Comparison of PIV and Other Immune Inflammation Markers of Oncological and Survival Outcomes in Patients Undergoing Radical Cystectomy

2. Materials and Methods

2.1. Patients

The records of 314 patients who underwent radical cystectomy and lymphadenectomy at the hospital of the researcher from January 2016 to November 2022 were examined. We only evaluated patients with non-metastatic urothelial bladder cancer and did not consider other histologic variants. The data collection followed the principles outlined in the Declaration of Helsinki and obtained endorsement from the Institutional Ethical Board. Of all patients who underwent RC, we excluded those who could have an altered inflammatory/immune profile independent of disease and surgery. In particular, patients with previous radiotherapy of the pelvis (3), previous surgery (1), neoadjuvant chemotherapy (115), and a history of autoimmune disease (2) were excluded from our analysis. Finally, this retrospective study included 193 enrolled patients, all of whom underwent both open and robotic radical cystectomy (RC) along with lymphadenectomy. The selection of lymphadenectomy extent and the method of urinary reconstruction were determined taking into consideration the individual’s medical traits and the doctor’s discernment. Prior to RC, we performed routine presurgical examinations, clinical and physical tests, and imaging studies (computer tomography or magnetic resonance) to rule out metastases. The RC specimens were examined by a specialized uropathologist who staged them a3ccording to the 2017 classification (8^th edition) of the Tumor Nodes Metastasis (TNM) staging. The histological grading was also reviewed using the 2004/2016 World Health Organization system.

2.2. Data gathering

Patient features prior to surgery were age, sex, smoking condition, diabetes, body composition index or BMI, clinical staging (cT), operative technique, urinary reconstruction, tumor grading (pT), nodal status (pN), lymph vascular infiltration (LVI) and adjuvant therapy. Laboratory examinations included neutrophil count (10⁹/ l), platelet count (10 ⁹/ l), monocyte count (10⁹/ l) and lymphocyte count (10⁹/ l). PIV, SII, and NLR were the results of "neutrophil count x platelet count x monocyte count /lymphocyte count," "neutrophil count x platelet count/lymphocyte count," neutrophil count/lymphocyte count, "respectively. The PIV value was derived through the construction of a diagnostic performance curve (ROC) curve, utilizing relapse as the outcome to optimize the Youden index. Subsequently, individuals were categorized into low PIV and high PIV groups, determined by the optimal cut-off PIV value (< 340.96 and > 340.96, respectively).

2.3. Outcomes

Association between inflammatory biomarkers and unfavorable pathological characteristics, including lymph node infiltration, advanced tumor grading (pTstage), and locoregionally extended state (defined as any progressed pT stage and lymphatic infiltration) at RC pathology, were established as the primary outcomes of our study. As secondary outcomes, we analyzed survival outcomes as relapse-free survival (RFS), cancer-specific survival (CSS), and overall survival (OS). We defined RFS as the time from radical cystectomy (RC) to the first occurrence of local relapse or systemic spread. CSS and OS were established as the duration from surgery to death from neoplasm and from general mortality, respectively.

2.4. Statistical methodology

Statistical analysis was performed using STATA/SE version 18 (StataCorp, College Station, TX, USA). With the ROC curve and the Youden index, we obtained the best cut offs for each of the biomarkers analyzed. Descriptive statistics were reported as medians and interquartile ranges (IQR) for continuous variables, while frequencies and percentages for categorical variables. The assessment of patient characteristics involved employing the Mann-Whitney U test for continuous data and utilizing the chi-square test and Fisher’s exact tests for categorical data. Statistical significance was defined as a p-value below 0.05. Correlation between preoperative features (lymph node infiltration, advanced tumor grading (pT3/pT4) and locoregionally extended state) and PIV, SII and NLR were examined through univariable and multivariable logistic regression models. To gauge the predictive efficacy of each model, we determined the Area under the Curve (AUC) from the ROC analysis. At the same time, we applied the Hosmer-Lemeshow test to appraise the model's effectiveness. Relapse-free survival rate (RFS), Cancer-specific survival rate (CSS), and Overall survival rate (OS) were established by the Kaplan-Meier, and the relationship of RFS, CSS, and OS was carried out using log-rank tests. Univariable and multivariable Cox proportional hazard regression models were conducted to determine the correlation between PIV and RFS, CSS, and OS. The Harrel’s C-index or concordance index was used to evaluate the discriminatory power of models and with likelihood ratio test we compared the goodness of fit of two statistical models, where one model was a restricted version of the other (model without PIV).

