On the prognostic value of post-load glucose in patients with coronary artery disease

This editorial refers to ‘Adjustment of the GRACE score by 2-hour post-load glucose improves prediction of long-termmajor adverse cardiac events in acute coronary syndrome in patients without known diabetes’†, by S. Chattopadhyay et al., on page 2740. In this issue of the European Heart Journal, Chattopadhyay et al.1 demonstrate the importance of 2 h post-load glucose (2hPG), derived form an oral glucose tolerance test (OGTT) when added as a continuous variable to the Global Registry of Acute Coronary Events (GRACE) score, in improving the prediction of death and recurrent non-fatal myocardial infarction in survivors of an acute coronary syndrome without known diabetes mellitus (DM). In contrast, fasting plasma glucose (FPG) did not increase the predictability of prognostic models containing the GRACE score. This score, based on a multitude of variables (e.g. age, gender, heart rate, systolic blood pressure, creatinine, congestive heart failure, history of myocardial infarction, ST-segment depression, elevated troponin, and in-hospital revascularization), is well established in helping guide clinicians regarding the type and intensity of therapy;2 but, as the authors rightly note, thus far does not contain glycaemic measures and accordingly it is unknown whether doing this could improve risk predictability. A first indication that 2hPG is a prognostic indicator for future cardiovascular events in patients with acute coronary events came from the Glucose in Acute Myocardial Infarction study.3 However, this was a small study with relatively few events and the results were in need of further confirmation. Within the auspices of European Action on Secondary and Primary Prevention by Intervention to Reduce Events (EUROASPIRE) IV, Shahim et al.4 recently reported data on cardiovascular events in 3775 patients with stable coronary artery disease (CAD) investigated at baseline with an OGTT and followed over an average of 2 years. The primary endpoint, defined as the first of cardiovascular death, non-fatal myocardial infarction or stroke, and hospitalization for heart failure, was reached by 246 patients (6.5%). Neither FPG nor glycated haemoglobin A1c (HbA1c) predicted the outcome, while the 2hPG dichotomized below or above 7.8 mmol/L, i.e. below or above the threshold for impaired glucose tolerance (IGT), was a significant predictor (adjusted hazard ratio 1.38, 95% confidence interval 1.07–1.78; P = 0.01). Adjustments were made for age and sex, and relevant variables including education level, current smoking, body mass index, systolic blood pressure, LDL-cholesterol, statin use, level of physical activity, and Hospital Anxiety and Depression Scale (HADS) anxiety and depression score. These results were consistent with the Silent Diabetes study, comparing the prognostic capacity of HbA1c with that of an OGTT in 1015 patients without previously known DM undergoing coronary angiography, which reported that a post-load glucose was closely related to the severity of CAD and mortality over 3 years and, in this respect, superior to FPG, whereas there was no association with HbA1c.5 In the present publication, the authors performed a retrospective cohort study on 1056 patients, who were discharged alive after an acute coronary syndrome and did not have previously known DM. They were subjected to an OGTT before hospital discharge and divided into quartiles according to their 2hPG (Q1 ≤ 6.5 mmol/L, Q2 6.6–8.1 mmol/L, Q3 8.2–10.4 mmol/L, and Q4 > 10.4 mmol/L). During a median follow-up of 3.4 years, there were 235 events (deaths, n = 112 or reinfarctions, n = 123). An increasing 2hPG—not, however, FPG—was an independent prognostic predictor even after adjusting for the GRACE score and the 2hPG significantly improved the accuracy of different models including the GRACE score in predicting long-term outcome. Of importance is that the 2hPG value was used as a continuous, not categorical, variable and that the contribution of the 2hPG to the GRACE score in improving prediction was rigorously assessed by various statistical approaches. Not surprisingly, the majority of patients did have dysglycaemia when classified according to categorical glycaemic terms, i.e. 370 out of 1056 (35%) had IGT and 229 of 1056 (22%) had undiagnosed DM. Unfortunately, HbA1c measurements were not available for this cohort, but it is unlikely that this test would have conferred significant further prognostic information. The presence of dysglycaemia, i.e. type 2 DM (T2DM) and IGT, has a negative prognostic implication for patients with CAD.3,4,6 Unfortunately it is not uncommon that an acute coronary event is the first manifestation of dysglycaemia, as also seen in the present publication, underlining the importance of screening for such conditions, as recommended by European guidelines.7 As demonstrated by EUROASPIRE IV, this recommendation is far from fulfilled across a number of European countries.8 This contrasts with recent achievements that have increased the importance of a timely and accurate identification of dysglycaemia in patients with CAD. As reviewed