Br. J. Surg. Vol. 63 (1976) 941-946
Clinical diagnosis of abdominal tuberculosis P R I T A M D A S A N D H. S . S H U K L A * Clinically, the patients were divided into two main One hundred and eighty-two cases of abdominal tuber- groups: those who presented with intestinal obstrucculosis admitted to Swaroop Rani Nehri ( S R N ) Hospi- tion (93 cases) and those who did not have intestinal tal, Allahabad, in the past 7 years have been reviewed. obstruction (89 cases). The 20 ascitic cases are The clinical diagnosis of abdominal tuberculosis was included in the non-obstructive group, although they made correctly only in 50 per cent of cases. About half had a distinctive clinical picture with fluid distension the cases presented with chronic or acute on chronic of the abdomen. intestinal Obstruction. The remaining patients had tiague pains, tender abdomen, constitutional symptoms or a mass in the abdomen. Diarrhoea was not frequent and fistula formation was rare. A chronic obstructive type loo of lesion was found not only in cases with a bowel lesion 80 but also in patients with chronic miliary peritonitis and tuberculous mesenteric adenitis. Similarly, a lump was present not only in hypertrophic bowel tuberculosis but 9 60 also in chronic miliary peritonitis and tuberculous J mesenteric adenitis. On radiological examination false r positive features such as fluid levels, bowel dilatation or 3 40 even the ‘string’ sign were encountered. Liver and endometrial biopsies were positive in only a very few cases. Peritoneal biopsy was of considerable help, 20 being positive in 88 per cent of ascitic cases and in 42.1 per cent of non-ascitic cases. Open peritoneal biopsy obtained after making a small incision in the right iliac fossa was found to be the most Age useful investigation in the diagnosis of abdominal tuberculosis. A n ascitic fluid protein content of 2.5g or more Fig. 1. Age incidence of abdominal tuberculosis nnd a predominantly lymphocytic count ojover 100/mm3 are diagnostic, but a cell count of 10/mm3was recorded Table I: INCIDENCE OF VARIOUS TYPES OF LESIONS in one tuberculous case. No. of Type of lesion cases % ABDOMINAL tuberculosis continues to be a common Stricture of ileum, jejunum or both 27.5 50 disease responsible for considerable morbidity and Hyperplastic ileocaecal lesion with or 20.8 38 without small bowel stricture mortality in India. However, even where it is relatively 5 2.8 common, the diagnosis of abdominal tuberculosis is Colonic stricture Mesenteric lymphadenitis 21 11.6 often obscure. The symptoms and signs are often quite Chronic miliary tuberculous peritonitis 48 26.4 vague and the laboratory investigations and radio- Tuberculous ascites 10.9 20 logical findings unhelpful. The present study has been Total 182 undertaken to assess the value of clinical features and laboratory and radiological investigations in proved Clinical features cases of abdominal tuberculosis. Accuracy of diagnosis Patients and methods The accuracy of clinical diagnosis was 50 per cent. The records of 182 cases of abdominal tuberculosis Other cases were provisionally diagnosed as chronic admitted to SRN Hospital, Allahabad, during the 7 or acute intestinal obstruction, abdominal pain of years 1967-73 were studied. The various types of uncertain aetiology, recurrent appendicitis, acute lesions found are shown in Table I. The cases were appendicitis, appendicular mass, pyloric stenosis, classified according to the dominant feature visible to peptic ulcer, renal mass, retroperitoneal sarcoma, the naked eye at operation. Mesenteric adenitis and faecal fistula, cirrhosis, malignant ascites and malabperitoneal tubercles were present in a number of sorption syndrome. cases along with the bowel lesion. However, as the main feature was the intestinal pathology, such cases * Department of Surgery, MLN Medical College, Allahabad, India. have been classified according to the type and site of Present address of H. S. Shukla: Surgical Unit, University the gut Iesion. Hospital of Wales, Cardiff. SUMMARY
1
-
941
Pritam Das and H. S. Shukla Table 11: INCIDENCE OF SYMPTOMS Stricture of ~ileum Symptom
0 33 31 19 4 7
NO
Hyperplastic i leocaeca I lesion ~ 0 23 18 13 5 1
NO 14
Pain 17 Vomiting 10 9 Constipation 3 5 Diarrhoea 1 0 Diarrhoea alternating 1 2 with constipation Weight loss I 5 2 3 2 Anorexia 1 9 3 5 4 Cough with expectoration 4 1 2 0 Moving lump inabdomen 18 7 7 0 Borborygmi 1 8 0 9 2 Post-prandial distress 13 2 9 3 Distension of abdomen 12 2 17 2 Menstrual disorders 1 1 4 6 3 Fever 8 2 6 8 0, Obstructive; NO, Non-obstructive. * Percentage of female patients. Table 111: SITE OF PAIN IN THE ABDOMEN Cases Cases with without obstruc- obstruc- No. of Site of pain tion tion cases
Stricture of colon ~ 0 NO
Mesenteric lymphadenitis 0 NO
Tubercular peritonitis _ 0 NO
0
Ascitic_ NO
_ Total
7;
94.0 69.6 46.7
2 2 2 0 0
3 2 1 0 0
8 8 5 2 0
9 7 3 1 0
25 22 2310 2 4 3 1 3 0 3
0 0 0 0 0
14 5 7 3 2
170 125 20 16
11.1
2 2 0 2 2 2 2 0 1
2
5 7
4 3
6 12 9 12 1 4 3 5 5 2 3 1 0 2 0 2 3 1 1 1 0 1 6
0 I0 0 15 0 3 0 0 0 2 0 2 0 1 4 0 7 0 1 6
63
35.0 444 8.8 28.8 25.5 27.2 45.0 35.6* 42.2
1 0 3 1 2 0 1 2
1
0
4 4 1 7 0 2
2 1 2 3 1 5
85
80
16 51 46 49 81 47 76
8.8
