Association between the Absolute Baseline ... - Health Advance

1 downloads 0 Views 987KB Size Report
[8] Templeton AJ, McNamara MG, Seruga B, et al. Prognostic role · of neutrophil-to-lymphocyte ratio in solid tumors: a sys- · tematic review and meta-analysis.
Clinical Oncology 28 (2016) 790e796 Contents lists available at ScienceDirect

Clinical Oncology journal homepage: www.clinicaloncologyonline.net

Original Article

Association between the Absolute Baseline Lymphocyte Count and Response to Neoadjuvant Platinum-based Chemotherapy in Muscle-invasive Bladder Cancer R. Leibowitz-Amit *yy, A. Israel y, M. Gal *, E.A. Atenafu z, Z. Symon *yy, O. Portnoy xyy, M. Laufer {, Z. Dotan {yy, J. Ramon {yy, D. Avni jj, E. Fridman **yy, R. Berger *yy * Department

of Oncology, Sheba Medical Center, Tel-Hashomer, Israel Department of Family Medicine, Clalit Health Services, Jerusalem, Israel z Department of Biostatistics, Princess Margaret Cancer Center, Toronto, Ontario, Canada x Department of Diagnostic Imaging, Sheba Medical Center, Tel-Hashomer, Israel { Department of Urology, Sheba Medical Center, Tel-Hashomer, Israel jj Laboratory of Molecular Cell Biology, Center for Cancer Research, Sheba Medical Center, Tel Hashomer, Israel ** Department of Pathology, Sheba Medical Center, Tel-Hashomer, Israel yy Faculty of Medicine, Tel-Aviv University, Tel-Aviv, Israel y

Received 3 March 2016; received in revised form 6 June 2016; accepted 22 June 2016

Abstract Aims: Platinum-based neoadjuvant chemotherapy (NAC) improves overall survival in muscle-invasive bladder cancer (MIBC). A pathological complete response (pCR) at radical cystectomy after NAC is associated with better overall survival, but there are no established predictive biomarkers of response to NAC in MIBC. The aim of this study was to find laboratory variables associated with pCR following NAC. Materials and methods: We carried out a retrospective review of MIBC patients treated with NAC followed by radical cystectomy at the Sheba Medical Center between 2005 and 2015. Overall survival was calculated using the KaplaneMeier product-limit method and compared between patients who achieved or did not achieve pCR using the Log-rank test. Baseline and pre-surgery laboratory values were collected and compared between patients who subsequently achieved pCR and those who did not using logistic regression. Results: Fifty-eight patients underwent radical cystectomy after NAC, with a median follow-up of 32 (range 4.8e111.4) months from diagnosis. Of 55 patients with documented pathological outcome on radical cystectomy, 17 (31%) achieved pCR (complete responders). Of the 15 complete responders with follow-up data, 13 (87%) were still alive at time of last follow-up for this study (July 2015). Patients who did not achieve pCR had a significantly worse overall survival than complete responders (P ¼ 0.0007). The baseline lymphocyte count, neutrophilelymphocyte ratio (NLR) and plateletelymphocyte ratio (PLR) were significantly associated with response (P ¼ 0.037, P ¼ 0.045, P ¼ 0.042, respectively) on univariate analysis, whereas baseline albumin, haemoglobin, neutrophils, platelets and the total white blood count were not significantly associated with response. Lymphocyte counts were significantly higher in responders than nonresponders throughout three time points (P ¼ 0.003 using a generalised linear mixed model). Conclusions: A high baseline level of lymphocytes is associated with the achievement of pCR at radical cystectomy after NAC, which, in turn, is associated with a significantly longer overall survival. Our results suggest that chemosensitivity in MIBC is associated with lymphocyte count. Ó 2016 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.