by Standl et al.,9 the institution of either an Sodium Glucose Transporter-2 (SGLT-2) inhibitor or a Glucagon Like Peptide-1 (GLP-1) receptor agonist in patients with established CAD and DM offers an opportunity to reduce major coronary events by about 14%, above that accomplished by means of traditional secondary prevention through coronary interventions, platelet stabilization, and blood pressure and lipid-lowering drugs. Accordingly, identifying patients suitable for these new drugs, i.e. with hyperglycaemia, is truly warranted, as depicted in Figure 1. A reasonable algorithm for dysglycaemia screening in low- and high-risk populations. See text for further explanation. ACS, acute coronary syndrome; FINDRISC, Finnish Diabetes Risk Score; GRACE, Global Registry of Acute Coronary Events. A reasonable algorithm for dysglycaemia screening in low- and high-risk populations. See text for further explanation. ACS, acute coronary syndrome; FINDRISC, Finnish Diabetes Risk Score; GRACE, Global Registry of Acute Coronary Events. Why are large parts of the cardiological community refractory to dysglycaemia screening of their patients? One reason may be that the relationship between hyperglycaemia and macrovascular disease is considered unclear, an assumption that gained support from clinical trials with older glucose-lowering drugs without showing evidence of cardiovascular benefit.7,9 In light of the mortality and/or cardiovascular event reducing capacities of newer drugs, this is no longer an acceptable argument against dysglycaemia screening. Even if the beneficial impact of the (as glucose-lowering drugs launched) SGLT-2 inhibitors empagliflozin and canagliflozin, and the GLP-1 receptor agonists lira- and semiglutide, are related to other pleiotropic effects rather than glucose control per se, the cardiovascular benefits are there.9 Another reason for the screening inertia may be uncertainties about accurate and feasible tests, as reflected in an ongoing debate regarding the best way to identify T2DM in the patient group involved: FPG combined with 2hPG obtained from an OGTT or eventually HbA1c.10,11 The OGTT is criticized for demanding overnight fasting and for taking somewhat more than 2 h to implement. However, both these objections must be considered obsolete if diagnostic and prognostic information derived by this test is superior to that obtained by an FPG and/or HbA1c. A reasonable question with this in mind would be: what is 2–3 h in the life of a patient if there are options available for preventing future complications including cardiovascular death, reinfarction, and heart failure? There is ample evidence that the information derived by an OGTT (FPG + 2hPG) identifies the largest number of undiagnosed patients with DM and that it is the only test that discloses the presence of IGT.12 In EUROASPIRE IV, Gyberg et al.7 screened 4004 patients with CAD without known DM. Using all screening tests together, 1158 (29%) had undetected DM. Of those patients, the proportion identified by FPG was 75%, by 2hPG 40%, by HbA1c 17%, by FPG + HbA1c 81%, and by OGTT (= FPG + 2hPG) 96%. When 4422 patients with CAD without known glucose perturbations were screened within EUROASPIRE V, 41% were dysglycaemic (IGT=25% and T2DM=16%). The proportion with T2DM identified by HbA1c was 19%, by FPG 58%, and with OGTT 91%. Thus, HbA1c only contributed a small proportion above that disclosed by FPG + 2hPG.13 These data favor the use of an OGTT if the ambition is to disclose as many dysglycaemic patients as possible. A prerequisite is of course that this test also contributes prognostic information of importance. Following the reports by Shahim et al. in patients with stable,4 and by Chattopadhyay et al. (this issue of the European Heart Journal) in patients with acute, coronary syndromes, this should be considered as established (Figure 1). Together with existing information, the investigation by Chattopadhyay et al. strongly supports the use of an OGTT and the incorporation of the 2hPG value in GRACE score-based risk prediction following an acute coronary event or, in principle, in all patients with CAD without known dysglycaemia. It has been claimed that in the presence of a diagnostic FPG and/or HbA1c further screening efforts are not required. As outlined, this seems to be an outdated statement. Among the presently available screening tests, HbA1c is the least suitable, leaving a majority of people with dysglycaemia undetected and devoid of prognostic information beyond what is available from the clinical scenario. An FPG will disclose up to 75% of patients with T2DM, but none of those with IGT, who also have a compromised prognosis of a severity that is almost the same as those with newly detected T2DM.3,6 In addition, an FPG does not seem to add independent prognostic information. The investment in the time and effort required to perform an OGTT seems without doubt worthwhile, both from the patients’ and the cardiologists’ perspective. The accumulating evidence should encourage those in charge of patients with CAD to understand the importance of screening and to use the most informative test for this purpose, i.e. an OGTT.