Abstract
Background
Inflammatory and nutritional status may refine risk assessment before coronary artery bypass grafting (CABG). This study evaluated the association between preoperative Naples Prognostic Score (NPS) and early and follow-up outcomes after isolated CABG.
Methods
This single-center retrospective cohort included adults who underwent isolated CABG between 2020 and 2023. Patients without a recorded preoperative NPS were excluded. The recorded NPS (0–4) was categorized as low (score 0), intermediate (scores 1–2), or high (scores 3–4). Outcomes were institutionally recorded binary event indicators. Logistic regression evaluated associations with 30-day major adverse cardiac and cerebrovascular events (MACCE), follow-up MACCE, and follow-up all-cause death.
Results
The analytic cohort included 332 patients: 16 low-NPS, 193 intermediate-NPS, and 123 high-NPS patients. Follow-up all-cause death occurred in 25 patients (7.5%) and was more frequent in the high-, intermediate-, and low-NPS groups, respectively (15.4%, 3.1%, and 0.0%; exact P < 0.001). Any follow-up MACCE occurred in 45 patients (13.6%; 18.7%, 11.4%, and 0.0% across the same groups; exact P = 0.050). Thirty-day MACCE occurred in 11 patients (3.3%); its analyses were exploratory. Per 1-point increase in NPS, the adjusted odds ratio for follow-up all-cause death was 1.69 (95% CI 1.00-2.84; P = 0.048) in the clinical model and 2.06 (95% CI 1.25–3.39; P = 0.005) in the EuroSCORE II complete-case model.
Conclusions
Higher preoperative NPS was associated with greater follow-up all-cause mortality after isolated CABG. Associations with composite MACCE were less consistent after adjustment. NPS may support inflammatory-nutritional risk interpretation alongside established surgical risk assessment, but the findings require confirmation in cohorts with standardized follow-up and time-to-event data.
Graphical abstract
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1186/s12872-026-06427-y.
Keywords: Coronary artery bypass grafting, Naples Prognostic Score, All-cause mortality, Inflammation, Nutritional status, EuroSCORE II, Retrospective cohort
Background
Coronary artery bypass grafting remains a central revascularization strategy for patients with left main disease, multivessel coronary artery disease, diabetes, and complex coronary anatomy [1, 2]. Established surgical risk tools such as the European System for Cardiac Operative Risk Evaluation II (EuroSCORE II) incorporate clinical and procedural risk factors and have been validated in CABG populations [3–5]. These tools do not directly measure systemic inflammation or nutritional reserve, both of which have been associated with adverse cardiovascular and surgical outcomes [6–8].
The Naples Prognostic Score is a composite inflammatory-nutritional index based on serum albumin, total cholesterol, neutrophil-to-lymphocyte ratio, and lymphocyte-to-monocyte ratio [9]. The score was first described in surgical oncology and has since been evaluated in cardiovascular settings including ST-segment elevation myocardial infarction and transcatheter aortic valve implantation [10–12]. CABG-specific evidence is emerging, but independent cohorts remain useful for examining reproducibility under differing data structures and adjustment strategies.
This study evaluated the association between recorded preoperative NPS and postoperative event indicators in patients undergoing isolated CABG, with primary analytical emphasis on follow-up all-cause mortality and secondary, cautious interpretation of composite MACCE findings.
Methods
Study design and population
This single-center retrospective cohort study included adults who underwent isolated CABG between 1 January 2020 and 31 December 2023 at Bezmialem Vakif University Medical Faculty Hospital. The available source file contained 350 records dated within this period. Eighteen records without a recorded preoperative NPS were excluded, leaving 332 records in the analytic cohort (Fig. 1). The recorded NPS field, rather than post hoc reconstruction from its component variables, defined exposure availability. The cohort flow reflects the verifiable analytic file; detailed counts from any screening conducted before creation of that file were unavailable.
Fig. 1.
Cohort flow diagram. CABG, coronary artery bypass grafting; NPS, Naples Prognostic Score
The study is reported according to the STROBE statement. The STROBE checklist is submitted as a related file [13]. Missingness, model-completeness, and sensitivity-analysis details are summarized in Additional file 1.
Naples prognostic score
The analysis used the preoperative NPS score recorded in the source dataset. NPS is conventionally assigned as one point for each of serum albumin < 4 g/dL, total cholesterol < = 180 mg/dL, neutrophil-to-lymphocyte ratio > 2.96, and lymphocyte-to-monocyte ratio < = 4.44, yielding a score from 0 to 4 [9]. The recorded score was categorized as low (0), intermediate (1–2), or high (3–4). Missing recorded scores were not reconstructed from component fields. Regression models used the recorded 0–4 score as a continuous one-point increment; the high-NPS comparison used low and intermediate NPS combined as the reference group.
