Clinical challenges in the management of hepatic sarcoidosis

INTRODUCTION Sarcoidosis is a multisystem disease of unknown aetiology, characterised by the presence of non-caseating granulomas, most commonly in the lungs and intrathoracic lymph nodes.1,2,3 Although hepatic involvement in systemic sarcoidosis has been reported in 50%–80% of autopsy studies,2,4,5 isolated hepatic sarcoid is rare. The clinical and diagnostic features of predominant hepatic sarcoidosis appear to be variable. While many cases in these studies were reported to be asymptomatic,2,6,7 some reported symptoms such as fatigue (12%), arthralgia (12%) and fever (4%–50%).4,8,9 Jaundice is rare (1.6%), and some patients (5%–15%) have hepatomegaly.9,10 Progression to cirrhosis (6%–26%) and portal hypertension (3%–20%) has been reported as well.4,7,11,12 The biochemical parameters of hepatic sarcoidosis are mainly increase in the activities of alkaline phosphatase (ALP) and/or gamma-glutamyltransferase (GGT), which are more than two times the upper limit of normal (ULN). Aminotransferase elevation was usually less than two times the ULN.2,3,7 Imaging studies may show normal liver, hepatomegaly or multiple hypointense liver nodules.2 Treatment of hepatic sarcoidosis is not well established.4,13,14,15,16 What can be extrapolated from the general treatment of sarcoidosis is that corticosteroids should be the first-line treatment. However, given the toxicity of long-term corticosteroids and limited data on their efficacy in hepatic sarcoidosis, other more potent treatments, namely methotrexate (MTX), azathioprine and tumour necrosis factor (TNF) inhibitors such as infliximab, are increasingly being used to achieve remission.2,4,15,17 With the ambiguity of the clinical manifestation and treatment of hepatic sarcoidosis, more data are needed to understand this condition further. In this study, we report a series of seven cases of hepatic sarcoidosis in Asian patients to illustrate the challenges in treatment and its outcome. METHODS Between 2011 and 2019, there were seven patients diagnosed with hepatic sarcoidosis in University of Malaya Medical Centre, a tertiary hospital in Malaysia. All patients were female and of Indian ethnicity. The median (interquartile range IQR) age was 50 (41–58) years, and the median (IQR) duration of follow-up was 6 (3–8) years. The prevalence of metabolic risk factors among the seven patients with hepatic sarcoidosis in our series was as follows: type 2 diabetes mellitus (n = 5), hypertension (n = 1), dyslipidaemia (n = 1) and obesity (n = 2). Weight loss was the predominant symptom, and deranged liver function tests (LFT) denoting cholestatic liver injury was observed during initial evaluation. This study was approved by the Medical Ethics Committee (MECID 2020612-8750). All participants provided written informed consent. RESULTS Clinical presentation, LFT and markers of hepatic synthetic function at the time of diagnosis are summarised in Table 1. In all the patients, GGT was elevated (>30 IU/L) and more than two times the ULN (ULN at our centre 30 IU/mL). In six (85.7%) patients, ALP was twice the ULN (normal range at our centre 45–140 IU/L). The activities of aspartate transaminase and alanine transaminase were elevated (>40 IU/ mL) in 85.7% and 71.4% of patients, respectively. Of the seven patients, 42.9% and 71.4% had hyperbilirubinaemia (>35 μmol/L) and hypoalbuminaemia (<30 mg/L), respectively. All patients had normal prothrombin time and international normalised ratio. Computed tomography (CT) had been performed in all cases. Coronal CT image of one of the patients Figure 1 shows typical features of hepatomegaly with heterogeneity and a 'nutmeg' appearance in the liver.Table 1: Summary of clinical presentation, liver function test and markers of the liver's synthetic function of patients with hepatic sarcoidosis from 2011 to 2019.Figure 1: Representative coronal CT image of a patient with hepatic sarcoidosis.All patients underwent percutaneous liver biopsy, and granulomatous inflammation was present in all seven cases. Figure 2 shows the representative histopathological images of three patients in this study. The extrahepatic manifestations observed were pulmonary and lymphadenopathy (n = 6) and splenic involvement (n = 4).Figure 2: Representative histopathological images of a liver biopsy sample from three patients with hepatic sarcoidosis show (a) sarcoid granulomas with no casseation (haematoxylin however, four