Association of naples prognostic score and systemic immune-inflammation index with left ventricular recovery in peripartum cardiomyopathy

Abstract

Background

Peripartum cardiomyopathy (PPCM) is a life-threatening form of heart failure characterized by substantial heterogeneity in left ventricular recovery. Identifying reliable and readily accessible biomarkers associated with myocardial recovery remains an important clinical challenge. Although systemic inflammatory indices such as the systemic immune-inflammation index (SII) have been associated with persistent left ventricular dysfunction in PPCM, they do not incorporate nutritional and immune-related parameters that may also influence clinical outcomes. The Naples Prognostic Score (NAPLES), which integrates inflammatory, immune, and nutritional parameters, has emerged as a promising prognostic index in cardiovascular disease; however, its role in PPCM has not yet been elucidated. In this context, this study aimed to investigate the association between the Naples Prognostic Score and left ventricular functional recovery in patients with peripartum cardiomyopathy, and to compare its discriminatory ability with that of the Systemic Immune-Inflammation Index (SII).

Methods

This retrospective single-center study included 30 patients diagnosed with PPCM between 2015 and 2025. Left ventricular functional recovery was defined as an LVEF ≥ 50% at the last available follow-up assessment, provided that follow-up duration was at least 12 months. Laboratory parameters obtained on the day of hospital admission (baseline) were used to calculate NAPLES and SII. Patients were categorized according to the presence or absence of left ventricular functional recovery. Receiver operating characteristic (ROC) curve analysis and risk stratification based on optimal cut-off values were performed.

Results

Sixteen patients (53.3%) achieved left ventricular functional recovery, whereas 14 (46.7%) did not. Patients without left ventricular functional recovery exhibited significantly higher SII values and higher NAPLES scores. SII demonstrated good discriminatory ability for differentiating patients according to left ventricular recovery status (AUC = 0.800), and NAPLES also showed a significant association with left ventricular recovery status (AUC = 0.731). A combined model incorporating both SII and NAPLES yielded the highest numerical AUC (0.818), although no statistically significant improvement compared with SII alone was observed. An SII > 1332 and a NAPLES score ≥ 3 were associated with a significantly increased risk of persistent left ventricular dysfunction.

Conclusions

Both SII and NAPLES were significantly associated with left ventricular functional recovery in patients with peripartum cardiomyopathy. By integrating inflammatory, nutritional, and immune-related parameters, the Naples Prognostic Score may provide complementary biological and clinical information beyond conventional inflammatory indices. These findings suggest that readily available biomarkers such as SII and NAPLES may provide additional clinical information and warrant further investigation in PPCM. However, larger prospective studies are required before their clinical applicability can be established.

Introduction

Peripartum cardiomyopathy (PPCM) is a rare but potentially life-threatening form of cardiomyopathy characterized by left ventricular systolic dysfunction (left ventricular ejection fraction ≤ 45%) occurring in the last month of pregnancy or within the first five months postpartum in women without previously known structural heart disease [1]. The clinical presentation ranges from mild symptoms of heart failure to severe manifestations such as cardiogenic shock, malignant arrhythmias, and sudden cardiac death [2]. Despite advances in contemporary heart failure therapy, PPCM remains associated with considerable morbidity and mortality [3].

One of the most distinctive features of PPCM is the marked heterogeneity in its clinical course and in the pattern of left ventricular functional recovery. While a substantial proportion of patients experience partial or complete improvement in left ventricular ejection fraction (LVEF), a considerable subgroup develops persistent systolic dysfunction, leading to adverse long-term clinical outcomes [4]. Therefore, identifying biomarkers associated with left ventricular functional recovery at an early stage is of major clinical importance for risk stratification, follow-up planning, and patient counseling. However, reliable and easily accessible biomarkers that adequately reflect recovery potential remain limited. Recent multicenter registry data from Türkiye (ARTEMIS registry), including 293 patients with PPCM, also demonstrated substantial heterogeneity in clinical presentation, treatment practices, and recovery outcomes, further emphasizing the need for improved early risk stratification strategies in this population [5].

In recent years, PPCM has been increasingly recognized not merely as an inflammatory condition, but as a complex disorder characterized by immune dysregulation, oxidative stress, and metabolic imbalance [6]. Pregnancy- and postpartum-related biological changes may contribute to disease onset and progression through multiple interconnected mechanisms affecting vascular and immune homeostasis [6]. Within this framework, not only the severity of inflammation but also nutritional and immune-related factors may play a critical role in determining prognosis.

