|Year : 2023 | Volume
| Issue : 1 | Page : 41-47
A critical appraisal of clinicopathological, imaging, and genexpert profiles of surgical referrals with pediatric abdominal tuberculosis
Archana Puri1, Peter Daniel S. Khairong1, Smita Singh2
1 Department of Pediatric Surgery, Lady Hardinge Medical College and Kalawati Saran Children's Hospital, New Delhi, India
2 Department of Pathology, Lady Hardinge Medical College and Kalawati Saran Children's Hospital, New Delhi, India
|Date of Submission||15-Sep-2021|
|Date of Decision||08-Jun-2022|
|Date of Acceptance||30-Jul-2022|
|Date of Web Publication||10-Jan-2023|
Tower 3, Flat 303, Common Wealth Games Village, New Delhi - 110 092
Source of Support: None, Conflict of Interest: None
| Abstract|| |
Aim: To study the clinicopathological, imaging, and GeneXpert profiles of surgical referrals with abdominal tuberculosis (TB) and to compare the utility of GeneXpert versus conventional diagnostic armamentarium.
Materials and Methods: This cohort study which was conducted over a study period of 8 years (2011–18) included seventy-seven children operated with a provisional diagnosis of abdominal TB and those who had either histological (n = 58; 75.3%) or GeneXpert (n = 9) confirmation or had miliary tubercles on exploration with supportive clinical and imaging findings (n = 17; 22.1%). GeneXpert testing was added to the diagnostic armamentarium only in the latter half of the study (2016–18, n = 31). Demographic details, symptomatology, prior antitubercular treatment, GeneXpert positivity, imaging, operative, and histological findings were recorded and analyzed using mean, standard deviation, and range for continuous variables and proportion for categorical variables.
Results: Perforation peritonitis (n = 26; 33.8%) and unrelieved obstruction (n = 51; 66.2%) were the main surgical indications. The mean age at presentation was 9.5 ± 3.6 years with a distinct female preponderance. The presence of right lower abdomen lump (n = 23; 29.9%), alternate diarrhea and constipation (n = 34; 44.1%), tubercular toxemia (n = 38; 49.4%), positive history of contact (n = 20; 25.9%), tuberculin positivity (n = 38; 49.4%), fibrocavitary pulmonary lesion (5.2%), clumped bowel loops with pulled-up cecum (n = 23; 29.9%), septated ascites (n = 17), mesenteric lymphadenopathy and omental thickening (n:18; 23.4% each) were the supportive tell-tale signs of the disease. The hallmark of pathological diagnosis was caseous necrosis with epithelioid granulomas (n = 43; 55.8%), nongranulomatous caseation (n = 15; 19.5%), and acid-fast bacilli positivity in 27.3% of patients. GeneXpert was positive in only nine patients with an overall sensitivity of 29% as compared to 75.3% for histopathology.
Conclusion: Bacteriological and histological confirmation of the disease eluded us in a significant proportion of patients, requiring a very high index of clinical suspicion to clinch the diagnosis. The current version of GeneXpert has low sensitivity in diagnosing pediatric abdominal TB.
Keywords: Abdominal tuberculosis, GeneXpert, pediatric, surgical referrals
|How to cite this article:|
Puri A, S. Khairong PD, Singh S. A critical appraisal of clinicopathological, imaging, and genexpert profiles of surgical referrals with pediatric abdominal tuberculosis. J Indian Assoc Pediatr Surg 2023;28:41-7
|How to cite this URL:|
Puri A, S. Khairong PD, Singh S. A critical appraisal of clinicopathological, imaging, and genexpert profiles of surgical referrals with pediatric abdominal tuberculosis. J Indian Assoc Pediatr Surg [serial online] 2023 [cited 2023 Feb 2];28:41-7. Available from: https://www.jiaps.com/text.asp?2023/28/1/41/367387
| Introduction|| |
Tuberculosis (TB) is a major public health problem in South Asia accounting for nearly one-fifth of the global burden of TB. Pediatric TB accounts for 10%–20% of the total disease burden. Abdominal TB in children constitutes 11.2% of all pediatric cases and poses a diagnostic dilemma due to nonspecific, atypical clinical presentation that mimics a variety of benign and malignant diseases, often leading to unwarranted delay in management., Diagnostic dilemma is further compounded by the low bacterial isolation rates in abdominal TB varying from 47% to 50%. Strong index of clinical suspicion along with supportive clinical, imaging, and pathological findings has so far been the cornerstone of diagnosis. In the last decade, however, a newer diagnostic assay like GeneXpert has shown promising results in diagnosing pediatric pulmonary TB. However, its diagnostic utility for pediatric abdominal TB is still unexplored. The present study aimed to study the clinicopathological, imaging, and GeneXpert profiles of surgical referrals of pediatric abdominal TB and to analyze their clinical utility and diagnostic pitfalls.
