Emily G Barr Fritcher1, Jesse S Voss1, Shannon M Brankley1, Michael B Campion1, Sarah M Jenkins2, Matthew E Keeney1, Michael R Henry1, Sarah M Kerr1, Roongruedee Chaiteerakij3, Ekaterina V Pestova4, Amy C Clayton1, Jun Zhang1, Lewis R Roberts5, Gregory J Gores5, Kevin C Halling1, Benjamin R Kipp6. 1. Department of Laboratory Medicine and Pathology, Mayo Clinic and Foundation, Rochester, Minnesota. 2. Division of Biomedical Statistics and Informatics, Mayo Clinic and Foundation, Rochester, Minnesota. 3. Division of Gastroenterology and Hepatology, Mayo Clinic and Foundation, Rochester, Minnesota; Department of Medicine, Faculty of Medicine, Chulalongkorn University and King Chulalongkorn Memorial Hospital, Thai Red Cross Society, Bangkok, Thailand. 4. Abbott Molecular, Inc, Des Plaines, Illinois. 5. Division of Gastroenterology and Hepatology, Mayo Clinic and Foundation, Rochester, Minnesota. 6. Department of Laboratory Medicine and Pathology, Mayo Clinic and Foundation, Rochester, Minnesota. Electronic address: kipp.benjamin@mayo.edu.
Abstract
BACKGROUND & AIMS: Pancreatobiliary cancer is detected by fluorescence in situ hybridization (FISH) of pancreatobiliary brush samples with UroVysion probes, originally designed to detect bladder cancer. We designed a set of new probes to detect pancreatobiliary cancer and compared its performance with that of UroVysion and routine cytology analysis. METHODS: We tested a set of FISH probes on tumor tissues (cholangiocarcinoma or pancreatic carcinoma) and non-tumor tissues from 29 patients. We identified 4 probes that had high specificity for tumor vs non-tumor tissues; we called this set of probes pancreatobiliary FISH. We performed a retrospective analysis of brush samples from 272 patients who underwent endoscopic retrograde cholangiopancreatography for evaluation of malignancy at the Mayo Clinic; results were available from routine cytology and FISH with UroVysion probes. Archived residual specimens were retrieved and used to evaluate the pancreatobiliary FISH probes. Cutoff values for FISH with the pancreatobiliary probes were determined using 89 samples and validated in the remaining 183 samples. Clinical and pathologic evidence of malignancy in the pancreatobiliary tract within 2 years of brush sample collection was used as the standard; samples from patients without malignancies were used as negative controls. The validation cohort included 85 patients with malignancies (46.4%) and 114 patients with primary sclerosing cholangitis (62.3%). Samples containing cells above the cutoff for polysomy (copy number gain of ≥2 probes) were classified as positive in FISH with the UroVysion and pancreatobiliary probes. Multivariable logistic regression was used to estimate associations between clinical and pathology findings and results from FISH. RESULTS: The combination of FISH probes 1q21, 7p12, 8q24, and 9p21 identified cancer cells with 93% sensitivity and 100% specificity in pancreatobiliary tissue samples and were therefore included in the pancreatobiliary probe set. In the validation cohort of brush samples, pancreatobiliary FISH identified samples from patients with malignancy with a significantly higher level of sensitivity (64.7%) than the UroVysion probes (45.9%) (P < .001) or routine cytology analysis (18.8%) (P < .001), but similar specificity (92.9%, 90.8%, and 100.0% respectively). Factors significantly associated with detection of carcinoma, in adjusted analyses, included detection of polysomy by pancreatobiliary FISH (P < .001), a mass by cross-sectional imaging (P < .001), cancer cells by routine cytology (overall P = .003), as well as absence of primary sclerosing cholangitis (P = .011). CONCLUSIONS: We identified a set of FISH probes that detects cancer cells in pancreatobiliary brush samples from patients with and without primary sclerosing cholangitis with higher levels of sensitivity than UroVysion probes. Cytologic brushing test results and clinical features were independently associated with detection of cancer and might be used to identify patients with pancreatobiliary cancers.
