The prognostic value of preoperative Naples prognostic score for overall survival and progression-free survival in colorectal cancer

Abstract

Background

This study aimed to explore the prognostic value of the Naples Prognostic Score (NPS) for progression-free survival (PFS), overall survival (OS) and sarcopenia in patients with colorectal cancer (CRC).

Methods

The Kaplan-Meier (KM) method was used to plot survival curves for PFS and OS, and the Log-rank test was applied to compare survival differences between the two groups. Receiver operating characteristic (ROC) curve analysis was performed to evaluate the predictive efficacy of NPS for PFS and OS in CRC patients. Univariate and multivariate Cox proportional hazards regression models were used to identify independent risk factors affecting PFS, OS, and sarcopenia. Nomograms were developed based on PFS and OS to forecast the 1-year, 3-year, and 5-year prognosis of CRC patients, while the concordance index (C-index) and calibration curves were used to assess the predictive performance of these nomograms.

Results

Kaplan-Meier analysis revealed that patients in the high-NPS (scores 3–4) group were significantly associated with poor prognosis (PFS: 48.9% vs. 66.0%, p < 0.001; OS: 48.9% vs. 66.0%, p < 0.001). Multivariate Cox regression confirmed that the high-NPS group was an independent risk factor for PFS (HR = 1.485, 95% CI: 1.241–1.777, p < 0.001), OS (HR = 1.452, 95% CI: 1.219–1.729, p < 0.001), and sarcopenia (HR = 1.876, 95% CI: 1.403–2.507, p < 0.001). For the nomograms developed based on the independent prognostic factors for PFS and OS, the areas under the ROC curve (AUC) for 1-year, 3-year, and 5-year outcomes were 0.807, 0.776, and 0.766 for PFS, and 0.772, 0.775, and 0.769 for OS, respectively. The C-indices of the nomograms were 0.795 for PFS and 0.815 for OS. Calibration curves demonstrated good agreement between predicted and actual values. Decision curve analysis (DCA) confirmed that the nomograms provided a better clinical net benefit than tumor staging alone.

Conclusions

The preoperative NPS is a promising prognostic scoring method for predicting poor outcomes in CRC patients. The prognostic nomogram based on NPS can serve as an effective tool for the comprehensive prognostic evaluation of CRC patients.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12885-026-15679-y.

Introduction

As the global population ages and its total size continues to grow, cancer has become one of the leading causes of death worldwide, posing a significant threat to human life expectancy. Colorectal cancer (CRC), among the most common malignancies globally, ranks third in terms of new diagnoses and second in cancer-related deaths [1]. In China, CRC incidence remains high: prior projections suggested that by 2022, it would already be the second most commonly diagnosed malignancy and the fifth leading cause of cancer-related deaths in the country [2]. Notably, early-stage CRC symptoms are subtle, with around 60% of patients diagnosed at an advanced stage. These advanced-stage patients face suboptimal treatment outcomes, high recurrence rates, and a 5-year survival rate below 40%. In contrast, early-stage patients who receive standardized treatment can achieve a 5-year survival rate exceeding 90% [3]. Identifying reliable predictive indicators is therefore crucial for assessing patient risk and formulating precise treatment regimens.