Key words: Bladder cancer; chemosensitivity; lymphocytes; MIBC; neoadjuvant chemotherapy; urothelial carcinoma

Introduction Muscle-invasive bladder cancer (MIBC) continues to pose significant therapeutic challenges. When the disease is Author for correspondence: R. Berger, Oncology Department, Sheba Medical Center, Ramat-Gan 52621, Israel. Tel: þ972-3-5302542. E-mail address: [email protected] (R. Berger).

localised to the bladder, standard treatment generally includes neoadjuvant chemotherapy (NAC) with a platinumbased combination followed by radical cystectomy [1]. Whereas this multidisciplinary approach was shown to increase survival in a pivotal phase III trial [2] and in several meta-analyses [3,4], the extent of benefit for the entire patient population is small. It is now well established that

http://dx.doi.org/10.1016/j.clon.2016.07.007 0936-6555/Ó 2016 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.

R. Leibowitz-Amit et al. / Clinical Oncology 28 (2016) 790e796

the achievement of pathological complete response (pCR) in the cystectomy specimen after NAC is associated with a significantly improved outcome [5], yet only about 30% of patients achieve this goal. Moreover, it is hard to specifically quantify the direct contribution of NAC to the achievement of pCR as the pre-treatment transurethral resection of the bladder tumour can also lead to a complete response in some of the patients. As there are no proven clinical, biochemical or molecular biomarkers that predict which patients are likely to respond to NAC to date, many patients receive futile toxic chemotherapy that postpones definitive surgery. Finding markers that predict the response to NAC in MIBC is therefore highly needed. In recent years, immunotherapy has emerged as a promising systemic approach in a wide range of solid cancers; specifically, there is extensive ongoing clinical research on the use of immunotherapy in bladder cancer [6], long thought to be linked to aberrant inflammatory/immune processes [7]. Currently, it is unclear which patients are more likely to respond to an immunotherapeutic approach. Similarly, little is known about the crosstalk between the immune system and chemosensitivity in bladder cancer. The neutrophilelymphocyte ratio (NLR) in the blood count has emerged as a strong prognostic factor in a wide range of solid malignancies, with high values evidently associated with a worse prognosis [8]. The plateletelymphocyte ratio (PLR) has also been linked to a poor prognosis in cancer [9]. A persistently high NLR throughout NAC was also associated with a lack of response to NAC in MIBC in a Canadian patient cohort [10]. Our aim here was to further study the relationship between the different components of the inflammatory/immune systems, as manifested in the complete blood count and the response to NAC in MIBC.

Materials and Methods Data Collection and Definitions This was a retrospective chart review of all patients treated with NAC followed by radical cystectomy for MIBC between October 2005 and December 2014 that were documented in the institutional electronic patient record at the Sheba Medical Center. This study received institutional review board approval. Inclusion criteria included: a histological confirmation of MIBC with predominant urothelial histology, no radiological evidence of distant or lymph node metastasis, the administration of combination platinumbased chemotherapy with neoadjuvant intent and a subsequent radical cystectomy. Patients who received NAC with adjuvant intent but did not subsequently undergo radical cystectomy (for any reason) were summarised for sake of database completeness but were not included in our primary or secondary end point analyses. Patients who underwent upfront radical cystectomy without prior NAC and patients diagnosed with MIBC that was not upfront amenable to curative therapy were excluded from this analysis as well.

