LAUSR

The most frequent adverse drug reactions (ADRs) of codeine and tramadol, the two main step 2 opioid analgesics, include nausea, dizziness, sedation, vomiting, and constipation [1]. The relative importance of vomiting and constipation remains discussed [1–3]. We performed a comparative analysis of ADRs in VigiBase®, the World Health Organization Global Individual Case Safety Report (ICSR) database [4]. We used the disproportionality method [5, 6] including reports (age ≥ 18 years and gender known) registered from 1969 to 7 October 2017 under the MedDRA preferred terms Bvomiting^ or Bconstipation^. Drug exposure was identified using the anatomical therapeutic and clinical codes N02AX and N02AA, with drugs defined as Bsuspected^ or Bconcomitant^ [7, 8]. Associations of codeine or tramadol with antispasmodics (N02AG), other drugs (caffeine, barbiturates...) as well as the use of codeine + tramadol simultaneously were not included in the study. A sensitivity analysis compared tramadol + paracetamol (N02AJ) versus codeine + paracetamol. Strength of the association was quantified by reporting odds ratios (ROR) with their 95% confidence interval. Among the 10,265,261 ICSRs, 224,788 were included, 74.8% with tramadol [19.2% with paracetamol] and 25.2% with codeine [66.6%with paracetamol].Mean age (56.3 years) and sex ratio (64.4%women) were similar. Vomiting and constipation were more frequently reported with tramadol than with codeine. The sensitivity analysis including the association with paracetamol found similar trends (Table 1). Although the two drugs are structurally related (tramadol is a synthetic codeine analog) [1], we described differences in reports of vomiting and constipation. There are very few published clinical data. For vomiting, a meta-analysis of 3453 postoperative patients [9] and a 4-week randomized trial [10] failed to find any difference. In contrast, Rodriguez found a higher rate of vomiting with tramadol in a double-blind trial with 177 painful cancer patients [11] and Zavareh a three times higher value on a visual analog score for vomiting with tramadol following cholecystectomy [12]. Mullican [10] and Duthie [13, 14] reported more constipation with codeine. The major strength of our work is its high power, due to the number of ICSRs. In order to minimize a putative indication bias, frequent in such studies [6, 15], we compared two drugs used in similar indications. The main limitation of the paper is underreporting, like in each pharmacovigilance study. However, underreporting does not modify results of case non-case studies, since ADRs underreporting is similar inside a same pharmacotherapeutic class, allowing direct comparisons [16]. For these dose-dependent ADRs, doses were not included since they are not systematically recorded in VigiBase®. The results could have two explanations. First, a difference in the pharmacodynamic characteristics of the two drugs [1]. Tramadol is, first, a mu agonist and, second, an inhibitor of serotonin reuptake whereas codeine only stimulates mu opioid receptors. Activation of bothmu and serotonin receptors in the chemoreceptor trigger zone by tramadol can explain the differences with codeine. Opioid-induced constipation is a multifactorial process explained by a decrease in both gastrointestinal motility and gastrointestinal secretions. The second and probably more relevant explanation could involve both a * Jean-Louis Montastruc [email protected]