| Materials and Methods|| |
This was a single-center descriptive cohort study on seventy-seven children of pediatric abdominal TB, over a study period of 8 years with a mean follow-up of 5 years (range: 2–9.5 years). For data collection, the study period was divided into two time periods, namely retrospective (2011–15) and prospective (2016–18). In the latter half of the study (2016–18), GeneXpert testing was added to the diagnostic armamentarium. We included all children who were operated in our hospital with a provisional diagnosis of abdominal TB and those who had either a postoperative histological confirmation or GeneXpert confirmation or acid-fast bacilli (AFB) positivity or had classical miliary tubercles on exploration with supportive clinical and radiological findings and associated with either positive history of TB contact or had positive tuberculin testing (Mantoux skin test) with skin induration of ≥10 mm. Demographic details, duration of illness, presenting symptoms (presence of pain, vomiting, bowel habits, fever, cough, anorexia, weight loss, and moving lump in abdomen, menstrual irregularities, and bleeding per rectum) were recorded from admission records, case files, discharge summaries, and follow-up files. A note was also made for the missing data if any. A family history of TB and a previous history of antitubercular therapy were also noted. Nutritional and hemodynamic status, any evidence of peritonitis, lump in abdomen, and ascites were noted on systemic examination. Preoperative management included intravenous fluids supplementation along with broad-spectrum antibiotics and nasogastric aspiration as most of the surgical referrals were either for subacute intestinal obstruction or for perforation peritonitis. Initial baseline hematological investigations included complete blood count, blood urea, serum creatinine, protein, albumin, and electrolytes. Tubercular workup included assessment of erythrocyte sedimentation rate (ESR), tuberculin test, and plain X-ray chest. GeneXpert testing in nasogastric aspirate and sputum (if available) was done in the latter half of the study as mentioned earlier. Standard protocol as described earlier was followed for GeneXpert testing. Sample (5 ml) was collected in a Falcon tube and twice the amount of buffer solution was added (1:2). After liquefaction, it was poured into a cartridge, which was stored at 5°C–10°C. These cartridges were kept at room temperature before usage. Radiological investigations included plain X-ray abdomen and ultrasonography (USG) of the abdomen. Other radiological investigations such as contrast meal follow-through and computed tomography (CT) abdomen were done if required. Surgical indications were perforation peritonitis and obstruction that were not relieved by conservative medical management. Resected bowel specimen, biopsy from the edge of perforation, peritoneum, and enlarged mesenteric lymph node were sent for histopathology. Light microscopy sections were stained by hematoxylin and eosin and Ziehl–Neelsen (ZN) stain for AFB. Pathological diagnosis of abdominal TB was considered definitive either in presence granulomatous inflammation with or without caseous necrosis or if ZN staining of AFB was positive. Descriptive statistics were done using mean, standard deviation, and range for continuous variables and proportion for categorical variables. Inferential statistics was done using Chi-square test.
| Results|| |
During the study period, 96 children were operated in our hospital with a provisional diagnosis of abdominal TB. Seventy-seven children who had either a postoperative histological confirmation (n = 58) or GeneXpert confirmation (n = 9) or supportive clinical and radiological findings with the presence of miliary tubercles on exploration (n = 17) were included in the study. Histologically confirmed cases (n = 58), however, showed AFB positivity in only 36.2% of patients (n = 21). Nineteen patients turned out to be having an appendicular perforation and were excluded from the study. Perforation peritonitis (n = 26; 33.8%) and obstruction not relieved by conservative measures (n = 51; 66.2%) were the two main indications for surgery.
