International Journal of Colorectal Disease

, Volume 25, Issue 3, pp 401–404

Preoperative infliximab treatment in patients with ulcerative and indeterminate colitis does not increase rate of conversion to emergent and multistep abdominal surgery

Authors

    • MGH Crohn’s & Colitis CenterHarvard Medical School
    • Department of SurgeryMassachusetts General Hospital
  • Hiroko Kunitake
    • MGH Crohn’s & Colitis CenterHarvard Medical School
    • Department of SurgeryMassachusetts General Hospital
  • Paul Shellito
    • MGH Crohn’s & Colitis CenterHarvard Medical School
    • Department of SurgeryMassachusetts General Hospital
  • Richard Hodin
    • MGH Crohn’s & Colitis CenterHarvard Medical School
    • Department of SurgeryMassachusetts General Hospital
Original Article

DOI: 10.1007/s00384-009-0800-5

Cite this article as:
Bordeianou, L., Kunitake, H., Shellito, P. et al. Int J Colorectal Dis (2010) 25: 401. doi:10.1007/s00384-009-0800-5

Abstract

Introduction

A recent study has raised concerns that infliximab treatment, by postpoing surgery for ulcerative and indeterminate colitis patients, may result in a greater need for high-risk emergent or multistep surgical procedures (subtotal colectomies). Our aim was to assess whether infliximab exposure affects rates of subotal colectomy in a large cohort of patients.

Methods

We evaluated 171 consecutive patients with ulcerative or indeterminate colitis who had a total proctocolectomy or a subtotal colectomy between 1993 and 2006 for symptoms of unremitting disease. Forty-four patients (25.7%) received infliximab prior to surgery. We compared the surgical procedures employed on these 44 patients to the surgical procedures employed on the 127 non-infliximab patients, using Fisher’s exact or Student’s t test.

Results

Infliximab exposure did not appear to affect the rate of emergent surgery (4.5% vs 4.4%, p = 0.98), rate of subtotal colectomy (19.2% vs. 18.0%, p = 0.99), or rate of ileoanal J pouch reconstruction (53.8% vs. 62%, p = 0.98). Nor did it affect intraoperative findings of perforation, toxic megacolon, and active disease. The infliximab and non-infliximab cohorts were similar in age, Charlson Comorbidity Index, concomitant steroid use, and albumin levels, although infliximab patients had higher rates of concomitant exposure to 6-mercaptopurine (34.1% vs 16.6%, p = 0.02) and azathioprine (40.9% vs 22.6%, p = 0.02).

Conclusion

Infliximab does not appear to increase rates of emergent surgery or multistep procedures in patients undergoing treatment for ulcerative or indeterminative colitis at our institution.

Keywords

InfliximabSubtotal colectomySurgical complicationsUlcerative colitis

Introduction

Overall, the use of infliximab has diminished the number of surgical procedures in patients with ulcerative and indeterminate colitis [1]. However, for some of these patients, the use of infliximab merely delays—but does not eliminate—the long-term need for surgery. This has led to concerns that infliximab treatment in these patients only serves to postpone surgery to a time when the patient is less clinically fit and that this, in turn, may result in an increased rate of emergent and/or salvage procedures [24]—thereby offsetting the clinical benefits of the treatment regimen.

One of the most troubling recent reports on this topic came from the Cleveland Clinic, where a retrospective review of 523 patients undergoing proctocolectomy determined that their rate of salvage subtotal colectomy was nearly 46% in the 85 patients treated with infliximab, versus 28% in those who were not [2]. These conversion rates are very high, which raises questions whether the emergent and/or salvage procedures are driven by complications relating to infliximab exposure or by some other factor, such as surgeon preference or bias.

In an attempt to address these concerns, we charted the intraoperative course of a large cohort of patients who were treated with abdominal surgery for complications of ulcerative and indeterminate colitis at Massachusetts General Hospital: examining the relationship between infliximab infusion and choice and timing of surgical procedure and the patients’ ultimate intraoperative findings.

Methods

The Institutional Review Board of the Massachusetts General Hospital (MGH) determined that this study was exempt from review.

The patients included in this study were initially identified using the Partners Healthcare System Research Patient Data Registry (RPDR), which can identify patients by specific demographics, diagnoses, laboratory tests, medications, molecular medicine, health history, microbiology, procedures, providers, and/or transfusion services. Using the RPDR Query Tool, we selected those patients who underwent abdominal surgery between January 15, 1993 and December 2007, for one of the following diagnoses: ulcerative colitis (ICD-9:556) or toxic gastroenteritis and colitis (ICD-9:558.2). This initial search identified 267 patients.

