Computational prediction and experimental validation associating FABP-1 and pancreatic adenocarcinoma with diabetes
1 Department of Gastroenterology and Hepatology, Stanford University School of Medicine, Alway Building, Room M211, 300 Pasteur Drive, MC: 5187, Stanford, CA, 94305-5187, USA
2 Department of Pediatrics, Stanford University School of Medicine, 300 Pasteur Drive, Room H310, Stanford, CA, 94305-5208, USA
3 Lucile Packard Children's Hospital, 725 Welch Road, Palo Alto, CA, 94304, USA
4 Biomedical Informatics Graduate Training Program, Stanford University School of Medicine, 251 Campus Drive, MSOB, x-215, MC: 5479, Stanford, CA, 94305-5479, USA
5 Department of Pathology, Stanford University School of Medicine, 300 Pasteur Drive, Lane 235, Stanford, CA, 94305-5324, USA
BMC Gastroenterology 2011, 11:5 doi:10.1186/1471-230X-11-5Published: 20 January 2011
Pancreatic cancer, composed principally of pancreatic adenocarcinoma (PaC), is the fourth leading cause of cancer death in the United States. PaC-associated diabetes may be a marker of early disease. We sought to identify molecules associated with PaC and PaC with diabetes (PaC-DM) using a novel translational bioinformatics approach. We identified fatty acid binding protein-1 (FABP-1) as one of several candidates. The primary aim of this pilot study was to experimentally validate the predicted association between FABP-1 with PaC and PaC with diabetes.
We searched public microarray measurements for genes that were specifically highly expressed in PaC. We then filtered for proteins with known involvement in diabetes. Validation of FABP-1 was performed via antibody immunohistochemistry on formalin-fixed paraffin embedded pancreatic tissue microarrays (FFPE TMA). FFPE TMA were constructed using148 cores of pancreatic tissue from 134 patients collected between 1995 and 2002 from patients who underwent pancreatic surgery. Primary analysis was performed on 21 normal and 60 pancreatic adenocarcinoma samples, stratified for diabetes. Clinical data on samples was obtained via retrospective chart review. Serial sections were cut per standard protocol. Antibody staining was graded by an experienced pathologist on a scale of 0-3. Bivariate and multivariate analyses were conducted to assess FABP-1 staining and clinical characteristics.
Normal samples were significantly more likely to come from younger patients. PaC samples were significantly more likely to stain for FABP-1, when FABP-1 staining was considered a binary variable. Compared to normals, there was significantly increased staining in diabetic PaC samples (p = 0.004) and there was a trend towards increased staining in the non-diabetic PaC group (p = 0.07). In logistic regression modeling, FABP-1 staining was significantly associated with diagnosis of PaC (OR 8.6 95% CI 1.1-68, p = 0.04), though age was a confounder.
Compared to normal controls, there was a significant positive association between FABP-1 staining and PaC on FFPE-TMA, strengthened by the presence of diabetes. Further studies with closely phenotyped patient samples are required to understand the true relationship between FABP-1, PaC and PaC-associated diabetes. A translational bioinformatics approach has potential to identify novel disease associations and potential biomarkers in gastroenterology.