3. Results

3.1. Patient features

The demographic features of the patients stratified into the two groups of PIV are outlined in Table 1. The median estimates of NLR, SII, and PIV (x10⁹) were 3.0 (quartile 2.1 - 4.3), 712.5 (quartile 460.5 - 1215.8), and 311.4 (quartile 185.5 – 644.5), respectively. The optimal cut-off value of NLR, SII, and PIV (x 10⁹) was 3.14, 640.27, and 340.96, respectively. The corresponding AUC for the three markers was 0.57, 0.63 and 0.65, respectivel (Figure 1). An elevated presurgical PIV was more frequently detected in individuals with conditions characterized by more advanced clinical (p < 0.001) and histological stages (p = 0.001), lymph node involvement (p = 0.004), lymph vascular invasion (p < 0.001), lcoregionally extended state ( p < 0.001) and disease prgression ( p = 0.004).

3.2. Oncological results

Table 2, Table 3 and Table 4 show the results of logistic regression analyses (both univariable and multivariable) concerning the prediction of oncological outcomes of three biomarkers. Elevated PIV linked with an increased likelihood of infiltration of lymph nodes (odds ratio (OR) 1.11, 95% confidence interval (95% CI): 0.34-1.89; p = 0.005), advanced Tumor grading (pT3/pT4) (OR 1.17, 95% CI: 0.58-1.76); p < 0.001), and locoregionally extended state (OR 1.32, 95% CI: 0.72-1.92; p < 0.001). We had different results about the NLR. High NLR correlated with aggressive pT stage (OR 0.63, 95% CI: 0.06-1.21; p = 0.02). High SII correlated with an increased likelihood of nodal invasion (OR 1.35, 95% CI: 0.47-2.23; p = 0.003), advanced Tumor grading (OR 0.81, 95% CI: 0.22-1.40; p = 0.007) and locoregionally extended state (OR 1.11, 95% CI: 0.52-1.71; p < 0.001).

In multivariable logistic regression models, adjustments were made for presurgical features (age, sex, smoke, and disease staging), an elevated PIV maintained an independent association with advanced tumor grading (OR 2.87, 95% CI: 1.36-6.04; p = 0.005) and locoregionally extended state (OR 3.30, 95% CI: 1.60-6.77; p = 0.001). SII also showed a statistically independent association with lymph node infiltration (OR 2.91, 95%CI: 1.09-7-72; p = 0.03), complex tumor grading (OR 2.09, 95% CI: 1.00-4.40; p = 0.05) and locoregionally extended state (OR 2.96, 95%CI: 1.44-6.08; p = 0.003). Table 3 shows no significant correlation between NLR and the three variables in the multivariable logistic regression analysis.

3.3. Long-term outcomes

Individuals with elevated PIV experienced a reduced of follow-up duration (36 vs 23 months, p < 0.001). Likewise, individuals exhibiting a high SII also exhibited a reduced median follow-up duration (24 vs. 37 months, p = 0.002). Although patients with high NLR exhibited a shorter median follow-up, the difference did not reach statistical significance (27 vs. 32 months, p = 0.14). Through the follow-up duration, recurrences occurred in 74 (38.3%) individuals, and 96 (49.7%) patients succumbed, with 52 (26.9%) of these fatalities linked to neoplasm. Figure 2 compares RFS between PIV, SII, and NLR groups. The Kaplan-Meier estimates curves revealed significantly poorer survival outcomes across all three biomarkers regarding relapse-free interval (RFS). Figure 3 illustrates a correlation regarding Cancer-specific Survival (CSS) for PIV, SII, and NLR. Individuals exhibiting elevated PIV and SII levels experienced a notably diminished Cancer-specific survival rate (CSS) compared to those with lower PIV and SII levels (p = 0.05 and p = 0.02), respectively. There wasn't a significant difference for CSS about the NLR biomarker between high NLR and low NLR. Lastly, we evaluated the overall survival among the three groups (Figure 4). As can be seen in Figure 4, individuals with an elevated PIV, SII, and NLR have a lower overall survival than those with lower values (p < 0.001, p < 0.001 and p = 0.04), respectively.