791

Clinical and pathological data were extracted from the electronic patient record. Laboratory variables including haemoglobin, neutrophil, lymphocyte, platelet and total white blood cell (WBC) counts and albumin levels were retrieved from blood tests taken within 1 week before NAC start (designated ’Pre-NAC’ or baseline), within a week before radical cystectomy (designated ’pre-surgery’) and 4e6 weeks after radical cystectomy (designated ’post-surgery’). The pre-surgery laboratory variables were taken from the closest time point available before radical cystectomy (and no longer than 1 month before radical cystectomy) as long as it was at least 3 weeks after completion of the last chemotherapy cycle. NLR and PLR were calculated as the ratio between the absolute neutrophil and lymphocyte count or the absolute platelet and lymphocyte count, respectively, in the peripheral blood. NLR measurements were not assessed for patients with known active infection, patients receiving steroid medication (20 mg of prednisone a day or the equivalent) or patients with known haematological malignancies. pCR was defined when no evidence of invasive cancer was reported in the cystectomy sample and when all lymph nodes in the surgical specimen were free of tumour. A pathological partial response was defined when there was unequivocal down-staging of the tumour in the cystectomy specimen with remaining evidence of muscle invasion. Stable and progressive disease were defined when there was no clear evidence of response or when there was unequivocal evidence of disease progression (such as T4 disease not evident on pre-NAC radiological staging), respectively. In the current analysis, patients who achieved pCR are designated as ’complete responders’, whereas all other patients are defined as ‘non-complete responders’ or ’non-responders’ for short. Primary End Point and Statistical Analysis Our primary end point was to find laboratory variables that are associated with the achievement of pCR at radical cystectomy after NAC. Our secondary end point was to study the trends in laboratory variables that are associated with the achievement of pCR throughout time. Categorical variables were summarised with counts and percentages. Continuous variables were summarised with medians/ranges and/or mean (standard deviation) accordingly. The Kolmogorov-Smirnov test was used for normality assumption of lab values and log transformation was used for those not satisfying the normality assumptions. Time to death/censoring was calculated in months from the diagnosis date to the death date or the date of the last follow-up, respectively. The overall survival rate was calculated using the KaplaneMeier product-limit method and the Log-rank test was used to assess the effect of pCR on overall survival. A logistic regression analysis was used to determine which laboratory values were associated with the primary outcome of pCR. A generalised linear mixed model was used to account for the time trend and co-linearity when assessing changes in laboratory variables between patients who achieved pCR (’complete responders’) and those who

792

R. Leibowitz-Amit et al. / Clinical Oncology 28 (2016) 790e796 Table 2 Pathological outcome at radical cystectomy after neoadjuvant chemotherapy

Fig 1. Patient treatment allocation.

did not (’non-responders’). P values were two-sided and P < 0.05 was considered statistically significant. A data analysis was carried out using SAS Version 9.4 (SAS Institute, Inc., Cary, NC).

Results Patient Characteristics and Outcome Eighty-one patients were started on NAC with curative intent for MIBC. The median age at diagnosis of MIBC was 67.6 years (range 41e87 years) and 65 (80%) of the patients were male. Sixty-two (76%) patients received a combination of cisplatinegemcitabine (CIS-GEM), 11 (14%) received a combination of carboplatinegemcitabine (Carbo-GEM), one received a combination of methotrexate, doxorubicin, vinblastine and cisplatin (MVAC) and for seven (9%) the regimen could not be retrieved from the medical records. The median time from the diagnosis of MIBC to the start of NAC was 1.9 months (range 0.3e14.7 months). Of these 81

Pathological outcome on radical cystectomy

N

% (of all known outcome)

Complete response Partial response Stable disease/progressive disease Unknown

17 6 32 3

31 11 58 e

patients, 58 (72%) went on to have radical cystectomy as initially planned. Twenty-three patients (28%) did not undergo radical cystectomy for the following reasons: five (22%) refused surgery, seven (30%) were ineligible for operation due to clinical deterioration/toxicity, in six (26%) the disease had progressed beyond definitive treatment while on chemotherapy and for five (22%) the reasons were unknown (Figure 1). Of these 23 patients, eight (35%) went on to have local therapy in the form of radiotherapy or chemoradiotherapy. The patient characteristics and baseline (’Pre-NAC’) laboratory values of the 58 patients who underwent NAC followed by radical cystectomy are given in Table 1. Forty-eight (83%) of these patient were male, with a median age of 67 years (range 41e78 years). Pathological response data are given in Table 2; of 55 patients with known pathological outcome on radical cystectomy, 17 patients (31%) had a complete response, six (11%) had a partial response and 32 (58%) had stable disease or progressive disease. For three patients, the pathological outcome from the radical cystectomy was not documented. The pathological response was dichotomised to complete responders (’responders’; 31%) and non-complete responders (partial response, stable disease and progressive disease, collectively; ’non-responders’; 69%), in accordance with the known positive prognostic significant of achieving pCR at radical cystectomy after NAC [5]. At the time of data analysis, 32 (55%) patients were still alive. The median follow-up for the entire cohort was 32.3 months (range 4.8e111.4 months). Of 15 ’responders’ with follow-up survival data, 13 (87%) were alive at the time of

