Sex-related differences in platelet aggregation: rebuttal to reply

The reply by Galli et al.¹ to my letter² commenting their study³ of sex differences in platelet aggregation failed to address my comments properly, probably because they were expressed with insufficient clarity. I wish to clarify better my insight into the issue of sex difference in platelet aggregation addressed in the study by Galli et al.³

In their reply, the Authors mostly focused on the results of the study by Kelton et al.⁴ (quoted in my letter²), which they erroneously perceived to be different from those of their study, which showed that platelet aggregation, when studied by light-transmission aggregometry in platelet-rich plasma (PRP) anticoagulated with fixed volumes of citrate, is higher in women than men.³ One of the main criticisms raised by the Authors was that the results of the small study by Kelton et al. (22 subjects) cannot be compared with their large study (1483 subjects), whose sample size allowed the detection of statistically significant differences between sexes. The perceived presence of discrepancies between the Authors’ and Kelton’s studies epitomizes the most relevant misunderstanding that influenced the essence of the Authors’ reply.¹ Indeed, the results of the study performed by Kelton et al. in their highly specialized laboratory, far from contrasting those reported by Galli et al.,³ perfectly matched them: the extent of platelet aggregation induced by adenosine diphosphate at concentrations ranging from 0.5 to 5 μM (comprising the 2 μM concentration used by Galli et al.³) was significantly higher in women than men.⁴ To the best of my knowledge, all the similar studies published in the last 50 years, including the large Northwick Park Heart Study (958 subjects) and Framingham Study (3429 subjects), reported similar results.² Therefore, the concept that in vitro platelet aggregation is higher in women than men is well established and accepted by the scientific community, although the real reason for this discrepancy is apparently not as widely known.

The great importance of the study by Kelton et al. lies actually in its elegant demonstration that the observed sex differences in platelet aggregation are determined by an artefact caused by the increase in volume distribution of citrate in women’s blood samples, as a consequence of their lower haematocrit levels compared with men’s blood samples. The study by Kelton et al., published in 1980, may be considered old (which is not in itself a flaw), but certainly not dated, as the Authors alluded in their reply,¹ because it was methodologically very sound and its results were confirmed by subsequent studies, some of which were quoted in my letter.² Therefore, the focus of my comment has never been on the validity of the results of the study by Galli et al.,³ which are perfectly in line with all previously published studies and also with my personal experience,² but, rather, on the real biological relevance of the observed sex difference in platelet aggregation, which is to be ascribed to an in vitro artefact.²

The relevance of the haematocrit- and citrate-dependent artefact affecting the results of in vitro platelet aggregation applies not only to differences in sex, but also to any conditions associated with abnormal haematocrit levels. Among these, ageing could be considered, as the lower haematocrit of elderly subjects could hypothetically partially contribute to the known ‘hyper-aggregability’ of their platelets⁵ and possibly explain the observation by Galli et al.^1,3 that womens’ ‘hyper-aggregability’ is particularly pronounced in subjects >50 years of age. The haematocrit-adjusted citrate concentration used in parallel with standard fixed citrate concentration in the same subjects allowed Kelton et al.⁴ to unravel the artefact in their small sample-size study. Although haematocrit-adjusted citrate concentration is not recommended for routine use,⁶ evaluation of results of platelet aggregation in citrate PRP must not overlook the potential interference of abnormal haematocrit levels. This is particularly relevant in the practice of antiplatelet treatment monitoring, where small differences in borderline results could assign the same patient to the group of ‘high’ or ‘low’ on treatment platelet reactivity, which would condition the implementation of different treatment strategies.⁷ For example, in 475 clopidogrel-treated patients who were monitored by VerifyNow (which uses citrate-anticoagulated blood), the prevalence of high on-treatment platelet reactivity decreased from 54.5% to 41.9% after correction of the results on the basis of the haematocrit levels.⁸ This and other methodological flaws might be responsible for the failure of platelet function tests (PFT) monitoring of antiplatelet therapy,⁷ which was demonstrated by a meta-analysis restricted to randomized controlled trials with correct and homogeneous study designs, which guided antiplatelet therapy uniquely based on the results of PFT.⁹

Funding

No funding was necessary for this work.

Data availability

No data were generated or analysed for this article.