The mean age at presentation varied from 2 to 17 years with a mean of 9.5 ± 3.6 years, median of 10 years. The majority of the patients (51.6%) presented between 10 and 15 years of age although 15% of our patients were ≤2 years of age (with 3.5% being infants). There was a distinct female preponderance with a male-to-female ratio of 1:1.8. The majority of our patients (61.3%) had acute on chronic presentation with a median duration of present illness being 75 days. Pathognomonic clinical findings of surgical referrals are shown in [Table 1]. Nearly one-third of the patients had a right lower abdominal lump, and a provisional diagnosis of appendicular lump was also considered in these patients besides abdominal TB. None of our patients had perianal or perineal fistula. Occasionally, a history of bleeding per rectum (2.5%) and menstrual irregularities (8.1%) was noted in this series.
|Table 1: Clinical, laboratory, and imaging profiles of surgical referrals of pediatric abdominal tuberculosis|
Click here to view
GeneXpert, radiological, and laboratory investigations
In the latter half of the study (2016–18), GeneXpert testing was sent in nasogastric aspirate (n = 31), in sputum (n = 10), and in peritoneal fluid or ascites (n = 7). GeneXpert was positive in nine patients (nasogastric aspirate: 6, sputum: 0, peritoneal or ascitic fluid: 4; one patient had GeneXpert positive both in nasogastric and purulent peritoneal fluid). In this series, none of the patients had multidrug resistance to rifampicin on GeneXpert. Imaging and laboratory findings of surgical referrals are depicted in [Table 1]. Although mesenteric lymph nodes were enlarged in eighteen patients, only seven patients (9.1%) had significant enlargement of more than 15 mm. In 23 patients who presented with right lower abdominal lump, the diagnosis of abdominal TB was further confirmed by contrast-enhanced CT whole abdomen. While thickened matted bowel loops with pulled-up cecum was noted in all of them, hepatosplenomegaly (n = 4), loculated pus collection with air pockets (n = 3), pleural effusion with abdominal and mediastinal lymphadenopathy was seen in four patient. Contrast meal follow-through was done in 22 patients with a history of moving lump in the abdomen and it revealed terminal ileum narrowing with proximal bowel dilatation in 19 and deformed or pulled-up cecum in 3 patients. Radiological findings are depicted in [Figure 1] and [Figure 2].
|Figure 1: Radiological findings in abdominal TB. (a) Plain X-ray chest and upper abdomen (erect) showing hilar lymphadenopathy (arrow) and gas under diaphragm (arrow) in a child with tubercular bowel perforation; contrast meal follow-through in children presenting with gurgling and moving lump in abdomen; (b) pathognomonic pulled up cecum (arrow) with distal ileal stricture (passable); (c) matted bowel loops; (d) dilated bowel loop (asterix) with ileal stricture. TB: Tuberculosis|
Click here to view
|Figure 2: Computed tomography abdomen (a) showing loculated collection (LC) with debris (asterix), air foci, and dilated thickened bowel loops in a child presenting with lump in the left lower abdomen with bowel perforation opening into the cavity of LC; (b) loculated collection with air foci (arrow) with dense adhesions and matted bowel loops in a child presenting with right iliac fossa lump with perforation of ileocecal region opening into the loculated cavity; (c) ascites with thickened dilated bowel loops|
Click here to view
Combined gastrointestinal and nodal disease was noted in 49 patients (63.6%), while 13 (16.9%) patients had peritoneal, gastrointestinal, and nodal involvement. Adjacent bowel loops, mesentery, and nodes adhered to form a mass termed as “intestinal cocoon” in 15 patients (19.5%), where the exact site of involvement could not be delineated [Figure 3]. Miliary tubercles over bowel, peritoneum, and omentum were present in 17 patients (22.1%). Ileum was the most common site of intestinal involvement, either alone (n = 24; 31.2%) or in combination with ileocecal junction (n = 24; 31.2%) or jejunum (n = 12; 15.6%). Two of our patients (2.6%) presented with gastric outlet obstruction and had gastroduodenal involvement. Pyoperitoneum was present in all patients with perforation peritonitis (n = 26). Ten patients had single-bowel perforation, while sixteen had multiple perforations. Of the 26 patients with perforation peritonitis, the majority had resection of the diseased segment with stoma formation (n = 20), one patient with single midileal perforation had resection of the diseased segment with end–end anastomosis, while five patients with old perforation with dense adhesions and matted bowel loops had laparotomy, drainage, and biopsy. Bowel strictures were noted in 34 patients, of whom two had perforation proximal to the stricture with associated ascariasis infection. Depending on the condition of the diseased bowel, those with single stricture (n = 13) had either stricturoplasty (n = 8) or resection of the diseased segment with end–end anastomosis. While those who had multiple bowel strictures present over a short segment had resection and anastomosis (n = 19). Two patients with multiple ileal strictures associated with proximal perforation had resection of the diseased bowel with stoma formation. In patients with abdominal cocoon or with dense adhesions, we resorted to laparotomy and either omental or peritoneal biopsy (n = 15). Thus, resection and end-to-end anastomosis (n = 25; 32.5%) was the most common operative procedure performed, followed by resection of the diseased segment and stoma formation (n = 22; 28.6%), laparotomy, biopsy, and drainage (n = 20; 25.9%), stricturoplasty (n = 8), and pyloroplasty (n = 2).
