Impact of contemporary patterns of chemotherapy utilization on survival in patients with advanced cancer of the urinary tract: a Retrospective International Study of Invasive/Advanced Cancer of the Urothelium (RISC)

Impact of contemporary patterns of chemotherapy utilization on survival in patients with advanced... Abstract Background Cisplatin-based combination chemotherapy is the standard treatment of advanced urinary tract cancer (aUTC), but 50% of patients are ineligible for cisplatin according to recently published criteria. We used a multinational database to study patterns of chemotherapy utilization in patients with aUTC and determine their impact on survival. Patients and methods This was a retrospective study of patients with: UTC (bladder, renal pelvis, ureter or urethra); advanced disease (stages T4b and/or N+ and/or M+); urothelial, squamous or adenocarcinoma histology. Primary objective was overall survival (OS). Eligibility-for-cisplatin was defined by Eastern Cooperative Oncology Group performance status ≤ 1, creatinine clearance ≥ 60 ml/min, no hearing loss, no neuropathy and no heart failure. Cox regression multivariate analyses were used to establish independent associations of cisplatin versus noncisplatin-based chemotherapy on OS. Results About 1794 patients treated between 2000 and 2013 at 29 centers were analyzed. Median follow-up was 29.1 months. About 1333 patients (74%) received first-line chemotherapy: the use of first-line chemotherapy was associated with longer OS: [hazard ratio (HR): 1.91, 95% confidence interval (CI): 1.67–2.20]. Type of first-line chemotherapy received was: cisplatin-based 669 (50%), carboplatin-based 399 (30%) and other 265 (20%). Cisplatin use was an independent favorable prognostic factor (HR: 1.54, 95% CI: 1.35–1.77). This benefit was independent of baseline characteristics or comorbidities but was associated with eligibility-for-cisplatin: eligible patients treated with cisplatin lived longer than those who were not (HR: 1.74, 95% CI: 1.36–2.21), while such benefit was not observed among ineligible patients. About 26% of patients who did not receive cisplatin were eligible for this agent. Median OS of ineligible patients was poor irrespective of the chemotherapy used. Conclusions The importance of applying published criteria of eligibility-for-cisplatin was confirmed in a multinational, real-world setting in aUTC. The reasons for deviations from these criteria set targets to improve adherence. Effective therapies for cisplatin-ineligible patients are needed. urothelial cancer, chemotherapy, cisplatin eligibility Key Message Cisplatin-based chemotherapy offers a survival advantage over non-cisplatin chemotherapy in patients with advanced urinary-tract-cancer who are eligible-for-cisplatin, according to recently established criteria. Adherence to eligibility-for-cisplatin criteria can be improved in everyday practice. Effective therapies for ineligible-for-cisplatin patients should be sought. Introduction Cisplatin-based chemotherapy is the treatment of choice for advanced urinary tract cancer (aUTC). Nevertheless, a (yet not precisely estimated) percentage of patients do not receive any systemic therapy [1]. Furthermore, about 50% of treated patients do not receive cisplatin [2] mainly due to concerns of increased toxicity and/or limited efficacy in certain patient groups. Instead, carboplatin is a preferred agent in the community due to its more favorable toxicity profile and ease of administration [3–8]. The absence of level I evidence establishing the superiority of cisplatin over carboplatin and lack of well-established criteria defining patients likely to benefit from cisplatin instead of carboplatin-based therapy may encourage this tendency. Recently, specific criteria for eligibility-for-cisplatin have been proposed by Galsky et al. [9]. Although these criteria are widely accepted, their impact has not yet been formally assessed. In addition, adherence to these criteria outside the context of clinical trials has not been studied. In a recent report, 25% of patients with aUTC, treated at 10 Greek Oncology centers, did not receive cisplatin-based chemotherapy, although they were eligible [10]. The effect of deviations from these criteria on patients’ outcomes remains unknown. We attempted to map the different practice patterns in aUTC and the reasons behind those practices as well as their impact on patient outcomes. We used the Retrospective International Study of Cancers of the Urothelium (RISC) database, the largest multinational database of UTC patients worldwide [11–13]. This database includes data from hospitals in the United States, Europe, Israel and Canada, thus ensuring an adequately wide representation of contemporary trends in the treatment of UTC. Patients and methods Patients Data were collected using a web-based electronic data capture tool. Data fields were centrally reviewed and queries were completed by each participating site. For the current study, selection criteria were: histopathologically confirmed UTC (defined as primary carcinoma of the urinary bladder, renal pelvis, ureter or urethra); advanced disease (defined as metastatic or nonmetastatic inoperable disease, i.e. clinical stages T4bN0M0, TanyN + M0 and TanyNanyM+); histological subtype of pure or mixed urothelial carcinoma, pure squamous or pure adenocarcinoma; survival data available. Administration of chemotherapy for radiosensitization only was not considered as first-line chemotherapy. Database was locked in October 2015. The study was approved by the ethics committees at each participating institution. Statistical analysis This study aimed to address the impact of the type of management on the outcome of patients with aUTC. Overall survival (OS) was the primary variable, while the study of types of chemotherapy used and the impact of eligibility-for-cisplatin on chemotherapy utilization were the secondary end points. Eligibility-for-cisplatin was defined according to the criteria of Galsky et al. [9]: Eastern Cooperative Oncology Group performance status (ECOG-PS) 0 or 1, calculated (Cockroft–Gault formula) creatinine clearance (CrCl) ≥ 60 ml/min, no hearing loss, no heart failure and no significant neuropathy. Definitions of groups regarding treating center and disease extent are included in supplementary Appendix S1, available at Annals of Oncology online. OS was defined as the time between the diagnosis of advanced disease and the date of death from any cause. Alive patients were censored at the date of last contact. Survival curves were estimated using the Kaplan–Meier method. Factors studied for their prognostic ability are listed in supplementary Table S1, available at Annals of Oncology online. To account for missing values (supplementary Table S2, available at Annals of Oncology online), multiple imputations were employed using the Markov Chain Monte Carlo method for Arbitrary Missing Data (number of imputations was 35). Average treatment effect (ATE) was calculated using regression adjustment; averages of treatment-specific predicted mean survival times to estimate mean survival times for each potential outcome. Further details on statistical methodology can be found in supplementary Appendix S1, available at Annals of Oncology online. Results Population included in the analysis The flow chart of our analyses is depicted in Figure 1. From 3024 patients included in the RISC database, 1794 patients with advanced UTC, managed at 29 centers between 2000 and 2013, fulfilled selection criteria (Table 1). Their median time from diagnosis of UTC to diagnosis of advanced disease is 0.41 years (range 0–20.4). About 1333 patients received first-line chemotherapy and had information about the type of chemotherapy and were included in the survival analyses according to type of chemotherapy. Objective tumor responses to first-line chemotherapy were reported in 41% of the 1039 assessable patients. Complete responses were significantly more frequent among cisplatin-treated patients (12% versus 5%, P < 0.001). Table 1. Baseline characteristics of patients included in the analyses Characteristic  Received first-line chemotherapy   P  All patients (n = 1794)  Yes  No    Age (median, range)a  67.6, 34–94.2   70.7, 20.5–93.3   <0.001      n (%)    Gendera      0.003   M  1060 (78.8)  318 (72)   F  285 (21.2)  124 (28)  Charlson Comorbidity Indexa      <0.001   0  603 (47.3)  121 (31.6)   1–2  370 (29)  145 (37.9)   >2  301 (23.7)  117 (30.6)  Histologya      <.001   Transitional  1174 (87.6)  337 (76.2)   Mixed  104 (7.8)  54 (12.2)   Other  62 (4.6)  51 (11.5)  Metastatic sites         Locoregional  265 (19.6)  111 (25.1)  0.015   Lymph nodes  863 (63.9)  202 (45.6)  <0.001   Bone  367 (27.2)  123 (27.8)  0.806   Lung  371 (27.5)  117 (26.4)  0.666   Liver  250 (18.5)  69 (15.6)  0.162   Brain  24 (1.8)  15 (3.4)  0.044   Adrenal  41 (3)  5 (1.1)  0.028   Peritoneum  62 (4.6)  31 (7)  0.047   Distantb  808 (59.8)  270 (61)  0.671  Center      <0.001   Low-volume (n ≤ 100)  570 (42.2)  