Table 1 Characteristics of patients who underwent neoadjuvant chemotherapy (NAC) followed by radical cystectomy (RC) Characteristic

Median value

Range

Age Time from start of NAC to RC Baseline laboratory values Haemoglobin (g/dl) White blood cell count (109/ml) Neutrophil count (109/ml) Lymphocyte count (109/ml) Platelet count (109/ml) NLR PLR Albumin (g/dl)

67 years 4.6 months

41e78 years 1.8e21.3 months

13.3 8.5 5.7 1.8 236 3.0 134 4.1

10.0e15.0 4.1e13.7 2.3e10.9 0.7e3.2 94e438 1.3e10.2 37e391 2.9e4.8

NLR, neutrophilelymphocyte ratio; PLR, plateletelymphocyte ratio.

Mean (standard deviation) 13.0 (1.3) 8.5 (2.1) 5.7 (1.8) 1.8 (0.6) 248.5 (73.8) 3.6 (2.0) 155.4 (69.5) 4.1 (0.4)

R. Leibowitz-Amit et al. / Clinical Oncology 28 (2016) 790e796

Fig 2. Overall survival curves of ’responders’ and ’non-responders’. Survival in patients who achieved a pathological complete response (dotted red line) or did not achieve a pathological complete response (blue line) was calculated using the KaplaneMeier analysis and compared using the Log-rank test.

the end of follow-up for this study versus 16 of 38 ’nonresponders’ (42%; P ¼ 0.0007 using the Log-rank test). At 1 year after diagnosis, 100% of the responders and 88% of the non-responders were alive; at 2 years after diagnosis, 93% and 70% of the responders and non-responders were alive, respectively. The overall survival curves of responders and non-responders are shown in Figure 2. Laboratory Variables Associated with the Achievement of Pathological Complete Response The results of the univariate analysis assessing pre-NAC laboratory variables and pathological response are given in Table 3. The pre-NAC lymphocyte count, PLR and NLR were each associated with pCR (P < 0.05 for each variable). A higher pre-NAC lymphocyte count and a lower PLR or NLR were associated with a higher likelihood of achieving pCR. There were no responders among the nine patients with a baseline lymphocyte count of less than 1.3109/l versus 14 responders among the 33 patients with a baseline

793

lymphocyte count that was higher than 1.3109/l (P ¼ 0.016 using the chi-squared test). The pre-NAC levels of haemoglobin, albumin, WBC, neutrophil count or platelet count were not significantly associated with pCR in our cohort. The lymphocyte counts and the PLR before surgery and after surgery were also both associated with a pathological response at radical cystectomy by univariate analysis (Table 4). Age, gender and the sub-histological type of the urothelial carcinoma (papillary, non-papillary/flat or sarcomatoid/undifferentiated) were also not associated with pCR (results not shown). Six patients progressed to metastatic or inoperable disease during NAC and did not ultimately undergo cystectomy (Figure 1). A repeat univariate analysis incorporating their pre-NAC laboratory variables into the non-responder groups yielded similar results (Supplementary Table S1). Most laboratory values satisfied normality assumption except NLR and PLR, where logarithmic transformation was used. The transformed variables satisfied normality assumption in both cases. We then analysed the time trend and co-linearity of the different laboratory variables using a generalised linear mixed model throughout time. All the time trends for laboratory values were statistically significant except for WBC count (which was marginal, P value ¼ 0.065). There was a statistically significant difference in the lymphocyte count between responders and non-responders throughout time (P ¼ 0.003), with the responders exhibiting a higher and constant mean lymphocyte count at all three time points (Figure 3F) and the nonresponders exhibiting a trend of a decrease in total lymphocyte count with time. The NLR and PLR also showed a statistically different difference between responders and non-responders (P < 0.01 for both), with both ratios being lower and relatively constant in responders than nonresponders at all three time points and increasing with time in non-responders (Figure 3G, H, respectively). The total WBC counts were constant throughout time (Figure 3C), whereas haemoglobin and albumin levels decreased with time irrespective of response (Figure 3A, B) and the levels of the neutrophil counts and the platelet count increased with time irrespective of response (P < 0.05 for each; Figure 3D, E).

