Correlation of Anti-mitochondrial Antibodies with Liver Histology and Outcomes

A recent article published in your journal got our attention and we wanted to make a comment [1]. The article reported among a select group of patients with nonalcoholic fatty liver disease (NAFLD) or alcoholic liver disease (ALD) that the presence of non-organ-specific antibodies (NOSA) does not impact either liver histology or severity of liver disease. Of the 15 liver biopsies available on patients with NAFLD and ALD, only one showed evidence of probable autoimmune hepatitis. Interestingly, of the three patients with positive AMA (anti-mitochondrial antibody), only one had evidence of bile duct injury, while one patient who tested negative for AMA had normal alkaline phosphatase, but showed bile duct injury on histology. AMA has been reported to be highly specific for the diagnosis of PBC [2]; however, the prevalence of AMA has been variably reported at 6.66 % (2/84) in NAFLD, 8 % (18/237) in hepatitis C, and 34 % (14/41) in type 1 autoimmune hepatitis. While AMA positivity correlated with the diagnosis of PBC among all type 1 autoimmune hepatitis patients, only 5/12 liver biopsies among HCV patients had evidence of PBC and the one patient with NAFLD who underwent liver biopsy did not have any evidence of bile duct injury [3–5]. We looked at our own data of 90 patients who tested positive for AMA (mitochondrial, M2), and 43 liver biopsies were available to assess the frequency of bile duct injury, NAFLD, and overlap of these histologic features. Overall, we found 42 % (18/43) patients with changes of NAFLD and 53 % (23/43) with evidence of bile duct injury on the liver biopsies. Overlap of fatty liver and bile duct injury was present in 16 % (7/43). One of our patients like the one reported by Ravi et al. had evidence of bile duct injury with normal alkaline phosphatase, but positive AMA. Further analysis of our data showed that there was no difference in alcohol consumption and BMI among NOSA-positive and NOSA-negative patients; however, a significant correlation of NOSA was seen with ascites, varices (imaging/endoscopy), and cirrhosis (liver biopsies) compared to patients without NOSA. Ascites and varices were observed in 10/40 (25 %) and 10/40 (25 %) patients with at least one of the autoimmune markers and 3/37 (8.1) and 4/34 (11.8 %) with neither ANA nor F-actin, respectively. Cirrhosis on liver biopsies was present in 7/23 (30.4 %) of patients with and 4/17 (23.5 %) without NOSA, respectively. Data reported by Ravi et al. and us suggest that NAFLD is not uncommonly seen among patients who test positive for NOSA. Moreover, AMA is infrequently positive among NAFLD patients, but patients who test positive for AMA have a higher frequency of NAFLD. The correlation of AMA with liver histology compatible with PBC was not shown to be as high in these reports as has been previously described. Additionally, the outcomes of liver disease were not shown to be impacted by NOSA among NAFLD patients of Ravi et al.; however, in our experience, NOSA are associated with worse liver disease outcomes among patients with primarily an underlying autoimmune disease such as PBC—a finding previously reported as well [6]. & Abdul Nadir Anadir786@aol.com