248 (56)   High-volume (n > 100)  781 (57.8)  195 (44)  Racea      <0.001   Caucasian  1246 (92.6)  377 (85.7)   Other  99 (7.4)  63 (14.3)  Ethnicitya      0.274   Hispanic/Latino  120 (9.3)  32 (7.6)   Other  1169 (90.7)  391 (92.4)  Neoadjuvant chemotherapya      <0.001   Yes  166 (12.9)  104 (24.5)   No  1121 (87.1)  321 (75.5)  Adjuvant therapya      <0.001   Yes  182 (14.3)  97 (22.8)   No  1092 (85.7)  329 (77.2)  Surgical removal of primary tumora      <0.001   Yes  732 (55.3)  312 (71.4)   No  592 (44.7)  125 (28.6)  Radiation therapy to primary tumora      <0.001   Yes  119 (9.4)  87 (20.5)   No  1145 (90.6)  338 (79.5)  Primary sitea      <0.001   Bladder  1085 (81.8)  402 (93.3)   Other  241 (18.2)  29 (6.7)    Characteristic  Received first-line chemotherapy   P  All patients (n = 1794)  Yes  No    Age (median, range)a  67.6, 34–94.2   70.7, 20.5–93.3   <0.001      n (%)    Gendera      0.003   M  1060 (78.8)  318 (72)   F  285 (21.2)  124 (28)  Charlson Comorbidity Indexa      <0.001   0  603 (47.3)  121 (31.6)   1–2  370 (29)  145 (37.9)   >2  301 (23.7)  117 (30.6)  Histologya      <.001   Transitional  1174 (87.6)  337 (76.2)   Mixed  104 (7.8)  54 (12.2)   Other  62 (4.6)  51 (11.5)  Metastatic sites         Locoregional  265 (19.6)  111 (25.1)  0.015   Lymph nodes  863 (63.9)  202 (45.6)  <0.001   Bone  367 (27.2)  123 (27.8)  0.806   Lung  371 (27.5)  117 (26.4)  0.666   Liver  250 (18.5)  69 (15.6)  0.162   Brain  24 (1.8)  15 (3.4)  0.044   Adrenal  41 (3)  5 (1.1)  0.028   Peritoneum  62 (4.6)  31 (7)  0.047   Distantb  808 (59.8)  270 (61)  0.671  Center      <0.001   Low-volume (n ≤ 100)  570 (42.2)  248 (56)   High-volume (n > 100)  781 (57.8)  195 (44)  Racea      <0.001   Caucasian  1246 (92.6)  377 (85.7)   Other  99 (7.4)  63 (14.3)  Ethnicitya      0.274   Hispanic/Latino  120 (9.3)  32 (7.6)   Other  1169 (90.7)  391 (92.4)  Neoadjuvant chemotherapya      <0.001   Yes  166 (12.9)  104 (24.5)   No  1121 (87.1)  321 (75.5)  Adjuvant therapya      <0.001   Yes  182 (14.3)  97 (22.8)   No  1092 (85.7)  329 (77.2)  Surgical removal of primary tumora      <0.001   Yes  732 (55.3)  312 (71.4)   No  592 (44.7)  125 (28.6)  Radiation therapy to primary tumora      <0.001   Yes  119 (9.4)  87 (20.5)   No  1145 (90.6)  338 (79.5)  Primary sitea      <0.001   Bladder  1085 (81.8)  402 (93.3)   Other  241 (18.2)  29 (6.7)    Patients who received first-line chemotherapy (n = 1333)  Characteristic  Cisplatin  Carboplatin  Other  P  (n = 669)  (n = 399)  (n = 265)  Age (median, range)a  65.1, 34–87  71, 38.8–92.9  68.6, 37.5–92  <0.001  Creatinine clearance (median, range)a  71.9, 20.2–242.5  50.9, 1.6–334.6  54.2, 8.6–164.8  <0.001     n (%)     Gendera        0.037   M  542 (81.6)  302 (75.7)  201 (76.1)   F  122 (18.4)  97 (24.3)  63 (23.9)  Charlson Comorbidity Indexa        <0.001   0  354 (55.1)  161 (43.2)  85 (35.4)   1–2  183 (28.5)  94 (25.2)  88 (36.7)   >2  106 (16.4)  118 (31.6)  67 (27.9)  Histologya        0.270   Transitional  591 (89)  348 (88.3)  221 (83.7)   Mixed  45 (6.8)  29 (7.43)  26 (9.9)   Other  28 (4.2)  17 (4.3)  17 (6.4)  Metastatic sites           Locoregional  132 (19.7)  80 (20.1)  50 (18.9)  0.930   Lymphnodes  440 (65.8)  258 (64.7)  159 (60)  0.248   Bone  155 (23.2)  128 (32.1)  78 (29.4)  0.004   Lung  183 (27.4)  105 (26.3)  79 (29.8)  0.608   Liver  128 (19.1)  76 (19.1)  44 (16.6)  0.645   Brain  14 (2.1)  7 (1.8)  3 (1.1)  0.607   Adrenal  17 (2.5)  15 (3.8)  9 (3.4)  0.507   Peritoneum  19 (2.8)  21 (5.3)  20 (7.6)  0.005   Distantb  374 (55.9)  247 (61.9)  175 (66)  0.010  Center        0.560   Low volume (n ≤ 100)  287 (42.9)  163 (40.9)  104 (39.3)   High volume (n > 100)  382 (57.1)  236 (59.1)  161 (60.8)  Racea        0.490   Caucasian  618 (93.1)  364 (91.2)  246 (93.2)   Other  46 (6.9)  35 (8.8)  18 (6.8)  Ethnicitya        0.049   Hispanic/Latino  57 (9.1)  29 (7.5)  34 (13.2)   Other  568 (90.9)  359 (92.5)  224 (86.8)  Neoadjuvant chemotherapya        <0.001   Yes  45 (7.1)  42 (11.3)  79 (30.5)   No  593 (92.9)  331 (88.7)  180 (69.5)  Adjuvant therapya        <0.001   Yes  48 (7.6)  48 (13)  82 (31.9)   No  582 (92.4)  321 (87)  175 (68.1)  Surgical removal of primary tumora        <0.001   Yes  334 (50.8)  205 (53.2)  180 (68.2)   No  323 (49.2)  180 (46.8)  84 (31.8)  Radiation therapy to primary tumora        0.011   Yes  44 (7)  41 (11.3)  32 (12.7)   No  587 (93)  323 (88.7)  220 (87.3)  Post chemotherapy surgerya        0.019   Yes  45 (7.3)  11 (3.1)  12 (4.9)   No  570 (92.7)  346 (96.9)  232 (95.1)  Primary sitea        0.533   Bladder  544 (82.7)  310 (79.9)  214 (81.7)   Other  114 (17.3)  78 (20.1)  48 (18.3)    ECOG-PSa        <0.001   0  232 (43.1)  65 (20.9)  49 (24.8)   1  256 (47.4)  154 (49.5)  107 (54)   2  44 (8.2)  75 (24.1)  35 (17.7)   3  6 (1.1)  14 (4.5)  7 (3.5)   4  1 (0.2)  3 (1)  0 (0)  Risk stratificationa,c        <0.001   Low  228 (42.3)  87 (28)  51 (25.7)   Intermediate  275 (51)  159 (51.1)  117 (59.1)   High  36 (6.7)  65 (20.9)  30 (15.2)  Eligibility for cisplatina,d        <0.001   Yes  275 (61)  61 (21.4)  63 (32.6)   No  176 (39)  224 (78.6)  130 (67.4)  Cycles of chemotherapy (median, range)a  5 (1–35)  4 (0–21)  4 (1–26)  <0.001  Patients who received first-line chemotherapy (n = 1333)  Characteristic  Cisplatin  Carboplatin  Other  P  (n = 669)  (n = 399)  (n = 265)  Age (median, range)a  65.1, 34–87  71, 38.8–92.9  68.6, 37.5–92  <0.001  Creatinine clearance (median, range)a  71.9, 20.2–242.5  50.9, 1.6–334.6  54.2, 8.6–164.8  <0.001     n (%)     Gendera        0.037   M  542 (81.6)  302 (75.7)  201 (76.1)   F  122 (18.4)  97 (24.3)  63 (23.9)  Charlson Comorbidity Indexa        <0.001   0  354 (55.1)  161 (43.2)  85 (35.4)   1–2  183 (28.5)  94 (25.2)  88 (36.7)   >2  106 (16.4)  118 (31.6)  67 (27.9)  Histologya        0.270   Transitional  591 (89)  348 (88.3)  221 (83.7)   Mixed  45 (6.8)  29 (7.43)  26 (9.9)   Other  28 (4.2)  17 (4.3)  17 (6.4)  Metastatic sites           Locoregional  132 (19.7)  80 (20.1)  50 (18.9)  0.930   Lymphnodes  440 (65.8)  258 (64.7)  159 (60)  0.248   Bone  155 (23.2)  128 (32.1)  78 (29.4)  0.004   Lung  183 (27.4)  105 (26.3)  79 (29.8)  0.608   Liver  128 (19.1)  76 (19.1)  44 (16.6)  0.645   Brain  14 (2.1)  7 (1.8)  3 (1.1)  0.607   Adrenal  17 (2.5)  15 (3.8)  9 (3.4)  0.507   Peritoneum  19 (2.8)  21 (5.3)  20 (7.6)  0.005   Distantb  374 (55.9)  247 (61.9)  175 (66)  0.010  Center        0.560   Low volume (n ≤ 100)  287 (42.9)  163 (40.9)  104 (39.3)   High volume (n > 100)  382 (57.1)  236 (59.1)  161 (60.8)  Racea        0.490   Caucasian  618 (93.1)  364 (91.2)  246 (93.2)   Other  46 (6.9)  35 (8.8)  18 (6.8)  Ethnicitya        0.049   Hispanic/Latino  57 (9.1)  29 (7.5)  34 (13.2)   Other  568 (90.9)  359 (92.5)  224 (86.8)  Neoadjuvant chemotherapya        <0.001   Yes  45 (7.1)  42 (11.3)  79 (30.5)   No  593 (92.9)  331 (88.7)  180 (69.5)  Adjuvant therapya        <0.001   Yes  48 (7.6)  48 (13)  82 (31.9)   No  582 (92.4)  321 (87)  175 (68.1)  Surgical removal of primary tumora        <0.001   Yes  334 (50.8)  205 (53.2)  180 (68.2)   No  323 (49.2)  180 (46.8)  84 (31.8)  Radiation therapy to primary tumora        0.011   Yes  44 (7)  41 (11.3)  32 (12.7)   No  587 (93)  323 (88.7)  220 (87.3)  Post chemotherapy surgerya        0.019   Yes  45 (7.3)  11 (3.1)  12 (4.9)   No  570 (92.7)  346 (96.9)  232 (95.1)  Primary sitea        0.533   Bladder  544 (82.7)  310 (79.9)  214 (81.7)   Other  114 (17.3)  78 (20.1)  48 (18.3)    ECOG-PSa        <0.001   0  232 (43.1)  65 (20.9)  49 (24.8)   1  256 (47.4)  154 (49.5)  107 (54)   2  44 (8.2)  75 (24.1)  35 (17.7)   3  6 (1.1)  14 (4.5)  7 (3.5)   4  1 (0.2)  3 (1)  0 (0)  Risk stratificationa,c        <0.001   Low  228 (42.3)  87 (28)  51 (25.7)   Intermediate  275 (51)  159 (51.1)  117 (59.1)   High  36 (6.7)  65 (20.9)  30 (15.2)  Eligibility for cisplatina,d        <0.001   Yes  275 (61)  61 (21.4)  63 (32.6)   No  176 (39)  224 (78.6)  130 (67.4)  Cycles of chemotherapy (median, range)a  5 (1–35)  4 (0–21)  4 (1–26)  <0.001  a Missing values apply and are listed in supplementary Table S2, available at Annals of Oncology online. b Distant metastases: metastases outside primary site and lymph nodes. c Risk stratification according to Bajorin et al. [8]. d Eligibility was defined according to Galsky et al. [9]. SD, standard deviation. Figure 1. View largeDownload slide Study flow chart. Figure 1. View largeDownload slide Study flow chart. Treatment patterns Patients who did not receive first-line chemotherapy were older, more frequently noncaucasians, had higher frequency of nonurothelial histology, higher Charlson Comorbidity Index (CCI), received more frequently perioperative chemotherapy and management of primary tumor, and were predominantly treated at low volume centers (Table 1). Cisplatin-based first-line chemotherapy was administered to 669 patients, carboplatin based to 399 and other therapies to 265. Detailed description of chemotherapy administered is presented in supplementary Table