|Figure 3: Operative photographs in children with abdominal TB. (a) Pathognomonic miliary tubercles on omentum; (b) multiple miliary tubercles on small bowel loops with mesenteric lymphadenopathy; (c) fibroadhesive variety (abdominal cocoon); (d) napkin ring multiple ileal strictures (arrow) with fat stranding and thickened bowel loops. TB: Tuberculosis|
Click here to view
Postoperative course, complications, and histopathology
Postoperative management was done as per the standard protocol. All patients received four-drug antitubercular therapy for 3 months (isoniazid (10 mg/kg), rifampicin (15 mg/kg), ethambutol (20 mg/kg), and pyrazinamide (35 mg/kg)), followed by isoniazid and rifampicin for six more months. The most common adverse outcome was wound infection and fecal fistula formation in eleven (14.3%), followed by mortality in seven (9.1%) and chest infection in five patients (6.5%). Resected bowel specimens (n = 47), biopsy from the edges of strictures (n = 8), peritoneum and omentum (n = 20), and mesenteric lymph nodes (n = 33) were sent for histopathology. Gross hypertrophy of the bowel specimens was seen in all resected bowel specimens. Histology slides of 108 specimens from 77 patients were available for review. In nine patients, the histopathology specimens were considered inadequate for opinion, leaving 68 patients with valid histology reports. On light microscopy examination, pathognomonic caseous necrosis with epithelioid granulomas was noted in 43 patients (55.8%) and nongranulomatous caseation in 15 patients (19.5%), while nonspecific inflammatory findings like sinus histiocytosis and follicular hyperplasia (n = 5; 6.5%) and fibrocollagenous tissue and congested blood vessels (n = 5; 6.5%) were also noted. Caseating epithelioid granulomas were usually located either in serosa or lamina propria of resected bowel specimens or in mesenteric lymph nodes. AFB positivity with ZN staining was noted in 21 patients (27.3%), mostly in lymph node samples (n = 18), followed by peritoneum (n = 5) and omentum and bowel (n = 10). All of them with positive ZN staining had pathognomonic caseous necrosis with epithelioid granulomas on histology. Histopathology was considered diagnostic of abdominal TB (n = 58, 75.3%), showed nonspecific inflammatory findings (n = 10), and was inconclusive in nine patients. The diagnosis of abdominal TB in those with equivocal or inconclusive histology (n = 19) was confirmed by GeneXpert testing (n = 2) and supportive clinical and radiological findings with the presence of miliary tubercles on exploration (n = 17).