BACKGROUND & AIMS:Pancreatobiliary cancer is detected by fluorescence in situ hybridization (FISH) of pancreatobiliary brush samples with UroVysion probes, originally designed to detect bladder cancer. We designed a set of new probes to detect pancreatobiliary cancer and compared its performance with that of UroVysion and routine cytology analysis. METHODS: We tested a set of FISH probes on tumor tissues (cholangiocarcinoma or pancreatic carcinoma) and non-tumor tissues from 29 patients. We identified 4 probes that had high specificity for tumor vs non-tumor tissues; we called this set of probes pancreatobiliary FISH. We performed a retrospective analysis of brush samples from 272 patients who underwent endoscopic retrograde cholangiopancreatography for evaluation of malignancy at the Mayo Clinic; results were available from routine cytology and FISH with UroVysion probes. Archived residual specimens were retrieved and used to evaluate the pancreatobiliary FISH probes. Cutoff values for FISH with the pancreatobiliary probes were determined using 89 samples and validated in the remaining 183 samples. Clinical and pathologic evidence of malignancy in the pancreatobiliary tract within 2 years of brush sample collection was used as the standard; samples from patients without malignancies were used as negative controls. The validation cohort included 85 patients with malignancies (46.4%) and 114 patients with primary sclerosing cholangitis (62.3%). Samples containing cells above the cutoff for polysomy (copy number gain of ≥2 probes) were classified as positive in FISH with the UroVysion and pancreatobiliary probes. Multivariable logistic regression was used to estimate associations between clinical and pathology findings and results from FISH. RESULTS: The combination of FISH probes 1q21, 7p12, 8q24, and 9p21 identified cancer cells with 93% sensitivity and 100% specificity in pancreatobiliary tissue samples and were therefore included in the pancreatobiliary probe set. In the validation cohort of brush samples, pancreatobiliary FISH identified samples from patients with malignancy with a significantly higher level of sensitivity (64.7%) than the UroVysion probes (45.9%) (P < .001) or routine cytology analysis (18.8%) (P < .001), but similar specificity (92.9%, 90.8%, and 100.0% respectively). Factors significantly associated with detection of carcinoma, in adjusted analyses, included detection of polysomy by pancreatobiliary FISH (P < .001), a mass by cross-sectional imaging (P < .001), cancer cells by routine cytology (overall P = .003), as well as absence of primary sclerosing cholangitis (P = .011). CONCLUSIONS: We identified a set of FISH probes that detects cancer cells in pancreatobiliary brush samples from patients with and without primary sclerosing cholangitis with higher levels of sensitivity than UroVysion probes. Cytologic brushing test results and clinical features were independently associated with detection of cancer and might be used to identify patients with pancreatobiliary cancers.
Authors: Vaidehi Avadhani; Ezgi Hacihasanoglu; Bahar Memis; Burcin Pehlivanoglu; Krisztina Z Hanley; Uma Krishnamurti; Alyssa M Krasinskas; Adeboye O Osunkoya; Lauren M Daniels; Alexa A Freedman; Michael Goodman; Volkan Adsay; Michelle D Reid Journal: Mod Pathol Date: 2017-06-30 Impact factor: 7.842
Authors: Sumera Rizvi; Shahid A Khan; Christopher L Hallemeier; Robin K Kelley; Gregory J Gores Journal: Nat Rev Clin Oncol Date: 2017-10-10 Impact factor: 66.675
Authors: Karan Kaura; Tarek Sawas; Fateh Bazerbachi; Andrew C Storm; John A Martin; Gregory J Gores; Barham K Abu Dayyeh; Mark D Topazian; Michael J Levy; Bret T Petersen; Vinay Chandrasekhara Journal: Dig Dis Sci Date: 2019-09-30 Impact factor: 3.199
Authors: Christian Brooks; Valerie Gausman; Chanthel Kokoy-Mondragon; Khushboo Munot; Sunil P Amin; Amit Desai; Claudine Kipp; John Poneros; Amrita Sethi; Frank G Gress; Michel Kahaleh; Vundavalli V Murty; Reem Sharaiha; Tamas A Gonda Journal: Dig Dis Sci Date: 2018-01-20 Impact factor: 3.199