Diagnostic implications of neutrophil-to-lymphocyte ratio, platelet-to-lymphocyte ratio, and systemic immune-inflammatory index for gastric carcinoma

Abstract

BACKGROUND

The diagnosis of gastric carcinoma (GC) is essential for improving clinical outcomes. However, the biomarkers currently used for GC screening are not ideal.

AIM

To explore the diagnostic implications of the neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and systemic immune-inflammatory index (SII) for GC.

METHODS

The baseline data of 133 patients with GC and 134 patients with precancerous gastric conditions admitted between January 2022 and December 2023 were retrospectively analyzed. The information on peripheral blood platelet, neutrophil, and lymphocyte counts in each patient was collected, and the NLR, PLR, and SII levels of both groups were calculated. Additionally, multivariate logistic regression analysis was conducted, and the diagnostic implications of NLR, PLR, and SII in differentiating patients with precancerous gastric conditions, compared with those with GC, were analyzed through receiver operating characteristic (ROC) curves.

RESULTS

The data indicated that NLR, PLR, and SII had abnormally increased levels in the patients with GC. Gender and body mass index were risk factors for the occurrence of GC. ROC data revealed that the areas under the curve of three patients with precancerous gastric conditions, who were differentiated from those with GC, were 0.824, 0.787, and 0.842, respectively.

CONCLUSION

NLR, PLR, and SII are all abnormally expressed in GC and have diagnostic implications, especially when used as joint indicators, in distinguishing patients with precancerous gastric conditions from those with GC.

Key Words: Gastric carcinoma; Neutrophil-to-lymphocyte ratio; Platelet-to-lymphocyte ratio; Precancerous gastric conditions; Systemic immune-inflammatory index

Core Tip: This study included 133 patients with gastric carcinoma (GC) and 134 with precancerous gastric conditions. Routine peripheral blood information of these patients was collected, and the systemic immune-inflammatory index (SII), neutrophil-to-lymphocyte ratio (NLR), and platelet-to-lymphocyte ratio (PLR) levels in both groups were calculated and compared. In-depth analysis confirmed that NLR, PLR, and SII exhibited abnormally increased levels in GC and, therefore, had diagnostic significance for distinguishing patients with precancerous gastric conditions from those with GC. Additionally, multivariate logistic regression analysis confirmed that gender and body mass index were risk factors for the occurrence of GC. The findings can provide convenient, low-cost, and highly efficient biological indicators for screening for early GC, as well as more clinical evidence and references for the prevention and management of early GC.

INTRODUCTION

Gastric carcinoma (GC) remains a major cause of cancer-related mortality worldwide. The risk of developing GC in some regions and populations has remained high or even increased over the past five decades[1,2]. There are 1.2 million new GC cases and nearly 900000 deaths globally each year, with a poor prognosis of a five-year survival rate lower than 30%[3,4]. GC is a heterogeneous disease with lesions confined to the mucosa or submucosa, atypical symptoms at the early stage, and no lymph node metastasis. It usually presents with specific symptoms only at the advanced stage[5], thus impeding timely intervention and significantly reducing overall survival[6].

The diagnosis of GC is valuable for improving the course of the disease[7]. Endoscopic or radiology-based screening has been used in high-incidence countries and has shown some benefits; however, they are not cost-effective in low-incidence populations[8]. Presently, commonly used biomarkers are not ideal for GC screening. Therefore, there is a need to search for new GC diagnostic biomarkers to improve GC outcomes[9].

This study aims to study the correlation between GC and various serum biomarkers that can be more easily obtained. For example, the neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio (PLR) can be calculated from the results of routine blood tests; both of them reflect the systemic inflammatory response and have a certain correlation with the prognosis of various malignancies[10,11]. Additionally, NLR can be utilized to predict tumor prognoses and lymph node metastases in GC, and PLR can be employed to predict the pathological progression of clinical TNM stage 3 diseases and lymph node dissection number < 28 in patients with GC[12,13]. Also, the systemic immune-inflammatory index (SII) is a biomarker indicating the body’s systemic inflammatory responses, and it may play a more effective role in indicating inflammation than NLR and PLR[14]. Furthermore, to some extent, it can be applied in patients with GC undergoing endoscopic submucosal dissection for assessing surgical risks[15]. In the study by Wang and Zhu[16], SII was used for predicting the prognosis of patients undergoing radical GC surgery and evaluating their disease-free and overall survival.

