In this pilot study, we show that sevoflurane anaesthesia impairs resting myocardial blood flow in type 2 diabetic patients, while it preserves myocardial blood flow in healthy controls. Furthermore, we observed a decrease in hyperaemic MBF (endothelium-independent vasodilation) during sevoflurane anaesthesia in both groups, with a trend towards a larger decrease in diabetic patients. Sympathetic stimulation under sevoflurane did not lead to changes in MBF in controls or diabetics when compared to baseline (endothelium-dependent vasodilation). These findings suggest different myocardial responses to sevoflurane anaesthesia in type 2 diabetic patients compared to healthy controls.
Type 2 diabetes mellitus leads to endothelial dysfunction of the coronary circulation due to reduced availability of nitric oxide, accumulation of glycation end products in the arterial wall, impaired insulin signalling, inflammation and other still unknown mechanisms[7, 21]. Picchi et al. showed that diabetic coronary dysfunction is clinically characterised by increased basal MBF and similar hyperaemic MBF compared to healthy patients. In contrast, Di Carli et al. reported similar resting MBF and lower hyperaemic MBF in diabetics compared to controls. The additional influence of sevoflurane anaesthesia on MBF and endothelial function in diabetic patients is largely unknown. Our type 2 diabetic patients had a lower resting MBF when compared to baseline and this sevoflurane effect on MBF was different from healthy controls. Interestingly, Cosyns et al. reported no difference in resting MBF in streptozotocin-induced type 1 diabetic rats compared to controls rats under thiopental anaesthesia. No studies on sevoflurane anaesthesia are available for comparison with our data.
Sevoflurane anaesthesia decreased MBF during adenosine-induced hyperaemia in both groups when compared to baseline conditions. This was in accordance with our previous study in healthy volunteers. Also, diabetics tended to have a larger decrease in MBF than healthy controls. Cosyns et al., showing higher hyperaemic MBF in controls than in diabetic rats, confirm the latter observation. Underlying mechanisms for this observation may include changes in perfusion pressure and myocardial oxygen demand, decreased capillary recruitment in diabetic myocardium and direct effect of sevoflurane on endothelial cells.
Endothelial-dependent vasodilation by the CPT during sevoflurane anaesthesia was not different from baseline in both controls and diabetic patients. It has been shown that the MBF response to sympathetic stimulation is primarily dependent on the integrity of the autonomic nervous system and not on hemodynamic changes or circulating catecholamine levels. Also, it was shown by the same investigators that impaired MBF responses to endothelium-dependent vasodilation in type 2 diabetic patients were related to the degree of cardiac autonomic dysfunction. In our population, only one out of six diabetic patients had clinical evidence of autonomic dysfunction, which hampers evaluation of MBF responses in this subpopulation under sevoflurane anaesthesia. Future study designs should also be aimed at type 2 diabetics with cardiac autonomic dysfunction.
For this study sevoflurane anaesthesia was used since it is widely available and allows both induction and maintenance of anaesthesia. Volatile anaesthetics are thought to possess cardioprotective properties although experimental evidence suggests that these properties are reduced in type 2 diabetes. Interestingly, a recent study reported that the widely used intravenous anaesthetic propofol attenuates hyperglycaemia-induced upregulation of endothelial adhesion molecules expression and mononuclear-endothelial adhesion. This implicates that propofol inhibits pathological leukocytes-endothelial adhesion and may subsequently prevent endothelial dysfunction and injury. Furthermore, comparing between sevoflurane and propofol, it has been reported that sevoflurane better preserved right ventricular function than propofol in patients receiving esophagectomy. However, propofol was better in improving oxygenation and shunt-fraction during one-lung ventilation. The results of these studies suggest that the choice of anaesthetic may be of interest in certain patient categories. Specifically in type 2 diabetics, influence of the two most widely used agents on myocardial perfusion and function should be further investigated.
All type 2 diabetic patients in this study received oral antidiabetic therapy (see Table 1), which are known to exert cardiovascular effects and may affect myocardial blood flow. Previous studies reported cardioprotective effects exerted by metformin mediated by protein kinase B, adenosine monophosphate-activated protein kinase and upregulation of insulin signalling pathways[29, 30]. Also, metformin improved endothelium-dependent vasodilation in both nondiabetic and diabetic subjects[31, 32]. In isolated rat hearts, sulfonylurea derivatives reduced myocardial perfusion under ischemic conditions, while perfusion under normal conditions was unaffected[33, 34]. In humans, these oral diabetics deteriorated both resting and hyperaemic myocardial blood flow. In type 2 diabetic patients, exenatide increased myocardial blood flow, while treatment with peroxisome proliferator-activated gamma agonists did not affect myocardial perfusion[35–37]. Whether and how antidiabetic therapy influences the results of this study cannot be concluded.
The small number of patients studied limits interpretation of the present study results. However, pilot studies are valuable and necessary to enable power calculation for future studies. Ideally, in subsequent studies age-matched controls should be used for comparison with diabetic subjects. The age difference between healthy controls and diabetic subjects in this study could be a potential confounder.
Already in this small study population we showed that sevoflurane reduces resting MBF in diabetic patients compared to healthy controls. Hyperaemic MBF decreased during sevoflurane anaesthesia in both groups. Next, an adequately powered study should be undertaken; for hyperaemia a sample size of 21 in each group will have 80% power to detect a significant difference using a two-group t-test with a 0.05 two-sided significance level. For cold pressor testing, a sample size of 92 in each group will have 80% power to detect a significant difference using a two-group t-test with a 0.05 two-sided significance level. Furthermore, assessment of effects of hypo- and hyperinsulinemia on vascular reactivity via nitric oxide and insulin signaling proteins such as Akt and endothelial nitric oxide synthase combined with altered plasma levels of endothelin-1, asymmetric dimethylarginine and nitric oxide may elucidate mechanisms behind sevoflurane-induced MBF changes in diabetic patients[38–41].