In recent years, advances in precision medicine research have led to accumulating evidence from clinical observations, molecular mechanism studies, and multi-center cohort analyses. This evidence strongly confirms the key role of peripheral blood biomarkers related to nutritional status and inflammatory responses in prognostic evaluation for cancer patients. These easily accessible blood indicators not only reflect the body’s overall physiological state but also connect closely to the tumor microenvironment, immune regulation, and treatment response through complex molecular networks—making them highly promising prognostic tools. When it comes to specific indicators, nutrition-related parameters include serum albumin, prealbumin, transferrin, cholesterol, and vitamin D. Among these, serum albumin is a classic marker for evaluating the body’s nutritional reserves; reduced levels often tie to tumor-related consumption, dietary disruptions, and impaired synthesis caused by chronic inflammation [4, 5]. In gastrointestinal malignancies like CRC and gastric cancer, low albumin levels have been confirmed to correlate significantly with shorter survival and higher risk of postoperative complications [6, 7]. Systemic inflammation, a key component of the tumor microenvironment, plays a pivotal role in tumorigenesis and progression [8]. Inflammation-associated indicators include white blood cell (WBC) count, neutrophil-to-lymphocyte ratio (NLR), monocyte-to-lymphocyte ratio (MLR), C-reactive protein (CRP), and cytokines such as interleukin-6 (IL-6) [9]. Neutrophils promote tumor angiogenesis and metastasis by secreting pro-inflammatory cytokines; lymphocytes, as core effector cells of the immune system, mediate the recognition and elimination of tumor cells; and monocytes play a critical role in immune regulation. Dysregulation of these blood cell ratios not only reflects the extent of impairment to the body’s anti-tumor immune function but also has been repeatedly confirmed as an independent adverse prognostic factor in various malignancies, including lung, breast, and liver cancer [10, 11]. A major advantage of these peripheral blood indicators is the convenience and minimal invasiveness of their detection. Compared with tumor tissue biopsy, blood sample collection is more tolerable for patients and can be repeated multiple times during treatment, enabling real-time monitoring of dynamic changes in disease status. The Naples Prognostic Score (NPS)—a novel scoring system developed in recent years—incorporates serum albumin (ALB), total cholesterol (TC), NLR, and lymphocyte-to-monocyte ratio (LMR). It comprehensively reflects the body’s inflammatory and nutritional status, making it a potential biomarker for predicting prognosis in various tumors [12, 13]. However, current research exploring the link between NPS and adverse prognosis in CRC patients remains relatively limited.

This retrospective cohort study was therefore designed to investigate the prognostic value of NPS for PFS and OS in CRC patients. The results will provide a scientific basis for the clinical application of NPS in CRC prognostic evaluation. Additionally, the study will develop a prediction model based on NPS, which can serve as a reference for individualized prognostic risk monitoring in CRC patients, help clinicians formulate more precise treatment regimens and management strategies, and thus provide more effective diagnostic and therapeutic support for this patient group.

Materials and methods

Study population

This study selected CRC patients who underwent surgical treatment at the Department of Colorectal and Anal Surgery, the First Affiliated Hospital of Guangxi Medical University, from January 2013 to December 2016 as the research subjects. The inclusion criteria were as follows: histopathologically confirmed colorectal cancer; complete albumin level, total cholesterol, neutrophils, lymphocytes, and monocytes, as well as other relevant clinical information; patients aged 18 years and above. The exclusion criteria were: receiving neoadjuvant radiotherapy and chemotherapy before surgery; coexisting with other malignant tumors; having severe renal insufficiency or immunodeficiency; suffering from acute or chronic inflammatory diseases. This study strictly adhered to the provisions of the Declaration of Helsinki during the research process and was approved by the Ethics Committee of the First Affiliated Hospital of Guangxi Medical University. Since a retrospective study design was adopted, informed consent was not required.

Data collection

Preoperative clinical data of patients were collected through the hospital’s electronic medical record system, covering aspects such as basic patient information, tumor - related information, laboratory test results, and treatment information. Basic patient information included gender, age, height, weight, and the presence of hypertension, diabetes, etc. Tumor - related information encompassed TNM staging, the presence of nerve or blood vessel invasion, tumor size, and tumor differentiation degree. Laboratory test data included tumor markers, albumin level, total cholesterol, neutrophils, lymphocytes, monocytes, and other indicators. Treatment information mainly recorded whether patients received radiotherapy, chemotherapy, or other treatment methods after surgery. The indicators used to calculate NPS — including serum albumin, total cholesterol, neutrophil-to-lymphocyte ratio (NLR), and lymphocyte-to-monocyte ratio (LMR) — as well as the indicators for calculating sarcopenia (height and weight), were measured on the day before surgery after the patient was admitted to the hospital.

Sarcopenia

The updated consensus on the diagnosis and treatment of sarcopenia in 2019 by the Asian Working Group for Sarcopenia (AWGS) was widely used as the diagnostic criteria for sarcopenia in the Asian population. That is, when the skeletal muscle mass index (SMI) is less than 6.92 Kg/㎡ in men and less than 5.13 Kg/㎡ in women, it is considered to have sarcopenia. The appendicular skeletal muscle mass (ASM) was estimated according to the anthropometric equation, and the calculation equation was as follows: 0.193×weight (kg) + 0.107×height (cm)-4.157×gender (male = 1, female = 2)-0.037×age (years)-2.631. The skeletal muscle index (SMI) was defined as ASM divided by the square of height (㎡), that is, SMI = ASM/Ht² (Kg/㎡).

