A Nomogram Based on Lactate to Predict Major Adverse Cardiovascular Events for Acute Coronary Syndrome after Percutaneous Coronary Intervention

Accurate prediction of major adverse cardiovascular events (MACE) is very important for the management of acute coronary syndrome (ACS) patients. We aim to develop and validate effective prognostic nomogram for individualized risk estimate of MACE in patients with ACS after percutaneous coronary intervention (PCI). We conducted a prospective assessment of patients with ACS after PCI from January 2013 to July 2019 (n = 1986). Based on the training set, single-factor and multi-factor Cox proportional hazard analysis method was used to determine the results of single-factor and multi-factor Cox proportional hazard analysis. The receiver operating characteristic (ROC) and calibration curve were used to evaluate the prediction accuracy and discriminability, we have compared nomogram with the classical cardiovascular risk scores. In the validation set, X-tile analysis and Kaplan-Meier curve were used to evaluate the value of clinical application.


Abstract
Background Accurate prediction of major adverse cardiovascular events (MACE) is very important for the management of acute coronary syndrome (ACS) patients. We aim to develop and validate effective prognostic nomogram for individualized risk estimate of MACE in patients with ACS after percutaneous coronary intervention (PCI).

Methods
We conducted a prospective assessment of patients with ACS after PCI from January 2013 to July 2019 (n = 1986). Based on the training set, single-factor and multi-factor Cox proportional hazard analysis method was used to determine the results of single-factor and multi-factor Cox proportional hazard analysis. The receiver operating characteristic (ROC) and calibration curve were used to evaluate the prediction accuracy and discriminability, we have compared nomogram with the classical cardiovascular risk scores. In the validation set, X-tile analysis and Kaplan-Meier curve were used to evaluate the value of clinical application.

Results
Independent prognostic factors included lactate, age, left anterior descending branch (LAD) stenosis ≥ 50%, right coronary artery (RCA) stenosis ≥ 50%, brain natriuretic peptide (BNP), and left ventricular ejection fraction (LVEF). The area under the ROC curve (AUC) of the training group were about 0.712 to 0.762. In the validation set, the nomogram still shows good differentiation (AUC were about 0.724 to 0.818). On the calibration plot, the predicted values of the statistical chart agree well with the actual observed values. In addition, participants can be divided into two different risk groups (low and high) according to the nomogram.

Background
According to the statistical results of the World Health Organization, coronary artery diseases such as acute coronary syndrome (ACS) have become one of the most important causes of death worldwide and also one of the diseases with the greatest social burden. [1] ACS includes acute ST-segment elevation myocardial infarction (STEMI), non-ST-segment elevation myocardial infarction (NSTEMI), and unstable angina (UA). [2]Percutaneous coronary intervention (PCI) remains the preferred treatment for ACS. [3] [4]However, the incidence of major adverse cardiovascular events(MACE) was approximately 5-6% at an average of 3.2 years after PCI, and about 29.8% after 10 years. [5] [6]The incidence of MACE in patients with different risk factors was vary.
A reliable method for predicting the risk of MACE may be valuable for selecting high-risk patients for new or aggressive treatment and for risk counseling with patients. A series of studies have shown that age, cardiac function indicators, coronary artery severity and laboratory testing indicators are independent prognostic factors for cardiovascular events to identify high-risk patients. [7,8]  [24] [25] However, their indicators such as heart rate, systolic blood pressure, myocardial enzyme and creatinine are dynamic. Killip class and ST-segment depression require the judgment of experienced physicians. These reasons may lead to a deviation in the prediction score. Another question, their maximum prediction time is generally not more than one year, thus it is unable to say what the long-term prognosis would be.
New and reliable biomarkers need to be added to prognosis models. Several recent studies have shown in disease states, lactate is independent prognostic factor to be useful for identifying patients at high-risk.
[26] The increase of lactate concentration is the secondary cause of anaerobic glycolysis caused by tissue hypoperfusion, hypoxia or both, in addition, stress hyperlactataemia is actually due to increased production of aerobic lactate. [27]Multifactor analysis of a latest study showed that for patients with ACS complicating refractory cardiogenic shock (CS) or refractory cardiac arrest (CA), level of lactate at extracorporeal life support implantation was the only independent predictor of survival.
[28]Another study showed that the peak of lactate under extracorporeal membrane oxygenation (ECMO) in the rst 24 h predicted 30-day mortality in patients with ACS complicated with CS and CA. [29]Therefore, the purpose of this study was to develop and validate a nomogram based on lactate for predicting short-term, midterm and long-term MACE in patients of ACS after PCI.

