Article Text

PDF

Original article
Preoperative biological therapy and short-term outcomes of abdominal surgery in patients with inflammatory bowel disease
  1. Matti Waterman1,2,
  2. Wei Xu3,4,
  3. Amreen Dinani1,2,
  4. A Hillary Steinhart1,2,3,4,
  5. Kenneth Croitoru1,2,4,
  6. Geoffrey C Nguyen1,2,4,
  7. Robin S McLeod2,4,5,
  8. Gordon R Greenberg1,2,4,
  9. Zane Cohen2,4,5,
  10. Mark S Silverberg1,2,4
  1. 1Division of Gastroenterology, Mount Sinai Hospital, Toronto, Ontario, Canada
  2. 2Zane Cohen Centre for Digestive Diseases, Toronto, Ontario, Canada
  3. 3Dalla Lana School of Public Health, Toronto, Ontario, Canada
  4. 4University of Toronto, Toronto, Ontario, Canada
  5. 5Department of Surgery, Mount Sinai Hospital, Toronto, Ontario, Canada
  1. Correspondence to Dr Mark S Silverberg, Division of Gastroenterology, Mount Sinai Hospital, 600 University Ave., Room 441, Toronto, ON, Canada M5G 1X5; msilverberg{at}mtsinai.on.ca

Abstract

Objective Previous investigations of short-term outcomes after preoperative exposure to biological therapy in inflammatory bowel disease (IBD) were conflicting. The authors aimed to assess postoperative outcomes in patients who underwent abdominal surgery with recent exposure to anti-tumour necrosis factor therapy.

Design A retrospective case-control study with detailed matching was performed for subjects with IBD with and without exposure to biologics within 180 days of abdominal surgery. Postoperative outcomes were compared between the groups.

Results 473 procedures were reviewed consisting of 195 patients with exposure to biologics and 278 matched controls. There were no significant differences in most postoperative outcomes such as: length of stay, fever (≥38.5°C), urinary tract infection, pneumonia, bacteraemia, readmission, reoperations and mortality. On univariate analysis, procedures on biologics had more wound infections compared with controls (19% vs 11%; p=0.008), but this was not significant in multivariate analysis. Concomitant therapy with biologics and thiopurines was associated with increased frequencies of urinary tract infections (p=0.0007) and wound infections (p=0.0045). Operations performed ≤14 days from last biologic dose had similar rates of infections and other outcomes when compared with those performed within 15–30 days or 31–180 days. Patients with detectable preoperative infliximab levels had similar rates of wound infection compared with those with undetectable levels (3/10 vs 0/9; p=0.21).

Conclusion Preoperative treatment with TNF-α antagonists in patients with IBD is not associated with most early postoperative complications. A shorter time interval from last biological dose is not associated with increased postoperative complications. In most cases, surgery should not be delayed, and appropriate biological therapy may be continued perioperatively.

  • Biological therapy
  • postoperative complications
  • inflammatory bowel disease
  • Crohn's disease
  • ulcerative colitis
  • genetic polymorphisms
  • infliximab
  • IBD
  • clinical trials
  • IBD clinical
  • Helicobacter pylori
  • mucosal immunology
  • health economics
  • health outcomes
  • health disparities
  • health service research
  • Crohns colitis
  • 5-aminosalicylic Acid (5-ASA)
  • inflammatory bowel disorders
  • genotype
  • genetics
  • genetic testing

Statistics from Altmetric.com

Significance of this study

What is already known on this subject?

  • Patients on biological therapy such as infliximab or adalimumab are at increased risk for infections.

  • Some studies in patients with inflammatory bowel disease who had abdominal operation with preoperative exposure to biologics suggested that they were at increased risk for short-term postoperative complications, while others did not find such an association.

  • Many studies that investigated complications of pelvic pouch reconstruction in patients with ulcerative colitis and preoperative exposure to infliximab found increased rates of pelvic infections and anastomotic leaks compared with unexposed controls.

What are the new findings?

  • In the largest series to date, 473 abdominal operations on patients with inflammatory bowel disease with 195/473 patients exposed to biologics, most of the postoperative infections and short-term complication rates were no different than matched controls.

  • Preoperative combination therapy with biologics and thiopurines was significantly associated with urinary tract infections and operative wound infection and possibly bacteraemia and increased antibiotics requirement.