Symptoms Abdominal pain was the commonest symptom and was present in 170 cases (94 per cent) (Table ZZ). The site of the pain is given in Table IIZ. % Analysis of the character of the pain revealed the Umbilical 35 2s 60 35.4 following : Generalized 29 26 55 32.3 1 . As might be expected, the pain was usually Right iliac fossa 17 18 35 20.6 Epigastrium 5 6 11 6.5 colicky in the obstructive group (89 per cent). In the Left hypochondrium 3 1 4 2.4 remaining cases in this group the pain was vague in 6, Right hypochondrium 1 2 3 1.8 gripping in 2 and like that of peptic ulcer in 2 cases. Left iliac fossa 1 0 1 0.5 2. In the non-obstructive type of cases colicky pain Right lumbar region 0 1 1 0.5 was present in 37 cases out of 89, while the pain was Total 91 79 170 vague in 29 and gripping in character in 21 cases. In 2 cases of mesenteric lymphadenitis the pain waslikethat of peptic ulcer. In the ascitic group the character of Age and sex The age range is shown in Fig. 1. The predominant the pain was vague in the majority of cases. Vomiting was present in 88 per cent of cases of the age group was 21-30 years (47 per cent of cases). The next most frequent period was between 11 and 20 obstructive group. However, it was also present in years (28 per cent). After 30 years there is a sudden 48 per cent of cases without intestinal obstruction. drop in the incidence, which diminishes further with Patients with ascites had the lowest incidence of vomiting (25 per cent). increasing age. Constipation was more common in the obstructive There were 132 females (72 per cent) and 50 males (28 per cent). While the overall male to female group (67.7 per cent) than in the non-obstructive group ratio was 1 : 2.6, for the different subgroups it was (24.8 per cent). A ‘moving lump’ in the abdomen was 1 :2.8 in cases with stricture of the ileum, 1 : 8.5 in twice as common in the obstructive group (36 per cent hyperplastic ileocaecal lesions, 1 : 1.3 in mesenteric compared with 19 per cent in the non-obstructive lymphadenitis, 1 : 1.5 in chronic miliary tubercular group). Fever was present in 42.2 per cent of the cases. The peritonitis and 1 : 3 in ascitic cases. incidence of fever was greater in the non-obstructive group (49 out of 89) than in the obstructive group (27 Associated tuberculosis Associated pulmonary tuberculosis was present in 13 out of 93). It is therefore important to emphasize that (15 per cent) out of the 86 cases who had a radio- quite a large number of obstructive cases do not logical examination of the chest. In 6 other patients suffer from pyrexia. The incidence of menstrual disorders was 35.6 per (6.9 per cent) pleural effusion was present. These figures probably underestimate the incidence, because cent of the female patients. only patients for whom there was a suspicion of a pulmonary lesion were X-rayed. A family history of Duration of symptoms tuberculosis was forthcoming in 4 cases only. There The duration of symptoms before attending hospital was one case with associated glandular tuberculosis varied considerably. The largest number of patients but no case with skeletal tuberculosis. (30.4 per cent) had had symptoms for 1-6 months,
942
-
Diagnosis of abdominal tuberculosis Table 1V: INCIDENCE OF PHYSICAL SIGNS HyperStricture plastic of ileocaecal ileum lesion 0 NO Sign 0 NO 15 5 Poorly nourished 20 2 Pallor 17 4 25 6 Cervical adenitis 0 1 0 0 21 8 Tender abdomen 22 I Distension of abdomen 28 6 17 1 1 0 Doughy feel to abdomen 3 1 Visible peristalsis 26 2 12 4 1 1 0 1 Ascites Palpable liver 2 6 2 0 6 1 1 0 Rieiditv Lump in abdomen 5 4 1 1 5 0, Obstructive; NO, non-obstructive.
Stricture of colon 0 NO 2 2 1 0 1 0 2 0 2 0 2
while 24.5 per cent of cases had had symptoms for 6 months to 1 year. A short history (up to 1 month) was recorded for 20,5 per cent, and a history of 1-3 years in 12 per cent, 3-5 years in 11 per cent and over 5 years in 1.6 per cent of cases. There was little difference between the obstructive and non-obstructive groups.
Physical signs Tenderness and distension of the abdomen were the most frequent physical signs and were present in 65.9 and 58.2 per cent of cases respectively (Table ZV). An abdominal mass was found in 28.6 per cent of cases. The mass most frequently occurred in the right iliac fossa (18.5 per cent). The other sites where a mass was found were the umbilical region in 4.2 per cent, the epigastrium in 3.2 per cent and the right hypochondrium in 1.6 per cent of cases. A vague mass was present in 1.1 per cent of cases. No mass was palpable in the 20 cases of tuberculous ascites. A mass was found in 42.1 per cent of the hyperplastic ileocaecal tuberculosis cases, in 43 per cent of the mesenteric lymphadenitis cases, in 31.2 per cent of the tuberculous peritonitis cases, in 60 per cent of the cases with stricture of the colon and in 18 per cent of the cases with stricture of the small bowel. There was no significant difference in the occurrence of a mass in obstructive and non-obstructive groups. It is evident that an ileocaecal lesion was not the only type of case in which a mass could be present, and it even occurred in cases with tuberculous peritonitis. Visible peristalsis was present in 35.1 per cent of cases. In none of the ascitic cases was peristalsis seen, and its incidence was significantly less in nonobstructive cases (10.1 per cent) compared with obstructive cases (59 per cent). Pallor was noticed in 56.5 per cent of cases, and 45.6per cent of the patients appeared poorlynourished. Ascites was found in all ascitic cases but rarely in the other types. It should be noted that a ‘doughy feel to the abdomen’, which is often quoted as typical of a tuberculous abdomen, was present in only 6.0 per cent of cases. A pus-discharging sinus at the umbilicus was found in 2 cases. The liver was palpable in 14.2 per cent of cases.
