Dyspnoea after AZD 6140: blaming the messenger

To the Editor: We thank Husted and van Giezen (1) for their ‘Letter to the Editor’ and their interest in our work (2). Given here the opportunity for a rebuttal, we think that three issues are worth mentioning. First, the high double-digit dose-dependent rates of dyspnea were first observed in the AZD6140 arms, while being absent in the clopidogrel treated patients enrolled in the DISPERSE study (3). We believe that this finding should have been followed by an evaluation and definition of the potential mechanism(s) responsible for such adverse association, and a reconsideration of the AZD6140 safety profile. This did not happen. Additional phase 2 (DISPERSE2), and phase 3 (PLATO) trials are now ongoing. The DISPERSE2 data confirmed a high rate of dyspnoea after AZD6140, also suggesting that clopidogrel causes dyspnoea (6.4%). Of course we were aware of this purported number. As a matter of fact, we deliberately avoided this controversy rather than “neglected to mention” increase of shortness of breath after clopidogrel. Of note, the reported adverse reactions in the clopidogrel insert are based on the broad-inclusion CAPRIE-trial. In this study, dyspnoea could be reasonably explained by a host of underlying pulmonary, cardiac, psychiatric and/or mixed-aetiology diseases, rather than representing a drug adverse reaction per se. Secondly, had dyspnoea after AZD6140 been appropriately and timely recognised in animal experiments and early human studies, there would not have been a need to propose different hypotheses on the genesis of this serious adverse event. Therefore, we are not ‘speculating’, but attempting to hypothesise why AZD6140 causes dyspnoea, filling the gap of the preclinical safety blemish. These ideas in the interest of patient safety are criticised by Husted and van Giezen (1), without, however, providing an own reasonable explanation, nor an in-depth analysis of the existing risks. While thousands of patients will be enrolled in the PLATO trial, and a considerable share of them presumably will experience dyspnoea, we still have no clue why this adverse event occurs, and what the immediate and long-term dangers are with regard to AZD6140 prescription and cardiovascular or/and pulmonary hazards. Finally, suggesting that AZD6140 is not exhibiting any ATP-analogue properties requires an alternative discovery of a distinct non-ADP dependent mechanism by which the drug inhibits platelet P2Y12 receptors. As such mechanism is not identified, and considering the very similar chemical structure of AZD6140 with the parent compound (ARC69931MX, or cangrelor) which is a classical ATP analogue, it seems reasonable to consider adenosine overload as a rational mechanism by which both drugs affect platelets. Moreover, a very recent review on ATP biology suggests similar activities of both agents via blocking of platelet G1 adenylyl cyclase, classifying them in one group (4).