67POR. A Cochrane analysis by Doherty et al. demonstrated a significant reduction in
both clinical recurrence (RR 0.75, 95% CI 0.62–0.94) and severe (Rutgeerts ≥ i3)
endoscopic recurrence (RR 0.50, 95% CI 0.29–0.84) compared to placebo but with
a number needed to treat (NNT) of 12 and 8, respectively [ 40 ]. A subsequent meta-
analysis by Ford et al. demonstrated that this effect was exclusive to mesalamine as
sulfasalazine was of no benefit to prevent POR compared to placebo in 448 patients
(RR = 0.97, 95% CI 0.72–1.31) [ 41 ]. The authors conclude that mesalamine is of
modest benefit in preventing POR but should only be used when immunosuppres-
sive therapy is either not warranted or contraindicated.
In the previously mentioned Cochrane meta-analysis, Doherty et al. also exam-
ined the impact of nitroimidazole (including metronidazole) antibiotics and found
that these agents significantly reduced the risk of clinical (RR 0.23, 95% CI 0.09–
0.57, NNT = 4) and 3-month endoscopic (Rutgeerts ≥ i2) (RR 0.44, 95% CI 0.26–
0.74, NNT = 4) recurrence compared to placebo [ 40 ]. However, these agents were
associated with significantly higher risk of serious adverse events (RR 2.39, 95% CI
1.5–3.7), and the clinical recurrence effect lost statistical significance after exclu-
sion of ornidazole. Thus the role of antibiotics in prevention of POR seems to be of
limited benefit and short-term due to adverse events.
Immunomodulators have also been studied in the prevention of POR. In the
aforementioned Cochrane meta-analysis, Doherty et al. examined two trials com-
paring thiopurines to placebo for prevention of POR and found that the use of aza-
thioprine (AZA)/6-mercaptopurine (6-MP) significantly reduced the risk of clinical
(RR 0.59, 95% CI 0.38–0.92, NNT = 7) and severe (Rutgeerts ≥ i3) endoscopic
recurrence (RR 0.64, 95% CI 0.44–0.92, NNT = 4) at 12 months [ 40 ]. Comparing
mesalamine to thiopurines, mesalamine carried a significantly higher risk of endo-
scopic recurrence at 12 months (RR 1.45, 95% CI 1.03–2.06) but had significantly
fewer serious adverse events (RR 0.51, 95% CI 0.30–0.89). Similar findings were
observed in a concurrent meta-analysis of the same studies by Peyrin-Biroulet et al.
but found the superiority of immunomodulators to placebo extended to 2 years in
prevention of clinical recurrence (mean difference 13%, 95% CI 2–24%, p = 0.0016,
NNT = 8) [ 42 ]. However, immunomodulators were not effective in prevention of
very severe (Rutgeerts i3–i4) recurrence. In a recent randomized, double-blind,
placebo-controlled, parallel-group trial of 6-MP in POR, Arnott et al. studied 240
CD patients undergoing intestinal resection and found that patients receiving pla-
cebo were more likely to have clinical recurrence (CDAI >150 plus 100-point rise)
(23.2% vs. 12.5%), but adjusted analysis was not statistically significant (p = 0.07)
[ 43 ]. Stratifying by smoking status showed a significant difference between placebo
and 6-MP in smokers in clinical recurrence (HR 0.127, 95% CI 0.04–0.46, NNT = 3)
but not in nonsmokers (HR 0.898, 95% CI 0.42–1.94, NNT = 31). Significantly
more patients receiving 6-MP maintained complete endoscopic remission (Rutgeerts
i0) at 1 year (29.7% vs. 14.4%, p = 0.006) and 3 years (22.5% vs. 12.5%, p = 0.041).
The authors concluded that thiopurines modestly reduce POR in CD with a signifi-
cant effect in smokers, but not in nonsmokers.
The combination of short-term metronidazole with AZA may improve outcomes
further. Postoperative CD patients treated with metronidazole for 3 months and
5 Use of Biologics in the Postoperative Management of Crohn’s Disease