Purpose:

To characterize the somatic mutational landscape, investigate associations between genetic alterations and clinical outcomes, and determine the prevalence of pathogenic germline mutations in low-grade serous ovarian carcinomas (LGSC).

Experimental Design:

Patients with LGSC tumors who underwent panel-based sequencing of up to 505 genes were identified. Data on somatic and germline mutations; copy-number alterations; and clinicopathologic features, including age at diagnosis, platinum sensitivity, and overall survival (OS), were collected.

Results:

Following central pathology rereview, 119 patients with LGSC were identified for analysis. Of these, 110 (92%) had advanced-stage disease (stages III/IV). Somatic KRAS (33%), NRAS (11%), EIF1AX (10%), and BRAF (11%) alterations were the most common; MAPK pathway alterations were found in 60% (n = 71) of LGSCs. KRAS mutations were significantly associated with age at diagnosis more than 50 years (P = 0.02) and platinum-sensitive disease (P = 0.03). On multivariate analysis, MAPK pathway alterations (P = 0.02) and platinum sensitivity (P = 0.005) were significantly associated with improved OS. Seventy-nine patients (66%) underwent germline genetic testing; seven pathogenic germline mutations were identified: MUTYH (n = 2), BAP1 (n = 1), RB1 (n = 1), CHEK2 (n = 1), APC (n = 1), and FANCA (n = 1). There were no germline BRCA1/2 mutations. One germline MUTYH-associated LGSC harbored loss-of-heterozygosity at the MUTYH locus, and the patient with the germline BAP1 mutation also harbored a somatic BAP1 frameshift mutation.

Conclusions:

This study showed that MAPK pathway alterations in LGSC, including KRAS mutations, are independently associated with platinum sensitivity and prolonged survival. Germline data, which were limited, identified few pathogenic germline mutations in patients with LGSC.

See related commentary by Veneziani and Oza, p. 4357

Translational Relevance

Low-grade serous ovarian cancer (LGSC) is molecularly distinct from high-grade serous ovarian cancer (HGSC), having rare somatic TP53 mutations and frequent MAPK alterations. While the association between germline BRCA1/2 mutations with increased risk of HGSC is well established, a genetic predisposition for LGSC remains unclear. In our institutional cohort of LGSCs subjected to clinical next-generation sequencing, the majority harbored somatic alterations affecting the MAPK pathway. KRAS mutations were associated with older age at diagnosis and platinum sensitivity. On multivariate analysis, MAPK pathway alterations were associated with platinum sensitivity and improved overall survival. No patients had a pathogenic BRCA1/2 germline mutation, and the overall rate of germline mutations was low. As targeted therapies against the MAPK pathway have recently shown benefit in prospective studies for treatment of recurrent LGSC, routine somatic tumor testing may guide prognostic information and aid in patient selection for targeted therapies.

Serous carcinomas of the ovary, which account for up to 80% of epithelial ovarian, tubal, and peritoneal cancers, are subclassified into high-grade (90%–95% of serous tumors) and low-grade (5%–10% of serous tumors) disease based on morphologic, immunophenotypic, and molecular features (1–3). From a molecular standpoint, high-grade serous carcinoma (HGSC) is characterized by high levels of chromosomal instability, recurrent TP53 mutations, and germline alterations in homologous recombination repair–related genes, including BRCA1/2 (4–9). Clinically, HGSC presents at a median age of 63 years, is highly platinum responsive, and often has an aggressive disease course (10). Meanwhile, low-grade serous carcinoma (LGSC) is characterized by low mutational burden, low frequency of TP53 mutations, and an unclear germline association (11–15). Clinically, LGSC has a bimodal age of onset distribution (20–30 and 50–60 years), is less platinum-responsive than HGSC, and typically exhibits indolent behavior, although some cases are aggressive (10, 16, 17).

Alterations in the MAPK signal transduction pathway occur in up to 60% of LGSCs (18–20), as reported in prior studies. The MAPK pathway regulates cell proliferation through a signal transduction cascade mediated by Ras family members (KRAS, NRAS, HRAS), resulting in activation of downstream RAF/MEK/ERK effectors, and can be inhibited by negative regulators, such as NF1 (21). Alterations in MAPK pathway genes are common in LGSC, with alterations affecting KRAS in 19% to 41%, NRAS in 9% to 26%, and BRAF in 2% to 16% of patients (12, 18, 20, 22, 23).

Given the frequency of RAS/RAF gene alterations in LGSC, recent efforts have focused on the inhibition of MAPK signaling as a potential therapeutic strategy. Findings from a phase III trial investigating the MEK1/2 inhibitor binimetinib for persistent or recurrent LGSC demonstrated a significant progression-free survival (PFS) advantage with binimetinib treatment in patients with KRAS-mutated LGSC (n = 43) compared with KRAS wild-type (WT) LGSC (n = 90). In patients who had received cytotoxic chemotherapy, a posthoc analysis showed a trend towards improved PFS for patients with KRAS-mutated tumors, which did not reach statistical significance; however, the trial was not powered to answer this question (24). Of note, improved overall survival (OS) has previously been reported in patients with KRAS- or BRAF-mutated LGSC (n = 21) compared with KRAS/BRAF WT LGSC (n = 58; ref. 25). We sought to further characterize the somatic mutational landscape of LGSC, investigate the association between specific genetic alterations and clinical outcomes, and define the prevalence of pathogenic germline mutations in a cohort of patients with LGSC who had undergone tumor-normal clinical sequencing analysis.

Case selection

This study was performed in accordance with the U.S. Common Rule and approved by the Institutional Review Board of Memorial Sloan Kettering Cancer Center (MSKCC; New York, NY), and written informed consent was obtained from all patients. All patients with epithelial ovarian carcinoma diagnosed between April 2015 and February 2021 whose tumors had undergone clinical targeted massively parallel sequencing using the FDA-authorized MSKCC-Integrated Mutation Profiling of Actionable Cancer Targets (MSK-IMPACT) platform were identified (n = 2,063; Fig. 1; ref. 26). The original diagnostic pathology reports of cases noted as LGSC were manually reviewed (n = 182). Eleven cases were excluded because they were duplicate entries and 47 were excluded because they were miscoded as LGSC [serous borderline tumor with or without noninvasive implants (n = 12) and HGSC (n = 35, of which 3 arose from high-grade transformation of LGSC)]. Of note, in addition to ovarian LGSCs, ovarian serous borderline tumors with invasive implants and primary peritoneal LGSCs were included. A subspecialty-trained gynecologic pathologist (M.H. Chui) rereviewed the tumor slides of the remaining 124 cases to confirm the diagnosis in accordance with criteria defined by the World Health Organization (WHO) Classification of Tumors, 5th edition (27). From this central pathology rereview, cases with low tumor cellularity (n = 1), high-grade serous histology (n = 2), seromucinous carcinoma (n = 1), and noninvasive serous borderline tumors without a component of LGSC (n = 1) were identified and excluded, yielding a final cohort of 119 patients (Fig. 2).

