Hyperglycemia is “poison in the blood,” at least for the endothelium. It is widely recognized that the characteristics of vasculopathy differ markedly between diabetic and non-diabetic patients. This distinction, in addition to the natural course of the disease, has an evident clinical impact. Despite extensive research efforts aimed at reversing the effects of microvascular loss, such attempts have so far been unsuccessful. Whether in surgery or interventional procedures, the only available target remains the epicardial vessels, that is, macrovascular disease. However, the absence of distal beds or dysregulated microvasculature likely translates into poorer outcomes compared to patients in whom diabetes mellitus has not left its mark.

Beyond this, although it is uncommon to conduct clinical studies on other forms of ischemia when epicardial vessel disease is present, I am convinced that in some patients, this condition coexists with non-obstructive forms (also known as MINOCA: myocardial ischemia with non-obstructive coronary artery disease). Among its various presentations, an imbalance between demand and supply in a hypertrophic and remodeled myocardium, the presence of vasospasm—often related to endothelial dysfunction—or the loss of capillary territory may be present alongside left main and/or multivessel disease that we intend to treat. Among the overall population diagnosed with myocardial injury, more than 75% have an obstructive cause. However, when discussing the coexistence of both phenomena, the impact of microvascular disease is likely underestimated.

Barili et al. provide yet another excellent study, based on the follow-up of a cohort of 10,989 patients included in the PRIORITY project (PRedictIng long-term Outcomes afteR Isolated coronary arTery bypass surgerY), a multicenter observational cohort designed to shed light on long-term outcomes after coronary revascularization surgery. In this case, the study focused on the impact of diabetes mellitus on survival and the need for repeat revascularization. The cohort was followed for up to 10 years, including 32.3% of patients who had diabetes at the time of revascularization. Notably, the study did not account for the possibility that some patients may have been diagnosed with diabetes over such a prolonged follow-up period. However, this factor is unlikely to have significantly altered the results, as the vascular damage associated with diabetes becomes clinically relevant approximately 10 years after diagnosis.

After adjusting patient groups through propensity score analysis, the presence of diabetes mellitus did not affect short-term survival. Paradoxically, it was associated with a lower incidence of cardiovascular and cerebrovascular events at 30 days (OR = 0.67), as well as lower rates of myocardial infarction (OR = 0.6) and stroke (OR = 0.47).

However, throughout follow-up, patients with diabetes mellitus experienced worse outcomes, with significantly higher rates of cardiovascular and cerebrovascular events (HR = 1.3), mortality (HR = 1.45), and recurrent myocardial infarction (HR = 1.39). Nevertheless, despite this 30–40% increase in adverse events—again paradoxically—this did not translate into a higher need for repeat revascularization at 10 years.

In conclusion, the authors determine that diabetic patients have worse long-term outcomes after coronary revascularization surgery. They emphasize the various pathophysiological mechanisms that may explain these findings, which we will analyze below.

COMMENTARY:

This article can be considered a hypothesis generator, as it employs an epidemiological methodology to explore explanations for phenomena observed over prolonged follow-ups. Many cardiology studies could benefit from adopting such methodologies, which, despite their simplicity, provide solid evidence to guide clinical decision-making and patient management in real-world practice.

The analysis of this study can be approached through two paradoxical findings observed at two different follow-up intervals: the perioperative period at 30 days and the extended follow-up at 10 years. Regarding the former, the better short-term outcomes in diabetic patients stand out as a paradoxical finding. The inherent limitations of a retrospective study should be acknowledged, yet this observation could be explained by a degree of selection bias within the diabetic patient population. In fact, in the general cohort, the presence of fewer than one-third of diabetic patients does not align with the clinical reality we are accustomed to, where, in our setting, they nearly represent half of the cases. Other factors, such as treatment type and the degree of glycemic control, as well as perioperative management strategies and complication prevention protocols—potentially differing across centers—may account for these findings. Finally, another possible explanation could lie in the pathophysiological characteristics of diabetic vascular disease itself. More advanced stages of vasculopathy in these patients may have led to better development of collateral circulation, providing greater protection against acute perioperative ischemic events.

Regarding long-term events, the study highlighted a similar rate of repeat revascularization between diabetic and non-diabetic patients, despite the diabetic cohort exhibiting worse outcomes in terms of recurrent cardiovascular and cerebrovascular events. The fact that diabetic patients have poorer long-term outcomes is well established, translating into an excess mortality of over 40%. A significant proportion of this excess mortality is likely driven by the progression of vasculopathy, as evidenced by the higher rate of events occurring outside the coronary circulation, particularly in the cerebrovascular territory. However, how can we explain the paradoxical finding of a higher incidence of recurrent myocardial infarctions without a corresponding increase in the need for repeat revascularization? The authors address this paradox, and I believe correctly, by stating that “it cannot be explained.” The impact of diabetic microvascular disease has likely been underestimated and could be responsible for some of these events, where, upon reviewing coronary anatomy and graft patency, no treatable lesions are found. This leads to a third paradox: the traditional notion that an increased need for repeat revascularization correlates with higher mortality does not hold true in this context.

Ultimately, this study aims to generate new hypotheses for further exploration of coronary artery disease. While the proposed hypotheses are pathophysiologically plausible, they require further validation. Additionally, they contribute to a better characterization of ischemic heart disease in our patients and the various therapeutic approaches available. Until then, we should bear in mind that, if diabetic vessels are already small and difficult to visualize, it is highly likely that other forms of ischemia are present, which we are simply not detecting.

REFERENCE:

Barili F, Vitale N, D’Errigo P, Porcedda F, Pollari F, Baglio G, et al. The effect of diabetes on long-term outcomes in patients following coronary artery bypass grafting. Eur J Cardiothorac Surg. 2025 Feb 4;67(2):ezaf024. doi: 10.1093/ejcts/ezaf024.