutant TNBC and HGSC patient-derived xenografts76, indicating additional treatment 599 options that exploit RB1 inactivation in these tumours. In this study, t he BRCA1-mutant cell 600 line AOCS7.2 with induced RB1 knockout was more sensitive to olaparib suggesting that RB1 601 loss may also predict responses to PARP inhibitors in HGSC. RB1 staining of tumour tissue 602 by IHC is a relatively low-cost pathology-based assay that could be used in prospective studies 603 to test whether RB1 expression is predictive of responses to PARP inhibitors, either alone or 604 in combination with approved HRD tests. 605 606 ACKNOWLEDGMENTS 607 We thank J. Beach and L. Bowes for their contributions to the study. This work was supported 608 by the National Health and Medical Research Council (NHMRC) of Australia (1186505 to 609 DWG; 1092856, 1117044 and 2008781 to DDLB; 2009840 to SJR), the National Institutes of 610 Health (NIH) / National Cancer Institute (R01CA172404 to SJR, P50 CA136393 to SHK) and 611 the U.S. Army Medical Research and Materiel Command Ovarian Cancer Research Program 612 (Award No. W81XWH -16-2-0010 and W81XWH -21-1-0401). DWG is supported by a 613 Victorian Cancer Agency / Ovarian Cancer Australia Low-Survival Cancer Philanthropic Mid-614 Career Research Fellowship ( MCRF22018). FAMS is supported by a Swiss National 615 Foundation Early Postdoc Mobility Fellowship (P2BEP3 -172246), a Swiss Cancer League 616 grant BIL KFS-3942-08-2016 and a Prof. Max Cloëtta foundation grant. KIP is supported by 617 a NHMRC CJ Martin Overseas Biomedical Fellowship (APP1111032). ELC is supported by a 618 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 26 Victorian Cancer Agency Mid -Career Fellowship ( MCRF21004). MW is supported by the 619 European Research Council under the European Union’s Horizon 2020 Research and 620 Innovation Programme grant agreement No 742432 (BRCA -ERC). KS is supported by the 621 Swedish Cancer Foundation. MSA is funded through a Michael Smith Health Research BC 622 Scholar Award (18274) and the Janet D. Cottrelle Foundation Scholars program managed by 623 the BC Cancer Foundation. 624 BC’s Gynecological Cancer Research team (OVCARE) receives support through the 625 BC Cancer Foundation and the VGH & UBC Hospitals Foundation. The Gynaecological 626 Oncology Biobank at Westmead was funded by the NHMRC (ID310670, ID628903); the 627 Cancer Institute NSW (12/RIG/1 -17, 15/RIG/1 -16); and acknowledges support from the 628 Department of Gynaecological Oncology, Westmead Hospital, and the Sydney West 629 Translational Cancer Research Centre (Cancer Institute NSW 15/TRC/1 -01). The Women's 630 Cancer Research Program at Cedars-Sinai Medical Center (LAX) is supported by The National 631 Center for Advancing Translational Sciences (NCATS) Grant UL1TR000124 . The Study of 632 Epidemiology and Risk Factors in Cancer Heredity (SEARCH) is funded by Cancer Research 633 UK (C490/A10119 C490/A10124 C490/A16561) and the UK National Institute for Health 634 Research Biomedical Research Centre at the University of Cambridge. The UKOPS study was 635 funded by The Eve Appeal (The Oak Foundation) with contribution to authors’ salary through 636 MRC core funding MC_UU_00004/01 and the National Institute for Health Research 637 University College London Hospitals Biomedical Research Centre. 638 The investigators also acknowledge generous contributions from the Border Ovarian 639 Cancer Awareness Group, the Peter MacCallum Cancer Foundation, the Graf Family 640 Foundation, Wendy Taylor, Arthur Coombs and family, and the Piers K Fowler Fund. The 641 contents of the published material are solely the responsibility of the authors and do not reflect 642 the views of the NHMRC, NIH, and other funders. 643 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 27 644 AUTHOR CONTRIBUTIONS 645 MK, SJR, DDLB and DWG conceived the study design. FAMS, KT, KP, JB and TH carried 646 out experiments, and analysed and interpreted results along with TB, AP, DA, TZ, NSM, SF, 647 AD, MK, SJR, DDLB and DWG. MK assessed and interpreted immunohistochemical scores. 648 All authors contributed through recruitment and consenting of patients, collection and 649 processing of biological samples, clinical care, abstraction and curation of clinical data and 650 maintenance of follow -up. DDLB and DWG supervised the study and together with FAMS 651 and KT wrote the manuscript. All authors contributed to writing, review and revision of the 652 manuscript and approved the final submitted version. 653 654 COMPETING INTERESTS 655 DDLB is an Exo Therapeutics advisor and has received research grant funding from 656 AstraZeneca, Genentech -Roche and BeiGene for unrelated work. SF, NT, KA, and ADeF 657 received grant funding from AstraZeneca for unrelated work . AGM and UM report funded 658 research collaborations for unrelated work with industry: Intelligent Lab on Fiber, RNA 659 Guardian, Micronoma and Mercy BioAnalytics. UM had stock ownership (2011 -2021) 660 awarded by University College London (UCL) in Abcodia, which held the licence for the Risk 661 of Ovarian Cancer Algorithm (ROCA). UM reports research collaboration contracts with 662 Cambridge University and QIMR Berghofer Medical Research Institute. UM holds patent 663 number EP10178345.4 for Breast Cancer Diagnostics. UM is a member of Tina's Wish 664 Scientific Advisory Board (USA) and Research Advisory Panel, Yorkshire Cancer Research 665 (UK). The remaining authors declared no conflicts of interest. 666 667 Figure legends: 668 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 28 Figure 1. Expression of RB1 and survival associations across ovarian cancer histotypes. 669 (A) Representative images of immunohistochemical detection of RB1 expression in ovarian 670 carcinoma tissues, showing examples of the three most common expression patterns: retained, 671 lost and subclonal loss. (B) Proportion of patients with loss or retention of RB1 protein 672 expression in tumour samples by ovarian cancer histotype s. Chi-square P value reported for 673 difference in proportion s across all histotypes . HGSC, tubo-ovarian high-grade serous 674 carcinoma; LGSC, low -grade serous carcinoma; MOC, mucinous ovarian cancer; ENOC, 675 endometrioid ovarian cancer; CCOC, clear cell ovarian cancer. (C) Boxplots show RB1 mRNA 676 expression (NanoString) by RB1 protein expression status; lines indicate median and whiskers 677 show range (Mann-Whitney test P value reported). Kaplan-Meier analysis of overall survival 678 in patients diagnosed with HGSC (D) and ENOC (E) stratified by tumour RB1 expression. (F) 679 Loss of RB1 tumour expression is more common in germline BRCA1 and BRCA2 mutation 680 carriers than retained RB1 expression. Chi -square P value is reported. (G) Kaplan-Meier 681 estimates of overall survival in HGSC patients by combined germline BRCA and tumour RB1 682 expression status. 683 684 Figure 2. Sensitivity to therapeutic agents in BRCA1-mutant cell lines with RB1 knockout. 685 (A) RB1 was knocked out using CRISPR/Cas9 in 3 patient-derived Australian Ovarian Cancer 686 Study (AOCS) HGSC cell lines with either wild -type or mutant BRCA1 background. 