To the Editor Although Kolesnikov et al.1 are to be congratulated for describing the effect of COMT and OPRM1 polymorphisms on clinical efficacy and side effects of IV morphine, several aspects of the study design, analysis, and, therefore, conclusions require discussion. The study was designed to identify differences in total IV patient-controlled analgesia morphine use over the first 48 hours postradical prostatectomy or women undergoing hysterectomy under general anesthesia. Given that the patient-controlled analgesia pump was set to deliver a background infusion of 1 mg/h morphine and a 1-mg bolus dose with a lockout time of 10 minutes, all patients were expected to receive a total dose of at least 48 mg over the study period. The authors report that the average morphine consumption was in the range of 62 to 66 mg over 48 hours and they found no difference between genotypic groups based on Val158Met polymorphism of COMT (Table 21) and A118G polymorphism of OPRM1 (Table 31). However, the continuous infusion may have masked a possible effect of COMT genotype on morphine analgesic efficacy preventing a significant difference in the patient-administered analgesic bolus from becoming apparent. In addition, the mixed-gender study population and 2 distinct surgical models may have increased variability in postoperative morphine consumption. A further observation is that the joint COMT/OPRM1 allelic distribution is not reported for the entire cohort of patients enrolled in the study. With 2 polymorphisms, there should be 9 possible combinations, with the exception that some combinations may be extremely rare among Caucasians.2 The frequency for possible allelic combinations is not reported; therefore, the joint allelic distribution in 70 patients (68% of the cohort) remains unknown. Consequently, when the authors analyzed the joint allelic effect on morphine dose, they only compared morphine dose among all AA individuals (n = 82) irrespective of COMT genotype to a selected subgroup of individuals with A118G/Val158Met (n = 11). The groups that preferably should have been analyzed are Met/Met and AA patients (n = 18; Table 31) with Met/Val and AG (n = 11; Table 31), and the difference between these 2 groups would most likely not be significant (66.2 ± 29.7 mg vs 51.9 ± 12.1 mg, respectively). Another analysis to be considered would be to compare Met/Met and AA (n = 18) versus all the other patients (n = 84; data not presented in Table 3). This comparison would be of interest because Met/Met and AA individuals have been found to need less oral morphine dosing for pain relief than patients that are neither Met/Met nor AA.2 Ruth Landau, MD Clemens Ortner, MD Department of Anesthesiology University of Washington Medical Center Seattle, Washington [email protected] Brendan Carvalho, MD Department of Anesthesia Stanford University Stanford, California