Introduction

Cervical cancer is the fourth most common type of cancer affecting women worldwide. It poses a serious threat to women’s health1. In 2020, there were approximately 604,000 new cases of cervical cancer globally. In China alone, there were nearly 110,000 new cases and 47,700 deaths from this disease. This accounted for approximately 18.3% of all global cases and 17.6% of all global deaths related to cervical cancer2. Despite the advancement of preventive HPV vaccines and effective screening and early detection methods, cervical cancer continues to be a significant burden worldwide. In cases of locally advanced cervical cancer, the standard treatment is platinum-based concurrent chemoradiotherapy, which has a 5-year survival rate ranging from 50–80%3,4. However, for advanced and metastatic cervical cancer, the 5-year survival rate drops dramatically to just 17%5.

In recent years, there have been significant breakthroughs in the treatment of cervical cancer with anti-PD-1 therapy. Studies such as KEYNOTE-1586 and CHECKMATE-3587 have demonstrated the effectiveness of pembrolizumab and nivolumab, respectively, as second-line treatments for cervical cancer. More recently, the results of the KEYNOTE-8268 study revealed that when used as a first-line treatment, pembrolizumab in combination with chemotherapy, with or without bevacizumab, significantly extends the overall survival (OS) of patients with advanced or metastatic cervical cancer. Additionally, the KEYNOTE-A189 study reported that, in cases of locally advanced cervical cancer, the combination of synchronous radiotherapy and chemotherapy with pembrolizumab greatly improves patient progression-free survival (HR = 0.70, 95% CI: 0.55–0.89; P = 0.0020). This represents a 30% reduction in the risk of disease progression. Based on these results, the FDA approved the combination of pembrolizumab with concurrent radiotherapy and chemotherapy as a treatment for newly diagnosed stage III-IVA cervical cancer patients.

However, it is important to note that not all patients benefit from immunotherapy. Currently, tumor tissue PD-L1 expression and tumor mutation burden (TMB) are widely accepted biomarkers for predicting the success of immunotherapy. Unfortunately, the clinical utility of these biomarkers is limited due to deviations in detection methods and expensive, uneven distribution of PD-L1 within tumors, and dynamic changes in PD-L1 expression in tumors. Additionally, the CLAP study showed that advanced cervical cancer patients can benefit from the combination of carlizumab and apatinib treatment, regardless of the PD-L1 status10. In a retrospective study of the real world, PD-L1 CPS ≥ 1 was not associated with OS of non-immunotherapy-treated patients with advanced cervical cancer11. However, only 25% of cervical cancer patients exhibit high TMB12. Hence, there is an urgent need to identify better predictive biomarkers. Inflammatory indicators are obtained through peripheral blood testing, which are easy to obtain, inexpensive and reproducible and can screen for patients who are most likely to benefit clinically.

The correlation between inflammation and tumors was first reported by Rudolf Virchow in 186313. Inflammation is a well-known characteristic of the tumor microenvironment, which can promote the occurrence, progression, and metastasis of tumors by inducing angiogenesis, cell proliferation, reactive oxygen species damaging DNA, and inhibiting cancer cell apoptosis14,15,16. Several studies have highlighted the predictive role of immune-inflammation systems, including the neutrophil-lymphocyte ratio (NLR), platelet-lymphocyte ratio (PLR), lymphocyte-to-monocyte ratio (LMR), and systemic immune-inflammation index (SII), in various malignant tumors17,18,19. Prognostic Nutritional Index (PNI) is calculated based on serum albumin and circulating peripheral blood lymphocyte count and has been used to assess the immune nutritional status of cancer patients. PNI has also been proven to be an effective prognostic biomarker for various cancers, including cervical cancer20. However, the predictive role of these immune-inflammation systems and nutritional status in cervical cancer patients undergoing immunotherapy remains unclear. Therefore, we conducted this study to analyze the relationship between NLR, PLR, LMR, SII, PNI, and the short-term outcomes of patients with locally advanced or recurrent/metastatic cervical cancer who underwent anti-PD-1 treatment.

