Colonic Strictures Following Successful Medical Management of Necrotizing Enterocolitis: A Prospective Study Evaluating Early Gastrointestinal Contrast Studies By Jayant Radhakrishnan,

Gary Blechman,

Charisse Shrader,

Mukesh K. Patel, Henry H. Mangurten,

and John C. McFadden Park Ridge, Illinois l This is a prospective

study of 50 patients with neonatal

necrotizing enterocolitis

(NEC) treated successfully by medi-

cal means. They were all screened with an upper gastrointestinal (GI) contrast study after 14 days of healing and prior to establishment

of feeding.

Thirty-six

normal upper GI examinations

patients

responded

(72%)

with

well to a gradu-

ated increase in feeding. Another 5 (10%) with questionable areas on their upper GI examination

had a normal follow-up

contrast enema. Feeding was successfully established group of infants also. The remaining demonstrable tive

in this

(18%)

had

strictures in both contrast studies. After elec-

resection

continuity,

9 patients

of strictures

they

were

with

also fed

restoration successfully.

of intestinal No

delayed

strictures were seen in any of the patients. We propose that this method of evaluation diagnosis

of strictures

is safe, efficient and reliable in the

that develop

in patients

recovering

from NEC. Copyright

o 1991 by W.B. Saunders Company

INDEX WORDS:

Necrotizing

enterocolitis;

colonic strictures.

N

ECROTIZING enterocolitis (NEC) is one of the most common acquired emergencies in neonatal intensive care units (NICUs).‘” Better understanding of its pathophysiology combined with early and aggressive medical management have reduced its overall mortality4; however, colonic strictures have been reported in 15% to 35% of cases that recover with medical management.‘.’ A number of investigatorS’*5-7 have reported perforations, sepsis, and death secondary to obstruction caused by missed strictures because they are difficult to differentiate clinically from feeding intolerance and gastroenteritis. In an attempt to avoid these catastrophic consequences, some investigators have suggested routine contrast enemas.‘~2~5*s 0thers3*9 are reluctant to carry out routine contrast studies because the incidence of symptomatic strictures is low and a contrast enema would not demonstrate small bowel strictures nor would an early study identify late-developing strictures. In an effort to resolve these issues, we performed a prospective study of infants managed medically for NEC. On completion of treatment, these infants were screened radiographically by performing an upper gastrointestinal (GI) examination with follow through. In patients with suspicious areas on upper GJ examination, a contrast enema was performed. JournalofPediatric Surgery, Vol 26, No 9 (September), 1991: pp 1043-1046

MATERIALS

AND METHODS

The study group consisted of infants in the NICU who were diagnosed as having NEC as evidenced by abdominal distension, hematochezia, pneumatosis intestinalis, and associated clinical problems. The entire triad was not necessarily present in all cases. After the diagnosis was made, a full “septic workup” was completed and the infants were kept NPO with nasogastric decompression. They were also placed on intravenous antibiotics for 14 days. Serial clinical and radiographic evaluations of the abdomen were carried out and, after both parameters returned to normal, another 14 days of improvement were counted off. In all but 1 case a single observer (J.R.) made the diagnosis and followed-up the clinical and x-ray findings to define improvement. After these 14 days of improvement, a nonionic, water-soluble radiographic contrast was used to obtain an upper GI series with delayed films that followed the contrast through the small and large intestines. If a suspicious area was noted, a contrast enema using a nonionic, water-soluble material was carried out. Almost all the x-rays and contrast studies were undertaken and read by a single radiologist (J.C.M.). Feeding was commenced in all patients with normal upper GI examinations and those with a questionable upper GI examination with a normal contrast enema. Patients with abnormal upper and lower GI examinations underwent elective resection of the stricture with primaIy anastomosis. They were fed on return of intestinal function (Fig 1). A telephone follow-up of these patients was undertaken at the end of December 1989. Where this was unsuccessful, charts from the follow-up clinics were used. The patients were questioned about any history of subsequent hospitalization or symptomatology suggestive of gastrointestinal illness. RESULTS

During a 4-year 3-month period extending from January 1985 through March 1989, a total of 82 cases were evaluated and treated by one of us (J.R.) for possible NEC. All but 1 of these cases were seen by the observer on first suspicion of the diagnosis. Thirty-two cases had to be excluded from the study because protocol criteria were not fulfilled, eg, 6 babies required immediate surgery, 22 either did not From the Division of Neonatology, Department of Pediattics, and the Depafiments of Surgery Radiology, Lutheran General Children’s Medical Center, Park Ridge, IL. Presented at the 42nd Annual Meeting of the Sutgicat Section of the American Academy of Pediatrics, Boston, Massachusetts, October 6- 7, 1990. Address reprint requests to Jayant Radhakrishnan, MD, 2454 E Dempster, Suite 406, Des Plaines, IL 60016. Copyright o 1991 by W.B. Saunders Company 0022-3468/9112609-0007$03.0010

1043

RADHAKRISHNAN

1044

Upper Gl with follow through up to rectum

+

+

Normal

Possible Stricture

I

Lower GI

+

Normal

4 Confirmedstricture

Resection and anastomosis

Fig 1.