Definition of NPS

NPS is defined by four clinical indicators, namely serum albumin (ALB), total cholesterol (TC), neutrophil-to-lymphocyte ratio (NLR), and lymphocyte-to-monocyte ratio (LMR) [14]. Specifically, a score of 0 is assigned to study participants who meet any of the following criteria: serum albumin ≥ 40 g/L, TC > 180 mg/dL, NLR < 2.96, or LMR > 4.44. In contrast, a score of 1 is assigned to those with serum albumin < 40 g/L, TC ≤ 180 mg/dL, NLR ≥ 2.96, or LMR ≤ 4.44. The total NPS is calculated as the sum of individual scores corresponding to the four aforementioned factors [14]. NPS is subjected to ROC curve analysis to identify the optimal cut-off value for discriminating prognostic disparities among patients with CRC patients. Furthermore, Youden’s index (defined as sensitivity + specificity − 1) is applied to screen out the cut-off threshold with maximum discriminatory efficacy for both PFS and OS. After determining the optimal cut-off value for NPS, patients are stratified into two groups: the low-NPS group (0–2 points) and the high-NPS group (3–4 points).

Follow - up

The follow - up methods for patients after surgery included telephone follow - up and outpatient review. Follow - up was conducted once every 3 months in the first 2 years after surgery, once every 6 months in the subsequent 3 years, and once a year after 5 years. The follow - up endpoint was the patient’s death or the last follow - up time (January 2024). The follow - up content mainly included inquiring about the patient’s symptoms and signs and conducting relevant examinations, such as tumor marker detection, imaging examinations, etc., to assess the disease recurrence and survival status of the patient. The primary endpoint indicators of the study were OS and PFS. OS was calculated from the date of surgery to the patient’s death or the last follow - up date; PFS was defined as the time from the date of surgery to local recurrence or distant metastasis of the disease.

Statistical analysis

Continuous variables were expressed as mean ± standard deviation (SD) or median (interquartile range), and categorical variables were expressed as number of cases (percentage). The chi - square test or Fisher’s exact test was used for the comparison of categorical variables, and the t - test was used for the comparison of continuous variables. The Kaplan - Meier method was used to draw survival curves, and the log - rank test was used to compare the OS and PFS of patients in different NPS groups to analyze the relationship between NPS and patient prognosis. Potential confounding factors incorporated into the Cox regression model were screened based on three criteria: (1) Factors with clinical relevance to the prognosis of colorectal cancer, including age, TNM stage, histological differentiation grade, and CEA level; (2) Variables with a P-value < 0.1 in univariate Cox regression analysis, so as to avoid omitting factors with potential prognostic effects; (3) Variables without significant multicollinearity confirmed by collinearity diagnosis (variance inflation factor [VIF] < 5). This screening strategy balanced clinical rationality with statistical rigor, and thus ensured the stability and interpretability of the multivariate Cox regression model. The Cox proportional hazards regression model was used to determine the independent association between NPS and OS and PFS in CRC patients. Univariate and multivariate analyses were performed using the Cox proportional hazards regression model to screen independent prognostic factors affecting OS and PFS. Based on the variables with statistical significance in the Cox regression model, a prediction model was constructed using R software, and a nomogram was generated. The discrimination ability and calibration of the prediction model were evaluated through internal validation curves, the concordance index (C - index), and the ROC curve. Decision curve analysis (DCA) was used to intuitively compare the benefits of the prediction model and traditional tumor staging in clinical applications. A p < 0.05 was considered as the standard for statistical significance. Statistical analysis was performed using the R statistical software package (version 4.0.2).

Results

Patient baseline characteristics

The study cohort had a mean age of 59.21 ± 12.65 years (SD), comprising 807 males (63.3%) and 467 females (36.7%). Of these, 625 cases (49.1%) were diagnosed with colon cancer and 649 (50.9%) with rectal cancer. There were 350 (27.5%) cases of disease recurrence and 519 (40.7%) deaths. Per clinicopathological staging, 679 cases (53.3%) were stage I-II and 595 (46.7%) stage III-IV. NPS scores for all enrolled patients ranged from 0 to 4. Statistical analysis showed the high-NPS group was significantly associated with male gender, advanced age, low Body Mass Index (BMI), higher colon cancer incidence, larger tumor diameter, elevated Carcinoembryonic Antigen (CEA) levels, and higher mortality. Compared with the low-NPS group, the high-NPS group had a significantly higher metastasis risk (13.1% vs. 7.3%, p = 0.001) and a 6.5-percentage-point higher overall recurrence rate (31.4% vs. 24.9%, p = 0.013). Comorbidity analysis revealed a significantly higher proportion of hypertension in the high-NPS group (21.3% vs. 15.0%, p = 0.005) and a higher proportion of diabetes (8.9% vs. 5.1%, p = 0.009). In terms of medical resource-related outcomes, high-NPS patients had a 2-day longer hospital stay and 3,415.32 RMB higher medical expenses than low-NPS patients (Table S1).