Materials and methods

Patients

This study included recurrent/metastatic cervical cancer patients and locally advanced cervical cancer patients in The Affiliated Suzhou Hospital of Nanjing Medical University and Yancheng City No.1 People’s Hospital from February 2019 to February 2023.The patients were included as follows (a) ≥ 18 years old, (b) available clinicopathological and pretreatment laboratory data, (c) Eastern Cooperative Oncology Group performance status (ECOG-PS) of 0 − 1, and at least one measurable lesion at baseline according to Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1. (d) patients who underwent at least two cycles of immunotherapy, immunotherapy used until the disease progresses or maintained for one year. (e) received anti-PD-1 therapy with or without other therapy, anti-PD-1 therapy contained Nivolumab, Pembrolizumab, Sintilimab, Camrelizumab, Tislelizumab or Toripalimab, chemotherapy regimens contained cisplatin, carboplatin, paclitaxel and albumin bound paclitaxel. The exclusion criterion was as follows: patients with fever, systemic infammation, blood disease, immune disease, cardiovascular or cerebrovascular events, or infection; withdrew immunotherapy because of intolerable toxicities, lacked follow-up data or received drugs that improved blood cell function within 2 weeks from the first dose of immunotherapy. According to these inclusion and exclusion criteria, 105 patients were finally enrolled in the study. The basic information of these patients was collected, including age, clinical stage, pathological classification, treatment regimen and hematological parameters. (Table 1). All procedures performed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1964 and later versions. This study was approved by the Institutional Review Board of The Affiliated Suzhou Hospital of Nanjing Medical University and Yancheng City No.1 People’s Hospital. Informed consent was obtained from all individual participants included in the study or their family.

Laboratory testing

The definition of each ratio is as follows: The NLR refers to the absolute neutrophil count divided by the lymphocyte count measured in peripheral blood, whereas the platelet-to-lymphocyte ratio (PLR) refers to the platelet count (cell/µL) divided by the lymphocyte count (cell/µL). The lymphocyte-to-monocyte ratio (LMR) refers to the lymphocyte count (cell/µL)divided by the monocyte count (cell/µL). PNI was calculated based on a peripheral blood sample and using the formula: serum albumin (g/L) + 5×peripheral blood lymphocyte count (×109 /L).SII was derived from platelet count (cell/µL)× neutrophil count (cell/µL)/lymphocyte count (cell/µL).

Follow up

PFS is defined as the time from PD-1 inhibitor treatment until disease progression or death or last follow-up. Patients were mainly followed up through medical record searches or telephone communications. The cut-off date was August 31, 2023.

Statistical analyses

All statistical tests were conducted using SPSS 26.0 (IBM Corp., Armonk, NY, USA).For descriptive analysis, continuous and categorical variables were expressed as medians (range) and percentages, respectively. Receiver operating characteristic (ROC) curve analysis was performed to identify the cut-off values of continuous variables. The Kaplan–Meier method was applied for survival analysis. A Cox proportional hazards regression model was used for univariate and multivariate analyses. Using statistically significant factors in univariate analysis for multivariate analysis. A forward selection stepwise procedure was applied for multivariate analysis. A two-sided p value of < 0.05 was considered statistically significant.

Results

Clinical characteristics of patients

A total of 105 patients with locally advanced, recurrent/metastatic cervical cancer treated with immunotherapy were enrolled in this study. The clinical characteristics of the patients, including age, BMI, albumin, pathological type, therapy regimen, and FIGO stage, were obtained from the medical records in Table 1. The median age was 57 years (ranging from 31 to 81 years). The median follow-up duration was 20 months (ranging from 2 to 44 months). Cervical squamous cell carcinoma accounted for most cases (95.2%), while cervical adenocarcinoma was observed in 4.8% of patient. 41% patients were locally advanced cervical cancer, 59% patients were recurrent or metastatic cervical cancer. All patients received combined treatment, including either 76.2% with chemoradiotherapy, 9.5% with radiotherapy, or 14.3% with chemotherapy. The median PFS was 19.0  months. The optimal cutoff values for NLR, PLR, LMR, SII, and PNI were determined as 3.76, 218.1, 3.34, 1147.7, 43.75, respectively. Other baseline characteristics are shown in Table 1.

