Pediatric PseudoTumoral Hepatic Tuberculosis. A Great Mimicker!!

Pediatric liver masses are rare and difficult to treat. Common liver masses in children include hepatoblastoma, hemangiomas, liver abscesses, and hydatid disease. Isolated liver tuberculosis (TB) is rare in children and can have variable clinical presentations. We report a child with isolated liver TB masquerading as a liver tumor. Copyright:

INTRODUCTION

Pediatric liver masses are uncommon and include tumors, vascular lesions, and infective pathologies such as hydatid disease in endemic areas.[1] Isolated tuberculosis (TB) of the liver, also known as pseudotumoral hepatic TB (PTHT), is a rare entity.[12] PTHT presenting as a liver mass is unusual in children, and the diagnosis may be delayed due to the nonspecific clinical features, and lesions being mistaken as malignancy.[2] Some patients even undergo hepatic resections in spite of the modern imaging modalities. We present one such case who was admitted with the working diagnosis of liver malignancy based on imaging. The patient was eventually diagnosed as isolated hepatic TB and managed successfully with antitubercular treatment (ATT).

CASE REPORT

A 10-year-old female child presented with a history of high-grade fever for 1 month and intermittent pain in the right upper quadrant of the abdomen and progressive weight loss. There was no history of vomiting, diarrhea, loss of appetite, or jaundice. There was a history of self-resolving skin abscess in the epigastric region around 3 months ago. Physical examination revealed low body weight of 23 kg (−1.5 z). There was pallor, a healed scar in the epigastrium, tenderness in the right hypochondrium, mild hepatomegaly, and skin changes suggestive of palmoplantar keratoderma with transgradience. Another examination was normal. Laboratory data revealed anemia (6g/dl), leukocytosis (155,10/μL), and elevated erythrocyte sedimentation rate and C-reactive protein (57mm/h and 125 mg/L, respectively). Liver and renal function tests were normal except for hypoalbuminemia (2.1 g/dL). Lactate dehydrogenase was mildly elevated, and serum alpha-fetoprotein and beta-human chorionic gonadotropin were normal. Hydatid serology and the Mantoux test were both nonreactive.

Ultrasonography showed multiple well-defined hypoechoic lesions involving both lobes of the liver with showing central anechoic components and the largest lesion measuring 7.7 cm × 6 cm, suggesting either evolving abscess or malignant necrosis. Contrast-enhanced CT scan (CECT) abdomen showed multiloculated heterogeneous mass lesion of size 12 cm × 7 cm × 11 cm involving segments IV and V with compression of the portovenous architecture [Figure 1a]. In view of the dilemma of differentiating between abscess and tumor, magnetic resonance imaging was done, which showed well-defined multiseptated solid cystic heterogeneously enhancing lesion of 12 cm × 9.5 cm × 6.5 cm predominantly involving left lobe of liver involving segments II, III, IV, and V reaching up to the capsule [Figure 1b]. The lesion was T2 hyperintense T1 hypointense with postcontrast enhancement within the septated areas and in the periphery suggestive of undifferentiated embryonal sarcoma of the liver (UESL).

Figure 1

(a) Multiloculated heterogeneous mass lesion of size 12 cm × 7 cm × 11 cm involving segments IV and V on Contrast-enhanced computed tomography scan. (b) T2 hyperintense T1 hypointense with postcontrast enhancement within the septated areas and in the periphery suggestive of malignancy

USG-guided biopsy of the lesion showed predominantly neutrophils, necrotic material, and lymphocyte infiltration. PAS and Gram stain were negative. Ziehl–Neelsen staining showed acid-fast bacilli suggestive of TB [Figure 2a and b]. The patient was initially managed conservatively with intravenous antibiotics, nutritional supplementation, and blood transfusion. She was then started on category I ATT. The child is asymptomatic at 1.5 years with good weight gain and resolution of the lesions in USS.