Further management

of NEC.

have the contrast studies done correctly or they were not done at the appropriate time. In 4 patients the diagnosis of NEC was never satisfactorily confirmed. The 50 patients who qualified for this study consisted of 21 boys and 29 girls with a mean gestational age of 32 weeks. At the onset of disease they ranged in age from 1 to 99 days (mean, 24 days). Their mean birth weight was 1,800 g (range, 775 to 3,880 g). Fifty percent of the patients had umbilical artery catheters but none had exchange transfusions. On an average the babies had been fed for 9 days (range, 0 to 30 days) prior to the development of symptoms. Antibiotic coverage consisted of Ampicillin and Gentamicin in 47 of the 50 patients. In 1 patient Vancomycin was added, whereas in 2 patients Ampicillin was combined with Ceftriaxone instead of Gentamicin. An average of 13 days (range, 7 to 21 days) of treatment was required for healing. Seventy-two percent (36 infants) had normal upper GI examinations; 28% (14 patients) had suspicious upper GI examinations. The second group of patients were subjected to a contrast enema that proved to be normal in 5 patients (10%). The 36 infants with a normal upper GI and the 5 in whom a lower GI did not confirm a stricture were started on a graduated feeding schedule. The lower GI confirmed a stricture in 9 patients (18%), who then underwent elective resection of the stricture with primary anastomosis. The lesion was located in the terminal ileum in 1 patient and in the left colon in the remaining 8. The surgical specimen showed transmural fibrosis in 7 and extensive ulceration and granulation in 2. They were all placed on a feeding schedule after intestinal function had returned. Forty-three of the 50 patients in the study (86%)

ET AL

had a follow-up ranging from 3 to 51 months (mean, 24 months) without rehospitalization for GI problems or any ongoing GI symptoms. Of the remaining 7 patients in the study, 2 died of other causes at 1 and 2 months after resolution of NEC; 3 were lost to follow-up after 1 month and 1 each after 1.5 and 2 months. These 7 patients also had no evidence of delayed stricture formation when last examined. We also followed-up the 28 patients excluded from this study. Of the 22 who were excluded for noncompliance with the protocol, 17 patients (77%) were seen for 5 to 51 months (mean, 28 months). One of these patients was unfortunately discharged from the hospital without contrast studies even though she had feeding problems. Within a month she returned with a rectal stricture, which we believe she had prior to discharge from the hospital. None of the others had late-developing symptoms. Five patients in this group were followed-up for less than 3 months. One died at 1 month and 2 each could not be contacted after 1 and 1.5 months. None of the 5 had GI symptoms when last contacted. Five of the 6 patients who required early surgery had an adequate follow-up ranging from 6 to 54 months (mean, 31 months). The remaining patient could not be contacted after 1 month. None of these 6 had GI symptoms at their last visit. DISCUSSION

In 1968, Rabinowitz et al” first reported the development of colonic strictures in patients who had recovered from acute NEC. It is believed that strictures occur due to fibrosis in an ischemic segment of bowel. Eighty percent of these strictures occur in the co10n,5,11usually on the left side. Strictures do occur in the small bowel, with the majority of these strictures occurring in the terminal ileum.1~2~8~12~‘5 The majority of strictures become symptomatic between 6 and 10 weekS,‘.2.‘2.14 although the interval can vary from the day of resolution of NEC’ to 20 months.” In 1988 Hartman et aI5 found that over 100 infants with post-NEC strictures reported in the literature had a recorded mortality of 8%. Schwartz et al2 attributed the increased incidence of strictures to the reduction in number of neonates who underwent acute surgical intervention. Although the reported incidences of strictures have varied from 3% to 44%,’ the consensus seems to indicate a range of 15% to 23%. We found an incidence of 18% among the 50 patients who fulfilled all the criteria of our protocol. This incidence would drop to 13.9% (10/72) if we were to include the 22 patients who responded to medical management but were excluded from the study.