Patients selection
The prospective study was approved by the Ethics Committee of the First A liated Hospital of Wenzhou Medical University, eliminating the need for informed consent. From January 2013 to July 2019, a total of 2,465 patients in the Cardiovascular Department of the First A liated Hospital of Wenzhou Medical University were diagnosed with ACS and treated with PCI.
Inclusion criteria were: (1) coronary artery disease con rmed by coronary angiography;(2) symptoms of myocardial ischemia; (3) electrocardiographic changes consistent with ACS. Exclusion criteria were: (1) chronic coronary syndrome;(2) Tumor history; (3) signi cant comorbidity, trauma, or surgery; (4) incomplete follow-up data. According to these inclusion and exclusion criteria, 1,986 patients were included in the study. Patients followed for 4 years were randomly divided into a training set (n = 1324) and a validated set (n = 662) based on a computer-generated randomly generated allocation sequence. All methods are carried out in accordance with approved guidelines.
Clinical Outcomes De nitions MACE is de ned as the end point of this study, which refers to all-cause mortality, clinically driven revascularization of target lesions, new or recurrent myocardial infarction, and Ischemic cerebral infarction.
Collection of demographic, clinical, and follow-up data All of the patient's study data was extracted from the electronic medical record system. Demographic data included sex and age. Clinical indicators including left anterior descending branch (LAD) stenosis ≥ 50%, left circum ex artery (LCX) stenosis ≥ 50%, right coronary artery (RCA) stenosis ≥ 50%, Three vessel disease(LAD, LCX and RCA all stenosis ≥ 50%), serum lactate level, brain natriuretic peptide (BNP) level, estimated glomerular ltration rate (EGFR), serum creatinine level, hemoglobin (HB), serum uric acid level, left ventricular ejection fraction (LVEF), hypertension, diabetes, peripheral artery stenosis, atrial brillation, prior stroke, kidney disease. The stenosis of LAD, LCX and RCA was determined by coronary angiography during hospitalization. All laboratory tests and auxiliary examination were performed 7 days before coronary angiography. Lactate and BNP were the maximum during hospitalization. EGFR was calculated by Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation. [30]LVEF was obtained by echocardiography. Regular medical follow-up data were obtained by telephone or clinic visits. Patients with the training set and the validation set were followed up until 1 April 2020.

Statistical analysis
Median (1st quartile, 3rd quartile) were utilized to describe the characteristics of continuous variables, comparisons in the two sets were carried out with Mann-Whitney U test. The categorical variables were expressed in frequency (proportion) and compared using the Chi-square test or the Fisher's exact test.
Univariate and multivariate Cox proportional hazards regression models were used to screen potential prognostic factors and estimate their weights. Multivariate Cox proportional hazard analysis was performed using forward step: LR. Results are reported as hazard ratios (HRs) and 95% CIs. Clinical variables with the P-value of ≤ 0.05 were included in the model. The identi ed variables based on the results of multivariate analysis were incorporated to construct the nomogram to predict the risk of 6month, 1-year and 4-year MACE after PCI using statistical software (rms in R, version 3.6.2; http://www.rproject.org). With the input of independent risk factors, the nomogram outputs a risk score for each patient.
To evaluate discrimination power of the nomogram, we calculated the area under the curve (AUC) of the time-dependent receiver operating characteristic (ROC) for both the training and validation sets of 6month, 1-year and 4-year, then compared it with Grace risk score and CADILLAC risk score. And we evaluated the predictive ability the nomogram by plotting calibration curves. In addition, we analyzed the possibility of nominal mapping for MACE risk strati cation in patients after PCI.
All data management and statistical analysis were performed using SPSS 20.0 and MedCalc 19.0.5 for windows. R 3.6.2, X-tile 3.6.1 and MedCalc 19.0.5 were used for analysis and mapping results. All tests were double-sided P 0.05 for the signi cant level.

Role of the funding source
This research has been funded by the National Natural Science Foundation of China (No. 81873468). The sponsor (ZH) has played a role in the research design and review.