More recently, composite prognostic scores derived from routinely available hematological and biochemical parameters have gained increasing attention in clinical practice [7–9]. By simultaneously reflecting multiple biological domains—including inflammation, immune response, and metabolic status—these indices provide a more comprehensive risk assessment compared with isolated laboratory parameters. Although inflammatory indices such as the Systemic Immune-Inflammation Index (SII) have shown significant associations with clinical outcomes in PPCM, they do not incorporate nutritional and immune-related parameters that may also be relevant to prognosis [10].

The Naples Prognostic Score (NAPLES) is a composite index integrating serum albumin, total cholesterol, neutrophil-to-lymphocyte ratio (NLR), and lymphocyte-to-monocyte ratio (LMR), thereby offering a multidimensional assessment of systemic inflammation, nutritional status, and immune-related parameters [11]. While albumin and cholesterol levels reflect nutritional status and chronic inflammatory burden, NLR and LMR serve as markers of acute inflammatory response and immune regulation. NAPLES has previously been identified as a prognostic marker in various cardiovascular conditions, including acute coronary syndromes and heart failure populations, where it has been associated with adverse outcomes and increased mortality [11, 12].

To the best of our knowledge, no previous study has specifically evaluated the association between the Naples Prognostic Score and left ventricular functional recovery in patients with PPCM. Given this gap in the literature, evaluating the relationship between NAPLES and recovery patterns in PPCM may provide insights beyond conventional inflammatory indices by integrating inflammatory, nutritional, and immune-related parameters, thereby enabling a more comprehensive clinical assessment.

The primary objective of this study was to assess the association between the Naples Prognostic Score and left ventricular functional recovery in patients diagnosed with peripartum cardiomyopathy. In addition, the comparative discriminatory ability of NAPLES and the Systemic Immune-Inflammation Index (SII), a more traditional inflammatory marker, was analyzed to further elucidate their relative contributions to clinical risk assessment.

Study population

This retrospective study included 30 consecutive patients aged 18–40 years who were diagnosed with peripartum cardiomyopathy (PPCM) and followed at the Department of Cardiology, Sakarya Training and Research Hospital between 2015 and 2025. A total of 3614 women younger than 50 years who were hospitalized in the cardiology department during the study period were screened. After detailed review of medical records, 36 patients with PPCM were identified. Two patients who died during the index hospitalization and four patients with insufficient clinical or echocardiographic data were excluded. The final study population consisted of 30 patients. The patient selection process is summarized in Fig. 1.

Fig. 1.

Flow diagram of patient screening, exclusions, and final study population. PPCM: peripartum cardiomyopathy; LVEF: left ventricular ejection fraction

In accordance with established diagnostic criteria, PPCM was defined as the development of left ventricular systolic dysfunction (left ventricular ejection fraction ≤ 45%) occurring in the last month of pregnancy or within the first five months postpartum in the absence of previously known structural heart disease [1]. All patients underwent transthoracic echocardiography at the time of diagnosis and during follow-up. Left ventricular ejection fraction (LVEF) was calculated using the Simpson biplane method.

Patients were excluded if they had a history of known structural heart disease prior to pregnancy, previously diagnosed cardiomyopathy or heart failure, or an identifiable alternative etiology that could explain left ventricular systolic dysfunction (including coronary artery disease, significant valvular heart disease, congenital heart disease, or myocarditis). Patients with acute or chronic active infection, autoimmune or rheumatologic disorders that could influence systemic inflammation, a history of malignancy or active malignancy, or incomplete clinical, laboratory, or echocardiographic data were also excluded.

Left ventricular functional recovery was defined as an LVEF ≥ 50% at the last available follow-up echocardiographic assessment, provided that the follow-up duration was at least 12 months [13]. Patients were subsequently categorized into two groups according to the presence or absence of left ventricular functional recovery. All patients received guideline-directed medical therapy for heart failure according to contemporary heart failure management recommendations, including beta-blockers, angiotensin-converting enzyme inhibitors or angiotensin receptor blockers after delivery, and diuretics when clinically indicated. Bromocriptine therapy was not routinely used during the study period, and none of the patients included in the present study received bromocriptine treatment. No patients in the study cohort received SGLT2 inhibitor therapy during the study period. None of the patients included in the study were breastfeeding at the time of diagnosis or during the early follow-up period.