  • Patients exposed to biologics within 14 days of surgery had similar complication rates compared with patients exposed to biologics 15–30 days and 31–180 days before surgery.

  • In a small subset of patients, those who had surgery with detectable serum infliximab levels had similar short-term complications compared with patients exposed to biologics with undetectable serum infliximab levels.

Significance of this study

How might it impact on clinical practice in the foreseeable future?

  • From these data, it appears that surgery should not be delayed in order to increase the time interval from the last dose of biological therapy to surgery.

  • Appropriate biological therapy may be continued perioperatively.

  • Combination immunosuppression of biologics with thiopurines may increase the risk of wound infections, urinary tract infections and bacteraemia postoperatively.

Introduction

Biological therapy for inflammatory bowel disease (IBD) has gained an ever-expanding role in the management of both Crohn's disease (CD) and ulcerative colitis (UC). It has been estimated that as of 2007, the world-wide use of infliximab has exceeded one million patients (who mostly had rheumatoid arthritis) and some 370 000 were prescribed adalimumab with smaller numbers receiving certolizumab and other agents.1 However, surgery is still an important mode of treatment for both forms of IBD. In fact, some recent population-based studies have shown that despite the use of highly efficacious biological treatments in UC and CD, surgery rates are still quite high. For example, a recent population-based study from Manitoba, Canada, revealed that among patients with IBD who were exposed to infliximab, surgery rates after 2 years were still approximately 30%.2 Another study from Northern California that investigated time-trends in therapies from 1998 to 2005 showed that although the colectomy rate in UC did diminish, surgery rates in CD have not declined.3 While biological therapy in IBD is highly efficacious, its use is associated with well-documented adverse events. For example, biological therapy is associated with higher rates of bacterial and opportunistic infections, especially in the setting of combination with corticosteroids (CS), such as urinary tract infections (UTIs), respiratory tract infections, latent tuberculosis reactivation, candidiasis and histoplasmosis among others.4 A recent review estimated that the RR for serious infections associated with all biologics is around 2–3-fold higher.5 As a result, concerns have been raised regarding the safety of biologics use in the setting of abdominal surgery.

Several studies have reported the outcomes of surgery in the setting of biologics use in IBD with conflicting results.6–12 In this study, subjects with IBD undergoing abdominal surgery with recent exposure to infliximab or adalimumab were evaluated to assess for 30 day surgical outcomes in a large surgical tertiary care referral centre.

Methods

Patient selection

The IBD Surgical Database at Mount Sinai Hospital in Toronto, Canada contains prospectively collected demographic and phenotypic data for all patients with IBD who underwent surgery since 1989. The dates and detailed description of the operative procedures are documented. The data on all IBD-affected individuals who underwent abdominal surgery during the biologic era from January 2000 to June 2010 were reviewed. Those individuals with documented preoperative exposure to infliximab and/or adalimumab within 180 days from the date of surgery were selected. Anti-tumour necrosis factor (anti-TNF) exposure was verified by reviewing both the inpatient and outpatient charts of each individual. Patients without adequate clinical records documenting the 30 day postoperative clinical outcomes were excluded. Date of the last preoperative dose of the biological agent was documented. For time-dependent analysis, the subjects were classified into three groups: exposure to biologics within ≤14 days of the surgery, exposure within 15–30 days and exposure within 31–180 days.

Data collection

For each procedure, we collected the following baseline data: gender, age at diagnosis, date of diagnosis, IBD subtype, systemic CS exposure within 30 days from surgery, prednisone exposure (≥20 mg/day) within 7 days of surgery, azathioprine/6-mercaptopurine (6-MP) exposure within 3 months of surgery, date of hospitalisation, date of surgery, age at surgery, main operative procedure, secondary operative procedure performed during the same operation, indication for surgery, laparoscopic approach, CD behaviour and anatomical location at the time of surgery (according to the Montreal classification).13

Outcomes measured

Postoperative length of stay (LOS), fever ≥38.5°C, UTI, pneumonia, other serious infections, septic shock, bacteraemia, requirement for systemic postoperative antibiotics, wound infection, anastomotic leak, reoperation, need for percutaneous abdominal abscess drainage, readmission, ileus/small bowel obstruction, intra-abdominal abscess and mortality.