1 1 1 1 0 0 1 1
0 0 1
Mesenteric lymphadenitis 0 3 I 0 4 8 0 I
NO 3 3 0 5 1 0 2
0
0
0 0 1
3 0 8
Tubercular peritonitis
Ascitic - -
0 NO 1 7 8 19 8 0
0
18 I5 22 8 1 3 8 0 5 4 1 5 6 2 4 1 1
0 0
NO
Total 83 103
0
9 11 0 9 14 2
0
0
0 0
0 0
0 2 0 0 4 0 0 0 0
3 110
106 11
64 33 25 16 52
% 45.6 56.5 1.6 65.9 58.2 6.0 35.1 18.6 14.2 8.7 28.6
Investigations Haematological findings The haemoglobin level was estimated in 165 cases. Eighteen per cent had haemoglobin levels above 12 g/dl, 60 per cent between 9 and 12 g/dl and 22 per cent below 9 g/dl. In the non-obstructive group the haemoglobin level was generally on the higher side. The ESR was found to be raised in 117 out of 126 cases examined (92.9 per cent). In 9 cases the ESR was normal although the disease was in the active stage. The total leucocyte count was normal in over half of the cases (55.5 per cent). In 43.8 per cent it was raised. There was no significant group variation. One case with ascites had leucopenia. The polymorphs were raised in 42.9 per cent, lymphocytes in 10.7 per cent and eosinophils in 7.9 per cent of cases. In 14.7 per cent of cases a low lymphocyte count was recorded. Chest X-ray Chest X-rays were taken in 86 cases. Thirteen (15.1 per cent) showed infiltration, 6 (6.9 per cent) showed pleural effusion and 5 (5.8 per cent) showed a healed tubercular focus. Thus, in 24 out of 86 cases evidence of pulmonary infection was present. The highest incidence (12 out of 24) of pulmonary involvement was noted in cases with tuberculous peritonitis and ascites. This is significant considering the aetiology of tuberculous peritoneal involvement which is believed to be haematogenous. The small bowel stricture group exhibited the next highest incidence of active or healed pulmonary tuberculosis (6 cases). Four cases with mesenteric adenitis and 2 with hyperplastic ileocaecal tuberculosis also had active or healed pulmonary tuberculosis. Plain X-ray of the abdomen A plain X-ray of the abdomen was taken in 54 cases. As expected, multiple fluid levels and gas shadows were present in 14 out of 29 cases of the obstructive group. In only one out of 25 cases belonging to the non-obstructive group was there demonstration of gas shadows and fluid levels. The majority of the nonobstructive cases had no positive findings. Calcified 943
Pritam Das and H. S. Shukla mesenteric lymph nodes were observed in 2 of the ascitic group and one of the colon stricture group. Gas under the diaphragm was found in one case belonging to the non-obstructive group and in another of the obstructive group. Barium meal examination Barium meal examination was done in 38 cases. In 15 cases the results were normal. In 34.2 per cent of cases areas of small bowel obstruction and dilatation were seen. An unusual finding was the presence of pyloric obstruction in one case and dilatation of tl-e first, second and third parts of the duodenum in 3 cases. A subsequent laparotomy revealed enlarged caseous mesenteric lymph nodes pressing on the pylorus and the third part of the duodenum respectively. Other positive findings were a ‘pulled-up’ caecum in one case and a filling defect in the caecum in 5 cases. Dilated bowel loops, narrowing of the terminal ileum, a filling defect in the caecum and even the ‘string’ sign were observed in some cases in which n o bomel lesion was found at exploratory laparotomy. Barium enema examination Five out of the 8 cases in whom this investigation was done showed a filling defect in the caecum, one case showed a stricture of the transverse colon and in the remaining 2 the bowel was normal. Intravenous pyelography Intravenous pyelography showed kinking of the left ureter in one case, which proved to be due to tubercular retroperitoneal fibrosis. Peritoneal biopsy Slides of 71 peritoneal biopsies were available for re-examination. Of these, 59 biopsies were done with an Abrahm’s needle, and 12 by making a small incision in the right iliac fossa, exposing the peritoneum and removing a small piece. In the ascitic and chronic miliary peritonitis groups, the biopsy material showed tubercular histopathology in 46 out of 52 cases. In the remaining 6 cases a non-specific picture was present, although these cases had other positive tutercular evidence. In cases of small bowel stricture, hyperplastic caecal lesions, colon stricture and mesenteric adenitis, 8 out of 19 biopsies revealed normal peritor,eum, 8 showed tubercular changes and 3 showed non-specific changes. Of the 12 cases (7 ascitic and 5 non-ascitic) having open peritoneal biopsy, tubercular histopathology was found in 1 1 . Liver biopsy Liver biopsy was done in 46 cases. Fifty per cent showed normal histopathology, 32.6 per cent showed fatty changes, 10 per cent revealed reactive hepatitis and 5 per cent showed caseation necrosis. In one case liver biopsy provided the diagnosis of abdominal tuberculosis. Two cases of hepatic tuberculosis occurred in the peritonitis group. Reactive hepatitis and fatty changes were predominantly seen in the cases with chronic miliary peritonitis. 944
Endometrial biopsy In ;i out of 21 endometrial biopsies performed tubQcular pathology was present. There were 4 failures to obtain endometrial tissue. It is significant to note that 4 out of the 5 cases with tubercular endometrium occurred in the miliary peritonitis and ascitic groups. The fifth case belonged to the tuberculous mesenteric lymphadenitis group. This could indicate that the endometrial lesion occurred first and later spread to the peritoneum. However, the possibility that the endometrium and peritoneum were both involved by blood-borne infection cannot be excluded. Ascitic fluid examination In 20 cases ascitic fluid was examined for cells and protein content. The cell count varied from 100 to 250 cells/mm3 in 8 cases. In 12 cases the ascitic fluid contained blood and therefore a cell count could not be done. The cells were predominantly lymphocytes. The protein level varied from 2.5 to 6.1 g per cent.
Discussion A correct clinical diagnosis of abdominal tuberculosis was made in 50 per cent of cases. Hoon et al. (1950) reported 34 per cent accuracy. The age incidence of the present series is similar to that reported by other workers ( A d a m and Miller, 1946; Bobrow and Friedman, 1956; Dutta Gupta, 1958; Paustian, 1966; Prakash et al., 1970; Sharma et al., 1972). The overall sex incidence showed a preponderance of the disease in females (72 per cent). However, the incidence varied according to the type of disease, with a female to male ratio of 8.5 : 1 in the hyperplastic ileocaecal tuberculosis group compared with 1.5 : 1 in the chronic miliary peritonitis group. Associated pulmonary tuberculosis was found in 15 per cent of cases and, considering the haematogenous route of infection in the aetiology of tuberculous peritonitis, it was interesting to note that half of the cases in this group had pulmonary tuberculosis. A much higher incidence of associated pulmonary tuberculosis as a whole has been reported by other workers (Hoon et al., 1950; Bobrow and Friedman, 1956; Singhai et al., 1964; Prakash et al., 1970). Paustian (1966) and Bhansali and Desai (1968) considered acute or chronic intestinal obstruction to be the commonest presentation of abdominal tuberculosis. While 50 per cent of our cases presented with intestinal obstruction, it was significant that the symptoms of pain, vomiting and constipation were present in some of the cases of the non-obstructive group as well. Therefore, the diagnosis of ‘obstructive lesion’ cannot be made with certainty on clinical grounds alone. Anorexia, loss of weight and a feeling of distension or of a mass in the abdomen were the other symptoms. Diarrhoea was present only in 1 1 . 1 per cent and was almost equally distributed between the obstructive and non-obstructive groups. Diarrhoea alternating with constipation was present in only 8.8 per cent of cases and was also equally common in obstructive and non-obstructive groups. Umbilical fistula was present in 2 cases, but perianal and internal
