The Effect of Biologics on Postoperative Complications in Children With Inflammatory Bowel Disease and Bowel Resection

Abstract

Objectives: 

There has been limited investigation of pediatric patients with inflammatory bowel disease (IBD) who have been treated with biologic agents and undergo operative management. Postoperative complications in the adult setting have been mixed and in the pediatric population the data have been limited. This study compares children with IBD treated with biologic agents to patients treated with nonbiologic therapy before bowel resection.

Methods: 

This is a single-center, retrospective chart review study of 62 children with IBD who underwent bowel resection between 2001 and 2017. Analysis included patient demographics, medications used before surgery, incidence of postoperative complications, indication for surgery, type of operation, and additional surgeries required. Postoperative complications were defined as superficial skin infection, leak at anastomotic site, intra-abdominal abscess, wound dehiscence, and so on. Complications were compared based on medical therapy.

Results: 

Of the 62 children reviewed, 21 carried the diagnosis of ulcerative colitis, 40 had Crohn disease, and 1 had IBD-unspecified. Thirty-seven of the patients were treated with infliximab, adalimumab, or vedolizumab before their bowel resection. There were 4 complications documented within 30 days of the operation, with an overall complication rate of 6.45%. There were 2 complications in each of the cohorts, including intra-abdominal abscess (2), abdominal wall abscess (1), and pouchitis (1).

Conclusion: 

The number of complications was the same between those who did and did not receive a preoperative biologic agent. This study suggests that biologics may be safe to use in patients undergoing bowel resection.

See “Biologics and Surgery in Inflammatory Bowel Disease: Learning at the Cutting Edge” by Rosh on page 297.

What Is Known

• Despite the introduction of biologic agents, some patients with inflammatory bowel disease still require bowel resection.
• There are many conflicting studies on early postoperative complications in patients with inflammatory bowel disease who are preoperatively treated with a biologic agent and then undergo bowel resection.

What Is New

• This study suggests that a biologic agent may be safe to use in the preoperative period in children with inflammatory bowel disease undergoing bowel resection.
• Our data found a shorter postoperative hospital length of stay in the group that received preoperative biologic therapy compared to the nonbiologic group.

Patients with inflammatory bowel disease (IBD) have traditionally been managed with immunomodulator medications. The natural history of pediatric IBD, including ulcerative colitis (UC) and Crohn disease (CD), is complicated and often requires surgical intervention. In 1997, Patel et al ⁽¹⁾ identified the probability for surgery in pediatric patients 3 years after diagnosis of CD as 28.8% and by 5 years to be 47.2%.

In the last decade, biologic agents have revolutionized the management for patients with both UC and CD. Current FDA-approved biologic agents used in pediatric patients with include infliximab (IFX) and adalimumab. IFX is a chimeric monoclonal antibody–based tumor necrosis factor alpha inhibitor, which was first approved for pediatric use with CD in 2006 and pediatric UC in 2011. Adalimumab is a humanized monoclonal antibody approved for pediatric CD in 2014. Vedolizumab is an anti-integrin antibody, which acts by blocking the interaction between gut-specific lymphocytes and their receptor on vascular endothelium, and was approved for adult use in both CD and UC in 2014. These agents have been helpful in both inducing and maintaining remission in children with chronically active luminal disease ^(2,3).

Despite these advancements, patients with medically refractory IBD require surgical intervention ⁽⁴⁾. The incidence of abdominal surgery for these pediatric patients remains notable. Recently, Vernier-Massouille et al ⁽⁵⁾ demonstrated the incidence of surgery at 3 years to be 20% and 34% at 5 years for patients with CD. In 2017, Rinawi et al ⁽⁶⁾ identified the cumulative probability for colectomy in children with UC to be 4% at 1 year and 17% at 10 years from diagnosis.

The immunosuppressive effects of biologic therapy have led to the concern for increased postoperative complications ⁽⁷⁾. The management consensus guidelines of the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition recommend avoiding biologic agents during the perioperative period due to these concerns ⁽⁸⁾. There have been conflicting data on postoperative morbidity in patients treated with a preoperative biologic agent who undergo surgical management. Adult studies have been mixed ^(9–28) and in the pediatric population the data have been both limited and conflicting ^(29–32). This study aims to compare early postoperative complications, defined as ≤30 days, in patients with IBD treated with biologic agents to patients treated with nonbiologic therapy before bowel resection.