Matched controls selection

Subjects were chosen from the same database as described above. Non-exposure to biologics within 180 days before the date of operation was verified. Only patients who had complete documentation as described above were reviewed. Care was taken to ensure that each control procedure was matched to only one case procedure. For each procedure on biologics, we carefully matched 1–2 control procedures using the following four criteria:

  1. Main operative procedure (eg, small bowel resection, subtotal colectomy (STC), etc)

  2. IBD subtype (CD vs UC/IBD unclassified)

  3. Exposure to preoperative prednisone ≥20 mg/day or equivalent CS within 7 days

  4. Patient age at surgery (<50 or ≥50 years).

Serum infliximab determination

A subset of patients with UC, with exposure to infliximab before surgery had serum infliximab levels and antibodies to infliximab (ATI) evaluated within 2 months from surgery. Serum infliximab and ATI (Prometheus Laboratories, Inc, San Diego, California, USA) were assessed, as described previously.14 Serum infliximab was measured by a microplate ELISA with a cut-off value of 1.4 μg/ml; serum concentrations below the cut-off value are reported as undetectable. Antibodies against infliximab were measured by a microplate ELISA based on the double-antigen format with a cut-off value of 1.69 μg/ml. Antibodies against infliximab are reported as positive when the concentration exceeds 1.69 μg/ml and the serum infliximab concentration was <1.4 μg/ml. Because infliximab interferes with the antibody against infliximab assay, antibody formation cannot be determined in the presence of detectable (>1.4 μg/ml) infliximab.

Statistical analysis

Descriptive statistics were reported as median and range for continuous variables, frequencies and proportions for categorical variables. A comparison was made between all procedures performed on patients with IBD (combining both UC and CD) with preoperative biologics, and their matched controls for each of the outcomes was performed using χ2 test or Fisher exact test. Wilcoxon rank-sum test was applied to compare for continuous variables. A combined end-point of postoperative leak, need for reoperation or percutaneous drainage among the two groups was separately evaluated. Multivariate analysis was applied using logistic regression adjusting for age at surgery, IBD subtype, main procedure performed, CS use, body mass index (BMI), and azathioprine/6-MP use.

The outcomes in the subgroups: small bowel resections, subtotal colectomy and all others were separately assessed as were the complication rates among patients on infliximab versus adalimumab. Finally, the complication rates among patients with procedures performed within 14 days of last biologics dose, within 15–30 days and procedures performed between 31 and 180 days from the last biologics dose were compared. In a small subset of patients with UC who had available preoperative infliximab levels, we compared complication rates between the group with detectable infliximab levels (>1.4 μg/ml) and the group with undetectable infliximab levels (serum infliximab <1.4 μg/ml, with or without ATI formation). Two-sided tests were applied. Results were considered significant if p value ≤0.05. Statistical analyses were performed using SAS system and user's guide 9.2 (SAS Institute).

Results

Of the 4075 abdominal operations performed at Mount Sinai Hospital on patients with IBD between January 2000 to June 2010, a total of 195 procedures were identified as meeting the entry criteria with exposure to infliximab and/or adalimumab within 180 days, and had complete records to review. One to two matched controls were picked from those unexposed to anti-TNF therapy for a total of 278 controls. A total of 473 procedures were reviewed and compared for 30 day postoperative complication rates. The baseline clinical characteristics and demographic data are depicted in table 1.

Table 1

Clinical characteristics of study population

As expected by the strict case-control matching, there were no significant differences between the groups in the age at surgery, IBD subtype distribution, operation type and proportion on systemic CS. The two groups were also similar in the distribution of elective procedures (defined as operation at the day of admission) and procedures done for inpatients (defined as at least one preoperative hospitalisation day). Moreover, the groups were similar in their gender distribution and median albumin levels and BMI. The proportions of patients with BMI extremes (≤19 or ≥ 30 kg/m2) were also similar. There were also no significant differences between the groups in the indications for surgery, IBD anatomical locations (except for perianal CD) at surgery and CD behaviour. However, patients in the group that received biologics before surgery were more frequently prescribed azathioprine or 6-MP (36% vs 20%, respectively, p=0.0001). The complication rates and surgical outcomes of the group on biologics and unexposed controls are outlined in table 2.

