Background: Low-dose acetylsalicylic acid (ASA) is recommended in current guidelines for a period of 3 months after surgical replacement of the aortic valve by a bioprosthesis as well as after surgical reconstruction of the mitral valve. Metamizole (M; dipyrone) and paracetamol (P, acetaminophen) are frequently used non-opioid analgesics after surgery. The active metabolite of metamizole (4-methylaminoantipyrine) is a reversible inhibitor of cyclooxygenase-1 and an interaction with the antiplatelet effect of ASA in more than 50 % of patients has been reported. Prospective randomized studies investigating the potential drug-drug interaction between ASA and high-dose metamizole in cardiac surgical patients are lacking so far.

Material and methods: We enrolled 162 patients in this prospective clinical study. Patients after surgical implantation of a bioprosthetic valve in aortic position or after mitral valve repair were eligible for enrollment. Patients were randomized at day 1 after surgery to either 4 x 1g metamizole iv/day or 4 x 1g paracetamol iv/day. All patients were on ASA 100 mg oid. The primary endpoint was the proportion of patients with adequate antiplatelet response to ASA assessed via determination of thromboxane B2 in platelet rich plasma after stimulation with arachidonic acid (AA) at day 4/5 after surgery (cutpoint ≤209.8ng/ml thromboxane B2). Key secondary endpoints comprised platelet reactivity assessed by light transmission aggregometry [LTA] and by multiple electrode impedance aggregometry [MEIA].

Results: 15 patients dropped out of the study. Thus, data from 71 patients randomized to metamizole and 76 patients randomized to paracetamol were available for data analysis. Patients had a mean age of 66.5±7.6 years and 97 were male and 50 female. Aortic valve was replaced in 92 patients, 54 underwent mitral valve reconstruction and combined surgery was performed in a single patient. The primary endpoint analyzed at day 4/5 after surgery is shown in the Figure. The proportion of non-responders in patients treated with metamizole was lower than expected (12.5%) which is 4-fold higher compared with patients on paracetamol. The difference between the treatment groups was, however, of borderline statistical significance only. Platelet reactivity at day 4/5 after surgery was assessed after stimulation with arachidonic acid by LTA or MEIA. No statistically significant difference between the treatment strata was determined in both assays: Maximum aggregation in LTA was 5.0% (4.0 - 8.0%) in M+ASA and 6.0% (5.0 - 11.0%) in P + ASA (median with 95% CI, p=0.528) and MEIA AUC was 345.5 AU*min (268.0 - 419.0) in M+ASA and 276.5 AU*min (241.0 - 330.0) in P + ASA (p=0.098).

Summary and conclusion: The proportion of non-responders to ASA is higher in patients treated with metamizole compared to paracetamol. The proportion of non-responders in the metamizole-treated group in our study is substantially lower than reported in previous studies.