Figure 1.

CONSORT diagram representing selection of the LGSC cohort. MSK-IMPACT was queried for all ovarian carcinomas that had undergone molecular analysis. All non-LGSCs were excluded (n = 1,881). Pathology reports for the remaining 182 cases were manually reviewed. Duplicates and cases with high-grade or noninvasive histology were excluded (n = 58). The remaining 124 cases underwent central pathology review of tumor diagnostic slides, resulting in a final cohort of 119 LGSCs.

Figure 1.

CONSORT diagram representing selection of the LGSC cohort. MSK-IMPACT was queried for all ovarian carcinomas that had undergone molecular analysis. All non-LGSCs were excluded (n = 1,881). Pathology reports for the remaining 182 cases were manually reviewed. Duplicates and cases with high-grade or noninvasive histology were excluded (n = 58). The remaining 124 cases underwent central pathology review of tumor diagnostic slides, resulting in a final cohort of 119 LGSCs.

Close modal
Figure 2.

Histomorphology of LGSC. A, A typical case characterized by small nests of monotonous tumor cells infiltrating through fibrotic stroma, and occasional psammomatous calcifications. B, Another case exhibiting the less common inverted macropapillary invasion pattern. C, Invasive LGSC in the peritoneum, previously termed “invasive implant,” associated with ovarian serous borderline tumor (inset). Representative excluded cases after central pathology review: D, Tumor with low cellularity precluding histologic subtyping and reliable molecular analysis. E, HGSC exhibiting marked nuclear pleomorphism and high mitotic activity. F, Seromucinous carcinoma, characterized by mucin-producing glands, and considered to be a variant of endometrioid carcinoma.

Figure 2.

Histomorphology of LGSC. A, A typical case characterized by small nests of monotonous tumor cells infiltrating through fibrotic stroma, and occasional psammomatous calcifications. B, Another case exhibiting the less common inverted macropapillary invasion pattern. C, Invasive LGSC in the peritoneum, previously termed “invasive implant,” associated with ovarian serous borderline tumor (inset). Representative excluded cases after central pathology review: D, Tumor with low cellularity precluding histologic subtyping and reliable molecular analysis. E, HGSC exhibiting marked nuclear pleomorphism and high mitotic activity. F, Seromucinous carcinoma, characterized by mucin-producing glands, and considered to be a variant of endometrioid carcinoma.

Close modal

Massively parallel sequencing analysis and genomic data extraction

Massively parallel sequencing was performed on the first available LGSC sample (primary, n = 81; recurrent, n = 38 LGSCs) and matched normal blood in the Clinical Laboratory Improvement Amendments (CLIA)-certified MSKCC Molecular Diagnostics Service Laboratory using the FDA-cleared MSK-IMPACT assay targeting 341 (n = 8), 410 (n = 23), 468 (n = 83), or 505 (n = 5) cancer-related genes, as well as fusions in 21 of these genes. The median sequencing depth was 588x (range, 94–1,231x). Genomic data, including somatic mutations and copy-number alterations (n = 119), MSIsensor score, and tumor mutational burden (TMB; somatic mutations per megabase), were extracted from MSK-IMPACT, and germline pathogenic mutations from custom panels (76 genes, n = 14; 84 genes, n = 1; 88 genes, n = 49; and 90 genes, n = 3; ref. 28). Germline variant calling was performed as previously described (29, 30). Identified variants were independently assessed and manually curated, applying current standards for variant classification by the American College of Medical Genetics and Genomics (31), to define pathogenic mutations (pathogenic/likely pathogenic variants). Germline data for 28 cases were collected from outside sources, including Myriad Genetics (n = 12), Ambry Genetics (n = 5), Invitae (n = 3), Caris Life Sciences (n = 2), GeneDx (n = 2), DNA Genotek (n = 1), FoundationOne (n = 1), Genekor (n = 1), and UCSF Cancer Gene Panel Test (n = 1). To assess whether the underlying germline alteration was the driver of a given tumor, biallelic inactivation was evaluated by inferring loss-of-heterozygosity (LOH) using the FACETS algorithm (32). Biallelic inactivation was defined as a loss of the WT allele in the tumor at the locus of a germline mutation or by the presence of a second somatic loss-of-function mutation in the respective tumor.

Association of genomic data with clinical outcomes

Clinicopathologic variables of interest, collected from the electronic medical record, included the following: body mass index (BMI), volume of residual disease at the time of initial cytoreduction [complete gross resection (no visible disease at end of the procedure); optimal debulking (residual tumor ≤1 cm in largest diameter); and suboptimal debulking (residual tumor >1 cm in largest diameter)], stage at diagnosis [per the International Federation of Gynaecology and Obstetrics (FIGO) 2014 staging system], adjuvant therapy (treatment within 6 weeks following surgery, including observation, chemotherapy, and/or endocrine therapy), and maintenance therapy (treatment following initial postsurgical treatment in the absence of active disease).

Associations between somatic genetic alterations and clinical variables of interest were determined, including age at diagnosis (≤50 years vs. >50 years as a surrogate for preoperative menopausal status) and platinum sensitivity. Platinum sensitivity was defined as tumor recurrence more than 6 months from date of last front-line platinum-based chemotherapy; platinum resistance was defined as tumor recurrence ≤6 months from date of last front-line platinum-based chemotherapy or progression on platinum-based chemotherapy.

PFS and OS

The association of clinicodemographic and molecular characteristics with PFS and OS were examined in univariate and multivariate settings. PFS was measured from the date of diagnosis to the date of progression, as determined by biopsy, surgical resection, or imaging confirmation of recurrence. OS was measured from the date of diagnosis to the date of death. Patients alive and disease-free or alive with disease were censored for PFS and OS, respectively, at the date of last-follow-up. Analyses for PFS and OS were performed using the entire cohort. For these analyses, “MAPK gene alterations” included pathogenic or likely pathogenic alterations involving any of the following genes in the MSK-IMPACT assay panel: BRAF, KRAS, HRAS, NRAS, NF1, NF2, MAP2K1, MAP2K2, MAP2K4, MAP3K1, MAP3K13, MAP3K14, MAPK1, MAPK3, MAPKAP1, DUSP4, ERBB2, RRAS, RRAS2, or RAF1. To investigate bias for patients referred to our center at recurrence versus at time of primary diagnosis, sensitivity analyses were performed for PFS and OS among patients seen at our center at diagnosis (n = 60).