Table 2

Complication rates and outcomes in procedures with preoperative biological therapy and unexposed controls

In the univariate analysis, the LOS, rates of postoperative fever ≥38.5°C, ileus/small bowel obstruction and readmission were similar. Most of the infectious complications (UTI, pneumonia, bacteraemia, septic shock) and the antimicrobial therapy requirements were similar among the groups. However, procedures on biologics were more likely to develop wound infections (p=0.008) by univariate analysis. The dehiscence rates in procedures on biologics and controls (1% vs 0.7%, respectively) were similar. The individual rates of anastomotic leaks, need for reoperation and need for percutaneous abdominal abscess drainage were similar; however, the combined endpoint for all these procedures did occur more frequently in those on biologics (20% vs 13%, respectively, p=0.03). The mortality rates were similar (2/195 vs 1/278, respectively, p=0.37). Two patients on biologics and one in the control group died from complications of sepsis. After multivariate analysis adjusting for age, disease type, main procedure type, preoperative prednisone requirements ≥20 mg/day, BMI, and preoperative azathioprine/6-MP, there was no significant difference between the two groups for any of the outcomes.

The rate of postoperative complications in patients receiving combinations of immunosuppressive therapy was assessed. Thirty-three procedures on combination of biologics and prednisone were compared with matched control procedures (n=43) matched by the main operative procedure, IBD subtype and age group at surgery. The groups had no significant differences in the median age at surgery, disease subtype distribution, BMI and surgery subtype distribution, rates of concomitant thiopurine therapy, but the group on biologics had lower mean albumin levels (28 vs 37 g/l, respectively). There were no differences in the postoperative LOS (8 days in both), re-hospitalisation (12/33 vs 10/43), postoperative fever (4/33 vs 2/46), UTI (4/33 vs 4/43), pneumonia (1/33 vs 0/43), deep wound infection (6/33 vs 6/43), sepsis (1/33 vs 0/43), postoperative ileus (6/33 vs 13/43), anastomotic leaks (0/33 vs 3/43), postoperative antibiotics requirements (6/33 vs 6/43) and need for reoperation (2/33 vs 2/43), abscess drainage (3/33 vs 3/43), and there were no cases of death.

Procedures on biologics in combination with azathioprine/6-MP (n=70) were compared with their matched control procedure matched by the main operative procedure, IBD subtype, CS exposure and age group at surgery (n=96). The groups had no significant differences in the median age at surgery, disease subtype distribution, BMI, serum albumin levels and surgery subtype distribution, but only 13% of the matched control group on biologics had concomitant thiopurine therapy. There were increased frequencies of the respective rates of postoperative UTI (8/70 vs 0/96, p=0.0007), deep wound infections (17/70 vs 8/96, p=0.0045), bacteraemia (5/70 vs 1/96, p=0.04) and postoperative antibiotics requirements (17/70 vs 10/96, p=0.02). There were no significant differences in the rates of death (1/70 vs 0/96), need for reoperation (4/70 vs 3/96), septic shock (2/70 vs 1/96), percutaneous abscess drainage (4/70 vs 3/96), postoperative ileus (15/70 vs 26/96), pneumonia (3/70 vs 2/96), postoperative fever (14/70 vs 11/96) and need for re-hospitalisation (14/70 vs 11/96). Ten procedures on triple immunosupression therapy (biologics, thiopurines and CS) were compared with 13 matched controls (CS and thiopurines). While this comparison was clearly under-powered, there were no differences in the postoperative LOS (8 days vs 10 days), re-hospitalisation (4/10 vs 1/13), postoperative fever 2/10 vs 0/13), UTI (2/10 vs 0/13), pneumonia (1/10 vs 0/13), deep wound infection (3/10 vs 2/13), bacteraemia (1/10 vs 0/13), postoperative ileus (1/10 vs 3/13), anastomotic leaks (0/10 vs 0/13), postoperative antibiotics requirements (2/10 vs 0/13) and need for reoperation (1/10 vs 1/13), abscess drainage (2/10 vs 0/13), and there were no cases of death.