patients developed liver cirrhosis despite therapy, and this included the two patients who were treated with MTX and infliximab. For the four patients who developed cirrhosis, there was no baseline exclusion of liver fibrosis. However, initial imaging did not show any evidence of cirrhosis or portal hypertension. All four patients had negative viral and autoimmune serology and no history of alcohol excess but had concurrent diabetes mellitus. Initial abdominal imaging did not reveal evidence of fatty liver disease, and the subsequent liver biopsies, which demonstrated features of sarcoidosis, did not have significant steatosis. Hence, we are confident that these patients developed cirrhosis due to hepatic sarcoidosis. Among the four patients who developed liver cirrhosis despite treatment, two patients progressed to portal hypertension, which was detected during routine hepatobiliary system ultrasound examination. On upper gastrointestinal endoscopy, only one patient was diagnosed with grade 1 oesophageal varices. None of the patients developed ascites or hepatocellular carcinoma during the study period. One patient (case 4) developed pulmonary tuberculosis (TB) 2 years after her diagnosis of sarcoidosis and completed anti-tuberculous therapy. The patient was initially managed by the respiratory team in 2013 with imaging abnormalities of both lung and liver. A transbronchial biopsy in 2013 did not reveal any evidence of TB, and she was diagnosed with sarcoidosis and commenced on prednisolone. Two years later, a repeat transbronchial biopsy confirmed TB infection. Hence, it is likely that TB developed as a consequence of immunosuppression. She was on a tapering dose of prednisolone of 10 mg/day at the point of diagnosis of TB. She died from a non-liver-related cause 3 years after diagnosis. Another patient (case 7) was treated for TB initially, but her TB diagnosis was subsequently revised and the appropriate therapy commenced after a month of anti-tuberculous therapy that showed no response. DISCUSSION The basic demographics of the patients with hepatic sarcoidosis in our series — Indian ethnicity, female gender and median age of 50 years — are consistent with previous reports on pulmonary sarcoidosis from Malaysia.18,19 There are no other series of predominant hepatic sarcoidosis published in Southeast Asia. Thus, we are unable to compare the demographics of our patients with hepatic sarcoidosis to any regional population. However, a similar female predominance in hepatic sarcoidosis has been widely reported in studies published in the west.2,9 A retrospective cohort study conducted in Minnesota, USA, reported a similar observation.20 The age distribution of our patients was between 30 and 59 years. This is in contrast to non-Asian data, where incidence peaks in the 20–40 years age group have been observed.9,21 The main clinical symptoms in our case series were fatigue, weight loss and abdominal pain. All patients had cholestatic liver injury at presentation, which was similar to other reports.14,22 A common feature (5/7 patients) observed in CT imaging was hepatomegaly, which has been recognised in hepatic sarcoidosis.6,15 Liver histopathological examination, with the typical features of granulomatous inflammation, remained vital in the diagnosis of hepatic sarcoidosis in our patients, usually after TB had been excluded.23 Nevertheless, the exclusion of TB is not always simple, and one of our patients was erroneously treated with anti-tuberculosis medications before she was switched to immunosuppressive therapy. The clinical manifestations and differential diagnoses of hepatic sarcoidosis are summarised in Box 1 and Table 3, respectively.Box 1: Clinical manifestations of hepatic sarcoidosis.Table 3: Differential diagnosis of hepatic sarcoidosis.The most pertinent observation in our series was the challenge of treating hepatic sarcoidosis. Due to a lack of randomised controlled trials, treatment of hepatic sarcoidosis has largely been based on case reports, retrospective cohort studies, expert opinions and extrapolation from the pulmonary sarcoidosis literature.4,24 Glucocorticoids are often the initial drug of choice when treatment is required in hepatic sarcoidosis.2,14,15 All seven patients were initiated with glucocorticoids as the first-line therapy. However, only one patient completely responded to the corticosteroid therapy without escalation of treatment. This patient who responded to prednisolone alone