METHODS

Patient Selection

Following approval from the Wayne State University institutional review board, a retrospective chart review was performed at the Children's Hospital of Michigan, a member of the Detroit Medical Center, and the only hospital providing pediatric surgical care. The review encompassed pediatric patients (≤18 years of age) with IBD who underwent bowel resection between 2001 and 2017 in our 228-bed tertiary facility. Patients were recognized via current procedural terminology, International Classification of Diseases (ICD)-9, and ICD-10 codes. The list was then cross-referenced with an IBD database and patients were excluded if they did not carry an IBD diagnosis. Charts were reviewed to identify documentation of the use of preoperative medications. Patients were excluded if the preoperative or postoperative care was performed at an outside institution or if they had undergone a permanent ileostomy.

Study Design

Data were collected by a single reviewer using a standardized form, including patient's age at diagnosis, age at surgery, sex, ethnicity/race, other comorbid illnesses, disease location, type of medications used, and duration of medications used before surgery, incidence and type of postoperative complications, indication(s) for resection, type of operation (open or laparoscopic, number of stages), length of hospital stay, and additional surgeries required. Patients with moderate to severe IBD were prescribed biologic therapy based on clinical guidelines and discretion of the primary gastroenterologist. Disease location was categorized by findings on endoscopy. As frequently used in literature, the preoperative administration of biologic therapy in this study was defined as use within 12 weeks of surgery ^{(13,14,29,33)}. The type and approach of the operation was chosen by the pediatric surgeon. Early postoperative complications were defined as surgical site infections (SSIs) including superficial SSI, deep incisional SSI, and organ space infection, as well as anastomotic leak, intra-abdominal abscess, and so on occurring <30 days from bowel resection. Complications were compared to biologic-naive patients on nonbiologic medical therapy.

Statistical Analyses

Demographic and surgical variables were compared between those who were treated with preoperative biologic agents to those who were never treated with biologic before surgery through means and standard deviations. Median postoperative length of stay was compared using means. Comparisons across groups were performed via t tests. Median and interquartile ranges (IQRs) were calculated for total duration and last administration of biologic agent and duration of preoperative disease. Univariate analyses were performed to evaluate association between preoperative biologic exposure and surgical outcomes. Statistical analysis was performed using IBM SPSS Statistics (version 24.0, Chicago, IL). A P value <0.05 was deemed statistically significant.

RESULTS

Patient Characteristics

Seventy potential subjects were identified for the study and 62 were included in the final analysis. Eight additional patients were excluded for incomplete medical records. Thirty-seven of the 62 patients were treated with a biologic within 12 weeks of their operation and 25 were biologic naive.

Demographic, disease phenotype, and clinical characteristics of the 62 subjects which met inclusion criteria are summarized in Table 1. Thirty-three patients (53%) were boys and 29 patients (47%) were girls. The patient's race was identified in 60 of the patients: 30 (50%) were Caucasian, 27 (45%) were African American, and 3 (5%) were identified as other. The median ages for diagnosis and resection were 12.43 and 14.55 years, respectively. Forty patients (64%) carried the diagnosis of CD, 21 (34%) had UC, and 1 (2%) had IBD-unclassified. The disease was isolated to the terminal ileum in 9 of the patients (15%). Thirty patients (48%) had large bowel disease. Twenty-three patients (37%) had ileocolonic disease. All patients underwent an intestinal resection performed in 1, 2, or 3 stages at the discretion of the surgeon. A 1-stage procedure, consisting of a resection with primary anastomosis, was chosen for 50% (n = 31) of the patients. This number includes partial small bowel resections, ileocecal resections, and ileal pouch anal anastomosis (IPAA). Patients with UC typically undergo 2-stage procedures, although some require 3 stages based on disease severity. A laparoscopic technique was used in 27 (43%) patients, whereas 35 (56%) patients had an open technique (P = 0.14). The indications for the resection included refractory disease (n = 32), perforating disease (n = 9), stricturing disease (n = 17), and fistulizing disease (n = 8). Thirty-seven (60%) of the patients were treated with IFX (n = 34), adalimumab (n = 2), or vedolizumab (n = 1) before their bowel resection (Table 2). The biologic was administered at a median of 35 days (IQR 19–51) before their operation. Of the 30 patients with data available on total duration of biologic use, the median duration of use was 3 months (IQR 1.5–6) before resection.

