Abstract
Background/Aim: This study aimed to determine useful predictive factors for selecting patients with advanced urothelial carcinoma (UC) who might benefit clinically from treatment with pembrolizumab. Patients and Methods: We retrospectively analyzed 54 patients who underwent pembrolizumab treatment for UC. The hemoglobin, albumin, lymphocyte and platelet (HALP) score, neutrophil-to-lymphocyte ratio (NLR), and platelet-to-lymphocyte ratio (PLR) were calculated as indices of systemic inflammatory response, and the relationships between these scores and the initial tumor response or overall survival, as well as other clinicopathological factors, were assessed. Results: High NLR and PLR were associated with a poor initial tumor response to pembrolizumab. A HALP score <30.05 and a PLR ≥173.73 were associated with worse overall survival. In the multivariate Cox regression analysis, a high PLR was a significant independent prognostic factor for unfavorable outcomes. Conclusion: A high pretreatment PLR may be a valuable indicator for choosing therapy other than pembrolizumab in patients with advanced UC.
In recent years, immune checkpoint inhibitors (ICIs) have considerably changed the strategy for the treatment of various advanced cancer types, including urothelial carcinoma (UC). Remarkably, pembrolizumab was associated with significantly longer overall survival (OS) than chemotherapy as second-line therapy for platinum-refractory advanced UC in the KEYNOTE-045 trial (1). However, considering the modest rate of the objective responses to pembrolizumab (21.1% in KEYNOTE-045), identification of optimal predictive biomarkers for response would be valuable for patient selection. While the immunohistochemistry-based categorization of the expression of programmed cell death-ligand 1 (PD-L1) in tumor and infiltrating immune cells, the so-called the PD-L1 combined positive score, is still an insufficient indicator for the efficacy of ICIs, several blood test-based indices of the systemic inflammatory response (SIR) related to treatment response have also arisen as candidate indicators of ICI efficacy. Some studies on UC reported a strong association of a high pretreatment neutrophil to lymphocyte ratio (NLR) and platelet-lymphocyte ratio (PLR) with worse survival outcomes and a poor response rate, respectively, in patients treated with radical cystectomy, including those receiving neoadjuvant chemotherapy (2, 3). Furthermore, a high PLR or low hemoglobin, albumin, lymphocyte and platelet (HALP) score is associated with worse OS in patients with bladder cancer after radical cystectomy (4). Notably, some multivariate analyses showed that NLR≥5 is an independent predictor of no clinical benefit from ICI therapy in patients with advanced UC (5); NLR>3.0 is associated with worse OS in patients with metastatic UC who are treated with anti-PD-L1 atezolizumab (6) or anti-PD-1 pembrolizumab (7); and NLR≥3.5 is associated with a worse prognosis in the treatment of UC with pembrolizumab in patients with impaired performance status (8). Notably, while the prognostic value of PLR has been reported in various cancer types treated with conventional cytotoxic chemotherapy, its association with the efficacy of ICI therapy in UC remains unclear (9). Here, we investigated HALP, NLR and PLR as candidate indicators that are easily accessible by routine blood tests and evaluated their predictive ability for responsiveness to induction pembrolizumab in patients with platinum-refractory advanced UC.
Patients and Methods
General information. We retrospectively analyzed the data of 54 patients with previously treated UC who received pembrolizumab treatment (200 mg pembrolizumab intravenously every 3 weeks) at Saitama Cancer Center from January 2018 to December 2020. The patient characteristics, clinicopathological factors, and representative indices of SIR, based on data from the routine blood test prior to initiation of pembrolizumab therapy, were evaluated. OS was defined as the number of days from the first dose of pembrolizumab to the last follow-up or mortality. The efficacy of pembrolizumab was assessed according to the Response Evaluation Criteria in Solid Tumors, version 1.1 (10). The Ethics Committee of Saitama Cancer Center approved the study (protocol no. 1132 and 1254). All of the patients provided informed consent to participate in this research.
Clinicopathological parameters. We investigated the following parameters from the clinical records: Patient age, gender, performance status, surgical history, metastatic sites, and upfront systemic therapy.
Indices of SIR. The HALP score was calculated as hemoglobin (g/l) × albumin (g/l) × lymphocyte count (n/l)/platelet count (n/l). The NLR and PLR are the ratio of the absolute neutrophil to lymphocyte counts and platelet to lymphocyte counts in the blood, respectively. The optimal cut-off values for each index were determined using receiver-operating characteristic curve analysis (BellCurve for Excel; Social Survey Research Information Co., Ltd., Tokyo, Japan).