NLR, PLR, and SII are hematological indexes composed of neutrophil (NEUT), lymphocyte (LYM), and platelet (PLT) counts, respectively, and their value in GC screening requires further analysis.

MATERIALS AND METHODS

General information

The inclusion criteria were: Diagnosis of GC or a precancerous gastric condition[17]; first diagnosis; no operation, radiotherapy, or chemotherapy before collection of blood specimens; complete clinical data. The exclusion criteria were: Administration of antiplatelet drugs or anticoagulant drugs, or both, within 3 months before the blood test; use of anti-inflammatory or blood transfusion treatment, or both, one month before the blood test; preoperative coexistence of severe cardiovascular diseases, infectious diseases, or autoimmune diseases; other malignant tumors; previous radiotherapy, chemotherapy, or surgery for GC; hematological disorders or hypersplenism; missing clinical data. The hospital’s Ethics Committee reviewed and ratified this study’s research protocols. Lastly, the baseline data of 133 patients with GC and 134 patients with precancerous gastric conditions admitted between January 2022 and December 2023 were retrospectively collected.

Patient data collection

The basic information about the patients, such as name, gender, age, height, weight, admission time, admission number, and discharge diagnosis, was collected, and their endoscopic examination results and pathological findings were recorded. The first peripheral blood routine (red cell distribution width, NEUT, LYM, and PLT counts) of the patients with GC and those with precancerous gastric conditions upon admission was collected and used to calculate SII, NLR, and PLR levels using the following formulas: SII = (PLT × NEUT)/LYM; NLR = NEUT/LYM; PLR = PLT/LYM.

Statistical analysis

All data were analyzed using SPSS version 21.0. The Kolmogorov-Smirnov test was employed to conduct normality testing on the measurements. Age, NLR, PLR, SII, and other measurements were represented as mean ± SD, and two-tailed t-tests were used for between-group comparisons. Cell counts, presented as rates (percentages), were compared between the groups using χ² tests. Multivariate logistic regression was adopted to analyze the risk factors influencing GC. The area under the curve (AUC) for the receiver operating characteristics (ROCs) was applied to assess the efficacy of the indicators in diagnosing GC. After calculation, 133 patients with GC and 134 patients with precancerous gastric conditions were included in this study. They underwent stringent screening with inclusion and exclusion criteria and fulfilled the minimum sample size requirement (approximately 71). All analyses used P < 0.05 as the threshold for statistical significance.

RESULTS

Analysis of clinical data of GC cases and those with precancerous gastric condition

The comparative analysis of the patient’s clinical data (Table 1) revealed significant differences in gender, age, body mass index (BMI), NEUT, and PLT between the patients with precancerous gastric conditions and those with GC; the GC group had a higher proportion of male patients, an older age, a lower BMI, and higher NEUT and PLT (P < 0.05). No significant inter-group difference was found for LYM (P > 0.05).

Table 1 Analysis of clinical data of gastric carcinoma and precancerous gastric condition groups.

Clinical data	Precancerous gastric condition (n = 134)	GC (n = 133)	χ2/t	P value
Gender, n (%)			9.504	0.002
Male	69 (51.49)	93 (69.92)
Female	65 (48.51)	40 (30.08)
Age (year)	61.76 ± 9.31	66.27 ± 10.54	3.706	< 0.001
BMI (kg/m2)	23.45 ± 3.01	22.26 ± 3.20	3.130	0.002
Neutrophil count	2.64 ± 0.78	3.44 ± 1.28	6.172	< 0.001
Lymphocyte count	1.88 ± 0.56	1.87 ± 0.55	0.147	0.883
Platelet count	199.46 ± 51.28	235.90 ± 76.38	4.580	< 0.001

GC: Gastric carcinoma; BMI: Body mass index.

Levels of NLR, PLR, and SII in the two groups

The results demonstrated markedly higher levels of NLR, PLR, and SII in the patients with GC than those with precancerous gastric conditions (P < 0.001; Figure 1).

Figure 1 Levels of neutrophil-to-lymphocyte ratio, platelet-to-lymphocyte ratio, and systemic immune-inflammatory index in the gastric carcinoma and precancerous gastric condition groups. A: Neutrophil-to-lymphocyte ratio levels in the two groups; B: Platelet-to-lymphocyte ratio levels in the two groups; C: Systemic immune-inflammatory index levels in the two groups. ^aP < 0.001 in the inter-group comparison. GC: Gastric carcinoma; NLR: Neutrophil-to-lymphocyte ratio; PLR: Platelet-to-lymphocyte ratio; SII: Systemic immune-inflammatory index.