had mild cholestatic liver derangement (only two or three times the ULN) compared to the other six cases who failed to respond (up to 10 times the ULN of serum ALP and GGT). Kennedy et al. had previously reported that among 63 patients with hepatic sarcoidosis who were administered corticosteroids, one-third showed complete response, one-third demonstrated partial response and one-third showed no response.7 Some other studies have indicated that steroids can improve symptoms such as pruritus, jaundice, abdominal pain and fatigue. However, treatment with corticosteroids has not been recognised to prevent progression to fibrosis, chronic cholestasis and portal hypertension.2,15 In our series, four patients developed liver cirrhosis despite being on high doses of steroid therapy. Steroid-sparing agents, also known as disease-modifying anti-sarcoid drugs, offer an alternative strategy for patients with chronic or complicated sarcoidosis.25 This second-line therapy, including MTX, azathioprine, leflunomide and TNF-α inhibitor, is reserved for patients with steroid-refractory disease or intolerable adverse effects.25,26 Our experience with azathioprine as a second-line agent showed promising results; three out of five patients exhibited clinical and biochemical improvement. All five patients were started on a dosage of 50 mg/day, which was gradually escalated within an interval of 3–6 months based on treatment response. Four patients were on a maximum dose of 100 mg/day, whereas one patient was on a maximum dose of 125 mg/day. However, poor response was seen in two patients, and their treatment was switched to MTX, and eventually to infliximab. Kennedy et al. had previously reported that augmentation with azathioprine and MTX showed 50% improvement of liver enzymes among 16 patients with hepatic sarcoidosis.7 Our experience with MTX was limited due to the small number of patients. Both patients treated with MTX were initiated on a dosage of 5 mg/week, which was gradually escalated within an interval of 3–6 months. Both were given a maximum dose of 15 mg/week. However, due to failure of response to treatment, MTX was discontinued and infliximab was initiated. Infliximab has been suggested as a therapeutic option because TNF has been implicated in the development of granulomatous inflammation in sarcoidosis.4,25 Judson et al. demonstrated in a randomised clinical trial that although infliximab was beneficial in extrapulmonary sarcoidosis, they were unable to draw conclusions about its efficacy in hepatic disease due to the small sample size.27 In our series, two patients were initiated on infliximab following suboptimal response to azathioprine and MTX. Both patients responded well in terms of liver biochemistry and symptoms. Our results are congruent with the findings by Doty et al., who reported marked improvement in LFT, symptoms and hepatomegaly in ten patients with liver sarcoidosis who were initiated on infliximab following failure of treatment with MTX and methylprednisolone.17,28 Although effective, TNF-α inhibitors are associated with a range of side effects, including infectious complications such as reactivation of TB. A workflow of the suggested treatment of hepatic sarcoidosis, adapted from Deutsch-Link et al.,4 is shown in Figure 3.Figure 3: Chart shows the suggested treatment algorithm for patients diagnosed with hepatic sarcoidosis (based on Deutsch-Link S, Fortuna D, Weinberg E. A Comprehensive Review of Hepatic Sarcoid. Semin Liver Dis. 2018;38(03):284-297).This study has several limitations. Due to the rarity of this condition, the study had only a small sample size. With the limited information on hepatic sarcoidosis and scarce literature resources, there is a knowledge gap to determine the optimum treatment strategy. In conclusion, hepatic sarcoidosis has a vague clinical presentation, and its diagnosis requires histopathological evaluation and the exclusion of TB. Based on reports from the literature, together with our limited experience, high-dose corticosteroids should be used as first-line therapy, but the treatment duration remains uncertain. A partial response will necessitate the use of steroid-sparing agents, of which TNF-α inhibitor may be the most effective. Further experience with such a proposed treatment algorithm may provide useful guidance to clinicians managing this challenging condition. Financial support and sponsorship Nil. Conflicts of interest There are no conflicts of interest.