INTRODUCTION

Currently, stratification of surgical risk in cardiac surgery is mainly evaluated through different risk scoring systems, such as the European System for Cardiac Operative Risk Evaluation (EuroSCORE) and the Society of Thoracic Surgeons (STS) scores ^[1-3]. However, these scores use variables collected preoperatively and do not consider multiple prognostic variables that are important in the early postoperative period. Scoring systems that are commonly used in general intensive care units, such as the Sequential Organ Failure Assessment and the Acute Physiology and Chronic Health Evaluation II, may be an alternative for determining the risk of patients in cardiac surgical care units ^[4,5]. Nevertheless, most of risk scores were not validated in special units. Specifically, cardiac surgery patients have been excluded from the development studies of general predictive scoring systems ^[6].

In relation to postoperative biomarkers, lactic acid and mixed venous oxygen saturation are classic predictors with prognostic value in patients undergoing cardiac surgery ^[7-10]. However, these isolated markers fail to predict the real patient’s risk. Hyperglycemia in the immediate postoperative period is also associated with increased morbidity and mortality ^[11]. In addition, there is evidence that after the use of cardiopulmonary bypass in cardiac surgery, plasma concentration of leukocyte mediators increases, contributing to the systemic inflammatory response syndrome and worse prognosis ^[12,13].

The leuko-glycaemic index (LGI) is an index that combines white blood cell count and blood glucose, calculated by multiplying both values and dividing them by a thousand. This index could be a marker of systemic inflammatory response syndrome and has already shown to be a good predictor of events in various scenarios of critical care medicine, such as acute myocardial infarction and stroke ^[14-19]. Seoane et al. ^[20] have demonstrated its prognostic value for the first time in cardiac surgery in 2017. However, LGI was evaluated in a heterogeneous population since it included different types of cardiac surgeries. Furthermore, high LGI was not associated with higher in-hospital death.

The aim of this study was to analyze the prognostic value of LGI in the postoperative period of coronary artery bypass grafting (CABG).

METHODS

RESULTS

The study evaluated 3,813 patients undergoing CABG, most of them male (68.5%). The mean age was 64.6 (SD 9.6) years. Regarding cardiovascular risk factors, 77% had arterial hypertension and 29% had diabetes. In relation to the presence of other comorbidities, 41% had prior acute myocardial infarction, 9% had history of atrial fibrillation, 16% had severe left ventricular dysfunction, 5% had chronic obstructive pulmonary disease, and 5% had chronic renal failure. The mean EuroSCORE II was 3.6 (SD 2.4). In the population analysis, according to LGI quartiles, there was no difference in the baseline characteristics, except for the highest percentage of diabetes in the highest quartile (Table 1).

Table 1 - Basal characteristics of the population according to LGI.

CABG procedures were mainly off-pump (88.8%) and elective surgeries (64.1%). Urgent procedures and number of bypasses were similar in each quartile of the population. Nevertheless, the use of cardiopulmonary bypass was higher in quartile 4 (Table 2).

Table 2 - Surgical and laboratory characteristics according to LGI.

The median LGI was 2035 mg/dl. mm³ (IQR 25-75%: 1641 - 2683). The rate of in-hospital mortality, LCO, or AKI was 11.4%. Presence of the primary endpoint significantly increased per LGI quartile (9.2%, 9.7%, 11.8%, and 15%; P<0.001) (Figure 1). The ROC curve value for this combined endpoint was 0.61 (95% CI 0.56 - 0.67). High LGI was associated with a statistically significant increased occurrence of in-hospital death, LCO, AKI, and postoperative atrial fibrillation (Table 3).

Table 3 - Primary and secondary outcomes according to LGI.

Fig. 1 - Incidence (%) of primary combined endpoint according to the leuko-glycaemic index (LGI). Q=quartile

Download

Fig. 1 - Incidence (%) of primary combined endpoint according to the leuko-glycaemic index (LGI). Q=quartile

The best prognostic cutoff value for the primary endpoint obtained by ROC curve was 2000 mg/dl. mm³. In a multivariate logistic regression model, high LGI was independently associated with in-hospital death, LCO, or AKI (OR: 1.50, 95% CI 1.21 - 1.87; P<0.001) (Table 4).

Table 4 - Multivariate logistic regression model for in-hospital mortality, LCO, or AKI.

DISCUSSION

In this large cohort of patients, high LGI was an independent predictor of in-hospital mortality, LCO, or AKI in the postoperative period of CABG. It was also associated with higher in-hospital death.

Hyperglycemia is very common in the postoperative period of cardiac surgery, affecting approximately 33% of patients, whether they are diabetic or not ^[22]. Furthermore, the immediate postoperative blood glucose was predictive of complications (prolonged in-hospital stay, higher infection rates, and increased postoperative morbidity), particularly in those patients with levels > 250 mg/dl causing a 10-fold increase in risk ^[23]. Thereafter, Doenst et al. ^[11] showed that hyperglycemia during cardiopulmonary bypass was an independent risk factor for mortality in patients undergoing cardiac surgery.

