Pathologic findings in breast, fallopian tube, and ovary specimens in non-BRCA hereditary breast and/or ovarian cancer syndromes: a study of 18 patients with deleterious germline mutations in RAD51C, BARD1, BRIP1, PALB2, MUTYH, or CHEK2

Highlights

• No genotype-phenotype trends were observed in patients with deleterious non-BRCA mutations.

• Two patients with frameshift mutations of PALB2 rapidly developed stage IV breast cancer.

• One patient with nonsense mutation of PALB2 developed stage IA breast cancer.

• Some forms of loss-of-function variants may lead to differences in clinical outcome.

Summary

Germline BRCA mutations account for a significant proportion of genetic/familial risk of breast and ovarian cancer (GBOC) susceptibility, but a broader spectrum of GBOC susceptibility genes has emerged in recent years. Genotype-to-phenotype correlations are known for some established forms of GBOC; however, whether such correlations exist for less common GBOC variants is unclear. We reviewed our institution's experience with non-BRCA GBOC, looking specifically for trends in pathologic and clinical features. Eighteen women with deleterious germline mutations in RAD51C (5 patients), BARD1 (1 patient), BRIP1 (2 patients), PALB2 (3 patients), MUTYH (2 patients), or CHEK2 (5 patients) were identified between January 2011 and December 2016. Thirteen (72%) of 18 patients developed carcinoma of the breast, fallopian tube, or ovary, with 1 patient developing 2 separate primary neoplasms. Twelve (86%) of 14 tumors occurred in the breast. One (7%) arose in the fallopian tube and another (7%) arose in the ovary. Evidence of genotype-phenotype correlation was not identified. However, some data suggest that the type of alteration in select genes may influence tumor behavior and patient outcome. In our PALB2 mutation cohort, 2 patients with frameshift mutations led to early onset and rapid progression to stage IV breast cancer in contrast to stage IA breast cancer in 1 patient with a nonsense mutation. Despite no apparent genotype-phenotype trends, our data indicate that some loss-of-function variants in PALB2 may lead to differences in tumor behavior and patient outcome.

Introduction

It is estimated that 12.4% and 1.3% of women in the United States will, respectively, develop breast or ovarian cancer in their lifetime [1]. Although the majority of these cases are sporadic, a small but substantial subset of patients develops breast or ovarian/fallopian tube cancer due to inheritance of one or more variably penetrant germline mutations, and early identification of at-risk patients may lead to life-saving prophylactic intervention. Mutations in BRCA1 and BRCA2 are the prototypes of genetic/familial risk of breast and ovarian cancer (GBOC), and although germline BRCA mutations account for a significant proportion of GBOC susceptibility, other genes have also been recognized to increase breast and/or ovarian cancer risk.

Many national oncologic societies encourage germline genetic testing: whether indicated by trends in pedigree or statistical model calculators that estimate an individual's risk based on factors such as patient age and family history [2], [3]. As a result, recognition of a broader spectrum of GBOC susceptibility genes has grown in recent years as risk assessment has become a routine component of patient evaluation and, in parallel, sequencing technology has become more accessible, affordable, and comprehensive [4], [5].

Most studies of the pathogenicity of non-BRCA GBOC mutations have surveyed affected patients on an epidemiologic scale to register shared clinical attributes to potentially identify additional unknown, but at risk, individuals [6], [7], [8]. Known genotype-to-phenotype correlations exist for BRCA1-associated breast carcinoma [9] such as syncytial morphology with expansile, pushing invasion; robust tumor-infiltrating lymphocytes; and “triple-negative” (TN) hormone receptor and HER2/neu status as well as solid, pseudoendometrioid and transitional-like (SET) pattern of BRCA1/2-associated high-grade serous carcinoma (HGSC) of the fallopian tube and ovary [10], [11]. However, few studies have increased resolution of surveys of non-BRCA GBOC mutations to the level of the individual patient in a detailed examination of pathologic findings in specimens. A recent review of 10 families with deleterious RAD51C mutations (predominantly missense, insertions, and nonsense mutations) noted that although there was no difference in the clinical or pathologic aspects of RAD51C-associated ovarian carcinomas when compared with sporadic carcinomas, their cohort of patients with breast carcinoma strongly tended to early stage, hormone receptor–positive, HER2-negative invasive ductal carcinoma (IDC) of no special type (NST), suggesting that RAD51C-associated breast carcinomas may have more favorable pathologic characteristics [12]. Yet, another study of 2 independent families with large heterozygous deletions of RAD51C found early onset, advanced-stage, and a high-grade TN phenotype IDC in a total of 3 individuals in those families [13]. A potential explanation for opposing phenotypes is in the downstream consequences of specific loss-of-function alterations. Alternate explanations include differing genetic backgrounds of the affected individuals studied such as additional underlying protective and/or predisposing variants in other genes/pathways or inadvertent participant selection bias. Both reports concluded that additional RAD51C mutation cohorts were necessary to clarify this relationship.

In an effort to improve understanding of the increasing spectrum of non-BRCA hereditary mutations and expand reported cases with comprehensive pathologic aspects of patient specimens, we queried our institutional experience with non-BRCA GBOC, looking specifically for trends in gross and morphologic features and their relationship, if any, to clinical outcome.