We employed a Cox regression model to explore the relationship between the PIV value and time to the event of interest. Initially, we assess the PIV’s effect considering preoperative variables, including smoke, gender, and clinical T stage. In this kind of analysis, an elevated PIV was significantly linked with RFS (hazard ratio (HR): 1.89, 95% CI:1.12-3.19; p = 0.017) and global survival (OS) (HR: 1.64, 95% CI: 1.05-2.56; p = 0.029) (Table 5.)

Subsequently, we examined how this association changes when introducing postoperative variables, such as pT stage, Lymph vascular invasion, lymph node infiltration, and adjuvant chemotherapy. Also, in this case, high PIV was intensely associated with RFS (HR 1.74, 95% CI: 1.04-2.92; p = 0.034) and OS (HR 1.58, 95% CI: 1.00-2.49; p = 0.048) (Table 6.). Despite the addition of a preoperative PIV to the reference model not yielding a significant discrimination potential of the model for CSS ( p = 0.67), it did, however, result in a considerable improvement of both RFS (by 1%, p = 0.03) and OS (by 1%, p =0.04).

4. Discussion

We analyzed the prognostic implications of preoperative PIV and compared its effects with other biomarkers in patients undergoing RC for BCa. The assessment and recognition of systemic variables that could indicate a poorer prognosis in patients with various diseases have been the subject of growing research [15]. Our study represents the first analysis in existing literature, encompassing an evaluation of PIV with other distinct systemic indices associated with an unfavorable prognosis among patients with BCa. In a previous article, it was reported that high PIV in individuals with non-metastatic MIBC was linked to worse global survival and cancer-specific survival [16]. Conforming to the findings of the exclusive article in the literature [16], an elevated PIV was related to increased aggressiveness of urothelial bladder cancer (worst clinical T stage, lymph node infiltration, lymph vascular involvement, progressive disease) and continued to be independently linked to pT3/T4 disease, locoregional advanced disease, and lymph node metastasis. Also, our analysis revealed that high PIV (> 340.96) demonstrated an independent correlation with survival results (both RFS and OS) in a model that accounted for both pre and post-surgical factors. The PIV and SII exhibited better predictive power than NLR. As far as our understanding goes, this study is the initial observation of an association within a logistic model that integrates postsurgical features. Furthermore, the study reveals the intricate interplay between inflammation and cancer, which is a pivotal aspect underpinning the complex dynamics of tumorigenesis and disease progression. Inflammatory responses mirror the systemic immune status, unveiling the intricacies of compromised immune surveillance induced by cancer cells [17]. The biomarkers analyzed in this study comprise cells that are key players in the body's immune and inflammatory processes. Neutrophils and platelets have been demonstrated to facilitate the advancement of neoplasm [18]. Neutrophils promote cellular cohesion, acting as a bridge as they bind circulating tumor cells (CTCs) to the endothelium of the organ in which they are destined (a contact-dependent mechanism) [19]. In addition, neutrophils can release soluble components that stimulate endothelial and parenchymal cells, thereby augmenting the attachment of CTCs in distant organs [20,21,22]. Platelets, through the release of dense adenine nucleotide granules in the tumor stroma, assist the migration of tumor cells across the endothelium and, in addition, support tumor cell survival in the bloodstream and induce a pro-metastatic phenotype [23]. Monocytes play a crucial role in tumor development and metastasis. Tumor-associated macrophages derived from peripheral blood mononuclear cells can suppress the acquired immune response [24]. Conversely, lymphocytes can exert an anti-tumoral function by promoting cellular cytotoxicity, impeding tumor cell replication, and bolstering the organism’s immune reaction against tumors [25,26]. It has already been shown in the literature that NLR is an unfavorable prognostic indicator in patients diagnosed with urothelial bladder cancer, and a high NLR negatively correlates with DFS [27]. Another meta-analysis by Tang et al., which included 5425 patients across 17 studies, showed that NLR had negative predictive value for OS, CSS, RFS, and PFS [28]. Our findings revealed a statistically significant negative