Abstract
Background: The prognostic significance of programmed death ligand 1 (PD-L1) has been reported in non-small cell lung cancer; however, the significance of PD-L1 expression in patients with resected small-cell lung cancer (SCLC) remains to be clarified. Materials and Methods: Forty patients with SCLC whose resected specimens were available for immunohistochemistry for PD-L1 were evaluated to determine the association between its expression and the clinicopathological factors and prognosis. Results: Among 40 patients, PD-L1 was expressed in tumor cells (TCs) of six (15%), tumor-infiltrating cells (ICs) of 16 (40%), and TCs and/or ICs cells of 18 (45%) patients. Patients with PD-L1-positve ICs and TCs and/or ICs exhibited significantly longer disease-free survival than those without PD-L1-expression (hazard ratio (HR)=0.268; 95% confidence interval (CI)=0.100-0.645; p=0.003 and HR=0.301; 95% CI=0.118-0.702; p=0.005, respectively). Conclusion: This study provides important evidence on the prognostic value of the PD-L1 expression in resected SCLC patients.
Small-cell lung cancer (SCLC) is a devastating subtype of lung cancer, and its biological, clinical and genetic characteristics are different from those associated with non-small-cell lung cancer (NSCLC), i.e., the cancer is more aggressive, more metastatic, more sensitive to chemotherapy and radiation, but more prone to relapse, has a poorer prognosis, and no genetic disorders targeted by specific kinase inhibitors exist at present (5, 20). Although chemotherapy and radiotherapy are considered to be the standard-of-care for patients with SCLC, surgical resection is considered to be one of the important modalities for a subset of SCLC (7, 13, 23, 26); however, further novel and promising treatment options should be urgently developed due to its aggressive features and poor prognosis. In addition, biomarkers to predict its prognosis and sensitivity to antitumor agents should also be identified.
Inhibition of the programmed death 1 (PD-1)/programmed death ligand 1 (PD-L1) immune checkpoint pathway has recently demonstrated a benefit in the survival of various types of cancers, such as melanoma and lung cancer (6). PD-1, a member of the immunoglobulin superfamily B7, is expressed on the surface of T cells and manipulates their activity through interaction with its ligands PD-L1 and PD-L2 (15). The interaction between PD-1 and PD-L1 or PD-L2 attenuates the T cell activity, which results in downregulation of the immune response against cancer cells (8), and inhibition of these interactions with PD-1 or PD-L1 blocking antibodies suppresses immune escape in cancer cells and induces immune response in T cells against cancer cells (15). Recent clinical trials, such as the CheckMate and KEYNOTE studies, demonstrated that nivolumab and pembrolizumab, representative PD-1 inhibitors, exhibited a survival benefit in comparison to conventional standard therapy in melanoma and lung cancer (3, 4, 16, 25). Thus, immune checkpoint inhibition has emerged as a novel and promising therapeutic option in the oncology field.
With regard to the significance of the PD-L1 expression, although PD-L1 is expected to be a potential biomarker to predict the response to immune checkpoint inhibition, definitive evidence has yet to be fully established (9). In NSCLC, controversial data that the PD-L1 expression can predict responses for patients with non-squamous NSCLC but not those with squamous NSCLC, have been reported (3, 4). In addition to its role as a potential predictive biomarker, the prognostic significance of PD-L1 has been suggested in several types of cancers, including NSCLC, SCLC, breast cancer, prostate cancer, and laryngeal squamous cell carcinoma (1, 2, 10, 12, 17, 18, 24, 27). With regard to PD-L1 expression in patients with SCLC, Ishii and colleagues demonstrated the association between its expression and a significantly favorable overall survival (OS), which was retained in a multivariate analysis (12). Because almost all patients examined in that study were non-surgical patients, the precise role of PD-L1 expression in surgically resected patients with SCLC remains unclear.
In this translational study, we examined PD-L1 expression in 40 patients with SCLC who underwent surgery to investigate its expression in association with clinicopathological characteristics and the clinical outcome, such as the disease-free survival (DFS) and OS.
Materials and Methods
Study patients. Among 62 patients with SCLC who underwent surgery at the Department of Surgery and Science, Graduate School of Medical Sciences, Kyushu University between April 1974 and August 2015, 40 patients, whose resected specimens were available for immunohistochemistry of PD-L1, were included in this translational study. Pathological staging was performed using the 7th edition of the TNM Classification of Malignant Tumors. In addition to the pathological stage, the patients' age, gender, performance status, smoking history, pathological tumor and pathological lymph nodal factors, pleural invasion (pl), lymphatic invasion (ly), vascular invasion (v) and surgical procedure were examined. The present study was approved by our institutional review board.