We then reviewed the medical record of each patient individually to confirm that the indication for surgery was ulcerative or indeterminate colitis. As a result, we excluded 42 patients who had Clostridium difficile colitis, appendicitis, necrotizing enterocolitis, and no history of IBD. Additional of 22 patients where miscoded and actually had Crohn’s disease. They were also excluded. Finally, 32 patients were excluded because they had a partial colectomy for malignancy in setting of long-standing quiescent UC or a completion proctectomy following a subtotal colectomy at another institution. This resulted in a cohort of 171 patients, of whom 126 (74%) had a definite preoperative diagnosis of ulcerative colitis. The remaining 45 (26.3%) patients had indeterminate colitis suspected to be ulcerative colitis clinically.

We conducted a careful review of the electronic medical record for each of these 171 patients. The medical record included gastroenterology office visit notes, surgical office visit notes, hospital admission and discharge notes, surgical operative notes, pathology reports, endoscopy reports, radiology studies, and admission and discharge medication lists. Patients treated with infliximab (IFX) within 12 weeks prior to surgery were placed in the IFX group. All others were categorized as non-IFX. Patient demographics and preoperative nutritional status were recorded and a cumulative age-adjusted Charlson Comorbidity Score was calculated for each patient [5, 6].

Finally, we reviewed the surgical procedure performed on each patient and abstracted the surgeon’s operative note to determine pertinent intraoperative findings: the presence of perforation, abscess, active disease, and toxic megacolon. We divided these procedures into two categories: optimal procedures and salvage procedures. An ileoanal pouch procedure or total proctocolectomy with ileostomy was considered an optimal procedure, while a subtotal colectomy was considered to be a salvage procedure.

Statistical analysis

Descriptive statistics were examined using frequency and percentage for categorical variables and mean and standard deviation (SD) or median and range for continuous variables. IFX and non-IFX patients were compared using Chi-square and Fisher exact tests and independent t tests, as appropriate. For two-sided tests, a p value less than 0.05 was considered statistically significant.

Results

Of the 171 patients with either ulcerative (N = 126) or indeterminate (N = 45) colitis who underwent attempted total proctocolectomy at MGH, 44 (25.7%) had been treated with IFX prior to surgery. The IFX and non-IFX groups were similar in age, sex, rate of preoperative comorbidities, preoperative nutritional status, rate of concomitant steroids exposure, and proportion of patients with the preoperative diagnosis of indeterminate colitis. However, IFX patients were somewhat more likely to have concomitant exposure to 6-mercaptopurine or azathioprine (Table 1).
Table 1

Comparison of patients based on IFX exposure

Characteristic

IFX (n = 44)

Non-IFX (n = 159)

p Value

Age (years)

31.6 (17.8)

38.8 (19.9)

0.38

Mean (SD)

Gender (%)

52.3

49.7

0.86

Male

CCI mean (SD)

5.7(2.8)

5.9 (2.9)

0.61

Preoperative albumin

3.2 (0.9)

3.1 (0.8)

0.28

Mean (SD)

Concomitant steroids (%)

79.5

77.3

0.75

Concomitant azathioprine (%)

40.9

22.6

0.02a

Concomitant 6-MP/imuran (%)

34.1

16.6

0.02a

Ulcerative colitis (%)

20.6

18.2

0.74

Indeterminate colitis (%)

20.1

18.6

0.86

CCI Charlson Comorbidity Index; 6-MP 6-mercaptopurine; SD standard deviation

aStatistically significant

Rates of emergent surgery were similar between the two groups (4.4% IFX vs. 4.5% IFX, p = 0.98). At the time of surgery, no statistically significant difference was found in the rate of active disease (61.3% IFX vs. 63.6% non-IFX, p = 0.86), toxic megacolon (2.27% IFX vs. 1.1% non-IFX, p = 0.0.48), or perforation (2.3% IFX vs. 3.8% non-IFX, p = 0.61). Ultimately, only 32 patients (18.7%) had a subtotal colectomy instead of a proctocolectomy, and the rate of subtotal colectomy was similar in both groups (19.2% IFX vs. 18.0 non-IFX, p = 0.99). One hundred thirty-nine patients had a total proctocolectomy, and 106 of these patients (76.3%) had an IPAA. Overall, the rate of IPAA was similar between the groups (Fig. 1). The remainder had a total proctocolectomy with a permanent-end ileostomy as decided upon preoperatively based on continence, age, and personal preference.
https://static-content.springer.com/image/art%3A10.1007%2Fs00384-009-0800-5/MediaObjects/384_2009_800_Fig1_HTML.gif
Fig. 1

Rates of subtotal colectomy and restorative ileoanal pouch reconstruction based on infliximab exposure in patients with ulcerative and indeterminate colitis. Abbreviations: IPAA (ileal pouch-anal anastomosis)

In this population, we did not find a statistically significant difference in the rates of deaths (2.2% IFX vs 0.5% non-IFX, p = 0.27), anastomotic or rectal stump leak (4% IFX vs 3.3% non-IFX, p = 0.85), or intra-abdominal abscess (0.1% IFX vs 3.3% non-IFX, p = 0.2) based on infliximab exposure.