Kaplan - Meier survival curves of NPS and prognosis

The Kaplan-Meier analysis showed that compared with the low-NPS group, the high-NPS group had a significantly poorer 5-year PFS (48.9% vs. 66.0%, p < 0.001) (Fig. 1A). In addition, the 5-year OS of patients in the high-NPS group was significantly worse than that in the low-NPS group (48.9% vs. 66.0%, p < 0.001) (Fig. 1B). Subgroup analysis using Kaplan-Meier curves revealed that in the TNM staging subgroups, whether for Stage I-II (PFS, 64.3% vs. 77.4%, p = 0.002; OS, 66.5% vs. 79.8%, p = 0.001) or Stage III-IV (PFS, 27.1% vs. 46.8%, p < 0.001; OS, 29.6% vs. 49.9%, p < 0.001), the PFS of the high-NPS group was shorter than that of the low-NPS group (Figs. 2A, C). The OS data also showed a similar trend (Figs. 1B, D). For patients with colon cancer, both the PFS (47.1% vs. 69.9%, p < 0.001) and OS (49.1% vs. 71.7%, p < 0.001) of the high-NPS group were shorter than those of the low-NPS group (Figures S1A, B). For patients with rectal cancer, compared with the low-NPS group, the high-NPS group had significantly poorer PFS (48.6% vs. 61.7%, p = 0.003) and OS (45.7% vs. 58.3%, p < 0.001) (Figures S1C, D).

Fig. 1.

Kaplan-Meier curve of NPS in patients with colorectal cancer. Notes: A, Progression-free survival; B, Overall survival

Fig. 2.

Stratified Kaplan-Meier curve of NPS based on TNM stage subgroup in patients with colorectal cancer. Notes: A, Progression-free survival (Stage I-II); B, Overall survival (Stage I-II); C, Progression-free survival (Stage III-IV); D, Overall survival (Stage III-IV)

The relationship between NPS and PFS/OS

Cox regression analysis was used to explore the link between NPS and CRC prognosis. Univariate Cox regression showed high NPS was closely associated with poor PFS (HR: 1.602, 95% confidence interval [CI]: 1.356–1.893, p < 0.001). After adjusting for confounding factors, high NPS remained an independent risk factor for poor PFS (HR: 1.485, 95% CI: 1.241–1.777, p < 0.001) (Table 1). For OS, univariate Cox regression indicated high-NPS patients had a 1.662-fold higher risk of poor outcomes than low-NPS patients (HR: 1.662, 95% CI: 1.399–1.974, p < 0.001). Multivariate Cox regression confirmed high NPS as an independent risk factor for poor OS (HR: 1.452, 95% CI: 1.219–1.729, p < 0.001) (Table 2). Multivariate forest plots for PFS/OS—used to explore NPS-HR relationships across subgroups—showed high NPS was an independent risk factor in most subgroups (Figures S2A, B), with the most notable effects seen in early-stage (T1-2) and elderly patients.

Table 1.

Univariate and multivariate Cox regression analysis of clinicopathological characteristics associated with Progression-free survival in CRC patients

