Diagnosis of pancreaticobiliary malignancy by detection of minichromosome maintenance proteins in cytological specimens
Lead Research Organisation:
University College London
Department Name: Medicine
Abstract
The diagnosis of pancreatic and bile duct cancer at an early stage of disease remains difficult and there are no established tests to screen patients at high-risk of developing these cancers such as those with chronic pancreatitis (20x increased risk) and primary sclerosing cholangitis (160x increased risk). Blood tests and scans are not usually accurate enough to confirm the presence of cancer so cells are usually obtained from the lining of a narrowed bile duct with a brush (brush cytology) to look for cancer under the microscope. Brush cytology however only detects cancer in 20-50% of cases which means that patients often have to undergo additional invasive procedures, leading to potential complications and delays in diagnosis. Better diagnostic tests are clearly needed.
The aim of this study is to test new potential markers called minichromosome maintenance proteins (mcm 2-7) which may improve the early detection of precancerous changes and cancers involving the bile duct, and thus enhance the longevity and quality of life of patients living with pancreatic and bile duct cancer. These proteins are necessary for DNA replication which is essential for all cancer cells to multiply and grow. We have shown that the levels of these proteins are increased in a wide variety of early and advanced cancers in comparison to benign tissues, and can be used as accurate diagnostic tests for cervical, oesophageal and bladder cancer.
We have recently completed a study of 102 patients which showed that Mcm 5 protein when detected in bile is more than three times better than brush cytology at detecting pancreatic and biliary tract cancer, with up to 80% of cancers correctly diagnosed. The accuracy if this test depends in part on how many cells there are available to test. Brush cytology contains many more cells than bile and we now have exciting early results indicating that when Mcm 5 is detected in brush cytology specimens, close to 100% of cancers may be detected. We therefore plan to confirm these results by studying this marker as part of a large multicentre study, including patients at high risk of developing pancreatic and bile duct cancer.
The aim of this study is to test new potential markers called minichromosome maintenance proteins (mcm 2-7) which may improve the early detection of precancerous changes and cancers involving the bile duct, and thus enhance the longevity and quality of life of patients living with pancreatic and bile duct cancer. These proteins are necessary for DNA replication which is essential for all cancer cells to multiply and grow. We have shown that the levels of these proteins are increased in a wide variety of early and advanced cancers in comparison to benign tissues, and can be used as accurate diagnostic tests for cervical, oesophageal and bladder cancer.
We have recently completed a study of 102 patients which showed that Mcm 5 protein when detected in bile is more than three times better than brush cytology at detecting pancreatic and biliary tract cancer, with up to 80% of cancers correctly diagnosed. The accuracy if this test depends in part on how many cells there are available to test. Brush cytology contains many more cells than bile and we now have exciting early results indicating that when Mcm 5 is detected in brush cytology specimens, close to 100% of cancers may be detected. We therefore plan to confirm these results by studying this marker as part of a large multicentre study, including patients at high risk of developing pancreatic and bile duct cancer.
Technical Summary
The diagnosis of pancreatic and biliary tract cancer at an early stage of disease remains difficult and there are currently no established screening tests available, even in the high-risk groups of chronic pancreatitis (~20-fold increased risk of pancreatic cancer) and primary sclerosing cholangitis (~160-fold increased risk of biliary tract cancer). Biliary brush cytology is the standard method of sampling a biliary stricture but has a low sensitivity (20-50%) for the detection of malignancy. We have previously shown that minichromosome maintenance (MCM) replication proteins (Mcm2-7) are markers of dysplasia and have utilised these novel biomarkers of growth for the diagnosis of cervical and bladder cancer (Stoeber et al. Lancet 1999; Stoeber et al. JNCI 2002).
Following on from these data in other solid organ tumours, we reported in a proof of principle prospective, blinded, single-centre study of 102 patients that Mcm5 levels in bile, detected by a two site time resolved immunofluorometric test with europium as a label, were significantly more sensitive (66% vs. 20%, p=0.004) than brush cytology for the detection of malignancy, with a comparable positive predictive value (97% vs. 100%, p=ns) (Ayaru et al. Br J Cancer 2008). In the same study, we also showed that Mcm 2 and 5 proteins detected in biopsy specimens are dysregulated in pancreaticobiliary malignancy. The percentage of nuclei positive for Mcm2 was higher in malignant tissue (median 76.5%, range 42-92%) than in benign tissue (median 5%, range 0-33%; p 0.0005), with similar results for Mcm5.
The sensitivity of the Mcm 5 test depends in part on the cellularity of bile, which can be acellular in up to 30% of samples, so that the sensitivity of bile aspirate Mcm5 cannot be expected to reach 100%. In contrast, the cellularity of brush cytology is much greater than bile aspirates. In a preliminary study of 20 patients with indeterminate biliary strictures, the Mcm 5 assay, when applied to biliary brush cytology specimens (n=20), had a sensitivity of 100% and an accuracy of 89% for the detection of malignancy (Ayaru et al, unpublished). We aim to extend these exciting preliminary data in a much larger group of patients with suspected pancreaticobiliary malignancy as part of a multicentre study, and to determine if the level of Mcm5 detected in brush cytology and bile specimens can distinguish accurately between the diagnostically challenging subgroups of: a) chronic pancreatitis and pancreatic cancer, and b) primary sclerosing cholangitis and biliary tract cancer.
Following on from these data in other solid organ tumours, we reported in a proof of principle prospective, blinded, single-centre study of 102 patients that Mcm5 levels in bile, detected by a two site time resolved immunofluorometric test with europium as a label, were significantly more sensitive (66% vs. 20%, p=0.004) than brush cytology for the detection of malignancy, with a comparable positive predictive value (97% vs. 100%, p=ns) (Ayaru et al. Br J Cancer 2008). In the same study, we also showed that Mcm 2 and 5 proteins detected in biopsy specimens are dysregulated in pancreaticobiliary malignancy. The percentage of nuclei positive for Mcm2 was higher in malignant tissue (median 76.5%, range 42-92%) than in benign tissue (median 5%, range 0-33%; p 0.0005), with similar results for Mcm5.
The sensitivity of the Mcm 5 test depends in part on the cellularity of bile, which can be acellular in up to 30% of samples, so that the sensitivity of bile aspirate Mcm5 cannot be expected to reach 100%. In contrast, the cellularity of brush cytology is much greater than bile aspirates. In a preliminary study of 20 patients with indeterminate biliary strictures, the Mcm 5 assay, when applied to biliary brush cytology specimens (n=20), had a sensitivity of 100% and an accuracy of 89% for the detection of malignancy (Ayaru et al, unpublished). We aim to extend these exciting preliminary data in a much larger group of patients with suspected pancreaticobiliary malignancy as part of a multicentre study, and to determine if the level of Mcm5 detected in brush cytology and bile specimens can distinguish accurately between the diagnostically challenging subgroups of: a) chronic pancreatitis and pancreatic cancer, and b) primary sclerosing cholangitis and biliary tract cancer.