Statistical analysis

Associations between continuous clinicopathologic variables were compared using the Wilcoxon rank-sum tests; associations between categorical variables were performed using two-tailed Fisher exact test. In PFS and OS analyses, for analyses of primary treatment after surgery, landmark analysis was applied with a 6-week landmark time for postoperative treatment (n = 0 patients excluded); another landmark time at 6 months was applied for analyzing maintenance therapy (9 early progression and short follow-up patients excluded for PFS and 2 short follow-up patients excluded for OS), and another at 12 months for platinum sensitivity (n = 7 patients excluded for OS). Survival analysis was performed by the Kaplan–Meier method (Fig. 3). P values were obtained by applying log-rank test for categorical variables and Wald test based on Cox proportional hazards (CoxPH) model for continuous variables. P < 0.05 was considered significant. Variables that showed significance in univariate PFS/OS analyses were selected for the multivariate CoxPH model. All statistical analyses were performed using R version 3.6.3 (https://cran.r-project.org/).

Figure 3.

OS of patients with MAPK pathway–driven LGSCs (n = 71, harboring somatic genetic alterations in BRAF KRAS, NRAS, and others) compared with those lacking MAPK pathway alterations (n = 48).

Figure 3.

OS of patients with MAPK pathway–driven LGSCs (n = 71, harboring somatic genetic alterations in BRAF KRAS, NRAS, and others) compared with those lacking MAPK pathway alterations (n = 48).

Close modal

Data availability

Targeted sequencing data supporting the findings of this study will be available at cBioPortal for Cancer Genomics [www.cbioportal.org “Low-Grade Serous Ovarian Cancer (MSK, Clin Cancer Res 2022)”] upon publication of this manuscript.

Clinicopathologic demographics

The median age at LGSC diagnosis was 48 years (range, 20–79 years); 65 patients (55%) were diagnosed at 50 years of age or younger, and 54 (45%) were diagnosed after 50 years of age (Table 1). 110 patients (92%) were diagnosed with advanced-stage (FIGO III/IV) disease. 16 patients (13%) self-identified as Ashkenazi Jewish. Among patients who had undergone initial surgical cytoreduction, 52 (57%) achieved a complete gross resection and 32 (35%) were optimally debulked. 107 patients (90%) received adjuvant platinum-based chemotherapy, while 6 patients (5%) received endocrine therapy alone. After primary treatment, 75 patients (63%) had no evidence of disease, while 44 patients (37%) had clinical evidence of disease. Among those with no evidence of disease, 53 (71%) underwent observation and had a median PFS of 43.1 months [95% confidence interval (CI), 30.2–54.8 months], 17 (23%) received endocrine maintenance therapy and had a median PFS of 19.6 months (95% CI, 12.8–not estimable), and 2 (3%) received bevacizumab maintenance therapy and had a median PFS of 12.1 months (95% CI, 6.6–not estimable; P = 0.11). 3 patients (4%) with no evidence of disease received maintenance therapies and were not included in comparison groups. Among patients with evidence of disease at the conclusion of postoperative treatment, 9 (20%) underwent observation and had a median PFS of 13.0 months (95% CI, 3.3–28.1), 21 (48%) received endocrine maintenance therapy and had a median PFS of 22.8 months (95% CI, 8.0–not estimable), and 4 (9%) received bevacizumab maintenance and had a median PFS of 37.8 months (95% CI, 2.9–not estimable; P = 0.17). 10 patients (23%) with evidence of disease at the conclusion of postoperative treatment received other maintenance therapies and were not included in comparison groups.

Table 1.

Clinicopathologic and treatment characteristics of the full cohort (N = 119).

CharacteristicPatients, n (%)
Age (continuous) 
Median, years (range) 48 (20–79) 
Age (bivariate) 
 ≤50 years 65 (55%) 
 >50 years 54 (45%) 
BMI at diagnosis 
 Median, kg/m2 (range) 25.5 (17.0–50.9) 
Stage 
 I 4 (3%) 
 II 4 (3%) 
 III 94 (79%) 
 IV 16 (14%) 
Race/ethnicity 
 White 88 (74%) 
 Asian 10 (8%) 
 Black/African-American 7 (6%) 
 Unknown 14 (12%) 
Ashkenazi Jewish 
 Yes 16 (13%) 
 No 88 (74%) 
 Unknown 15 (13%) 
Initial cytoreduction 
 Complete gross resection 52 (57%) 
 Optimal 32 (35%) 
 Suboptimal 8 (9%) 
 Unknown/missing 27 
Postoperative treatment 
 Chemotherapy 107 (90%) 
 Endocrine therapy 6 (5%) 
 Observation 6 (5%) 
Platinum sensitivity 
 Sensitive 66 (73%) 
 Resistant 24 (27%) 
 Unknown/unable to classify 29 (24%) 
CharacteristicPatients, n (%)
Age (continuous) 
Median, years (range) 48 (20–79) 
Age (bivariate) 
 ≤50 years 65 (55%) 
 >50 years 54 (45%) 
BMI at diagnosis 
 Median, kg/m2 (range) 25.5 (17.0–50.9) 
Stage 
 I 4 (3%) 
 II 4 (3%) 
 III 94 (79%) 
 IV 16 (14%) 
Race/ethnicity 
 White 88 (74%) 
 Asian 10 (8%) 
 Black/African-American 7 (6%) 
 Unknown 14 (12%) 
Ashkenazi Jewish 
 Yes 16 (13%) 
 No 88 (74%) 
 Unknown 15 (13%) 
Initial cytoreduction 
 Complete gross resection 52 (57%) 
 Optimal 32 (35%) 
 Suboptimal 8 (9%) 
 Unknown/missing 27 
Postoperative treatment 
 Chemotherapy 107 (90%) 
 Endocrine therapy 6 (5%) 
 Observation 6 (5%) 
Platinum sensitivity 
 Sensitive 66 (73%) 
 Resistant 24 (27%) 
 Unknown/unable to classify 29 (24%) 

Among the 119 patients included in the study, there were 27 progression-free survivors after a median length of follow-up of 16.3 months (range, 3.8–52.4 months). There were 86 overall survivors after a median length of follow-up of 62.7 months (range, 3.8–611.9 months). Among recurrences, 66 tumors (73%) were considered clinically platinum sensitive. Recurrences were treated as follows: 27 with chemotherapy alone, 7 with chemotherapy and bevacizumab, 20 with endocrine therapy alone, 4 with MEK inhibitors, 3 with surgery alone, 22 with surgery and chemotherapy, and 9 with surgery followed by endocrine therapy. Among 4 patients who underwent MEK inhibitor treatment for recurrence, 3 had a MAPK-altered LGSC. None of the 4 patients had a progression event on MEK inhibitor treatment, and all patients ultimately discontinued MEK inhibitor therapy due to toxicity after a median duration of 6.7 months (range, 0.7–22.6 months).