Covariates and operative variables
Clinical variables included age, sex, body mass index, hypertension, diabetes mellitus, recent myocardial infarction within 3 months, previous percutaneous coronary intervention, chronic kidney disease, dialysis, previous stroke, hyperlipidemia, chronic obstructive pulmonary disease, peripheral arterial disease, atrial fibrillation, and EuroSCORE II. Echocardiographic assessment included left ventricular ejection fraction, recorded according to routine preoperative transthoracic echocardiography practice [14]. Operative variables included cardiopulmonary bypass time, aortic cross-clamp time, distal anastomosis count, intensive care unit stay, and hospital stay. Intensive care unit and hospital length of stay were reported descriptively and were not included as covariates in outcome models because they may occur after perioperative complications.
The available dataset did not contain a discrete acute coronary syndrome (ACS) variable. Therefore, recent myocardial infarction within 3 months was used as the clinically available ischemic-presentation variable. Structured systolic pulmonary artery pressure values were not retained in the main table because the analytic field did not contain clinically interpretable values across the cohort.
Outcomes
The early outcome was 30-day MACCE, defined as the composite of all-cause death, stroke, myocardial infarction, or repeat revascularization within 30 days. Follow-up outcomes were binary indicators derived from the available institutional 1-6-month and > 6-month fields and included all-cause death and MACCE. Follow-up MACCE was recalculated once as the patient-level union of myocardial infarction, stroke, repeat revascularization, and all-cause death across both follow-up fields; a patient with events in both fields contributed once to the Any follow-up outcome. Event dates, censoring dates, a common closed observation horizon, external vital-status linkage, and a separate loss-to-follow-up flag were unavailable. Accordingly, these variables represent institutionally recorded events and are not cumulative risks, incidence rates, or survival outcomes. No additional telephone or outpatient follow-up verification or independent event adjudication was available. Perioperative myocardial infarction was interpreted using available clinical event coding and contemporary myocardial infarction definitions [15]. Blank component fields were coded as non-events under the locked analysis rule; because unknown and event-free status could not be distinguished, this assumption may underestimate events, including those occurring outside institutional capture. Cause-specific mortality was unavailable, and all mortality analyses therefore used all-cause death.
Statistical analysis
Continuous variables are presented as median (interquartile range), and categorical variables as number (percentage). Continuous variables were compared across NPS categories using nonparametric tests. Sparse 3-by-2 outcome comparisons used two-sided Fisher-Freeman-Halton exact tests. Logistic regression modeled the binary event indicators. All score-based models used the same recorded NPS scale (0–4), and odds ratios are reported per 1-point increase. Covariates were selected on clinical grounds: recent myocardial infarction within 3 months represented measured ischemic presentation, while age, left ventricular ejection fraction, and estimated glomerular filtration rate represented major preoperative risk domains. Because 30-day MACCE occurred in only 11 patients, no adjusted 30-day model was reported; unadjusted analyses are exploratory. Follow-up models used the NPS score alone, a clinical adjustment model, and a complete-case sensitivity model containing recent myocardial infarction and EuroSCORE II. EuroSCORE II was missing in 52 patients; its models were therefore qualified as complete-case sensitivity analyses, and characteristics of complete and incomplete cases are reported in Additional file 1. Multiple imputation was not used because EuroSCORE II is a derived composite and the source components needed for defensible reconstruction were unavailable. An exploratory follow-up sensitivity model added troponin above the cohort median to the clinical model. Exploratory combined EuroSCORE II and NPS event rates used a EuroSCORE II threshold of 2%. Complete-case discrimination analyses compared apparent AUCs for NPS score alone, EuroSCORE II alone, and both variables; bootstrap confidence intervals refitted each model in 2,500 resamples. These were descriptive, internally unvalidated analyses rather than time-dependent ROC analyses. Firth penalized logistic regression assessed follow-up mortality because the low-NPS group had zero deaths. P values were two-sided; estimates and confidence intervals were emphasized over threshold-based interpretation.
Results
Cohort and baseline characteristics
Among 350 patients who underwent isolated CABG during 2020–2023, 18 were excluded because NPS was missing, yielding an analytic cohort of 332 patients (Fig. 1). NPS was low in 16 patients, intermediate in 193, and high in 123.
Baseline characteristics are shown in Table 1. Median age was 62.0 years, and 61 patients (18.4%) were female. Patients with high NPS were older and had a greater burden of chronic kidney disease than the other groups. The high-NPS group had lower albumin, lower total cholesterol, higher neutrophil-to-lymphocyte ratio, lower lymphocyte-to-monocyte ratio, and lower estimated glomerular filtration rate. Median EuroSCORE II was numerically higher in the high-NPS group.
Table 1.
