Background
As the first-line therapy of type 2 diabetes mellitus (T2DM), metformin is applying to nearly one-third of diabetes patients worldwide [
1]. Therefore, in urgent situations, such as acute myocardial infarction, many of them cannot withhold the metformin prior to percutaneous coronary intervention (PCI)-related contrast medium administration [
2]. Cleared by the kidneys, metformin is accumulated in circumstances when kidney function is deteriorated, such as contrast-induced acute kidney injury (CI-AKI), which may lead to lactic acid accumulation. Although metformin associated with lactic acidosis (MALA) has a reported mortality of 30–50%, it is a rare disease with an estimated incidence of 1–5 cases per hundred thousand people [
3]. However, previous studies have shown that in most cases, metformin therapy may be merely concomitant and may not have a causal role at all [
4].
ST-elevation myocardial infarction (STEMI) patients with CI-AKI after coronary stent implantation had longer hospital stays, increased adverse cardiovascular outcomes, and higher mortality [
2]. As the direct toxicity of contrast media, the high thrombogenic state, inflammation, and the decrease in renal perfusion, patients submitted to primary PCI for STEMI are at high-risks for CI-AKI, [
5]. According to different definition, the incidence of CI-AKI can range from approximately 6.4 to 27.7% in STEMI patients after primary PCI [
6].
Because of the post-procedural risk-MALA and CI-AKI in diabetic patients submitted to PCI for STEMI, the guidelines are inconsistent on whether to continue metformin in patients undergoing PCI [
7‐
10]. Currently, studies had concluded that initiation of metformin ≤ 3 h after primary PCI and chronic metformin therapy before primary PCI had no influence on CI-AKI in STEMI patients [
2,
11]. However, there is no clinical data about the continuation of metformin treatment during the primary PCI period in diabetic patients with STEMI, and the effects of metformin on the kidneys are still widely debated in patients exposed to contrast agents [
12,
13].
The objective of the study to was to evaluate the influence of continuous metformin therapy on kidney function after coronary intervention for STEMI. In the meantime, the independent predictors of CI-AKI were explored.
Discussion
In this study, the continuous use of metformin for T2DM treatment in STEMI patients, who were submitted to primary PCI, was found reliable concerning the development of CI-AKI. It was also found eGFR ≤ 60 ml/min/1.73 m2 and the contrast volume were risk factors of CI-AKI. Besides, discontinuation of metformin results in interruption of hypoglycemic treatment or change of hypoglycemic regimen, and patients presented glucose fluctuation after admission. It was the first study to demonstrate that diabetic patients continuing metformin treatment during primary PCI period were not correlated with CI-AKI.
T2DM is correlated with poor prognosis in patients who have non-obstructive or obstructive stable coronary artery disease. Recent studies demonstrated that diabetes increase mortality and adverse cardiac outcomes in patients with non-obstructive coronary artery stenosis (NOCS)-acute myocardial infarction (AMI) [
15,
16]. Diabetes may favour the plaque instability in the context of NOCS through pathogenetic mechanisms including inflammation, endothelial dysfunction and coronary vasospasm [
15]. As a first-line drug for T2DM, metformin has been shown to have cardiovascular benefits and fewer adverse reactions [
17]. Current guidelines suggest that prediabetes should also be treated with metformin to mitigate the risk of developing diabetes. A recent study by Celestino et al. found that metformin therapy may improve adverse cardiovascular outcomes in prediabetes patients by reducing coronary endothelial dysfunction. The improvement in endothelial dysfunction is attributed to metformin downregulating the inflammation/ oxidative stress in the context of NOCS [
18]. Besides, metformin therapy for prediabetes may improve outcomes by a reduction of inflammatory tone and leptin to adiponectin rate in peri-coronary fat in AMI patients [
19]. There is increasing evidence that specific hypoglycemic drugs with pleiotropic effect on inflammatory tone and oxidative stress may affect the control of atherosclerotic plaque progression in AMI patients [
15].
Patients presenting with STEMI are prone to develop CI-AKI after primary PCI, and high thrombogenic state and inflammation are among the leading causes [
5]. Hyperglycemic patients presenting with STEMI have higher coronary thrombus burden compared with thrombi from normoglycemic counterparts. Evidence showed that hyperglycemia causes overproduction of reactive oxygen species and inflammation from thrombus plaque, favouring thrombotic embolization and poor myocardial infarction outcomes. The miR33/ sirtuin 1 pathway have been demonstrated to play a part in promoting inflammatory and coagulation of coronary thrombi in STEMI patients during hyperglycaemia [
20,
21]. In this context, whether metformin in diabetic patients with STEMI undergoing coronary angiography should be discontinued, has been discussed because of its post-procedural risks, including CI-AKI and MALA. Advice on the discontinuation of metformin differs between guidelines [
7‐
10,
22]. European Society of Urogenital Radiology guideline recommends stopping metformin directly from the time of contrast media administration [
10], while European Society of Cardiology guidelines recommends checking renal function after angiography for at least 3 days and withhold metformin when renal function is deteriorated [
8]. There are few studies about metformin used in patients with mildly impaired kidney function after the administration of contrast agents. In the two latest randomized controlled studies, the patients continuing metformin during peri-angiography does not carry the excess risk for renal dysfunction. No lactic acidosis is observed in both studies [
23,
24]. Currently, the influence of metformin on the kidney function in STEMI patients is widely discussed. The GIPS III trial has supported the idea that metformin is safe to use after STEMI and contrast agent exposure. Non-diabetic patients started metformin therapy within 3 h after the coronary intervention had no harmful effect on kidney functions [
11]. In addition, a multi-centre observational study has shown that chronic metformin treatment before PCI has no significant effect on CI-AKI in T2DM patients with STEMI [
2]. Those studies strongly suggested that metformin is not related to an increased risk of renal dysfunction after coronary angiography. The hypothesis that AMI patients may use metformin safely during the peri-angiography period was further reinforced. Our research is consistent with the above studies. The present data indicated that both absolute and relative creatinine change after PCI were similar between the patients continuing metformin therapy and those suspending metformin therapy. By multivariate analysis, metformin was not related to CI-AKI, whereas contrast volume (
p = 0.002) and eGFR
< 60 ml/min/1.73 m
2 (
p < 0.025) were indicated to be predictive factors of CI-AKI.