Somatic genetic alterations of LGSC

Clinical targeted next-generation sequencing (NGS) analysis using the MSK-IMPACT platform of 341 to 505 cancer-related genes revealed LGSCs have a low TMB (median, 1.8 mutations/Mb; range, 0–17.8), with a median of 1 (range, 0–16) nonsynonymous somatic mutation. All of the 102 evaluable cases were microsatellite stable (median MSIsensor score, 0.12; range, 0–3.83; Supplementary Fig. S1). This cohort of LGSCs had low levels of chromosomal instability (median fraction of the genome altered, 0.16; range, 0–0.74).

MAPK pathway alterations were found in the following genes: BRAF (n = 13), KRAS (n = 39), HRAS (n = 1), NRAS (n = 13), NF1 (n = 2), NF2 (n = 1), MAP3K1 (n = 1), and ERBB2 (n = 5; Supplementary Fig. S2). There were no alterations found in MAP2K1, MAP2K2, MAP2K4, MAP3K13, MAP3K14, MAPK1, MAPK3, MAPKAP1, DUSP4, RRAS, RRAS2, and RAF1. Overall, 71 cases (60%) had somatic alterations affecting MAPK pathway genes (Fig. 4A). The most common recurrent alterations detected included KRAS hotspot mutations (33%; G12V, n = 19; G12D, n = 14; G12C, n = 3; G12R, n = 2; and G12A, n = 1), NRAS hotspot mutations (11%; Q61R, n = 9; and Q61K, n = 4), EIF1AX alterations (10%, n = 12), BRAF alterations (11%; hotspot mutations in V600E, n = 8; D594N, n = 2; and N581I, n = 1; as well as fusions of BRAF-MKRN1, n = 1; and BRAF-AGAP3, n = 1), and CDKN2A homozygous deletions (8%, n = 9; Fig. 4A). There were two cases with ERBB2 amplification, and one hotspot mutation (Y772_A775dup). KRAS, NRAS, and BRAF mutations were mutually exclusive, except for one case that had coexisting BRAF and KRAS mutations (KRAS vs. BRAF: P = 0.034, q = 1.0; KRAS vs. NRAS: P = 0.004, q = 0.46; and BRAF vs. NRAS: P = 0.204, q = 1.0).

Figure 4.

Somatic mutational landscape for the full cohort of LGSCs (n = 119). A, Percentages of mutated samples are presented on the right. Heatmap comparisons of mutations within clinical groups: age at diagnosis ≤50 years versus more than 50 years (B), and platinum-sensitive versus platinum-resistant disease (C). *Statistically significant associations. SNV, single-nucleotide variant.

Figure 4.

Somatic mutational landscape for the full cohort of LGSCs (n = 119). A, Percentages of mutated samples are presented on the right. Heatmap comparisons of mutations within clinical groups: age at diagnosis ≤50 years versus more than 50 years (B), and platinum-sensitive versus platinum-resistant disease (C). *Statistically significant associations. SNV, single-nucleotide variant.

Close modal

No somatic mutations or gene copy-number alterations were identified in 21 LGSCs (18%; Fig. 4A). Of note, the median tumor purity for samples with and without identified somatic mutations was similar [mutations, 40% (range, 10%–80%) vs. 30% (range, 10%–80%), respectively; P = 0.12], suggesting that these were not genetic alterations missed due to technically inadequate samples.

Eight patients had NGS performed on metachronous tumors (paired primary and recurrent tumors, n = 3; and multiple recurrent tumors only, n = 5; Supplementary Fig. S3). In 5 patients, the first tumor analyzed did not have somatic genetic alterations. Subsequent recurrences in 2 of these patients harbored driver genetic alterations (one case with a CDKN2A/B deletion and the other with BRAF G469V). In the remaining three cases, there were shared genetic alterations across primary/recurrent tumors. While some samples harbored mutations exclusive to recurrent samples, overall the findings are consistent with a clonal relationship.

Somatic genetic alterations in LGSC and associations with outcome

We next assessed the association of the specific major RAS/RAF genetic alteration, i.e., KRAS hotspot mutations, NRAS hotspot mutations, and BRAF hotspot mutations, which are generally mutually exclusive, with age at diagnosis and platinum sensitivity. KRAS mutations were found to be significantly associated with age at diagnosis more than 50 years (P = 0.03; Fig. 4B) and platinum-sensitive disease (P = 0.04; Fig. 4C). NRAS and BRAF alterations were not statistically associated with age or platinum sensitivity. For example, 62% (n = 8) of patients with NRAS-altered tumors were more than 50 years of age at diagnosis, compared with 43% (n = 46) of patients with NRAS WT tumors (P = 0.25); and 90% (n = 9) of patients with NRAS-altered tumors were platinum sensitive, compared with 71% (n = 57) of patients with NRAS WT tumors (P = 0.28). Meanwhile, 31% (n = 4) of patients with BRAF-altered tumors were more than 50 years of age at diagnosis, compared with 47% (n = 50) of patients with BRAF WT tumors (P = 0.38); and 50% (n = 4) of patients with BRAF-altered tumors were platinum sensitive, compared with 76% (n = 62) of patients with BRAF WT tumors (P = 0.20).

Univariate and multivariate survival analyses were performed to further investigate molecular and clinicodemographic features associated with OS (Table 2). For this analysis, KRAS, NRAS, and BRAF mutations, and other rare somatic pathogenic mutations (listed previously) affecting the MAPK pathway were grouped together, as they likely represent a biologically distinct group of MAPK pathway-driven tumors. On univariate analysis, platinum sensitivity and MAPK pathway alterations were significantly associated with improved OS (P = 0.001 and P = 0.005, respectively), whereas age and residual disease after initial cytoreduction were not significantly associated with OS (P > 0.99 and P = 0.29, respectively). Of 90 patients in the population included in landmark multivariate analysis, 12 (50%) of 24 patients with platinum-resistant disease died, compared with 20 (30%) of 66 patients with platinum-sensitive disease. Meanwhile, 14 (26%) of 53 patients with a MAPK pathway–altered tumor died, compared with 18 (49%) of 37 patients without a MAPK pathway–altered tumor. Both MAPK pathway alterations and platinum sensitivity remained significant on multivariate analysis (P = 0.02 and P = 0.005, respectively). On sensitivity analysis, restricted to only patients presenting to our institution at the time of diagnosis (n = 60), no variables were significantly associated with OS, likely due to the small numbers.