Baseline, laboratory, and operative characteristics according to NPS category. Continuous variables are median (interquartile range); categorical variables are n (%)
| Variable | Overall (N=332) | Low NPS (n=16) | Intermediate NPS (n=193) | High NPS (n=123) | P value |
|---|---|---|---|---|---|
| Age, years | 62.0 (56.0-67.2) | 58.5 (53.8-62.2) | 61.0 (56.0-67.0) | 65.0 (57.0-70.0) | 0.008 |
| Female sex | 61 (18.4) | 3 (18.8) | 40 (20.7) | 18 (14.6) | 0.395 |
| BMI, kg/m2 | 27.6 (25.3-30.6) | 27.4 (24.9-30.8) | 27.7 (25.3-31.1) | 27.2 (25.3-29.7) | 0.443 |
| Hypertension | 240 (72.3) | 11 (68.8) | 139 (72.0) | 90 (73.2) | 0.926 |
| Diabetes mellitus | 179 (53.9) | 8 (50.0) | 101 (52.3) | 70 (56.9) | 0.692 |
| Recent MI within 3 months | 115 (34.6) | 5 (31.2) | 61 (31.6) | 49 (39.8) | 0.311 |
| Previous PCI | 67 (20.2) | 2 (12.5) | 37 (19.2) | 28 (22.8) | 0.544 |
| Chronic kidney disease | 60 (18.1) | 0 (0.0) | 30 (15.5) | 30 (24.4) | 0.022 |
| Dialysis | 7 (2.1) | 0 (0.0) | 1 (0.5) | 6 (4.9) | 0.026 |
| Previous stroke | 9 (2.7) | 2 (12.5) | 1 (0.5) | 6 (4.9) | 0.003 |
| Hyperlipidemia | 206 (62.0) | 14 (87.5) | 131 (67.9) | 61 (49.6) | <0.001 |
| COPD | 23 (6.9) | 0 (0.0) | 12 (6.2) | 11 (8.9) | 0.347 |
| PAD | 56 (16.9) | 5 (31.2) | 29 (15.0) | 22 (17.9) | 0.232 |
| Atrial fibrillation | 92 (27.7) | 5 (31.2) | 50 (25.9) | 37 (30.1) | 0.684 |
| EuroSCORE II | 1.1 (0.8-1.7) | 1.0 (0.8-1.4) | 1.0 (0.7-1.5) | 1.4 (0.9-2.0) | <0.001 |
| Neutrophil, x10^3/uL | 5.2 (4.3-6.6) | 5.0 (4.2-6.0) | 4.8 (3.9-5.9) | 6.2 (4.9-8.1) | <0.001 |
| Lymphocyte, x10^3/uL | 2.2 (1.6-2.8) | 3.1 (2.5-4.2) | 2.4 (1.9-3.1) | 1.6 (1.2-2.3) | <0.001 |
| Monocyte, x10^3/uL | 0.6 (0.5-0.8) | 0.6 (0.4-0.7) | 0.6 (0.5-0.7) | 0.7 (0.6-0.8) | 0.008 |
| NLR | 2.2 (1.7-3.2) | 1.5 (1.3-1.9) | 1.9 (1.6-2.5) | 3.4 (2.7-5.4) | <0.001 |
| LMR | 3.7 (2.6-4.9) | 5.5 (5.1-6.7) | 4.1 (3.3-5.2) | 2.6 (1.9-3.5) | <0.001 |
| Albumin, g/dL | 3.5 (3.2-3.9) | 4.2 (4.2-4.5) | 3.6 (3.3-4.0) | 3.3 (3.1-3.6) | <0.001 |
| Total cholesterol, mg/dL | 194.2 (162.2-233.0) | 235.5 (218.4-263.5) | 212.0 (177.0-243.0) | 172.0 (146.0-196.5) | <0.001 |
| eGFR, mL/min/1.73m2 | 85.0 (67.0-95.0) | 90.0 (79.2-91.2) | 86.0 (70.0-97.0) | 78.0 (56.5-91.5) | 0.002 |
| Troponin, ng/L | 22.9 (5.0-840.1) | 11.1 (4.5-987.1) | 19.4 (3.9-388.8) | 54.1 (7.0-1927.2) | 0.052 |
| Glucose, mg/dL | 125.0 (101.0-181.2) | 135.5 (104.5-250.8) | 122.0 (98.0-175.0) | 127.0 (101.5-185.5) | 0.322 |
| LVEF, % | 58.5 (50.0-60.0) | 55.0 (50.0-60.0) | 60.0 (50.0-60.0) | 55.0 (45.0-60.0) | 0.034 |
| Cross-clamp time, min | 66.0 (55.0-80.0) | 70.0 (61.5-79.0) | 68.0 (56.0-81.0) | 64.0 (52.0-79.0) | 0.282 |
| Distal anastomoses | 4.0 (3.0-4.0) | 4.0 (3.8-4.0) | 4.0 (3.0-4.0) | 4.0 (3.0-4.0) | 0.100 |
| ICU stay, days | 3.0 (2.0-4.0) | 3.0 (2.0-4.0) | 3.0 (2.0-4.0) | 3.0 (2.0-4.0) | 0.754 |
| Hospital stay, days | 12.0 (9.0-15.0) | 12.5 (9.8-15.2) | 12.0 (9.0-15.0) | 13.0 (10.0-16.0) | 0.065 |
BMI Body Mass Index, COPD Chronic Obstructive Pulmonary Disease, eGFR Estimated Glomerular Filtration Rate, ICU Intensive Care Unit, LMR Lymphocyte-to-Monocyte Ratio, LVEF Left Ventricular Ejection Fraction, MI Myocardial Infarction, NLR Neutrophil-to-Lymphocyte Ratio, PAD Peripheral Artery Disease, PCI Percutaneous Coronary Intervention
Early and follow-up outcomes
Outcome frequencies are shown in Table 2. Thirty-day death occurred in 7 patients (2.1%), and 30-day MACCE occurred in 11 patients (3.3%). Thirty-day MACCE occurred in 1 low-NPS patient (1/16, 6.2%), 2 intermediate-NPS patients (2/193, 1.0%), and 8 high-NPS patients (8/123, 6.5%; exact P = 0.021). Because only 11 events were observed and the low-NPS group was small, this early composite outcome was interpreted as exploratory.