Diagnosis of abdominal tuberculosis fistulas were not found. This symptomatology is in marked contrast to that of Crohn’s disease in which diarrhoea, lower abdominal pain and external and internal fistulas are the dominating features along with the obstructive signs. Diarrhoea has been reported to be present in 90 per cent, external abdominal fistula in 21.3 per cent, perianal and perirectal fistulas in 28-50 per cent and internal fistula in 11-1 7.5 per cent of cases of Crohn’s disease. Fever was present in 42.2 per cent of our cases. An incidence of 37.9-58 per cent has been reported by other workers. The incidence of menstrual disorder was 35.6 per cent. Other authors have reported an incidence varying from 15 to 91.5 per cent (Singhai et al., 1964; Bhansali and Desai, 1968; Levine, 1968; Sharma et al., 1972). The duration of symptoms before presentation was up to 1 year in most of the patients (54.6 per cent). Other workers (Levine, 1968; Sharma et al., 1972) found a duration of 1-3 years to be the commonest. Bhansali and Desai (1968) reported a duration of less than 1 year in most of their cases. Physical signs Tenderness and abdominal distension were the most frequent physical signs (65.9 and 58.2 per cent respectively). Other workers have reported the incidence to vary from 28 to 84 per cent. An abdominal mass has been reported in nearly half of the cases of many series while it was found in only 28.6 per cent of our cases. Only 18.5 per cent of our patients presented with a lump in the right iliac fossa, while others have reported it in the majority of cases. The reason for this difference appears to be that we have included all types of abdominal tuberculosis in our series. Visible peristalsis was present in 35.1 per cent of cases. A far lower incidence (10 per cent) has been recorded by other workers (Taylor, 1945; Anand and Pathak, 1961; Bhansali and Desai, 1968; Levine, 1968; Sharma et al., 1972). Its incidence was significantly less in non-obstructive cases (10.1 per cent) compared with obstructive cases (59 per cent). Thus, visible peristalsis does not necessarily indicate an obstructive lesion. Investigations A moderate degree of anaemia, a raised ESR and a normal total leucocyte count were the commonest haematological findings. Other workers (Dutta Gupta, 1958; Levine, 1968; Prakash et al., 1970; Sharma et al., 1972) have reported similar findings. Chest X-ray evidence of pulmonary tuberculosis was present in 24 out of 86 cases. Other authors have found pulmonary involvement in abdominal tuberculosis in 3.7 per cent (Sharma et al., 1972) to 77 per cent (Hoon et al., 1950). Hoon et al. (1950) considered radiological examination to be the single most important investigation for establishing the diagnosis of abdominal tuberculosis. However, it was helpful in the diagnosis of only 29 out of 46 cases in this series. Moreover, false positive results were seen. It is emphasized that similar radiological features
were found in the obstructive and non-obstructive groups. Open peritoneal biopsy through the right iliac fossa gave a positive result in 11 out of 12 cases. Adequate peritoneal tissue was available for examination and the procedure itself was simple and consisted of making a small grid-iron incision in the right iliac fossa under local anaesthesia. There were no complications and even in cases with ascites the wound healed. In a review of the literature no mention of a similar procedure was found. Our finding of fatty change in the liver in 32.6 per cent of cases is about half of the 64 per cent reported by Gambhir et al. (1972). Taylor (1945) and Burack and Hollister (1960) considered a case tuberculous if the ascitic fluid contained more than 250 cells/mm3 and a protein level of more than 2.5 g per cent. In the present study in cases where the cell count was 100/mm3 the peritoneal biopsy showed tuberculosis. However, in one case with a positive peritoneal biopsy the cell count was only 10/mm3and the protein 3.5 g per cent . It is the general impression that bacteriological examination of ascitic fluid is unreliable. Burack and Hollister (1960) and Mehrotra et al. (1964) found one positive culture each in 41 and 13 cases respectively. Levine (1968), however, obtained a positive culture in 30 per cent of the cases studied. In the present series positive bacteriological results were obtained in 60 per cent (3 out of 5 studied). In one case each, tubercle bacilli were demonstrated by smear examination, culture for AFB and guineapig inoculation. In one case anonymous mycobacteria were grown from the ascitic fluid.
Acknowledgement We are grateful to Professor L. E. Hughes, Department of Surgery, Welsh National School of Medicine, Cardiff, for his critical remarks in the preparation of this paper. References ADAMS R. and MILLER w. H. (1946) Surgical treatment of intestinal tuberculosis. Surg. Clin. North Am. 26, 656-664. ANAND s. s. and PATHAK I. c. c. (1961) Surgical treatment of abdominal tuberculosis with special reference to ileocaecal tuberculosis. A record of one hundred cases treated surgically. J. Indian Med. Assoc. 37, 423429. BHANSALI s. K . and DESAI A. N. (1968) Abdominal tuberculosis, clinical analysis of 135 cases. Indian J . Surg. 30, 216. BOBROW M. L. and FRlEDMAN S. (1956) Tuberculous appendicitis. Am. J. Surg. 91, 389-393. BURACK w. R. and HOLLISTER R. M. (1960) Tuberculous peritonitis. A study of forty-seven proved cases encountered by a general medical unit in twentyfive years. Am, J . Med. 28, 510-523. DUTTA GUPTA A. K . (1958) Intestinal tuberculosis. Indian J. Surg. 20, 396. 945
Pritam Das and H. S . Shukla and RAWAT M. L. (1972) Hepatic involvement in abdominal tuberculosis (a clinical, biochemical and histopathological study). J. Assoc. Phys. India 20, 843. HOON J. R., DOCKERTY M. B., PEMBERTON J. DE J . (1950) Collective review; ileocaecal tuberculosis including comparison of this disease with nonspecific regional enterocolitis and noncaseous tuberculated enterocolitis. Znt. Abstr. Surg. 60, 417440. LEVINE H. (1968) Needle biopsy diagnosis of tuberculous peritonitis. Am. Rev. Rrsp. Dis.97, 889894. MEHROTRA M. P., MATHUR K. s., WAHI P. N. et al. (1964) Percutaneous biopsy of peritoneum as a diagnostic aid in cases of ascites. J. Indian Men. Assoc. 43, 319-321.
GAMBHIR M. S., GOYA S. K.
946
(1966) In: BOCKUS H. L. (ed.) Gastroenterology. Vol. 11, 2nd ed. Philadelphia, Saunders, p. 3 1 1 . PRAKASH A., TANDON H. D., NIRMALA L. et al. (1970) Chronic ulcerative lesions of the bowel. Indian J. Surg. 30, 1 . SHARMA G . C., KALA P. C. and BHARGAVA M. (1972) Obstructive tubercular lesions of the large gut. Am. J . Proctol. 23, 218-227. SINGHAI s. L., TANDON P. L., HAFIZ M. A . et al. (1964) Abdominal tuberculosis. Indian J . Surg. 26, 440. TAYLOR A . w. (1945) Chronic hypertrophic ileocacal tuberculosis, and its relation to regional ileitis (Crohn’s disease). Br. J. Snrg. 33, 178-181. PAUSTIAN F. F.