We then compared complication rates and operative outcomes in patients who underwent small bowel resections (mostly for CD). The group with preoperative biologics and unexposed controls were similar in their gender distribution, age at procedure, proportion on CS, IBD subtype distribution, preoperative azathioprine/6-MP, BMI and albumin levels. The complication rates and 30-day outcomes were mostly similar (table 3) except for bacteraemia, which occurred significantly more frequently among the group exposed to preoperative biologics (3/69 vs 0/94, respectively; p=0.04) and postoperative antibiotics requirements (14/69 vs 9/91; p=0.05).

Table 3

Complication rates and outcomes in SB resections with preoperative biological therapy and unexposed controls

One hundred and eight STC procedures were evaluated mostly for UC (94/108). The baseline characteristics were similar in the group on biologics (n=51) and controls (n=57) including age at procedure, gender, preoperative CS and azathioprine/6-MP, IBD subtype distribution, median albumin level and BMI, gender distribution and UC disease extent. Preoperative biological therapy was not associated with increased rates of infectious complications or poor healing (table 4).

Table 4

Complication rates and outcomes in STC and ileostomy in preoperative biological therapy and unexposed controls

Looking only at procedures performed with preoperative biological therapy, we compared the postoperative complications in procedures exposed to infliximab (n=156) and to adalimumab (n=33). We found no differences in the rates of any of the infectious complications and any of the other complications including LOS, reoperation, readmission, anastomotic leak, poor wound healing, postoperative fever and antibiotic requirement.

The influence of the time interval between the last biological exposure and surgery was also investigated. Overall, the median time interval between the last biological dose and surgery was 27 days. There were 56 procedures performed within 14 days from the last dose, 48 within 15–30 days and 89 within 31–180 days (median 50 days) from the last dose. The baseline characteristics including age at procedure, gender distribution, IBD subtype distribution, type of biological therapy used, proportion on azathioprine/6-MP, median albumin and BMI were all similar between the 3 time-interval groups. However, the group in which biologics were stopped >30 days before surgery were less frequently on prednisone ≥20 mg/day before their surgery than the groups where biologics were stopped within 15–30 days and <15 days before surgery (10% vs 26% and 21%, respectively; p=0.05). As shown in table 5, there were no differences in the rates of any of the postoperative complications between the three groups.

Table 5

Time-interval from last biological dose and surgical outcomes

The association between serum infliximab levels and short-term complications was also assessed. Nineteen procedures were carried out in patients with UC, where preoperative infliximab and ATI levels were tested within 60 days of surgery. Ten patients had detectable serum levels of infliximab and nine patients had undetectable infliximab.

The groups were similar in age at procedure, gender distribution and proportion on prednisone ≥20 mg/day. The group with detectable infliximab levels had a shorter time interval between the last dose of infliximab and surgery (median 18 days vs 34 days; p=0.03). There were no differences in the overall infectious complication rates between the two groups. Wound infections were more frequent in the group with detectable infliximab although the results were not statistically significant (3/10 vs 0/9, respectively; p=0.21).

