American Journal of Gastroenterology ISSN 0002-9270 C  2007 by Am. Coll. of Gastroenterology doi: 10.1111/j.1572-0241.2007.01136.x Published by Blackwell Publishing Overlap of Autoimmune Hepatitis and Primary Biliary Cirrhosis: Long-Term Outcomes Marina G. Silveira, M.D., Jayant A. Talwalkar, M.D., Paul Angulo, M.D., and Keith D. Lindor, M.D. Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester, Minnesota BACKGROUND: The coexistence of primary biliary cirrhosis (PBC) and autoimmune hepatitis (AIH) has been called “overlap syndrome,” but diagnosis is challenging and the natural history of this syndrome has not been demonstrated. The importance of the diagnosis of PBC-AIH overlap is due to potential therapeutic options. Patients with PBC should receive ursodeoxycholic acid (UDCA); the role of and response to additional immunosuppressive therapy are unknown when AIH overlaps PBC. METHODS We reviewed 135 patients with PBC according to a revised scoring system proposed by the AND RESULTS: International Autoimmune Hepatitis Group (IAHG). Twenty-six patients had features of PBC-AIH overlap and 109 did not. Mean follow-up was 6.1 yr for overlap syndrome patients and 5.4 yr in PBC patients. There was a higher rate of portal hypertension (P = 0.01), esophageal varices (P < 0.01), gastrointestinal (GI) bleeding (P = 0.02), ascites (P < 0.01), and death and/or orthotopic liver transplantation (OLT) (P < 0.05) in the overlap group. CONCLUSION: In conclusion, esophageal varices, GI bleeding, ascites, and death and/or OLT were more common in the overlap group. The higher risk of symptomatic portal hypertension and worse outcomes in patients with PBC overlap syndrome may justify the risks of immunosuppressive therapy. Large randomized studies are necessary to establish optimal therapeutic strategies. (Am J Gastroenterol 2007;102:1244–1250) INTRODUCTION The coexistence of primary biliary cirrhosis (PBC) and au- toimmune hepatitis (AIH) occurring simultaneously has been called “overlap syndrome” (1), but accurate diagnosis is chal- lenging. The diagnostic criteria for this variant form of PBC have not been standardized, nor has its frequency and ap- propriate treatment strategy been established (2). Without using universal diagnostic criteria, the prevalence of typical PBC with features of AIH has been reported in the litera- ture to range from 9 to 19% (2–4). A scoring system has been proposed by the International Autoimmune Hepatitis Group (IAHG) for the diagnosis of definite AIH (5) and its usefulness to identify patients with AIH has been previously validated (6). The use of a revised version of the IAHG scor- ing system (7) as a diagnostic tool for identifying overlap of AIH and PBC was evaluated recently (3). In addition to the obvious purposes of classification, the importance of making a clear distinction between PBC and PBC-AIH overlap is due to therapeutic options (4). Ur- sodeoxycholic acid (UDCA) is a safe and life-extending ther- apy for most patients with PBC (8), whereas corticosteroids with or without azathioprine markedly improve survival in patients with AIH (8, 9). In contrast, a benefit from immuno- suppressive therapy in PBC has not been demonstrated to date, whereas the effectiveness of UDCA in patients with AIH is also less clear. The combination of UDCA with immuno- suppressive therapy, nevertheless, could potentially provide benefits among patients with PBC and definite overlap (4, 10). However, standard corticosteroid therapy increases the risk of developing osteoporosis, particularly in peri- and postmenopausal women, which is especially a problem for middle-aged women with PBC (11). To date, it is generally accepted that patients with PBC and AIH overlap should con- tinue to receive UDCA, but it is unclear if the degree of AIH overlap of these cases may justify the addition of immunosup- pressive therapy. A previous study indicated that the features of AIH in patients with PBC may be transient and the outcome of such individuals treated only with UDCA is no different from that of those patients with PBC without features of AIH (12). Therefore, studies are needed to define the outcomes of PBC patients compared with PBC-AIH overlap patients in order to define optimal therapeutic strategies among PBC subjects with AIH overlap. In this study, we re-evaluate the PBC patients reported in a previous study of AIH overlap, which used the revised IAHG scoring system, and describe therapy and outcome of these patients (3). MATERIALS AND METHODS Patient Population and Data Collection One hundred thirty-seven patients with PBC evaluated at the Mayo Clinic in Rochester, MN, were included in the original 1244