Table 2.

Risk of adenoma according to celecoxib treatment, stratified by aspirin use and UGT1A6 genotypea

PlaceboCelecoxib, 200-mg twice dailyCelecoxib, 400-mg twice daily
All genotypes
All patients, no. at riskb493497494
 Cumulative incidence, 3 y, % ± SE59.8 ± 2.343.3 ± 2.336.8 ± 2.3
 RR (95% CI)c1.00.68 (0.59–0.79)0.54 (0.46–0.64)
Pd<0.001<0.001
Aspirin users, no. at riskb162160152
 Cumulative incidence, 3 y, % ± SE61.0 ± 4.046.0 ± 4.138.7 ± 4.1
 RR (95% CI)e1.00.69 (0.54–0.89)0.54 (0.41–0.73)
Pd0.004<0.001
Non-aspirin users, no. at riskb331337342
 Cumulative incidence, 3 y, % ± SE59.2 ± 2.842.0 ± 2.836.0 ± 2.7
 RR (95% CI)e1.00.67 (0.56–0.81)0.54 (0.44–0.66)
Pd<0.001<0.001
Patients with wild-type genotypes
All wild-type genotypes, no. at riskb212213221
 Cumulative incidence, 3 y, % ± SE62.3 ± 3.442.0 ± 3.536.3 ± 3.4
 RR (95% CI)f1.00.63 (0.50–0.78)0.51 (0.40–0.65)
Pd<0.001<0.001
Aspirin users, no. at riskb746969
 Cumulative incidence, 3 y, % ± SE64.3 ± 3.444.0 ± 6.235.7 ± 6.1
 RR (95% CI)g1.00.64 (0.44–0.92)0.51 (0.33–0.77)
Pd0.013<0.001
Non-aspirin users, no. at riskb138144152
 Cumulative incidence, 3 y, % ± SE61.4 ± 4.241.0 ± 4.336.5 ± 4.0
 RR (95% CI)g1.00.62 (0.47–0.82)0.51 (0.38–0.69)
Pd<0.001<0.001
Patients with variant genotypes
All variant genotypes, no. at riskb281284273
 Cumulative incidence, 3 y, % ± SE57.9 ± 3.144.2 ± 3.137.1 ± 3.0
 RR (95% CI)f1.00.72 (0.59–0.88)0.57 (0.45–0.71)
Pd0.001<0.001
Aspirin users, no. at riskb889183
 Cumulative incidence, 3 y, % ± SE58.4 ± 5.447.4 ± 5.140.7 ± 5.6
 RR (95% CI)g1.00.74 (0.53–1.04)0.57 (0.38–0.86)
Pd0.0840.005
Non-aspirin users, no. at riskb193193190
 Cumulative incidence, 3 y, % ± SE57.6 ± 3.742.6 ± 3.735.6 ± 3.6
 RR (95% CI)g1.00.71 (0.56–0.91)0.56 (0.44–0.73)
Pd0.006<0.001
  • aCumulative risk of adenoma through follow-up colonoscopy at year 3. Patients were stratified at study entry according to the use or nonuse of low-dose aspirin (≤325 mg every other day or ≤162.5 mg every day). Patients not taking aspirin at baseline were required to abstain from taking it during the trial. Wild-type UGT1A6 genotypes include individuals with no T181A or R184S alleles. Variant UGT1A6 genotypes include individuals with ≥1 variant (T181 or R184S) allele.

  • bNo. at risk includes patients who underwent a follow-up colonoscopy at year 1 and/or 3.

  • cRisk ratio is estimated by the Mantel–Cox method, with stratification for genotype, aspirin use, time, age, and sex, with the placebo group as the referent group.

  • dP value is the Cochran–Mantel–Haenszel test of general association compared with the placebo group.

  • eRisk ratio is estimated by the Mantel–Cox method, with stratification for genotype, time, age, and sex, with the placebo group as the referent group.

  • fRisk ratio is estimated by the Mantel–Cox method, with stratification for aspirin use, time, age, and sex, with the placebo group as the referent group.

  • gRisk ratio is estimated by the Mantel–Cox method, with stratification for time, age, and sex, with the placebo group as the referent group.