COLONIC STRICTURES FOLLOWING NEC

1045

There are two major areas of controversy in the management of post-NEC strictures. The first is whether all documented strictures should be operated on and the second is whether routine contrast studies are indicated for diagnosis in all patients. The two areas of controversy overlap. Although everyone agrees that only symptomatic strictures should be treated, investigators such as Hartman et als feel that there is “a risk of delayed diagnosis due to overlap of symptoms with more common conditions such as formula intolerance and gastroenteritis.” Furthermore, they also have indicated that the first presentation of a stricture may be sudden and associated with life-threatening sepsis or perforation. Therefore, they advise routine diagnostic contrast studies while patients who are treated medically for NEC are recovering from the disease. Kosloske et al’ and Schwartz et a12*8agree with Hartman’s approach. On the other hand, Kliegman and Fanarot? and Born et al9 recommend avoiding routine contrast studies because there is a low incidence of symptomatic strictures in most series (15% to 23%). This is countered by Schwartz et al,’ who identified colonic stenosis in 7 of 28 infants (25%) responding to medical management. Six of these infants were asymptomatic at the time of the study and 3 were managed without an operation. They felt that in the 3 patients that did develop symptoms, prompt intervention and an uncomplicated postoperative course could be attributed to the fact that they had prior knowledge of this stenosis. Hartman et al5 further emphasize that “occurrence of such life threatening complications suggest that clinical observation alone is not adequate in the management of many of these infants.” Opposition to the routine use of contrast enemas is based on the following reasons: (1) false-negative studies could occur in patients who have small bowel strictures or late-developing strictures; (2) a general reluctance to subject asymptomatic infants to a relatively invasive procedure; (3) the possibility of spontaneous resolution of strictures2,s.‘6.‘7;(4) a potential risk of taking an

unstable child to the radiology suite; (5) exposure to radiation; (6) risk of perforation after an enema; and (7) the increased cost of performing routine contrast studies. We believe that our results preclude most of the reasons for not doing routine studies for strictures: (1) a nonionic, water-soluble contrast is essentially innocuous and this contention has been borne out in 2 reports’8,‘9; (2) by using an upper GI study with a follow-through as the screening procedure, the patient could be studied in the NICU; (3) the followthrough films on the upper GI examination demonstrate the terminal ileum and avoid the possibility of missed ileal strictures; (4) if the forward passage of contrast is impeded one can not expect the patient to be asymptomatic once oral feedings are started; and (5) because all patients are studied while in the hospital, the experience of Born et al9 with missed strictures due to failed follow-up would be completely avoided. Although increased cost is a concern, the potentially catastrophic complications and their attendant additional expense suggest that it would be prudent to err on the side of overinvestigation rather than miss a stricture. Although our protocol of early contrast studies could result in a missed latedeveloping stricture, that has not been our experience to date. We believe that the so-called late-developing strictures are in fact strictures that were diagnosed late. In this institution, prior to the commencement of this study, one patient who had recovered from NEC presented with a bowel obstruction secondary to a late-developing colonic stricture. This .patient had a routine upper GI with follow through at the time of discharge from the hospital, which showed a suspicious area that was not investigated further because the patient was asymptomatic at that time. We propose that a diagnostic protocol of early upper GI contrast studies followed by a contrast enema in suspicious cases is a safe, efficient, and reliable method of diagnosing strictures in patients who are treated medically for neonatal NEC.

REFERENCES 1. Kosloske AM, Burstein J, Bartow SA: Intestinal obstruction due to colonic stricture following neonatal necrotizing enterocolitis. Ann Surg 192:202-207,198O 2. Schwartz MZ, Hayden CK, Richardson CJ, et al: A prospective evaluation of intestinal stenosis following necrotizing enterocolitis. J Pediatr Surg 17:764-770, 1982 3. Kliegman RM, Fanaroff AA: Necrotizing enterocolitis. N Engl J Med 310:1093-1103,1984 4. Beasley SW, Auldist AW, Ramanujan TM, et al: The surgical management of neonatal necrotizing enterocolitis, 1975~1984. Pediatr Surg Int 1:210-217, 1986 5. Hartman GE, Drugas GT, Shochat SJ: Post-necrotizing

enterocolitis strictures presenting with sepsis or perforation: Risk of clinical observation. J Pediatr Surg 23:562-566,1988 6. Kosloske AM: Necrotizing enterocolitis in the neonate. Surg Gynecol Obstet 148:259-269,1979 7. Bell MJ, Temberg JL, Askin FB: Intestinal stricture necrotizing enterocolitis. J Pediatr Surg 11:319-327,1976

in

8. Schwartz MZ, Richardson CJ, Hayden CK, et al: Intestinal stenosis following successful medical management of necrotizing enterocolitis. J Pediatr Surg 15:890-899,198O 9. Born M, Holgersen LO, Shahrivar F, et al: Routine contrast enemas for diagnosing and managing strictures following non-