Table 1 General characteristics of patients with cervical cancer treated with combination immunotherapy.
Full size table

The Optimal Cut-Off Values for NLR, PLR, LMR, SII and PNI

As revealed in Fig. 1, the areas under the ROC curve for NLR, PLR, LMR, SII, and PNI were 0.663 (95% CI: 0.553–0.773), 0.675 (95% CI: 0.565–0.785), 0.549 (95% CI: 0.434–0.665), 0.702 (95% CI: 0.576–0.794), and 0.640 (95% CI: 0.530–0.750), respectively. The optimal cut-off values of NLR, PLR, LMR, SII, and PNI that predicted survival results were 3.76 (with a sensitivity of 0.615 and specificity of 0.697), 218.1(with a sensitivity of 0.692 and specificity of 0.682), 3.34(with a sensitivity of 0.379 and specificity of 0.744), 1147.7(with a sensitivity of 0.513 and specificity of 0.879), and 43.75(with a sensitivity of 0.712 and specificity of 0.564). According to the optimal cut-off values, patients were divided into high and low groups.

Fig. 1
figure 1

Receiver operating characteristic (ROC) analysis for progression-free survival prediction of neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), lymphocyte-to-monocyte ratio (LMR), systemic immune-inflammation index (SII), and prognostic nutritional index (PNI).

Full size image

Survival analysis

The Kaplan–Meier method was used to generate survival curves and the log rank test to compare the differences. Compared with the SII High group, patients before treatment in the SII Low group had longer PFS (P < 0.001, Fig. 2A). Compared with the NLR High group, patients before treatment in the NLR Low group had longer PFS (P < 0.001, Fig. 2B). Compared with the PLR High group, patients before treatment in the PLR Low group had longer PFS (P < 0.001, Fig. 2C). Compared with the PNI Low group, patients before treatment in the PNI High group had longer PFS (P = 0.002, Fig. 2D).

Fig. 2
figure 2

Kaplan-Meier curve of association between inflammation markers and PFS. (A) SII, (B) NLR, (C) PLR, (D) PNI. SII, systemic immune-inflammation index; NLR, neutrophil-to-lymphocyte ratio; PLR, platelet-to-lymphocyte ratio; PNI, prognostic nutritional index.

Full size image

Univariate and Multivariate Analyses

The Cox proportional hazards regression model was applied to perform univariate and multivariate analyses of PFS. The results of univariate and multivariate analyses are presented in Table 2. In univariate analysis, we indicated age, pathological type, FIGO Stage, treatment and LMR was not related with PFS. Low SII(HR = 4.671, 95% CI = 2.457–8.879, P<0.001), Low NLR (HR = 3.139, 95% CI = 1.632–6.037, P = 0.001), Low PLR (HR = 3.804, 95% CI = 1.911–7.572, P<0.001) and High PNI (HR = 0.379, 95% CI = 0.200-0.719, P = 0.003) were related to longer PFS. However, the results of multivariate analyses displayed that NLR, PLR, PNI were not significantly risk factors for PFS, only Low SII (HR = 3.539, 95% CI = 1.256–9.976, P = 0.017) was independently associated with longer PFS.

