The first-line setting represents the only opportunity for cure in advanced OC. 5-y PFS rates may predict cure as most pts not relapsing at 5 years are potentially cured. In the final overall survival (OS) analysis of the Phase III PAOLA-1/ENGOT-ov25 trial (NCT02477644), maintenance olaparib + bev provided a clinically meaningful OS benefit vs placebo (pbo) + bev in pts with newly diagnosed advanced OC who tested positive for homologous recombination deficiency (HRD+; tumour BRCA1/2 mutation [tBRCAm] and/or genomic instability) regardless of tBRCAm status (Ray-Coquard et al. Ann Oncol 2022:LBA29). We evaluated 5-y PFS by molecular subgroup to help inform the potential for cure.
Pts in response after platinum-based chemotherapy + bev were randomized 2:1 to olaparib (300 mg bid for up to 24 months) + bev (15 mg/kg q3w for up to 15 months) or pbo + bev. A post hoc, updated descriptive PFS analysis by molecular subgroup was conducted at the final OS data cutoff (DCO).
537 pts were randomized to olaparib + bev and 269 to pbo + bev (median PFS follow-up in censored pts was 56.7 and 57.8 months, respectively). At DCO (22 March 2022), the 5-y PFS rate was 46% with olaparib + bev vs 19% with pbo + bev in the HRD+ group, 50% vs 25%, respectively, in the tBRCAm group, 41% vs 15%, respectively, in the HRD+ excluding tBRCAm group, 13% vs 13%, respectively, in the HRD−/unknown group, and 8% vs 12%, respectively, in the HRD− group (Table). Updated data for time to first or second subsequent therapy or death will be presented.
No. of events/no. of pts | 5-y Median PFS | HR (95% CI) | 5-y PFS rate,* (%) | ||||
Olaparib + bev | Pbo + bev | Olaparib + bev | Pbo + bev | Olaparib + bev | Pbo + bev | ||
ITT | 366/537 | 222/269 | 22.9 | 16.6 | 0.63 (0.53–0.74) | 29 | 16 |
HRD+ | 136/255 | 104/132 | 46.8 | 17.6 | 0.41 (0.32–0.54) | 46 | 19 |
tBRCAm | 78/157 | 58/80 | 60.7 | 21.7 | 0.45 (0.32–0.64) | 50 | 25 |
HRD+ excluding tBRCAm | 58/97 | 46/55 | 30.0 | 16.6 | 0.47 (0.32–0.70) | 41 | 15 |
HRD−/unknown | 230/282 | 118/137 | 17.3 | 16.0 | 0.90 (0.72–1.13) | 13 | 13 |
HRD− | 167/192 | 74/85 | 16.6 | 16.2 | 1.01 (0.77–1.33) | 8 | 12 |
Investigator assessed (RECIST 1.1). tBRCAm status by central labs and HRD status by Myriad MyChoice HRD Plus. |
The 5-y updated descriptive PFS rates suggest the benefit of adding olaparib to bev is sustained beyond end of treatment and may indicate the potential of cure in HRD+ newly diagnosed advanced OC, supporting use of maintenance olaparib + bev as a standard of care in pts with HRD+ tumours, regardless of tBRCAm status.
NCT02477644 (23 June 2015)
Medical writing assistance was provided by Abbie Newman BSc, at Cence, funded by AstraZeneca and Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA.
The PRIME (NCT03709316) study was a randomized, double-blind, placebo-controlled, phase 3 trial. Eligible patients were randomised (2:1) to receive niraparib or placebo. Niraparib maintenance therapy with an individualized starting dose (ISD) improved PFS vs placebo (HR 0.45; 95% CI, 0.34–0.60) in Chinese patients with newly diagnosed, advanced ovarian cancer and CR or PR to 1L platinum-based chemotherapy. This post-hoc analysis aims to evaluate niraparib efficacy by the timing of debulking surgery and postoperative residual disease status.
This analysis reports PFS and HRs based on surgical timing (primary debulking surgery [PDS] and interval debulking surgery [IDS]) and residual disease status [optimal(R0/R1) vs suboptimal(R2)]. Median PFS (assessed by BICR) and HRs were obtained using the Kaplan-Meier method and stratified Cox proportional hazards models, respectively.