Laboratory measurements

Laboratory data were retrospectively retrieved from the hospital’s electronic medical record system. Complete blood count and biochemical parameters obtained from routinely collected venous blood samples at the time of initial hospital admission were retrospectively analyzed.

Complete blood count parameters included leukocyte, neutrophil, lymphocyte, monocyte, and platelet counts. Biochemical analyses comprised serum albumin, total cholesterol, C-reactive protein (CRP), creatinine, and other routine laboratory parameters.

For the calculation of inflammatory and prognostic indices, laboratory values obtained on the day of hospital admission were used to ensure a standardized time point. This approach was adopted to enhance comparability and minimize variability related to temporal changes in inflammatory markers.

Using these parameters, composite indices reflecting systemic inflammation, immune response, and nutritional status were calculated, and their associations with left ventricular functional recovery were evaluated.

The Naples Prognostic Score (NAPLES) was calculated as a composite index integrating systemic inflammation, immune response, and nutritional status. The neutrophil-to-lymphocyte ratio (NLR) was derived by dividing the absolute neutrophil count by the absolute lymphocyte count, while the lymphocyte-to-monocyte ratio (LMR) was calculated as the absolute lymphocyte count divided by the absolute monocyte count. Based on previously established cut-off values, one point was assigned for each of the following: NLR > 2.96, LMR < 4.44, serum albumin < 4.0 g/dL, and total cholesterol < 180 mg/dL. The total NAPLES score ranged from 0 to 4, with higher scores indicating a more unfavorable inflammatory–metabolic–immune profile [14].

Statistical analysis

All statistical analyses were performed using SPSS software (version 22.0; IBM Corp., Armonk, NY, USA). The distribution of continuous variables was assessed using the Shapiro–Wilk test. Normally distributed variables were expressed as mean ± standard deviation, whereas non-normally distributed variables were presented as median (interquartile range). Categorical variables were reported as counts and percentages. The choice of statistical test for each continuous variable was based on the results of the Shapiro–Wilk normality test.

Comparisons between patients with and without left ventricular functional recovery were conducted using the independent samples t-test or the Mann–Whitney U test for continuous variables, depending on distribution characteristics. The chi-square test or Fisher’s exact test was used for categorical variables, as appropriate.

The discriminative performance of the Naples Prognostic Score (NAPLES) and the Systemic Immune-Inflammation Index (SII) for differentiating patients according to left ventricular recovery status was evaluated using receiver operating characteristic (ROC) curve analysis, and the area under the curve (AUC) was calculated. To generate the ROC curve of the combined model, binary logistic regression analysis including SII and NAPLES as continuous variables was performed, and the predicted probabilities derived from this model were used.

Optimal cut-off values were determined using the Youden index. For NAPLES, the optimal threshold identified by the Youden index was 2.5. Because NAPLES is an integer-based score ranging from 0 to 4, this threshold was categorized as ≤ 2 and ≥ 3 for clinical applicability.

Based on the determined cut-off values for NAPLES and SII, patients were stratified into low- and high-risk groups, and 2 × 2 contingency tables were constructed. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated. Because of the retrospective design of the study and the rarity of peripartum cardiomyopathy, no a priori sample size calculation was performed. Instead, all consecutive eligible patients diagnosed at our center between 2015 and 2025 were included in the analysis. In addition, a post-hoc power analysis was performed based on the observed between-group difference in SII values. The calculated statistical power was approximately 83%, supporting the adequacy of the present exploratory analysis.

Given the limited sample size and the small number of outcome events, the analyses were considered exploratory and hypothesis-generating. To partially address potential confounding by baseline disease severity, additional exploratory logistic regression models adjusted for baseline left ventricular ejection fraction (LVEF) were performed. Owing to the limited number of events, more complex multivariable models including multiple clinical covariates were intentionally avoided to minimize the risk of overfitting.

Results

A total of 30 patients with peripartum cardiomyopathy were included in the study. All included patients had a follow-up duration of at least 12 months. At the end of the follow-up period, left ventricular functional recovery (LVEF ≥ 50%) was observed in 16 patients (53.3%), whereas 14 patients (46.7%) did not achieve recovery.

The baseline clinical and laboratory characteristics of patients with and without left ventricular functional recovery are presented in Table 1.

Table 1.

Baseline clinical and laboratory characteristics of patients with and without left ventricular functional recovery

Variable	No Recovery (n = 14)	Recovery (n = 16)	p value
Age (years)	30.80 ± 5.12	31.93 ± 4.71	0.533
Antepartum diagnosis, n (%)	3 (21)	6 (38)	0.427
Pregnancy number	1 (1–2)	1 (1–1)	0.383
Time from delivery to diagnosis, days*	16 (8.5–21.3)	8 (7–10)	0.24
Baseline LVEF (%)	32.00 (28.50–38.50)	41.00 (28.50–44.00)	0.146
Baseline LVEDD (mm)	54.2 ± 9.7	51.9 ± 7.7	0.45
Baseline LVESD (mm)	45.0 ± 9.5	43.5 ± 8.6	0.53
Neutrophils (×10⁹/L)	10.58 ± 3.85	8.04 ± 3.33	0.064
Lymphocytes (×10⁹/L)	1.63 (1.14–1.88)	1.50 (1.14–2.25)	0.619
Monocytes (×10⁹/L)	0.53 (0.36–0.83)	0.38 (0.30–0.52)	0.213
Platelets (×10⁹/L)	334.00 ± 109.78	251.53 ± 88.28	0.032
Albumin (g/dL)	3.04 ± 0.54	3.11 ± 0.52	0.732
Creatinine (mg/dL)	0.74 ± 0.33	0.69 ± 0.29	0.688
Total cholesterol (mg/dL)	184.80 ± 59.04	226.80 ± 43.63	0.036
C-reactive protein (mg/L)	51.54 (13.10–86.33)	19.10 (8.82–49.75)	0.237
SII	2345.75 ± 1170.09	1283.82 ± 725.15	0.006
NAPLES score	3.00 (3.00–4.00)	2.00 (2.00–3.00)	0.025
Beta-blocker use, n (%)	13 (93)	14 (88)	1
ACEi/ARB use, n (%)	14 (100)	13 (81)	0.228
MRA use, n (%)	10 (71)	7 (43)	0.462
Follow-up duration, months	14 (13-15.5)	13 (12-15.5)	0.688

Values are presented as mean ± standard deviation for normally distributed variables and median (interquartile range) for non-normally distributed variables. Left ventricular functional recovery was defined as a follow-up LVEF ≥50%. Follow-up duration is expressed in months

ACEi Angiotensin-converting enzyme inhibitor, ARB Angiotensin receptor blocker, LVEDD Left Ventricular End-Diastolic Diameter, LVEF Left Ventricular Ejection Fraction, LVESD Left Ventricular End-Systolic Diameter, MRA mineralocorticoid receptor antagonist, NAPLES Naples Prognostic Score, SII Systemic Immune-Inflammation Index

*Calculated only among patients diagnosed in the postpartum period

The proportions of antepartum versus postpartum diagnosis, beta-blocker use, ACEi/ARB use, and MRA use were comparable between the recovery and non-recovery groups (all p > 0.05). Similarly, baseline serum creatinine levels, pregnancy number, and time from delivery to diagnosis did not differ significantly between the groups. Follow-up duration was comparable between patients with and without left ventricular recovery [14 (13–15.5) vs. 13 (12–15.5) months, respectively; p = 0.688].

There were no significant differences between the two groups in terms of age and baseline left ventricular ejection fraction (LVEF) (p = 0.533 and p = 0.146, respectively).

Among laboratory parameters, platelet count was significantly higher in patients without recovery (p = 0.032). Total cholesterol levels were significantly higher in patients who achieved recovery (p = 0.036). Serum albumin and C-reactive protein (CRP) levels did not differ significantly between the groups.

Patients without recovery had significantly higher SII values compared with those who achieved recovery (2345.8 ± 1170.1 vs. 1283.8 ± 725.2; p = 0.006). Similarly, the NAPLES score was significantly higher in the no-recovery group (median 3 [3–4] vs. 2 [2–3]; p = 0.025).

Receiver operating characteristic (ROC) curve analysis suggested that the Systemic Immune-Inflammation Index (SII) showed reasonable discriminatory ability for differentiating patients according to left ventricular recovery status (AUC = 0.800, 95% CI: 0.63–0.97). The Naples Prognostic Score (NAPLES) also showed discriminatory ability for recovery status (AUC = 0.731, 95% CI: 0.55–0.91). The combined model incorporating both SII and NAPLES yielded the highest numerical AUC (0.818, 95% CI: 0.65–0.98). However, DeLong analysis demonstrated that the difference between the combined model and SII alone was not statistically significant (p = 0.550). The corresponding ROC curves are presented in Fig. 2.

Fig. 2.

Receiver operating characteristic (ROC) curves of SII, Naples Prognostic Score, and their combined model for differentiating patients according to left ventricular recovery status in patients with peripartum cardiomyopathy. SII: Systemic Immune-Inflammation Index; NAPLES: Naples Prognostic Score

The optimal cut-off values determined by the Youden index were 1332 for SII and 2.5 for NAPLES. For clinical applicability, the NAPLES score was categorized as ≤ 2 and ≥ 3.

Risk analyses based on these cut-off values are presented in Table 2.

Table 2.

Association between SII and NAPLES cut-off values and lack of left ventricular functional recovery

Parameter	OR (95% CI)	p value
SII > 1332	16.0 (1.7–147.8)	0.014
NAPLES ≥ 3	9.75 (1.6–58.9)	0.021

Odds ratios were calculated from 2 × 2 contingency tables using ROC-derived optimal cut-off values (SII > 1332 and NAPLES ≥ 3). SII Systemic Immune-Inflammation Index, NAPLES Naples Prognostic Score

Based on the identified cut-off values, patients with SII > 1332 had a significantly higher risk of lack of left ventricular recovery (OR = 16.0, p = 0.014). Similarly, a NAPLES score ≥ 3 was associated with an increased risk of persistent left ventricular dysfunction (OR = 9.75, p = 0.021).

To evaluate whether these associations persisted after accounting for baseline left ventricular systolic function, exploratory logistic regression analyses adjusted for baseline LVEF were performed. In these exploratory models, higher SII values showed a statistically significant association with persistent left ventricular dysfunction (OR 4.50, 95% CI 1.30–15.52, p = 0.017). Similarly, higher NAPLES scores showed a statistically significant association with persistent left ventricular dysfunction after adjustment for baseline LVEF (OR 2.82, 95% CI 1.05–7.61, p = 0.040).

Patients presenting with concomitantly elevated SII and high NAPLES scores showed a higher frequency of persistent left ventricular dysfunction, whereas patients with lower values for both parameters demonstrated a higher rate of left ventricular functional recovery.

Discussion

In the present study, we evaluated the association of the Naples Prognostic Score (NAPLES) and the Systemic Immune-Inflammation Index (SII) with left ventricular functional recovery in patients with peripartum cardiomyopathy (PPCM). Our findings suggest that both elevated SII and higher NAPLES scores are associated with the absence of left ventricular functional recovery. These associations remained statistically significant in exploratory analyses adjusted for baseline LVEF; however, given the limited sample size and low number of outcome events, these findings should be interpreted cautiously and considered hypothesis-generating rather than confirmatory. Nevertheless, because more comprehensive multivariable adjustment including additional clinical and hemodynamic variables was not feasible, residual confounding related to overall disease severity cannot be completely excluded. In addition, no significant differences were observed between the recovery groups regarding baseline serum creatinine levels, timing of diagnosis (antepartum versus postpartum), baseline ventricular dimensions, or guideline-directed medical therapies, suggesting that these factors were unlikely to fully explain the observed associations. Baseline LVEF was also comparable between the groups and did not show a significant association with recovery in univariable analysis. However, given the limited sample size, this finding should be interpreted cautiously, since even modest baseline differences in left ventricular systolic function may potentially influence recovery outcomes in PPCM. Although the combined use of these two parameters yielded the highest numerical discriminative performance, DeLong analysis demonstrated no statistically significant improvement compared with SII alone.

The pathophysiology of PPCM cannot be explained by inflammation alone. Growing evidence suggests that PPCM is characterized by immune dysregulation, oxidative stress, and endothelial dysfunction [15–17]. Toxic effects of prolactin fragments on the microvascular endothelium and abnormal immune responses have been implicated as central mechanisms in the development and progression of PPCM [15, 16]. These mechanisms reflect a multidimensional pathophysiological process that encompasses not only inflammatory activation but also alterations in nutritional status, immune regulation, and endothelial function.

In the pathogenesis of PPCM, oxidative stress–mediated cleavage of prolactin results in the formation of a 16-kDa prolactin fragment with pronounced anti-angiogenic and cytotoxic effects on the microvascular endothelium [15–17]. This fragment promotes the release of microvesicles and microRNAs from endothelial cells, activating secondary signaling pathways that disrupt cardiomyocyte metabolism and impair capillary integrity. The resulting microvascular dysfunction and compromised oxygen delivery are considered key mechanisms limiting myocardial recovery capacity.

The prognostic significance of inflammatory activation in PPCM has also been demonstrated through various biochemical markers. Previous studies have suggested that inflammatory activation plays an important role in the pathophysiology of PPCM, and inflammatory biomarkers such as C-reactive protein (CRP), tumor necrosis factor-α (TNF-α), and interleukin-6 (IL-6) have been reported to be elevated in affected patients [18]. These findings support the concept that PPCM is not merely a structural cardiomyopathy, but rather a condition characterized by substantial systemic and local inflammatory activity. Although CRP levels were numerically higher in patients without recovery, this difference did not reach statistical significance. This finding may partly be related to the limited sample size and may also suggest that isolated inflammatory markers such as CRP incompletely reflect the complex multidimensional inflammatory processes involved in PPCM when compared with composite inflammatory indices. In addition, platelet counts were significantly higher in patients without recovery. Although the biological significance of this finding remains uncertain, elevated platelet counts may reflect an enhanced inflammatory and prothrombotic milieu. Activated platelets can amplify inflammatory pathways and contribute to thrombotic processes, both of which have been implicated in adverse cardiovascular outcomes [7].

Within this context, SII primarily reflects the acute inflammatory burden, whereas the NAPLES score provides a more integrated assessment by incorporating serum albumin and total cholesterol as markers of nutritional and metabolic status, together with NLR and LMR as indicators of immune response and regulation. In the setting of heart failure, hypoalbuminemia and reduced total cholesterol levels have been associated with chronic inflammation and adverse prognosis [19, 20]. This phenotype may reflect a less favorable clinical profile and has been associated with poorer cardiovascular outcomes in affected patients [20]. More broadly, growing evidence suggests that persistent systemic inflammation plays a central role in heart failure development and progression by promoting endothelial dysfunction, oxidative stress, myocardial fibrosis, and adverse ventricular remodeling. Pro-inflammatory mediators such as CRP, TNF-α, and IL-6 have been implicated not only as biomarkers of inflammatory activation but also as active contributors to myocardial dysfunction and long-term disease progression in heart failure populations [21]. These observations support the concept that inflammatory biomarkers may reflect underlying biological processes directly linked to myocardial recovery potential rather than simply representing nonspecific inflammatory activation. It should also be noted that the component thresholds used for the calculation of NAPLES were adopted from previously published studies and have not been specifically validated in PPCM populations. Therefore, their applicability to peripartum cardiomyopathy should be interpreted cautiously and warrants further investigation. Although SII has previously been investigated in patients with PPCM [10], data regarding the clinical relevance of NAPLES in this population remain lacking. Therefore, the present study extends the existing literature by evaluating a composite score that integrates inflammatory, nutritional, metabolic, and immune-related parameters.

Among the immune components of the NAPLES score, the neutrophil-to-lymphocyte ratio (NLR) and lymphocyte-to-monocyte ratio (LMR) have been identified as potentially important prognostic markers in various cardiovascular conditions [22–24]. An elevated NLR reflects enhanced pro-inflammatory neutrophil activation coupled with suppressed adaptive immune response, whereas a reduced LMR is associated with monocyte-driven chronic inflammation and immune dysregulation [23, 24]. Therefore, NAPLES may more comprehensively capture a hyperinflammatory and immunologically imbalanced phenotype that potentially limits myocardial recovery in PPCM.

Our ROC analyses suggested that SII showed reasonable discriminatory ability for differentiating patients according to recovery status (AUC = 0.800), while the NAPLES score also showed discriminatory ability for differentiating patients according to recovery status (AUC = 0.731). The combined model incorporating both SII and NAPLES yielded the highest AUC (0.818). Although the combined model yielded the highest numerical AUC, DeLong analysis demonstrated no statistically significant improvement compared with SII alone. Therefore, the potential additive value of NAPLES should be interpreted cautiously and requires validation in larger prospective cohorts. Nevertheless, although NAPLES provides a broader multidimensional assessment incorporating inflammatory, metabolic, and immune-related parameters, whether this additional complexity translates into meaningful incremental clinical benefit beyond SII alone remains uncertain. It should also be acknowledged that both SII and NAPLES partly reflect overlapping inflammatory pathways, which may partially explain the relatively modest incremental clinical information obtained when both indices were combined.

In risk stratification based on the identified cut-off values, patients with SII > 1332 had a significantly increased risk of persistent left ventricular dysfunction. Similarly, patients with a NAPLES score ≥ 3 exhibited a lower likelihood of recovery. These findings suggest that, beyond inflammation alone, nutritional, metabolic, and immune-related factors may contribute to clinical outcomes in PPCM. In the ARTEMIS registry, complete recovery was reported in only approximately one-third of patients, further highlighting the highly heterogeneous clinical course of PPCM and the potential importance of identifying additional clinical markers associated with recovery patterns that may improve early risk stratification [5]. Future prospective multicenter studies including larger PPCM cohorts are warranted to validate these findings and to determine whether integrating inflammatory and metabolic indices may improve current risk stratification strategies in peripartum cardiomyopathy. Because the clinical presentation, management strategies, and outcomes of PPCM may vary across different geographic regions and populations, external validation in diverse patient cohorts is particularly important. It should also be acknowledged that these cut-off values were derived from the same small study cohort and therefore may be susceptible to overfitting, limiting their immediate generalizability to external populations.

Although biomarker-based risk assessment may provide additional clinical information, patients with peripartum cardiomyopathy should be evaluated within the broader context of established clinical predictors. Previous studies, including large registry-based analyses such as the ARTEMIS registry from Türkiye, have identified baseline left ventricular systolic function, ventricular dilatation, right ventricular involvement, and hemodynamic instability at presentation as important determinants of clinical outcomes in PPCM [5]. Therefore, biomarkers such as SII and NAPLES should currently be considered complementary tools rather than substitutes for established clinical risk assessment approaches.

In addition, differential diagnosis remains an important consideration in pregnancy-associated cardiomyopathies. Recent evidence suggests that pregnancy-associated Takotsubo syndrome may be underrecognized and can clinically mimic peripartum cardiomyopathy, since both conditions may present with acute heart failure symptoms and left ventricular systolic dysfunction. However, Takotsubo syndrome is typically characterized by sudden onset, regional wall motion abnormalities, and more rapid recovery, whereas PPCM more commonly presents with global ventricular dysfunction and a less predictable recovery course. Careful differentiation between these entities remains important because prognosis and management strategies may differ substantially [25].

From a clinical perspective, early identification of patients at increased risk of persistent left ventricular dysfunction may warrant closer follow-up and careful optimization of guideline-directed heart failure therapy. Although the present study was not designed to evaluate treatment strategies, contemporary heart failure therapies such as sodium-glucose cotransporter-2 (SGLT2) inhibitors and vericiguat have demonstrated favorable effects on cardiovascular outcomes and cardiac remodeling in broader heart failure populations [26, 27]. In addition, structured treatment approaches and strategies aimed at improving treatment adherence may be particularly relevant in patients identified as being at higher risk [28, 29]. Whether inflammatory and metabolic risk profiles such as SII and NAPLES can help guide therapeutic decision-making in PPCM remains to be determined in future prospective studies.

NAPLES has previously been associated with mortality and adverse outcomes in several cardiovascular conditions, including coronary artery disease, acute coronary syndromes, and heart failure populations [14, 30]. Data regarding the clinical relevance of the Naples Prognostic Score in patients with peripartum cardiomyopathy remain limited. Our findings suggest that composite indices incorporating inflammatory, metabolic, and immune parameters may offer a more comprehensive framework for risk stratification in complex cardiomyopathy phenotypes such as PPCM.

Limitations

This study has several limitations. First, its retrospective and single-center design may limit the generalizability of the findings to broader populations of patients with peripartum cardiomyopathy. In addition, geographic and population-related differences in the clinical characteristics and management of PPCM may affect the applicability of our findings to other patient populations. Although established diagnostic criteria for peripartum cardiomyopathy were applied, previously unrecognized asymptomatic cardiomyopathies such as dilated cardiomyopathy cannot be completely excluded because of the retrospective design and the lack of pre-pregnancy baseline cardiac imaging. Furthermore, given the rarity of PPCM, the relatively small sample size may have reduced statistical power, particularly in subgroup analyses. Due to the limited number of events (n = 14), including multiple predictors within the same multivariable logistic regression model would have violated the recommended events-per-variable rule, potentially resulting in model overfitting and unstable estimates. Moreover, the relatively wide confidence intervals observed for odds ratio estimates reflect the limited sample size and event distribution, and the magnitude of these associations should therefore be interpreted with caution. In addition, the relatively small sample size may have increased the risk of type I error, and therefore the observed associations should be interpreted as preliminary findings requiring external validation. Accordingly, the present findings should be considered exploratory and hypothesis-generating rather than confirmatory. To partially address potential confounding by baseline disease severity, exploratory logistic regression analyses adjusted for baseline LVEF were performed. Nevertheless, more comprehensive multivariable adjustment including multiple clinical covariates was not feasible because of the limited number of outcome events and the risk of overfitting. Although baseline LVEF, renal function, timing of diagnosis, and major heart failure therapies were evaluated, residual confounding from unmeasured variables cannot be completely excluded because of the retrospective design and limited sample size. In addition, because of the exploratory nature of the study and the limited sample size, no formal adjustment for multiple comparisons was performed. Therefore, the findings should be interpreted cautiously and require confirmation in larger prospective studies.

Second, laboratory parameters used to calculate SII and NAPLES were obtained from routine clinical practice. Although admission-day measurements were selected to ensure temporal standardization, dynamic changes in inflammatory markers during hospitalization could not be assessed due to the retrospective nature of the study.

Third, although patients with overt infection, malignancy, or systemic inflammatory diseases were excluded, subclinical inflammatory or metabolic conditions may still have influenced inflammatory and nutritional markers. In addition, certain potentially relevant variables—such as detailed nutritional assessments, micronutrient levels, cytokine profiles, natriuretic peptide levels, prothrombotic markers, right ventricular functional parameters, and myocardial strain imaging—could not be consistently evaluated because of the retrospective design and incomplete data availability. Although troponin measurements were available for all patients, different assay kits with varying upper reference limits were used during the study period, preventing reliable standardization and meaningful comparative analysis across the study cohort. Furthermore, clinical outcomes such as mortality, rehospitalization for heart failure, thromboembolic complications, or other major adverse cardiovascular events were not systematically assessed, limiting evaluation of the potential clinical relevance of these biomarkers beyond left ventricular recovery alone.

Fourth, follow-up echocardiographic intervals were not fully standardized across all patients, which may have influenced the timing of left ventricular recovery assessment. However, all patients had a minimum follow-up duration of 12 months, which ensured adequate evaluation of recovery status. In addition, the ROC analyses were performed in a relatively small cohort and were not externally validated. Therefore, the reported AUC values and optimal cut-off thresholds should be interpreted cautiously, as they may be susceptible to optimism bias and may not be generalizable to other PPCM populations. Formal internal validation techniques such as bootstrap resampling were not performed because the very limited sample size and low number of outcome events could still produce unstable and potentially unreliable estimates under such conditions. However, follow-up duration was similar between patients with and without recovery, reducing the likelihood that differences in follow-up length substantially influenced the observed associations.

Finally, although SII and NAPLES demonstrated significant associations with left ventricular recovery, this study was not designed to establish causality. Accordingly, the observed relationships should be interpreted as associative findings rather than evidence of direct mechanistic effects on myocardial recovery. Therefore, prospective, multicenter studies incorporating larger cohorts and serial biomarker measurements are warranted to validate these findings and to further clarify their role in clinical risk stratification.

Conclusion

In this study evaluating the association of the Naples Prognostic Score in peripartum cardiomyopathy, both NAPLES and SII were found to be associated with left ventricular functional recovery. The concomitant presence of elevated SII and high NAPLES scores may be associated with a lower likelihood of recovery.

By incorporating inflammatory, nutritional, and immune-related parameters, NAPLES reflects a broader multidimensional inflammatory, nutritional, and immune-related profile compared with traditional inflammatory markers in a complex condition such as PPCM. When interpreted together with SII, these biomarkers may offer complementary insight into biological factors potentially associated with myocardial recovery in PPCM.

These preliminary findings should be interpreted as exploratory and hypothesis-generating observations requiring validation in larger prospective multicenter studies before definitive conclusions regarding clinical applicability can be drawn.

Authors’ contributions

B.S.Ç. contributed to the conceptualization of the study, data collection, and drafting of the manuscript. A.C.Ç. contributed to the study design, supervision of the research process, and critical revision of the manuscript. Y.C. performed the statistical analysis and contributed to data interpretation. P.V. contributed to the literature review and manuscript editing. All authors read and approved the final version of the manuscript.

Funding

This research received no external funding.

Data availability

The data presented in this study are not publicly available due to privacy and ethical restrictions but are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

The study protocol was approved by the Sakarya University Faculty of Medicine Health Sciences Scientific Research Ethics Committee (Approval Number: E-43012747-050.04-567598-124, Date: 25 February 2026) and was conducted in accordance with the principles of the Declaration of Helsinki. Written informed consent was obtained from all participants included in the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.