Legacy effect of intensive glucose control on major adverse cardiovascular outcome: Systematic review and meta-analyses of trials according to different scenarios

Background: Early and intensive glycaemic control provides long-term protection against the development of microvascular complications, a phenomenon defined legacy effect. Whether a legacy effect of high glucose exists also on macrovascular endpoints is uncertain. Methods and findings: We performed a systematic review of both randomized clinical trials (RCT)s and observational studies pertinent to the research question. We searched PubMed, Embase, Scopus and the Cochrane database up to January 31th 2020. Eligibility criteria for RCTs were: 1 - efficacy assessment of intensive glucose lowering treatment vs a less-stringent/conventional treatment; 2 - the inclusion of a post-active phase, observational follow-up; 3 - enrolment of patients with T1DM, pre-diabetes, and T2DM; and 4 - data report on major adverse cardiovascular events (MACE) incidence, which was the primary endpoint of this meta-analysis. We performed multiple meta-analyses of the available RCTs according to different scenarios considering the type of diabetes, diabetes duration, the presence of previous cardiovascular events, follow-up extension, and the incidence of MACE recorded only during the observational, post-active phase of the trials. Results from observational studies reporting the association between HbA1c levels during the first year after diabetes diagnosis and subsequent MACE incidence were also collected and are reported narratively. We included data from 7 RCTs and 40,346 patients. The intensive glucose-lowering approach significantly decrease the incidence of MACE compared with conventional treatment (OR 0.86, CI 0.77–0.96; p = 0.007) when considering all the available studies, with a more consistent effect (OR 0.73, CI 0.56–0.94; p = 0.01) in the case of RCTs enrolling patients with diabetes duration <10 years, and an even more pronounced protection (OR 0.64, 48 CI 0.48, 0.86; p = 0.003) when analysing only RCTs enrolling patients without previous cardiovascular events at baseline. Considering only RCTs with a post-trial follow-up >10 years also yielded a relevant beneficial effect of the intensive approach (OR 0.71, CI 0.57, 0.88; p = 0.002). On the other hand, no effect was observed (OR 0.99, CI 0.92, 1.06; p = 0.81) when considering only the events recorded during the post-active, observational phases of the trials. Observational studies showed that HbA1c values >6.5% or 7% during the first year of diabetes diagnosis are associated with a higher incidence of late MACE with increased risk ranging from 19 up to 64%, according to the different study design and HbA1c stratification. Conclusions: These results support the recommendation regarding glucose-lowering treatment intensification in order to decrease the probability of having a macrovascular event in patients with short diabetes duration, no prevalent cardiovascular diseases, and long life-expectancy. On the other side, data from RCTs do not support the existence of a protective legacy effect on the macrovasculature beyond the period of intensive glycaemic treatment.