Table 2.
Early and follow-up outcomes according to NPS category. Follow-up MACCE was recalculated as the union of myocardial infarction, stroke, repeat revascularization, and all-cause death across the available institutional 1-6-month and >6-month fields. Blank component fields were coded as non-events under the locked analysis rule. Period-specific rows are not mutually exclusive, whereas each patient is counted once in the Any follow-up row. Fisher-Freeman-Halton exact P values are reported. Because a common follow-up horizon, censoring dates, external vital-status linkage, and a loss-to-follow-up flag were unavailable, follow-up values are institutionally recorded-event indicators rather than cumulative risks
| Outcome | Overall (N=332) | Low NPS (n=16) | Intermediate NPS (n=193) | High NPS (n=123) | P value |
|---|---|---|---|---|---|
| 30-day death | 7 (2.1) | 1 (6.2) | 1 (0.5) | 5 (4.1) | 0.019 |
| 30-day stroke | 4 (1.2) | 1 (6.2) | 0 (0.0) | 3 (2.4) | 0.022 |
| 30-day MI | 1 (0.3) | 0 (0.0) | 1 (0.5) | 0 (0.0) | 1.000 |
| 30-day revascularization | 1 (0.3) | 0 (0.0) | 1 (0.5) | 0 (0.0) | 1.000 |
| 30-day MACCE | 11 (3.3) | 1 (6.2) | 2 (1.0) | 8 (6.5) | 0.021 |
| 1-6 month death | 5 (1.5) | 0 (0.0) | 2 (1.0) | 3 (2.4) | 0.517 |
| 1-6 month MACCE | 15 (4.5) | 0 (0.0) | 10 (5.2) | 5 (4.1) | 0.901 |
| >6 month death | 20 (6.0) | 0 (0.0) | 4 (2.1) | 16 (13.0) | <0.001 |
| >6 month MACCE | 31 (9.3) | 0 (0.0) | 13 (6.7) | 18 (14.6) | 0.032 |
| Any follow-up death | 25 (7.5) | 0 (0.0) | 6 (3.1) | 19 (15.4) | <0.001 |
| Any follow-up MACCE | 45 (13.6) | 0 (0.0) | 22 (11.4) | 23 (18.7) | 0.050 |
MI Myocardial Infarction, MACCE Major Adverse Cardiovascular and Cerebrovascular Events
Follow-up all-cause death was recorded in 25 patients (7.5%): 19/123 (15.4%) high-NPS, 6/193 (3.1%) intermediate-NPS, and 0/16 (0.0%) low-NPS patients (exact P < 0.001). Any follow-up MACCE was recorded in 45 patients (13.6%): 23/123 (18.7%), 22/193 (11.4%), and 0/16 (0.0%), respectively (exact P = 0.050). The low-NPS zero-event strata and small denominator limit inference. Outcome rates are shown in Fig. 2.
Fig. 2.
Outcome rates according to preoperative NPS category. Bars show the percentage of patients with 30-day MACCE, follow-up all-cause death, and follow-up MACCE in each NPS category. Percentages should be interpreted with attention to the small low-NPS group size. MACCE, major adverse cardiac and cerebrovascular events; NPS, Naples Prognostic Score
Logistic regression analyses
Model results are summarized in Table 3. For 30-day MACCE, the unadjusted OR was 1.54 (95% CI 0.82–2.89; P = 0.177) per 1-point increase in NPS. High NPS, compared with the combined low- and intermediate-NPS reference group, had an unadjusted OR of 4.78 (95% CI 1.24–18.36; P = 0.023). No adjusted 30-day model was reported because only 11 events occurred.
Table 3.