Table 2.

Univariate and multivariate analysis of factors associated with OS.

Factors associated with OSTotal (n)Median OS (mo)Univariate HR (95% CI)Univariate PMultivariate HR (95% CI)Multivariate P
Age at initial diagnosis    0.99   
 ≤50 years 65 263 —    
 >50 years 54 213 1.0 (0.5–2.1)    
BMI   1.0 (1.0–1.1) 0.36   
Stage    0.09   
 III 94 246 —    
 IV 16 117 2.2 (0.9–5.3)    
Initial cytoreduction    0.29   
 CGR 52 NYR —    
 Optimal 32 106 1.8 (0.8–4.5)    
 Suboptimal NYR 2.1 (0.6–7.8)    
 Adjuvant treatment    0.32   
 None/observation NYR —    
Chemotherapy 107 244 2.5 (0.3–18.6)    
Endocrine therapy 138 6.0 (0.5–69.3)    
Platinum sensitivity    0.001a  0.005a 
 Resistant 24 45 —  —  
 Sensitive 66 234 0.3 (0.2–0.7)  0.4 (0.2–0.7)  
MAPK pathway alterationa    0.005a  0.019a 
 Present 71 339 —  —  
 Absent 48 125 2.7 (1.3–5.6)  2.5 (1.2–5.2)  
Factors associated with OSTotal (n)Median OS (mo)Univariate HR (95% CI)Univariate PMultivariate HR (95% CI)Multivariate P
Age at initial diagnosis    0.99   
 ≤50 years 65 263 —    
 >50 years 54 213 1.0 (0.5–2.1)    
BMI   1.0 (1.0–1.1) 0.36   
Stage    0.09   
 III 94 246 —    
 IV 16 117 2.2 (0.9–5.3)    
Initial cytoreduction    0.29   
 CGR 52 NYR —    
 Optimal 32 106 1.8 (0.8–4.5)    
 Suboptimal NYR 2.1 (0.6–7.8)    
 Adjuvant treatment    0.32   
 None/observation NYR —    
Chemotherapy 107 244 2.5 (0.3–18.6)    
Endocrine therapy 138 6.0 (0.5–69.3)    
Platinum sensitivity    0.001a  0.005a 
 Resistant 24 45 —  —  
 Sensitive 66 234 0.3 (0.2–0.7)  0.4 (0.2–0.7)  
MAPK pathway alterationa    0.005a  0.019a 
 Present 71 339 —  —  
 Absent 48 125 2.7 (1.3–5.6)  2.5 (1.2–5.2)  

Note: BMI was evaluated as a continuous variable.

Abbreviations: mo, months; CGR, complete gross resection; NYR, not yet reached.

aMAPK pathway genes included in targeted sequencing panel: BRAF, KRAS, HRAS, NRAS, NF1, NF2, MAP2K1, MAP2K2, MAP2K4, MAP3K1, MAP3K13, MAP3K14, MAPK1, MAPK3, MAPKAP1, DUSP4, ERBB2, RRAS, RRAS2, RAF1.

Survival analyses were also performed to investigate molecular and clinicodemographic features associated with PFS (Supplementary Table S1). Univariate analyses demonstrated that age and residual disease after initial cytoreductive surgery were significantly associated with PFS (P = 0.04 and P = 0.04, respectively) while on multivariate analysis, only residual disease after initial cytoreduction remained significant (P = 0.04). On sensitivity analysis, residual disease after initial cytoreduction was the only factor that remained significantly associated with PFS on univariate analysis (Supplementary Table S2).

Pathogenic germline mutations in patients with LGSC

Of the 119 patients with LGSC included in this study, 79 (66%) underwent germline genetic testing. 7 patients (9%) carried a pathogenic germline mutation (Supplementary Table S3), including RB1 (n = 1), BAP1 (n = 1), MUTYH (n = 2), CHEK2 (n = 1), APC (n = 1), and FANCA (n = 1). MUTYH was present in 2 patients, both harboring a c.1187G > A (p.Gly396Asp) alteration. LOH of the WT allele was only observed in the tumor from 1 patient with a pathogenic MUTYH germline mutation. The patient with a BAP1 pathogenic germline mutation also had a somatic frameshift BAP1 mutation in their LGSC. No germline pathogenic BRCA1/2 mutations were found.

The varying clinical course and challenging treatment of LGSC underscore the importance of identifying genetic alterations that may be associated with clinical outcomes or could be leveraged for targeted therapy. Unfortunately, the relative rarity of this histologic subtype has limited the ability of investigators to demonstrate these associations. This study examined a large cohort of patients with LGSC with tumor molecular profiling by panel-based sequencing and detailed long-term clinical follow-up. We characterized the somatic mutational landscape of this cohort of patients and examined the association between somatic alterations and age at diagnosis, platinum sensitivity, and survival outcomes. This study echoes prior findings in demonstrating the frequency of MAPK pathway alterations, which are present in approximately 60% of LGSCs (18–20). The presence of a KRAS mutation or other MAPK pathway alteration within our cohort was associated with improved outcomes. We demonstrate, by multivariate analysis, significantly improved OS in patients with MAPK pathway driven LGSC compared with those lacking somatic genetic alterations in MAPK pathway genes.

Prior studies have associated KRAS with improved outcomes in LGSC; however, these studies have not controlled for other variables such as platinum sensitivity, often due to sample size limitations (25). We found KRAS alterations were associated with platinum sensitivity, and multivariate analysis including both platinum sensitivity and MAPK alteration demonstrated independent associations between these variables and OS. Findings from the MILO-ENGOT-ov11 study, a phase III trial of a MEK1/2 inhibitor in persistent or recurrent LGSC, demonstrated a significant PFS advantage for patients with KRAS-mutated (n = 43) compared with KRAS WT LGSC who received the MEK1/2 inhibitor (n = 90; P = 0.006), as well as a nonsignificant trend towards improved PFS in the smaller cohort of patients treated with chemotherapy (24). Our findings on the independent association of MAPK pathway alteration with both platinum sensitivity and improved OS suggest that future studies, including upcoming trials comparing upfront cytotoxic therapies to endocrine therapies, should present distributions of MAPK alterations within patient subgroups, as this may affect response to therapy and overall prognosis.