Discussion

With increasing use of biologic therapy, more patients with IBD may require surgery after recent exposure to these agents. This may be as a result of failure of biological therapy or, in the case of CD, where ongoing therapy is continued peri-operatively, to treat residual disease. Additionally, there are no standard recommendations regarding whether it is advisable to modify biological treatment schedules in order to minimise exposure prior to surgery. It is therefore of interest to evaluate whether abdominal surgery performed after exposure to biological therapy results in an increase in postoperative complications. Prior studies on this topic have been relatively small and with conflicting results. A retrospective study from the Mayo clinic evaluating patients with CD (52 on infliximab) found an overall rate of 19% for septic complications but did not find an association between such complications and peri-operative infliximab, thiopurines or CS.6 Another study from Leuven, Belgium (40 on infliximab) reported similar rates of minor complications (15.0% vs 12.8%), major complications (12.5% vs 7.7%) and mean hospital stay between patients exposed to infliximab and controls, respectively.10 The largest study to date, performed at the Massachusetts General Hospital in IBD (101 on infliximab) found similar rates of death (2% vs 0.3%), anastomotic leaks (3.0% vs 2.9%), infections (5.97% vs 10.1%), thrombotic complications (3.6% vs 3.0%) and ileus/small bowel obstructions7 compared with controls. Another study from the Cleveland Clinic on CD (60 on infliximab) did show a significant association between preoperative biological exposure and postoperative readmission (p=0.045), sepsis (p=0.027) and intra-abdominal abscess (p=0.005).8 In patients with UC undergoing restorative operations after proctocolectomy, preoperative biological therapy has generally been associated with increased rates of postoperative complications. For example, patients on infliximab undergoing ileal-pouch anal anastomosis at the Cleveland Clinic (85 on infliximab) had early complication rates 3.5 times higher than that of controls (p=0.004; 95% CI, 1.51 to 8.31), and the odds of sepsis were 13.8 times greater (p=0.011; 95% CI 1.82 to 105).9 A recently published meta-analysis of five studies evaluating a total of 132 subjects on infliximab undergoing ileal-pouch anal anastomosis found that its use was associated with an increase in total postoperative complications (OR 1.80, 95% CI 1.12 to 2.87) but not with short-term infectious complications (OR 2.24, 95% CI 0.63 to 7.95) or non-infectious complications (OR 0.85, 95% CI 0.50 to 1.45).11 A recent review concluded that specific recommendations regarding the need to stop or modify preoperative biological therapy in patients with IBD with planned surgery could not be made.12 It is also unknown whether the time interval between exposure to biological agents and surgery may influence the predisposition to postoperative complications.

Our study is the largest to date with 195 anti-TNF exposed procedures evaluated, and compares short-term outcomes of abdominal surgery in patients with IBD with preoperative biological therapy to matched, unexposed controls. Preoperative biological therapy was not associated with increased rates of 30 day postoperative infectious complications, prolonged hospital stay, readmission, reoperation or ‘poor’ healing of surgical anastomoses. However, combination therapy of thiopurines and biologics was associated with significantly increased rates of postoperative infections including UTI, wound infections, bacteraemia and postoperative antibiotics. Similar to other studies, preoperative infliximab in our cohort of small bowel resections was not associated with anastomotic leaks or intra-abdominal abscess formation.6 In patients who had STC on biologics we did not find increased rates of any of the postoperative complication (including infections and complications of poor healing). Previous series that investigated the influence of preoperative biological therapy in patients with UC concentrated mainly on pouch procedures, and the majority suggest that preoperative infliximab exposure increased the complication rate.9 ,15 However, studies that looked specifically at STC were too small to arrive at definite conclusions. In one series, two cases of STC with preoperative infliximab were compared with 38 without infliximab exposure16 and while 1/2 on infliximab had complications compared with 7/37 in the controls, the sample size is clearly too small to draw any conclusions. Another study performed in patients with UC found that total colectomy with creation of pelvic pouch with preoperative infliximab was associated with pelvic complications and infectious complications15; however, the complication rates in patients who initially had STC (n=49, 6 on infliximab) were not reported. In accordance with our findings, a study that investigated the outcomes of minimally invasive colectomy in patients with CD did not find preoperative infliximab to be associated with increased complications.17 Similarly, a study from Belgium where 22 patients with UC who underwent restorative proctocolectomy on infliximab were compared with 119 UC who had surgery without exposure to infliximab showed that the postoperative complication rates were not increased with infliximab.18

Whether biological therapy administered very shortly before an operation confers higher risk for postoperative infections and poor healing has not been adequately investigated in any of the previous reports. We have shown for the first time that a shorter time interval between the last dose of biologics and surgery is not associated with increased rates of complications. Specifically, biologics administered within 2 weeks of surgery were not associated with increased postoperative complications compared with procedures performed within 15–30 days from last dose and procedures performed more than 30 days and up to 6 months (a median of 50 days) from the last biologics dose. We have also shown that patients with UC who had surgery with detectable infliximab in their serum are no more likely to have complications compared with patients with undetectable infliximab. We did find a slightly higher rate of wound infections in the group with detectable infliximab, but the results were not statistically different (3/10 vs 0/9, respectively). Though these data support the assertion that dosage of anti-TNF closer to the time of surgery does not result in an overall increase in complications, we cannot exclude that infliximab trough levels are related to an increase in postoperative complications due to the small sample size.