Regardless of hyperglycemia, diabetic patients after cardiovascular surgery present higher perioperative morbidity and mortality. Insulin deficiency and insulin resistance are aggravated by surgery and anesthesia and may lead to lipolysis, metabolic acidosis, electrolyte changes, and increased protein catabolism. Insulin administration is high among CABG patients and could reverse most of these metabolic disorders. However, Ranney et al. ^[24] showed that insulin use did not change prognosis in diabetic patients and that it was associated with a higher risk of adverse outcomes in patients without diabetes.

Non-diabetic patients also frequently present hyperglycemia in the postoperative period of cardiovascular surgery, which is a risk factor for mortality. In these patients, preoperative screening with Homeostasis Model Assessment of Insulin Resistance might provide an insight to the glucometabolic state of individuals before CABG, and could be a predictive value for postoperative outcomes ^[25].

The metabolic stress of the surgery increases levels of glucagon, growth hormone, and cortisol and is enhanced by prolonged fasting and decreased insulin levels. Within this response to trauma, insulin resistance and hyperglycemia are manifested, being more intense in the first postoperative days and in major surgeries. In 2012, Feguri et al. ^[26] showed that preoperative fasting abbreviation with administration of liquid enriched with carbohydrates in CABG was safe and improved the glycemic control in the intensive care unit. Afterwards, the same author described that this strategy added to intraoperative infusion of Omega-3 polyunsaturated fatty acids supported faster postoperative recovery in patients undergoing on-pump CABG ^[27].

Leukocytosis is also a frequent finding in cardiac surgery. Inflammation is produced by humoral and cellular interactions with numerous pathways and multiple mediators ^[12,28]. It is known that cardiopulmonary bypass induces a systemic inflammatory reaction, that may contribute to postoperative complications, including myocardial dysfunction, AKI, respiratory failure, bleeding disorders, altered liver function, and multiorgan failure ^[13,29]. Neutrophil to lymphocyte ratio and platelet to lymphocyte ratio are clinical parameters, which can be found in a simple hemogram and increase in systemic inflammation. These ratios were independent biomarkers directly related to development of AKI in the early postoperative period of isolated CABG. Parlar et al. ^[30] showed that increased postoperative first-day leukocyte count was also associated with AKI in these patients.

LGI is an index that combines white blood cell count and blood glucose levels. It is calculated by multiplying both values and dividing them by a thousand. It was first described in 2010 by Quiroga Castro et al. ^[14] to determine prognostic value in patients with acute myocardial infarction. In that study, LGI > 1600 mg/dl. mm³ was associated with in-hospital complications (acute myocardial infarction, heart failure, and post-infarction angina). Hirschson Prado et al. ^[15] also showed that high LGI was an independent predictor of poor evolution in acute myocardial infarction (in hospital death or Killip Kimball 3-4). Although it was a novel prognostic index, it was evaluated in a small number of patients. Afterwards, León et al. ^[16] and Rodriguez Jimenez et al. ^[17] also proved that high LGI in the course of an acute myocardial infarction could be associated with a higher in-hospital death. In addition, Caldas et al. ^[18] and García Alvarez et al. ^[19] showed that LGI was a risk marker for predicting mortality in patients with ischemic stroke.

We previously demonstrated its prognostic value for the first time in cardiac surgery in 2017. In that study, a large number of patients were evaluated (n=2,743) and different types of cardiac surgeries were included ^[20]. Although high LGI was associated with an increased occurrence of in-hospital death or LCO, a more heterogeneous population was evaluated. Furthermore, high LGI was not associated with higher in-hospital death or atrial fibrillation. In this study, more patients were evaluated (n=3,813), and only CABG surgeries were included, analyzing a more homogeneous population, mainly off-pump. Population characteristics and cutoff value of LGI for primary combined endpoint were similar in both studies. In contrast, in this study, high LGI predicted in-hospital death and atrial fibrillation as secondary events.

LGI, being determined by blood glucose levels and white blood cell count, could be a marker of systemic inflammation. Therefore, it is logically associated with poor outcomes in various scenarios of critical care medicine, including cardiac surgery. Even in this population with CABG mainly off-pump, in which systemic inflammatory reaction is lower, this index could predict in-hospital death, LCO, and AKI. Values of LGI > 2000 mg/dl. mm³ may be the best predictive values for the primary combined endpoint.

Concerning clinical relevance, LGI being a simple index available in any intensive care unit, easy to calculate, and low cost, could be a valuable prognostic tool to stratify patients in the immediate postoperative period of CABG. Those patients with high LGI values may benefit from greater monitoring and from using early therapeutic strategies to reduce LCO, AKI, and atrial fibrillation.

CONCLUSION

High LGI was an independent predictor of in-hospital mortality, LCO, or AKI in the postoperative period of CABG. It was also associated with a significant increased occurrence of in-hospital death, LCO, AKI, and atrial fibrillation. Due to these findings, the use of this simple and novel index could be an attractive tool to stratify risk in patients undergoing CABG.