Integrating Molecular Testing With Clinical Criteria and Histopathology Improves Diagnostic Precision in Immune-Mediated Liver Diseases

Abstract

Autoimmune hepatitis (AIH) and primary biliary cholangitis (PBC) are immune-mediated liver diseases (IMLDs) that are diagnosed by a combination of clinical, serologic, and histologic features. Diagnosis may be challenging, particularly when patients have mixed features of both AIH and PBC, a disease often called overlap syndrome (OS). In addition, many patients have refractory disease. We hypothesized that adding molecular testing to the current diagnostic criteria would provide an additional tool that could assist in correctly classifying patients. RNA was isolated from liver biopsies from patients with AIH (n = 16), PBC (n = 13), OS (n = 8), drug-induced/serology-negative AIH (AIH DI/Ser-neg, n = 6), or controls (n = 10). Gene expression was determined using an nCounter Sprint Profiler, and principal component analysis delineated distinct clusters for patients with an inflammatory profile due to AIH, PBC, and AIH DI/Ser-neg. Two patients with minimal histologic features of PBC clustered with the control group, and 2 patients with predominantly AIH and minimal PBC features clustered with the PBC group. A patient with OS who received treatment for both conditions showed no disease progression, whereas a patient with OS treated solely for AIH failed to respond. Conversely, one of the gene signatures from a patient diagnosed with PBC fell within the AIH group. This patient did not respond to treatment with ursodiol and ultimately required liver replacement. These findings suggest that the IMLD initially diagnosed in these patients may have been incorrectly classified. As expected, molecular analysis could not identify a distinct cluster for patients diagnosed with OS, and these had variable gene signatures that fell throughout the identified AIH or PBC groups. Cluster analysis was also able to distinguish patients with disease progression from non-progressors with mild disease who responded to treatment. In summary, gene expression analysis may assist in confirming the type of IMLD, especially when the diagnosis is unclear. Combining molecular testing with existing criteria could provide an additional diagnostic tool, improving patient care and response to treatment.