Perforation into gut by ventriculoperitoneal shunts: A report of two cases and review of the literature.

We report two cases of gastrointestinal perforation by ventriculoperitoneal (VP) shunts and review the literature on the topic. The time interval between shunt surgery and detection of bowel perforation is minimum in infants and increases with age. Sigmoid and transverse colon followed by stomach are the most frequent sites of gastrointestinal perforations by VP shunts.

INTRODUCTION

The peritoneal end of the ventriculoperitoneal (VP) shunt has been associated with complications such as pseudocyst formation, perforations of hollow viscus, penetration into solid organs and abdominal wall and protrusion outside body. We report two cases of bowel perforation by VP shunts.

CASE REPORTS

Case 1

A 9-month-old male had undergone the right-sided VP shunt (Chhabra-slit-in-spring silicone shunt) procedure 7 months back for congenital hydrocephalus. He presented to us with complaints that the child protruded a white tube per anus on defecation for past 15 days with clear fluid dripping from it. On examination, the child was afebrile, alert and had no neck rigidity, and the abdomen was soft. On rectal examination, there was a white tube coming from beyond the reach of finger. Total leukocyte count (TLC) was 7200/cumm. An ultrasonography (USG) abdomen was normal. The sigmoidoscopy showed VP shunt protruding into the colon at 18 cm from the anal verge. The child was operated and the shunt was cut at abdominal surface through a small incision. The rest of the distal tube was extracted per rectum. The proximal tube was taken out as external drainage. On antibiotics, the child improved. The cerebrospinal fluid (CSF) culture done after 1 month was sterile and so a revision of shunt was done on the left side. The child was asymptomatic at 3 years follow-up.

Case 2

A 3-year-old male, who underwent VP shunt 1 year ago for congenital hydrocephalus presented with similar complaints as the above patient [Figure 1]. On examination, the child was stable except that proctoscopy showed VP shunt protruding into the rectum at about 12 cm from anal verge. TLC was 6400/cumm, CSF was sterile and the USG abdomen was normal. The child was operated and managed similarly as the above patient with successful outcome and is on regular follow-up.

Figure 1

VP shunt protruding from anus of patient in case 2

DISCUSSION

Perforation of bowel by VP shunts is rare and the incidence is only 0.1-0.7% of shunt surgery.[1] A search of the literature revealed a total of 94 patients. More than half, i.e., 49 cases were reported in the age group 0-10 years. Our both patients belonged to this group. Thirty-four were male, 26 were females while in 34 cases sex was not mentioned. The duration of time between VP shunt surgery and detection of bowel perforation was found minimum (mean, 4.86 months) in the 0-1 year age group and found increasing with age with an overall mean duration of 24.8 months [Table 1].

Table 1

Mean duration of time from VP shunt surgery to GI perforation

The possible factors responsible for this complication are thin bowel wall in children, sharp and stiff end of the VP shunt,[23] use of trocar by operating surgeons,[4] chronic irritation by the shunt,[5] previous surgery, infection and silicone allergy.[6] Congenital hydrocephalus (n = 33) was the most common diagnosis followed by infective etiology (10), cysts/tumors (5), normal pressure hydrocephalus (5), intracerebral hemorrhage (4), tubercular (3), and trauma (1) while in 33 cases diagnosis was not mentioned.

The total number of patients found to have anal protrusion of VP shunts was 55 (58.5%) including our two patients [Table 2]. Similarly nine cases (9.6%) of oral extrusion were seen.[27-12] While 7 (7.45%) patients had peritonitis and 12 (12.8%) had meningitis separately, three patients developed both (3.2%).[413]

Table 2

Patient characteristics in the literature

The diagnosis was obvious in patients presenting with spontaneous extrusion of a whitish tube while defecating through which clear fluid dripped. Eight patients were subjected to a contrast study through distal shunt (shuntogram) which opacified the bowel confirming the diagnosis[1] while computed tomogram (CT) scan of abdomen showed the shunt in bowel lumen along with peritonitis if present.[61415] CT scan of head showing pneumocephalus in patients with VP shunts could clinch the diagnosis of bowel perforation when other causes were ruled out.[16] Patients with VP shunts who developed meningitis by “unusual pathogens” such as gram-negative and anaerobic organisms as confirmed by CSF examinations were suspected to have bowel perforation and were further investigated.[13] In patients with oral extrusion, upper GI endoscopy helped localize the site of perforation.[2] Finally in undiagnosed patients, an exploratory laparotomy was done.[1011]

Out of 94 patients, the exact site of perforation in gastrointestinal tract (GIT) was mentioned in 49 (52.1%) patients. In six out of nine patients with oral extrusion of VP shunt, the site was localized to stomach[27-10] and in two patients, jejunum.[11] In rest of the cases without oral extrusion,[1616-20] the site of perforation in bowel was as follows: Caecum (2), ascending colon (1), transverse colon (6), splenic flexure (1), descending colon (5), sigmoid (7), rectum (4), and unknown sites in colon (12).[1421] The cerebrospinal fluid (CSF) cultures were positive in most of the cases.[117]

Bowel perforation in patients with VP shunt should be considered with gram-negative meningitis or abdominal symptoms. The optimum treatment of such a patient would be decided by the presence of features of sepsis, perforation peritonitis, or intraperitoneal abscess. In a patient with simple bowel perforation and no other complications like our both patients, a formal exploratory laparotomy is not required. The shunt should be disconnected at abdominal wall and the lower end should be removed through the rectum by colonoscopy or sigmoidoscopy/proctoscopy.[15] The distal end of VP shunt should not be pulled back into the peritoneal cavity to prevent contamination of the tract. External ventriculostomy should be established at least for 3 weeks and patient should be put on broad spectrum antibiotics to prevent infection of CSF.[2415] After repeated CSF cultures are sterile, patient should undergo repeat VP shunt on the opposite side.[22] In patients with bowel perforation peritonitis, they should undergo exploratory laparotomy with removal of shunt, thorough lavage and primary closure of the bowel wall.[141016212023]

In our review, a total of five patients (5.32%) died. One patient each died of peritonitis[13] and intractable seizures,[23] two of continued bacterial ventriculitis[7] and in one patient who died of meningitis, autopsy revealed stomach perforation by VP shunt. The time relationship and the site of perforation in GI tract by VP shunt have never been studied in the past.

CONCLUSION

In symptomatic VP shunted patients, suspicion of bowel perforation should be kept high, if they develop abdominal symptoms or gram-negative or anaerobic meningitis. The duration of time between shunt surgery and detection of perforation was found minimum in infants and increasing with age. Most of the patients presented with asymptomatic passage of a tube per anus. Sigmoid and transverse colon followed by stomach are the most frequent sites of GI perforations. In a patient with simple bowel perforation and no other complications, a formal laparotomy is not required while in patients with intraabdominal complications, urgent laparotomy should be undertaken. If detected on time and managed properly, the results are good.