All patients from the RESERVOIR trial7 evaluated by OCT at 9 months of follow-up were included in the current study. A full description of the trial has been published previously.7 Briefly, the RESERVOIR trial was a multicenter, prospective, randomized clinical trial that aimed to compare the endothelialization of 2 different DES by OCT in 112 patients with DM.8 The target population consisted of diabetic patients with documented silent ischemia, stable angina, unstable angina or non—ST-segment elevation myocardial infarction. Major exclusion criteria included patients with DM treated only with diet and lifestyle changes, ST-segment elevation myocardial infarction, left main or ostial left descending artery stenosis, large bifurcations, severely calcified lesions, stent restenosis, chronic renal failure, or left ventricular ejection fraction < 30%. From a total of 112 patients randomized, 98 patients (101 lesions) underwent angiographic and OCT follow-up surveillance at 9 months. OCT results demonstrated that formulation of the drug with an amphiphilic carrier (amphilimus-eluting stents, CID, Saluggia, Italy)9 resulted in noninferior efficacy compared with the classic elution from durable polymers (everolimus-eluting stents, Abbott Vascular, Illinois, United States).

All patients were followed up by an endocrinologist during the study follow-up. A complete metabolic profile including body mass index, creatinine, hemoglobin A1c, high-density lipoproteins and low-density lipoproteins cholesterol was assessed at baseline and at the moment of the angiographic follow-up. Diabetes mellitus management was performed according to recommendations for standard of care of European clinical practice guidelines on DM.10

Patients were divided into 3 groups according to the glucose-lowering agents received during the study follow-up: a) patients not treated with MF; patients receiving MF were subsequently divided depending on the exogenous insulin status; b) patients treated with MF but not receiving exogenous insulin, and c) insulin-treated patients also treated with MF.

The synergy between MF and mTOR inhibitors inhibiting neointimal healing was evaluated in this study by the relative frequency of uncovered struts by OCT at 9 months of follow-up (primary endpoint). Struts were classified as uncovered if any part of the strut was visibly exposed to the lumen, or covered if a layer of tissue was visible over all the reflecting surfaces.11 The NI volume obstruction was used as the efficacy endpoint, inhibiting the NI hyperplasia. Neointimal volume obstruction was defined as the NI volume (mm3) divided by the stent volume (mm3) multiplied by 100.11 Other endpoints included the relative frequency of malapposed struts, the relative frequency of uncovered and malapposed struts, NI thickness, maximal NI area obstruction assessed by OCT, and angiographic in-stent late loss.8

Quantitative coronary analysis was performed at a central core laboratory (BARCICORE Lab, Barcelona, Spain)12 by experienced analysts, blinded to the medical treatment received for DM. The analysis was performed using dedicated coronary angiography analysis software (CAAS version 5.9, Pie Medical BV, Maastricht, the Netherlands).

Optical coherence tomography data were also analyzed at the same core laboratory (BARCICORE Lab, Barcelona, Spain) by experienced analysts, who were also blinded to the clinical data, using proprietary offline software (LightLab Imaging, St. Jude Medical Inc.). Cross-sections at 1-mm intervals within the stent segment and 5mm proximal and distal to the stent edges were analyzed as previously described. Frames with overlapped stents or side-branches take off were not considered for analysis. For totally occluded vessels that were not associated with stent thrombosis, it was estimated that the entire length of the stent was filled with NI hyperplasia.13

Statistical analysis was performed using SPSS version 23.0 (Chicago, Illinois, United States). Continuous variables are shown as mean±standard deviation unless otherwise noted. Categorical variables are expressed as frequencies and group percentages. Baseline characteristics were compared using the chi-square test or Fischer exact test for categorical variables and ANOVA (analysis of variance) for continuous variables. Bonferroni adjustment was used for multiple comparisons between groups. To evaluate the synergic effect of mTOR inhibitors and MF taking into account the nonindependence of struts within lesions, generalized estimating equation models as well as nonparametric analysis of aggregated data were used.14 Generalized estimating equation models were conducted nesting struts within lesions, and using treatment groups and the stent type received by randomization as covariates. Statistical significance was set at the .05 level for all comparisons.

Angiographic characteristics are shown in Table 3. After stent implantation, reference vessel diameter, minimal lumen diameter and acute gain were similar for all groups. At 9 months of follow-up, in-segment late lumen loss was significantly lower for non-ITDM treated with MF compared with ITDM patients treated with MF (P < .01), whereas there were no differences between non-ITDM patients with MF vs patients without MF (P = 1.0). Insulin-treated patients had significantly higher late loss vs those never treated with insulin (P = .02, Figure 2).

Figure 2.

Cumulative frequency of in-segment late luminal loss by quantitative coronary angiography depending on whether the patient was treated with metformin (left panel) or exogenous insulin (right panel). MF, metformin.

(0.17MB).

The OCT results are presented in Table 4, Figure 3 and Figure of the supplementary material. Analyzable neointimal volume data were available in all patients; however, strut coverage could not be assessed in 1 patient in the no MF group due to technical issues. The primary endpoint, the relative frequency of uncovered struts, was similar for all 3 groups (3.07±4.80% for patients without MF vs 2.23±4.73% for non-ITDM + MF group vs 3.43±6.69% for ITDM + MF group, P = .48). Generalized estimating equations models confirmed no association between MF and the relative frequency of uncovered struts, regardless the insulin status or the stent type allocated during the study randomization (OR, 1.49; 95%CI, 0.71-3.08; P = .29). Similarly, the NI volume obstruction (endpoint of efficacy of the combination of drugs) was also similar for all groups (P = .87).

Figure 3.

Cases of the 3 groups of therapy received: A: No MF; B: MF + non-ITDM; C: MF + ITDM. Column 1 represents angiography images pre-PCI, column 2 post-PCI, column 3 angiography at 9 months, and the last column represents OCT images at 9 months. Green arrows point to covered struts. Red arrows represent uncovered struts. The asterisk represents the wire artefact. FU, follow-up; ITDM, insulin-treated diabetes mellitus; MF, metformin; non-ITDM, noninsulin treated diabetes mellitus; OCT, optical coherence tomography; PCI, percutaneous coronary intervention.

(0.54MB).

Clinical follow-up was obtained for all patients at 12 months (Table 5). No significant differences were observed in any of the prespecified endpoints between groups, although numerically there were more target vessel revascularizations in the no MF and ITDM + MF groups (15.8% and 14.3%, respectively) vs 5.6% in non-ITDM + MF group; P = .28).

Our study design has some limitations. This study is a posthoc analysis of a randomized trial designed to compare 2 different DES. The sample size was calculated based on a different hypothesis, and MF was not randomly assigned. Therefore, the results should be interpreted with caution.

However, our study has also some strengths. First, the population was prospectively included and followed up, and data were exhaustively collected and checked by an external contract research organization; second, because glucose-lowering agents were prescribed in our study by a local endocrinologist according to the current recommendations rather than randomly assigned, the conclusions of the present study may be more representative of a real population of diabetic patients.