687 Representative Western Blots show protein levels of RB1 and phosphorylated RB1 (pRB1) 688 compared to GAPDH loading control in single cell cloned, homozygous RB1 wildtype (WT) 689 and knockout (KO) colonies in comparison to heterogeneous populations with a scramble 690 single guide RNA (sgRNA). Independent blots were used for RB1 and pRB1. (B) Cell viability 691 was compared between RB1 WT and KO clones following treatment with cisplatin (72 hours), 692 paclitaxel (72 hours) or olaparib (120 hours). Nonlinear regression drug curves are shown; P 693 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 29 values of a curve fit, extra sum -of squares F test (ns, not significant; ** P < 0.01; **** P < 694 0.0001; n = 3). Error bars indicate ± SEM ; for some values error bars are shorter than the 695 symbols and thus are not visible. (C) Proportion of surviving colonies following 16 days of 696 treatment with cisplatin, paclitaxel or a combination of both (with half of the IC50 determined 697 per drug and cell line respectively) relative to DMF vehicle control (n = 3 replicates). Data are 698 presented as mean ± SEM. Mean values were compared by student's t-test (ns, not significant; 699 *P < 0.05; **P < 0.01). Representative scans of the fixed cell colonies stained with crystal 700 violet are shown for each condition. 701 702 Figure 3. Genomic landscape of high -grade serous ovarian tumours with co -occurring 703 BRCA and RB1 alterations. 704 (A) Pathogenic germline and somatic alterations in homologous recombination (HR) and DNA 705 repair genes detected by whole -genome sequencing and DNA methylation analysis of 126 706 primary HGSC samples26 are shown, as well as alterations in immune genes and CCNE1. 707 Samples are grouped by HR D and RB1 status (wt, wild-type; mut, mutation). Bars at the top 708 indicate the number of alterations in each listed gene per patient. Patients are annotated with 709 survival group (LTS, long -term survivor, OS >10 years; MTS, mid -term survivor, OS 2 -10 710 years; STS, short-term survivor, OS <2 years), tumour CHORD40 scores, and the proportion of 711 structural variant (SV) type ( DUP, duplication; DEL, deletion; INV, inversion; ITX, intra -712 chromosomal translocation ). (B) Kaplan -Meier estimates of progression -free and overall 713 survival of patients with according to HR status ( BRCA1-type HRD, BRCA2-type HRD or 714 homologous recombination proficient tumours) and RB1 status (mut, mutation; wt, wild-type). 715 716 Figure 4. Characterisation of HGSC with co-loss of RB1 and BRCA. 717 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 30 (A) Gene set enrichment analysis indicating up - and downregulated pathways in tumours 718 according to BRCA and RB1 status. HRP, homologous recombination proficient; HRD, 719 homologous recombination deficient; RB1wt, RB1 wild-type; RB1m, RB1 altered. (B) 720 Proportion of tumour infiltrating lymphocytes (TILs) in HGSC tumours grouped by RB1 721 expression and BRCA germline mutation status (Chi-square P value is indicated). (C) 722 Proportion of tumours classified as each HGSC molecular subtype 12 grouped by RB1 723 expression and BRCA germline mutation status (Chi -square P value is indicated; C5.PRO, 724 C5/proliferative subtype ; C4.DIF, C4/differentiated subtype ; C2.IMM, C2/immunoreactive 725 subtype; C1.MES, C1/mesenchymal subtype). 726 727 Supplementary Figure S1. Patients and tumour samples analysed in this study. 728 Number of patients included in each molecular analysis. HGSC, tubo -ovarian high -grade 729 serous ovarian carcinoma; ENOC, endometrioid ovarian carcinoma; OS, overall survival. 730 731 Supplementary Figure S2. Combined p53 and RB1 protein expression in ENOC. 732 (A) Correlation between RB1 and p53 tumour expression in patients with endometrioid ovarian 733 carcinoma (ENOC). Chi -square P value is reported. (B) Kaplan -Meier estimates of overall 734 survival in patients with ENOC by combined RB1 and p53 tumour expression status. 735 736 Supplementary Figure S3. Drug sensitivity in HGSC cell lines with innate RB1 and/or 737 BRCA1 alterations. 738 (A) Summary of the molecular features of innate HGSC cell models, including mutations in 739 key genes (TP53, CDKN2A, BRCA1, BRCA2), copy number alterations in CCNE1, and protein 740 expression of RB1 and p16. (B) IC50 of high grade serous ovarian cancer cell lines after 741 treatment with cisplatin (72 hours), paclitaxel (72 hours), or olaparib (120 hours). ND, Not 742 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 31 determined. (C) Viability of high -grade serous ovarian cancer cell lines after treatment with 743 cisplatin (72 hours), paclitaxel (72 hours), or olaparib (120 hours). Data are expressed as mean 744 (n = 3 replicates) ± standard error of the mean (SEM). For some points, error bars are shorter 745 than the height of the symbol and are not visible. 746 747 Supplementary Figure S4. Cell proliferation and cell cycle distribution of HGSC cell lines 748 with RB1 knockout. 749 (A) CRISPR/Cas9 knockout of RB1 in 3 patient-derived ovarian cancer cell lines with different 750 BRCA1/2 and p16 background s. The bar graph indicates RB1 mRNA expression levels 751 determined by RT -PCR ( n = 3) in single -cell clones confirming RB1 wildtype (WT) and 752 knockout (KO) compared to heterozygous colonies without gene editing (Scramble). 753 Representative Western Blots show p16 protein levels compared to GAPDH loading controls 754 in each cell line and clone. Images of p16 IHC in AOCS parental cell lines are included 755 confirming the respective p16 status. (B) Proliferative capacity of 3 patient-derived HGSC cell 756 lines (RB1 wild-type, WT and RB1 knockout, KO clones) measured by IncuCyte Zoom live -757 cell imaging. Data represent mean ± SEM confluency after 20 -25% starting confluency from 758 three to six independent experiments. Dashed line denotes 75% confluency. (C) Cell cycle 759 distribution following RB1 CRISPR knockout. Proportion of cells in G0G1, S or G2/M phase 760 24 hours after treatment with DMF, cisplatin or paclitaxel at half the IC50 determined per cell 761 line and drug, analysed by flow cytometry. Mean proportion ± SEM of three independently 762 performed experiments are shown. Distribution was compared between RB1 WT and KO 763 clones using unpaired t test (ns, not significant; *P < 0.05). 764 765 Supplementary Figure S5. Mutational signatures in homologous recombination 766 deficiency and RB1 subgroups. 767 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 32 Boxplots show the relative proportion (y -axis) of genome -wide mutational signatures 26 768 according to homologous recombination deficiency (HRD) and RB1 status. Boxes show the 769 interquartile range (25-75th percentiles), central lines indicate the median, dots represent each 770 sample, whiskers show the smallest and largest values within 1.5 times the interquartile range, 771 red triangles indicate the mean, and dotted lines join the mean of each subgroup to visualise 772 the trend. The Kruskal–Wallis test P values displayed are Benjamini -Hochberg adjusted and 773 the signatures are ordered by their significance. Pair-wise Mann -Whitney-Wilcoxon test 774 adjusted P values are also reported. HRP, homologous recombination proficient. 