TABLE 1: Patient characteristics stratified by use of preoperative biologic agent

TABLE 2: Preoperative biologic agents used by diagnosis

Clinical Outcomes

Girls were more likely to be treated with a biologic than nonbiologic therapy (P = 0.0008). There was not a significant difference between boys treated with biologics and their nonbiologic therapy counterparts (P = 0.75). Caucasian patients were more likely to be treated with biologic agents than nonbiologic therapy (P = 0.0004). There was not a significant difference between the age at diagnosis (P = 0.27) nor age at resection (P = 0.41) between the 2 treatment groups. Subanalysis of preoperative biologic exposure identified patients with refractory disease more likely to have biologic treatment (P = <0.0001), whereas patients with perforating disease were less likely to have been treated with a biologic agent (P = 0.02). There were 4 complications documented within 30 days of the operation, with an overall complication rate of 6.45% (Table 3). Preoperative biologic therapy did not significantly increase the risk of early postoperative complications. There were 2 complications in each of the cohorts, including 2 cases of intra-abdominal abscess, 1 case of abdominal wall abscess, and a single case of pouchitis. There was not a statistically significant difference in postoperative hospital length of stay: 7.17 days in biologic cohort and 8.56 days in nonbiologic group (P = 0.31) (Table 4). When subanalysis was performed based on underlying diagnosis, the UC group that was pretreated with biologics had notably shorter length of stay (LOS) compared to the nonbiologic group (7.58 and 13 days, respectively), although not reaching statistical significance (P = 0.08). Subanalysis was also performed based on surgical technique and identified median LOS for laparoscopic cases as 6 days (IQR 4–8) and open cases as 6.5 days (IQR 5–9) (P = 0.6).

TABLE 3: Details of postoperative complications experienced within 30 days of resection

TABLE 4: Median hospital length of stay by use of preoperative biologic agent and diagnosis

DISCUSSION

In this retrospective study, we analyzed the postoperative complication rate in 62 pediatric patients with IBD who underwent bowel resection in our institution. Children preoperatively treated with a biologic agent were not more likely to experience a postoperative complication than patients not treated with this group of medications.

Previous adult reports have suggested mixed outcomes when evaluating this question. Syed et al ⁽¹²⁾ found an association between IFX use and infections after abdominal surgery in patients with CD. Conversely, Regueiro et al, Krane et al, Norgard et al, Waterman et al, and Myrelid et al ^(24–28) determined no such association. Kunitake et al ⁽¹³⁾, Schluender et al ⁽²⁰⁾, Ferrante et al ⁽¹⁷⁾, Coquet-Reinier et al ⁽²³⁾, and Gainsbury et al ⁽²²⁾ evaluated adults with UC and did not find an association with use of IFX and increased rate of postoperative complication either. However, Selvasekar et al ⁽¹⁸⁾ found there was a substantially increased risk of postoperative pouch-related and infectious complications in patients pretreated with IFX before IPAA. Similarly, Mor et al ⁽¹⁹⁾ noted the odds of early complication to be 3.54 times greater, and odds of sepsis 13.8 times greater, for patients treated with IFX before 2- or 3-stage restorative proctocolectomies. In 2017, Ferrante et al ⁽²¹⁾ evaluated patients with UC who were treated with vedolizumab before colectomy and did not find an increase risk of postoperative infectious complication. Even adult meta-analyses have provided opposing conclusions. Three meta-analyses determined that the use of preoperative IFX was associated with an increased risk of total and infectious complications in adult patients with CD ^(9,14,15). However, Rosenfeld et al ⁽¹⁰⁾ established no significant difference in complication rate, reoperation rate, or 30-day mortality rate between IFX and control groups. Billioud et al ⁽¹⁶⁾ evaluated both CD and UC adult patients and found that preoperative biologic use increases the prevalence of infectious postoperative complications in patients with CD (odds ratio of 1.45) but does not have the same effect in patients with UC.

From the pediatric standpoint, only 2 major studies have been published which evaluate this question in patients with CD ^(29,30), and another 2 examining patients with UC ^(31,32). Zimmerman et al ⁽³⁰⁾ evaluated children with CD who underwent bowel resection with primary anastomosis. Of the 24 patients who had received IFX before the operation, 3 had major complications including anastomotic leak, acute renal failure, and intra-abdominal abscess. In the non-IFX group there were 9 major complications identified. This study also evaluated length of stay and did not identify significant differences between the 2 cohorts. Abbas et al ⁽²⁹⁾ similarly studied children with CD requiring bowel resection and found 7 total postoperative complications among the 73 patients studied, 5 of which were superficial SSIs. Of the 22 patients who had been treated with preoperative IFX, there were 2 postoperative complications. Authors concluded that outcomes of patients stratified by IFX use were not different and that refractory disease was associated with higher preoperative IFX use. In 2012, Kennedy et al ⁽³²⁾ retrospectively evaluated 1-year postoperative complications in children with UC who had or had not received preoperative IFX. The authors found that all complications were more frequent in patients treated with IFX before proctocolectomy with IPAA than the control group, especially small bowel obstructions which were significantly higher (55% vs 7%). Conversely, Schaufler et al ⁽³¹⁾ concluded preoperative exposure to IFX within 90 days of colectomy was not associated with increased risk of postoperative complications in a retrospective chart review of 51 pediatric patients with UC or IBD-unclassified .

In our study, there were a total of 4 complications documented within 30 days of the operation, indicating an overall complication rate of 6.45% . Because of their underlying anemia, malnutrition, and immunosuppression, these patients are thought to be inherently poor surgical candidates ⁽³⁴⁾. Historically, the complication rate for pediatric patients with IBD undergoing bowel resection has been variable. In 2014, Blackburn et al ⁽³⁵⁾ identified the early postoperative complication rate of 22% in a study evaluating 69 children undergoing intra-abdominal surgery for CD. Outcomes were not stratified by preoperative use of biologic agent. Uchida et al ⁽³⁶⁾ identified SSI rate as high as 26% in study of 136 patients. Too few patients had received preoperative biologic therapy, so no conclusion was drawn between administration and SSI occurrence in that study either. Zimmerman et al ⁽³⁰⁾ identified a major complication rate of 13% in IFX group and 9% in the non-IFX group. Patton et al ⁽³⁷⁾ studied outcomes for pediatric patients with UC and reported an early complication rate of 55%, most commonly small intestinal obstruction. Their study did not look at biologic agents. The complications identified in our study were 2 cases of intra-abdominal abscess, a single case of abdominal wall abscess, and a single case of pouchitis.

Historically, surgery may have been delayed if patient received biologic agent and there was subsequent concern for postoperative complications based on previous mixed data. Clinically, this meant that there was sometimes a hesitation to use the biologic agent if the patient was not responding to other medical therapies and the patient appeared to be heading toward surgical intervention. However, this study supports the findings of Zimmerman et al and Abbas et al. Actually, it may be worthwhile to give the biologic agent because it does not increase risk of complications and may delay or prevent surgery ^(35,38,39).

In addition, our data found a notable shorter postoperative hospital LOS in the group that received preoperative biologic therapy compared to the nonbiologic group (7.58 and 13 days, respectively), although this did not reach clinical significance (P = 0.08). The LOS was similar between open and laparoscopic techniques (6.5 and 6 days, respectively) (P = 0.6). LOS is correlated with a considerable burden to healthcare systems in terms of cost and resources ^(40,41). Children undergoing colectomy for UC who have postoperative complications require longer length of stay and higher hospital charges compared to patients without complications ⁽⁴²⁾. A review of the literature suggests costs from postoperative complications in patients with UC ranging from $18,650 per patient over 6 months to $34,714 per patient over a 5-year period ⁽⁴³⁾. It is important to note that surgical complications represent a substantial burden in terms of cost, including reoperations and additional hospital stays, but also related to quality of life.

As with similar studies, this study is limited by its retrospective design and reliance on accurate and complete documentation. In addition, there are inherent confounding effects of the concurrent therapies used in our pediatric patients with IBD. Our study focused on the outcomes described above, and did not investigate body mass index, nutritional markers, or corticosteroid use. Nor did we study the urgency of surgery, training of surgeon, length of bowel resection, operative blood loss, operative time, or rate of late complications. These are variables that could be examined in future studies.

In conclusion, although multiple previous studies suggested conflicting data concerning the association between preoperative biologic use and postoperative complications, this study suggests that a biologic agent may be safe to use in the preoperative period in children with IBD undergoing bowel resection. Thus, surgery should not be delayed and patients may continue appropriate perioperative use of biologic therapy.