Our results also confirm prior studies reporting that LGSC harbors few genetic alterations in cancer-related genes (23, 33). Notably, no somatic gene alterations were observed in 18% of LGSCs (n = 21). While the institutional panel-based sequencing (MSK-IMPACT) covers 341 to 505 cancer-related genes, it is possible that these LGSCs may have carried an alteration that was not included in this panel, such as in USP9X, which has been reported as altered in up to 27% of LGSCs (33). Future collaborative studies using whole-exome sequencing of LGSCs lacking alterations on MSK-IMPACT are warranted and may elucidate the underlying molecular drivers of cases lacking alterations in the tested cancer genes. It is also possible that tumors developed additional alterations not present in the tested sample. We conducted panel genetic testing on 8 patients with sequential tumor samples (either primary and recurrent tumors, n = 3; or samples of multiple recurrences, n = 5; Supplementary Fig. S3). This testing over multiple sequential samples revealed the presence of driver genetic alterations in two of five cases without somatic genetic alterations in the first tested tumor sample (Supplementary Fig. S3). This suggests that there may be a role for sequencing of tumors throughout a disease course, as a potentially actionable target may be acquired and present only at a recurrence or may be clonal and detectable in a recurrence. A previous study reported a similar finding, in which a patient with a BRAF/KRAS-WT ovarian micropapillary serous borderline tumor subsequently developed metastatic LGSC harboring a BRAF-V600E alteration (34).

Our rate of observed pathogenic germline mutations (9%) was higher than that of prior studies (6%), likely due to the larger number of genes assessed in our study, but lower than what has been observed in HGSC (20%; ref. 9). In order to determine if the germline mutations identified were associated with the development of LGSC, LOH was evaluated. A biallelic alteration was observed in the tumor of a patient with a MUTYH pathogenic germline mutation. Germline MUTYH mutations are associated with an increased risk of colon cancer, especially in homozygous state (35). 2 patients had the same germline MUTYH pathogenic variant (c.1187G > A; p.Gly396Asp), but only 1 demonstrated LOH of the WT allele, and the association with LGSC remains unclear. Ovarian cancers have been reported in patients with germline BAP1 pathogenic mutations (36). The LGSC of the patient with a germline BAP1 pathogenic mutation did not show LOH at that variant in the tumor; however, a somatic frameshift BAP1 mutation was identified.

We did not identify pathogenic germline BRCA1/2 mutations, even though 13% (n = 16) of our cohort identified as Ashkenazi Jewish. Previously published data on germline mutations in LGSC by Norquist and colleagues found that 3.5% to 5.7% of patients carried a pathogenic germline BRCA1 or BRCA2 mutation (9). In our study, we did not include LGSCs transformed to high grade at any point in the disease course, and all cases underwent rigorous central pathology review. It is possible that serous carcinomas with high-grade features or transformation were included in some of the prior studies, and that those were associated with BRCA1/2 germline mutations. In addition, our study may exhibit ascertainment bias as all of the women were selected through a database of patients who underwent tumor-normal clinical sequencing, and patients who had germline-only assessments or testing through outside laboratories entirely may have been missed. Although this study represents one of the largest cohorts of LGSC to undergo comprehensive mutational profiling, multi-institutional studies in diverse populations are needed to truly characterize the frequency of germline alterations.

The strengths of our study include the large sample size and central pathology review to ensure inclusion of only low-grade histology. All patients were seen at our institution and underwent rigorous clinical follow-up, which allowed us to accurately determine outcomes and associations with molecular factors. Our study has limitations in addition to those inherent to its retrospective design. Given the inclusion of patients from a tertiary referral center, there is likely an ascertainment bias to somatic testing, as evidenced by the high rate of advanced-stage disease (94%). To minimize the implications of this, we conducted sensitivity analyses for PFS and OS on the patients who presented at the time of diagnosis; however, these analyses were limited by sample size. In addition, both primary and recurrent tumor samples were tested, and it is possible that molecular alterations may change over time as demonstrated in our analysis of eight cases with panel genetic testing of tumors over several time points (37).

Traditionally, LGSC has been considered a disease with an indolent course compared with other epithelial ovarian cancers, but there is a subset of patients with LGSC who have aggressive disease and poor outcomes. Our findings may help identify patients with more aggressive LGSC and further define important subtypes of LGSC that may respond differently to systemic therapies. Our study showed that MAPK pathway alterations are independently associated with platinum sensitivity and prolonged survival in LGSC. Although limited in sample size, we identified few pathogenic germline alterations among our patients, and most were likely passengers rather than drivers of disease. Studies like ours are critical for understanding the implications of molecular findings on outcomes and for the development of new therapies for rare tumors.

Y.L. Liu reports grants from AstraZeneca, GlaxoSmithKline, and Repare Therapeutics outside the submitted work. D.S. Chi reports personal fees from AstraZeneca, Apyx Medical Corp., and Biom 'Up and other support from Moderna, BioNTech SE, Doximity, and Apyx Medical Corp. outside the submitted work. R.E. O'Cearbhaill reports personal fees from Tesaro/GSK, Regeneron, Seattle Genetics, Fresenius Kabi, Gynecologic Oncology Foundation, Bayer HealthCare, and Curio and other support from Hitech Health outside the submitted work. C. Aghajanian reports personal fees from Eisai/Merck, Mersana Therapeutics, Roche/Genentech, AbbVie, AstraZeneca/Merck, and Repare Therapeutics and grants from AstraZeneca, Genentech, AbbVie, Clovis, and AstraZeneca outside the submitted work. B. Weigelt reports personal fees from Repare Therapeutics outside the submitted work. M.H. Chui reports personal fees from Roche outside the submitted work. R.N. Grisham reports personal fees from GSK, Novartis, Natera, Corcept, and SpringWorks outside the submitted work. No disclosures were reported by the other authors.

