A 29-year-old male originating from the Indian subcontinent (a region with a high incidence of TB) and currently residing in Qatar for the past three years presented to the emergency department with a three-day history of worsening epigastric and right upper quadrant pain. The pain was exacerbated by food intake and associated with dyspepsia and self-induced vomiting. He also reported passing dark urine for three days and one episode of pale stool. He denied any fever, pruritus, or constitutional symptoms such as night sweats or weight loss. He had no known sick contacts, no family history of TB, and no history of travel outside Qatar in the past two years. He was a non-smoker and consumed alcohol only occasionally (approximately one to two units per month). He worked as a construction worker and lived with his family in shared accommodation. He had received the Bacillus Calmette-Guérin or BCG vaccination as a child. He had no history of diabetes mellitus, immunosuppressive medication use, or previous TB treatment.

On physical examination, his vital signs were: temperature 36.8°C, heart rate 78 beats per minute, blood pressure 125/78 mmHg, respiratory rate 16 breaths per minute, and oxygen saturation 98% on room air. He was alert and oriented, in no acute distress. He was noted to have scleral icterus. No conjunctival pallor, jugular venous distention, or peripheral lymphadenopathy (cervical, supraclavicular, axillary, or inguinal) was present. Cardiovascular examination revealed regular rate and rhythm with no murmurs, rubs, or gallops. Respiratory examination demonstrated clear breath sounds bilaterally without wheezes or crackles. Abdominal examination revealed mild tenderness in the epigastrium and right upper quadrant, without guarding or rigidity. Murphy's sign was negative. No palpable masses, organomegaly, or hepatosplenomegaly were noted. Bowel sounds were normal. Skin examination showed no rash, though jaundice was noted. Body mass index was 22.4 kg/m² (within the normal range).

Initial laboratory investigations revealed a mixed cholestatic and hepatocellular injury pattern. Comprehensive laboratory testing is presented in Table 1.

An abdominal ultrasound demonstrated a dilated CBD measuring 10 mm, an overdistended gallbladder measuring 11.5 × 4.2 cm with sludge layering dependently, and a 7.6 mm sessile polyp along the posterior wall; no gallstones or wall thickening were identified. The intrahepatic biliary radicals were mildly prominent. The pancreas appeared normal without a mass or a focal lesion. The visualized portions of the liver were homogeneous without focal lesions. Peripancreatic and portacaval lymph nodes were not clearly visualized due to bowel gas (Figure 1).

Given suspicion for distal CBD obstruction, magnetic resonance imaging (MRI) and magnetic resonance cholangiopancreatography (MRCP) were performed and interpreted by a board-certified radiologist with 12 years of experience in abdominal imaging. This revealed a conglomerate of necrotic lymph nodes in the precaval and portacaval region, extending into the pancreatic head and encasing the intrapancreatic portion of the CBD. This resulted in smooth narrowing of the distal CBD with significant upstream biliary dilatation (Figure 2). The largest nodal mass measured 44 × 25 mm. The radiological features, including the conglomerate morphology and necrotic appearance of the nodes, with central low signal on T1 and high signal on T2, were highly suggestive of tuberculous lymphadenitis, making a malignant etiology, such as pancreatic cancer or lymphoma, less likely.

A contrast-enhanced CT scan of the abdomen was also performed, which confirmed conglomerated necrotic lymphadenopathy in the portacaval and precaval regions with central low attenuation (necrosis) and peripheral rim enhancement, a finding characteristic of tuberculous lymphadenitis. No pancreatic mass was identified, and the visualized lung bases showed no pulmonary parenchymal lesions.

MRCP demonstrated smooth tapering of the distal CBD with significant upstream dilatation of the CBD and intrahepatic biliary radicals. The pancreatic duct was normal in caliber, consistent with extrinsic compression without intraluminal pathology.