Statistical analysis. Statistical analyses were performed using BellCurve for Excel, version 3.2 (Social Survey Research Information Co., Ltd.). Continuous variables are presented as the mean±standard deviation for the entire range of values. The differences between two groups in the indices of SIR were assessed using a t-test. Two-group comparison of OS was assessed using the Kaplan–Meier method with the log-rank test. Cox proportional hazards regression models for univariate and multivariate analyses were used to calculate the hazard ratios of clinicopathological parameters. Two-sided p-values of less than 0.05 were considered statistically significant.
Results
The characteristics of patients and clinical findings. Fifty-four patients with advanced UC who had relapse after one or more regimens of systemic chemotherapy and were about to begin pembrolizumab therapy were included. Patients’ characteristics are shown in Table I. The average patient age at the initiation of pembrolizumab treatment was 70±6.8 years. Thirty-seven (68.5%) patients were male and 17 (31.5%) were female. The Karnofsky Performance status was 80 or more in 50 (92.6%) patients. Twenty-three (42.6%) patients underwent cystectomy or nephroureterectomy. Twenty-six (48.1%) patients only had lymph node metastasis at the time of pembrolizumab initiation. Forty-nine (90.7%) patients underwent one regimen of platinum-based combination therapy with gemcitabine and five (9.3%) patients underwent two or more regimens. Representative indexes of SIR were calculated as the HALP score, NLR, and PLR. In particular, the mean and range of NLR (4.49 and 0.75-22.0, respectively) were quite comparable with those of the previous study (5.0 and 0.5-18.3, respectively) (5).
Patient characteristics (n=54).
The NLR and PLR were associated with the initial tumor response to pembrolizumab. There were 5 patients with a complete response (CR), 11 with partial response (PR), eight with stable disease (SD), and 30 progressive disease (PD) cases. The patients were divided into two groups according to the initial response: CR/PR (n=19) and SD/PD (n=38). The possible associations between the two groups and the indices of SIR were evaluated. As shown in Figure 1, a high NLR and high PLR were significantly associated with poor initial tumor response (p=0.041 and p=0.018, respectively). In addition, a low HALP score tended to be associated with an unfavorable response (p=0.064).
The association between the pretreatment indices of systemic inflammatory response and initial tumor response to pembrolizumab in 54 patients with advanced urothelial cancer. The hemoglobin, albumin, lymphocyte, and platelet score (HALP), neutrophil-to-lymphocyte ratio (NLR), and platelet-to-lymphocyte ratio (PLR) were calculated from blood test data that were collected just before the initiation of treatment with pembrolizumab. The patients were divided into two groups based on the initial tumor response as partial or complete response (PR/CR) and no response (progressive disease or stable, PD/SD). Statistical difference is presented as the p-value.
The association between indices of SIR and OS of patients treated with pembrolizumab. To explore potential predictive indicators for the efficacy of pembrolizumab, the patients were divided into two groups according to the cut-off values for the HALP score [≥30.05 (n=19) vs. <30.05 (n=35)], NLR [≥3.33 (n=30) vs. <3.33 (n=24)], and PLR [≥173.73 (n=30) vs. <173.73 (n=24)], and the possible correlations of indices of SIR with OS were assessed. As shown in Figure 2, the OS of patients with an HALP score <30.05 or PLR≥173.73 was significantly shorter than an HALP score≥30.05 or PLR<173.73 at a median of 644 [95% confidence interval (CI)=339-949] versus 1,036 (95% CI=705-1,367) days (p=0.045) and 586 (95% CI=292-880) versus 1,036 (95% CI=952-1,120) days (p=0.016), respectively. There was no significant difference (p=0.599) between groups with NLR ≥3.33 and NLR <3.33 at a median of 818 (95% CI=559-1,077) and 788 (95% CI=379-1,197) days, respectively.
Kaplan–Meier analysis of overall survival of 54 patients with advanced urothelial cancers treated with pembrolizumab. The patients were stratified by high or low hemoglobin, albumin, lymphocyte, and platelet score (HALP), neutrophil-to-lymphocyte ratio (NLR), and platelet-to-lymphocyte ratio (PLR) based on each cut-off value. Statistical difference is presented as the p-value.
To determine independent predictors of OS in patients with advanced UC treated with pembrolizumab, univariate and multivariate analyses were performed. As shown in Table II, a PLR≥73.73 was highlighted as a significant prognostic factor for worse OS in the univariate proportional regression analysis (HR=2.33, 95% CI=1.10-4.91, p=0.027). The multivariate analysis also highlighted a high PLR as a prognostic factor that was significantly associated with worse OS (HR=2.56, 95% CI=1.16-5.65, p=0.020).
Clinicopathological parameters affecting the overall survival of patients treated with pembrolizumab (n=54).
