Twenty-Five Percent of Breast Cancer Patients Referred for BRCA Testing Have Pancreatic Cystic Lesions on Magnetic Resonance Imaging at Time of Cancer Diagnosis

Abstract

Introduction: Recognition of early pancreatic ductal adenocarcinoma (PDAC) or precancerous pancreatic lesions may improve PDAC morbidity and mortality. Patients with hereditary breast and ovarian cancer syndrome (BRCA1/2 germline mutations) are at increased risk of PDAC. Germline mutations in Partner and Localizer of BRCA2 (PALB2) have been identified in both breast and PDAC families. Pancreatic cystic lesions (PCLs) represent PDAC precursor lesions that can be identified on computed tomography (CT) or magnetic resonance imaging (MRI). In this study, we determined the prevalence of PCLs in breast and gynecologic cancer patients. Methods: We identified patients with breast or gynecologic cancer who had an abdominal CT or MRI within 9 months of cancer diagnoses. We compared the prevalence of PCLs between BRCA-positive and BRCA-negative patients, as well as between those tested for BRCA mutations and those not referred for testing. Results: We identified 170 patients with breast cancer and CT imaging and 33 with gynecologic cancer and CT imaging. Forty-six patients had breast cancer and MRI, and 2 patients had gynecologic cancer and MRI. Eighty-three patients had BRCA testing and CT imaging, and 20 had BRCA testing and MRI. We found no difference in the prevalence of PCLs between BRCA carriers vs non-carriers on CT (0/32 vs 3/51; p=0.28) or MRI (2/9 vs 3/11; p=1.00). No differences were noted between the BRCA-tested and untested cohorts on CT (3/83 or 3.6% vs. 3/114 or 2.6%; p=1.00) or MRI (5/20 or 25% vs. 5/27 or 18.5%; p=.72). We noted a trend toward more PCLs found on CT and MRI combined in patients with breast cancer compared to gynecologic cancer (16/216 or 7.4% vs. 0/35; p=.14). In our overall cohort, PCLs were more frequently demonstrated on MRI than CT (10/47 or 21.3% vs. 6/197 or 3.0%; p<.0001). Conclusion: MRI is superior to CT for detecting PCLs in breast cancer patients. Twenty-five percent of patients referred for BRCA testing had PCLs on MRI at the time of cancer diagnosis, suggesting these patients may require surveillance of these precancerous lesions following management of breast or gynecologic tumors. We also found a trend toward more PCLs in those with breast cancer compared to those with gynecologic cancer, irrespective of BRCA status or imaging modality. Germline mutations such as PALB2 that predispose to breast and pancreatic, but not gynecologic cancer, may explain this trend. Our study was limited by a small sample size and radiology data extracted from radiology reports, not expert radiology review. We hope that the addition of patients to our database and analysis, as well as review of the radiologic imaging looking specifically for PCLs will further elucidate the trends seen in this study.