The DCO date was 30 Sep. 2021. In total, 255 patients were randomized and treated to niraparib (134 PDS, 121 IDS) and 129 to placebo (70 PDS, 59 IDS). The PFS median follow-up was 27.5 months. The efficacy results are shown in Table 1. Niraparib significantly prolonged PFS compared with placebo irrespective of surgical timing. For PFS, HR 0.63 (95% CI 0.42–0.94, median PFS was not reached in niraparib arm vs. 12.0 months in placebo arm) in patients who underwent PDS; HR = 0.32 (95% CI 0.21-0.48, median PFS in niraparib and placebo arm was 22.3 vs. 5.6 months) in patients undergoing IDS. In niraparib-treated patients, those who underwent PDS and IDS experienced similar rates of grade ≥3 adverse events (TEAEs) (50.7% vs 58.7%) and treatment discontinuation due to AEs (6.7% vs 6.6%).
Table 1. Efficacy of niraparib maintenance therapy in patients with newly diagnosed advanced ovarian cancer by surgical timing and residual disease status | |||||
Subgroups | Niraparib | Placebo | Hazard ratio (95% CI) | ||
n | mPFS, months | n | mPFS, months | ||
Overall | 255 | 24.8 | 129 | 8.3 | 0.45 (0.34–0.60) |
PDS | 134 | NR | 70 | 12.0 | 0.63 (0.42–0.94) |
Optimal | 101 | NR | 56 | 13.8 | 0.74 (0.46–1.17) |
Suboptimal | 24 | 13.8 | 11 | 8.3 | 0.28 (0.10–0.84) |
IDS | 121 | 22.3 | 59 | 5.6 | 0.32 (0.21–0.48) |
Optimal | 92 | 24.7 | 49 | 5.6 | 0.26 (0.17–0.41) |
Suboptimal | 12 | 16.5 | 3 | 5.5 | 0.20 (0.02–2.22) |
CI: confidence interval; IDS: interval debulking surgery; mPFS: median progression-free survival; PDS: primary debulking surgery; NR: not reached. |
This subgroup analysis demonstrated that niraparib improved PFS regardless of surgical timing compared with placebo in patients with newly diagnosed advanced ovarian cancer.
Clinical trial identifier: NCT03709316
To determine the prevalence of somatic homologous recombination (HR) gene mutations in uterine serous cancer (USC) and compare these with rates among high-grade serous ovarian cancer (HGSOC).
The American Association for Cancer Research’s (AACR) Project Genomics Evidence Neoplasia Information Exchange (GENIE) database version 12.0 was queried via cBioPortal (http://genie.cbioportal.org). This is a publicly available, multi-institutional database of next-generation sequencing genomic profiles of tumor samples. Mutation frequencies for 8 homologous recombination deficiency (HRD) genes are reported and compared between uterine serous cancer (USC) and high-grade serous ovarian cancer (HGSOC) samples using chi-squared or Fischer’s exact tests. The threshold used for statistical significance was a two-sided alpha of 0.05.
Mutation frequency among USC | Mutation frequency among HGSOC | P-value | |
---|---|---|---|
BRCA1 | 12/776, 1.5% | 313/3125, 10.0% | p<0.001 |
BRCA2 | 40/776, 5.2% | 218/3125, 7.0% | p<0.001 |
BRIP1 | 4/752, 0.5% | 57/2974, 1.9% | p<0.01 |
CHEK2 | 6/768, 0.8% | 19/3040, 0.6% | 0.82 |
BARD1 | 12/664, 1.8% | 55/2615, 2.1% | 0.74 |
RAD51C | 3/671, 0.4% | 27/2663, 1.0% | 0.25 |
RAD51D | 4/671, 0.6% | 24/2658, 0.9% | 0.59 |
PALB2 | 12/766, 1.6% | 37/3045, 1.2% | 0.58 |
Among 844 USC samples, the most prevalent somatic HR gene mutations were BRCA2 (5.2%), BARD1 (1.8%), PALB2 (1.6%) and BRCA1 (1.5%). When compared to 3304 HGSOC samples, higher HR gene mutation frequencies among HGSOC were only found for BRCA1 (10.0% vs 1.5%, p<0.001), BRCA2 (7.0% vs 5.2%, p<0.001) and BRIP1 (1.9% vs 0.5%, p<0.01); however, there was no difference between HGSOC and USC samples in mutation frequencies of BARD1, CHEK2, RAD51C, RAD51D or PALB2 (Table 1).