Table 3 Univariate logistic regression of pre-neoadjuvant chemotherapy laboratory variables and pathological complete response Effect

n

Pathological complete response [mean (standard deviation)] No

Yes

Lymphocyte count PLR NLR Albumin Haemoglobin Neutrophil count Platelet count White blood cell count

42 41 41 22 41 42 42 42

1.67 (0.58) 166.87 (72.32) 3.97 (2.10) 3.99 (0.38) 12.79 (1.35) 5.82 (1.86) 252.32 (81.91) 8.47 (2.11)

2.10 (0.56) 119.57 (36.26) 2.58 (0.76) 4.25 (0.43) 13.36 (1.19) 5.55 (1.67) 243.57 (59.21) 8.64 (2.06)

NLR, neutrophilelymphocyte ratio; PLR, plateletelymphocyte ratio.

P value

Odds ratio

95% confidence interval

0.04 0.04 0.05 0.18 0.19 0.64 0.72 0.80

3.63 0.98 0.48 9.66 1.44 0.92 0.99 1.04

1.12e12.24 0.97e0.99 0.23e0.98 0.34e273.2 0.83e2.51 0.63e1.33 0.99e1.01 0.76e1.43

794

R. Leibowitz-Amit et al. / Clinical Oncology 28 (2016) 790e796

Table 4 Univariate logistic regression of pre-surgery and post-surgery laboratory variables and pathological complete response Effect

Mean Lymphocyte count (standard deviation) Mean PLR (standard deviation)

Time point

Pre-surgery Post-surgery Pre-surgery Post-surgery

Pathological complete response [mean (standard deviation)] No

Yes

1.59 (0.65) 1.31 (0.58) 165.99 (89.62) 277.13 (137.73)

2.29 (0.57) 2.11 (0.69) 81.52 (24.34) 137.50 (42.53)

Discussion In this retrospective review of all patients treated with NAC followed by radical cystectomy at the Sheba Medical Center between October 2005 and December 2014, 17 of 55 patients (31%) achieved pCR. We show here that a higher absolute baseline lymphocyte count before the initiation of neoadjuvant platinum-based chemotherapy for MIBC was associated with a higher likelihood of achieving a pCR in the subsequent radical cystectomy. A higher baseline NLR and PLR were associated with a lower likelihood of achieving a pCR, but as both the absolute neutrophil or platelet counts were not significantly associated with pCR, these observations were probably driven by the lymphocyte count comprising the denominator of both ratios. The lymphocyte count continued to be higher in responders immediately before surgery and upon recovery from surgery. Accordingly, the NLR and PLR continued to be lower in responders than in non-responders. The retrospective nature of this analysis clearly cannot determine whether a higher lymphocyte count contributes to the achievement of a pCR or is an unrelated or non-specific phenomenon. The observations that the median albumin and haemoglobin levels did not differ between responders and non-responders suggest that these groups do not, a priori, significantly differ in terms of general health. It is therefore less likely, in our opinion, that the lower lymphocyte count in non-responders is a non-specific manifestation of decreased reserve. Whether the significant difference in lymphocyte count drives the difference in responsiveness to a platinum-based chemotherapy regimen, or is unrelated to it, is yet to be studied. With the recent realisation that some cancerous tumours actively suppress the adaptive arm of the immune system, there is an ongoing revolution in the treatment of solid cancers with the increasing use of novel T-cell-based immunotherapies (reviewed in [11]). Bladder cancer is thought to be a promising target for immune modulation, and there are currently several ongoing and completed clinical trials assessing a few different immunotherapeutic approaches in this malignancy [6,12]. Almost 20 years ago, CD8þ lymphocyte infiltrates were recognised within urothelial carcinomas and suggested to have anti-neoplastic activity [13], and the extent of tumour infiltrating lymphocytes has repeatedly been associated with better prognosis in MIBC (reviewed in [14]). In parallel, it is becoming clear that some chemotherapies exert part of their anti-neoplastic effect by eliciting an ’immunogenic cell death’ (ICD) of tumour cells,