Table 2 Univariate and multivariate analyses of progression-free survival in patients with cervical cancer treated with combination immunotherapy.
Full size table

Discussion

The majority of cervical cancer cases are caused by a persistent high-risk HPV infection21. Such persistent infections can integrate into the host genome, leading to the overexpression of the oncoproteins E6 and E7, thereby interfering with natural immune responses by downregulating key pathways22. Research has demonstrated that a persistent HPV infection upregulates the expression levels of PD-1 and PD-L1 in cervical cancer cells as well as in infiltrating immune cell23. One study found that 85% of patients with cervical cancer exhibit positive expression of PD-L1 in tumor tissues12. These findings underscore the potential therapeutic effect of immune checkpoint inhibitors (ICIs) for HPV-infected cervical cancer. Consequently, exploring biomarkers to predict the efficacy of ICIs is essential.

Clinical trials have demonstrated that PD-L1 expression, one of the most common predictive markers, cannot perfectly predict the efficacy of immunotherapy. Patients with low or negative PD-L1 expression can still benefit from monotherapy or combination immunotherapy. KEYNOTE-189 and KEYNOTE-407 have shown that, for PD-L1-negative non-small cell lung cancer, the combination of pembrolizumab and chemotherapy can improve PFS and OS compared to chemotherapy alone24,25. EMPOWER Cervical 1 compared the effects of cemiplimab monotherapy versus chemotherapy in advanced cervical cancer. In the overall population, the OS of the cemiplimab group was significantly longer than that of the chemotherapy group (12.0 months vs. 8.5 months) and the treatment reduced the risk of death by 31%. In the PD-L1-negative group, the ORR of the cemiplimab group was 11.4% (95% CI, 4-25%), while the chemotherapy group was only 8.3% (95% CI, 2.3-20.0%)26. The CheckMate 358 study evaluated the efficacy and safety of nivolumab ± ipilimumab in the treatment of recurrent and metastatic cervical cancer. The median follow-up time was 30.4 months, and the data indicated that regardless of the tumor PD-L1 status, lesion remission was observed in the nivolumab monotherapy, N3 + I1(nivolumab 3 mg/kg + ipilimumab 1 mg/kg), and N1 + I3(nivolumab 1 mg/kg + ipilimumab 3 mg/kg) groups27. The following reasons may account for this: firstly, due to the inherent bias of the detection methods, there is currently a lack of unified standards for detecting PD-L1 expression28. Secondly, the expression of PD-L1 in tumor tissues depends on their biological characteristics. PD-L1 expression is not uniformly distributed within the tumor; within the same tumor tissue, some parts may express positively while others may express negatively29. Thirdly, the expression of PD-L1 in tumor tissue is not sufficiently stable and is influenced by many molecular signals, which can undergo dynamic changes30. Therefore, the results of sampling at a specific time may not represent the overall PD-L1 expression level of the tumor tissue.

Previous studies have explored that inflammation is crucial in the occurrence and development of malignant tumors, and it also impacts the tumor immune microenvironment and treatment response31,32. Mingxia Cheng et al. reported an adverse association between NLR and prognosis in cervical cancer treated with combination immunotherapy33. Miaomiao Gou et al. have suggested that pre-treatment PLR correlated significantly with PFS and OS in gastric cancer patients who received immunotherapy34. Jingjing et al. found that pretreatment SII, NLR, and PLR are significant prognostic predictors of PFS and OS in advanced NSCLC patients receiving nivolumab35. In our study, univariable analysis indicated that pretreatment SII, NLR, and PLR were related to PFS in cervical cancer patients, but multivariate Cox analysis indicated that only SII is an independent prognostic factor for patients with cervical cancer(Table 2). Meanwhile, the ROC curve results showed that SII was more effective and accurate in predicting patient prognosis among these biomarkers (Fig. 1).The reason may be that compared to NLR and PLR, SII is composed of three types of blood cells, which is more comprehensive and can better reflect the body’s inflammation and immunity status. Therefore, SII may have more important clinical significance. In this study, univariate analysis showed a correlation between PNI and PFS, but multivariate analysis showed no correlation(Table 2). This may be due to that all patients included in this study received immunotherapy, which is different from the treatment methods used in previous studies. Furthermore, relevant clinical factors such as age and pathology show no significant correlation with prognosis. The reason may be that the majority of the study participants were older, and the pathological type was predominantly squamous cell carcinoma, which may introduce bias.