Correlation of GeneXpert and histopathology
Based on histopathology and GeneXpert status, patients were divided in two groups, namely those with GeneXpert positive and negative as shown in [Table 2]. The overall sensitivity of GeneXpert in diagnosing abdominal TB was 29% as compared to 75.3% for histopathology.
|Table 2: Distribution of patients according to GeneXpert status and histopathology|
Click here to view
| Discussion|| |
Abdominal TB denotes involvement of the gastrointestinal tract, peritoneum, lymph nodes, and solid viscera, namely liver, spleen, and pancreas. Abdominal TB poses a unique management challenge due to its atypical, nonspecific clinical presentation, and paucibacillary nature of the disease, leading to poor mycobacterial diagnostic yields. Diagnosis of abdominal TB is considered definitive if Mycobacterium TB could be isolated on culture or demonstrated by polymerase chain reaction or by ZN staining of AFB (microbiological case definition). However, such bacterial isolation was possible in only 27.3% of our patients, much lower than 47% incidence reported in literature. Histopathology confirmed the diagnosis in nearly three-fourths of the patients (pathological case definition; n = 58; 75.3% by presence of epithelioid granuloma with or without caseation and giant cells), much higher than 19% reported earlier by a medical series. In the absence of bacteriological and histological confirmation, the diagnosis of abdominal TB was confirmed by GeneXpert (n = 2, 6.5%) and by the presence of strongly supportive imaging, clinical, and operative findings with pathognomonic miliary tubercles (n = 17; 22.1%).
One needs to have a very high index of clinical suspicion, especially in endemic areas to clinch an early diagnosis. The presence of vague colicky abdominal pain associated with bilious vomiting, gurgling, alteration of bowel habits, positive history of tubercular contact, and an ill-defined lower abdominal lump are some of the tell-tale signs of this disease. Thickened clumped bowel loops, pulled-up cecum, septate ascites with internal echoes, mesenteric lymphadenopathy with rim enhancement, and omental thickening are some of the supportive radiological clues. Right lower abdominal lumps, seen in nearly one-third of our patients, were usually composed of matted, thickened bowel loops, caseating mesenteric lymph nodes and thickened omentum. Features suggestive of TB toxemia like low-grade fever with evening rise of temperature, significant weight loss, and anorexia were additional nonspecific tell-tale signs of this disease. Unlike adults, signs of tubercular toxemia occur more frequently (66%–75%) in pediatric abdominal TB as was also noted in this series.,, Positive contact history was not forthcoming in majority of our patients due to illiteracy, ignorance, and social stigma attached to the disease.,,, Other supportive diagnostic evidence like tuberculin positivity of ≥10 mm and elevated ESR of more than 40 mm in the 1st h was observed in nearly half of our patients, similar to that reported in literature.,,,, Nearly one-third of our patients had empirically received antitubercular treatment (ATT) prior to presentation. Acute presentation in abdominal TB occurs rarely in 20%–40% of patients.,, However, being a surgical series, most of our patients had acute on chronic presentation. Of the 96 cases clinically suspected to be suffering from abdominal TB, prior to laparotomy or histology, 77 were confirmed to be having the disease. Thus, the diagnostic accuracy of preoperative clinical diagnosis in this series was 80.2%. This is much higher than 39% accuracy of clinical diagnosis reported in previous studies. Abdominal TB needs to be differentiated from Crohn's disease as they both present with lower abdominal pain. Absence of persistent diarrhea, hematochezia, perianal and internal fistulae, and extra-intestinal manifestations like uveitis, arthritis, and vasculitis helps in differentiating Crohn's disease from abdominal TB.
Active pulmonary lesion in abdominal TB varies from 6% to 90%. However, in nearly 30%–50% of patients with pediatric abdominal TB, chest X-ray may be reported as normal, as also noted in this study., In this series, although hilar prominence was present in majority (55.8%), fibrocavitary lesion was seen in only four patients (5.2%). While USG accurately picked up ascites and nodal disease, it was not the modality of choice for bowel lesions., Those presenting with ill-defined lower abdominal lumps had CT scan. The presence of irregular soft tissue densities in omentum, ascites with internal echoes, low-density nodal masses with rim enhancement, and presence of poorly defined masses of bowel loops, omentum, nodes, and entrapped fluid were highly diagnostic imaging findings of abdominal TB.,, Characteristic granulomas were often seen just beneath the ulcer bed, mainly in the submucosal layer and in mesenteric lymph nodes. We had diagnostic dilemma in 12.9% of patients who had nonspecific round cell infiltration in the bowel wall and mesenteric lymph nodes. This may be caused due to prior empirical treatment of ATT.