The patient was admitted for observation and started on intravenous cefuroxime and metronidazole due to a concern for developing cholangitis. Endoscopic retrograde cholangiopancreatography (ERCP) was not performed. The clinical rationale for avoiding ERCP included: (1) the patient had mild cholangitis without systemic features (no fever, hypotension, or organ dysfunction), (2) definitive diagnosis was pursued via endoscopic ultrasound (EUS)-guided biopsy, which also allowed tissue acquisition, (3) EUS-guided fine needle biopsy (EUS-FNB) is preferred over ERCP when tissue diagnosis is needed and cholangitis is mild, and (4) ERCP with stenting is reserved for severe cholangitis, failed medical therapy, or when biliary drainage is urgently needed, consistent with current guidelines.

To obtain a definitive diagnosis, an EUS was performed, and a peripancreatic lymph node biopsy was obtained using a 22-gauge Acquire™ fine needle biopsy (FNB) needle, which provided core tissue for histopathological analysis. The obtained samples were sent for cytology, histopathology, TB polymerase chain reaction (Xpert MTB/RIF Ultra assay), mycobacterial culture, and flow cytometry to rule out lymphoma. Histopathological examination revealed necrotizing granulomatous inflammation with central caseous necrosis, epithelioid histiocytes forming granulomas, multinucleated giant cells (Langhans type), and lymphocytic cuffing. Ziehl-Neelsen staining for acid-fast bacilli (AFB) was negative, which can occur in paucibacillary TB. The Xpert MTB/RIF Ultra assay returned positive for Mycobacterium tuberculosis complex DNA, with no rifampicin resistance mutations detected. Mycobacterial culture (MGIT system, Beckton-Dickinson, USA) was sent and remains pending at the time of submission (the eight-week incubation period is not yet complete). Concurrently, an AFB smear and sputum analysis were negative. QuantiFERON-TB Gold Plus assay was performed and returned positive. The patient was also tested for HIV, which was negative. The final diagnosis was obstructive jaundice secondary to tuberculous peripancreatic lymphadenitis with extrinsic compression of the CBD.

Given the significant baseline hepatocellular injury (with alanine aminotransferase (ALT) peaking at 465 U/L, representing >5 times the upper limit of normal), a hepatoprotective anti-tubercular regimen was initiated to minimize the risk of drug-induced liver injury. According to the World Health Organization (WHO) and American Thoracic Society (ATS) guidelines, patients with a baseline ALT >3 times the upper limit of normal should avoid hepatotoxic first-line agents (isoniazid, rifampicin, and pyrazinamide). This decision was made in accordance with guidelines for managing anti-tubercular therapy (ATT) in patients with significant baseline hepatitis or those at high risk of hepatotoxicity. After consultation with the infectious diseases team, the patient was started on moxifloxacin 400 mg daily, ethambutol 1200 mg daily, amikacin 15 mg/kg intravenously (transitioned to intramuscular after two weeks), and linezolid 600 mg daily, along with pyridoxine 50 mg daily. This regimen was chosen due to concerns for hepatotoxicity with standard first-line therapy (isoniazid, rifampicin, and pyrazinamide) in the setting of significant baseline liver dysfunction. A transition to a standard rifampicin-based regimen was planned once liver enzyme levels normalized (ALT <40 U/L). Serial liver function tests showed a progressive downward trend (Table 2). The patient demonstrated significant clinical improvement with resolution of his pain and jaundice within seven days of initiating therapy.

The patient was discharged home on the same regimen after 10 days of hospitalization with close outpatient follow-up. Serial liver function tests at two weeks, four weeks, three months, and six months showed progressive normalization, with all parameters within normal range by three months. Follow-up abdominal ultrasound at three months demonstrated a normalized CBD at 6 mm, persistent but smaller peripancreatic lymph nodes (largest 28 mm), and resolution of gallbladder sludge. Also, follow-up abdominal ultrasound at six months demonstrated a normal CBD (5 mm), significant regression of lymphadenopathy (the largest node now 12 mm), and normal gallbladder appearance. Follow-up MRI at six months confirmed near-complete resolution of the previously observed conglomerate lymphadenopathy with no residual biliary dilatation and no new lesions.

The patient remained clinically well at six-month follow-up, with complete resolution of symptoms and return to normal activities. He will continue ATT for a total of six months (two months of the intensive phase with the current regimen, followed by four months of continuation phase with rifampicin and isoniazid once liver enzymes have fully normalized).