Clinical diagnosis of abdominal tuberculosis P R I T A M D A S A N D H. S . S H U K L A * Clinically, the patients were divided into two main One hundred and eighty-two cases of abdominal tuber- groups: those who presented with intestinal obstrucculosis admitted to Swaroop Rani Nehri ( S R N ) Hospi- tion (93 cases) and those who did not have intestinal tal, Allahabad, in the past 7 years have been reviewed. obstruction (89 cases). The 20 ascitic cases are The clinical diagnosis of abdominal tuberculosis was included in the non-obstructive group, although they made correctly only in 50 per cent of cases. About half had a distinctive clinical picture with fluid distension the cases presented with chronic or acute on chronic of the abdomen. intestinal Obstruction. The remaining patients had tiague pains, tender abdomen, constitutional symptoms or a mass in the abdomen. Diarrhoea was not frequent and fistula formation was rare. A chronic obstructive type loo of lesion was found not only in cases with a bowel lesion 80 but also in patients with chronic miliary peritonitis and tuberculous mesenteric adenitis. Similarly, a lump was present not only in hypertrophic bowel tuberculosis but 9 60 also in chronic miliary peritonitis and tuberculous J mesenteric adenitis. On radiological examination false r positive features such as fluid levels, bowel dilatation or 3 40 even the ‘string’ sign were encountered. Liver and endometrial biopsies were positive in only a very few cases. Peritoneal biopsy was of considerable help, 20 being positive in 88 per cent of ascitic cases and in 42.1 per cent of non-ascitic cases. Open peritoneal biopsy obtained after making a small incision in the right iliac fossa was found to be the most Age useful investigation in the diagnosis of abdominal tuberculosis. A n ascitic fluid protein content of 2.5g or more Fig. 1. Age incidence of abdominal tuberculosis nnd a predominantly lymphocytic count ojover 100/mm3 are diagnostic, but a cell count of 10/mm3was recorded Table I: INCIDENCE OF VARIOUS TYPES OF LESIONS in one tuberculous case. No. of Type of lesion cases % ABDOMINAL tuberculosis continues to be a common Stricture of ileum, jejunum or both 27.5 50 disease responsible for considerable morbidity and Hyperplastic ileocaecal lesion with or 20.8 38 without small bowel stricture mortality in India. However, even where it is relatively 5 2.8 common, the diagnosis of abdominal tuberculosis is Colonic stricture Mesenteric lymphadenitis 21 11.6 often obscure. The symptoms and signs are often quite Chronic miliary tuberculous peritonitis 48 26.4 vague and the laboratory investigations and radio- Tuberculous ascites 10.9 20 logical findings unhelpful. The present study has been Total 182 undertaken to assess the value of clinical features and laboratory and radiological investigations in proved Clinical features cases of abdominal tuberculosis. Accuracy of diagnosis Patients and methods The accuracy of clinical diagnosis was 50 per cent. The records of 182 cases of abdominal tuberculosis Other cases were provisionally diagnosed as chronic admitted to SRN Hospital, Allahabad, during the 7 or acute intestinal obstruction, abdominal pain of years 1967-73 were studied. The various types of uncertain aetiology, recurrent appendicitis, acute lesions found are shown in Table I. The cases were appendicitis, appendicular mass, pyloric stenosis, classified according to the dominant feature visible to peptic ulcer, renal mass, retroperitoneal sarcoma, the naked eye at operation. Mesenteric adenitis and faecal fistula, cirrhosis, malignant ascites and malabperitoneal tubercles were present in a number of sorption syndrome. cases along with the bowel lesion. However, as the main feature was the intestinal pathology, such cases * Department of Surgery, MLN Medical College, Allahabad, India. have been classified according to the type and site of Present address of H. S. Shukla: Surgical Unit, University the gut Iesion. Hospital of Wales, Cardiff. SUMMARY
1
-
941
Pritam Das and H. S. Shukla Table 11: INCIDENCE OF SYMPTOMS Stricture of ~ileum Symptom
0 33 31 19 4 7
NO
Hyperplastic i leocaeca I lesion ~ 0 23 18 13 5 1
NO 14
Pain 17 Vomiting 10 9 Constipation 3 5 Diarrhoea 1 0 Diarrhoea alternating 1 2 with constipation Weight loss I 5 2 3 2 Anorexia 1 9 3 5 4 Cough with expectoration 4 1 2 0 Moving lump inabdomen 18 7 7 0 Borborygmi 1 8 0 9 2 Post-prandial distress 13 2 9 3 Distension of abdomen 12 2 17 2 Menstrual disorders 1 1 4 6 3 Fever 8 2 6 8 0, Obstructive; NO, Non-obstructive. * Percentage of female patients. Table 111: SITE OF PAIN IN THE ABDOMEN Cases Cases with without obstruc- obstruc- No. of Site of pain tion tion cases
Stricture of colon ~ 0 NO
Mesenteric lymphadenitis 0 NO
Tubercular peritonitis _ 0 NO
0
Ascitic_ NO
_ Total
7;
94.0 69.6 46.7
2 2 2 0 0
3 2 1 0 0
8 8 5 2 0
9 7 3 1 0
25 22 2310 2 4 3 1 3 0 3
0 0 0 0 0
14 5 7 3 2
170 125 20 16
11.1
2 2 0 2 2 2 2 0 1
2
5 7
4 3
6 12 9 12 1 4 3 5 5 2 3 1 0 2 0 2 3 1 1 1 0 1 6
0 I0 0 15 0 3 0 0 0 2 0 2 0 1 4 0 7 0 1 6
63
35.0 444 8.8 28.8 25.5 27.2 45.0 35.6* 42.2
1 0 3 1 2 0 1 2
1
0
4 4 1 7 0 2
2 1 2 3 1 5
85
80
16 51 46 49 81 47 76
8.8