Baseline characteristics of patients and outcomes
A total of 1,986 patients of ACS treated with PCI were included in this study. The training set comprised 1,324 patients with 662 patients in the validation set. Baseline characteristics of patients in the training and validation set are shown in Table 1. The baseline characteristics were similar between the two cohorts, except for gender. The male rate of the training set was higher than that of the validation set(81.3% vs. 76.4%, P = 0.012). Nomogram screening depending on the training set According to the single factor survival analysis, a total of 15 factors statistically signi cant in the single factor survival analysis. (Table 2). According to the multivariate. The multivariate Cox regression analysis indicated age, LAD stenosis (≥ 50%), RCA stenosis (≥ 50%), lactate, BNP and LVEF as independent prognostic factors in the training dataset (P < 0.05), and were used to construct the nomogram. (Fig. 1) Each predictor corresponds to a speci c point by drawing the straight line upwards to the point axis (e.g. Age ≥ 75) and corresponds to 75 points on the integral axis. The sum represents the incidence of MACE, and a straight line is plotted down to the total point axis. We reported a case of 75 years old (75 points), The degree of LAD stenosis is 50% (78 points), The degree of RCA stenosis is 10% (0 point), lactate 1 mmol/L (0 point), BNP 100 pg/ml(0 point) and LVEF 38%(58 points). The total score is 211 points, and the MACE rate after 6 months,1 year and 4 years is 3% ,4% and 30% respectively.
We compared the discrimination of the nomogram with that of other already available risk scores such as CADILLAC score and GRACE score in the training and validation sets. The time-dependent ROC curve was found to be consistently more favorable in both training and validation sets. (Table 3) The calibration curves for the MACE probability at 6 months, 1 year and 4 years after PCI showed favorable agreement between the predicted probability and actual observation, demonstrating good calibration of the nomogram (Fig. 3) Performance of the prognostic nomogram in stratifying risk The total prognostic scores calculated by the nomogram were categorized into two risk groups to predict MACE: 'low-risk' (score ≤ 285.1) and 'high-risk' (score > 285.1) based on the cut-off value calculated using the X-tile software (Fig. 4).
The Kaplan-Meier curves for both sets clearly show that nomogram is stable in differentiating between high-risk and low-risk patients (Fig. 5). The HR for 'high-risk' category was found to be 4.11 (95%CI,3.08-5.49) compared to the 'low-risk' category in the training set and 4.01 (95%CI,2.68-6.00) in the validation set.