Several studies have demonstrated that the contrast agent volume is associated with the morbidity of acute kidney injury [
25]. The nephrotoxicity of iodinated contrast media may be proportional to the dose for coronary angiography. The main causes of CI-AKI have been proposed, including renal medullary hypoxia caused by hemodynamic instability, oxidative stress and direct toxicity on kidney tubular epithelial cells [
26‐
28]. It was important to understand that using a lower dose of contrast agent may substantially reduce the CI-AKI risk of patients. Patients with chronic kidney disease have fewer nephron units than normal, so exposure to the same volume of contrast media will significant increased proportionally. Because of their low adaptive capacity and increased contrast agent exposure, they are more susceptible to develop CI-AKI [
27].
As metformin is eliminated by the kidneys, there are concerns that in patients with the reduced kidney function, the lactic acidosis will be accumulated and precipitated [
29]. However, the strength of the relationship between metformin and lactic acidosis has been dramatically overstated [
4]. Several clinical studies have shown that there is no significant correlation between metformin concentration and lactic acidosis [
1,
4,
30‐
32]. Furthermore, the current study and meta-analyses show that the morbidity of lactic acidosis using metformin is not significantly different from other hypoglycemic treatments, such as sulfonylureas, insulin, and other oral hypoglycemic agents [
33‐
39]. Other studies have also shown that metformin concentrations remain in a therapeutic range in mildly to moderately renal impaired patients [
40]. At the same time, growing evidence suggests that the underlying disease associated with the tissue hypoxia rather than metformin use is related to lactic acidosis in diabetes [
41‐
43]. In the present research, no case of lactic acidosis was observed during hospitalization for both groups.
Hyperglycemia has been linked with more complications during hospitalization and poor outcomes in AMI patients. Lazzeri et al. [
44] concluded that in STEMI patients, in-hospital peak glycemia is negatively correlated with long-term survival. Besides that, another study had also confirmed that acute hyperglycemia is a predictor of CI-AKI and in-hospital mortality [
45]. Peri-procedural tight glycemic control has been shown to significantly increase the area of myocardial salvage following a great recovery of left ventricular function in hyperglycemic patients undergoing emergency coronary intervention for STEMI. These observations strongly suggest that the tight glycemic control at the time of the PCI may be pursued in the STEMI patients to improve their prognosis [
46]. Therefore, strict glucose management in STEMI patients with mildly impaired renal function is recommended during a hospital stay [
47]. In this research, we found the patients who discontinued metformin treatment, were more likely to initiate insulin therapy and had higher peak glycemia. Our data indicated that the patients who stopped metformin were inclined to have blood glucose fluctuation and changes in hypoglycemic regimens after admission.
Unlike previous studies, our study focused on the effect of perioperative use of metformin on renal function, with one group continued metformin treatment after primary PCI and another discontinued it > 48 h after the primary PCI, which is different from Zeller’s [
2] work, and this is a major contribution of our work. The application of metformin in the peri-angiography period in the existing guidelines is not yet consistent; our study focused on this situation and target at patients who are inclined to develop CI-AKI. The result of the study indicated that continued use of metformin did not impair renal function compared to discontinuation of metformin during primary PCI period. In real clinical practice, we found that discontinuous use of metformin caused problems in blood glucose management in STEMI patients after admission. The patients who stopped metformin were more likely to have blood glucose fluctuation and changes in hypoglycemic regimens after PCI. The peak values of fasting and postprandial blood glucose in patients with discontinuous metformin treatment were significantly higher than those of patients receiving metformin continuously. Previous studies have confirmed that peak glycemia during hospitalization is negatively correlated with the long-term survival in diabetic STEMI patients [
39]. To summarize the above findings, continuous use of metformin during coronary angiography may not raise the risk of CI-AKI, and the blood glucose of patients after admission will be better controlled, which is conducive to the prognosis of STEMI patients.
This study had the following two limitations. Firstly, it was a retrospective cohort study, conducted at a single centre, based on a relatively small size of populations. Secondly, since the high-risk patients such as those who needed intra-aortic balloon pump, or had respiratory failure were excluded, the results in this paper may not be adapted to these subgroups of patients. However, it was believed that the findings are of clinical significance in most patients.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.