Symptoms Abdominal pain was the commonest symptom and was present in 170 cases (94 per cent) (Table ZZ). The site of the pain is given in Table IIZ. % Analysis of the character of the pain revealed the Umbilical 35 2s 60 35.4 following : Generalized 29 26 55 32.3 1 . As might be expected, the pain was usually Right iliac fossa 17 18 35 20.6 Epigastrium 5 6 11 6.5 colicky in the obstructive group (89 per cent). In the Left hypochondrium 3 1 4 2.4 remaining cases in this group the pain was vague in 6, Right hypochondrium 1 2 3 1.8 gripping in 2 and like that of peptic ulcer in 2 cases. Left iliac fossa 1 0 1 0.5 2. In the non-obstructive type of cases colicky pain Right lumbar region 0 1 1 0.5 was present in 37 cases out of 89, while the pain was Total 91 79 170 vague in 29 and gripping in character in 21 cases. In 2 cases of mesenteric lymphadenitis the pain waslikethat of peptic ulcer. In the ascitic group the character of Age and sex The age range is shown in Fig. 1. The predominant the pain was vague in the majority of cases. Vomiting was present in 88 per cent of cases of the age group was 21-30 years (47 per cent of cases). The next most frequent period was between 11 and 20 obstructive group. However, it was also present in years (28 per cent). After 30 years there is a sudden 48 per cent of cases without intestinal obstruction. drop in the incidence, which diminishes further with Patients with ascites had the lowest incidence of vomiting (25 per cent). increasing age. Constipation was more common in the obstructive There were 132 females (72 per cent) and 50 males (28 per cent). While the overall male to female group (67.7 per cent) than in the non-obstructive group ratio was 1 : 2.6, for the different subgroups it was (24.8 per cent). A ‘moving lump’ in the abdomen was 1 :2.8 in cases with stricture of the ileum, 1 : 8.5 in twice as common in the obstructive group (36 per cent hyperplastic ileocaecal lesions, 1 : 1.3 in mesenteric compared with 19 per cent in the non-obstructive lymphadenitis, 1 : 1.5 in chronic miliary tubercular group). Fever was present in 42.2 per cent of the cases. The peritonitis and 1 : 3 in ascitic cases. incidence of fever was greater in the non-obstructive group (49 out of 89) than in the obstructive group (27 Associated tuberculosis Associated pulmonary tuberculosis was present in 13 out of 93). It is therefore important to emphasize that (15 per cent) out of the 86 cases who had a radio- quite a large number of obstructive cases do not logical examination of the chest. In 6 other patients suffer from pyrexia. The incidence of menstrual disorders was 35.6 per (6.9 per cent) pleural effusion was present. These figures probably underestimate the incidence, because cent of the female patients. only patients for whom there was a suspicion of a pulmonary lesion were X-rayed. A family history of Duration of symptoms tuberculosis was forthcoming in 4 cases only. There The duration of symptoms before attending hospital was one case with associated glandular tuberculosis varied considerably. The largest number of patients but no case with skeletal tuberculosis. (30.4 per cent) had had symptoms for 1-6 months,
942
-
Diagnosis of abdominal tuberculosis Table 1V: INCIDENCE OF PHYSICAL SIGNS HyperStricture plastic of ileocaecal ileum lesion 0 NO Sign 0 NO 15 5 Poorly nourished 20 2 Pallor 17 4 25 6 Cervical adenitis 0 1 0 0 21 8 Tender abdomen 22 I Distension of abdomen 28 6 17 1 1 0 Doughy feel to abdomen 3 1 Visible peristalsis 26 2 12 4 1 1 0 1 Ascites Palpable liver 2 6 2 0 6 1 1 0 Rieiditv Lump in abdomen 5 4 1 1 5 0, Obstructive; NO, non-obstructive.
Stricture of colon 0 NO 2 2 1 0 1 0 2 0 2 0 2
while 24.5 per cent of cases had had symptoms for 6 months to 1 year. A short history (up to 1 month) was recorded for 20,5 per cent, and a history of 1-3 years in 12 per cent, 3-5 years in 11 per cent and over 5 years in 1.6 per cent of cases. There was little difference between the obstructive and non-obstructive groups.
Physical signs Tenderness and distension of the abdomen were the most frequent physical signs and were present in 65.9 and 58.2 per cent of cases respectively (Table ZV). An abdominal mass was found in 28.6 per cent of cases. The mass most frequently occurred in the right iliac fossa (18.5 per cent). The other sites where a mass was found were the umbilical region in 4.2 per cent, the epigastrium in 3.2 per cent and the right hypochondrium in 1.6 per cent of cases. A vague mass was present in 1.1 per cent of cases. No mass was palpable in the 20 cases of tuberculous ascites. A mass was found in 42.1 per cent of the hyperplastic ileocaecal tuberculosis cases, in 43 per cent of the mesenteric lymphadenitis cases, in 31.2 per cent of the tuberculous peritonitis cases, in 60 per cent of the cases with stricture of the colon and in 18 per cent of the cases with stricture of the small bowel. There was no significant difference in the occurrence of a mass in obstructive and non-obstructive groups. It is evident that an ileocaecal lesion was not the only type of case in which a mass could be present, and it even occurred in cases with tuberculous peritonitis. Visible peristalsis was present in 35.1 per cent of cases. In none of the ascitic cases was peristalsis seen, and its incidence was significantly less in nonobstructive cases (10.1 per cent) compared with obstructive cases (59 per cent). Pallor was noticed in 56.5 per cent of cases, and 45.6per cent of the patients appeared poorlynourished. Ascites was found in all ascitic cases but rarely in the other types. It should be noted that a ‘doughy feel to the abdomen’, which is often quoted as typical of a tuberculous abdomen, was present in only 6.0 per cent of cases. A pus-discharging sinus at the umbilicus was found in 2 cases. The liver was palpable in 14.2 per cent of cases.