According to a recent estimate, only 15%–20% of children suffering from abdominal TB would warrant surgery, as majority would respond to conservative measures., Acute perforation of tuberculous ulcer is uncommon due to reactive thickening of peritoneum with subsequent adhesions formation, and the reported incidence of free perforations in abdominal TB varies from 1% to 10%., The diagnosis of tuberculous perforations is often delayed, due to absence of generalized peritonitis and pneumoperitoneum, in a significant proportion of patients. In our study, although nearly one-third of children (n = 26) had bowel perforations, gas under diaphragm was seen in only eight children (10.4%). Dense peritoneal and interbowel adhesions probably prevented the ascent of gas in the rest. Primary closure of tubercular perforation is usually not advocated due to high chances of subsequent fecal fistula formation and was not attempted in any of our patients. Unlike the earlier published reports recommending resection and end–end anastomosis as the preferred treatment for tuberculous bowel perforation, majority of our patients had resection of the diseased segment with stoma formation in. We would recommend stoma formation in a moribund, hemodynamic unstable patient with gross fecal contamination. Thus, nearly one-third of our patients had stoma formation, which helped us immensely in reducing our mortality to <10%. However, in those presenting with subacute obstruction and bowel strictures either resection of the diseased segment with end–end anastomosis or stricturoplasty was performed. Sometimes, in the presence of dense adhesions and abdominal cocoon, it is advisable to resort to simple lavage, drainage, and biopsy rather than extensive dissection which often increases the chances of serosal tear, fecal fistulae formation, and postoperative morbidity and mortality.
GeneXpert is a novel automated, semiquantitative nucleic acid amplification test. It not only detects the DNA of mycobacterium TB but also assesses rpo8 mutation associated with rifampicin resistance., It was once considered a “Game changer” in the management of TB. Although GeneXpert test is recommended for respiratory samples, cerebrospinal fluid, and lymph node, due to lack of evidence, it is sparingly advocated for urine, stool, and blood samples., A recent meta-analysis showed that GeneXpert assay provides an excellent sensitivity of 95% in smear-positive children for both gastric lavage (aspirate) and expectorated or induced sputum and an acceptable sensitivity of 62% in smear-negative culture-positive children., Specificity for all respiratory specimens varied from 70% to 100%. An assessment of GeneXpert in EPTB revealed that detection rates were highest in TB lymphadenitis (33.3%), followed by pleural effusion (11.9%) and least in peritoneal fluid and CSF (6.7% and 5.4%), respectively., Concerns have been raised about the low sensitivity and negative predictive value of GeneXpert in EPTB., This series is unique as it compared the sensitivity of GeneXpert in abdominal TB in nasogastric aspirate, sputum, and pus or peritoneal fluid. We compared its efficacy with the existing established gold standard of diagnosis, namely histopathology, unlike the previous reported reference standards of smear microscopy, cultures, clinical, and radiological findings., GeneXpert positivity in pediatric abdominal TB varied in nasogastric aspirate (19.4%), sputum (0%), and peritoneal or ascitic fluid (57%). The sensitivity of GeneXpert was much lower than histopathology in pediatric abdominal TB (35% vs. 75.3%). This low diagnostic yield can be attributed to paucibacillary nature of pediatric abdominal TB with tendency of Mycobacterium TB to form clumps, leading to uneven distribution of bacilli. The use of GeneXpert is further limited by its vulnerability to heat and dust and it cannot differentiate between live and dead bacilli. Any lapses in storage and faulty collection as occurred in two of our patients may lead to false negative results. Prior intake of ATT especially for more than 60 days, as seen in 24.9% of patients in this series, also leads to lower positivity rate for GeneXpert. We noted a comparatively better diagnostic yield in pus and would advocate its sampling whenever possible. Enlarged lymph nodes reportedly have high mycobacterial load and should be sampled to improve the diagnostic accuracy of GeneXpert. Diagnostic yield of GeneXpert is reported to be enhanced using the newer advanced version, increasing sample volume (5–15 ml), and by concentrating the samples and using sediments.
| Conclusion|| |
To conclude, despite being a surgical series, the bacteriological and histological confirmation of the disease eluded us in a significant proportion of patients. Thus, suggestive clinical and imaging findings as described herein are needed to clinch an early diagnosis. The current version of GeneXpert has low sensitivity in diagnosing pediatric abdominal TB. Faulty sample collection, lapses in storage of thermolabile cartridges, and prior empirical intake of ATT contributed to a significant false negative rate in this series. Thus, the search for an ideal diagnostic tool for pediatric abdominal TB continues.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Vadwai V, Boehme C, Nabeta P, Shetty A, Alland D, Rodrigues C. Xpert MTB/RIF: A new pillar in diagnosis of extrapulmonary tuberculosis?. J Clin Microbiol 2011;49:2540-5.