Discussion
To date, various indices of SIR have been investigated as prognostic factors for treating UC with either radical surgery or systemic chemotherapy, including ICI therapy (2-8). In this study, we evaluated the HALP score, NLR, and PLR as representative indices of SIR that might potentially predict the efficacy of pembrolizumab in patients with advanced UC and found that PLR holds promise as an indicator for treatment with pembrolizumab. Although a high NLR was associated with a poor initial response to pembrolizumab treatment, it was not an independent prognostic factor of OS in the multivariate analysis. This finding is partially inconsistent with previous reports (5, 7, 8), especially in lung cancer in which the NLR was recently shown to have predictive value for the efficacy of ICIs (11, 12). However, the NLR might strongly reflect a patient’s tolerability to treatment, and it might be affected by physical impairment related to tumor invasion, regardless of the efficacy of pembrolizumab in certain other types of cancer. The NLR has long been reviewed as an independent predictor of patient outcomes after surgery or conventional chemotherapy for various solid tumor types (13). In fact, Ito et al. illustrated that a high NLR had the strongest association with poor OS in patients with a performance status of 2 or more and may largely reflect systemic inflammation caused by cancer (8).
The PLR has also been systematically reviewed and an association between an elevated PLR and poor survival of patients with various tumors and therapies has been demonstrated (9, 14). Regarding the efficacy of ICIs, an increasing number of reports suggest that the PLR is an independent predicator for survival in small or non-small cell lung cancer (15, 16), hepatocellular carcinoma (17), melanoma (18), gastric and colorectal cancer (19), and neuroendocrine neoplasms (20), whereas it is still inconsistent in UC. There is only one study showing an association between a high PLR and a shorter duration of SD in patients treated with PD-1/PD-L1 therapy for UCs and renal cell carcinomas (21). Notably, although several previous reports showed that both NLR and PLR were equivalently valuable as independent predictors for survival (15-17, 20), we observed a remarkable difference in their predictive value. Therefore, this might have important implications with regard to the future application of the PLR as a biomarker for the efficacy of pembrolizumab in UC. Mechanistically, neutrophilia has long been recognized as having the ability to inhibit the cytotoxic activity of lymphocytes through the secretion of myeloperoxidase and H2O2 that react with halides to form immunosuppressive products (22). Furthermore, tumor-associated neutrophils can maintain reactive oxygen species-mediated suppression of T-cells in the tumor microenvironment (23). Although tumor-associated neutrophils have a pro-tumorigenic function through the immunosuppression of tumor-infiltrating T-cells, the direct contribution of neutrophils to the efficacy of anti-PD-1/PD-L1 therapy remains inconclusive. Platelets have also been commonly recognized as having pro-tumorigenic involvement, protecting tumor cells against the immune system and facilitating metastasis, and promoting tumor growth with angiogenesis (24). Thrombocytosis predicts the invasiveness of bladder cancer and may adversely affect survival (25). Most prominently, regarding a study of upper tract UCs, among patients with high platelet counts, tumor PD-L1 positivity was significantly associated with shorter OS, suggesting the presence of a functional relationship between platelets and PD-L1 (26). Therefore, the PLR may be more tightly linked to the efficacy of pembrolizumab than the NLR in patients with UC.
Considering previous reports and our present results together, patients with advanced UC who have increasing platelet counts may not favorably respond to a conventional cytotoxic chemotherapy, but may be included in subgroups, such as patients with tumor PD-L1 expression, who would benefit from therapy with pembrolizumab. However, among these subgroups, individuals with a high PLR may benefit more from therapeutics other than pembrolizumab. Therefore, the PLR is the most easily accessible index and would provide valuable information for selecting therapy by routine blood tests.
This study has several limitations in providing strong evidence of a correlation between the PLR and the efficacy of pembrolizumab. It was a retrospective study at a single institution with a limited number of patients; possible relationships between the PLR and the proposed biomarkers such as the PD-L1 expression status, tumor mutational burden, and microsatellite instability in tumors were not evaluated. Further consecutive studies are required to establish the PLR as a predictive biomarker for the efficacy of pembrolizumab treatment in patients with advanced UC.
Acknowledgements
The Authors thank Edanz Group (www.edanzediting.com/ac) for editing a draft of this article.
Footnotes
↵* These Authors are designated as co-first authors.
Conflicts of Interest
None of the Authors declare any conflicts of interest.
Authors’ Contributions
RK and KA conceived and designed this study. RK, MI, KI, RM, KM, TI, HK and YK collected samples and recorded the general data and the observational indicators of patients. KA analyzed the data and wrote the first draft of the article. RK, HT, MH, RS, HK, TK, and YK critically reviewed and corrected the article. All Authors reviewed and approved the final version of the article.
- Received November 30, 2021.
- Revision received December 16, 2021.
- Accepted December 20, 2021.
- Copyright © 2022 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