S3, available at Annals of Oncology online. Cisplatin administration was associated with younger age, male gender, less frequent perioperative chemotherapy, better PS and renal function, and less frequent distant metastases. Cisplatin-treated patients also had a lower CCI: specifically, hypertension, diabetes, chronic obstructive pulmonary disease (COPD), coronary artery disease, congestive heart failure (CHF), hyperlipidemia and cerebrovascular disease were less frequent among cisplatin-treated patients (supplementary Table S4, available at Annals of Oncology online). Eligibility-for-cisplatin Adequate data to assess eligibility-for-cisplatin according to Galsky’s criteria [9] were available for 929 patients: 429 patients (46%) had at least one criterion for ineligibility (supplementary Figure S1, available at Annals of Oncology online). The most common criterion was CrCl < 60 ml/min (97.4%) followed by PS ≥ 2 (43.1%). The reason for not using cisplatin was provided by the investigators in 340 cases (supplementary Table S5, available at Annals of Oncology online). Established ineligibility criteria accounted for only 56% of these cases. Advanced age and comorbidities were the most frequent reasons for not administering cisplatin to patients who were otherwise eligible. Since reasons for not administering cisplatin were not reported in the majority of patients, we carried out analyses on all 929 patients who were assessable for cisplatin-eligibility to more accurately examine deviations from these criteria. One hundred twenty-four patients (26%) who did not receive cisplatin were eligible for this agent (Table 1). This deviation from Galsky’s criteria was more frequent in patients with: CCI ≥ 3 (49% versus 25%, P < 0.001), Memorial Sloan Kettering Cancer Center (MSKCC) intermediate/high risk (36% versus 24%, P = 0.012), hypertension (47% versus 23%, P < 0.001), diabetes (56% versus 28%, P < 0.001) and COPD (60% versus 29%, P = 0.001). Interestingly, 176 patients (39%) who received cisplatin fulfilled at least one criterion for ineligibility: CHF (n = 9), ECOG-PS > 1 (n = 51), CrCl < 60 ml/min (n = 129). Survival analyses Overall population Among the 1794 selected patients, 638 (36%) were alive, 1004 patients died of disease progression, 10 died from treatment-related complications, 41 from other causes and 101 from unknown reason, at the time of analysis. The median follow-up was 29.1 months [95% confidence interval (CI) 26.4–31.4]. Patients who received systemic chemotherapy had significantly longer median OS (Figure 2). Multivariate analysis confirmed the independent significance of chemotherapy administration (supplementary Table S7, available at Annals of Oncology online). This result remained essentially unchanged when patients who received perioperative chemotherapy were excluded. Nevertheless, these analyses should be viewed with caution, since PS was not included in the cox regression model due to high number of missing values among nontreated patients. Figure 2. View largeDownload slide Kaplan–Meier curves of overall survival (OS) according to the type of first-line chemotherapy (A); Forest plot of subgroup analysis according to the type of first-line chemotherapy (B); Kaplan–Meier curves of OS according to: the type of first-line chemotherapy stratified by eligibility-for-cisplatin (C). Maximum analysis time to be graphed was selected at 6 years. Patients alive at 6 years (regardless of subsequent death) were censored at that time. Blue dots in forest plot denote hazard ratios for the treatment type and horizontal bars 95% confidence interval. Interaction refers to potential interactions of the effects of treatment type (cisplatin versus carboplatin + other) and each of the covariates listed on the left-hand side of the figure. P-value of interaction refers to the P-value of the interaction term from a Cox model, which includes treatment type and the respective covariate. CCI, Charlson Comorbidity Index; COPD, chronic obstructive airway disease; PS, performance status; Hgb, hemoglobin; WBC, while blood cells. Figure 2. View largeDownload slide Kaplan–Meier curves of overall survival (OS) according to the type of first-line chemotherapy (A); Forest plot of subgroup analysis according to the type of first-line chemotherapy (B); Kaplan–Meier curves of OS according to: the type of first-line chemotherapy stratified by eligibility-for-cisplatin (C). Maximum analysis time to be graphed was selected at 6 years. Patients alive at 6 years (regardless of subsequent death) were censored at that time. Blue dots in forest plot denote hazard ratios for the treatment type and horizontal bars 95% confidence interval. Interaction refers to potential interactions of the effects of treatment type (cisplatin versus carboplatin + other) and each of the covariates listed on the left-hand side of the figure. P-value of interaction refers to the P-value of the interaction term from a Cox model, which includes treatment type and the respective covariate. CCI, Charlson Comorbidity Index; COPD, chronic obstructive airway disease; PS, performance status; Hgb, hemoglobin; WBC, while blood cells. Patients who received first-line chemotherapy Among the 1333 patients, 872 were dead at the time of analysis (disease: 782, toxicity: 9, other causes: 24, unknown reason: 57). The median follow-up time was 31.7 months. Cisplatin-based chemotherapy was associated with significantly longer median OS, (Table 2 and Figure 2A). Univariate analysis showed that bone, liver, lung and distant metastases, smoking history, CCI ≥ 3, ECOG-PS (0 versus 1 + 2 and 0 + 1 versus 2), white blood cell (WBC) > 10 000/mm3, plt > 450 000/mm3, hemoglobin ≤ institutional lower limit, albumin ≤ median, age > median and increasing MSKCC score were adverse prognostic factors, while the presence of lymph node metastases and BMI > median was favorable prognostic factors. BMI, WBC, plt, hemoglobin and albumin, studied as continuous variables, were also significantly associated with prognosis. Origin of the center, histology and taxane use were not associated with OS. When significant factors were entered into a multivariate cox regression model including therapy group (cisplatin versus noncisplatin), the latter retained its significance (Table 3). ATE analysis showed that death would occur 6.9 months earlier if no patient received cisplatin-based chemotherapy, compared with the outcome if everybody received such treatment (supplementary Table S6, available at Annals of Oncology online). The independent prognostic significance of treatment type was also retained when multiple imputations for missing values were applied, after stratification for eligibility-for-cisplatin (supplementary Table S8, available at Annals of Oncology online). Since other therapies or different histological subtypes may have confounded our results, we also carried out multivariate analyses excluding patients who had received perioperative chemotherapy, postchemotherapy surgery or had nontransitional histology. The independent prognostic significance of cisplatin-based chemotherapy remained unchanged. Table 2. Unadjusted hazard ratios and 95% CI for death according to the use of first-line chemotherapy   n  Median OS (95% CI)  HR (95% CI)  P  All patients (n = 1794)    Received chemotherapy        <0.001   Yes  1351  14.8 (13.8–15.6)  1     No  443  5.7 (4.6–6.8)  1.91 (1.67–2.2)    Patients who received first-line chemotherapy (n = 1333)    Therapy group        <0.001   Cisplatin  669  18 (16.1–19.8)  1     Carboplatin  399  12.5 (10.9–13.4)  1.58 (1.36–1.83)     Other  265  12.6 (9.9–15.4)  1.5 (1.25–1.79)    Eligible-for-cisplatin        <0.001   Cisplatin treated  275  20.3 (17.1–22.7)  1     Carboplatin treated  61  12.9 (8.7–15.7)  2.3 (1.67–3.17)     Other  63  16.7 (11.8–21)  1.5 (1.04–2.1)    Ineligible-for-cisplatin  176  12.7 (11.2–15.1)  1  0.083   Cisplatin treated  224  11.6 (10.5–13.4)  1.12 (0.88–1.43)     Carboplatin treated  130  8.7 (6.9–10.1)  1.37 (1.05–1.8)    Other      n  Median OS (95% CI)  HR (95% CI)  P  All patients (n = 1794)    Received chemotherapy        <0.001   Yes  1351  14.8 (13.8–15.6)  1     No  443  5.7 (4.6–6.8)  1.91 (1.67–2.2)    Patients who received first-line chemotherapy (n = 1333)    Therapy group        <0.001   Cisplatin  669  18 (16.1–19.8)  1     Carboplatin  399  12.5 (10.9–13.4)  1.58 (1.36–1.83)     Other  265  12.6 (9.9–15.4)  1.5 (1.25–1.79)    Eligible-for-cisplatin        <0.001   Cisplatin treated  275  20.3 (17.1–22.7)  1     Carboplatin treated  61  12.9 (8.7–15.7)  2.3 (1.67–3.17)     Other  63  16.7 (11.8–21)  1.5 (1.04–2.1)    Ineligible-for-cisplatin  176  12.7 (11.2–15.1)  1  0.083   Cisplatin treated  224  11.6 (10.5–13.4)  1.12 (0.88–1.43)     Carboplatin treated  130  8.7 (6.9–10.1)  1.37 (1.05–1.8)    Other    OS, overall survival; HR, hazard ratio; CI, confidence interval. Table 3. Multivariate analysis for OS in patients who received first-line chemotherapy Factor  Subgroups  HR  95% CI  P-value  Distant metastases and PS in regression