Genomic profiling of USC tumor samples demonstrates a substantial rate of somatic gene mutations conferring HR deficiency, with rates for most HR gene mutations similar to those found among HGSOC samples. This demonstrates the need for genetically targeted clinical trials of poly-ADP ribose polymerase (PARP) inhibitors in patients with USC, and further research into the germline genetic profile of these patients to establish if this rare uterine cancer could be a phenotype of the hereditary breast and ovarian cancer syndrome.
The bevacizumab(bev)/olaparib(ola) maintenance regimen was approved for women affected with BRCA-mutated (BRCAm) and HRD HGOC, based on a greater progression-free survival (PFS) compared to bevacizumab alone in the PAOLA-1/ENGOT-ov25 trial (NCT02477644). HRD status in this trial was determined by the centralized MyChoice CDx test by Myriad Genetics (MG test). Belonging to the second part of the ENGOT HRD European initiative we evaluated the analytical and clinical validity of the ShallowHRDv2 test, an assay that analyzes the genome-wide copy number alterations, as compared with the MG test on PAOLA-1/ENGOT-ov25 tumor samples.
We performed shallow Whole Genome Sequencing (sWGS) with a mean coverage of ~1x (XT-HS capture kit; Agilent) on 449 tumor DNAs from the PAOLA-1/ENGOT-OV25 trial. HRD status was determined using the shallowHRDv2 bio-informatics pipeline with a single cut-off at 20 Large scale Genomic Alterations. We further evaluated our test on 109 prospective samples from our routine practice.
In the PAOLA-1 cohort, positive agreement for the ShallowHRDv2 vs. MG test status was 95% (196/206), negative agreement was 92% (173/188) and the overall agreement was 94% (369/394). Patients with shallowHRDv2-HRD-positive (including BRCAm) tumors showed a significantly prolonged PFS with bev/ola versus bev/placebo (median PFS: 44.8 vs. 20.3 months, hazard ratio (HR): 0.35 [95% CI, 0.23–0.54]). This benefit was still significant after exclusion of BRCAm tumors (41.7 vs. 18.6 months, HR: 0.47 [95% CI, 0.27–0.84]). Less non-contributive analyses were observed with the ShallowHRDv2 test (15/449; 3%) than with MG test (51/449; 11%). We confirmed the high overall percentage agreement of 91% (86/94) between ShallowHRDv2 and MG tests in the prospective cohort, with less non contributive results for ShallowHRDv2 (5% vs. 12%).
The ShallowHRDv2 test is a robust, cost-effective, clinically validated & highly performant test for HRD determination and is a valid alternative to MG to detect HRD in ovarian cancers.
NCT02477644
The PAOLA-1/ENGOT-ov25 phase 3 trial is a first line treatment study for advanced ovarian cancer which showed that the maximum benefit of the PARP inhibitor olaparib plus bevacizumab maintenance is observed in patients with Homologous Recombination Deficient (HRD) tumors. Using PAOLA-1 data, we report a retrospective clinical evaluation of SeqOne HRD solution based on low-coverage and cost-effective whole genome sequencing (WGS).
SeqOne HRD solution combines genomic instability and CCNE1 gene amplification features extracted from WGS, as well as pathogenic mutations in BRCA 1/2 genes. We conducted biological and computational experiments to evaluate the lower limit for sequencing coverage and tumoral content. The clinical evaluation on 368 patients from the PAOLA-1 trial included Progression-Free Survival (PFS) data and comparisons with the MyChoice® CDx test (Myriad genetics). All samples were prepared using Agilent’s SureSelect reagents and sequenced on Illumina’s NextSeq or NovaSeq platforms.