Logistic regression analyses for binary event indicators. Odds ratios for score-based models are per 1-point increase in recorded NPS (0-4). Because 30-day MACCE had 11 events, only unadjusted 30-day estimates are shown. High NPS uses the combined low- and intermediate-NPS groups as reference. EuroSCORE II models are complete-case sensitivity analyses (280/332)
| Outcome | Model | N/events | NPS term | OR (95% CI) | P value |
|---|---|---|---|---|---|
| 30-day MACCE | Unadjusted NPS score | 332/11 | NPS score, per 1-point increase | 1.54 (0.82-2.89) | 0.177 |
| 30-day MACCE | Unadjusted high vs non-high NPS | 332/11 | High NPS vs low/intermediate | 4.78 (1.24-18.36) | 0.023 |
| Follow-up MACCE | Unadjusted NPS score | 332/45 | NPS score, per 1-point increase | 1.52 (1.09-2.11) | 0.013 |
| Follow-up MACCE | Unadjusted high vs non-high NPS | 332/45 | High NPS vs low/intermediate | 1.95 (1.04-3.68) | 0.038 |
| Follow-up MACCE | Clinical model adjusted for recent MI, age, LVEF, and eGFR | 332/45 | NPS score, per 1-point increase | 1.34 (0.95-1.88) | 0.096 |
| Follow-up MACCE | Complete-case model adjusted for recent MI and EuroSCORE II | 280/43 | NPS score, per 1-point increase | 1.51 (1.07-2.13) | 0.019 |
| Follow-up all-cause death | Unadjusted NPS score | 332/25 | NPS score, per 1-point increase | 2.18 (1.38-3.46) | <0.001 |
| Follow-up all-cause death | Unadjusted high vs non-high NPS | 332/25 | High NPS vs low/intermediate | 6.18 (2.40-15.95) | <0.001 |
| Follow-up all-cause death | Clinical model adjusted for recent MI, age, LVEF, and eGFR | 332/25 | NPS score, per 1-point increase | 1.69 (1.00-2.84) | 0.048 |
| Follow-up all-cause death | Complete-case model adjusted for recent MI and EuroSCORE II | 280/23 | NPS score, per 1-point increase | 2.06 (1.25-3.39) | 0.005 |
eGFR Estimated Glomerular Filtration Rate, LVEF Left Ventricular Ejection Fraction, MACCE Major Adverse Cardiovascular and Cerebrovascular Events, MI Myocardial Infarction, NPS Naples Prognostic Score
For follow-up MACCE, the OR per 1-point NPS increase was 1.52 (95% CI 1.09–2.11; P = 0.013) unadjusted, 1.34 (95% CI 0.95–1.88; P = 0.096) in the clinical model, and 1.51 (95% CI 1.07–2.13; P = 0.019) in the EuroSCORE II complete-case model. For follow-up all-cause death, the corresponding adjusted ORs were 1.69 (95% CI 1.00-2.84; P = 0.048) and 2.06 (95% CI 1.25–3.39; P = 0.005). In the troponin sensitivity model, the NPS estimate was OR 1.69 (1.01–2.84) for death (P = 0.048) and OR 1.34 (0.95–1.88) for MACCE (P = 0.096). Firth estimates were directionally similar but borderline in the clinical model (OR 1.64, 95% CI 1.00-2.69; P = 0.050). Adjusted and sensitivity estimates are shown in Fig. 3 and Additional file 1.
Fig. 3.
Adjusted NPS score estimates for institutionally recorded follow-up outcomes. Points and horizontal lines show odds ratios and 95% confidence intervals per 1-point increase in the recorded NPS score (0–4) from conventional and Firth penalized logistic models. CI, confidence interval; eGFR, estimated glomerular filtration rate; LVEF, left ventricular ejection fraction; MACCE, major adverse cardiac and cerebrovascular events; MI, myocardial infarction; NPS, Naples Prognostic Score; OR, odds ratio
In complete cases with available EuroSCORE II, exploratory combined risk stratification suggested higher follow-up event rates among patients with both EuroSCORE II > = 2% and high NPS. Follow-up all-cause death occurred in 11 of 25 patients (44.0%) in this combined category, and follow-up MACCE occurred in 12 of 25 patients (48.0%). The full cross-tabulation is provided in Additional file 1.
Exploratory complete-case apparent AUC analyses are provided in Additional file 1. For 30-day MACCE (280/10), AUCs were 0.66 (0.48–0.80) for NPS, 0.68 (0.43–0.83) for EuroSCORE II, and 0.69 (0.56–0.85) for the combined model. For follow-up MACCE (280/43), the corresponding AUCs were 0.62 (0.54–0.71), 0.62 (0.52–0.71), and 0.64 (0.56–0.74); for follow-up all-cause death (280/23), they were 0.71 (0.59–0.82), 0.82 (0.75–0.88), and 0.80 (0.72–0.89). Each bootstrap sample refitted its model. These descriptive, internally unvalidated estimates do not support time-to-event discrimination claims.
Discussion
Higher recorded preoperative NPS was associated with institutionally recorded follow-up all-cause death after isolated CABG. Death was recorded in 19 of 123 high-NPS, 6 of 193 intermediate-NPS, and 0 of 16 low-NPS patients. The score-based estimate persisted after clinical adjustment and in the EuroSCORE II complete-case model, but the conventional clinical-model result was borderline and the Firth estimate was P = 0.050. The zero-event low-NPS stratum, incomplete follow-up ascertainment, and absence of time-to-event data limit causal and prognostic interpretation.
This mortality pattern is consistent with cardiovascular studies in which higher NPS was associated with adverse outcomes after ST-segment elevation myocardial infarction and transcatheter aortic valve implantation [10–12]. The present analysis should be viewed as an independent, constrained replication in isolated CABG rather than a first demonstration. Its contribution lies in testing the association in a separate cohort while explicitly showing the effects of sparse events, complete-case EuroSCORE II adjustment, and limited follow-up structure.