1 1 1 1 0 0 1 1
0 0 1
Mesenteric lymphadenitis 0 3 I 0 4 8 0 I
NO 3 3 0 5 1 0 2
0
0
0 0 1
3 0 8
Tubercular peritonitis
Ascitic - -
0 NO 1 7 8 19 8 0
0
18 I5 22 8 1 3 8 0 5 4 1 5 6 2 4 1 1
0 0
NO
Total 83 103
0
9 11 0 9 14 2
0
0
0 0
0 0
0 2 0 0 4 0 0 0 0
3 110
106 11
64 33 25 16 52
% 45.6 56.5 1.6 65.9 58.2 6.0 35.1 18.6 14.2 8.7 28.6
Investigations Haematological findings The haemoglobin level was estimated in 165 cases. Eighteen per cent had haemoglobin levels above 12 g/dl, 60 per cent between 9 and 12 g/dl and 22 per cent below 9 g/dl. In the non-obstructive group the haemoglobin level was generally on the higher side. The ESR was found to be raised in 117 out of 126 cases examined (92.9 per cent). In 9 cases the ESR was normal although the disease was in the active stage. The total leucocyte count was normal in over half of the cases (55.5 per cent). In 43.8 per cent it was raised. There was no significant group variation. One case with ascites had leucopenia. The polymorphs were raised in 42.9 per cent, lymphocytes in 10.7 per cent and eosinophils in 7.9 per cent of cases. In 14.7 per cent of cases a low lymphocyte count was recorded. Chest X-ray Chest X-rays were taken in 86 cases. Thirteen (15.1 per cent) showed infiltration, 6 (6.9 per cent) showed pleural effusion and 5 (5.8 per cent) showed a healed tubercular focus. Thus, in 24 out of 86 cases evidence of pulmonary infection was present. The highest incidence (12 out of 24) of pulmonary involvement was noted in cases with tuberculous peritonitis and ascites. This is significant considering the aetiology of tuberculous peritoneal involvement which is believed to be haematogenous. The small bowel stricture group exhibited the next highest incidence of active or healed pulmonary tuberculosis (6 cases). Four cases with mesenteric adenitis and 2 with hyperplastic ileocaecal tuberculosis also had active or healed pulmonary tuberculosis. Plain X-ray of the abdomen A plain X-ray of the abdomen was taken in 54 cases. As expected, multiple fluid levels and gas shadows were present in 14 out of 29 cases of the obstructive group. In only one out of 25 cases belonging to the non-obstructive group was there demonstration of gas shadows and fluid levels. The majority of the nonobstructive cases had no positive findings. Calcified 943
Pritam Das and H. S. Shukla mesenteric lymph nodes were observed in 2 of the ascitic group and one of the colon stricture group. Gas under the diaphragm was found in one case belonging to the non-obstructive group and in another of the obstructive group. Barium meal examination Barium meal examination was done in 38 cases. In 15 cases the results were normal. In 34.2 per cent of cases areas of small bowel obstruction and dilatation were seen. An unusual finding was the presence of pyloric obstruction in one case and dilatation of tl-e first, second and third parts of the duodenum in 3 cases. A subsequent laparotomy revealed enlarged caseous mesenteric lymph nodes pressing on the pylorus and the third part of the duodenum respectively. Other positive findings were a ‘pulled-up’ caecum in one case and a filling defect in the caecum in 5 cases. Dilated bowel loops, narrowing of the terminal ileum, a filling defect in the caecum and even the ‘string’ sign were observed in some cases in which n o bomel lesion was found at exploratory laparotomy. Barium enema examination Five out of the 8 cases in whom this investigation was done showed a filling defect in the caecum, one case showed a stricture of the transverse colon and in the remaining 2 the bowel was normal. Intravenous pyelography Intravenous pyelography showed kinking of the left ureter in one case, which proved to be due to tubercular retroperitoneal fibrosis. Peritoneal biopsy Slides of 71 peritoneal biopsies were available for re-examination. Of these, 59 biopsies were done with an Abrahm’s needle, and 12 by making a small incision in the right iliac fossa, exposing the peritoneum and removing a small piece. In the ascitic and chronic miliary peritonitis groups, the biopsy material showed tubercular histopathology in 46 out of 52 cases. In the remaining 6 cases a non-specific picture was present, although these cases had other positive tutercular evidence. In cases of small bowel stricture, hyperplastic caecal lesions, colon stricture and mesenteric adenitis, 8 out of 19 biopsies revealed normal peritor,eum, 8 showed tubercular changes and 3 showed non-specific changes. Of the 12 cases (7 ascitic and 5 non-ascitic) having open peritoneal biopsy, tubercular histopathology was found in 1 1 . Liver biopsy Liver biopsy was done in 46 cases. Fifty per cent showed normal histopathology, 32.6 per cent showed fatty changes, 10 per cent revealed reactive hepatitis and 5 per cent showed caseation necrosis. In one case liver biopsy provided the diagnosis of abdominal tuberculosis. Two cases of hepatic tuberculosis occurred in the peritonitis group. Reactive hepatitis and fatty changes were predominantly seen in the cases with chronic miliary peritonitis. 944
Endometrial biopsy In ;i out of 21 endometrial biopsies performed tubQcular pathology was present. There were 4 failures to obtain endometrial tissue. It is significant to note that 4 out of the 5 cases with tubercular endometrium occurred in the miliary peritonitis and ascitic groups. The fifth case belonged to the tuberculous mesenteric lymphadenitis group. This could indicate that the endometrial lesion occurred first and later spread to the peritoneum. However, the possibility that the endometrium and peritoneum were both involved by blood-borne infection cannot be excluded. Ascitic fluid examination In 20 cases ascitic fluid was examined for cells and protein content. The cell count varied from 100 to 250 cells/mm3 in 8 cases. In 12 cases the ascitic fluid contained blood and therefore a cell count could not be done. The cells were predominantly lymphocytes. The protein level varied from 2.5 to 6.1 g per cent.
Discussion A correct clinical diagnosis of abdominal tuberculosis was made in 50 per cent of cases. Hoon et al. (1950) reported 34 per cent accuracy. The age incidence of the present series is similar to that reported by other workers ( A d a m and Miller, 1946; Bobrow and Friedman, 1956; Dutta Gupta, 1958; Paustian, 1966; Prakash et al., 1970; Sharma et al., 1972). The overall sex incidence showed a preponderance of the disease in females (72 per cent). However, the incidence varied according to the type of disease, with a female to male ratio of 8.5 : 1 in the hyperplastic ileocaecal tuberculosis group compared with 1.5 : 1 in the chronic miliary peritonitis group. Associated pulmonary tuberculosis was found in 15 per cent of cases and, considering the haematogenous route of infection in the aetiology of tuberculous peritonitis, it was interesting to note that half of the cases in this group had pulmonary tuberculosis. A much higher incidence of associated pulmonary tuberculosis as a whole has been reported by other workers (Hoon et al., 1950; Bobrow and Friedman, 1956; Singhai et al., 1964; Prakash et al., 1970). Paustian (1966) and Bhansali and Desai (1968) considered acute or chronic intestinal obstruction to be the commonest presentation of abdominal tuberculosis. While 50 per cent of our cases presented with intestinal obstruction, it was significant that the symptoms of pain, vomiting and constipation were present in some of the cases of the non-obstructive group as well. Therefore, the diagnosis of ‘obstructive lesion’ cannot be made with certainty on clinical grounds alone. Anorexia, loss of weight and a feeling of distension or of a mass in the abdomen were the other symptoms. Diarrhoea was present only in 1 1 . 1 per cent and was almost equally distributed between the obstructive and non-obstructive groups. Diarrhoea alternating with constipation was present in only 8.8 per cent of cases and was also equally common in obstructive and non-obstructive groups. Umbilical fistula was present in 2 cases, but perianal and internal
