Hepatic tuberculosis masquerading as malignancy

Background and aims: Hepatic tuberculosis (HTB) is rare and mimics neoplastic liver lesions clinico-radiologically leading to misdiagnosis and even unnecessary surgery.

Methods and material: We analysed 43 cases of HTB diagnosed at a referral cancer centre over 10 years. Clinical details, investigations and treatment received were noted.

Results: The median age was 46 years with a female preponderance (58%). HTB was diagnosed incidentally in 28% cases during surveillance imaging for a previous cancer. Constitutional symptoms (31, 72%), abdominal pain (25, 58%), fever (12, 28%), hepatomegaly (22, 51%), elevated alkaline phosphatase (34, 79%), elevated aminotransferases (18, 42%) and hypoalbuminemia (19, 45%) were common features. All cases had negative HIV serology and normal tumor markers. Twenty-two (52.5%) had solitary liver lesion and lesions > 2 cm in 28 (65%). Ultrasound showed hypoechoic lesions in 31 of 33 cases. Computed tomography showed hypodense lesions (43,100%) with mild peripheral enhancement (32, 74%). Calcifications (5, 12%) and capsular retraction (8, 19%) was uncommon. MRI was performed in seven cases commonly showed T1 hypointense, T2 hyperintense lesions with restricted diffusion. Histopathology showed granulomatous inflammation (42, 97.5%), Langhan's giant cells (41, 95%) and caseation necrosis (35, 85%). Acid-fast stain and PCR positivity was uncommon. Extrahepatic organs were involved in 20 (46.5%). HTB mimicked cholangiocarcinoma (25, 58%), liver metastasis (11, 26%) and lymphoma (3, 7%). Six patients underwent liver resection with a presumptive diagnosis of cancer without a preoperative biopsy. All patients received antitubercular therapy, 37 had clinico-radiological response, there were 3 deaths and 3 patients were lost to follow-up.

Conclusion: HTB is rare and can mimic a malignancy clinico-radiologically. Calcifications and pseudocapsule appearance on multiphase CT scan may help in differentiating HTB from hepatic malignancy. Tumor markers are normal while histopathology is generally diagnostic. A high index of suspicion is required to avoid unnecessary surgery as the patients respond well to ATT.

Trial registration: This is a retrospective and observational study hence clinical trial registration is not applicable.