model (n = 727)  WBC  ≤10 000  1    <0.001  >10 000  1.49  1.23–1.8  Hb  ≤ ILL  1    0.001  > ILL  1.38  1.13–1.68  Distant metastases  No  1    0.002  Yes  1.33  1.11–1.6  Treatment  Cisplatin  1    0.003  Non Cisplatin  1.32  1.1–1.58  ECOG-PS  0  1    0.007  >0  1.32  1.08–1.6  MSKCC risk groups in Cox regression model (n = 727)  MSKCC risk  Low  1    <0.001  Intermediate  1.37  1.12–1.66  High  2.17  1.6–2.94  WBC  ≤10 000  1    <0.001  >10 000  1.42  1.17–1.73  Hb  ≥ILL  1    0.002  <ILL  1.36  1.12–1.66  Treatment  Cisplatin  1    0.007  Non Cisplatin  1.29  1.07–1.55  Factor  Subgroups  HR  95% CI  P-value  Distant metastases and PS in regression model (n = 727)  WBC  ≤10 000  1    <0.001  >10 000  1.49  1.23–1.8  Hb  ≤ ILL  1    0.001  > ILL  1.38  1.13–1.68  Distant metastases  No  1    0.002  Yes  1.33  1.11–1.6  Treatment  Cisplatin  1    0.003  Non Cisplatin  1.32  1.1–1.58  ECOG-PS  0  1    0.007  >0  1.32  1.08–1.6  MSKCC risk groups in Cox regression model (n = 727)  MSKCC risk  Low  1    <0.001  Intermediate  1.37  1.12–1.66  High  2.17  1.6–2.94  WBC  ≤10 000  1    <0.001  >10 000  1.42  1.17–1.73  Hb  ≥ILL  1    0.002  <ILL  1.36  1.12–1.66  Treatment  Cisplatin  1    0.007  Non Cisplatin  1.29  1.07–1.55  WBC, white blood cells; ILL, institutional lower limit; ECOG-PS, Eastern Cooperative Oncology Group performance status; MSKCC, Memorial Sloan Kettering Cancer Center. Subgroup analyses Benefit from cisplatin-based therapy was obtained regardless of the administration of perioperative chemotherapy and across all prognostic groups or comorbidities (Figure 2B) with the exception of the eligibility-for-cisplatin groups. Cisplatin-based chemotherapy was associated with a survival benefit, only among eligible patients (Figure 2C and Table 2). The interaction test between treatment type and eligibility-for-cisplatin yielded a significant result (P = 0.024). Among ineligible patients, carboplatin-based therapy was associated with longer median OS compared with nonplatinum chemotherapy (Table 2). Discussion This is the first study to describe international practice patterns and their effect on the outcome of patients with aUTC. The study has certain limitations associated with its retrospective nature. Inaccuracies in reporting cannot be completely excluded, in spite of the carried out quality control of the data. Our analyses are also subject to selection biases. Although we attempted to mitigate them by using multivariate analyses, residual unmeasured confounding factors may have impacted our results. In spite of these limitations, our data are original and potentially useful for clinical practice and future research. Apart from the depiction of practice patterns across countries with different health and insurance systems, the RISC database enabled the study of so far unexplored subjects, such as the comparison between cisplatin and carboplatin, as well as the value of published criteria of eligibility-for-cisplatin and the impact of deviations from these criteria on outcome. Given that a randomized study, directly comparing cisplatin with carboplatin is unlikely, this analysis provides the highest possible level of evidence to this question, while it also provides useful benchmarks of the outcome of the various subgroups of patients with aUTC. Patients who received chemotherapy lived significantly longer than those who did not. This OS benefit was confirmed in a multivariate analysis, which, however, is limited by the lack of PS data for untreated patients. Although the reasons for not offering chemotherapy are unclear, our analysis suggests that a combination of frailty as well as experience of the center in the treatment of aUTC may be the most relevant underlying causes. Nevertheless, other reasons, due to variations in referral patterns, which cannot be addressed by this study, cannot be excluded. Cisplatin administration was associated with OS benefit. An inherent bias exists in this analysis due to distinct patient profiles of better PS and fewer distant metastases in the cisplatin group. To overcome these imbalances, we used multivariate analyses, which showed that this benefit was independent of other prognostic factors. The OS benefit was restricted to the eligible-for-cisplatin population, confirming the validity of this feature as a selection criterion for cisplatin-based chemotherapy. Eligible patients treated with cisplatin had a median OS of 20.3 months, which compares favorably with those reported in randomized trials [14–18] and suggests that this therapy will remain a valuable option in spite of the important emerging changes in the treatment paradigm in aUTC. In contrast, eligible patients not treated with cisplatin had a significantly shorter median OS. This result underlines the importance of adhering to Galsky’s criteria and promptly identifying and correcting potentially reversible causes of ineligibility, such as recently established renal impairment due to urinary tract obstruction by primary tumor, in order to ensure optimal outcome. We found that 26% of patients not receiving cisplatin were eligible for this agent. This deviation was mainly due to advanced age and comorbidities, especially hypertension, diabetes and COPD. This practice may not be justified as our analysis showed that patients with these comorbidities derived the same benefit from cisplatin administration as their counterparts. These findings set clear targets to improve our management of advanced UTC. Strong emphasis on the eligibility criteria should be considered by Urology and Oncology Scientific and Professional Societies as well as academic institutions and health care providers at national and international level. Referral to centers with experience in treating UTC will improve utilization of cisplatin-based chemotherapy. It should be noted that patients included in this analysis were treated before the publication of Galsky et al. [9]. It could be speculated that adherence to eligibility criteria has been improved since. Nevertheless, recent observational studies suggested that considerable deviations still exist in everyday practice [10]. Interestingly, perioperative chemotherapy was associated with less frequent use of first-line chemotherapy as well as of cisplatin-based chemotherapy in otherwise eligible patients. This suggests that prior exposure to systemic chemotherapy may represent a dissuasive factor for offering similar therapy at progression. Although our results suggested similar benefit by cisplatin regardless of previous perioperative therapy, further analyses, i.e. according to the time from perioperative chemotherapy to progression or the type of chemotherapy, were not possible due to the small number of patients receiving perioperative therapy. Further study is warranted to accurately evaluate the potential of this factor as an additional criterion for cisplatin-ineligibility, especially in the context of the development of novel effective options in first-line therapy of aUTC. Fifty-seven percent of patients were ineligible-for-cisplatin. The distribution of criteria of ineligibility is similar to those of a national retrospective study [10] as well as of a recently reported phase II study [19]. This similarity supports the validity of these criteria and their applicability in the context of clinical trials. Optimal therapy for cisplatin-ineligible patients suffering from aUTC remains an unmet medical need. Importantly, cisplatin was used in 176 ineligible patients but this was not associated with an equally favorable effect as among eligible patients, which implies that cisplatin-ineligibility may be a surrogate for a more aggressive tumor behavior and not merely for unfitness for this particular agent. Currently, carboplatin-based chemotherapy is considered the standard for ineligible-for-cisplatin patients [20]. Our results support this recommendation, since carboplatin-based chemotherapy produced a longer median OS than nonplatinum therapy. Nevertheless, the median OS of 11.6 months emphasizes the need for developing more effective choices for this population. In this context, the median OS of 15.9 months recently reported for the programmed death ligand 1 inhibitor atezolizumab in first-line treatment of ineligible-for-cisplatin patients appears particularly promising [14]. Conclusion In conclusion, all efforts should be made to treat eligible-for-cisplatin patients with aUTC with cisplatin-based chemotherapy. The field should recognize that cisplatin may be underutilized and that adherence to published criteria can help increase appropriate cisplatin use. Novel, more effective therapies for ineligible-for-cisplatin patients should be sought through clinical trials specifically focused in this population. Funding None declared. Disclosure The authors have declared no conflicts of interest. References 1 Sonpavde G, Watson D, Tourtellott M et al.   Administration of cisplatin-based chemotherapy for advanced urothelial carcinoma in the community. Clin Genitourin Cancer  2012; 10( 1): 1– 5. Google Scholar CrossRef Search ADS PubMed  2 Dash A, Galsky MD, Vickers AJ et al.   Impact of renal impairment on eligibility for adjuvant cisplatin-based chemotherapy in patients with urothelial carcinoma of the bladder. Cancer  2006; 107( 3): 506– 513. Google Scholar CrossRef Search ADS PubMed  3 Linardou H, Aravantinos G, Efstathiou E et al.   Gemcitabine and carboplatin combination as first-line treatment in elderly patients and those unfit for cisplatin-based chemotherapy with advanced bladder carcinoma: phase II study of the Hellenic Co-operative Oncology Group. Urology  2004; 64( 3): 479– 484. Google Scholar CrossRef Search ADS PubMed  4 Bamias A, Lainakis G, Kastritis E et al.   Biweekly carboplatin/gemcitabine in patients with advanced urothelial cancer who are unfit for cisplatin-based chemotherapy: report of efficacy, quality of life and geriatric assessment. Oncology  2007; 73( 5–6): 290– 297. Google Scholar CrossRef Search ADS PubMed  5 Bellmunt J, de Wit R, Albanell J et al.   A feasibility study of carboplatin with fixed dose of gemcitabine in “unfit” patients with advanced bladder cancer. Eur J Cancer  2001; 37( 17): 2212– 2215. Google Scholar CrossRef Search ADS PubMed  6 Carles J, Nogue M, Domenech M et al.   Carboplatin-gemcitabine treatment of patients with transitional cell carcinoma of the bladder and impaired renal function. Oncology  2000; 59( 1): 24– 27. Google Scholar CrossRef Search ADS PubMed  7 De Santis M, Bellmunt J, Mead G et al.   Randomized phase II/III trial assessing gemcitabine/carboplatin and methotrexate/carboplatin/vinblastine in patients with advanced urothelial cancer “unfit” for cisplatin-based chemotherapy: phase II–results of EORTC study 30986. J Clin Oncol  2009; 27( 33): 5634– 5639. Google Scholar CrossRef Search ADS PubMed  8 Bajorin DF, Dodd PM, Mazumdar M et al.   Long-term survival in metastatic transitional-cell carcinoma and prognostic factors predicting outcome of therapy. J Clin Oncol  1999; 17( 10): 3173– 3181. Google Scholar CrossRef Search ADS PubMed  9 Galsky MD, Hahn NM, Rosenberg J et al.   Treatment of patients with metastatic urothelial cancer “unfit” for Cisplatin-based chemotherapy. J Clin Oncol  2011; 29( 17): 2432– 2438. Google Scholar CrossRef Search ADS PubMed  10 Bamias A, Peroukidis S, Stamatopoulou S et al.   Utilization of Systemic Chemotherapy in Advanced Urothelial Cancer: a Retrospective Collaborative Study by the Hellenic Genitourinary Cancer Group (HGUCG). Clin Genitourin Cancer  2016; 14( 2): e153– e159. Google Scholar CrossRef Search ADS PubMed  11 Necchi A, Sonpavde G, Lo Vullo S et al.   RISC Investigators. Nomogram-based prediction of overall survival in patients with metastatic urothelial carcinoma receiving first-line platinum-based chemotherapy: Retrospective International Study of Invasive/Advanced Cancer of the Urothelium (RISC). Eur Urol  2017; 71( 2): 281– 289. Google Scholar CrossRef Search ADS PubMed  12 Ramos JD, Casey MF, Bamias A et al.   Retrospective International Study of Cancers of the Urothelium (RISC) Investigators. The Khorana Score in predicting venous thromboembolism for patients with metastatic urothelial carcinoma and variant histology treated with chemotherapy. Clin Appl Thromb Hemost  2017; 23( 7): 755– 760. Google Scholar CrossRef Search ADS PubMed  13 Ramos JD, Casey MF, Crabb SJ et al.   RISC Investigators. Venous thromboembolism in metastatic urothelial carcinoma or variant histologies: incidence, associative factors, and effect on survival. Cancer Med  2017; 6( 1): 186– 194. Google Scholar CrossRef Search ADS PubMed  14 Sternberg CN, de Mulder PH, Schornagel JH et al.   Randomized phase III trial of high-dose-intensity methotrexate, vinblastine, doxorubicin, and cisplatin (MVAC) chemotherapy and recombinant human granulocyte colony-stimulating factor versus classic MVAC in advanced urothelial tract tumors: European Organization for Research and Treatment of Cancer Protocol no. 30924. J Clin Oncol  2001; 19: 2638– 2646. Google Scholar CrossRef Search ADS PubMed  15 von der Maase H, Hansen SW, Roberts JT et al.   Gemcitabine and cisplatin versus methotrexate, vinblastine, doxorubicin, and cisplatin in advanced or metastatic bladder cancer: results of a large, randomized, multinational, multicenter, phase III study. J Clin Oncol  2000; 18( 17): 3068– 3077. Google Scholar CrossRef Search ADS PubMed  16 Bamias A, Aravantinos G, Deliveliotis C et al.   Docetaxel and cisplatin with granulocyte colony-stimulating factor (G-CSF) versus MVAC with G-CSF in advanced urothelial carcinoma: a multicenter, randomized, phase III study from the Hellenic Cooperative Oncology Group. J Clin Oncol  2004; 22: 220– 228. Google Scholar CrossRef Search ADS PubMed  17 Bamias A, Dafni U, Karadimou A et al.   Prospective, open-label, randomized, phase III study of two dose-dense regimens MVAC versus gemcitabine/cisplatin in patients with inoperable, metastatic or relapsed urothelial cancer: a Hellenic Cooperative Oncology Group study (HE 16/03). Ann Oncol  2013; 24( 4): 1011– 1017. Google Scholar CrossRef Search ADS PubMed  18 Bellmunt J, von der Maase H, Mead GM et al.   Randomized phase III study comparing paclitaxel/cisplatin/gemcitabine and gemcitabine/cisplatin in patients with locally advanced or metastatic urothelial cancer without prior systemic therapy: EORTC Intergroup Study 30987. J Clin Oncol  2012; 30( 10): 1107– 1113. Google Scholar CrossRef Search ADS PubMed  19 Balar AV, Galsky MD, Rosenberg JE et al.   Atezolizumab as first-line treatment in cisplatin-ineligible patients with locally advanced and metastatic urothelial carcinoma: a single-arm, multicentre, phase 2 trial. Lancet  2017; 389( 10064): 67– 76. Google Scholar CrossRef Search ADS PubMed  20 Bellmunt J, Orsola A, Leow JJ et al.   Bladder cancer: ESMO Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol  2014; 25(Suppl 3): iii40– iii48. Google Scholar CrossRef Search ADS PubMed  © The Author 2017. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For Permissions, please email: journals.permissions@oup.com. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Annals of Oncology Oxford University Press

Impact of contemporary patterns of chemotherapy utilization on survival in patients with advanced cancer of the urinary tract: a Retrospective International Study of Invasive/Advanced Cancer of the Urothelium (RISC)

Loading next page...
 
/lp/ou_press/impact-of-contemporary-patterns-of-chemotherapy-utilization-on-PsP0OfzxEq
Publisher
Oxford University Press
Copyright
© The Author 2017. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For Permissions, please email: journals.permissions@oup.com.
ISSN
0923-7534
eISSN
1569-8041
D.O.I.
10.1093/annonc/mdx692
Publisher site
See Article on Publisher Site

Abstract

Abstract Background Cisplatin-based combination chemotherapy is the standard treatment of advanced urinary tract cancer (aUTC), but 50% of patients are ineligible for cisplatin according to recently published criteria. We used a multinational database to study patterns of chemotherapy utilization in patients with aUTC and determine their impact on survival. Patients and methods This was a retrospective study of patients with: UTC (bladder, renal pelvis, ureter or urethra); advanced disease (stages T4b and/or N+ and/or M+); urothelial, squamous or adenocarcinoma histology. Primary objective was overall survival (OS). Eligibility-for-cisplatin was defined by Eastern Cooperative Oncology Group performance status ≤ 1, creatinine clearance ≥ 60 ml/min, no hearing loss, no neuropathy and no heart failure. Cox regression multivariate analyses were used to establish independent associations of cisplatin versus noncisplatin-based chemotherapy on OS. Results About 1794 patients treated between 2000 and 2013 at 29 centers were analyzed. Median follow-up was 29.1 months. About 1333 patients (74%) received first-line chemotherapy: the use of first-line chemotherapy was associated with longer OS: [hazard ratio (HR): 1.91, 95% confidence interval (CI): 1.67–2.20]. Type of first-line chemotherapy received was: cisplatin-based 669 (50%), carboplatin-based 399 (30%) and other 265 (20%). Cisplatin use was an independent favorable prognostic factor (HR: 1.54, 95% CI: 1.35–1.77). This benefit was independent of baseline characteristics or comorbidities but was associated with eligibility-for-cisplatin: eligible patients treated with cisplatin lived longer than those who were not (HR: 1.74, 95% CI: 1.36–2.21), while such benefit was not observed among ineligible patients. About 26% of patients who did not receive cisplatin were eligible for this agent. Median OS of ineligible patients was poor irrespective of the chemotherapy used. Conclusions The importance of applying published criteria of eligibility-for-cisplatin was confirmed in a multinational, real-world setting in aUTC. The reasons for deviations from these criteria set targets to improve adherence. Effective therapies for cisplatin-ineligible patients are needed. urothelial cancer, chemotherapy, cisplatin eligibility Key Message Cisplatin-based chemotherapy offers a survival advantage over non-cisplatin chemotherapy in patients with advanced urinary-tract-cancer who are eligible-for-cisplatin, according to recently established criteria. Adherence to eligibility-for-cisplatin criteria can be improved in everyday practice. Effective therapies for ineligible-for-cisplatin patients should be sought. Introduction Cisplatin-based chemotherapy is the treatment of choice for advanced urinary tract cancer (aUTC). Nevertheless, a (yet not precisely estimated) percentage of patients do not receive any systemic therapy [1]. Furthermore, about 50% of treated patients do not receive cisplatin [2] mainly due to concerns of increased toxicity and/or limited efficacy in certain patient groups. Instead, carboplatin is a preferred agent in the community due to its more favorable toxicity profile and ease of administration [3–8]. The absence of level I evidence establishing the superiority of cisplatin over carboplatin and lack of well-established criteria defining patients likely to benefit from cisplatin instead of carboplatin-based therapy may encourage this tendency. Recently, specific criteria for eligibility-for-cisplatin have been proposed by Galsky et al. [9]. Although these criteria are widely accepted, their impact has not yet been formally assessed. In addition, adherence to these criteria outside the context of clinical trials has not been studied. In a recent report, 25% of patients with aUTC, treated at 10 Greek Oncology centers, did not receive cisplatin-based chemotherapy, although they were eligible [10]. The effect of deviations from these criteria on patients’ outcomes remains unknown. We attempted to map the different practice patterns in aUTC and the reasons behind those practices as well as their impact on patient outcomes. We used the Retrospective International Study of Cancers of the Urothelium (RISC) database, the largest multinational database of UTC patients worldwide [11–13]. This database includes data from hospitals in the United States, Europe, Israel and Canada, thus ensuring an adequately wide representation of contemporary trends in the treatment of UTC. Patients and methods Patients Data were collected using a web-based electronic data capture tool. Data fields were centrally reviewed and queries were completed by each participating site. For the current study, selection criteria were: histopathologically confirmed UTC (defined as primary carcinoma of the urinary bladder, renal pelvis, ureter or urethra); advanced disease (defined as metastatic or nonmetastatic inoperable disease, i.e. clinical stages T4bN0M0, TanyN + M0 and TanyNanyM+); histological subtype of pure or mixed urothelial carcinoma, pure squamous or pure adenocarcinoma; survival data available. Administration of chemotherapy for radiosensitization only was not considered as first-line chemotherapy. Database was locked in October 2015. The study was approved by the ethics committees at each participating institution. Statistical analysis This study aimed to address the impact of the type of management on the outcome of patients with aUTC. Overall survival (OS) was the primary variable, while the study of types of chemotherapy used and the impact of eligibility-for-cisplatin on chemotherapy utilization were the secondary end points. Eligibility-for-cisplatin was defined according to the criteria of Galsky et al. [9]: Eastern Cooperative Oncology Group performance status (ECOG-PS) 0 or 1, calculated (Cockroft–Gault formula) creatinine clearance (CrCl) ≥ 60 ml/min, no hearing loss, no heart failure and no significant neuropathy. Definitions of groups regarding treating center and disease extent are included in supplementary Appendix S1, available at Annals of Oncology online. OS was defined as the time between the diagnosis of advanced disease and the date of death from any cause. Alive patients were censored at the date of last contact. Survival curves were estimated using the Kaplan–Meier method. Factors studied for their prognostic ability are listed in supplementary Table S1, available at Annals of Oncology online. To account for missing values (supplementary Table S2, available at Annals of Oncology online), multiple imputations were employed using the Markov Chain Monte Carlo method for Arbitrary Missing Data (number of imputations was 35). Average treatment effect (ATE) was calculated using regression adjustment; averages of treatment-specific predicted mean survival times to estimate mean survival times for each potential outcome. Further details on statistical methodology can be found in supplementary Appendix S1, available at Annals of Oncology online. Results Population included in the analysis The flow chart of our analyses is depicted in Figure 1. From 3024 patients included in the RISC database, 1794 patients with advanced UTC, managed at 29 centers between 2000 and 2013, fulfilled selection criteria (Table 1). Their median time from diagnosis of UTC to diagnosis of advanced disease is 0.41 years (range 0–20.4). About 1333 patients received first-line chemotherapy and had information about the type of chemotherapy and were included in the survival analyses according to type of chemotherapy. Objective tumor responses to first-line chemotherapy were reported in 41% of the 1039 assessable patients. Complete responses were significantly more frequent among cisplatin-treated patients (12% versus 5%, P < 0.001). Table 1. Baseline characteristics of patients included in the analyses Characteristic  Received first-line chemotherapy   P  All patients (n = 1794)  Yes  No    Age (median, range)a  67.6, 34–94.2   70.7, 20.5–93.3   <0.001      n (%)    Gendera      0.003   M  1060 (78.8)  318 (72)   F  285 (21.2)  124 (28)  Charlson Comorbidity Indexa      <0.001   0  603 (47.3)  121 (31.6)   1–2  370 (29)  145 (37.9)   >2  301 (23.7)  117 (30.6)  Histologya      <.001   Transitional  1174 (87.6)  337 (76.2)   Mixed  104 (7.8)  54 (12.2)   Other  62 (4.6)  51 (11.5)  Metastatic sites         Locoregional  265 (19.6)  111 (25.1)  0.015   Lymph nodes  863 (63.9)  202 (45.6)  <0.001   Bone  367 (27.2)  123 (27.8)  0.806   Lung  371 (27.5)  117 (26.4)  0.666   Liver  250 (18.5)  69 (15.6)  0.162   Brain  24 (1.8)  15 (3.4)  0.044   Adrenal  41 (3)  5 (1.1)  0.028   Peritoneum  62 (4.6)  31 (7)  0.047   Distantb  808 (59.8)  270 (61)  0.671  Center      <0.001   Low-volume (n ≤ 100)  570 (42.2)  248 (56)   High-volume (n > 100)  781 (57.8)  195 (44)  Racea      <0.001   Caucasian  1246 (92.6)  377 (85.7)   Other  99 (7.4)  63 (14.3)  Ethnicitya      0.274   Hispanic/Latino  120 (9.3)  32 (7.6)   Other  1169 (90.7)  391 (92.4)  Neoadjuvant chemotherapya      <0.001   Yes  166 (12.9)  104 (24.5)   No  1121 (87.1)  321 (75.5)  Adjuvant therapya      <0.001   Yes  182 (14.3)  97 (22.8)   No  1092 (85.7)  329 (77.2)  Surgical removal of primary tumora      <0.001   Yes  732 (55.3)  312 (71.4)   No  592 (44.7)  125 (28.6)  Radiation therapy to primary tumora      <0.001   Yes  119 (9.4)  87 (20.5)   No  1145 (90.6)  338 (79.5)  Primary sitea      <0.001   Bladder  1085 (81.8)  402 (93.3)   Other  241 (18.2)  29 (6.7)    Characteristic  Received first-line chemotherapy   P  All patients (n = 1794)  Yes  No    Age (median, range)a  67.6, 34–94.2   70.7, 20.5–93.3   <0.001      n (%)    Gendera      0.003   M  1060 (78.8)  318 (72)   F  285 (21.2)  124 (28)  Charlson Comorbidity Indexa      <0.001   0  603 (47.3)  121 (31.6)   1–2  370 (29)  145 (37.9)   >2  301 (23.7)  117 (30.6)  Histologya      <.001   Transitional  1174 (87.6)  337 (76.2)   Mixed  104 (7.8)  54 (12.2)   Other  62 (4.6)  51 (11.5)  Metastatic sites         Locoregional  265 (19.6)  111 (25.1)  0.015   Lymph nodes  863 (63.9)  202 (45.6)  <0.001   Bone  367 (27.2)  123 (27.8)  0.806   Lung  371 (27.5)  117 (26.4)  0.666   Liver  250 (18.5)  69 (15.6)  0.162   Brain  24 (1.8)  15 (3.4)  0.044   Adrenal  41 (3)  5 (1.1)  0.028   Peritoneum  62 (4.6)  31 (7)  0.047   Distantb  808 (59.8)  270 (61)  0.671  Center      <0.001   Low-volume (n ≤ 100)  570 (42.2)  248 (56)   High-volume (n > 100)  781 (57.8)  195 (44)  Racea      <0.001   Caucasian  1246 (92.6)  377 (85.7)   Other  99 (7.4)  63 (14.3)  Ethnicitya      0.274   Hispanic/Latino  120 (9.3)  32 (7.6)   Other  1169 (90.7)  391 (92.4)  Neoadjuvant chemotherapya      <0.001   Yes  166 (12.9)  104 (24.5)   No  1121 (87.1)  321 (75.5)  Adjuvant therapya      <0.001   Yes  182 (14.3)  97 (22.8)   No  1092 (85.7)  329 (77.2)  Surgical removal of primary tumora      <0.001   Yes  732 (55.3)  312 (71.4)   No  592 (44.7)  125 (28.6)  Radiation therapy to primary tumora      <0.001   Yes  119 (9.4)  87 (20.5)   No  1145 (90.6)  338 (79.5)  Primary sitea      <0.001   Bladder  1085 (81.8)  402 (93.3)   Other  241 (18.2)  29 (6.7)    Patients who received first-line chemotherapy (n = 1333)  Characteristic  Cisplatin  Carboplatin  Other  P  (n = 669)  (n = 399)  (n = 265)  Age (median, range)a  65.1, 34–87  71, 38.8–92.9  68.6, 37.5–92  <0.001  