Immunohistochemical analysis. Immunohistochemistry was performed in 40 patients with surgically resected SCLC, using formalin-fixed tissue sections. Sections were cut at widths of 4 μm from paraffin-embedded material, dewaxed with xylene, and rehydrated through a graded series of ethanol. After the inhibition of endogenous peroxidase for 30 min with 3% H2O2 in methanol, the sections were pretreated with Target Retrieval Solution (Dako, Glostrup, Denmark) in a decloaking chamber at 110°C for 15 min before being incubated with a monoclonal antibody at 4°C overnight. The immune complex was detected using a DAKO EnVision Detection System (Dako). The sections were finally reacted in 3,3’-diaminobenzidine, counterstained with hematoxylin, and mounted.
The primary antibody used was an anti-human PD-L1 rabbit monoclonal antibody (clone SP142, dilution 1:100; Spring Bioscience, Ventana, Tucson, AZ, USA). Tumor cells (TCs) and tumor-infiltrating immune cells (ICs), such as macrophages, showed membranous staining for PD-L1, that was evaluated as positive staining. The proportion of PD-L1 positive TCs and ICs was estimated as the percentage of total TCs and ICs, respectively. Because the distribution of PD-L1 positive cells was very focal, positivity for PD-L1 was defined as ≥5% of TCs or ICs stained for PD-L1 (11). Sections from human placentas were used as positive controls.
Patients' characteristics.
Statistical analysis. The associations between the PD-L1 expression and patient characteristics were analyzed using the Fisher's exact test. The OS was defined as the time from the initial surgery until death from any cause, while the DFS was defined as the time from the initial surgery until recurrence. The Kaplan-Meier method was used to estimate survival probabilities. The curves of the two groups were statistically compared using the log-rank test. Univariate and multivariate analyses were performed using a Cox proportional hazard model. All statistical analyses were conducted using the JMP version 12 software program (SAS Institute, Cary, NC, USA). p-Values of <0.05 were considered to indicate statistically significant differences.
Results
Clinicopathological characteristics of patients enrolled in this study. Table I shows the characteristics of patients included in this translational study. The median age of the patients was 69 years (range=48-84 years), and the number of male and female patients was 34 (85%) and six (15%), respectively. The performance status of the enrolled patients was 0 or 1. Most patients (85%) had a history of smoking. Pathological tumor and nodal stages were as follows: T1, 16 (40%)/T2, 18 (45%)/T3, 4 (10%)/T4, 2(5%); N0, 23 (57.5%)/N1, 8 (20%)/N2, 7 (17.5%)/N3, 2 (5%). Twenty (50%), nine (22.5%), and 11 (27.5%) patients were diagnosed with pathological stage I, II, and III, respectively. Pathological examinations revealed pl, ly, and v in 15 (37.5%), 11 (27.5%), and 21 (52.5%) patients, respectively. Twenty-nine patients (72.5%) underwent surgical resection of more than one lobe, while six (15%) received sublobar resection.
Representative images of PD-L1-positive TCs (A) and ICs (B). Scale bar: 100 μm.
PD-L1 expression and its association with clinicopathological features. PD-L1 expression was analyzed in both TCs and ICs, and Figure 1A and B denotes representative images of PD-L1-positive TCs and ICs. Immunoreactivity of TCs and ICs for PD-L1 was observed in six (15%) and 16 (40%) SCLC patients, respectively, and 18 (45%) patients were positive for PD-L1 in TCs and/or ICs. Fisher's exact test was conducted to determine the associations between the PD-L1 expression in TCs, ICs, and TCs and/or ICs and the abovementioned clinicopathological factors in patients with resected SCLC; the findings are summarized in Tables II, III and IV. The PD-L1 expression in TCs was significantly associated with the absence of nodal metastasis and the presence of v, while that in ICs was significantly associated with an advanced tumor stage. No significant associations between the PD-L1 expression in TCs and/or ICs and patient characteristics were identified.
Survival analysis according to PD-L1 expression. At the time of the investigation, the median follow-up time was 14.1 months (range=0.4-167.0 months). The log-rank test showed that PD-L1 positivity in ICs and TCs and/or ICs was associated with a significantly prolonged DFS in comparison to PD-L1 negativity (median 83.7 vs. 9.3 months, p=0.003; median 83.7 vs. 9.0 months, p=0.005, respectively), while the PD-L1 expression status in TCs did not significantly affect the DFS (p=0.080; Figures 2A, 3A, and 4A). Positive PD-L1 staining in ICs was identified to be significantly associated with the improved DFS by the univariate Cox proportional hazard model (hazard ratio (HR), 0.268; 95% confidence interval (CI), 0.100-0.645; p=0.003), and retained significance in the multivariate analysis (HR, 0.123; 95% CI, 0.035-0.363; p<0.001), as well as higher age and the presence of nodal metastasis (Table V). Similar results were observed for the PD-L1 expression in TCs and/or ICs (HR=0.301; 95% CI, 0.118-0.702; p=0.005 in the univariate analysis and HR=0.151; 95% CI=0.047-0.418; p<0.001 in the multivariate analysis, respectively; Table VI). In contrast, a univariate examination revealed PD-L1 expression in TCs not to be significantly associated with a favorable DFS (HR=0.349; 95% CI=0.081-1.035; p=0.059). Alhough not significant, a Kaplan-Meier analysis and univariate Cox proportional model highlighted the trend of superior OS outcomes in patients overexpressing PD-L1 in TCs, ICs, and TCs and/or ICs (p=0.130, 0.264, and 0.109, respectively; Figures 2B, 3B, and 4B).