Characteristics	Progression-free survival
	Univariate analysis		Multivariate analysis
	HR (95%CI)	P value	HR (95%CI)	P value
Age	0.973 (0.818–1.158)	0.757	1.302 (1.089–1.558)	0.004
T stage	2.421 (1.907–3.073)	< 0.001	1.527 (1.167–1.999)	0.002
N stage		< 0.001		< 0.001
N0	Ref.		Ref.
N1	1.856 (1.52–2.267)	< 0.001	1.52 (1.224–1.887)	< 0.001
N2	4.101 (3.342–5.033)	< 0.001	2.626 (2.072–3.327)	< 0.001
M stage	5.328 (4.309–6.588)	< 0.001	3.225 (2.566–4.054)	< 0.001
Perineural invasion (Yes)	1.835 (1.449–2.323)	< 0.001	1.151 (0.877–1.51)	0.311
Vascular invasion (Yes)	2.058 (1.697–2.496)	< 0.001	1.23 (0.979–1.544)	0.075
Differentiation (high/medium)	0.671 (0.533–0.844)	0.001	0.746 (0.587–0.949)	0.017
Tumor size (≥ 5 cm)	1.211 (1.025–1.431)	0.025	1.011 (0.844–1.212)	0.904
CEA (≥ 5ng/ml)	1.917 (1.621–2.267)	< 0.001	1.439 (1.199–1.728)	< 0.001
NPS (3–4 score)	1.602 (1.356–1.893)	< 0.001	1.485 (1.241–1.777)	< 0.001

CRC colorectal cancer

Table 2.

Univariate and multivariate Cox regression analysis of clinicopathological characteristics associated with overall survival in CRC patients

Characteristic	Overall survival
	Univariate analysis		Multivariate analysis
	HR (95%CI)	P value	HR (95%CI)	P value
Age	1.31 (1.1–1.56)	0.002	1.277 (1.073–1.518)	0.006
T stage	2.552 (1.983–3.283)	< 0.001	1.478 (1.145–1.907)	0.003
N stage		< 0.001		< 0.001
N0	Ref.
N1	1.849 (1.503–2.275)	< 0.001	1.516 (1.231–1.868)	< 0.001
N2	4.123 (3.341–5.087)	< 0.001	2.758 (2.192–3.471)	< 0.001
M stage	5.499 (4.434–6.819)	< 0.001	3.11 (2.482–3.897)	< 0.001
Perineural invasion (Yes)	1.792 (1.404–2.286)	< 0.001	1.187 (0.913–1.542)	0.200
Vascular invasion (Yes)	2.065 (1.695–2.516)	< 0.001	1.213 (0.971–1.514)	0.089
Differentiation (high/medium)	0.616 (0.489–0.775)	< 0.001	0.825 (0.651–1.046)	0.113
Tumor size (≥ 5 cm)	1.329 (1.119–1.579)	0.001	0.905 (0.759–1.08)	0.268
CEA (≥ 5ng/ml)	1.958 (1.647–2.328)	< 0.001	1.447 (1.212–1.728)	< 0.001
NPS (3–4 score)	1.662 (1.399–1.974)	< 0.001	1.452 (1.219–1.729)	< 0.001

CRC colorectal cancer

The relationship between NPS and sarcopenia

A total of 247 patients were diagnosed with sarcopenia: 113 (14.66%) in the low-NPS group and 134 (26.64%) in the high-NPS group. Logistic regression was used to investigate sarcopenia risk factors in CRC patients. Univariate logistic regression showed high-NPS patients had a 2.262-fold higher sarcopenia risk than low-NPS patients (OR: 2.262, 95% CI: 1.721–2.973, p < 0.001). Multivariate logistic regression identified high NPS as the strongest independent risk factor for sarcopenia among all clinicopathological factors (OR: 1.876, 95% CI: 1.403–2.507, p < 0.001) (Table 3).

Table 3.

Association between NPS and Complication of CRC patients

Characteristic	sarcopenia
	Univariate analysis		Multivariate analysis
	HR (95%CI)	P value	HR (95%CI)	P value
Age	1.647 (1.249-2.172)	<0.001	1.542 (1.147 - 2.071)	0.004
Perineural invasion (Yes)	1.764 (1.184-2.629)	0.005	1.765 (1.113 - 2.798)	0.016
Vascular invasion (Yes)	1.439 (1.028-2.013)	0.034	1.153 (0.782 - 1.700)	0.471
Hypertension (Yes)	1.449 (1.038-2.024)	0.029	1.160 (0.812 - 1.658)	0.414
Tumor size (≥5cm)	1.397 (1.067-1.831)	0.015	1.164 (0.875 - 1.550)	0.296
Surgical method (laparoscopy)	0.549 (0.418-0.720)	<0.001	0.661 (0.496 - 0.881)	0.005
Operating time (240 min)	1.605 (1.199-2.149)	0.001	1.490 (1.095 - 2.028)	0.011
Blood loss (100mL)	1.741 (1.294-2.344)	<0.001	1.396 (1.018 - 1.915)	0.038
NPS (3-4 score)	2.262 (1.721-2.973)	<0.001	1.876 (1.403 - 2.507)	<0.001

