Increased Incidence of Autoimmune Hepatitis Is Associated With Wider Use of Biological Drugs

Kjartan B. Valgeirsson; Jóhann P. Hreinsson; Einar S. Björnsson


Liver International. 2019;39(12):2341-2349. 

In This Article


A total of 515 possible cases were identified of which 71 received the diagnosis of AIH within the study period. The reasons for exclusion are shown in Figure 1. The average annual incidence of AIH was 2.2 per 100 000 inhabitants during the study period, 3.8 per 100 000 for women and 0.6 for men. The mean annual incidence was higher in the last 2 years of the study, 3.97 per 100 000, compared to 1.78 per 100 000 in the period from 2006 to 2013 (risk ratio 2.2, 95% confidence interval 1.4–3.6, P = .002) (Figure 2). Point prevalence, including 21 patients previously diagnosed, still alive, on the 31 December 2015 was 26.9 per 100 000.

Figure 1.

Patient identification and reasons for exclusion

Figure 2.

Annual incidence of autoimmune hepatitis in Iceland per 100 000 inhabitants

Overall 61 (86%) of the patients were of female gender. The median age at diagnosis was 56 years (IQR 44–69), 56 years for women (IQR 43.5–68.5) and 58 years for men (IQR 43–72). Females were most frequently diagnosed in their sixth decade with a lesser top in their twenties whereas frequency was similar between age groups in men (Figure 3).

Figure 3.

Gender-specific age at diagnosis among autoimmune hepatitis patients in Iceland

The vast majority of patients had markedly elevated liver tests at presentation with 79% and 72% having values more than five times upper normal limit of ALT, AST respectively. Overall 59% and 61% had ALT and AST elevations more than ten times upper limits of normal. A total of 80% patients were ANA positive, 45% were SMA positive but only one out of eleven tested was positive for anti-LKMA. Antimitochondrial antibodies (AMA) were positive in 4/63 (6%) of patients tested. In all, 23 (32%) were both positive for ANA and SMA but only four patients (6%) were seronegative (Table 2). IgG was above upper normal limits in 39/61 (64%) patients in whom results were available. All but one patient (whose results were unavailable) had negative serological tests for Hepatitis B and C virus. Histology was obtained at baseline in 61/71 (86%) patients of whom 18/61 (30%) had features typical for AIH, 42/61 (69%) had features compatible with AIH. A total of eight (13%) patients had histological features of cirrhosis at diagnosis. Drugs were suspected to have induced AIH in 13 of the 71 patients (18%); eight cases were caused by infliximab, two caused by nitrofurantoin and one each caused by etanercept, imatinib and natalizumab. Overlap syndrome was suspected with PBC in six cases and PSC in one case. Patient characteristics and laboratory data are further summarized in Table 2 and Table 3.

Overall 25/71 (35%) patients met the new simplified criteria for definite AIH and 25/71 (35%) for probable AIH but 21/71 (30%) did not fulfill the simplified criteria; of them 15/21, (71%) had a score of 5, 5/21 (23%) had score of 4 and 1/21 (5%) had a score of 3. If only patients who had available data for all four components of the score were considered 25/50 (50%) and 15/50 (30%) met the criteria for definite and probable AIH respectively but 10/50 (20%) did not fulfill the criteria.

Among those 10 patients four had DIAIH and two patients overlap with PBC. The revised original score for the diagnosis of AIH[3] was calculated for all the patients who did not meet the new simplified criteria. Of those 7/21 (33%) had a post-treatment score of >17, suggestive of definite AIH and 14/21 (67%) had a post-treatment score of 12–17, suggestive of probable AIH. None had a post-treatment score of less than 12. All 21 patients who did not fulfill the simplified criteria were started on immunosuppressive therapy because of clinical diagnosis of AIH.

Compared to patients with classical form of AIH, patients with DIAIH had lower IgG levels, had a lower frequency of SMA positive, lower NSC score and were less likely to be treated with azathioprine (Table 4). Other demographics, biochemical and clinical features were otherwise similar in the two groups (Table 4).

Treatment with either corticosteroids azathioprine or both was initiated in 69/71 (97%) cases and both medications were used in 48/71 (68%) cases, steroids only in 19/71 cases (27%) and Azathioprine only in 2/71 (3%) cases. Two patients did not receive any treatment because of spontaneous remission of symptoms and liver tests; one of those cases was considered to have been induced by Nitrofurantoin. Biochemical remission was achieved in 64 out of 69 cases (93%) where treatment was initiated according to ALT/AST values at last follow-up. A trial of treatment discontinuation was performed in 24/69 (35%) patients; of which 17/24 (71%) remained in remission but 7/24 (29%) relapsed during the follow-up time. Treatment discontinuation was successful in seven of 15 cases (43%) of classical AIH where applicated and in all nine cases of DIAIH (Table 4). Median follow-up time was 4.8 years (IQR 3.2–7.9 years). At the end of follow-up 66/71 (93%) patients were alive, one had died of liver-related mortality (hepatocellular carcinoma) but four died of unrelated causes. None of the patients in this study underwent liver transplantation during the study period. Two of the eight patients (25%) with cirrhosis at diagnosis died during follow-up compared to 3/63 (5%) patients without cirrhosis at diagnosis (P = .042).