PD-L1 levels, TP53 mutation profiles, and survival outcomes in pancreatic cancer differ by immune-nutritional status

Abstract

Background

Pancreatic ductal adenocarcinoma (PDAC) frequently exhibits an immunosuppressive microenvironment coupled with malnutrition status. These features are instrumental in clinical management strategies for PDAC.

Methods

Immune-nutrition status of patients was evaluated by integrating systemic immune-inflammatory index (SII) and prognostic nutritional index (PNI). Individuals were divided into SII-PNI Status positive (SPS⁺) group and SPS negative (SPS⁻) group. Morphology of tissues was evaluated by hematoxylin-eosin (H&E) staining. Expression of PD-L1 and p53 was detected using immunohistochemistry (IHC).

Results

In this study, 530 eligible patients (mean ± SD age, 60.5 ± 9.17 years, 296 males [55.8%], 74 SPS⁺ [14.0%]) were included. These patients exhibited a median survival of 24 months (1-, 3- and 5-year survival rate; 72.9%, 34.7% and 25.1%, respectively). In the multivariate analysis, independent indicators for outcomes were identified as tumor size, lymph node metastasis and SPS (all p <.01). After matching and adjusting, patients with SPS⁺ exhibited a notably reduced overall survival compared to those with SPS⁻ (14 vs. 25 months, p <.001), with hazard ratio (95% CI) of 1.79 (1.25–2.56). IHC revealed markedly elevated positive cell proportion of PD-L1 in SPS⁺ group (p <.01) and distinct p53 mutation patterns between SPS⁺ and SPS⁻ groups (p =.03). Morphology demonstrated a dissimilar trend of differentiation levels between the two groups (p =.08).

Conclusion

The findings suggest poorer outcome, higher PD-L1 expression and distinct p53 mutation status of patients with SPS⁺. These patterns may contribute to PDAC management and strategic deployment of immunotherapy and targeted therapy.

Introduction

Pancreatic ductal adenocarcinoma (PDAC) stands as a formidable malignancy, with approximately 467,000 new deaths estimated in 2022. The incidence of PDAC is on an alarming upward trajectory, with forecasts predicting it will ascend to the second most frequent cause of cancer-related mortality by the year 2030 [1]. The rising incidence of early-onset PDAC also emphasizes the urgent need for effective prevention strategies to curb its impact on public health [2]. Treatment strategies for PDAC include surgical resection, followed by adjuvant chemotherapy, radiotherapy, immunotherapy, targeted therapies, and so on. Yet, these approaches are often constrained by the intricate nature of the PDAC’s immune-inflammatory profile and the nutritional status of the patients [3].

The tumor microenvironment (TME) in PDAC is notably unique, featuring an abundance of desmoplastic stroma that not only provides physical barriers to drug delivery but also actively participates in tumor progression and immune evasion [4]. This stroma, rich in cancer-associated fibroblasts (CAFs) and immune-suppressive cells [5, 6], forms an intricate network with tumor cells, contributing to the disease’s resistance to the treatment. Additionally, pancreatic cancer can disrupt both the exocrine and endocrine functions of the pancreas, leading to a cascade of gastrointestinal symptoms [7]. Approximately 34–71% of individuals with PDAC suffer from malnutrition, which is associated with increased toxicity from chemotherapy, higher postoperative risks, reduced survival, and lower quality of life [8]. Consequently, identifying of immune-nutritional indicators can aid in risk stratification and clinical decision making of PDAC.

The systemic immunoinflammatory index (SII) is an innovative marker of immune-inflammation, derived from the levels of peripheral blood neutrophils, platelets, and lymphocytes. It reflects a variety of in vivo inflammatory and immune responses and exhibits enhanced stability [9]. A multitude of studies have demonstrated that SII is a predictive indicator for the prognosis of malignant tumors such as colorectal cancer [10], hepatocellular carcinoma [11], lung cancer [12], and so on. Furthermore, the expression of programmed death receptor 1 ligand (PD-L1) on tumour cells and inflammatory tumour microenvironment has demonstrated substantial prognostic significance on partial patients with cancer, indicating the potential crosstalk between SII and PD-L1 [13]. The prognostic nutritional index (PNI) is a straightforward and practical nutritional assessment approach. It measures a person’s nutrition level and is widely used to predict cancer outcomes and assess the risk of surgery complications [14]. Research has found that the expression of p53 in adipocytes can affect metabolic activity and insulin sensitivity by regulating inflammation and immune related pathways, demonstrating the interaction between p53 expression and immune nutritional status of patients [15]. A prior prospective study demonstrated that the pre-treatment SII-PNI status (SPS) serves as a crucial predictor for assessing the chemosensitivity of locally advanced gastric cancer patients following sintilimab immunotherapy in combination with XELOX chemotherapy. SPS can aid in identifying high-risk cohorts and forecasting patient outcomes [16]. A previous study also revealed a notable association between SII and PNI, which were utilized as indicators to assess the severity of COVID-19 [17]. To the best of our knowledge, there is no study reported to investigate the clinicopathologic patterns of PDAC patients stratified by SPS.