B. Manning-Geist: Conceptualization, data curation, formal analysis, investigation, visualization, methodology, writing–original draft, writing–review and editing. S. Gordhandas: Conceptualization, data curation, formal analysis, investigation, visualization, methodology, writing–original draft, writing–review and editing. Y.L. Liu: Supervision, methodology, writing–original draft, writing–review and editing. Q. Zhou: Resources, data curation, software, formal analysis, validation, investigation, visualization, writing–review and editing. A. Iasonos: Resources, data curation, software, formal analysis, validation, investigation, visualization, writing–review and editing. A. Da Cruz Paula: Resources, data curation, software, formal analysis, validation, investigation, visualization, writing–review and editing. D. Mandelker: Resources, data curation, formal analysis, investigation, visualization, writing–review and editing. K. Long Roche: Visualization, methodology, writing–review and editing. O. Zivanovic: Visualization, methodology, writing–review and editing. A. Maio: Resources, data curation, formal analysis, investigation. Y. Kemel: Resources, data curation, formal analysis, investigation. D.S. Chi: Visualization, methodology, writing–review and editing. R.E. O'Cearbhaill: Visualization, methodology, writing–review and editing. C. Aghajanian: Visualization, methodology, writing–review and editing. B. Weigelt: Conceptualization, data curation, supervision, visualization, methodology, writing–original draft, project administration, writing–review and editing. M.H. Chui: Conceptualization, data curation, supervision, visualization, methodology, writing–original draft, project administration, writing–review and editing. R.N. Grisham: Conceptualization, formal analysis, supervision, investigation, methodology, writing–review and editing.

This work was funded in part by the NIH/NCI Cancer Center Support Grant P30 CA008748. B. Weigelt is funded in part by NIH/NCI P50 CA247749, as well as the Breast Cancer Research Foundation and Cycle for Survival grants. M.H. Chui is funded in part by PaigeAI. R.N. Grisham is funded in part by the Ovarian Cancer Research Fund Alliance.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Note: Supplementary data for this article are available at Clinical Cancer Research Online (http://clincancerres.aacrjournals.org/).

