Occult Cirrhosis can be Uncovered by Evaluation of Unexplained Thrombocytopenia

Abstract

Purpose: To evaluate unexplained thrombocytopenia as a marker for detecting previously undiagnosed liver cirrhosis. Methods: A retrospective database was created by a search of the VA New York Harbor Health System electronic medical software from 2008 until 2010, using a combination of cirrhosis-related ICD9 codes and thrombocytopenia of <150K. From July through October of 2010, the identified patients were classified as either cirrhosis by ICD9 code, chart documentation or clinical evaluation; no cirrhosis; or indeterminate patients requiring further workup. Over time, a portion of the indeterminate patients were seen in either the hepatology or the gastroenterology clinics and were sent for further workup to evaluate for underlying chronic liver disease with cirrhosis. After 20 months, an analysis was performed to examine rates of cirrhosis in these patients. Results: Of the 497 patients originally grouped as indeterminate, 382 were randomly chosen for analysis. Of these, 112 patients were seen in a GI or hepatology clinic, while 259 had not. Of the patients seen in clinic, 51 (45.5%) have not yet been fully worked up. Of the 62 patients who have been fully evaluated, 50% were found to have cirrhosis (diagnosed via liver-spleen scan, other radiological studies, EGD displaying esophageal varices, clinical impression, or biopsy), while 50% did not. Of the 31 patients found to have cirrhosis, 4 (12.9%) have since been diagnosed with hepatocellular carcinoma. All of the cases were due to either hepatitis C, alcoholic liver disease, or non-alcoholic fatty liver disease. Analysis of differences in the AST to platelet ratio index (APRI) displayed a mean of 1.41 (95% CI 1.00-1.82) amongst the cirrhotics, in contrast to a mean of 0.64 (95% CI 0.51-0.78) amongst those without cirrhosis (p<0.001). While 77.4% of the patients who were noted to have cirrhosis had a persistent pattern of thrombocytopenia, only 58.1% of the patients who did not have cirrhosis displayed a similar pattern, although this difference was not noted to be statistically significant (p=0.11). Conclusion: Many patients with cirrhosis, who have few other clinical clues beyond unexplained thrombocytopenia, go unrecognized. Our data collection is ongoing but at this time, half of the patients with unexplained thrombocytopenia have displayed evidence of cirrhosis. In patients with unexplained thrombocytopenia and clinical suspicion for liver disease, APRI is an excellent tool to identify occult cirrhotics. In addition to better tools for identification of cirrhotics, educational programs are needed to raise awareness in primary care providers and hematologists as to the importance of persistent thrombocytopenia in patients with risk factors for liver disease.