The biological interpretation of NPS in this CABG cohort should remain specific to the available data. The inflammatory component should not be interpreted as evidence of active infection, autoimmune disease, or malignancy, which were not the intended clinical constructs. It more plausibly reflects a combination of ischemic burden, recent myocardial injury, renal dysfunction, comorbidity, and baseline host response. Troponin was numerically higher in the high-NPS group, supporting the possibility that myocardial injury contributed to the observed inflammatory-nutritional profile. The troponin-adjusted score estimate was directionally similar for follow-up death, but median-based categorization does not resolve the ACS construct limitation. Low albumin may reflect impaired physiologic reserve, systemic illness, or inflammatory burden, while NLR and LMR characterize leukocyte patterns associated with adverse cardiovascular outcomes [6, 7, 16–20]. In adult cardiac surgery, nutritional indices and inflammatory scores have been associated with postoperative morbidity and mortality, supporting the concept that preoperative host status contributes information beyond anatomic revascularization variables [16–18].
The clinical role of NPS should be interpreted alongside established surgical risk tools. EuroSCORE II captures age, ventricular function, renal function, and procedural risk [3–5]. NPS retained an association with recorded death in the EuroSCORE II complete-case model, but EuroSCORE II was unavailable in 52 patients and missingness was differential. Apparent AUCs did not show incremental discrimination for death when NPS was added to EuroSCORE II (0.80 vs. 0.82). This is statistically possible because logistic likelihood, not AUC, is optimized; the small difference may also reflect sampling variability with only 23 complete-case deaths. These internally unvalidated estimates do not establish clinical improvement.
Composite MACCE findings were less consistent than mortality findings. The score estimate attenuated in the clinical model, which may reflect outcome heterogeneity because repeat revascularization, stroke, myocardial infarction, and death do not share a single pathway. The 30-day composite had only 11 events; adjusted models were therefore omitted rather than interpreted from an event-limited fit.
The ischemic-presentation adjustment also requires caution. The dataset did not contain a discrete field for acute coronary syndrome, so recent myocardial infarction within 3 months was used as the available ischemic-presentation variable. This distinction matters because recent myocardial infarction does not capture all urgent coronary presentations, including unstable angina or non-ST-segment elevation presentations without a documented infarction field. Recent myocardial infarction should therefore be interpreted as a measured covariate rather than a complete descriptor of acute coronary presentation, and subgroup analyses by acute coronary syndrome status could not be performed.
Operative variables were reported descriptively and should not be used to infer surgical intent. High-NPS patients had numerically shorter cardiopulmonary bypass and cross-clamp times, while distal anastomosis counts were similar across groups. Prior studies link prolonged cross-clamp and cardiopulmonary bypass durations with adverse outcomes in cardiac surgery [21–23], but this retrospective dataset cannot determine whether operative time differences reflected case selection, technical strategy, anatomy, surgeon preference, or documentation patterns. Shorter operative times should therefore not be interpreted as evidence of deliberate risk-reduction strategies.
A recent 1,195-patient isolated-CABG study also reported associations between higher NPS and 30-day and 1-year mortality and assessed addition of NPS to EuroSCORE II [24]. Compared with that larger cohort, the present study has fewer events and weaker follow-up ascertainment. Its results therefore provide limited external reproducibility evidence but should not be used to claim superior or novel prognostic performance.
Several limitations constrain interpretation. First, the retrospective single-center design limits generalizability and leaves residual confounding. The available source file did not contain the full pre-extraction screening history or a stable identifier audit; Fig. 1 therefore reports only verifiable file-level flow. Second, exposure was defined by the recorded preoperative NPS field; missing scores were not reconstructed from components. Third, the low-NPS group was small and had zero recorded follow-up deaths, creating sparse-data and quasi-separation concerns. The conventional clinical-model estimate was fragile, although Firth analyses were directionally similar. Fourth, statin and lipid-lowering therapy were unavailable, so treatment may have influenced the cholesterol component. Fifth, cause-specific mortality was unavailable. Sixth, institutional follow-up fields lacked event dates, censoring dates, a common observation horizon, external linkage, and a loss-to-follow-up indicator. Blank fields were coded as non-events under the locked rule, although missingness was substantial in later fields; recorded-event proportions may therefore underestimate events and must not be interpreted as cumulative risk. Seventh, discrete ACS/CCS presentation, SYNTAX score, detailed anatomy, and conduit data were unavailable. Recent myocardial infarction and median-based troponin adjustment do not replace adjudicated ACS status. Eighth, EuroSCORE II was missing in 52 patients and complete cases differed from incomplete cases; these models and AUCs are sensitivity analyses. Multiple imputation was not used because the source components needed to reconstruct the composite score were unavailable. Finally, isolated implausible entries were identified in several non-model laboratory and operative fields. Affected variables were removed from the main descriptive table when source verification was unavailable and were not used in outcome models.
Conclusions
Higher recorded preoperative NPS was associated with institutionally recorded follow-up all-cause death after isolated CABG, although the adjusted estimate was statistically fragile. MACCE associations were less consistent, and the 30-day analysis was event-limited. NPS may characterize inflammatory-nutritional status alongside established risk assessment, but these data do not establish incremental clinical utility.
Supplementary Information
Acknowledgements
Not applicable.
Abbreviations
- ACS
Acute coronary syndrome
- AI
Artificial intelligence
- AUC
Area under the receiver operating characteristic curve
- CABG
Coronary artery bypass grafting
- CI
Confidence interval
- COPD
Chronic obstructive pulmonary disease
- CPB
Cardiopulmonary bypass
- eGFR
Estimated glomerular filtration rate
- EuroSCORE II
European System for Cardiac Operative Risk Evaluation II
- ICU
Intensive care unit
- LMR
Lymphocyte-to-monocyte ratio
- LVEF
Left ventricular ejection fraction
- MACCE
Major adverse cardiac and cerebrovascular events
- NLR
Neutrophil-to-lymphocyte ratio
- NPS
Naples Prognostic Score
- OR
Odds ratio
- PCI
Percutaneous coronary intervention
- ROC
Receiver operating characteristic
- STROBE
Strengthening the Reporting of Observational Studies in Epidemiology
- SYNTAX
Synergy Between PCI with Taxus and Cardiac Surgery
Authors’ contributions
M.S.: Conception, Design, Resource, Analysis and/or Interpretation, Literature Review, Writer. A.H.: Conception, Design, Analysis and/or Interpretation. R.A.: Conception, Data Collection and/or Processing, Analysis and/or Interpretation. S.M.: Design, Resource, Data Collection and/or Processing. A.M.: Materials, Data Collection and/or Processing, Literature Review. M.C.: Materials, Analysis and/or Interpretation, Literature Review.B.G.S.: Materials, Analysis and/or Interpretation, Literature Review. B.I.: Supervision, Resource, Critical Review. U.A: Conception, Supervision, Resource, Critical Review.
Funding
The authors received no specific funding for this work.
Data availability
No datasets were generated or analysed during the current study.
Declarations
Ethics approval and consent to participate
The study was conducted in accordance with the principles of the Declaration of Helsinki. The study protocol was approved by the Bezmialem Foundation University Medical Faculty Hospital. Informed consent was waived due to the retrospective nature of the study.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Deb S, Wijeysundera HC, Ko DT, Tsubota H, Hill S, Fremes SE. Coronary artery bypass graft surgery vs percutaneous interventions in coronary revascularization: a systematic review. JAMA. 2013;310(19):2086–95. 10.1001/jama.2013.281718. [DOI] [PubMed] [Google Scholar]
- 2.Beerkens FJ, Claessen BE, Mahan M, Gaudino MFL, Tam DY, Henriques JPS, et al. Contemporary coronary artery bypass graft surgery and subsequent percutaneous revascularization. Nat Rev Cardiol. 2022;19(3):195–208. 10.1038/s41569-021-00612-6. [DOI] [PubMed] [Google Scholar]
- 3.Fukui T, Uchimuro T, Takanashi S. EuroSCORE II with SYNTAX score to assess risks of coronary artery bypass grafting outcomes. Eur J Cardiothorac Surg. 2015;47(1):66–71. 10.1093/ejcts/ezu045. [DOI] [PubMed] [Google Scholar]
- 4.Biancari F, Vasques F, Mikkola R, Martin M, Lahtinen J, Heikkinen J. Validation of EuroSCORE II in patients undergoing coronary artery bypass surgery. Ann Thorac Surg. 2012;93(6):1930–5. 10.1016/j.athoracsur.2012.02.064. [DOI] [PubMed] [Google Scholar]
- 5.Silverborn M, Nielsen S, Karlsson M. The performance of EuroSCORE II in CABG patients in relation to sex, age, and surgical risk: a nationwide study in 14,118 patients. J Cardiothorac Surg. 2023;18(1):40. 10.1186/s13019-023-02141-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.He F, Huang H, Xu W, Cui K, Ruan Y, Guo Y, et al. Prognostic impact of malnutrition in patients with coronary artery disease: a systematic review and meta-analysis. Nutr Rev. 2024;82(8):1013–27. 10.1093/nutrit/nuad108. [DOI] [PubMed] [Google Scholar]
- 7.He L, Xie X, Xue J, Xie H, Zhang Y. Association of the systemic immune-inflammation index with all-cause mortality in patients with arteriosclerotic cardiovascular disease. Front Cardiovasc Med. 2022;9:952953. 10.3389/fcvm.2022.952953. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Kim CH, Kang Y, Kim JS, Sohn SH, Hwang HY. Association between the frailty index and clinical outcomes after coronary artery bypass grafting. J Chest Surg. 2022;55(3):189–96. 10.5090/jcs.21.147. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Galizia G, Lieto E, Auricchio A, Cardella F, Mabilia A, Podzemny V, et al. Naples Prognostic Score, based on nutritional and inflammatory status, is an independent predictor of long-term outcome in patients undergoing surgery for colorectal cancer. Dis Colon Rectum. 2017;60(12):1273–84. 10.1097/DCR.0000000000000961. [DOI] [PubMed] [Google Scholar]