Discussion
The long-term clinical outcomes of ACS patients after PCI vary, so an accurate predictive model is required for identi cation. Risk prediction can help clinicians identify high-risk groups, guide follow-up and individualized treatment. In addition, the prediction contributes to the development of health care and clinical guidelines for ACS. Nomogram is evidence-based and fully personalized tool to regulate clinical decision-making and provides patient friendly, accurate and repeatable predictions without the need for computer software to interpret. [31] Therefore, we have developed and validated a nomogram with satisfying stability and accuracy. The most important 6 factors-lactate, age, LAD stenosis, RCA stenosis, BNP and LVEF-contained most of the prognostic information has been included.
Harjola et al. found lactate level ( 2 mmol/L) independently associated with an increased short-term mortality in patients with cardiogenic shock [32] A meta-analysis showed a greater reduction in lactate concentrations in survivors than in non-survivors, whether following cardiac surgery, cardiogenic shock, or cardiac arrest. [33] In STEMI patients, higher lactate levels were independently associated with 30-day mortality and overall adverse reactions to PCI (in particular, lactate ≥ 1.8 mmol/L). [34] Besides, in a study of 1,865 patients with ACS, elevated lactate levels (≥ 1.8 mmol/L) at admission were an independent predictor of 30-day and 180-day all-cause mortality.
[26] In terms of energy supply for heart,in a normal heart, at rest, β-oxidation of fatty acids provides about 60%-90% of energy while pyruvate produces 10%-40%. [35] Lactate produced by dehydrogenation of pyruvate which synthesized from glycolysis, is also an important fuel for the stressed heart.
[36] [37] During exercise, the uptake and use of lactate in the myocardium increases, as does the stimulation of β-adrenergic stimulation and shock. [27] Hyperlactatemia can be seen as part of the stress response, including increased metabolic rate, sympathetic nervous system activation, accelerated glycolysis, and improved bioenergy supply. [38] Hyper-lactate after ACS may be caused by hypoxia following hemodynamic disorders or by catecholamine-induced aerobic glycolysis in response to stress. [27,39] These studies suggest that lactate may play an important role in the course of ACS. To the best of our knowledge, however, there has been no such risk prediction tool for MACE containing lactate so far. Therefore, it is signi cant to set up this risk model.
For the other ve variables, TIMI ow grades reported signi cant coronary stenosis as an independent predictor. [24] In a study of 6,755 patients after PCI, Iqbal et al. found that in patients with multivessel disease, untreated proximal LAD and RCA were associated with increased mortality. [7] BNP level was a strong independent predictor of short-term postoperative mortality.
[8] Grabowski et al. have improved their predictive power by adding BNP to the Killip class and TIMI ow grades. [40] The possible explanation is that the elevated BNP level re ects a larger infarct size and progressive left ventricular remodeling, thus more obviously re ecting the degree of cardiac insu ciency. [41] The same to BNP, LVEF also serves as a reference index for cardiac function, to supply important prognostic information and should be included in approaches for stratifying risk after myocardial infarction. [9] Many studies have reported that age is a signi cant risk factor for clinical events (cardiac death, target vessel myocardial infarction, and clinically driven target vessel revascularization) after PCI. [10] [42] The predictive ability of simple age cutoff points of 65 and 75 are similar to that of a more complex model age as a continuous variable. [24] The COX regression analysis results of our study consistent with the above results, that is, lactate, LAD stenosis, RCA stenosis, BNP, LVEF and age were important predictors. In order to overcome or avoid the limitations of a single predictor and achieve high prediction accuracy, we combined six detected predictors to construct a nomogram model. Because of dynamic, the nomogram did not include clinical symptoms and signs, such as Killip class, heart rate, and systolic blood pressure, which are signi cantly associated with ACS mortality. TIMI Risk Score, published in 2000, predicted primary end point (all-cause mortality, MI, or severe recurrent ischemia requiring urgent revascularization) through 14 days after randomization for UA/ NSTEMI. [44]GRACE risk score has been established to predict the risk of death during hospitalization and at 6 months in patients with ACS. [11] To predict 30-day and 1-year mortality risk after PCI in AMI,PAMI risk score and CADILLAC risk score established successively. [21] [22] Several studies has proved that in predicting the 30-day and 1-year mortality, CADILLAC risk score showed slight superiority than Grace, TIMI, and PAMI risk scores. [45] The probable reason is CADILLAC risk score considers LVEF and three-vessel disease.
[46] Our nomogram also takes LVEF and speci c coronary angiography results into account. Furthermore, data from both our training set and validation set con rmed that nomogram was superior in predicting the MACE in ACS patients after PCI than the above several risk scores.
In order to achieve better results in the actual prediction, we used the nomogram to calculate the total score of MACE risk. ACS Patients after PCI can be classi ed into the "high risk" group (score ≥ 285.1) and the "low risk" group (score < 285.1) based on the cutoff values determined by X-tile analysis. Kaplan-Meier analysis showed that the incidence of MACE was statistically different between the two groups. It helps more accurately monitoring of high-risk patients to personalized health management and increase costeffectiveness.
Several strengths could be found in the study. In the past, risk scores were almost based on western populations, while the population of patients with ACS after PCI in the East, especially in China, was much larger, requiring a specialized prediction model. Our nomogram uses the latest clinical data from the past 7 years to re ect current cardiovascular medical standards. Between PCI, drug therapy, and coronary bypass surgery, our study evaluated patient outcomes solely treated with PCI, with fewer uncontrolled variables and more stability and accuracy. Our nomogram has combined an independent and new risk factor, lactate, that is an easily accessible indicator. Unlike traditional forecasting models, a follow-up period of up to 4 years, which is conducive to the evaluation of both short-term and long-term prognosis. The most attractive aspect of the nomogram is its good discrimination and calibration power.
Limitations also existed in this study. Our data came from the same medical center. Independent external validation is required to con rm the performance of the nomogram before the clinical application. Although lactate has certain predictive ability, the detection time and collection method of lactate are not uni ed and clear. Some clinical drugs may cause changes in lactate without the improvement of the prognosis.

Conclusions
In conclusion, a novel prognostic nomogram based on lactate with other ve easily determined and objective variables were developed and validated to predict short-term and long-term MACE in ACS patients after PCI. Availability of data and materials The datasets used and/or analyed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.

Funding:
This research has been funded by the National Natural Science Foundation of China (No. 81873468).
Authors' contribution:  Nomogram for predicting MACE for patients with ACS after PCI. Points are assigned for age, LAD stenosis ≥ 50%, RCA stenosis ≥ 50%, lactate, BNP, and LVEF. The score for each value is assigned by drawing a line upward to the points line, and the sum of the six scores is plotted on the total points line. Finally, the probability line to determine the probability of MACE.

Figure 2
Time-dependent ROC of the nomogram, CADILLAC score and GRACE score for 1-year MACE. AUC, Area under the curve; CADILLAC, The controlled abciximab and device investigation to lower late angioplasty complications; CI, Con dence interval; GRACE, The global registry of acute coronary events; ROC, Receiver operating characteristics. X-tile analysis of the total risk score in the training set and cut-off value.