In this study, we displayed distinct clinicopathologic features of PDAC patients within different SPS groups, aiming to explore whether SII and PNI reflect immune dysregulation linked to PD-L1 expression and/or p53-driven tumorigenesis, thereby providing novel insights into therapeutic stratification for PDAC patients.

Methods

Study design and patients

In this retrospective study, patients were included who underwent curative surgery for PDAC at the National Cancer Center of China, Chinese Academy of Medical Sciences, Cancer Hospital, from January 1, 2006, to December 31, 2023. To validate the findings, we retrospectively reviewed the PDAC patients at Beijing Hospital from 2013 to 2020. The Ethics Committee of the National Cancer Center of China (NCC) and Beijing Hospital granted approval for the study. Informed consent was secured from participants, authorizing the utilization for potential research. The median follow-up time of the NCC cohort was 61 months (ranges: 1 ~ 200 months). The study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines.

Pathologically confirmed PDAC patients without distant metastasis treated with curative-intent resection were included in the study. According to the exclusion criteria, patients without hematological data, as well as those with missing clinical pathological or follow-up information, were excluded. Variables including patient sex, age, symptom, body mass index (BMI), American society of anesthesiologists (ASA) physical status classification, pathological characteristics, and treatment modalities were collected from the medical records. Lymphovascular invasion (LVI) was defined by the presence of neoplastic cells within lymphatic spaces and muscular vessels. Pathologic staging (TNM) was classified based on the American Joint Committee on Cancer Staging Manual (8th edition).

Blood parameters

The immune-nutritional status was reflected by pre-operative systemic immune-inflammatory index (SII, platelets*neutrophils/lymphocytes) and prognostic nutritional index (PNI, albumin + 5*lymphocytes). By using X-tile software (v3.6.1), optimal SII and PNI cut offs were determined as 835.33 and 50.6, respectively. The method was similar with the one reported in a previous study [18]. The SII-PNI status (SPS) was subsequently categorized into two groups: SPS + and SPS-. The SPS + group, which accounted for 14% of the cases, was defined by an SII value of ≥ 835.33 and a PNI value of < 50.6. The SPS- group included all other cases.

Pathological analysis

Paraffin-embedded samples from PDAC patients diagnosed after January 1, 2022 were cut into serial slices with a thickness of 4 µm. Immunohistochemistry (IHC) staining was performed to evaluate the expression of PD-L1 (Dako22C3) and p53. The standard procedures include tissue dewaxing, rehydration, antigen retrieval, and peroxidase blockade. Slices were blocked and incubated with primary antibodies at 4°C overnight. After washing, they were treated with peroxidase-tagged secondary antibodies. The IHC staining was completed with 3,3’-diaminobenzidine solution (DAB) color development and hematoxylin counterstaining.

The proportion of positive cells on IHC staining was quantitatively assessed using QuPath software (v0.5.1) on digitized whole slide images (WSI). We detected 10 random fields for each sample and calculated the average value of positive cell proportion. Then, the differentiation levels of PDAC were identified based on the morphology on H&E staining slides.

Statistical analysis

A 1:2 ratio propensity score matching (PSM) was conducted based on clinicopathologic factors including gender, age, tumor size, lymph node metastasis, differentiation, surgical margin, invasion, and SPS. The primary outcome of the study, overall survival (OS), was defined as the time elapsed from the date of surgery to either the date of death or the last contact.

Spearman test was performed to identify the correlation among SII, PNI, and OS. Descriptive statistics were provided as frequencies with percentages for categorical variables and mean/median for continuous variables. Pearson χ² test was used to test the distribution of categorical variables. Estimates of survival rates were calculated using Kaplan-Meier method, and the distributions of survival were compared between groups by log-rank test. Univariate and multivariate Cox proportional hazard models were applied to identify the prognostic factors of OS (HR, hazard ratio; 95%CI, confidence interval). Pathologic differences were evaluated using unpaired nonparametric Mann–Whitney test or ANOVA, appropriately. Significance levels were defined as two-sided p <.05. All statistical analyses were conducted using R software (v4.4.1), GraphPad Prism (v8.0.1) and IBM SPSS software (v25.0). Data analysis was conducted from May, 2024 to August, 2024.