Analysis of risk factors affecting GC

We treated the occurrence of GC as the dependent variable and gender, age, BMI, NEUT, LYM, PLT, NLR, PLR, and SII as independent variables. After assigning values to these variables and conducting binary logistic regression analysis, we found that rather than age, NEUT, LYM, PLT, NLR, PLR, and SII (P > 0.05), gender and BMI were the risk factors for the occurrence of GC (P < 0.05; Table 2 and Table 3).

Table 2 Variable assignment.

Variable	Assignment method
Gender	Female = 0, male = 1
Age (year)	Continuous variable
BMI (kg/m2)	Continuous variable
Neutrophil count	Continuous variable
Lymphocyte count	Continuous variable
Platelet count	Continuous variable
NLR	Continuous variable
PLR	Continuous variable
SII	Continuous variable

NLR: Neutrophil-to-lymphocyte ratio; PLR: Platelet-to-lymphocyte ratio; SII: Systemic immune-inflammatory index; BMI: Body mass index.

Table 3 Analysis of risk factors affecting gastric carcinoma.

Variable	β	SE	Wald	P value	Exp (β)	95%CI
Gender	0.974	0.365	7.111	0.008	2.649	1.295-5.420
Age (year)	0.033	0.017	3.755	0.053	1.034	1.000-1.069
BMI (kg/m2)	-0.145	0.063	5.325	0.021	0.865	0.765-0.978
Neutrophil count	-0.457	0.704	0.421	0.517	0.633	0.159-2.519
Lymphocyte count	0.718	0.400	3.213	0.073	2.049	0.935-4.491
Platelet count	-0.004	0.012	0.106	0.745	0.996	0.972-1.020
NLR	0.293	1.428	0.042	0.837	1.340	0.082-22.007
PLR	-0.002	0.016	0.012	0.911	0.998	0.968-1.030
SII	0.012	0.008	2.118	0.146	1.012	0.996-1.029

95%CI: 95% confidence interval; NLR: Neutrophil-to-lymphocyte ratio; PLR: Platelet-to-lymphocyte ratio; SII: Systemic immune-inflammatory index; BMI: Body mass index.

Diagnostic implications of NLR, PLR, and SII for GC

By analyzing the potential of each indicator in diagnosing GC with the ROC curve, we found that the AUC of NLR in diagnosing GC was 0.824, with a specificity, sensitivity, and optimal cutoff of 91.04%, 65.41%, and 1.975, respectively. The AUC of PLR in diagnosing GC was 0.787, with a specificity, sensitivity, and optimal cutoff of 81.34%, 63.16%, and 136.000, respectively. The AUC of SII in diagnosing GC was 0.842, with the specificity, sensitivity, and optimal cutoff of 82.84%, 74.44%, and 372.700, respectively. Lastly, the AUC of the combined diagnosis (LR, PLR, and SII) of GC was 0.848, with a specificity, sensitivity, and optimal cutoff of 81.34%, 76.69%, and 0.433, respectively (Figure 2 and Table 4).

Figure 2 Receiver operating characteristic analysis of the diagnostic implications of neutrophil-to-lymphocyte ratio, platelet-to-lymphocyte ratio, and systemic immune-inflammatory index for gastric carcinoma. A: The receiver operating characteristic (ROC) curve of neutrophil-to-lymphocyte ratio (NLR) in the diagnosis of gastric carcinoma (GC); B: The ROC curve of platelet-to-lymphocyte ratio (PLR) in the diagnosis of GC; C: The ROC curve of systemic immune-inflammatory index (SII) in the diagnosis of GC; D: The ROC curve of NLR + PLR + SII in the diagnosis of GC. NLR: Neutrophil-to-lymphocyte ratio; PLR: Platelet-to-lymphocyte ratio; SII: Systemic immune-inflammatory index.

Table 4 Diagnostic implications of neutrophil-to-lymphocyte ratio, platelet-to-lymphocyte ratio, and systemic immune-inflammatory index for gastric carcinoma.