In our cohort, combination therapy with biologics and thiopurines was associated with increased rates of infections in the early postoperative period. Previous studies did not find azathioprine to be associated with early postoperative infections in patients with IBD.6 ,19 However, in another study by Kunitake et al, patients with IBD who underwent surgery on preoperative biologicals had an increased frequency of postoperative infections. This may have been due to more frequent concomitant therapy with thiopurines compared with controls.7

Our study has certain limitations. As a single-centre study, our results reflect the preoperative and postoperative care typical for a large IBD tertiary-care surgical centre and may not necessarily apply to other centres. Despite this limitation, our study has generally reflected the findings of other studies as to the safety of preoperative biological therapy. The series of patients with one-stage pelvic pouch procedures after preoperative anti-TNF therapy was small (n=11) and, thus, an adequately powered comparative analysis with and without biologics could not be performed. This surgical approach reflects the usual standard of practice in our institution of performing two- or three-stage procedures for patients with steroid-refractory moderately severe to severe UC who require colectomy. Additionally, our study is retrospective, and though every effort was made to meticulously review the postoperative period, under-reporting of complications and inaccuracies in the documentation of the details of biological therapy may still have occurred. Although our study reports on the largest series to date in total procedures on biologicals versus controls, and in the number of patients undergoing STC and small bowel resections, it may still be underpowered to evaluate some of the rare complications. We have compiled data on both UC and CD for our analysis. However, our multivariate analysis corrected for disease subtype among other factors. Furthermore, we performed subgroup analysis by disease diagnosis on the groups that had preoperative biologicals, where no differences in the complication rates were observed in patients with UC and CD. In the subgroup analysis, we investigated the complication rates after segmental small bowel resection (mostly patients with CD) and STC (mostly patients with UC). This provided further insight to the frequency of postoperative complications by disease subtype. Our data suggest that factors such as nutritional status, indication for surgery, urgency, postoperative care and particularly the procedure type are those that have the maximum impact on postoperative outcomes rather than the specific type of IBD.

Confounding variables that may predispose to infections and poor healing is also an issue of concern. However, the careful case-control matching by age at procedure, steroid exposure, operation type and IBD subtype, and multivariate analysis adjusting for potential confounders such as BMI, azathioprine/6-MP exposure, was performed to mitigate these factors. Another potential limiting factor for this study is the large percentage of patients who were hospitalised prior to surgery. At a rate of 25%, this is a significant number and may have impacted some of the postoperative findings (high wound infection rates in both groups). This rate likely reflects the indications for surgery where acute obstruction, penetrating complication and perianal disease were the indication in some 27.2% of the surgeries with preoperative biologics exposure. It should be noted that in many cases, biological therapy is indicated in patients with acute complications of IBD such as intravenous CS failure, penetrating complications of CD and perianal disease.

In summary, in a large cohort of patients with IBD who underwent abdominal surgery, preoperative biological therapy, regardless of the time interval from surgery, was not associated with postoperative infectious complications, anastomostic leaks and other peri-operative complications. In a small subgroup of patients, we did not find a relationship between the presence or absence of serum infliximab before surgery and rates of complications. Our findings do not support the practice of some physicians to delay surgery until anti-TNF-α agents are ‘washed out’.12 Based on these results, we do not recommend altering anti-TNF therapy or the scheduling of surgery to prolong the time interval to operation.

Acknowledgments

The authors would like to thank Brenda O'Connor and Harden Huang for their assistance in maintaining the Mount Sinai IBD Surgical Database and in case finding and matching. The authors are also grateful to Joanne Stempak and Yifat Moreh-Waterman for assistance in data compilation and analysis.

References

View Abstract

Footnotes

  • Funding MW was supported by a Fellowship Award from the Canadian Institute of Health Research, Canadian Association of Gastroenterology and the Crohn's and Colitis Foundation of Canada and by the Joseph Lebovic Charitable Foundation. MSS is supported in part by the Gale and Graham Wright Research Chair in Digestive Disease. Partial funding for this study was provided by the Zane Cohen Centre for Digestive Diseases.

  • Competing interests None.

  • Ethics approval Ethics approval was granted by Mount Sinai Hospital Research Ethics Board, Toronto, ON, Canada

  • Provenance and peer review Not commissioned; internally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles

  • Digest
    Emad El-Omar William Grady Alexander Gerbes