Results

Baseline characteristics

In this single-center study, a total of 704 patients with PDAC undergoing curative resection were screened in the studied period. After exclusion of patients without essential data (n = 174), 530 patients (mean ± SD age, 60.5 ± 9.17 years, 296 males [55.8%], 74 SPS⁺ [14.0%]) met the inclusion criteria (Fig. 1). A majority of patients were staged II (270 [50.9%]), with moderate differentiation (385 [72.6%]), with R0 resection (512 [96.6%]), with invasion of nerve (376 [70.9%]) and capsule (409 [77.2%]), without invasion of bile duct (365 [68.9%]), duodenum (402 [75.8%]), spleen, and lymphatic vessel (376 [70.9%]). More individuals received adjuvant therapy after operation (277 [52.3%]). Several distinctions between SPS- and SPS + groups were observed and were summarized in Table 1.

Fig. 1

Flowchart of patient inclusion and exclusion

Table 1 Baseline characteristics of unmatched 530 PDAC patients

After 1:2 ratio PSM adjusting based on the differences among groups, 203 patients were screened out (mean ± SD age, 60.5 ± 8.86 years, 106 males [52.2%], mean ± SD BMI, 23.2 ± 3.13, 71 SPS⁺ [35.0%]). The baseline characteristics in SPS- group and SPS + group were balanced and comparable, as shown in Table 2.

Table 2 Baseline characteristics of 203 PDAC patients after 1:2 PSM

The metrics of surgically treated PDAC patients were overviewed and collected from 2006 to 2023. Then, a report from National Cancer Center of China was presented. The median survival of all surgically resected PDAC patients was 24 months, with 1-, 3-, and 5-year survival rate of 72.9%, 34.7%, and 25.1%, respectively. All patients demonstrated an average hospital stay of 18 days. The average operation time for Whipple is 320 min. The average blood loss during surgery was 419 milliliters, with an average of 360 milliliters of blood transfused. An overview of Kaplan-Meier survival curve was displayed to show the survival pattern of PDAC patients in National Cancer Center of China (Fig. 2).

Fig. 2

Overview of survival metrics in the National Cancer Center of China

Distributive analysis of SII and PNI

Aiming to determine the correlation between immune-nutrition parameters and survival outcomes, Spearman analysis was conducted. SII showed a significantly reverse relation with OS (R = −.09, p =.038), while there was no significant relationship between PNI and OS (R =.01, p =.081). Notably, SII was negatively correlated with PNI (R = −.228, p <.001) (Fig. 3A). X-tile analysis showed that the optimal cutoff values for SII and PNI were 835.33 and 50.6, respectively. Then the SII-PNI status (SPS) was divided into SPS⁺ and SPS⁻ groups. SPS⁺, accounting for 14% cases, was defined as SII ≥ 835.33 and PNI < 50.6, and SPS⁻ encompassing all other cases (Fig. 3B). The distributive landscapes of SII and PNI of each sample were displayed in Fig. 3C, which can be utilized to quantitatively evaluate the immune-nutritional status.

Fig. 3

Distributive analysis of immune-nutrition status index. A, the association between overall survival (OS) and SII, as well as OS and PNI, and relationship between SII and PNI. B, the categorization of SII-PNI Status (SPS) C, distributive landscape of SII and PNI of each sample

Survival analysis

In the primary 530-patient cohort, multivariate Cox proportional hazard model identified the independent prognostic indicators including tumor size (HR = 1.11, 95%CI: 1.05, 1.18, p <.01), lymph node metastasis (HR = 1.32, 95%CI: 1.15, 1.45, p <.01), well differentiation (HR = 0.65, 95%CI: 0.43, 0.98, p =.04), R1 resection (HR = 1.74, 95%CI: 1.06, 2.86, p =.03), capsule invasion (HR = 1.32, 95%CI: 1.01, 1.74, p =.04), and SPS (HR = 1.98, 95%CI: 1.48, 2.64, p <.01) (Table 3). The K-M survival analysis demonstrated significant shorter OS for patients with SPS⁺ compared to SPS⁻ (p <.001) (Fig. 4A). The survival distinctions still existed after adjusting by influence of prognostic factors (p <.001) (Fig. 4B).