Creatinine clearance (median, range)a  71.9, 20.2–242.5  50.9, 1.6–334.6  54.2, 8.6–164.8  <0.001     n (%)     Gendera        0.037   M  542 (81.6)  302 (75.7)  201 (76.1)   F  122 (18.4)  97 (24.3)  63 (23.9)  Charlson Comorbidity Indexa        <0.001   0  354 (55.1)  161 (43.2)  85 (35.4)   1–2  183 (28.5)  94 (25.2)  88 (36.7)   >2  106 (16.4)  118 (31.6)  67 (27.9)  Histologya        0.270   Transitional  591 (89)  348 (88.3)  221 (83.7)   Mixed  45 (6.8)  29 (7.43)  26 (9.9)   Other  28 (4.2)  17 (4.3)  17 (6.4)  Metastatic sites           Locoregional  132 (19.7)  80 (20.1)  50 (18.9)  0.930   Lymphnodes  440 (65.8)  258 (64.7)  159 (60)  0.248   Bone  155 (23.2)  128 (32.1)  78 (29.4)  0.004   Lung  183 (27.4)  105 (26.3)  79 (29.8)  0.608   Liver  128 (19.1)  76 (19.1)  44 (16.6)  0.645   Brain  14 (2.1)  7 (1.8)  3 (1.1)  0.607   Adrenal  17 (2.5)  15 (3.8)  9 (3.4)  0.507   Peritoneum  19 (2.8)  21 (5.3)  20 (7.6)  0.005   Distantb  374 (55.9)  247 (61.9)  175 (66)  0.010  Center        0.560   Low volume (n ≤ 100)  287 (42.9)  163 (40.9)  104 (39.3)   High volume (n > 100)  382 (57.1)  236 (59.1)  161 (60.8)  Racea        0.490   Caucasian  618 (93.1)  364 (91.2)  246 (93.2)   Other  46 (6.9)  35 (8.8)  18 (6.8)  Ethnicitya        0.049   Hispanic/Latino  57 (9.1)  29 (7.5)  34 (13.2)   Other  568 (90.9)  359 (92.5)  224 (86.8)  Neoadjuvant chemotherapya        <0.001   Yes  45 (7.1)  42 (11.3)  79 (30.5)   No  593 (92.9)  331 (88.7)  180 (69.5)  Adjuvant therapya        <0.001   Yes  48 (7.6)  48 (13)  82 (31.9)   No  582 (92.4)  321 (87)  175 (68.1)  Surgical removal of primary tumora        <0.001   Yes  334 (50.8)  205 (53.2)  180 (68.2)   No  323 (49.2)  180 (46.8)  84 (31.8)  Radiation therapy to primary tumora        0.011   Yes  44 (7)  41 (11.3)  32 (12.7)   No  587 (93)  323 (88.7)  220 (87.3)  Post chemotherapy surgerya        0.019   Yes  45 (7.3)  11 (3.1)  12 (4.9)   No  570 (92.7)  346 (96.9)  232 (95.1)  Primary sitea        0.533   Bladder  544 (82.7)  310 (79.9)  214 (81.7)   Other  114 (17.3)  78 (20.1)  48 (18.3)    ECOG-PSa        <0.001   0  232 (43.1)  65 (20.9)  49 (24.8)   1  256 (47.4)  154 (49.5)  107 (54)   2  44 (8.2)  75 (24.1)  35 (17.7)   3  6 (1.1)  14 (4.5)  7 (3.5)   4  1 (0.2)  3 (1)  0 (0)  Risk stratificationa,c        <0.001   Low  228 (42.3)  87 (28)  51 (25.7)   Intermediate  275 (51)  159 (51.1)  117 (59.1)   High  36 (6.7)  65 (20.9)  30 (15.2)  Eligibility for cisplatina,d        <0.001   Yes  275 (61)  61 (21.4)  63 (32.6)   No  176 (39)  224 (78.6)  130 (67.4)  Cycles of chemotherapy (median, range)a  5 (1–35)  4 (0–21)  4 (1–26)  <0.001  Patients who received first-line chemotherapy (n = 1333)  Characteristic  Cisplatin  Carboplatin  Other  P  (n = 669)  (n = 399)  (n = 265)  Age (median, range)a  65.1, 34–87  71, 38.8–92.9  68.6, 37.5–92  <0.001  Creatinine clearance (median, range)a  71.9, 20.2–242.5  50.9, 1.6–334.6  54.2, 8.6–164.8  <0.001     n (%)     Gendera        0.037   M  542 (81.6)  302 (75.7)  201 (76.1)   F  122 (18.4)  97 (24.3)  63 (23.9)  Charlson Comorbidity Indexa        <0.001   0  354 (55.1)  161 (43.2)  85 (35.4)   1–2  183 (28.5)  94 (25.2)  88 (36.7)   >2  106 (16.4)  118 (31.6)  67 (27.9)  Histologya        0.270   Transitional  591 (89)  348 (88.3)  221 (83.7)   Mixed  45 (6.8)  29 (7.43)  26 (9.9)   Other  28 (4.2)  17 (4.3)  17 (6.4)  Metastatic sites           Locoregional  132 (19.7)  80 (20.1)  50 (18.9)  0.930   Lymphnodes  440 (65.8)  258 (64.7)  159 (60)  0.248   Bone  155 (23.2)  128 (32.1)  78 (29.4)  0.004   Lung  183 (27.4)  105 (26.3)  79 (29.8)  0.608   Liver  128 (19.1)  76 (19.1)  44 (16.6)  0.645   Brain  14 (2.1)  7 (1.8)  3 (1.1)  0.607   Adrenal  17 (2.5)  15 (3.8)  9 (3.4)  0.507   Peritoneum  19 (2.8)  21 (5.3)  20 (7.6)  0.005   Distantb  374 (55.9)  247 (61.9)  175 (66)  0.010  Center        0.560   Low volume (n ≤ 100)  287 (42.9)  163 (40.9)  104 (39.3)   High volume (n > 100)  382 (57.1)  236 (59.1)  161 (60.8)  Racea        0.490   Caucasian  618 (93.1)  364 (91.2)  246 (93.2)   Other  46 (6.9)  35 (8.8)  18 (6.8)  Ethnicitya        0.049   Hispanic/Latino  57 (9.1)  29 (7.5)  34 (13.2)   Other  568 (90.9)  359 (92.5)  224 (86.8)  Neoadjuvant chemotherapya        <0.001   Yes  45 (7.1)  42 (11.3)  79 (30.5)   No  593 (92.9)  331 (88.7)  180 (69.5)  Adjuvant therapya        <0.001   Yes  48 (7.6)  48 (13)  82 (31.9)   No  582 (92.4)  321 (87)  175 (68.1)  Surgical removal of primary tumora        <0.001   Yes  334 (50.8)  205 (53.2)  180 (68.2)   No  323 (49.2)  180 (46.8)  84 (31.8)  Radiation therapy to primary tumora        0.011   Yes  44 (7)  41 (11.3)  32 (12.7)   No  587 (93)  323 (88.7)  220 (87.3)  Post chemotherapy surgerya        0.019   Yes  45 (7.3)  11 (3.1)  12 (4.9)   No  570 (92.7)  346 (96.9)  232 (95.1)  Primary sitea        0.533   Bladder  544 (82.7)  310 (79.9)  214 (81.7)   Other  114 (17.3)  78 (20.1)  48 (18.3)    ECOG-PSa        <0.001   0  232 (43.1)  65 (20.9)  49 (24.8)   1  256 (47.4)  154 (49.5)  107 (54)   2  44 (8.2)  75 (24.1)  35 (17.7)   3  6 (1.1)  14 (4.5)  7 (3.5)   4  1 (0.2)  3 (1)  0 (0)  Risk stratificationa,c        <0.001   Low  228 (42.3)  87 (28)  51 (25.7)   Intermediate  275 (51)  159 (51.1)  117 (59.1)   High  36 (6.7)  65 (20.9)  30 (15.2)  Eligibility for cisplatina,d        <0.001   Yes  275 (61)  61 (21.4)  63 (32.6)   No  176 (39)  224 (78.6)  130 (67.4)  Cycles of chemotherapy (median, range)a  5 (1–35)  4 (0–21)  4 (1–26)  <0.001  a Missing values apply and are listed in supplementary Table S2, available at Annals of Oncology online. b Distant metastases: metastases outside primary site and lymph nodes. c Risk stratification according to Bajorin et al. [8]. d Eligibility was defined according to Galsky et al. [9]. SD, standard deviation. Figure 1. View largeDownload slide Study flow chart. Figure 1. View largeDownload slide Study flow chart. Treatment patterns Patients who did not receive first-line chemotherapy were older, more frequently noncaucasians, had higher frequency of nonurothelial histology, higher Charlson Comorbidity Index (CCI), received more frequently perioperative chemotherapy and management of primary tumor, and were predominantly treated at low volume centers (Table 1). Cisplatin-based first-line chemotherapy was administered to 669 patients, carboplatin based to 399 and other therapies to 265. Detailed description of chemotherapy administered is presented in supplementary Table S3, available at Annals of Oncology online. Cisplatin administration was associated with younger age, male gender, less frequent perioperative chemotherapy, better PS and renal function, and less frequent distant metastases. Cisplatin-treated patients also had a lower CCI: specifically, hypertension, diabetes, chronic obstructive pulmonary disease (COPD), coronary artery disease, congestive heart failure (CHF), hyperlipidemia and cerebrovascular disease were less frequent among cisplatin-treated patients (supplementary Table S4, available at Annals of Oncology online). Eligibility-for-cisplatin Adequate data to assess eligibility-for-cisplatin according to Galsky’s criteria [9] were available for 929 patients: 429 patients (46%) had at least one criterion for ineligibility (supplementary Figure S1, available at Annals of Oncology online). The most common criterion was CrCl < 60 ml/min (97.4%) followed by PS ≥ 2 (43.1%). The reason for not using cisplatin was provided by the investigators in 340 cases (supplementary Table S5, available at Annals of Oncology online). Established ineligibility criteria accounted for only 56% of these cases. Advanced age and comorbidities were the most frequent reasons for not administering cisplatin to patients who were otherwise eligible. Since reasons for not administering cisplatin were not reported in the majority of patients, we carried out analyses on all 929 patients who were assessable for cisplatin-eligibility to more accurately examine deviations from these criteria. One hundred twenty-four patients (26%) who did not receive cisplatin were eligible for this agent (Table 1). This deviation from Galsky’s criteria was more frequent in patients with: CCI ≥ 3 (49% versus 25%, P < 0.001), Memorial Sloan Kettering Cancer Center (MSKCC) intermediate/high risk (36% versus 24%, P = 0.012), hypertension (47% versus 23%, P < 0.001), diabetes (56% versus 28%, P < 0.001) and COPD (60% versus 29%, P = 0.001). Interestingly, 176 patients (39%) who received cisplatin fulfilled at least one criterion for ineligibility: CHF (n = 9), ECOG-PS > 1 (n = 51), CrCl < 60 ml/min (n = 129). Survival analyses Overall population Among the 1794 selected patients, 638 (36%) were alive, 1004 patients died of disease progression, 10 died from treatment-related complications, 41 from other causes and 101 from unknown reason, at the time of analysis. The median follow-up was 29.1 months [95% confidence interval (CI) 26.4–31.4]. Patients who received systemic chemotherapy had significantly longer median OS (Figure 2). Multivariate analysis confirmed the