775 776 Supplementary Figure S6. Genomic and clinical characteristics by combined homologous 777 recombination deficiency and RB1 status. 778 Boxplots show numerical clinical and genomic features (y-axis) according to homologous 779 recombination deficiency (HRD) and RB1 status. Boxes show the interquartile range (25 -75th 780 percentiles), central lines indicate the median, dots represent each sample, whiskers show the 781 smallest and largest values within 1.5 times the interquartile range, red triangles indicate the 782 mean, and dotted lines join the mean of each subgroup to visualise the trend . The Kruskal–783 Wallis test P values displayed are Benjamini -Hochberg adjusted and the features are ordered 784 by their significance. Pair-wise Mann -Whitney-Wilcoxon test adjusted P values are also 785 reported. Features include BRCA1- and BRCA2-type CHORD ( Classifier of HOmologous 786 Recombination Deficiency) scores; mean HRD scores ( scarHRD); absolute numbers of 787 structural variants (SVs), including deletions (DEL), duplications (DUP), intrachromosomal 788 rearrangements (ITX), and inversions (INV); relative expression levels of PCNA and MCM3; 789 proportion of whole-genome loss-of-heterozygosity (LOH); number of predicted neoantigens 790 and variants per megabase (Mb); age of patients at diagnosis ; progression-free and overall 791 survival; cancer cell purity and ploidy; absolute CIBERSORTx scores; proportion of Ki -67 792 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 33 positive tumour cells were available for n = 59 primary tumours as previously measured by 793 immunohistochemistry7. HRP, homologous recombination proficient. 794 795 Supplementary Figure S7. Gene alterations across BRCA and RB1 altered subgroups. 796 Proportion of tumours with alterations in genes of interest for each subgroup. WT, wild -type; 797 MUT, mutation ; HRP, homologous recombination proficient . Genes are ordered by 798 significance using Fisher's exact test; Benjamini-Hochberg adjusted P values are reported. 799 800 Supplementary Figure S8. Differentially expressed genes. 801 Bars indicate the number of differentially expressed genes (Benjamini-Hochberg adjusted P 802 value < 0.05) between HGSC tumours grouped by HRD and/or RB1 status as 803 shown. Differential gene expression analysis was performed using DESeq2 to determine fold 804 change of gene expression between groups (see Supplementary Table 7 for full DESeq2 805 results). HRP, homologous recombination proficient; HRD, homologous recombination 806 deficient; RB1wt, RB1 wild-type; RB1m, RB1 altered. 807 808 Supplementary Table captions: 809 Supplementary Table S1. 810 Details of participating Ovarian Tumor Tissue Analysis (OTTA) consortium studies and ethics 811 approval. 812 Supplementary Table S2. 813 Number of patients by study and histotype. 814 Supplementary Table S3. 815 Clinical characteristics of patients diagnosed with high-grade serous ovarian cancer. 816 Supplementary Table S4. 817 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 34 Clinical features of patients with endometrioid ovarian cancer 818 Supplementary Table S5. 819 Clinical characteristics of patients with high -grade serous ovarian cancer according to BRCA 820 and RB1 status. 821 Supplementary Table S6. 822 Relative expression of BRCA1 and RB1 by qPCR in AOCS30 CRISPR knockout model. 823 Supplementary Table S7. 824 Differential gene expression analysis comparing transcriptomes of tumours based on BRCA 825 and RB1 alteration status. 826 Supplementary Table S8. 827 Summary of cell lines used in this study. 828 Supplementary Table S9. 829 Summary of gene alterations and expression found in cell lines. 830 Supplementary Table S10. 831 Sequence of single guide RNA used for CRISPR gene knockout. 832 Supplementary Table S11. 833 Antibodies and reagents used for this project. 834 Supplementary Table S12. 835 List of primer sequences used in the study. 836 837

838 1. Bowtell DD, Böhm S, Ahmed AA, et al. Rethinking ovarian cancer II: reducing 839 mortality from high-grade serous ovarian cancer. Nature Reviews Cancer 2015; 15(11): 668-840 79. 841 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 35 2. Norquist B, Wurz KA, Pennil CC, et al. Secondary somatic mutations restoring 842 BRCA1/2 predict chemotherapy resistance in hereditary ovarian carcinomas. Journal of 843 Clinical Oncology 2011; 29(22): 3008-15. 844 3. Patch AM, Christie EL, Etemadmoghadam D, et al. Whole-genome characterization of 845 chemoresistant ovarian cancer. Nature 2015; 521(7553): 489-94. 846 4. Christie EL, Pattnaik S, Beach J, et al. Multiple ABCB1 transcriptional fusions in drug 847 resistant high-grade serous ovarian and breast cancer. Nature communications 2019; 10(1): 848 1295-. 849 5. Gockley A, Melamed A, Bregar AJ, et al. Outcomes of Women With High -Grade and 850 Low-Grade Advanced -Stage Serous Epithelial Ovarian Cancer. Obstetrics and gynecology 851 2017; 129(3): 439-47. 852 6. Dao F, Schlappe BA, Tseng J, et al. Characteristics of 10-year survivors of high-grade 853 serous ovarian carcinoma. Gynecologic Oncology 2016; 141(2): 260-3. 854 7. Garsed DW, Alsop K, Fereday S, et al. Homologous recombination DNA repair 855 pathway disruption and retinoblastoma protein loss are associated with exceptional survival in 856 high-grade serous ovarian cancer. Clinical Cancer Research 2018; 24(3): 569-80. 857 8. Saner FAM, Herschtal A, Nelson BH, et al. Going to extremes: determinants of 858 extraordinary response and survival in patients with cancer. Nature Reviews Cancer 2019; 859 19(6): 339-48. 860 9. du Bois A, Reuss A, Pujade -Lauraine E, Harter P, Ray-Coquard I, Pfisterer J. Role of 861 surgical outcome as prognostic factor in advanced epithelial ovarian cancer: A combined 862 exploratory analysis of 3 prospectively randomized phase 3 multicenter trials. Cancer 2009; 863 115(6): 1234-44. 864 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 36 10. Wallace S, Kumar A, Mc Gree M, et al. Efforts at maximal cytoreduction improve 865 survival in ovarian cancer patients, even when complete gross resection is not feasible. 866 Gynecologic Oncology 2017; 145(1): 21-6. 867 11. Harter P, Sehouli J, Vergote I, et al. Randomized Trial of Cytoreductive Surgery for 868 Relapsed Ovarian Cancer. The New England journal of medicine 2021; 385(23): 2123-31. 869 12. Tothill RW, Tinker AV, George J, et al. Novel molecular subtypes of serous and 870 endometrioid ovarian cancer linked to clinical outcome. Clinical Cancer Research 2008; 871 14(16): 5198-208. 872 13. Liu Z, Beach JA, Agadjanian H, et al. Suboptimal cytoreduction in ovarian carcinoma 873 is associated with molecular pathways characteristic of increased stromal activation. 874 Gynecologic Oncology 2015; 139(3): 394-400. 875 14. Wang C, Armasu SM, Kalli KR, et al. Pooled Clustering of High-Grade Serous Ovarian 876 Cancer Gene Expression Leads to Novel Consensus Subtypes Associated with Survival and 877 Surgical Outcomes. Clinical cancer research : an official journal of the American Association 878 for Cancer Research 2017; 23(15): 4077-85. 879 15. Torres D, Kumar A, Bakkum-Gamez JN, et al. Mesenchymal molecular subtype is an 880 independent predictor of severe postoperative complications after primary debulking surgery 881 for advanced ovarian cancer. Gynecologic Oncology 2019; 152(2): 223-7. 882 16. Zhang L, Conejo-Garcia JR, Katsaros D, et al. Intratumoral T Cells, Recurrence, and 883 Survival in Epithelial Ovarian Cancer. New England Journal of Medicine 2003; 348(3): 203-884 13. 885 17. Hwang WT, Adams SF, Tahirovic E, Hagemann IS, Coukos G. Prognostic significance 886 of tumor-infiltrating T cells in ovarian cancer: A meta -analysis. Gynecologic Oncology 2012; 887 124(2): 192-8. 888 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 37 18. Fong PC, Yap TA, Boss DS, et al. Poly(ADP) -ribose polymerase inhibition: frequent 889 durable responses in BRCA carrier ovarian cancer correlating with platinum -free interval. 890 Journal of clinical oncology : official journal of the American Society of Clinical Oncology 891 2010; 28(15): 2512-9. 892 19. The Cancer Genome Atlas Research Network. Integrated genomic analyses of ovarian 893 carcinoma. Nature 2011; 474(7353): 609-15. 894 20. Pennington KP, Walsh T, Harrell MI, et al. Germline and somatic mutations in 895 homologous recombination genes predict platinum response and survival in ovarian, fallopian 896 tube, and peritoneal carcinomas. Clinical cancer research : an official journal of the American 897 Association for Cancer Research 2014; 20(3): 764-75. 898 21. Bolton KL, Chenevix -Trench G, Goh C, et al. Association between BRCA1 and 899 BRCA2 mutations and survival in women with invasive epithelial ovarian cancer. JAMA 2012; 900 307(4): 382-90. 901 22. Alsop K, Fereday S, Meldrum C, et al. BRCA mutation frequency and patterns of 902 treatment response in BRCA mutation-positive women with ovarian cancer: A report from the 903 Australian ovarian cancer study group. Journal of Clinical Oncology 2012; 30(21): 2654-63. 904 23. Candido-dos-Reis FJ, Song H, Goode EL, et al. Germline mutation in BRCA1 or 905 BRCA2 and ten -year survival for women diagnosed with epithelial ovarian cancer. Clinical 906 cancer research 2015; 21(3): 652-7. 907 24. Wang Y, Bernhardy AJ, Cruz C, et al. The BRCA1 -Δ11q alternative splice isoform 908 bypasses germline mutations and promotes therapeutic resistance to PARP inhibition and 909 cisplatin. Cancer Research 2016; 76(9): 2778-90. 910 25. Maxwell KN, Wubbenhorst B, Wenz BM, et al. BRCA locus -specific loss of 911 heterozygosity in germline BRCA1 and BRCA2 carriers. Nature communications 2017; 8(1): 912 319-. 913 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 38 26. Garsed DW, Pandey A, Fereday S, et al. The genomic and immune landscape of long-914 term survivors of high-grade serous ovarian cancer. Nature genetics 2022; 54(12): 1853-64. 915 27. Stefansson OA, Jonasson JG, Olafsdottir K, et al. CpG island hypermethylation of 916 BRCA1 and loss of pRb as co -occurring events in basal/triple -negative breast cancer. 917 Epigenetics 2011; 6(5): 638-49. 918 28. Jönsson G, Staaf J, Vallon -Christersson J, et al. The Retinoblastoma Gene Undergoes 919 Rearrangements in BRCA1 -Deficient Basal -like Breast Cancer. Cancer Research 2012; 920 72(16): 4028-36. 921 29. Chakraborty G, Armenia J, Mazzu YZ, et al. Significance of BRCA2 and RB1 co-loss 922 in aggressive prostate cancer progression. Clinical Cancer Research 2020; 26(8): 2047-64. 923 30. Chen WS, Alshalalfa M, Zhao SG, et al. Novel RB1-Loss Transcriptomic Signature Is 924 Associated with Poor Clinical Outcomes across Cancer Types. Clinical Cancer Research 2019; 925 25(14): 4290-9. 926 31. Burkhart DL, Sage J. Cellular mechanisms of tumour suppression by the retinoblastoma 927 gene. Nature Reviews Cancer 2008; 8(9): 671-82. 928 32. Knudsen ES, Knudsen KE. Tailoring to RB: tumour suppressor status and therapeutic 929 response. Nature reviews Cancer 2008; 8(9): 714-24. 930 33. Vélez-Cruz R, Manickavinayaham S, Biswas AK, et al. RB localizes to DNA double -931 strand breaks and promotes DNA end resection and homologous recombination through the 932 recruitment of BRG1. Genes and Development 2016; 30(22): 2500-12. 933 34. Millstein J, Budden T, Goode EL, et al. Prognostic gene expression signature for high-934 grade serous ovarian cancer. Annals of Oncology 2020; 31(9): 1240-50. 935 35. Milea A, George SHL, Matevski D, et al. Retinoblastoma pathway deregulatory 936 mechanisms determine clinical outcome in high -grade serous ovarian carcinoma. Modern 937 Pathology 2014; 27(7): 991-1001. 938 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 39 36. Sieh W, Köbel M, Longacre TA, et al. Hormone-receptor expression and ovarian cancer 939 survival: An Ovarian Tumor Tissue Analysis consortium study. The Lancet Oncology 2013; 940 14(9): 853-62. 941 37. Köbel M, Kang EY, Weir A, et al. p53 and ovarian carcinoma survival: an Ovarian 942 Tumor Tissue Analysis consortium study. The Journal of Pathology: Clinical Research 2023; 943 9(3): 208-22. 944 38. Talhouk A, George J, Wang C, et al. Development and Validation of the Gene 945 Expression Predictor of High -grade Serous Ovarian Carcinoma Molecular SubTYPE 946 (PrOTYPE). Clinical Cancer Research 2020; 26(20): 5411-23. 947 39. Ovarian Tumor Tissue Analysis (OTTA) Consortium, Goode EL, Block MS, et al. 948 Dose-Response Association of CD8+ Tumor -Infiltrating Lymphocytes and Survival Time in 949 High-Grade Serous Ovarian Cancer. JAMA oncology 2017; 3(12): e173290-e. 950 40. Nguyen L, W. M. Martens J, Van Hoeck A, Cuppen E. Pan -cancer landscape of 951 homologous recombination deficiency. Nature Communications 2020; 11(1): 1-12. 952 41. Domcke S, Sinha R, Levine DA, Sander C, Schultz N. Evaluating cell lines as tumour 953 models by comparison of genomic profiles. Nature Communications 2013; 4(2126). 954 42. Köbel M, Piskorz AM, Lee S, et al. Optimized p53 immunohistochemistry is an 955 accurate predictor of TP53 mutation in ovarian carcinoma. The Journal of Pathology: Clinical 956 Research 2016; 2(4): 247-58. 957 43. Hollis RL, Thomson JP, Stanley B, et al. Molecular stratification of endometrioid 958 ovarian carcinoma predicts clinical outcome. Nature Communications 2020; 11(1). 959 44. Weinberg RA. The retinoblastoma protein and cell cycle control. Cell 1995; 81(3): 323-960 30. 961 45. Genovese C, Trani D, Caputi M, Claudio PP. Cell cycle control and beyond: emerging 962 roles for the retinoblastoma gene family. Oncogene 2006; 25(38): 5201-9. 963 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 40 46. Findlay GM, Daza RM, Martin B, et al. Accurate classification of BRCA1 variants with 964 saturation genome editing. Nature 2018; 562(7726): 217-22. 965 47. Degasperi A, Amarante TD, Czarnecki J, et al. A practical framework and online tool 966 for mutational signature analyses show intertissue variation and driver dependencies. Nature 967 Cancer 2020; 1(2): 249-63. 968 48. Alexandrov LB, Kim J, Haradhvala NJ, et al. The repertoire of mutational signatures in 969 human cancer. Nature 2020; 578(7793): 94-101. 970 49. Kang EY, Weir A, Meagher NS, et al. CCNE1 and survival of patients with tubo -971 ovarian high-grade serous carcinoma: An Ovarian Tumor Tissue Analysis consortium study. 972 Cancer 2022; 54(4): 538-45. 973 50. da Costa AABA, do Canto LM, Larsen SJ, et al. Genomic profiling in ovarian cancer 974 retreated with platinum based chemotherapy presented homologous recombination deficiency 975 and copy number imbalances of CCNE1 and RB1 genes. BMC Cancer 2019; 19(1): 422-. 976 51. Mandigo AC, Tomlins SA, Kelly WK, Knudsen KE. Relevance of pRB Loss in Human 977 Malignancies. Clin Cancer Res 2022; 28(2): 255-64. 978 52. Ku SY, Rosario S, Wang Y, et al. Rb1 and Trp53 cooperate to suppress prostate cancer 979 lineage plasticity, metastasis, and antiandrogen resistance. Science 2017; 355(6320): 78-83. 980 53. Mu P, Zhang Z, Benelli M, et al. SOX2 promotes lineage plasticity and antiandrogen 981 resistance in TP53 - and RB1 -deficient prostate cancer. Science 2017; 355(6320): 84-8. 982 54. Palafox M, Monserrat L, Bellet M, et al. High p16 expression and heterozygous RB1 983 loss are biomarkers for CDK4/6 inhibitor resistance in ER(+) breast cancer. Nat Commun 2022; 984 13(1): 5258. 985 55. Wander SA, Cohen O, Gong X, et al. The Genomic Landscape of Intrinsic and 986 Acquired Resistance to Cyclin -Dependent Kinase 4/6 Inhibitors in Patients with Hormone 987 Receptor-Positive Metastatic Breast Cancer. Cancer Discov 2020; 10(8): 1174-93. 988 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 41 56. Derenzini M, Donati G, Mazzini G, et al. Loss of Retinoblastoma Tumor Suppressor 989 Protein Makes Human Breast Cancer Cells More Sensitive to Antimetabolite Exposure. 990 Clinical Cancer Research 2008; 14(7): 2199-209. 991 57. Treré D, Brighenti E, Donati G, et al. High prevalence of retinoblastoma protein loss in 992 triple-negative breast cancers and its association with a good prognosis in patients treated with 993 adjuvant chemotherapy. Annals of Oncology 2009; 20(11): 1818-23. 994 58. Patel JM, Goss A, Garber JE, et al. Retinoblastoma protein expression and its predictors 995 in triple-negative breast cancer. NPJ breast cancer 2020; 6(1): 19-. 996 59. Bowtell DD. The genesis and evolution of high -grade serous ovarian cancer. Nat Rev 997 Cancer 2010; 10(11): 803-8. 998 60. Cancer Genome Atlas N. Comprehensive molecular portraits of human breast tumours. 999 Nature 2012; 490(7418): 61-70. 1000 61. Köbel M, Kalloger SE, Boyd N, et al. Ovarian carcinoma subtypes are different 1001 diseases: implications for biomarker studies. PLoS medicine 2008; 5(12): e232-e. 1002 62. Kang EY, Millstein J, Popovic G, et al. MCM3 is a novel proliferation marker 1003 associated with longer survival for patients with tubo -ovarian high-grade serous carcinoma. 1004 Virchows Archiv 2022; 480(4): 855-71. 1005 63. Zhao Y, Wang Y, Zhu F, Zhang J, Ma X, Zhang D. Gene expression profiling revealed 1006 MCM3 to be a better marker than Ki67 in prognosis of invasive ductal breast carcinoma 1007 patients. Clinical and Experimental Medicine 2020; 20(2): 249-59. 1008 64. Velez-Cruz R, Manickavinayaham S, Biswas AK, et al. RB localizes to DNA double -1009 strand breaks and promotes DNA end resection and homologous recombination through the 1010 recruitment of BRG1. Genes Dev 2016; 30(22): 2500-12. 1011 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 42 65. Westphalen CB, Fine AD, André F, et al. Pan -cancer Analysis of Homologous 1012 Recombination Repair –associated Gene Alterations and Genome -wide Loss -of-1013 Heterozygosity Score. Clinical Cancer Research 2022; 28(7): 1412-21. 1014 66. Meng J, Liu X, Zhang P, et al. Rb selectively inhibits innate IFN -β production by 1015 enhancing deacetylation of IFN -β promoter through HDAC1 and HDAC8. Journal of 1016 Autoimmunity 2016; 73: 42-53. 1017 67. Manzano RG, Catalan-Latorre A, Brugarolas A. RB1 and TP53 co-mutations correlate 1018 strongly with genomic biomarkers of response to immunity checkpoint inhibitors in urothelial 1019 bladder cancer. BMC Cancer 2021; 21(432). 1020 68. Molinero L, Li Y, Chang C -W, et al. Tumor immune microenvironment and genomic 1021 evolution in a patient with metastatic triple negative breast cancer and a complete response to 1022 atezolizumab. Journal for ImmunoTherapy of Cancer 2019; 7(274). 1023 69. Ahmed AA, Etemadmoghadam D, Temple J, et al. Driver mutations in TP53 are 1024 ubiquitous in high grade serous carcinoma of the ovary. Journal of Pathology 2010; 221(1): 1025 49-56. 1026 70. Landen CN, Molinero L, Hamidi H, et al. Influence of Genomic Landscape on Cancer 1027 Immunotherapy for Newly Diagnosed Ovarian Cancer: Biomarker Analyses from the 1028 IMagyn050 Randomized Clinical Trial. Clinical Cancer Research 2023; 29(9): 1698-707. 1029 71. Kandalaft LE, Odunsi K, Coukos G. Immune Therapy Opportunities in Ovarian 1030 Cancer. American Society of Clinical Oncology Educational Book 2020; 3(40): e228-e40. 1031 72. Bulanova D, Akimov Y, Senkowski W, et al. A synthetic lethal dependency on casein 1032 kinase 2 in response to replication-perturbing drugs in RB1-deficient ovarian and breast cancer 1033 cells. bioRxiv 2022: 1-22. 1034 73. Gong X, Du J, Parsons SH, et al. Aurora A Kinase Inhibition Is Synthetic Lethal with 1035 Loss of the RB1 Tumor Suppressor Gene. Cancer Discov 2019; 9(2): 248-63. 1036 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint 43 74. Lyu J, Yang EJ, Zhang B, et al. Synthetic lethality of RB1 and aurora A is driven by 1037 stathmin-mediated disruption of microtubule dynamics. Nat Commun 2020; 11(1): 5105. 1038 75. Oser MG, Fonseca R, Chakraborty AA, et al. Cells Lacking the RB1 Tumor Suppressor 1039 Gene Are Hyperdependent on Aurora B Kinase for Survival. Cancer Discov 2019; 9(2): 230-1040 47. 1041 76. Serra V, Wang AT, Castroviejo -Bermejo M, et al. Identification of a Molecularly -1042 Defined Subset of Breast and Ovarian Cancer Models that Respond to WEE1 or ATR 1043 Inhibition, Overcoming PARP Inhibitor Resistance. Clinical Cancer Research 2022; 28(20): 1044 4536-50. 1045 1046 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint HGSC ( n = 4482) LGSC ( n = 181) MOC ( n = 326) ENOC ( n = 908) CCOC ( n = 667) 0 20 40 60 80 100Proportion (%) RB1 loss RB1 retained Chi-square P < 0.0001 RB1 retained ( n = 2183) RB1 loss ( n = 369) -8 -6 -4 -2 0 RB1 mRNA expression Mann-Whitney P < 0.0001 ✱✱✱✱ Figure 1. B C D E F G RB1 loss ( n = 218) RB1 retained ( n = 916) 0 20 40 60 80 100Proportion (%) BRCA1 mutation carriers BRCA2 mutation carriers Non-carriers Chi-square P < 0.0001 RB1 retained RB1 loss RB1 subclonal loss A . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint Table 1. Multivariate analysis of molecular alterations and overall survival in patients with HGSC and ENOC Histotype Feature Category No. patients (events, %) HR (95% CI) P P for interaction HGSCa,b RB1 Retained 3453 (71.3) 1 [Reference] Loss 686 (61.1) 0.74 (0.66-0.83) 6.8 x 10-7 ENOCa RB1 Retained 649 (22.7) 1 [Reference] Loss 28 (39.3) 2.17 (1.17-4.03) 0.014 HGSCa,b RB1 and BRCA status RB1 retained & non-carrier 714 (76.3) 1 [Reference] 0.24 RB1 loss & non-carrier 135 (60.7) 0.74 (0.57-0.96) 0.023 RB1 retained & BRCA carrier 159 (67.9) 0.69 (0.55-0.86) 0.001 RB1 loss & BRCA carrier 70 (42.9) 0.38 (0.25-0.58) 5.2 x 10-6 ENOCa RB1 and p53 RB1 retained & p53 normal 492 (17.5) 1 [Reference] 0.698 RB1 retained & p53 abnormal 58 (36.2) 2.26 (1.38-3.71) 0.001 RB1 loss & p53 normal 11 (27.3) 1.77 (0.56-5.65) 0.332 RB1 loss & p53 abnormal 12 (58.3) 5.34 (2.43-11.8) <0.001 aAdjusted for stage and age at diagnosis. bStratified by study. HR, hazard ratio, CI, confidence interval; HGSC, tubo-ovarian high-grade serous carcinoma; ENOC, endometrioid ovarian cancer. . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint C Figure 2. A B RB1 WT RB1 KO RB1 WT RB1 KO RB1 WT RB1 KO DMF Cisplatin Paclitaxel Cis/Pac -6 -4 -2 0 0.0 0.5 1.0 Log[Cisplatin], µM Viability AOCS1 (BRCA1/2 WT, p16 normal) -6 -4 -2 0 0.0 0.5 1.0 Log[Cisplatin], µM Viability AOCS16 (BRCA1 mut, p16 absent) -6 -4 -2 0 0.0 0.5 1.0 Log[Cisplatin], µM Viability AOCS7.2 (BRCA1 mut, p16 normal) -8 -6 -4 -2 0 0.0 0.5 1.0 1.5 Log[Paclitaxel], µM Viability -8 -6 -4 -2 0 0.0 0.5 1.0 1.5 Log[Paclitaxel], µM Viability -8 -6 -4 -2 0 0.0 0.5 1.0 1.5 Log[Paclitaxel], µM Viability ns ** ns ns **** **** -6 -4 -2 0 0.0 0.5 1.0 Log[Olaparib], µM Viability RB1 WT RB1 KO -6 -4 -2 0 0.0 0.5 1.0 Log[Olaparib], µM Viability RB1 WT RB1 KO -6 -4 -2 0 0.0 0.5 1.0 Log[Olaparib], µM Viability RB1 WT RB1 KO ns **** ** DMF Cisplatin Paclitaxel Cis/Pac 0 50 100% Clonogenic Survival DMF Cisplatin Paclitaxel Cis/Pac 0 50 100% Clonogenic Survival DMF CisplatinPaclitaxel Cis/Pac 0 50 100% Clonogenic Survival AOCS1 BRCA1/2 WT, p16 normal AOCS7.2 BRCA1 mut, p16 normal AOCS16 BRCA1 mut, p16 absent ✱✱ ✱ ns ns ns ns ns ns ns RB1 WT RB1 KO RB1 GAPDH Scramble RB1 WT RB1 KO AOCS1 (BRCA1/2 WT p16 normal) pRB1 Scramble RB1 WT RB1 KO AOCS7.2 (BRCA1 mut p16 normal) Scramble RB1 WT RB1 KO AOCS16 (BRCA1 mut p16 absent) RB1 GAPDH pRB1 RB1 GAPDH pRB1 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint HRP & RB1wild−type (n=45) HRP & RB1altered (n=4) BRCA1 −HRD & RB1wild−type (n=36) BRCA1 −HRD & RB1altered (n=13) BRCA2 −HRD & RB1wild−type (n=20) BRCA2 −HRD & RB1altered (n=8) 2 4 6 Survival group CHORD score SV % 0.25 0.5 0.75 1 25 50 75 100 41% 14% 12% 6% 5% 4% 3% 2% 2% 2% 2% 2% 2% 1% 1% 1% 1% 1% 1% 1% 20% 17% 11% 5% 3% 3% 2% 2% BRCA1 BRCA2 RAD51B RAD51C BRIP1 A TM P ALB2 BLM FANCD2 A TR BARD1 FANCI FANCM CHEK2 FANCA FANCE MSH2 MSH6 PMS1 RAD51D RB1 CCNE1 PTEN CDK12 CXCL9 CXCL10 CXCL11 IFNG 0 25 50 0 50 100 Alteration Germline duplication Germline deletion Germline inversion Somatic duplication Somatic deletion Somatic inversion Somatic interchromosomal translocation Somatic amplification Germline frameshift indel Germline nonsense Germline missense Somatic nonsense Somatic frameshift indel Somatic in−frame indel Somatic splice site Promoter methylation Survival group LTS MTS STS CHORD None BRCA2 −type BRCA1 −type SV % DUP DEL INV ITX Alteration count Alteration % Figure 3. A B . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint E2f Targets G2m Checkpoint Interferon Gamma Response Interferon Alpha Response Oxidative Phosphorylation Estrogen Response Early Estrogen Response Late Tnfa Signaling Via Nfkb Reactive Oxygen Species Pathway Adipogenesis Myc Targets V1 P53 Pathway Fatty Acid Metabolism Mtorc1 Signaling Allograft Rejection Inflammatory Response Complement Il6 Jak Stat3 Signaling Epithelial Mesenchymal Transition Protein Secretion Mitotic Spindle Spermatogenesis Coagulation Peroxisome Il2 Stat5 Signaling Xenobiotic Metabolism Apoptosis Pi3k Akt Mtor Signaling Cholesterol Homeostasis Myc Targets V2 Uv Response Up Dna Repair Glycolysis Kras Signaling Dn Bile Acid Metabolism Hypoxia Uv Response Dn Kras Signaling Up Wnt Beta Catenin Signaling Myogenesis Apical Junction Heme Metabolism HRD/HRP (sans RB1 sig) BRCA2/HRP (sans RB1 sig) BRCA1/HRP (sans RB1 sig) BRCA2.RB1w/HRP.RB1wt BRCA1.RB1w/HRP.RB1wt BRCA2.RB1m/HRP.RB1wt BRCA1.RB1m/HRP.RB1wt RB1m/RB1wt (sans HRD type) −log10(padj) 10 20 30 40 −4 −2 0 2 4 NES Direction Up Down MSigDB HALLMARK fGSEA Results (P.adj <= 0.05) A Figure 4. B E2f Targets G2m Checkpoint Interferon Gamma Response Interferon Alpha Response Oxidative Phosphorylation Estrogen Response Early Estrogen Response Late Tnfa Signaling Via Nfkb Reactive Oxygen Species Pathway Adipogenesis Myc Targets V1 P53 Pathway Fatty Acid Metabolism Mtorc1 Signaling Allograft Rejection Inflammatory Response Complement Il6 Jak Stat3 Signaling Epithelial Mesenchymal Transition Protein Secretion Mitotic Spindle Spermatogenesis Coagulation Peroxisome Il2 Stat5 Signaling Xenobiotic Metabolism Apoptosis Pi3k Akt Mtor Signaling Cholesterol Homeostasis Myc Targets V2 Uv Response Up Dna Repair Glycolysis Kras Signaling Dn Bile Acid Metabolism Hypoxia Uv Response Dn Kras Signaling Up Wnt Beta Catenin Signaling Myogenesis Apical Junction Heme Metabolism HRD/HRP (sans RB1 sig) BRCA2/HRP (sans RB1 sig) BRCA1/HRP (sans RB1 sig) BRCA2.RB1w/HRP.RB1wt BRCA1.RB1w/HRP.RB1wt BRCA2.RB1m/HRP.RB1wt BRCA1.RB1m/HRP.RB1wt RB1m/RB1wt (sans HRD type) −log10(padj) 10 20 30 40 −4 −2 0 2 4 NES Direction Up Down MSigDB HALLMARK fGSEA Results (P.adj <= 0.05) E2f Targets G2m Checkpoint Interferon Gamma Response Interferon Alpha Response Oxidative Phosphorylation Estrogen Response Early Estrogen Response Late Tnfa Signaling Via Nfkb Reactive Oxygen Species Pathway Adipogenesis Myc Targets V1 P53 Pathway Fatty Acid Metabolism Mtorc1 Signaling Allograft Rejection Inflammatory Response Complement Il6 Jak Stat3 Signaling Epithelial Mesenchymal Transition Protein Secretion Mitotic Spindle Spermatogenesis Coagulation Peroxisome Il2 Stat5 Signaling Xenobiotic Metabolism Apoptosis Pi3k Akt Mtor Signaling Cholesterol Homeostasis Myc Targets V2 Uv Response Up Dna Repair Glycolysis Kras Signaling Dn Bile Acid Metabolism Hypoxia Uv Response Dn Kras Signaling Up Wnt Beta Catenin Signaling Myogenesis Apical Junction Heme Metabolism HRD/HRP (sans RB1 sig) BRCA2/HRP (sans RB1 sig) BRCA1/HRP (sans RB1 sig) BRCA2.RB1w/HRP.RB1wt BRCA1.RB1w/HRP.RB1wt BRCA2.RB1m/HRP.RB1wt BRCA1.RB1m/HRP.RB1wt RB1m/RB1wt (sans HRD type) −log10(padj) 10 20 30 40 −4 −2 0 2 4 NES Direction Up Down MSigDB HALLMARK fGSEA Results (P.adj <= 0.05) E2f Targets G2m Checkpoint Interferon Gamma Response Interferon Alpha Response Oxidative Phosphorylation Estrogen Response Early Estrogen Response Late Tnfa Signaling Via Nfkb Reactive Oxygen Species Pathway Adipogenesis Myc Targets V1 P53 Pathway Fatty Acid Metabolism Mtorc1 Signaling Allograft Rejection Inflammatory Response Complement Il6 Jak Stat3 Signaling Epithelial Mesenchymal Transition Protein Secretion Mitotic Spindle Spermatogenesis Coagulation Peroxisome Il2 Stat5 Signaling Xenobiotic Metabolism Apoptosis Pi3k Akt Mtor Signaling Cholesterol Homeostasis Myc Targets V2 Uv Response Up Dna Repair Glycolysis Kras Signaling Dn Bile Acid Metabolism Hypoxia Uv Response Dn Kras Signaling Up Wnt Beta Catenin Signaling Myogenesis Apical Junction Heme Metabolism HRD/HRP (sans RB1 sig) BRCA2/HRP (sans RB1 sig) BRCA1/HRP (sans RB1 sig) BRCA2.RB1w/HRP.RB1wt BRCA1.RB1w/HRP.RB1wt BRCA2.RB1m/HRP.RB1wt BRCA1.RB1m/HRP.RB1wt RB1m/RB1wt (sans HRD type) −log10(padj) 10 20 30 40 −4 −2 0 2 4 NES Direction Up Down MSigDB HALLMARK fGSEA Results (P.adj <= 0.05) C RB1 retained & non-carrier ( n = 602) RB1 retained & BRCA carrier ( n = 94) RB1 loss & non-carrier ( n = 123) RB1 loss & BRCA carrier ( n = 49) 0 20 40 60 80 100Proportion Negative Low Moderate High Chi-square P = 0.0264 Number of TILs RB1 retained & non-carrier ( n = 414) RB1 retained & BRCA carrier ( n = 86) RB1 loss & non-carrier ( n = 64) RB1 loss & BRCA carrier ( n = 37) 0 20 40 60 80 100Proportion C1.MES C2.IMM C4.DIF C5.PRO Chi-square P = 0.0271 Molecular subtype . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint Supplementary Figure S1. 7436 patients with ovarian carcinomas assessed by immunohistochemistry for tumour RB1 protein expression 6564 patients classified as having retained or lost RB1 tumour protein expression 872 excluded 66 had subclonal RB1 loss 17 had cytoplasmic RB1 expression 789 uninterpretable expression 4482 HGSC patients with retained or lost RB1 protein expression Survival analyses: 4256 univariate 4139 multivariate 2552 patients with tumour RB1 mRNA expression determined by NanoString 1134 BRCA germline mutation status established Survival analyses: 1119 univariate 1078 multivariate 908 ENOC patients with retained or lost RB1 protein expression Survival analyses: 718 univariate 677 multivariate 868 patients with CD8+ tumour infiltrating lymphocyte counts 601 patients with HGSC molecular subtypes from adnexal and presumed adnexal specimens Ovarian Tumor Tissue Analysis (OTTA) consortium Multidisciplinary Ovarian Cancer Outcomes Group (MOCOG) study 126 patients with advanced stage (IIIC/IV) HGSC assessed by whole-genome sequencing and RNA sequencing of primary tumours 34 short-term survivors (OS <2 years) 32 moderate-term survivors (OS ≥2 and <10 years) 60 long-term survivors (OS ≥10 years) 759 with p53 abnormal or normal protein expression Survival analyses: 609 univariate 573 multivariate . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint Supplementary Figure S2. A B RB1 loss ( n = 28) RB1 retained ( n = 731) 0 20 40 60 80 100Proportion (%) p53 abnormal p53 normal Chi-square P < 0.0001 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint A B C Supplementary Figure S3. . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint IHC p16 0.0 0.5 1.0 1.5 RB1 ex- pression Scramble RB1 WT RB1 KO AOCS1 (BRCA1/2 WT) Scramble RB1 WT RB1 KO AOCS7.2 (BRCA1 mut) Scramble RB1 WT RB1 KO AOCS16 (BRCA1 mut) p16 GAPDH A B C Supplementary Figure S4. 0 24 48 72 96 120 144 168 192 216 0 25 50 75 100 AOCS7.2 (BRCA1 mut, p16 normal) Time (hrs) Confluence (%) 0 24 48 72 96 120 144 168 192 216 0 25 50 75 100 AOCS1 (BRCA1/2 WT, p16 normal) Time (hrs) Confluence (%) 0 24 48 72 96 120 144 168 192 216 0 25 50 75 100 AOCS16 (BRCA1 mut, p16 absent) Time (hrs) Confluence (%) RB1 Wildtype RB1 Knockout RB1 WT RB1 KO 0 20 40 60 80 100% cells DMF AOCS1 (BRCA1/2 WT, p16 normal) RB1 WT RB1 KO 0 20 40 60 80 100% cells AOCS16 (BRCA1 mut, p16 absent) RB1 WT RB1 KO 0 20 40 60 80 100% cells AOCS7.2 (BRCA1 mut, p16 normal) RB1 WT RB1 KO 0 20 40 60 80 100% cells Cisplatin RB1 WT RB1 KO 0 20 40 60 80 100% cells RB1 WT RB1 KO 0 20 40 60 80 100% cells RB1 WT RB1 KO 0 20 40 60 80 100% cells Paclitaxel RB1 WT RB1 KO 0 20 40 60 80 100% cells RB1 WT RB1 KO 0 20 40 60 80 100% cells Sub G0G1 % G0G1 % S % G2M % Sub G2M % ns ns * ns ns ns ns ns ns . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint Supplementary Figure S5. <0.0001 <0.0001 <0.0001 <0.0001 0.5990 <0.0001 <0.0001 <0.0001 <0.0001 0.2420 Kruskal, P < 0.0001 0.0 0.5 1.0 1.5 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment id.ID6 <0.0001 <0.0001 0.1960 0.0290 0.0390 <0.0001 <0.0001 <0.0001 0.0003 0.2350 Kruskal, P < 0.0001 0 1 2 3 4 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment sv.Ovary_G <0.0001 <0.0001 <0.0001 <0.0001 0.3180 0.0004 0.1390 0.0004 0.0280 0.1510 Kruskal, P < 0.0001 0.0 0.3 0.6 0.9 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment id.ID8 <0.0001 <0.0001 <0.0001 <0.0001 0.6460 0.4870 0.1850 0.4770 0.1850 0.4870 Kruskal, P < 0.0001 0.3 0.6 0.9 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment id.ID1 <0.0001 <0.0001 <0.0001 <0.0001 0.9020 0.1180 0.7700 0.1560 0.9720 0.1000 Kruskal, P < 0.0001 0.0 0.2 0.4 0.6 0.8 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment sbs.SBS3 <0.0001 <0.0001 <0.0001 <0.0001 0.9860 0.9860 0.9860 0.9860 0.9860 0.9860 Kruskal, P < 0.0001 0.0 0.2 0.4 0.6 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment sbs.SBS1 0.1110 0.5160 <0.0001 <0.0001 0.5160 <0.0001 <0.0001 <0.0001 0.0003 0.5160 Kruskal, P < 0.0001 0 1 2 3 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment sv.Ovary_A <0.0001 <0.0001 <0.0001 0.0008 0.7270 0.8670 0.3380 0.8700 0.8670 0.7270 Kruskal, P < 0.0001 0.5 1.0 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment sbs.SBS5 <0.0001 0.0060 0.0500 0.1380 0.0820 0.0010 0.0020 0.3460 0.5070 0.8590 Kruskal, P < 0.0001 0.0 0.4 0.8 1.2 1.6 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment dbs.DBS2 <0.0001 0.0008 <0.0001 <0.0001 0.9220 0.3010 0.2800 0.2800 0.2800 0.9800 Kruskal, P < 0.0001 0.0 0.2 0.4 0.6 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment id.ID4 0.0002 <0.0001 0.0002 0.0002 0.4250 0.7670 0.2680 0.8420 0.4250 0.4250 Kruskal, P < 0.0001 0.4 0.8 1.2 1.6 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment id.ID2 0.0007 0.0010 0.0007 0.0100 0.8840 0.9280 0.8840 0.8840 0.8840 0.8840 Kruskal, P < 0.0001 0 1 2 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment sv.Ovary_C 0.0040 0.0850 0.0004 0.0030 0.8380 0.0850 0.0850 0.1030 0.0850 0.8380 Kruskal, P < 0.0001 0.0 0.2 0.4 0.6 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment id.ID10 0.0140 0.4870 <0.0001 0.0090 0.3980 0.0090 0.2030 0.0060 0.1060 0.6360 Kruskal, P < 0.0001 0.00 0.25 0.50 0.75 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment id.ID12 0.0170 0.7220 0.0010 0.0130 0.0730 0.0030 0.0730 0.0050 0.0210 0.5120 Kruskal, P < 0.0001 0.0 0.1 0.2 0.3 0.4 0.5 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment dbs.DBS4 0.0001 0.0130 0.1530 0.1420 0.6630 0.0130 0.3820 0.1420 0.6440 0.3820 Kruskal, P < 0.0001 0.0 0.5 1.0 1.5 2.0 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment sv.Ovary_F 0.0050 0.0040 0.0690 0.1080 0.0690 0.8580 0.8580 0.0700 0.1080 0.8580 Kruskal, P = 0.0003 0 1 2 3 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment sv.Ovary_B 0.0010 0.0180 0.1010 0.2610 0.8730 0.2610 0.2610 0.2610 0.2610 0.7210 Kruskal, P = 0.0006 0 1 2 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment sv.Ovary_D 0.0070 0.4220 0.0630 0.3030 0.4220 0.6840 1.0000 0.4640 0.4640 0.6960 Kruskal, P = 0.0089 0.0 0.1 0.2 0.3 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment id.ID5 0.5190 0.0820 0.2500 0.0820 0.0850 0.5900 0.0820 0.2500 0.2500 0.0820 Kruskal, P = 0.0258 0.0 0.2 0.4 0.6 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment dbs.DBS6 0.1770 0.1770 0.1770 0.2280 0.6210 0.6210 0.6210 1.0000 1.0000 1.0000 Kruskal, P = 0.0464 0 1 2 3 4 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment sv.Ovary_E 0.1470 0.1470 0.1470 0.1470 0.9310 0.9790 0.8600 0.9310 0.8600 0.8600 Kruskal, P = 0.0637 0.0 0.2 0.4 0.6 0.8 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment dbs.DBS7 0.0800 0.5950 0.5150 0.5950 0.5950 0.5950 0.9070 1.0000 1.0000 1.0000 Kruskal, P = 0.0984 0.0 0.4 0.8 1.2 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment dbs.DBS11 0.5480 0.6970 0.1030 0.6970 0.9200 0.1800 0.9760 0.2890 0.9200 0.5320 Kruskal, P = 0.0987 0.0 0.1 0.2 0.3 0.4 0.5 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment sbs.SBS39 0.0810 0.8330 0.5570 0.5570 0.5570 0.5760 0.8330 0.8330 0.8330 0.8620 Kruskal, P = 0.0987 0.0 0.1 0.2 0.3 0.4 0.5 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment sbs.SBS8 0.0750 0.5630 0.9450 0.7860 0.9450 0.5450 0.7860 0.7270 0.7860 0.7860 Kruskal, P = 0.1059 0.0 0.1 0.2 0.3 0.4 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment sbs.SBS40 0.9500 0.5550 0.5550 0.5550 0.5550 0.5550 0.5550 0.9500 0.9500 0.9500 Kruskal, P = 0.5160 0.00 0.25 0.50 0.75 1.00 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Signature Relative Enrichment dbs.DBS9 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint Supplementary Figure S6. <0.0001 <0.0001 <0.0001 <0.0001 0.2100 <0.0001 <0.0001 <0.0001 0.0002 0.4160 Kruskal, P < 0.0001 0 1 2 3 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Value CHORD BRCA1 Signature proportion <0.0001 <0.0001 <0.0001 <0.0001 0.2670 <0.0001 <0.0001 <0.0001 <0.0001 0.1630 Kruskal, P < 0.0001 0 1 2 3 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Value CHORD BRCA2 Signature proportion <0.0001 <0.0001 <0.0001 <0.0001 0.5280 0.5280 0.0130 0.9560 0.0450 0.0860 Kruskal, P < 0.0001 0 1 2 3 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Value CHORD Total HRD proportion <0.0001 <0.0001 <0.0001 0.0010 0.8560 0.0940 0.0180 0.2090 0.0650 0.5290 Kruskal, P < 0.0001 20 40 60 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Value scarHRD Mean <0.0001 <0.0001 <0.0001 <0.0001 0.6340 0.0020 0.0110 0.0140 0.0110 0.6210 Kruskal, P < 0.0001 500 1000 1500 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Value Number of SVs (DEL) <0.0001 0.0010 0.0340 0.0210 0.7000 <0.0001 <0.0001 <0.0001 0.0004 0.5640 Kruskal, P < 0.0001 500 1000 1500 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Value Number of SVs (DUP) <0.0001 0.0003 0.0004 0.0160 0.9710 0.9710 0.5840 0.9710 0.3140 0.5840 Kruskal, P < 0.0001 500 1000 1500 2000 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Value Number of Not Clustered SVs 0.0005 0.0005 0.0740 0.1200 0.1200 0.5160 0.4460 0.0870 0.0870 0.8980 Kruskal, P = 0.0001 0 100 200 300 400 500 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Value Number of Clustered SVs 0.4830 <0.0001 0.1320 0.0280 <0.0001 0.0650 0.0250 0.1830 0.5000 0.4760 Kruskal, P = 0.0001 6 8 10 12 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Value PCNA expression 0.0003 0.0020 0.0110 0.1870 0.6500 0.6500 0.3720 0.6500 0.0810 0.5940 Kruskal, P = 0.0001 100 200 300 400 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Value Number of SVs (ITX) 0.4190 <0.0001 0.1320 0.0270 <0.0001 0.2880 0.0270 0.0100 0.4160 0.2600 Kruskal, P = 0.0001 6 8 10 12 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Value MCM3 expression 0.0005 0.0005 0.1040 0.1570 0.2260 0.2260 0.3660 0.0900 0.0900 0.7460 Kruskal, P = 0.0002 50 100 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Value % Whole Genome LOH 0.0005 0.0080 0.0060 0.0760 0.9550 0.9300 0.6800 0.9300 0.6800 0.6800 Kruskal, P = 0.0002 500 1000 1500 2000 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Value Total SVs 0.0190 0.1360 0.0040 0.0220 0.5480 0.1270 0.2320 0.1360 0.0970 0.5120 Kruskal, P = 0.0006 200 400 600 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Value Number of Neoantigens 0.0002 0.1520 0.2490 0.2240 0.2240 0.1520 0.2490 0.7390 0.7710 0.7390 Kruskal, P = 0.0008 50 75 100 125 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Value Age at diagnosis 0.0910 0.1130 0.0040 0.0460 0.5160 0.0460 0.1850 0.5160 0.6080 0.9800 Kruskal, P = 0.0018 20 40 60 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Value Overall Survival (years) 0.0060 0.0280 0.0060 0.0690 0.8850 0.8850 0.8850 0.7060 0.8850 0.8850 Kruskal, P = 0.0019 50 100 150 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Value % Whole Genome Loss 0.0250 0.0250 0.8190 0.3130 0.2450 0.1380 0.6540 0.0370 0.1510 0.8190 Kruskal, P = 0.0052 100 200 300 400 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Value Number of SVs (INV) 0.1400 0.1400 0.0600 0.0600 0.3580 0.1400 0.0640 0.9860 0.3730 0.3580 Kruskal, P = 0.0075 10 20 30 40 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Value Progression−free Survival (years) 0.2190 0.3470 0.0300 0.2620 0.4090 0.2190 0.4090 0.2940 0.3080 0.4090 Kruskal, P = 0.0214 5 10 15 20 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Value Variants Per Mb 0.3620 0.1030 0.0660 0.3620 0.6520 0.6520 0.7630 0.6830 0.6840 0.9800 Kruskal, P = 0.0536 2 4 6 8 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Value Ploidy 0.2100 0.2100 0.8160 0.8160 0.5860 0.5860 0.5860 0.3500 0.3500 0.8160 Kruskal, P = 0.1298 2.5 5.0 7.5 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Value CIBERSORTx Absolute score 0.3420 0.6400 0.6400 0.6400 0.9720 0.9720 0.9720 0.9720 0.9720 0.9720 Kruskal, P = 0.3297 0 50 100 150 200 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Value % Ki67 positive 0.9490 0.9490 0.9490 0.9490 0.9490 0.9490 0.9490 0.9490 0.9720 0.9490 Kruskal, P = 0.7540 0.5 1.0 1.5 HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) Group Value Purity . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint Supplementary Figure S7. Fisher, p = 0.000009 Fisher, p = 0.074913 Fisher, p = 0.545260 Fisher, p = 0.749909 Fisher, p = 1.000000 Fisher, p = 0.000009 Fisher, p = 0.260249 Fisher, p = 0.567777 Fisher, p = 0.812293 Fisher, p = 1.000000 Fisher, p = 0.000009 Fisher, p = 0.545260 Fisher, p = 0.567777 Fisher, p = 0.812293 Fisher, p = 1.000000 Fisher, p = 0.000607 Fisher, p = 0.545260 Fisher, p = 0.567777 Fisher, p = 0.827347 Fisher, p = 0.012501 Fisher, p = 0.545260 Fisher, p = 0.576173 Fisher, p = 0.854705 Fisher, p = 0.057852 Fisher, p = 0.545260 Fisher, p = 0.729854 Fisher, p = 0.865845 FANCA FANCE FANCI KMT2C MSH6 CHEK2 ATR ATM FANCD2 BARD1 RAD51D MSH2 PMS1 CDK12 BLM PTEN NF1 PIK3CA P ALB2 RAD51B FANCM BRCA1 BRCA2 RB1 CCNE1 BRIP1 RAD51C HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) HRP & RB1 wild−type (n=45) BRCA1−HRD & RB1 wild−type (n=36) BRCA1−HRD & RB1 altered (n=13) BRCA2−HRD & RB1 wild−type (n=20) BRCA2−HRD & RB1 altered (n=8) 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 Status WT MUT . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint Supplementary Figure S8. HRD/HRP (sans RB1 sig) BRCA2/HRP (sans RB1 sig)BRCA1/HRP (sans RB1 sig)BRCA2.RB1wt/HRP.RB1wtBRCA1.RB1wt/HRP.RB1wtBRCA2.RB1m/HRP.RB1wtBRCA1.RB1m/HRP.RB1wt RB1m/RB1wt (sans HRD type) 0 1000 2000 3000 4000 Comparison groups Number of differentially expressed genes (adjusted P < 0.05) Direction Down Up . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted November 10, 2023. ; https://doi.org/10.1101/2023.11.09.23298321doi: medRxiv preprint