Figure 2

(a) Predominantly neutrophils, necrotic material, and lymphocyte infiltration. PAS and Gram stain negative. (b) Ziehl–Neelsen staining showed acid-fast bacilli suggestive of tuberculosis

DISCUSSION

Pediatric liver masses are rare and constitute approximately 2% of all malignant tumors in children. Only a third of these primary liver lesions are benign and the rest are malignant. Most of these lesions are asymptomatic hence, making the clinical diagnosis an arduous task.

Hepatic TB can be seen in up to 70% of patients with miliary TB, however, isolated hepatic involvement is rare with an incidence of <1%. Most of the hepatic involvement is underreported due to its asymptomatic clinical presentation.[3] Low oxygen tension in the liver parenchyma is an unfavorable for the growth of the mycobacteria.[2] Hepatic spread of the bacteria is by either hematogenous from other organs or though portal venous route from the ingested bacilli in bowel and rarely by direct spread from adjacent organs such as lung or pleura. Granulomata formation is commonly seen in the periportal area (Zone 1 of Rappaport), due to the best-oxygenated blood supply. Tuberculomas may coalesce and form large areas of caseation with central necrosis as in our case. Levine classified hepatic TB as five different patterns. These included miliary TB, concomitant hepatic and pulmonary disease, primary (isolated) hepatic TB, tubercular hepatic abscess, and tubercular cholangitis.[3] Focal lesions >2mm are considered macronodular and pseudotumoral TB. Tubercular liver abscess is found in 0.3% of hepatic TB.

Clinical presentation in isolated hepatic TB, in an immunocompetent child, is nonspecific and may include symptoms such as fever, weight loss, abdominal pain, and rarely jaundice.[2] Lesions may mimic malignancies of the liver in the radiological imaging and can even end up with an unnecessary hepatic resection.[4] HCC and UESL are differentials in patients between 6–10 years like in our case. However, noncirrhotic liver and normal AFP were misleading. The high water content with abundant myxoid stroma in UESL may give low attenuation than that of soft tissue on CT scans and the high signal intensity on T2W MRI images like in our case.[5] A differential diagnosis of hemangiosarcoma and complex liver abscess was also kept in mind due to the multiseptated solid cystic components on CECT and USS. The hyperkeratotic lesions in the palms and soles of these children were misleading, and the possibility of Papillon–Lefèvre syndrome was also entertained, which may present as liver abscess.

The diagnostic criteria for hepatic TB, include liver tissue culture for Mycobacterium tuberculosis, acid-fast bacilli on smear, or nucleic acid positivity for TB. Clinical symptoms and imaging findings of low-density hepatic nodules in patient with either pulmonary TB or patient from endemic regions; and resolution of elevated liver enzymes with ATT form weak evidence criteria.

Imaging done for evaluation of pediatric liver lesions may prove to be at times misleading as in the index case. Image-guided biopsy establishes the diagnosis. Surgery may be required if the biopsy from the lesion is inadequate and fails to demonstrate the bacilli. In the past, the diagnosis was established either in the surgical specimen or in the autopsy. Difficulty thus lies in the diagnosis due to the various liver lesions mimicking each other, rather than the management which involves starting ATT.

Treatment options available include ATT along with percutaneous aspiration, pigtail drainage, intralesional infusion of isoniazid or combination of isoniazid and rifampicin, and surgical excision.[1] Most patients improve with the above management along with ATT for up to an year or until the resolution of symptoms as seen in the index case.

CONCLUSION

PTHT is one of the rare presentations of Tb liver in children. The possibility of liver TB in pediatric liver masses should be kept, especially in endemic regions. A major surgical resection may be avoided with well-timed suspicion and fine-needle aspiration cytology findings demonstrating acid-fast bacilli.

Declaration of patient consent

The authors certify that they have obtained all appropriate patient consent forms. In the form the patient(s) has/have given his/her/their consent for his/her/their images and other clinical information to be reported in the journal. The patients understand that their names and initials will not be published and due efforts will be made to conceal their identity, but anonymity cannot be guaranteed.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.