SII, which consists of neutrophil, platelet and lymphocyte, is a prognostic indicator for various malignant tumors36,37. This composite marker uses neutrophils and lymphocytes counting and platelets to quantify systemic inflammation and reflect the balance between host inflammation and immune status. Neutrophils can change tumor microenvironment through both external and internal pathways and then promote tumor cell proliferation and distant metastasis38. In addition, neutrophils are also one of the targets of tumor immunotherapy and participate in the ICI resistance mechanism39,40. Platelets can directly interact with tumor cells, activating tumors cytokine TGF- β and NF- β signal pathways that promote epithelial mesenchymal transition in tumors and tumor metastasis41. Besides, platelets can bind to tumor cells, protecting them from immune system attacks and then affecting the effectiveness of immunotherapy42. Several studies showed that the increase of tumor infiltrating lymphocytes (TILs) is associated with better efficacy and prognosis of immunotherapy in solid tumor patients43. Low lymphocyte counts can reduce the immune system, leading tumors occurrence and development44 and weakening the effectiveness of immune checkpoint inhibitors (ICIs). Notably, 89% of cervical cancer patients will experience lymphocyte depletion after radiotherapy45. Therefore, both high NLR and high PLR were associated with worse prognosis. Elevated NLR and PLR together may imply a more intense inflammatory response and an underlying hypercoagulable state of the blood. In the tumor microenvironment, cytokines released by neutrophils activate platelets, which in turn promote neutrophil aggregation and activation, together promoting tumor cell growth. And lower SII, which is a combination of NLR and PLR, may benefit from cervical cancer with immunotherapy. Jiahong Yi studied that higher SII predicted worse outcomes in MSI-H mCRC patients undergoing immunotherapy46. De Giorgi et al. found that SII is one of the key prognostic factors for OS in patients receiving nivolumab treatment. Lower ORR and DCR are associated with higher baseline SII values, and SII ≥ 1375 can independently predict OS47. Nevertheless, the predictive role of SII in cervical cancer treated with combination immunotherapy has not been investigated. The results of our study indicated that pretreatment higher SII ≥ 1147.7 was significantly associated with worse PFS, which was independently predictive of inferior PFS in patients with cervical cancer after combination immunotherapy. At present, in the relevant research on inflammation indicators, most of the studies on cutoff values use the ROC curve method, which can show the sensitivity and specificity of the cutoff value of the results, and thus is more scientific and accurate, and has a better clinical utility value. Besides, the cut-off value for SII was similar to the value reported in the aforementioned studies.

Notably, single indicators of inflammation may have limitations in predicting prognosis. The multifactorial prediction model can combine several statistically significant predictive factors, making the prediction more accurate and meaningful. One study combined NLR, MLR, PNI, and albumin-alkaline phosphatase ratio (AAPR), to construct the Inflammation and Nutritional Prognostic Score (INPS), and the results showed that the INPS is an independent prognostic indicator for breast cancer48. Yang et al. found that LMR-NLR scoring system predicts prognosis in gliomas49. Therefore, constructing a multifactorial model based on multiple predictors may be a hot spot for future research on predictive markers of tumor immunotherapy efficacy and prognosis.

However, this study had some limitations. First, it is a retrospective study with a small number of patients, so a prospective study with a larger sample size should be needed to further validate the current conclusions. Second, this study lacks the detection of PD-L1 and it needs to be addressed by further research. Third, the optimal cut-off values of these indexes remain unknown. Finally, due to limitations in conditions, no overall survival data was obtained. We will conduct long-term follow-up to obtain the overall survival time.

Conclusion

In conclusion, our study demonstrated that SII was an independent prognostic factor for cervical cancer patients undergoing immunotherapy. However, further large and multi-center prospective studies should be performed to confirm these findings.