1.
Ledermann
JA
,
Raja
FA
,
Fotopoulou
C
,
Gonzalez-Martin
A
,
Colombo
N
,
Sessa
C
, et al
.
Newly diagnosed and relapsed epithelial ovarian carcinoma: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up
.
Ann Oncol
2018
;
29
Suppl 4
:
iv259
.
2.
Kurman
RJ
,
Shih Ie
M
.
The origin and pathogenesis of epithelial ovarian cancer: a proposed unifying theory
.
Am J Surg Pathol
2010
;
34
:
433
43
.
3.
Malpica
A
,
Deavers
MT
,
Lu
K
,
Bodurka
DC
,
Atkinson
EN
,
Gershenson
DM
, et al
.
Grading ovarian serous carcinoma using a two-tier system
.
Am J Surg Pathol
2004
;
28
:
496
504
.
4.
Kurman
RJ
,
Shih Ie
M
.
Molecular pathogenesis and extraovarian origin of epithelial ovarian cancer–shifting the paradigm
.
Hum Pathol
2011
;
42
:
918
31
.
5.
Ahmed
AA
,
Etemadmoghadam
D
,
Temple
J
,
Lynch
AG
,
Riad
M
,
Sharma
R
, et al
.
Driver mutations in TP53 are ubiquitous in high grade serous carcinoma of the ovary
.
J Pathol
2010
;
221
:
49
56
.
6.
Meinhold-Heerlein
I
,
Bauerschlag
D
,
Hilpert
F
,
Dimitrov
P
,
Sapinoso
LM
,
Orlowska-Volk
M
, et al
.
Molecular and prognostic distinction between serous ovarian carcinomas of varying grade and malignant potential
.
Oncogene
2005
;
24
:
1053
65
.
7.
Senturk
E
,
Cohen
S
,
Dottino
PR
,
Martignetti
JA
.
A critical re-appraisal of BRCA1 methylation studies in ovarian cancer
.
Gynecol Oncol
2010
;
119
:
376
83
.
8.
Nakayama
K
,
Nakayama
N
,
Jinawath
N
,
Salani
R
,
Kurman
RJ
,
Shih Ie
M
, et al
.
Amplicon profiles in ovarian serous carcinomas
.
Int J Cancer
2007
;
120
:
2613
7
.
9.
Norquist
BM
,
Harrell
MI
,
Brady
MF
,
Walsh
T
,
Lee
MK
,
Gulsuner
S
, et al
.
Inherited mutations in women with ovarian carcinoma
.
JAMA Oncol
2016
;
2
:
482
90
.
10.
Plaxe
SC
.
Epidemiology of low-grade serous ovarian cancer
.
Am J Obstet Gynecol
2008
;
198
:
459.e1
8
;
discussion 459.e8–9
.
11.
Bonome
T
,
Lee
JY
,
Park
DC
,
Radonovich
M
,
Pise-Masison
C
,
Brady
J
, et al
.
Expression profiling of serous low malignant potential, low-grade, and high-grade tumors of the ovary
.
Cancer Res
2005
;
65
:
10602
12
.
12.
Singer
G
,
Oldt
R
, 3rd,
Cohen
Y
,
Wang
BG
,
Sidransky
D
,
Kurman
RJ
, et al
.
Mutations in BRAF and KRAS characterize the development of low-grade ovarian serous carcinoma
.
J Natl Cancer Inst
2003
;
95
:
484
6
.
13.
Grisham
RN
,
Iyer
G
,
Garg
K
,
Delair
D
,
Hyman
DM
,
Zhou
Q
, et al
.
BRAF mutation is associated with early stage disease and improved outcome in patients with low-grade serous ovarian cancer
.
Cancer
2013
;
119
:
548
54
.
14.
Hsu
CY
,
Bristow
R
,
Cha
MS
,
Wang
BG
,
Ho
CL
,
Kurman
RJ
, et al
.
Characterization of active mitogen-activated protein kinase in ovarian serous carcinomas
.
Clin Cancer Res
2004
;
10
:
6432
6
.
15.
Vineyard
MA
,
Daniels
MS
,
Urbauer
DL
,
Deavers
MT
,
Sun
CC
,
Boerwinkle
E
, et al
.
Is low-grade serous ovarian cancer part of the tumor spectrum of hereditary breast and ovarian cancer?
Gynecol Oncol
2011
;
120
:
229
32
.
16.
Grabowski
JP
,
Harter
P
,
Heitz
F
,
Pujade-Lauraine
E
,
Reuss
A
,
Kristensen
G
, et al
.
Operability and chemotherapy responsiveness in advanced low-grade serous ovarian cancer. an analysis of the AGO study group metadatabase
.
Gynecol Oncol
2016
;
140
:
457
62
.
17.
Gershenson
DM
,
Sun
CC
,
Lu
KH
,
Coleman
RL
,
Sood
AK
,
Malpica
A
, et al
.
Clinical behavior of stage II-IV low-grade serous carcinoma of the ovary
.
Obstet Gynecol
2006
;
108
:
361
8
.
18.
Emmanuel
C
,
Chiew
YE
,
George
J
,
Etemadmoghadam
D
,
Anglesio
MS
,
Sharma
R
, et al
.
Genomic classification of serous ovarian cancer with adjacent borderline differentiates RAS pathway and TP53-mutant tumors and identifies NRAS as an oncogenic driver
.
Clin Cancer Res
2014
;
20
:
6618
30
.
19.
Anglesio
MS
,
Arnold
JM
,
George
J
,
Tinker
AV
,
Tothill
R
,
Waddell
N
, et al
.
Mutation of ERBB2 provides a novel alternative mechanism for the ubiquitous activation of RAS-MAPK in ovarian serous low malignant potential tumors
.
Mol Cancer Res
2008
;
6
:
1678
90
.
20.
Chui
MH
,
Chang
JC
,
Zhang
Y
,
Zehir
A
,
Schram
AM
,
Konner
J
, et al
.
Spectrum of BRAF mutations and gene rearrangements in ovarian serous carcinoma
.
JCO Precision Oncology
2021
;
5
:
1480
92
.
21.
Smalley
I
,
Smalley
KSM
.
ERK inhibition: a new front in the war against MAPK pathway-driven cancers?
Cancer Discov
2018
;
8
:
140
2
.
22.
Wong
KK
,
Tsang
YT
,
Deavers
MT
,
Mok
SC
,
Zu
Z
,
Sun
C
, et al
.
BRAF mutation is rare in advanced-stage low-grade ovarian serous carcinomas
.
Am J Pathol
2010
;
177
:
1611
7
.
23.
Hunter
SM
,
Anglesio
MS
,
Ryland
GL
,
Sharma
R
,
Chiew
YE
,
Rowley
SM
, et al
.
Molecular profiling of low grade serous ovarian tumours identifies novel candidate driver genes
.
Oncotarget
2015
;
6
:
37663
77
.
24.
Monk
BJ
,
Grisham
RN
,
Banerjee
S
,
Kalbacher
E
,
Mirza
MR
,
Romero
I
, et al
.
MILO/ENGOT-ov11: binimetinib versus physician's choice chemotherapy in recurrent or persistent low-grade serous carcinomas of the ovary, fallopian tube, or primary peritoneum
.
J Clin Oncol
2020
;
38
:
3753
62
.
25.
Gershenson
DM
,
Sun
CC
,
Wong
KK
.
Impact of mutational status on survival in low-grade serous carcinoma of the ovary or peritoneum
.
Br J Cancer
2015
;
113
:
1254
8
.
26.
Zehir
A
,
Benayed
R
,
Shah
RH
,
Syed
A
,
Middha
S
,
Kim
HR
, et al
.
Mutational landscape of metastatic cancer revealed from prospective clinical sequencing of 10,000 patients
.
Nat Med
2017
;
23
:
703
13
.
27.
World Health Organization
.
Female genital tumours: WHO classification of tumours
,
5th ed. Volume 4
.
Lyon (France)
:
International Agency for Research on Cancer
;
2020
.
28.
Li
MM
,
Datto
M
,
Duncavage
EJ
,
Kulkarni
S
,
Lindeman
NI
,
Roy
S
, et al
.
Standards and guidelines for the interpretation and reporting of sequence variants in cancer: a joint consensus recommendation of the association for molecular pathology, american society of clinical oncology, and college of american pathologists
.
J Mol Diagn
2017
;
19
:
4
23
.
29.
Cheng
DT
,
Prasad
M
,
Chekaluk
Y
,
Benayed
R
,
Sadowska
J
,
Zehir
A
, et al
.
Comprehensive detection of germline variants by MSK-IMPACT, a clinical diagnostic platform for solid tumor molecular oncology and concurrent cancer predisposition testing
.
BMC Med Genet
2017
;
10
:
33
.
30.
Mandelker
D
,
Zhang
L
,
Kemel
Y
,
Stadler
ZK
,
Joseph
V
,
Zehir
A
, et al
.
Mutation detection in patients with advanced cancer by universal sequencing of cancer-related genes in tumor and normal DNA vs guideline-based germline testing
.
JAMA
2017
;
318
:
825
35
.
31.
Richards
S
,
Aziz
N
,
Bale
S
,
Bick
D
,
Das
S
,
Gastier-Foster
J
, et al
.
Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the american college of medical genetics and genomics and the association for molecular pathology
.
Genet Med
2015
;
17
:
405
24
.
32.
Shen
R
,
Seshan
VE
.
FACETS: allele-specific copy number and clonal heterogeneity analysis tool for high-throughput DNA sequencing
.
Nucleic Acids Res
2016
;
44
:
e131
.
33.
Cheasley
D
,
Nigam
A
,
Zethoven
M
,
Hunter
S
,
Etemadmoghadam
D
,
Semple
T
, et al
.
Genomic analysis of low-grade serous ovarian carcinoma to identify key drivers and therapeutic vulnerabilities
.
J Pathol
2021
;
253
:
41
54
.
34.
Chui
MH
,
Xing
D
,
Zeppernick
F
,
Wang
ZQ
,
Hannibal
CG
,
Frederiksen
K
, et al
.
Clinicopathologic and molecular features of paired cases of metachronous ovarian serous borderline tumor and subsequent serous carcinoma
.
Am J Surg Pathol
2019
;
43
:
1462
72
.
35.
Nielsen
M
,
Infante
E
,
Brand
R
.
MUTYH polyposis
.
In
:
Adam
MP
,
Ardinger
HH
,
Pagon
RA
,
Wallace
SE
,
Bean
LJH
,
Mirzaa
G
, et al
.,
editors
.
GeneReviews(®) [Internet]
.
Seattle (WA)
:
University of Washington, Seattle
,
1993–2022
.
36.
Adam
MP
,
Ardinger
HH
,
Pagon
RA
,
Wallace
SE
,
Bean
LJH
,
Gripp
KW
, et al
.,
editors
.
GeneReviews(®) [Internet]
.
Seattle (WA)
:
University of Washington, Seattle
,
1993–2022
.
37.
Walpole
S
,
Pritchard
AL
,
Cebulla
CM
,
Pilarski
R
,
Stautberg
M
,
Davidorf
FH
, et al
.
Comprehensive study of the clinical phenotype of germline BAP1 variant-carrying families worldwide
.
J Natl Cancer Inst
2018
;
110
:
1328
41
.
38.
Hollis
RL
,
Gourley
C
.
Genetic and molecular changes in ovarian cancer
.
Cancer Biol Med
2016
;
13
:
236
47
.

Supplementary data