RADHAKRISHNAN

1046

operative treatment of necrotizing enterocolitis. .J Pediatr Surg 20:461-463, 1985 10. Rabinowitz JG, Wolf BS, Feller MR, et al: Colonic changes following necrotizing enterocolitis in the newborn. AJR Am J Roentgen01 103:359-364,1968 11. Janik JS, Ein SH, Mancer K: Intestinal stricture necrotizing enterocolitis. J Pediatr Surg 16:438-443, 1981

after

12. O’Neill JA Jr, Stahlman MT, Meng HC: Necrotizing enterocolitis in the newborn: Operative indications. Ann Surg 182:274279,1975 13. Reid WD, Shannon MP: Necrotizing enterocolitis-A medical approach to treatment. Can Med Assoc J 108:573-576,1973 14. Kosloske AM, Martin LW: Surgical complications of neonatal necrotizing enterocolitis. Arch Surg 107:223-228,1973

ET AL

15. Costin BS, Singleton EB: Bowel stenosis as a late complication of acute necrotizing enterocolitis. Radiology 128:435-438,1978 16. Tonkin ILD, Bjelland JC, Hunter TB, et al: Spontaneous resolution of colonic strictures caused by necrotizing enterocolitis: Therapeutic implications. AJR Am J Roentgen01 130:1077-1081, 1978 17. Pokorny WJ, Harr VL, McGill CW, et al: Intestinal stenosis resulting from necrotizing enterocolitis. Am J Surg 142:721-724, 1981 18. Cohen MD, Schreiner R, Grosfeld J, et al: A new look at the neonatal bowel-contrast studies with metrizamide (amipaque). J Pediatr Surg 18:442-448, 1983 19. Dutton RV, Singleton EB: Use of low osmolar contrast for gastrointestinal studies in low birth-weight infants. Am J Dis Child 141:635-638,1987

Discussion K. West (Indianapolis, IN): Were any of these patients who by study had strictures at all clinically symptomatic? Did they have palpable abdominal masses, bilious gastric drainage, or lack of stools? Did you feed any of these children to evaluate if these were really going to be clinically significant strictures? We have all seen patients that we have explored for other reasons such as ostomy closures and have seen clinical evidence of strictures that are not causing any kind of obstructions now. J. Raffensperger (Chicago, IL): It is better if we stand at the bedside of every child who has had NEC and teach. If, once we start feeding this child he throws up or gets distended, stop the feedings and think of a stricture. That seems to me to make more sense. It results in thoughtful medicine, it doesn’t cost much, and it is better for the baby. C. Priebe (Stony Brook, NY): We also have to look at the issue of the cost of this study. Fifty upper GI studies using metrizamide, an expensive material, costs $60.00 for about 20 mL. It is questionable to do this in babies who most of us feel should be observed and fed to see if they become symptomatic and then need an operation. I think this is unwarranted expense. T. Lobe (Memphis, TN): This paper emphasizes the concept that Marshall Schwartz and I and some others have suggested: that these infants all needed to be studied before discharge and that often we are going to find strictures that may or may not be clinically significant. My concern is that maybe you are studying these patients too early. You have an acute inflammatory process at about 3 weeks of age in these infants, and several weeks must pass for scarring to occur. Do you think that maybe you do your

contrast evaluation before you have had a chance for the inflammatory process to mature, allowing wound contraction to occur thus manifesting strictures at 6 weeks of age, which you might miss with the early studies? Dr Spence: I agree that these studies were done too early in the course of the disease, but you did not to tell us about the histopathology of the resected specimen. S. Shochat (Stanford, CA): Do you operate on all the patients? We have had a couple of children who we have treated with a similar protocol. We don’t operate immediately if we see a stricture immediately. We will then continue to use total parenteral nutrition for at least another 2 to 3 weeks, sometimes up to 1 month, and then repeat the study. J. Radhakrishnan (response): Dr West and Dr Raffensperger commented on whether these patients were symptomatic or not. Our study was started after we had a couple of asymptomatic patients who had been followed clinically and then returned with problems. This was the lead off to see if we could identify them. With regard to Dr West’s comment about feeding these babies, we felt that an upper GI with follow-through would simulate an oral feeding. In terms of doing the study too early, Dr Lobe and Dr Shochat, we saw no delayed strictures, so we feel that we can pick them up at 2 weeks after resolution of NEC and waiting is not required. At this time, metrizamide is expensive but there are other contrast materials awaiting FDA approval that would drop the price. With regard to the pathology, 7 showed fibrosis in the intestinal wall and there was granulation tissue in 2 of the patients.

Colonic strictures following successful medical management of necrotizing enterocolitis: a prospective study evaluating early gastrointestinal contrast studies.

This is a prospective study of 50 patients with neonatal necrotizing enterocolitis (NEC) treated successfully by medical means. They were all screened...
538KB Sizes 0 Downloads 0 Views