correlation between NLR and OS but not for CSS and DFS. Moreover, it has lower predictive power among the three biomarkers used than SII and PIV for oncological and survival outcomes. Another factor examined is the Systemic Immune inflammation Index, which has become increasingly popular recently. The SII has been employed as a prognostic marker in urological cancers and various other malignancies. Jian-Hui C. et al. showed a relationship between SII and colorectal cancer [29], Wang et al. detected a correlation among elevated SII and lung neoplasm [30], and Zhong et al. discovered a strong connection between an elevated SII and a poorer prognosis in solid tumors [31]. In particular, in bladder cancer, an elevated SII was linked to worse survival and oncological outcomes [32,33]. In addition, the superiority of SII over other biomarkers, such as NLR and PLR, has already been demonstrated in the prognosis of oncological disease [34,35,36]. Our findings are consistent with the existing literature, illustrating that an elevated SII adversely impacts both RFS and OS. Finally, the last biomarker evaluated and compared with the others is the Pan-immune inflammation value, a newly defined inflammation-related index. In 2020, the investigation of PIV commenced in patients with metastatic colorectal cancer, revealing its substantial prognostic value for patient survival. Furthermore, PIV's predictive efficiency was significantly superior to previously recognized inflammation-related markers [37]. A meta-analysis conducted by Qi et al., involving eight articles and 2953 patients with breast cancer, showed that the patients with elevated PIV correlated with shorter OS and PFS [38]. In the context of bladder cancer, specifically, Kayar et al. discovered that an elevated PIV (> 406.29) was associated with diminished OS and DFS [16]. Our analysis reveals that patients with higher PIV (>340.96) had a worse RFS and OS in both Cox regression models incorporating pre and post-operative characteristics. In addition, the biomarker PIV correlated very similarly with SII with oncological and survival outcomes. This data holds the capability to improve predictive models, resulting in a more precise prognosis for patients undergoing RC. Notwithstanding, establishing a clinical utility for PIV as a biomarker goes beyond merely associating high PIV with aggressive disease or compromised survival results in multivariable models [39]. Furthermore, high PIV value may result from other immune/inflammatory pathologies such as chronic inflammation, autoimmune disease, or infectious disease. Nevertheless, further examination of PIV is justified, given its immediate and economical accessibility. The key significance of this analysis resides in its potential benefit for practitioners as a supplementary indicator to assess the prognosis of BCa. This can improve the precision of risk assessment and contribute to more precise treatment planning decisions, including the evaluation of adjuvant therapy, neoadjuvant therapy, or bladder-sparing therapies.

While our study provides valuable insights for the first time, it is essential to recognize its limitations. The primary challenges revolve around the limitations associated with retrospective data collection, small sample size, and monocentric cohort study. This analysis is strengthened by the deliberate exclusion of potential confounding factors related to PIV, the consistent use of the same laboratory for PIV evaluations, and histological examinations conducted by a consistent uropathologist. Although akin to approaches applied in previous studies, determining the optimal PIV cut-off value for BC patients still needs to be discovered. It is crucial to validate our findings externally using independent cohorts to confirm the relevance and applicability of our results.

5. Conclusions

In conclusion, our study suggests that both PIV and SII can serve as independent predictors of aggressive disease and prognosis in bladder cancer patients following radical cystectomy. Looking ahead, the incorporation of SII and PIV into clinical practice holds promise for enhancing diagnostic and prognostic strategies. The ability to identify patients with a more aggressive disease profile early on may guide tailored therapeutic interventions and improve overall clinical results. Further investigation and validation studies are justified to strengthen the applicability of these biomarkers and to explore their potential in personalized medicine and treatment optimization.