Profiling the Acute Effects of Modified Risk Products: Evidence from the SUR-VAPES (Sapienza University of Rome-Vascular Assessment of Proatherosclerotic Effects of Smoking) Cluster Study.

Modified risk products (MRP) are promoted as a safer alternative to traditional combustion cigarettes (TCC) in chronic smokers. Evidence for their lower hazardous profile is building, despite several controversies. Yet, it is unclear whether individual responses to MRP differ among consumers. We hypothesized that different clusters of subjects exist in terms of acute effects of MRP. Pooling data from a total of 60 individuals, cluster analysis identified at least three clusters (labelled 1 to 3) of subjects with different electronic vaping cigarettes (EVC) effects and at least two clusters (labelled 4 to 5) of subjects with different heat-not-burn cigarettes (HNBC) effects. Specifically, oxidative stress, platelet aggregation, and endothelial dysfunction after EVC were significantly different cluster-wise (all p < 0.05), and oxidative stress and platelet aggregation after HNBC were significantly different (all p < 0.05). In particular, subjects belonging to Cluster 1 appeared to have less detrimental responses to EVC usage than subjects in Cluster 2 and 3, as shown by non-significant changes in flow-mediated dilation (FMD) and less marked increase in Nox2-derived peptide (NOX). Conversely, those assigned to Cluster 3 had the worst reaction in terms of changes in FMD, NOX, and P-selectin. Furthermore, individuals belonging to Cluster 4 responded unfavorably to both HNBC and EVC, whereas those in Cluster 5 interestingly showed less adverse results after using HNBC than EVC. Results for main analyses were consistent employing different clusters, tests, and bootstrap. Individual responses to MRP differ and smokers aiming at using EVC or HNBC as a risk reduction strategy should consider trying different MRP aiming at finding the one which is less detrimental, with subjects resembling those in Cluster 1 preferably using EVC and those resembling Cluster 5 preferably using HNBC.