P value

Odds ratio

95% confidence interval

0.01 0.03 0.01 0.02

7.3 6.6 0.95 0.98

1.6e33.4 1.3e34.4 0.92e0.99 0.966e0.997

which in turn activates the adaptive arm of the immune system against the tumour [15]. The mechanisms through which ICD causes the activation of the immune system are still being studied. There is experimental pre-clinical evidence suggesting that oxaliplatin, but not cisplatin, can lead to ICD [16]. Gemcitabine has not been associated with ICD to the best of our knowledge. It is tempting to speculate, based on our results, that a low baseline lymphocyte count impedes the ability of the adaptive immune system to contribute to the anti-neoplastic effect of chemotherapy, thus precluding the ability to achieve a complete pathological response. Should this hypothesis prove to be true, then potential methods to increase the function and activity of the adaptive immune response may prove to be useful in increasing chemosensitivity in MIBC. Our analysis has several limitations; first, its retrospective nature can merely point to statistically significant associations between the different laboratory variables and outcome, but cannot determine cause-effect relationships. Second, our small cohort size may not have allowed probing of clinically significant associations that may only have been discovered with larger cohorts. Third, the patients analysed in this current work were treated over a long period of time, potentially introducing variance in treatment and hence in outcome; in this regard, it is important to note that there have not been any significant changes in the approach to local MIBC in the timeframe depicted in this manuscript. Notwithstanding these important limitations, we believe that our current analysis points to intricate cross-talks between adaptive immunity, chemosensitivity and outcome in MIBC. Recently, pre-treatment lymphocytopenia was shown to be an independent adverse prognostic factor in both local MIBC and in advanced bladder cancer. The authors suggested that lymphopenia may be a manifestation of cancer-induced immune suppression driving tumour progression [17]. Our current results support and add to these previous observations, suggesting that lymphopenia may also impede the response to chemotherapy. Clearly our work should be further validated in independent patient cohorts and in prospective clinical trials.

Conclusions Our analysis shows that a higher baseline lymphocyte count before NAC for MIBC is associated with a higher chance of achieving a pCR at radical cystectomy, and that

R. Leibowitz-Amit et al. / Clinical Oncology 28 (2016) 790e796

795

Fig 3. Time trend of laboratory variables in ’responders’ and ’non-responders’. The mean value at three time points (pre-neoadjuvant chemotherapy, pre-surgery and post-surgery) are shown with lines (blue: non-responders; red: responders), whereas individual values are depicted in dots.

796

R. Leibowitz-Amit et al. / Clinical Oncology 28 (2016) 790e796

’complete responders’ are characterised by higher lymphocyte counts than non-responders before and after cystectomy. Our results may suggest that the adaptive arm of the immune response plays a role in determining chemosensitivity and may, if verified prospectively, aid in better patient selection to NAC or upfront cystectomy in bladder cancer.

Acknowledgment This work was funded by generous donations from the Israeli Jack Craps Foundation.

Appendix A. Supplementary data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.clon.2016.07.007.