Diagnosis of abdominal tuberculosis fistulas were not found. This symptomatology is in marked contrast to that of Crohn’s disease in which diarrhoea, lower abdominal pain and external and internal fistulas are the dominating features along with the obstructive signs. Diarrhoea has been reported to be present in 90 per cent, external abdominal fistula in 21.3 per cent, perianal and perirectal fistulas in 28-50 per cent and internal fistula in 11-1 7.5 per cent of cases of Crohn’s disease. Fever was present in 42.2 per cent of our cases. An incidence of 37.9-58 per cent has been reported by other workers. The incidence of menstrual disorder was 35.6 per cent. Other authors have reported an incidence varying from 15 to 91.5 per cent (Singhai et al., 1964; Bhansali and Desai, 1968; Levine, 1968; Sharma et al., 1972). The duration of symptoms before presentation was up to 1 year in most of the patients (54.6 per cent). Other workers (Levine, 1968; Sharma et al., 1972) found a duration of 1-3 years to be the commonest. Bhansali and Desai (1968) reported a duration of less than 1 year in most of their cases. Physical signs Tenderness and abdominal distension were the most frequent physical signs (65.9 and 58.2 per cent respectively). Other workers have reported the incidence to vary from 28 to 84 per cent. An abdominal mass has been reported in nearly half of the cases of many series while it was found in only 28.6 per cent of our cases. Only 18.5 per cent of our patients presented with a lump in the right iliac fossa, while others have reported it in the majority of cases. The reason for this difference appears to be that we have included all types of abdominal tuberculosis in our series. Visible peristalsis was present in 35.1 per cent of cases. A far lower incidence (10 per cent) has been recorded by other workers (Taylor, 1945; Anand and Pathak, 1961; Bhansali and Desai, 1968; Levine, 1968; Sharma et al., 1972). Its incidence was significantly less in non-obstructive cases (10.1 per cent) compared with obstructive cases (59 per cent). Thus, visible peristalsis does not necessarily indicate an obstructive lesion. Investigations A moderate degree of anaemia, a raised ESR and a normal total leucocyte count were the commonest haematological findings. Other workers (Dutta Gupta, 1958; Levine, 1968; Prakash et al., 1970; Sharma et al., 1972) have reported similar findings. Chest X-ray evidence of pulmonary tuberculosis was present in 24 out of 86 cases. Other authors have found pulmonary involvement in abdominal tuberculosis in 3.7 per cent (Sharma et al., 1972) to 77 per cent (Hoon et al., 1950). Hoon et al. (1950) considered radiological examination to be the single most important investigation for establishing the diagnosis of abdominal tuberculosis. However, it was helpful in the diagnosis of only 29 out of 46 cases in this series. Moreover, false positive results were seen. It is emphasized that similar radiological features
were found in the obstructive and non-obstructive groups. Open peritoneal biopsy through the right iliac fossa gave a positive result in 11 out of 12 cases. Adequate peritoneal tissue was available for examination and the procedure itself was simple and consisted of making a small grid-iron incision in the right iliac fossa under local anaesthesia. There were no complications and even in cases with ascites the wound healed. In a review of the literature no mention of a similar procedure was found. Our finding of fatty change in the liver in 32.6 per cent of cases is about half of the 64 per cent reported by Gambhir et al. (1972). Taylor (1945) and Burack and Hollister (1960) considered a case tuberculous if the ascitic fluid contained more than 250 cells/mm3 and a protein level of more than 2.5 g per cent. In the present study in cases where the cell count was 100/mm3 the peritoneal biopsy showed tuberculosis. However, in one case with a positive peritoneal biopsy the cell count was only 10/mm3and the protein 3.5 g per cent . It is the general impression that bacteriological examination of ascitic fluid is unreliable. Burack and Hollister (1960) and Mehrotra et al. (1964) found one positive culture each in 41 and 13 cases respectively. Levine (1968), however, obtained a positive culture in 30 per cent of the cases studied. In the present series positive bacteriological results were obtained in 60 per cent (3 out of 5 studied). In one case each, tubercle bacilli were demonstrated by smear examination, culture for AFB and guineapig inoculation. In one case anonymous mycobacteria were grown from the ascitic fluid.
Acknowledgement We are grateful to Professor L. E. Hughes, Department of Surgery, Welsh National School of Medicine, Cardiff, for his critical remarks in the preparation of this paper. References ADAMS R. and MILLER w. H. (1946) Surgical treatment of intestinal tuberculosis. Surg. Clin. North Am. 26, 656-664. ANAND s. s. and PATHAK I. c. c. (1961) Surgical treatment of abdominal tuberculosis with special reference to ileocaecal tuberculosis. A record of one hundred cases treated surgically. J. Indian Med. Assoc. 37, 423429. BHANSALI s. K . and DESAI A. N. (1968) Abdominal tuberculosis, clinical analysis of 135 cases. Indian J . Surg. 30, 216. BOBROW M. L. and FRlEDMAN S. (1956) Tuberculous appendicitis. Am. J. Surg. 91, 389-393. BURACK w. R. and HOLLISTER R. M. (1960) Tuberculous peritonitis. A study of forty-seven proved cases encountered by a general medical unit in twentyfive years. Am, J . Med. 28, 510-523. DUTTA GUPTA A. K . (1958) Intestinal tuberculosis. Indian J. Surg. 20, 396. 945
Pritam Das and H. S . Shukla and RAWAT M. L. (1972) Hepatic involvement in abdominal tuberculosis (a clinical, biochemical and histopathological study). J. Assoc. Phys. India 20, 843. HOON J. R., DOCKERTY M. B., PEMBERTON J. DE J . (1950) Collective review; ileocaecal tuberculosis including comparison of this disease with nonspecific regional enterocolitis and noncaseous tuberculated enterocolitis. Znt. Abstr. Surg. 60, 417440. LEVINE H. (1968) Needle biopsy diagnosis of tuberculous peritonitis. Am. Rev. Rrsp. Dis.97, 889894. MEHROTRA M. P., MATHUR K. s., WAHI P. N. et al. (1964) Percutaneous biopsy of peritoneum as a diagnostic aid in cases of ascites. J. Indian Men. Assoc. 43, 319-321.
GAMBHIR M. S., GOYA S. K.
946
(1966) In: BOCKUS H. L. (ed.) Gastroenterology. Vol. 11, 2nd ed. Philadelphia, Saunders, p. 3 1 1 . PRAKASH A., TANDON H. D., NIRMALA L. et al. (1970) Chronic ulcerative lesions of the bowel. Indian J. Surg. 30, 1 . SHARMA G . C., KALA P. C. and BHARGAVA M. (1972) Obstructive tubercular lesions of the large gut. Am. J . Proctol. 23, 218-227. SINGHAI s. L., TANDON P. L., HAFIZ M. A . et al. (1964) Abdominal tuberculosis. Indian J . Surg. 26, 440. TAYLOR A . w. (1945) Chronic hypertrophic ileocacal tuberculosis, and its relation to regional ileitis (Crohn’s disease). Br. J. Snrg. 33, 178-181. PAUSTIAN F. F.