- 10.Erdogan A, Genc O, Ozkan E, Goksu MM, Ibisoglu E, Bilen MN, et al. Impact of Naples Prognostic Score at admission on in-hospital and follow-up outcomes among patients with ST-segment elevation myocardial infarction. Angiology. 2023;74(10):970–80. 10.1177/00033197231151559. [DOI] [PubMed] [Google Scholar]
- 11.Hakgor A, Dursun A, Kahraman BC, Yazar A, Savur U, Akhundova A, et al. The impact of the Naples Prognostic Score on the short- and long-term prognosis of patients undergoing transcatheter aortic valve implantation. J Cardiovasc Med (Hagerstown). 2024;25(7):519–28. 10.2459/JCM.0000000000001637. [DOI] [PubMed] [Google Scholar]
- 12.Gitmez M, Guzel T, Kis M, Coskun F, Isik MA, Aktan A, et al. The performance of the NAPLES prognostic score in predicting one-year mortality and major adverse cardiovascular events after transcatheter aortic valve implantation in patients with severe aortic stenosis. Kardiol Pol. 2025;83(3):287–94. 10.33963/v.phj.103230. [DOI] [PubMed] [Google Scholar]
- 13.von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. PLoS Med. 2007;4(10):e296. 10.1371/journal.pmed.0040296. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Galderisi M, Cosyns B, Edvardsen T, Cardim N, Delgado V, Di Salvo G, et al. Standardization of adult transthoracic echocardiography reporting in agreement with recent chamber quantification, diastolic function, and heart valve disease recommendations. Eur Heart J Cardiovasc Imaging. 2017;18(12):1301–10. 10.1093/ehjci/jex244. [DOI] [PubMed] [Google Scholar]
- 15.Thygesen K. Ten Commandments for the Fourth Universal Definition of Myocardial Infarction 2018. Eur Heart J. 2019;40(3):226. 10.1093/eurheartj/ehy856. [DOI] [PubMed] [Google Scholar]
- 16.Lee SI, Ko KP, Choi CH, Park CH, Park KY, Son KH. Does the prognostic nutritional index have a predictive role in the outcomes of adult cardiac surgery? J Thorac Cardiovasc Surg. 2020;160(1):145–e1533. 10.1016/j.jtcvs.2019.08.069. [DOI] [PubMed] [Google Scholar]
- 17.Parmana IMA, Boom CE, Poernomo H, Gani C, Nugroho B, Cintyandy R, et al. High preoperative systemic immune-inflammation index values significantly predicted poor outcomes after on-pump coronary artery bypass surgery. J Inflamm Res. 2024;17:755–64. 10.2147/JIR.S449795. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Dey S, Kashav R, Kohli JK, Magoon R, ItiShri, Walian A, et al. Systemic immune-inflammation index predicts poor outcome after elective off-pump CABG: a retrospective, single-center study. J Cardiothorac Vasc Anesth. 2021;35(8):2397–404. 10.1053/j.jvca.2020.09.092. [DOI] [PubMed] [Google Scholar]
- 19.Gibson PH, Cuthbertson BH, Croal BL, Rae D, El-Shafei H, Gibson G, et al. Usefulness of neutrophil/lymphocyte ratio as predictor of new-onset atrial fibrillation after coronary artery bypass grafting. Am J Cardiol. 2010;105(2):186–91. 10.1016/j.amjcard.2009.09.007. [DOI] [PubMed] [Google Scholar]
- 20.Sawant AC, Adhikari P, Narra SR, Srivatsa SS, Mills PK, Srivatsa SS. Neutrophil to lymphocyte ratio predicts short- and long-term mortality following revascularization therapy for ST elevation myocardial infarction. Cardiol J. 2014;21(5):500–8. 10.5603/CJ.a2013.0148. [DOI] [PubMed] [Google Scholar]
- 21.Shultz B, Timek T, Davis AT, Heiser J, Murphy E, Willekes C, et al. Outcomes in patients undergoing complex cardiac repairs with cross clamp times over 300 minutes. J Cardiothorac Surg. 2016;11(1):105. 10.1186/s13019-016-0501-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Al-Sarraf N, Thalib L, Hughes A, Houlihan M, Tolan M, Young V, et al. Cross-clamp time is an independent predictor of mortality and morbidity in low- and high-risk cardiac patients. Int J Surg. 2011;9(1):104–9. 10.1016/j.ijsu.2010.10.007. [DOI] [PubMed] [Google Scholar]
- 23.Nissinen J, Biancari F, Wistbacka J, Peltola T, Loponen P, Tarkiainen P, et al. Safe time limits of aortic cross-clamping and cardiopulmonary bypass in adult cardiac surgery. Perfusion. 2009;24(5):297–305. 10.1177/0267659109354656. [DOI] [PubMed] [Google Scholar]
- 24.Sayarer C, Dayanir OB, Eken EN, Gencpinar T, Bayrak S. The impact of the Naples prognostic score on outcomes in patients undergoing coronary artery bypass grafting. Turk Gogus Kalp Damar Cerrahisi Derg. 2026;34(2):116–20. 10.4274/tjtcs.2025.28487. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
No datasets were generated or analysed during the current study.