Telomerase Activity for the Preoperative Diagnosis of Pancreatic Cancer

Among patients with cancer, those with pancreatic cancer have one of the lowest 5-year survival rates ( 1 ). Pancreatic cancer is usually diagnosed by ultrasonography, computed tomography, endoscopic ultrasonography, endoscopic retrograde cholangiopancreatography (ERCP), and cytology ( 2 ). However, it is sometimes difficult to distinguish benign from malignant tumors with the use of these diagnostic tools ( 3 , 4 ). K-ras mutations in pancreatic secretions have been reported to be useful for the diagnosis of cancer ( 5 ) but were also detected in chronic pancreatitis ( 6 ). Therefore, K-ras mutations are not considered to be a specific marker for pancreatic cancer. Telomerase is an enzyme that contains an RNA template complementary to the short DNA sequence repeats (GGTTAG in humans) present at chromosomal ends (i.e., telomeres); the enzyme is believed to be involved in the de novo synthesis at those sites ( 7 ). A highly sensitive polymerase chain reaction-based telomerase assay called TRAP (Telomeric Repeat Amplification Protocol) was developed ( 8 , 9 ), and led to the detection of telomerase activity in almost all cancer tissues ( 10 – 12 ). Recently, in surgically resected pancreatic tissue specimens, we detected telomerase activity in 41 (95%) of 43 cancer specimens but none (0 of 11) in benign tumor specimens ( 13 ). These findings suggest that telomerase activity in pancreatic duct cells may be useful for cancer diagnosis prior to surgery. Telomeric Repeat Amplification Protocol (TRAP) assay of pancreatic duct cells. For each tissue sample, an aliquot of the extract derived from 1000 cells with (+) or without (−) RNase pretreatment was used in each TRAP assay. An extract of a human lung cancer cell line (SBC-1) having telomerase activity was used as a positive control. Telomerase activity was detected as a 6-nt processive ladder. A 150-base-pair internal telomerase assay standard (ITAS) was used to identify noninformative specimens because of inhibitors of Taq polymerase affecting the TRAP assay. Most pancreatic cancer samples (case subjects 6, 7, 13, 14, 2, and 12) had detectable telomerase activity, whereas samples derived from two pancreatic cancers (case subjects 1 and 8), three chronic pancreatitis (case subjects 15, 17, and 18), and a benign islet cell tumor (case subject 31) did not have detectable telomerase activity. Therefore, in this study, telomerase was assayed by use of pancreatic duct cells obtained preoperatively. In almost all cases of suspected pancreatic tumors, ERCP is performed and pancreatic duct cells are collected by use of endoscopic retrograde pancreatic duct brushing (ERPDB). ERPDB has been shown to be a useful method for obtaining cells from the sites of pancreatic duct abnormalities ( 4 ). From January 1996 through June 1997, pancreatic duct cells were collected by use of ERPDB from 32 patients with pancreatic duct abnormalities at the Hiroshima University Hospital under informed consent. Institutional guidelines for the use of patients' materials were followed. The 21 male and 11 female patients ranged in age from 34 to 78 years (mean, 62 years). Part of each specimen was processed for cytologic analysis with the use of Pap staining and was then examined by a pathologist (F. Shimamoto) who made a diagnosis according to guidelines reported by Robins et al. ( 14 ). After washing the remaining cells with phosphate-buffered saline, the TRAP assay was performed as described previously ( 13 ). In a preliminary study, we examined telomerase activity without making an adjustment for cell number, and weak telomerase activity was detected in three of three chronic pancreatitis cases with massive infiltration of lymphocytes when we used more than 10 4 cells (data not shown). We reported previously that the extracts from 10 3 adult lymphocytes did not show detectable telomerase activity but those from 10 4 lymphocytes sometimes did ( 15 ). To avoid false-positive results that can occur when infiltrating lymphocytes are present, the number of cells was counted and extract derived from 10 3 cells was used for each TRAP assay in all subsequent experiments ( Fig. 1 ). Table 1 shows the TRAP results of 32 case subjects that consisted of 15 pancreatic cancers and 17 benign pancreatic diseases (12 chronic pancreatitis, two pancreatic cysts, one islet cell tumor, one pancreatic abscess, and one solid and cystic tumor). Telomerase activity was undetectable in all 17 subjects with benign pancreatic diseases, whereas 13 (87%) of 15 subjects with pancreatic cancer had detectable telomerase activity. In these case subjects with pancreatic cancer, all cancer tissues obtained by subsequent surgical resection had detectable telomerase activity. Cytologically, no atypical cells were observed in a cell specimen (stained by the Pap method) that has been obtained from case subject 1 ( Table 1 ), and the sample did not have detectable telomerase activity. However, the diagnosis of pancreatic cancer in that patient was based on roentgenographic abnormalities and was subsequently determined to be cancer upon surgery. This false-negative result in case subject 1 was likely because of the failure of the ERPDB procedure to collect cancer cells. The case subject 8 did not show detectable telomerase activity, whereas it was determined by cytologic examination to be malignant, indicating that the collected cells used in the TRAP assay may not have had sufficient telomerase-positive cancer cells. Specimens from six case subjects (Nos. 2 through 7), originally diagnosed as benign by cytology, showed detectable telomerase activity and on surgery were determined to be pancreatic cancer. The specimens from case subjects 2 through 4 had cellular degeneration and were difficult to diagnose cytologically. The specimens from case subjects 5 through 7 had several mild atypical cells but were also difficult to diagnose as cancer. In these six cytologically false-negative specimens from case subjects 2 through 7, telomerase activity could have potentially helped in the final diagnosis of pancreatic cancer. Interestingly, for case subject 2 who was determined by postoperative pathologic examination to have carcinoma in situ , we detected telomerase activity, even though other diagnostic methods were negative. Carcinoma in situ of the pancreas has a good prognosis when surgically removed. Our overall results suggest that telomerase activity of pancreatic duct cells obtained by ERPDB may be of value as a supplemental test for pancreatic cancer diagnosis. In the patients with pancreatic abnormalities, it is very difficult to diagnose early pancreatic cancers, especially carcinoma in situ , by roentgenographic findings and cytology. If future studies confirm and extend the present findings, the detection of telomerase activity of pancreatic duct cells may become a useful tool for diagnosis of pancreatic cancer, perhaps at early stages.