Discussion
PD-L1 expression appears to serve as a potential biomarker to predict the prognosis of various types of cancers, including lung cancer, breast cancer, prostate cancer, and laryngeal squamous cell carcinoma (1, 2, 10, 12, 17, 18, 24, 27). Some groups have indicated a positive impact of its expression on survival, while others have indicated a negative impact. This translational study highlighted the prognostic significance of the PD-L1 expression in patients with resected SCLC. The expression of PD-L1 in ICs and TCs and/or ICs was shown to be associated with better outcome in the DFS by univariate and multivariate Cox proportional hazard models, and a non-significant trend for a superior OS was observed in patients overexpressing PD-L1 in TCs, ICs, and TCs and/or ICs, which was similar to previous results that the expression of PD-L1 is an independent predictor of a better outcome in patients with SCLC (12). Furthermore, our data in agreement with the results of several previous studies for other types of cancer, including NSCLC (2, 18, 24). Several reasons can explain why the expression of PD-L1 resulted in improved survival, such as methodologic differences in the antibodies and criteria applied in each study and possible immune-related mechanisms. The latter is thought to be due to the findings that CD8+ T cells up-regulate the PD-L1 expression, and contribute to local antitumor immune control, by releasing cytokines, such as interferon γ (17, 22).
Association between PD-L1 expression in tumor cells and clinicopathological factors in patients with resected SCLC.
Association between PD-L1 expression in tumor-infiltrating immune cells and clinicopathological factors in patients with resected SCLC.
Association between PD-L1 expression in tumor cells and tumor-infiltrating immune cells and clinicopathological factors in patients with resected SCLC.
With regard to the PD-L1 expression in patients with SCLC, PD-L1 was reportedly expressed in approximately 70% of the membrane and/or cytoplasm of tumor and stromal lymphocytes (12). In addition, Schultheis et al. reported that 17 out of 92 tumors (18.5%) overexpressed PD-L1 in stromal cells adjacent to the tumor, but not in tumor cells (19), and the KEYNOTE-028 trial reported the frequency of PD-L1 expression in tumor nests or stroma to be 28.6% (14). Our study demonstrated that among 40 patients, PD-L1 was expressed in TCs of six (15%), ICs of 16 (40%), and TCs and/or ICs of 18 (45%) patients, respectively. Differences in the antibodies used and criteria for PD-L1 positive cells might explain the difference in the rate of PD-L1 expression in SCLC between our study and previous studies. Importantly, our report appears to be more appropriate for the evaluation of the PD-L1 expression than other studies because whole-tissue section samples were used, whereas almost all samples examined in other reports were biopsied samples (12, 14, 19). Thus, the data obtained in this study might provide insights into the selection of SCLC patients who would benefit from immune checkpoint inhibitors, if they would be approved for SCLC (21).
Kaplan-Meier curves of the DFS and OS according to the PD-L1 expression in TCs.
Kaplan-Meier curves of the DFS and OS according to the PD-L1 expression in ICs.
There exist several limitations associated with the present retrospective study. First, this study examined a small cohort of patients with resected SCLC. However, it should be emphasized that all specimens analyzed for PD-L1 expression were surgically resected whole-tissue section samples. Another limitation is that PD-L1 immunohistochemistry was conducted using only one antibody, which was different from the antibodies used in previous studies. Several antibodies for the detection of PD-L1 should be evaluated in future investigations. Since the kind of antibodies and cut-offs used to detect PD-L1 expression might affect the results, examination of agreement rates for various combinations between different antibodies and selected cutoffs would help circumvent this obstacle.
In conclusion, our results showed that PD-L1 was expressed in TCs (15%), ICs (40%), and TCs and/or ICs (45%). The assessment of the PD-L1 expression in patients with resected SCLC could be useful to predict a better DFS, which should be investigated by further studies with a larger cohort.
Kaplan-Meier curves of the DFS and OS according to the PD-L1 expression in TCs and ICs.
Univariate and multivariate analyses for the disease-free survival according to the PD-L1 expression in tumor-infiltrating cells.
Univariate and multivariate analyses for the disease-free survival according to the PD-L1 expression in tumor cells and tumor-infiltrating cells.
Acknowledgements
The Authors would like to thank Brian T. Quinn for his critical comments on the manuscript.
Footnotes
Conflicts of Interest
The Authors declare no conflicts of interest associated with this study.
Funding
This work was not supported by any funding sources.
- Received June 11, 2016.
- Revision received July 4, 2016.
- Accepted July 6, 2016.
- Copyright© 2016 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved