The Impact of Preoperative Serum Anti-TNFα Therapy Levels on Early Postoperative Outcomes in Inflammatory Bowel Disease Surgery

    loading  Checking for direct PDF access through Ovid


To the Editor:
We have read with great interest the study presented by Lau et al1 on the role of preoperative serum anti-TNFα levels in early postoperative outcomes among inflammatory bowel disease (IBD) patients. The authors conclude that increased serum preoperative anti-TNF levels are associated with adverse postoperative outcomes in Crohn's disease but not in ulcerative colitis. Given previous conflicting results as to whether preoperative exposure to anti-TNFs impacts on postoperative surgical complications, including infections and sepsis, this study is timely.2–6 Anti-TNF levels to optimize clinical responses are now part of the clinical practice in most centers and this study goes one step further by using more precise measures on which to test surgical outcomes. However, we wish to raise few points that may influence the interpretation of the results.
Only 21% of patients undergoing surgery during the study period were included, and we would be interested in further information about the remaining 79% of surgical patients, namely disease phenotype, severity, and complications, to determine the potential impact of selection bias.
In the final analysis, 217 included patients seem adequate to provide interpretable results. However, when the population heterogeneity is considered both in terms of primary disease (Crohn's disease and ulcerative colitis) and diversity of surgical procedures, the number of patients in each group diminishes, raising the possibility of a type I error for the subgroup results.
We noted that the group labeled as “undetectable” preoperative anti-TNFα levels also includes those not receiving anti-TNF treatment preoperatively. It's unclear why both groups were included. This distinction is important because it may reflect different clinical state (eg, refractory disease, less aggressive disease, and declined or adverse effect to anti-TNFs, which may affect outcome).
Moreover, several confounders known to be associated with poor outcomes7 were omitted. Therefore, the conclusion that high levels were associated with poorer outcomes may reflect more disease severity, emergency instead of elective surgery,8 previous surgery, and premorbid status. Most are not considered in multivariate analyses.
As this was a retrospective study, the serum anti-TNFα levels results were analyzed from stored frozen serum samples; however, there is no data about the stability of serum concentrations of anti-TNFs over long periods. Although it is reasonable to presume sample stability for short intervals, acceptance of stability for 13 years could have easily been supported by comparison of serum anti-TNFα concentration at different time intervals along the 13-year period. A similar prevalence of undetectable levels in the 13-year sample versus 1-year sample would assuage the concern. The cut-off value of 3 μg/mL as the commonly used therapeutic value seems arbitrary in the context of surgery, as the anti-TNFα concentration that guides dose adjustment is based on trough level (taken before anti-TNFα administration) and not a random sample unrelated to timing of anti-TNFα administration (interval is not quantified in the current study). In this study, only 177/217 (80%) had a sample collected the day before surgery and the results would have been more meaningful if the analysis was restricted to this group. As it stands, high levels done long before surgery may have been lower before surgery and are therefore overstating the impact of the high level. Short of this it would be useful to inform readers how many patients from the high levels group had samples drawn before the eve of the surgery.
In conclusion, the work presented contributes new knowledge by moving from a dichotomous (exposed vs nonexposed to anti-TNFα) to a continuous (cut-off level of anti-TNFα concentration) outcome. Future studies could improve on this through prospective recruitment on a more uniform patient population, for example, acute severe ulcerative colitis or emergency/elective surgery.

Related Topics

    loading  Loading Related Articles