Shah I, Uppuluri R. Clinical profile of abdominal tuberculosis in children. Indian J Med Sci 2010;64:204-9. [Full text]
Malik R, Srivastava A, Yachha SK, Poddar U, Lal R. Childhood abdominal tuberculosis: Disease patterns, diagnosis, and drug resistance. Indian J Gastroenterol 2015;34:418-25.
Maynard-Smith L, Larke N, Peters JA, Lawn SD. Diagnostic accuracy of the Xpert MTB/RIF assay for extrapulmonary and pulmonary tuberculosis when testing non-respiratory samples: A systematic review. BMC Infect Dis 2014;14:709.
Indumathi CK, Sethuraman A, Jain S, Krishnamurthy S. Revised antituberculosis drug doses and hepatotoxicity in HIV negative children. Indian J Pediatr 2019;86:229-32.
Kapoor VK. Abdominal tuberculosis. Postgrad Med J 1998;74:459-67.
Thapa BR, Yachha SK, Mehta S. Abdominal tuberculosis. Indian Pediatr 1991;28:1093-100.
Talwar BS, Talwar R, Chowdhary B, Prasad P. Abdominal tuberculosis in children: An Indian experience. J Trop Pediatr 2000;46:368-70.
Davies MR. Abdominal tuberculosis in children. S Afr J Surg 1982;20:7-19.
Talwar S, Talwar R, Prasad P. Tuberculous perforations of the small intestine. Int J Clin Pract 1999;53:514-8.
Bhansali SK. Abdominal tuberculosis. Experiences with 300 cases. Am J Gastroenterol 1977;67:324-37.
Das P, Kumar P, Gupta CK, Indrayan A. Clinical patterns of abdominal tuberculosis. Am J Proctol 1975;26:75-86.
Gurzu S, Molnar C, Contac AO, Fetyko A, Jung I. Tuberculosis terminal ileitis: A forgotten entity mimicking Crohn's disease. World J Clin Cases 2016;4:273-80.
Sheikh M, Abu-Zidan F, al-Hilaly M, Behbehani A. Abdominal tuberculosis: Comparison of sonography and computed tomography. J Clin Ultrasound 1995;23:413-7.
Engin G, Acunaş B, Acunaş G, Tunaci M. Imaging of extrapulmonary tuberculosis. Radiographics 2000;20:471-88.
Abro A, Siddiqui FG, Akhtar S, Memon AS. Spectrum of clinical presentation and surgical management of intestinal tuberculosis at tertiary care hospital. J Ayub Med Coll Abbottabad 2010;22:96-9.
Vasantha VC, Habibullah CM, Kumar A. Abdominal tuberculosis with special reference to pathological aspects. J Indian Med Assoc 1975;65:302-3.
Sarfaraz S, Iftikhar S, Memon Y, Zahir N, Hereker FF, Salahuddin N. Histopathological and microbiological findings and diagnostic performance of genexpert in clinically suspected tuberculous lymphadenitis. Int J Infect Dis 2018;76:73-81.
Walzl G, McNerney R, du Plessis N, Bates M, McHugh TD, Chegou NN, et al.
Tuberculosis: Advances and challenges in development of new diagnostics and biomarkers. Lancet Infect Dis 2018;18:e199-210.
Metaferia Y, Seid A, Fenta GM, Gebretsadik D. Assessment of extrapulmonary tuberculosis using GeneXpert MTB/RIF assay and fluorescent microscopy and Its risk factors at dessie referral hospital, Northeast Ethiopia. Biomed Res Int 2018;2018:8207098.
[Figure 1], [Figure 2], [Figure 3]
[Table 1], [Table 2]