Table 3 Univariate and multivariate Cox regression models before matching

Fig. 4

Overall survival stratified by SPS before and after PSM. A, unadjusted Kaplan-meier analysis before PSM. B, Kaplan-meier analysis before PSM adjusted by prognostic factors. C, unadjusted Kaplan-meier analysis after PSM. D, Kaplan-meier analysis after PSM adjusted by prognostic factors

After 1:2 ratio PSM, multivariate Cox model revealed the prognostic indicators including tumor size (HR = 1.21, 95%CI: 1.06, 1.38, p <.01), lymph node metastasis (HR = 1.4, 95%CI: 1.14, 1.58, p <.01), and SPS (HR = 1.79, 95%CI: 1.25, 2.56, p <.01) (Table 4). The K-M survival analysis revealed markedly reduced OS for patients with SPS⁺ compared to those with SPS⁻ (p <.001), as depicted in Fig. 4C. Even after accounting for the impact of prognostic factors, these survival differences persisted (p =.001), as shown in Fig. 4D.

Table 4 Univariate and multivariate Cox regression models after 1:2 PSM

In the validation cohort, 96 PDAC patients at Beijing Hospital were screened. After exclusion, 58 eligible patients were enrolled. By using the same classification method mentioned above, 20 (34.5%) and 38 (65.5%) patients were grouped into SPS⁺ and SPS⁻, respectively. The K-M survival analysis also demonstrated significant shorter OS for patients with SPS⁺ compared to SPS⁻ (p =.0026) (Supplementary Fig. S1).

Pathological staining analysis

Paraffin-embedded tissues of 52 patients diagnosed from 2022 to 2023 were screened. Among the patients, 12 (23.1%) were identified as SPS⁺, while 40 (76.9%) were SPS⁻. Hematoxylin-eosin (H&E) and immunohistochemistry (IHC) staining procedures were conducted to pathologically assess the differentiation and expression levels of p53 and PD-L1.

A typical section was selected from each SPS group for presentation and analysis. In the SPS⁺ group, the PDAC section exhibited poor differentiation, featured a nonsense mutation in the p53, and displayed a high expression of PD-L1, with a combined positive score (CPS) reaching 70. In contrast, the sample from the SPS⁻ group demonstrated moderate differentiation, harbored a missense mutation in p53, and showed low expression levels of PD-L1, with a CPS of 5, as shown in Fig. 5A.

Fig. 5

Pathological assessment of PDAC patients within distinct SPS groups. A, H&E staining, p53 status, and PD-L1 expression of PDAC in the two SPS groups. B, Quantitative analysis of differentiation, p53 mutation, and PD-L1 expression Well, well differentiation; Mod, moderate differentiation; Poor, poor differentiation; Mis-m, missense mutation; Wild, wild type; Non-m, nonsense mutation

Further analysis revealed distinct proportions of differentiation within the SPS⁺ and SPS⁻ groups (p =.08). Specifically, the SPS⁺ group showed 83% moderate differentiation and 17% poor differentiation. Conversely, the SPS⁻ group exhibited a higher proportion of well- and moderately differentiated cases. In terms of p53, the mutation patterns were markedly distinct among the two groups (p =.03). Patients within SPS⁺ group exhibited a higher likelihood of being with mutational type of p53 expression (67% vs. 55%). Based on Qupath software analysis, the proportion of PD-L1 positivity in IHC was significantly higher in the SPS⁺ group (23.5% vs. 9.5%, p <.01), as depicted in Fig. 5B.

Discussions

Pancreatic ductal adenocarcinoma (PDAC) is a highly lethal form of cancer characterized by a distinctive immune-suppressive tumor microenvironment (TME) and a propensity for malnutrition. The study employed the systemic immune-inflammation index (SII) and prognostic nutritional index (PNI) to assess the SII-PNI status (SPS), thereby providing a reflection of the patients’ immune-nutritional profile. We synthesized clinical features and pathological information, uncovering that (i) PDAC patients with distinct immune nutrition status could be practically stratified by SPS categorization approach; (ii) patients within different SPS groups exhibited varying clinical characteristics; (iii) patients with SPS⁺ featured a significant shorter overall survival (OS) compared to SPS⁻; (iv) SPS⁺ was associated with higher expression of PD-L1 and higher proportion of mutational type p53 expression compared to SPS⁻.