Discussion

Infliximab is a chimeric monoclonal antibody that targets tumor necrosis factor [7]. In 2006, the FDA approved its use in treating ulcerative colitis. Since then, it is thought to have changed the clinical course of this disease [8]. Studies report that infliximab appears to induce and maintain remission in ulcerative colitis patients [8]. It has also demonstrated success in reducing the need for corticosteroids and in healing colonic mucosa [8]. Other studies postulate that infliximab may also reduce the number of hospitalizations and surgeries in patients with ulcerative colitis, though this assertion is based on data extrapolated from patients with Crohn’s or colitis [1, 9].

Nonetheless, even with infliximab treatment, surgery will be required in approximately 30–40% of patients with ulcerative colitis at some point during their lifetime. This has led to concerns about surgical outcomes in patients who have been treated preoperatively with infliximab [10]. One of the most troubling recent reports on this topic, by Mor et al., involved a retrospective review of 523 patients undergoing proctocolectomy. The study determined that the rate of salvage subtotal colectomy was nearly 46% in the 85 patients who had been treated with infliximab, versus 28% in those who were not [2]. These conversion rates are very high, which raises questions whether the emergent and/or salvage procedures are driven by complications relating to infliximab exposure [2].

Our study, on the contrary, did not find a statistically significant association between infliximab treatment and rates of emergent operations. We also found no statistically significant association between infliximab treatment and relevant intraoperative findings, such as active disease, toxic megacolon or perforation, or in the rates of multistep or salvage procedures. Moreover, our overall rate of conversion to subtotal colectomy—under 20%—was substantially lower than the rates reported by Mor et al., both in the control group and in the infliximab group.

Interestingly, our 18% conversion rate—while much lower than reported by Mor et al.—was consistent with conversion rates previously reported by other authors, both in the IFX and control groups [3]. For example, Selvasekar et al., studying 341 patients treated with proctocolectomy at a different IBD referral center, reported a conversion rate of 13% in the IFX group versus 17% in the non-IFX group. As in our study, the Selvasekar conversion rates were not statistically different between the groups, and the rates of important postoperative complications such as pelvic abscess and anastomotic leaks were statistically indistinguishable (this study did report an increase in postoperative wound infection rates in the infliximab cohort, and, as a result, a statistically significant increase in overall complication rates) [3, 11].

Another study by Schluender et al. described 151 patients with fulminant ulcerative colitis, of whom 39 required conversion to a subtotal colectomy. In eight of the 39 patients (21%), the conversion was due to the presence of toxic megacolon or intestinal perforation. In 24 patients (62%), conversion was secondary to poor clinical condition and fragile tissues precluding a safe stretch of the pouch to the dentate line [11]. Interestingly, their reported rate of conversion to a subtotal colectomy in the non-infliximab patients (28%) was higher than the rate observed by us and Selvasekar et al. and was similar to the non-infliximab rates reported by Mor et al. However, as in our study, the Schluender study did not find a statistically significant increase in the rate of subtotal colectomy or the rate of postoperative complications in the infliximab cohort [11].

At bottom, our study—like the earlier studies of Selvasekar and Schluender—appears to show significantly lower rates of conversion than Mor et al. with respect to salvage three-step procedures following infliximab exposure. This suggests the possibility that different IBD centers may have different thresholds of conversion to subtotal colectomy and that these thresholds are not necessarily data driven. Clearly, further prospective studies are needed to sort out these conversion rates in a definitive fashion. It is possible that any meaningful answer will require that surgeons be blinded to a patient’s infliximab exposure. In the meantime, however, our data suggest that infliximab exposure should not necessarily be taken into account—in isolation from a patient’s overall clinical presentation—while making a decision to convert to a subtotal colectomy.

Conclusions

In our study of 171 consecutive patients who underwent abdominal surgery for complications of either ulcerative or indeterminate colitis, preoperative exposure to infliximab did not appear to increase the rate of emergent surgery or the rates of surgical salvage and multistep procedures.

Disclosures

None.

Copyright information

© Springer-Verlag 2009