References [1] Seah JA, Blais N, North S, et al. Neoadjuvant chemotherapy should be administered to fit patients with newly diagnosed, potentially resectable muscle-invasive urothelial cancer of the bladder (MIBC): a 2013 CAGMO Consensus Statement and Call for a Streamlined Referral Process. Can Urol Assoc J 2013;7(9e10):312e318. [2] Grossman HB, Natale RB, Tangen CM, et al. Neoadjuvant chemotherapy plus cystectomy compared with cystectomy alone for locally advanced bladder cancer. N Engl J Med 2003;349(9):859e866. [3] Sternberg CN, Bellmunt J, Sonpavde G, et al. ICUD-EAU International Consultation on Bladder Cancer 2012: chemotherapy for urothelial carcinoma-neoadjuvant and adjuvant settings. Eur Urol 2013;63(1):58e66. [4] Advanced Bladder Cancer Meta-analysis Collaboration. Neoadjuvant chemotherapy in invasive bladder cancer: update of a systematic review and meta-analysis of individual patient data advanced bladder cancer (ABC) meta-analysis collaboration. Eur Urol 2005;48(2):202e206. [5] Rosenblatt R, Sherif A, Rintala E, et al. Pathologic downstaging is a surrogate marker for efficacy and increased survival following neoadjuvant chemotherapy and radical cystectomy for muscle-invasive urothelial bladder cancer. Eur Urol 2012;61(6):1229e1238.

[6] Pembrolizumab (MK-3475) for advanced urothelial cancer: updated results and biomarker analysis from KEYNOTE-012. ASCO Annual Meeting. Available at: http://meetinglibrary. asco.org/content/147894-156; 2015. [7] Masson-Lecomte A, Guy L, Pedron P, et al. A switch from GnRH agonist to GnRH antagonist in castration-resistant prostate cancer patients leads to a low response rate on PSA. World J Urol 2013;31:339e343. [8] Templeton AJ, McNamara MG, Seruga B, et al. Prognostic role of neutrophil-to-lymphocyte ratio in solid tumors: a systematic review and meta-analysis. J Natl Cancer Inst 2014;106(6). [9] Templeton AJ, Ace O, McNamara MG, et al. Prognostic role of platelet to lymphocyte ratio in solid tumors: a systematic review and meta-analysis. Cancer Epidemiol Biomarkers Prev 2014;23(7):1204e1212. [10] Seah J-A, Leibowitz-Amit R, Atenafu EG, et al. Neutrophillymphocyte ratio (NLR) and pathological response to neoadjuvant chemotherapy in muscle-invasive bladder cancer. Clin Gen Cancer 2015;13:e229ee233. [11] Houot R, Schultz LM, Marabelle A, Kohrt H. T-cell-based immunotherapy: adoptive cell transfer and checkpoint inhibition. Cancer Immunol Res 2015;3(10):1115e1122. [12] Powles T, Eder JP, Fine GD, et al. MPDL3280A (anti-PD-L1) treatment leads to clinical activity in metastatic bladder cancer. Nature 2014;515(7528):558e562. [13] Housseau F, Zeliszewski D, Roy M, et al. MHC-dependent cytolysis of autologous tumor cells by lymphocytes infiltrating urothelial carcinomas. Int J Cancer 1997;71(4):585e594. [14] Liakou CI, Narayanan S, Ng Tang D, Logothetis CJ, Sharma P. Focus on TILs: prognostic significance of tumor infiltrating lymphocytes in human bladder cancer. Cancer Immun 2007;7:10. [15] Zitvogel L, Galluzzi L, Smyth MJ, Kroemer G. Mechanism of action of conventional and targeted anticancer therapies: reinstating immunosurveillance. Immunity 2013;39(1):74e88. [16] Martins I, Kepp O, Schlemmer F, et al. Restoration of the immunogenicity of cisplatin-induced cancer cell death by endoplasmic reticulum stress. Oncogene 2011;30(10):1147e1158. [17] Joseph N, Dovedi SJ, Thompson C, et al. Pre-treatment lymphocytopaenia is an adverse prognostic biomarker in muscleinvasive and advanced bladder cancer. Ann Oncol 2016;27:294e299.