The biological and computational validation shows that the test provides consistent HRD status for a sequencing coverage of at least 0.3X and a percentage of tumoral content of at least 20%. A concordance above 90% was found between MyChoice® CDx test and SeqOne HRD solution with WGS data. The ability to detect patients who benefit from olaparib maintenance was similar for SeqOne HRD solution or MyChoice test. In HRD positive patients treated with olaparib the PFS Hazard Ratio (HR) was 0.39 (0.27-0.55) and 0.33 (0.23-0.48) with the SeqOne solution or MyChoice test, and was 0.89 (0.67-1.41) and 1.08 (0.73-1.59) respectively in HRD negative patients. In patients with BRCA wild type tumors, both SeqOne and MyChoice positive patients benefit from Olaparib maintenance with HR respectively of 0.49 (0.30-79) and 0.41 (0.24-0.69).
The SeqOne HRD solution brings a cost-effective and clinically validated Whole Genome Sequencing approach to detect HRD and to select advanced ovarian cancer patients to be treated in first-line with olaparib plus bevacizumab maintenance. This solution is efficient with a low tumoral content while being flexible by accommodating different lab configurations.
50% HGSOC patients harbour defects in homologous recombination repair (HR), where there are many proteins involved eg. BRCA. Mutations in BRCA genes (BRCAm), affect ~15% of the HGSOC population. RAD51 is another key protein in HR, binding to broken strands, forming nucleoprotein filaments to facilitate repair. This process can be identified via RAD51 foci, subnuclear complexes which can be viewed microscopically. RAD51 is overexpressed in many cancers including HGSOC; higher levels in FFPE tissue of patients with HGSOC are associated with improved DNA repair efficiency and poorer survival (8, 9). Here we sought to identify RAD51 foci in circulating cancer-associated cells correlating them with BRCA status and survival outcomes
124 HGSOC patients had blood samples taken pre chemo of whom 99 were suitable for analysis. 53 of these had further samples taken during and after treatment, and in follow up (see CONSORT Fig 1). These were processed using an immunofluorescence protocol. CCs were stained with WT1/RAD51 antibodies and viewed under LEICA microscopy, where RAD51 foci were counted. Results were correlated with clinical data and outcomes
Of 40 patients (5 BRCAm/HRD+) with paired samples, 32 responded according to CA125 (GCIG criteria). These patients had a significant reduction in CCs (p = 0.02) and RAD51 foci (p = 0.0007). By contrast, 8 non-responding patients had no change in RAD51 foci or CC.
The numbers of RAD51 foci seen in platinum resistant patients (n=15) were significantly greater compared with those (n= 56) who were platinum sensitive (PS) both at the beginning and end of chemotherapy. There were no RAD51 foci in 7 BRCA1m/HRD+ compared to BRCAwt (n=42) patients (p=0.009).
Median progression free survival (mPFS) had not been reached in those with <3 RAD51 foci (n=39) versus 13 months >3 RAD51 foci (n=32) at the start of treatment, p=0.04. mPFS survival was 12 months in those with <3 RAD51 foci (n=18) versus 3 months for those with >3 RAD51 foci at the end of treatment (n=13), p=0.001.
Here we demonstrate RAD51 foci in CCs from HGSOC patients. Numbers of cells and RAD51 foci change with treatment; fewer CC/RAD51 foci are associated with improved PFS
To characterize the mutational landscape of mucinous epithelial ovarian cancer (MOC).
Data were extracted from the American Association for Cancer Research’s (AACR) Project Genomics Evidence Neoplasia Information Exchange (GENIE) database version 12.1 via cBioPortal (http://genie.cbioportal.org). This publicly available, multi-institutional database provides next generation sequencing genomic profiles of tumors. We queried this database for MOC samples and descriptively report mutation frequencies of genes for which targeted therapies are either approved or in clinical trials for other cancer types. We include the following genes in our analysis: ERBB2 (HER2) amplifications, homologous recombination (HR) and mismatch repair (MMR) genes, BRAF, KRAS, NRAS, RNF43, ARID1A, PIK3CA, ERBB3, and PTEN mutations.