CRC colorectal cancer

Establishment of NPS - based prediction nomograms

Multivariate Cox regression identified 7 independent prognostic factors for PFS: age, T stage, N stage, M stage, differentiation degree, CEA, and NPS. For OS, 6 independent prognostic factors were determined: age, T stage, N stage, M stage, CEA, and NPS (Table 2). Based on these key factors, NPS-based nomograms were constructed to predict PFS and OS in CRC patients. By summing the scores of each variable, predicted probabilities of 1-year, 3-year, and 5-year PFS and OS were calculated—with higher total scores corresponding to lower probabilities (Figs. 3 and 4). The AUC values of the PFS/OS nomograms for 1-year, 3-year, and 5-year outcomes were 0.807, 0.776, 0.766 (PFS) and 0.772, 0.775, 0.769 (OS) respectively (Figures S3A, B). The C-indices were 0.728 (95% CI: 0.685, 0.771) for the PFS nomogram and 0.723 (95% CI: 0.680, 0.766) for the OS nomogram. Calibration curves showed good agreement between predicted and actual probabilities of 1-year, 3-year, and 5-year PFS (Figure S4A) and OS (Figure S4B). DCA—used to compare clinical benefits of the NPS-based nomogram versus traditional tumor staging—showed the nomogram outperformed traditional staging for both PFS and OS over 1–5 years (Figures S5A, B). Additionally, patients were divided into high-score and low-score groups based on the nomogram’s median score. Kaplan-Meier survival analysis showed high-score patients had significantly worse PFS and OS than low-score patients (Figures S6A, B).

Fig. 3.

Construction the PFS nomogram in CRC patients

Fig. 4.

Construction the OS nomogram in CRC patients

Discussion

Since the 19th century, the link between cancer and inflammation has remained a core focus in oncology research. Numerous observational studies have shown tumors often arise in areas of chronic inflammation, and inflammatory cell infiltration is widespread in the tumor microenvironment—further highlighting inflammation’s key role in tumor initiation and progression [15]. Timely assessment of inflammatory status in cancer patients is essential for understanding disease progression and selecting appropriate treatment strategies [16]. The interaction between cancer, inflammation, and nutrition underscores the importance of comprehensive evaluation of these factors in guiding cancer treatment [17]. The impacts of malnutrition and systemic inflammation on cancer occurrence, tumor growth, and progression have been confirmed in various cancers—including CRC—spurring the search for biomarkers and development of prognostic scoring systems [18]. NPS stands out for its comprehensiveness and practicality, covering two key dimensions: nutrition and inflammation. By integrating the advantages of inflammatory cells, albumin, and cholesterol, it emerges as an ideal indicator for CRC prognosis.

The biological mechanisms underlying the prognostic value of NPS in CRC are closely associated with the complex interactions among systemic inflammation, malnutrition, and tumor progression. Specifically, an elevated NLR reflects an imbalance between proinflammatory neutrophils and antitumor lymphocytes: neutrophils secrete matrix metalloproteinases (MMPs) and vascular endothelial growth factor (VEGF), which promote tumor angiogenesis and extracellular matrix degradation, thereby facilitating tumor invasion and metastasis. In contrast, a decreased lymphocyte count impairs the tumor cell recognition ability and cytotoxic effect of immune cells, thus compromising the antitumor immune response. A reduction in serum ALB levels not only indicates systemic nutritional depletion but also reflects a state of chronic inflammation, as proinflammatory cytokines (interleukin-6, tumor necrosis factor-α) inhibit hepatic albumin synthesis; this dual effect exacerbates tumor-associated cachexia and simultaneously impairs immune cell functions (T cell proliferation and cytokine secretion). Hypocholesterolemia (total cholesterol ≤ 180 mg/dL) further impairs immune surveillance function through a mechanism that reduces the cell membrane fluidity of immunocompetent cells (T cells, natural killer cells), which weakens their capacity to migrate to the tumor microenvironment and eliminate malignant cells [19]. For sarcopenia, its association with a high NPS is mediated by inflammation-driven muscle catabolism: proinflammatory cytokines (TNF-α, IL-6) activate the ubiquitin-proteasome system (atrogin-1, muscle RING-finger protein 1) and simultaneously inhibit the mammalian target of rapamycin (mTOR) pathway, ultimately leading to skeletal muscle protein degradation. Meanwhile, nutritional deficiency (hypoalbuminemia, hypocholesterolemia) reduces muscle energy supply and protein synthesis, thereby forming a vicious cycle of systemic inflammation, malnutrition, sarcopenia, tumor progression, which ultimately exacerbates poor prognosis in CRC patients [20].