Indicators	AUC	SE	P value	Specificity (%)	Sensitivity (%)	Optimal cutoff
NLR	0.824	0.026	< 0.001	91.04	65.41	1.975
PLR	0.787	0.028	< 0.001	81.34	63.16	136.000
SII	0.842	0.025	< 0.001	82.84	74.44	372.700
Combined detection	0.848	0.024	< 0.001	81.34	76.69	0.433

AUC: Area under the curve; GC: Gastric carcinoma; NLR: Neutrophil-to-lymphocyte ratio; PLR: Platelet-to-lymphocyte ratio; SII: Systemic immune-inflammatory index.

DISCUSSION

The occurrence of GC is strongly linked to Helicobacter pylori infection, and its tumor pathology is highly associated with inflammation[18,19]. Inflammation not only participates in the onset and progression of diseases, such as tumors, but may also mediate tumor metastasis and other processes of disease deterioration[20]. Hence, we believe that tracking the levels of inflammation-related indicators may help facilitate GC screening.

Our results indicated that compared with the NLR, PLR, and SII values in patients with precancerous gastric conditions, those in patients with GC were significantly increased. Our observation suggests that all indicators may be involved in GC tumor lesions and is consistent with the results by Nguyen et al[21] that the NLR and PLR of patients with GC are statistically elevated compared to those in healthy subjects. After analyzing the clinical data of patients with precancerous conditions and those with GC, we found that gender, age, BMI, NEUT, and PLT had a certain significant correlation with GC. According to the binary logistic regression analysis, age, NEUT, LYM, PLT, NLR, PLR, and SII were not significant factors for the occurrence of GC, whereas gender and BMI were. This finding suggests that male patients and those with a low BMI are more predisposed to GC.

Finally, we evaluated the diagnostic performance of NLR, PLR, and SII (combined diagnosis) in GC. It was discovered that the AUCs of NLR, PLR, and SII for screening GC were 0.824, 0.787, and 0.842, respectively, and the AUC of the combination of the three was 0.848. Although the combined diagnosis had the largest AUC, it was similar to that of SII, which had a larger AUC of SII than NLR and PLR. SII had the largest AUC for single-indicator screening, suggesting that SII has the highest potential screening value for GC and might even be comparable to that of the combined screening. Meanwhile, NLR had the highest specificity, and the combined diagnosis had the highest sensitivity. Nevertheless, the comprehensive data of specificity and sensitivity of SII in differentiating GC were relatively more ideal. The AUCs of NLR and PLR in diagnosing GC in Karra et al[22] are similar to our research results; however, the AUCs of NLR and PLR in diagnosing GC in this study are higher. In addition, Zhang et al[23] suggest that NLR, PLR, and SII can all be used to screen GC, with SII having the highest diagnostic efficacy; combining the three indicators can further increase the efficacy, consistent with our findings. Also, Lai et al[24] suggest that SII, as a non-invasive indicator, can predict changes in gastric mucosal metaplasia (especially for the middle-aged and elderly population); this observation may partially explain the efficacious diagnostic potential of SII for GC. Previous evidence also indicates that NLR, PLR, and SII have relatively high efficacy for predicting advanced GC and that NLR is a poor prognosis indicator after immunotherapy for advanced GC[25].

While this study has confirmed that the three indicators, NLR, PLR, and SII, all possess high diagnostic implications for GC, there remain questions to be investigated. First, this study is not a prospective design; conducting a prospective study in the future may reduce the collection biases. Second, follow-up and prognosis analyses should be performed. Third, conducting a larger-scale multicenter study will enhance universality and validate these findings in different populations. Fourth, exploring the causal relationship or mechanism related to the elevated levels of biomarkers in patients with GC and the biological pathways underlying these associations will enhance our understanding of the mechanisms of GC. Fifth, an in-depth analysis of the potential interactions between gender, BMI, and indicators, such as NLR, PLR, and SII, will help identify potential factors or effect modifiers. These analyses will supplement and deepen our understanding of the prognostic predictive value of NLR, PLR, and SII in GC and their relationship with prognosis. In the future, this research project will be expanded based on the points mentioned above.

CONCLUSION

In summary, NLR, PLR, and SII are all anomalously elevated in GC and have varying degrees of diagnostic implications. As a single indicator, SII demonstrates the best overall performance for GC screening and may even be comparable to that of the combined three indicators. Although NLR, PLR, and SII are not confirmed to be risk factors for the occurrence of GC, we have discovered that gender and BMI may, to a certain extent, influence the occurrence of GC. We have identified convenient, low-cost, and efficient biomarkers for the screening of GC and provided more evidence and references for the prevention and management of GC.