Among 145 MOC samples, the most common somatic mutations were KRAS (98/145, 67.6%), ERBB2 (HER2) amplification (13/90, 14.4%), ARID1A (16/116, 13.8%), CDKN2A (17/141, 12.1%), PIK3CA (13/145, 9.0%), ERBB3 (10/120, 8.3%), RNF43 (7/102, 6.9%), BRAF (4/145, 2.8%), PTEN (4/145, 2.8%), and NRAS (2/145, 1.4%). Mutations in HR and MMR genes were relatively rare, with the most common HR mutations in BRIP 1 (5/114, 4.4%) and BRCA2 (5/122, 4.1%); and the most common MMR mutations in MLH3 (1/44, 2.3%) and MSH3 (1/48, 2.1%).
No. of samples with ≥1 mutation | No. of samples profiled for mutation | Mutation frequency | |
BRCA1 | 5 | 122 | 1.6% |
BRCA2 | 5 | 122 | 4.1% |
BRIP1 | 5 | 114 | 4.4% |
CHEK2 | 1 | 116 | 0.9% |
PALB2 | 2 | 116 | 1.7% |
MSH3 | 1 | 48 | 2.1% |
MSH6 | 1 | 120 | 0.8% |
MLH3 | 1 | 44 | 2.3% |
PMS2 | 1 | 113 | 0.9% |
TP53 | 91 | 145 | 62.8% |
Advanced stage MOC carries a poor prognosis, with only platinum/taxane, anti-VEGF agents, and 5-fluouracil currently available as current standard of care therapies. Here, we demonstrate high mutation rates in targetable genes among MOC samples. Most commonly, these mutations were identified among KRAS/NRAS (targetable with MEK inhibitors), ERBB2 (HER2) (targetable with anti-HER2 agents), ARID1A (targetable with epigenetic modifiers and ATR inhibitors), and PI3CA (targetable with PI3-kinase and AKT inhibitors). Given the frequency of each mutation, genetically targeted clinical trials evaluating the clinical efficacy of their respective agents are warranted.
Pelvic node-positive (PNP) vulvar carcinoma has historically had poor outcomes; 2-year DFS was 27% in GOG 37. Studies in this cohort were performed prior to widespread PET staging and intensity modulated radiotherapy (IMRT). We assessed clinical outcomes and prognostic factors in FIGO IVb patients in the PET and IMRT era, hypothesizing that PET staging and IMRT treatment may improve outcomes.
Following IRB approval, we reviewed all patients with PNP vulvar squamous cell carcinoma (SCC) who received IMRT from 2012-2022. Patients treated palliatively were excluded. Eligible patients met stage IVb criteria and received simultaneous integrated boost (SIB) to radiologically positive lymph nodes (LN). Tumor size, grade, PET primary and LN SUVmax, largest LN size, LN anatomic location, p16 status, chemotherapy (CHT), and RT completion were recorded, as were survival (OS), disease free survival (DFS), local (LRFS)/regional (RRFS) recurrence and distant metastasis (DMFS). Patterns of recurrence and survival were examined via Kaplan Meier, with univariate analysis (UVA) via log-rank t-test. Multivariate analysis (MVA) was performed via Cox regression.
198 patients received RT for vulvar SCC, and 31 met inclusion criteria. Median follow up was 26 months (IQR 9-49) for patients completing RT (n=25) and 3 months (IQR 1-9) for patients not completing RT (n=6). 93.5% (n=29) had PET staging, with avid LN at external iliac (74.2%, n=23), common iliac (16.1%, n=5) and para-aortic (9.7%, n=3) echelons. RT was majority definitive/neoadjuvant (83.9%, n=26); 93.5% (n=29) received concurrent CHT. 22.5% (n=7) underwent LN dissection, with median 3 LN resected and 2 LN positive. With RT completion, 3-year OS and DFS were 64.1% and 69.6%, with 82.6% LRFS, 83.1% RRFS, and 77.5% DMFS. On UVA, LN SUVmax correlated with worse DFS, RRFS, and DMFS, and p16- status with worse OS, RRFS, and DMFS (p<0.05). LN SUVmax retained significance for DFS (p=0.045) and DMFS (p=0.041) on MVA.
Clinical outcomes for PNP vulvar SCC have substantially improved in the PET and IMRT era, challenging the FIGO IVb designation. PET LN SUVmax may help predict patients at higher risk of DM, and p16+ status those with more favorable outcomes.