However, since albumin concentration can be affected by liver function changes and fluid volume shifts [21], some researchers have suggested incorporating plasma cholesterol levels to optimize nutritional status evaluation [22]. Multiple studies have shown low cholesterol often predicts poor prognosis in tumors like CRC [23]. Hypocholesterolemia impairs cell membrane fluidity, reducing the mobility of cell surface receptors and their ability to transmit transmembrane signals [24]. This prevents immunocompetent cells from effectively eliminating cancer cells with membrane alterations [25]. Additionally, NLR and LMR—key indicators of systemic inflammation and immune status—connect closely to tumor progression, metastasis, and treatment resistance. For instance, previous studies have noted inflammatory burden as a prognostic biomarker in various cancers [19, 26].

This study is the first to systematically confirm, in a large CRC cohort, that high NPS is an independent predictor of PFS and OS. Compared with the low-NPS group, high-NPS patients had a 1.485-fold higher risk of disease progression and 1.452-fold higher risk of death. Given NPS’s simple calculation and consistent standards across studies [27], our research offers promising insights for its prospective use in CRC prognostic assessment. The study also found high NPS associated with advanced age, low BMI, high CEA levels, high recurrence rates, and high metastasis rates—suggesting systemic inflammatory status may connect closely to tumor biological behavior, particularly in promoting distant metastasis. The underlying mechanism may involve inflammatory factors facilitating epithelial-mesenchymal transition (EMT) and angiogenesis. Notably, NPS showed stable predictive value across subgroup analyses, with higher hazard ratios seen especially in T1-2 stage and elderly patients. This may stem from more prominent baseline inflammation and poorer nutritional reserves in elderly patients, highlighting NPS’s unique value in identifying occult early-stage tumor progression and risk stratification for vulnerable populations.

We found high NPS (3–4 points) to be one of the strongest independent risk factors for sarcopenia—providing clinical evidence for the “inflammation-driven cancer-related sarcopenia” theory. In chronic inflammation, pro-inflammatory factors (e.g., tumor necrosis factor-α, TNF-α) can activate the ubiquitin-proteasome system to accelerate muscle catabolism [28], while low albumin and cholesterol further worsen muscle energy metabolism disorders—forming an “inflammation → sarcopenia → poor prognosis” cascade. This aligns with findings from Masashi Utsumi in pancreatic cancer, where the synergistic effect of an inflammation-nutrition composite score and sarcopenia significantly amplified poor prognostic signals [29]. In clinical practice, this suggests preoperative nutritional support combined with anti-inflammatory intervention may improve prognosis by addressing sarcopenia in high-NPS patients—a strategy showing potential in CRC preoperative prehabilitation studies [30]. It is important to note that no single indicator can fully assess cancer patients’ condition changes or prognosis. Comprehensive evaluation of multi-dimensional data—including clinical symptoms, signs, imaging, and histopathology—is therefore necessary. To address this, we used Cox regression to screen independent prognostic variables in CRC patients and built a standardized NPS-based nomogram, further enhancing its clinical translational value. After integrating traditional indicators like age and TNM stage, the 1–5-year PFS and OS AUC values remained above 0.76, with C-indices of 0.728 and 0.723 respectively. Good agreement on calibration curves indicates superior predictive accuracy compared with traditional TNM staging. Furthermore, DCA confirmed the nomogram outperforms traditional TNM staging and is suitable for guiding individualized CRC treatment decisions.