According to previous studies, the resistance of PDAC to various treatments such as immunotherapy, chemotherapy, and targeted therapies is largely due to three key factors: (i) the dense fibrotic stroma made up of cancer-associated fibroblasts (CAFs); (ii) the TME with myeloid-derived suppressor cells (MDSCs), M2-like macrophages that have been alternatively activated, and regulatory T cells (Tregs). These cell types release numerous cytokines that suppress the immune response, thereby reducing the effectiveness of cytotoxic CD8 + T cells and natural killer (NK) cells that could potentially combat the tumor effectively [19, 20]; and (iii) nutritional status and strategic interventions, which play a pivotal role in addressing the cachexia, malnutrition, and weight loss experienced by PDAC patients [21]. Starting from a clinical perspective, we approached the matter with a focus on real-world patient care and medical practice. In this study, individuals categorized as SPS⁺ exhibit a heightened immune-inflammatory response, yet they tend to have a more compromised nutritional status when contrasted with the SPS⁻. This can also be partially explained by the fact that, in clinical settings, PDAC with SPS⁺ tended to exhibit symptoms and was commonly located in the pancreatic head.

After performing 1:2 Propensity Score Matching (PSM) based on prognostic factors, the clinical features are balanced and comparable among SPS⁺ and SPS⁻ groups. Survival analysis demonstrated a markedly reduced OS in SPS⁺ PDAC patients compared to SPS⁻, a finding that corroborates prior research across various cancer types, including gastric cancer [9, 16], colorectal cancer [10], hepatocellular carcinoma [11], esophageal cancer [14], and medulloblastoma [22]. The identification of these diverse states through the SPS aid in a more nuanced understanding of a PDAC patient’s overall health and immune-nutritional function, suggesting a potential relationship between SPS and immunotherapy and targeted therapies. Zhang et al. revealed that baseline SII served as an independent prognostic marker for patients with advanced pancreatic cancer [23]. However, the study did not account for nutritional status and failed to elucidate the pathological patterns necessary for a more comprehensive interpretation. Integrating the findings regarding the immune status of PDAC and the limitations of previous studies, we further explored the association between SPS and the potential responses to immunotherapy and targeted therapies.

Experimental evidence from preclinical research indicated that blocking the PD-L1 function can suppress the growth of pancreatic cancer within experimental animal models, thereby highlighting the potential of the PD-1/PD-L1 pathway as a viable target for PDAC immunotherapeutic strategies [24]. In addition, overexpression of PD-L1 was confirmed to be associated with a poor prognosis [25]. Mutations in the tumor suppressor gene TP53, commonly known as p53, are prevalent in the majority of PDAC tumors [26]. There were numerous findings regarding the functional consequences of p53 mutations. Recently, a two-drug treatment strategy for p53-deficient PDAC has been reported, marking a targeted effort to address p53 dysfunction [27]. In the current study, we demonstrated that PD-L1 expression and proportion of mutational type of p53 were significantly higher in SPS⁺ PDAC patients relative to the SPS⁻ group. This finding may partially account for the diminished OS observed among the SPS⁺ patients, suggesting that these patients might derive greater benefits from immunotherapy and targeted treatments. This suggestion aligns with findings from a study on intrahepatic cholangiocarcinoma [28].

Strengths and limitations

The advantages of the study have been distinctly observed. We have provided a comprehensive report on the surgical data of PDAC patients who have sought treatment at the National Cancer Center of China over the past two decades, aiming to facilitate and enhance communication in the management of PDAC. Additionally, the study firstly synthesized clinical features, outcomes, and pathological data to reveal the PDAC patterns stratified by immune-nutritional status. To the best of our knowledge, apart from this study, no existing research explored the relationship between SPS and immunotherapy and targeted therapy in PDAC patients. The findings in this study can provide novel insights for PDAC treatment.

Nevertheless, there were also several limitations. Firstly, although 1:2 PSM and adjustment analysis were applied, the retrospective nature of the article implied the possibility of systematic errors such as selection bias and recall bias, which could influence the interpretation of the findings. Secondly, While the combination of the SII and PNI offers a convenient and applicable method to assess patients’ immune-nutritional status, the actual immune-nutritional condition of the patients may differ from what is indicated by these indices. In addition to clinical assessments, achieving a more precise evaluation of the status may necessitate the conduct of fundamental laboratory tests. Thirdly, the number of patients with SPS⁺ was relatively small, accounting for 14% of all patients. Furthermore, due to concerns about the quality of paraffin blocks, only samples from 2022 to 2023 were subjected to staining and analysis. Out of these 52 cases, only 12 were identified as SPS⁺. It is anticipated that future studies with larger sample sizes will provide further validation.

Conclusions

According to experience from the National Cancer Center of China, pancreatic ductal adenocarcinoma (PDAC) patients with varying immunonutrient statuses exhibited distinct clinical features, survival outcomes, and expressions of PD-L1 and p53. These differences in clinicopathologic patterns may offer novel insights for risk stratification and inform the potential employment of immunotherapy and targeted therapy for PDAC. However, larger-scale clinical trials are warranted to further validate these findings.