This study’s strengths include a large sample size, strict inclusion/exclusion criteria, and comprehensive subgroup analyses. This study adopted a single-center retrospective study design, which inherently harbors selection bias. The reason is that all study subjects were recruited from a Grade A tertiary hospital in Guangxi, China, and their demographic and clinical characteristics exhibit relative homogeneity (regional disease prevalence). This may restrict the generalization and application of the study findings to diverse populations (different ethnic groups, various healthcare settings). Some unmeasured confounding factors that may simultaneously influence NPS and CRC prognosis failed to be adequately adjusted for. The key limitation is that this study has not performed external validation of the nomogram constructed based on the NPS; thus, there is an urgent need to conduct multicenter prospective cohort studies enrolling heterogeneous study populations to verify the discriminatory efficacy and calibration accuracy of this model. External validation represents a key step in translating prognostic models into clinical practice, and its core function is to evaluate whether the model can sustain stable efficacy in settings beyond the original study population. Without the support of external validation, the clinical application value of this nomogram in non-local medical institutions or diverse healthcare systems remains to be elucidated. Future studies may verify the generalizability of the NPS through prospective multicenter cohorts; optimize the assessment method of muscle mass by combining radiomics technology; and explore the clinical efficacy of targeted anti-inflammatory therapy combined with nutritional intervention under the guidance of the NPS. In addition, integrating the NPS with other molecular biomarkers (circulating tumor DNA) may further improve the accuracy of prognostic assessment — this also constitutes a prominent research direction in the era of precision oncology.

To fully unlock the clinical potential of NPS in the diagnosis and management of CRC, several specific directions for future research are outlined as follows: First, perform dynamic monitoring of NPS throughout the treatment course to investigate whether temporal changes in NPS correlate with treatment response and long-term clinical outcomes. Such dynamic assessment would facilitate the identification of patients at high risk for early recurrence, thereby enabling timely adjustment of individualized treatment strategies. Second, explore the integrated application of NPS with novel molecular biomarkers; combined detection of inflammatory-nutritional scores and molecular features may yield a synergistic effect on prognostic stratification efficacy, outperforming any single biomarker alone. For example, a composite prognostic model integrating NPS with postoperative ctDNA status could stratify patients into low-, moderate-, and high-recurrence risk groups with enhanced precision. Third, it is imperative to conduct interventional studies to evaluate whether targeted interventions for patients with a high NPS can mitigate sarcopenia risk, improve treatment tolerance, and ultimately optimize long-term survival outcomes. Fourth, prospective multicenter cohort studies enrolling participants of diverse ethnicities, age strata, and treatment modalities are needed to validate the external generalizability of the NPS-based nomogram and extend its utility to a broader spectrum of clinical settings.

Conclusion

This study confirms that preoperative NPS—a scoring method integrating inflammation and nutritional status assessment—holds significant promise for predicting PFS and OS in CRC patients. It not only accurately forecasts survival outcomes but also effectively identifies patients at high sarcopenia risk, providing a reliable quantitative basis for formulating individualized clinical treatment strategies. The NPS-based prognostic nomogram serves as an effective tool for comprehensive CRC prognostic evaluation. Its clinical application is expected to help clinicians accurately identify high-risk patients, optimize treatment decisions, and ultimately improve patient survival outcomes.

Abbreviations

NPS

Naples Prognostic Score

PFS

Progression-free survival

OS

Overall survival

CRC

Colorectal cancer

KM

Kaplan-Meier

ROC

Receiver operating characteristic

C-index

Concordance index

AUC

areas under the ROC curve

DCA

Decision curve analysis

WBC

White blood cell

NLR

Neutrophil-to-lymphocyte ratio

MLR

Monocyte to-lymphocyte ratio

CRP

C-reactive protein

IL-6

Interleukin-6

ALB

Albumin

TC

Total cholesterol

LMR

Lymphocyte-to-monocyte ratio

AWGS

Asian Working Group for Sarcopenia

SMI

Skeletal muscle mass index

ASM

Appendicular skeletal muscle mass

SD

Standard deviation

CEA

Carcinoembryonic Antigen

EMT

Epithelial-mesenchymal transition

Authors’ contributions

Jialiang Gan, and Hailun Xie carried out the design of this study, analyses of statistics and draft the manuscript. Taiqi Chen, Bin Lan, Qi Zhou, Hailun Xie, and Shuangyi Tang carried out collection of the statistics and prepared the manuscript. All authors read and approved the final manuscript.

Funding

This study was supported by the Young Elite Scientist Sponsorship Program by Cast (YESS20220687), the Youth Science Foundation Project of Guangxi Medical University (GXMUYSF 202548), and the 18th batch of Special Funding from the China Postdoctoral Science Foundation (2025T180639).

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study followed the Helsinki declaration. All participants signed an informed consent form and this study was approved by the Institutional Review Board of the hospital (Registration number: NO.2022-KY-(043)).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.