independent significance of chemotherapy administration (supplementary Table S7, available at Annals of Oncology online). This result remained essentially unchanged when patients who received perioperative chemotherapy were excluded. Nevertheless, these analyses should be viewed with caution, since PS was not included in the cox regression model due to high number of missing values among nontreated patients. Figure 2. View largeDownload slide Kaplan–Meier curves of overall survival (OS) according to the type of first-line chemotherapy (A); Forest plot of subgroup analysis according to the type of first-line chemotherapy (B); Kaplan–Meier curves of OS according to: the type of first-line chemotherapy stratified by eligibility-for-cisplatin (C). Maximum analysis time to be graphed was selected at 6 years. Patients alive at 6 years (regardless of subsequent death) were censored at that time. Blue dots in forest plot denote hazard ratios for the treatment type and horizontal bars 95% confidence interval. Interaction refers to potential interactions of the effects of treatment type (cisplatin versus carboplatin + other) and each of the covariates listed on the left-hand side of the figure. P-value of interaction refers to the P-value of the interaction term from a Cox model, which includes treatment type and the respective covariate. CCI, Charlson Comorbidity Index; COPD, chronic obstructive airway disease; PS, performance status; Hgb, hemoglobin; WBC, while blood cells. Figure 2. View largeDownload slide Kaplan–Meier curves of overall survival (OS) according to the type of first-line chemotherapy (A); Forest plot of subgroup analysis according to the type of first-line chemotherapy (B); Kaplan–Meier curves of OS according to: the type of first-line chemotherapy stratified by eligibility-for-cisplatin (C). Maximum analysis time to be graphed was selected at 6 years. Patients alive at 6 years (regardless of subsequent death) were censored at that time. Blue dots in forest plot denote hazard ratios for the treatment type and horizontal bars 95% confidence interval. Interaction refers to potential interactions of the effects of treatment type (cisplatin versus carboplatin + other) and each of the covariates listed on the left-hand side of the figure. P-value of interaction refers to the P-value of the interaction term from a Cox model, which includes treatment type and the respective covariate. CCI, Charlson Comorbidity Index; COPD, chronic obstructive airway disease; PS, performance status; Hgb, hemoglobin; WBC, while blood cells. Patients who received first-line chemotherapy Among the 1333 patients, 872 were dead at the time of analysis (disease: 782, toxicity: 9, other causes: 24, unknown reason: 57). The median follow-up time was 31.7 months. Cisplatin-based chemotherapy was associated with significantly longer median OS, (Table 2 and Figure 2A). Univariate analysis showed that bone, liver, lung and distant metastases, smoking history, CCI ≥ 3, ECOG-PS (0 versus 1 + 2 and 0 + 1 versus 2), white blood cell (WBC) > 10 000/mm3, plt > 450 000/mm3, hemoglobin ≤ institutional lower limit, albumin ≤ median, age > median and increasing MSKCC score were adverse prognostic factors, while the presence of lymph node metastases and BMI > median was favorable prognostic factors. BMI, WBC, plt, hemoglobin and albumin, studied as continuous variables, were also significantly associated with prognosis. Origin of the center, histology and taxane use were not associated with OS. When significant factors were entered into a multivariate cox regression model including therapy group (cisplatin versus noncisplatin), the latter retained its significance (Table 3). ATE analysis showed that death would occur 6.9 months earlier if no patient received cisplatin-based chemotherapy, compared with the outcome if everybody received such treatment (supplementary Table S6, available at Annals of Oncology online). The independent prognostic significance of treatment type was also retained when multiple imputations for missing values were applied, after stratification for eligibility-for-cisplatin (supplementary Table S8, available at Annals of Oncology online). Since other therapies or different histological subtypes may have confounded our results, we also carried out multivariate analyses excluding patients who had received perioperative chemotherapy, postchemotherapy surgery or had nontransitional histology. The independent prognostic significance of cisplatin-based chemotherapy remained unchanged. Table 2. Unadjusted hazard ratios and 95% CI for death according to the use of first-line chemotherapy   n  Median OS (95% CI)  HR (95% CI)  P  All patients (n = 1794)    Received chemotherapy        <0.001   Yes  1351  14.8 (13.8–15.6)  1     No  443  5.7 (4.6–6.8)  1.91 (1.67–2.2)    Patients who received first-line chemotherapy (n = 1333)    Therapy group        <0.001   Cisplatin  669  18 (16.1–19.8)  1     Carboplatin  399  12.5 (10.9–13.4)  1.58 (1.36–1.83)     Other  265  12.6 (9.9–15.4)  1.5 (1.25–1.79)    Eligible-for-cisplatin        <0.001   Cisplatin treated  275  20.3 (17.1–22.7)  1     Carboplatin treated  61  12.9 (8.7–15.7)  2.3 (1.67–3.17)     Other  63  16.7 (11.8–21)  1.5 (1.04–2.1)    Ineligible-for-cisplatin  176  12.7 (11.2–15.1)  1  0.083   Cisplatin treated  224  11.6 (10.5–13.4)  1.12 (0.88–1.43)     Carboplatin treated  130  8.7 (6.9–10.1)  1.37 (1.05–1.8)    Other      n  Median OS (95% CI)  HR (95% CI)  P  All patients (n = 1794)    Received chemotherapy        <0.001   Yes  1351  14.8 (13.8–15.6)  1     No  443  5.7 (4.6–6.8)  1.91 (1.67–2.2)    Patients who received first-line chemotherapy (n = 1333)    Therapy group        <0.001   Cisplatin  669  18 (16.1–19.8)  1     Carboplatin  399  12.5 (10.9–13.4)  1.58 (1.36–1.83)     Other  265  12.6 (9.9–15.4)  1.5 (1.25–1.79)    Eligible-for-cisplatin        <0.001   Cisplatin treated  275  20.3 (17.1–22.7)  1     Carboplatin treated  61  12.9 (8.7–15.7)  2.3 (1.67–3.17)     Other  63  16.7 (11.8–21)  1.5 (1.04–2.1)    Ineligible-for-cisplatin  176  12.7 (11.2–15.1)  1  0.083   Cisplatin treated  224  11.6 (10.5–13.4)  1.12 (0.88–1.43)     Carboplatin treated  130  8.7 (6.9–10.1)  1.37 (1.05–1.8)    Other    OS, overall survival; HR, hazard ratio; CI, confidence interval. Table 3. Multivariate analysis for OS in patients who received first-line chemotherapy Factor  Subgroups  HR  95% CI  P-value  Distant metastases and PS in regression model (n = 727)  WBC  ≤10 000  1    <0.001  >10 000  1.49  1.23–1.8  Hb  ≤ ILL  1    0.001  > ILL  1.38  1.13–1.68  Distant metastases  No  1    0.002  Yes  1.33  1.11–1.6  Treatment  Cisplatin  1    0.003  Non Cisplatin  1.32  1.1–1.58  ECOG-PS  0  1    0.007  >0  1.32  1.08–1.6  MSKCC risk groups in Cox regression model (n = 727)  MSKCC risk  Low  1    <0.001  Intermediate  1.37  1.12–1.66  High  2.17  1.6–2.94  WBC  ≤10 000  1    <0.001  >10 000  1.42  1.17–1.73  Hb  ≥ILL  1    0.002  <ILL  1.36  1.12–1.66  Treatment  Cisplatin  1    0.007  Non Cisplatin  1.29  1.07–1.55  Factor  Subgroups  HR  95% CI  P-value  Distant metastases and PS in regression model (n = 727)  WBC  ≤10 000  1    <0.001  >10 000  1.49  1.23–1.8  Hb  ≤ ILL  1    0.001  > ILL  1.38  1.13–1.68  Distant metastases  No  1    0.002  Yes  1.33  1.11–1.6  Treatment  Cisplatin  1    0.003  Non Cisplatin  1.32  1.1–1.58  ECOG-PS  0  1    0.007  >0  1.32  1.08–1.6  MSKCC risk groups in Cox regression model (n = 727)  MSKCC risk  Low  1    <0.001  Intermediate  1.37  1.12–1.66  High  2.17  1.6–2.94  WBC  ≤10 000  1    <0.001  >10 000  1.42  1.17–1.73  Hb  ≥ILL  1    0.002  <ILL  1.36  1.12–1.66  Treatment  Cisplatin  1    0.007  Non Cisplatin  1.29  1.07–1.55  WBC, white blood cells; ILL, institutional lower limit; ECOG-PS, Eastern Cooperative Oncology Group performance status; MSKCC, Memorial Sloan Kettering Cancer Center. Subgroup analyses Benefit from cisplatin-based therapy was obtained regardless of the administration of perioperative chemotherapy and across all prognostic groups or comorbidities (Figure 2B) with the exception of the eligibility-for-cisplatin groups. Cisplatin-based chemotherapy was associated with a survival benefit, only among eligible patients (Figure 2C and Table 2). The interaction test between treatment type and eligibility-for-cisplatin yielded a significant result (P = 0.024). Among ineligible patients, carboplatin-based therapy was associated with longer median OS compared with nonplatinum chemotherapy (Table 2). Discussion This is the first study to describe international practice patterns and their effect on the outcome of patients with aUTC. The study has certain limitations associated with its retrospective nature. Inaccuracies in reporting cannot be completely excluded, in spite of the carried out quality control of the data. Our analyses are also subject to selection biases. Although we attempted to mitigate them by using multivariate analyses, residual unmeasured confounding factors may have impacted our results. In spite of these limitations, our data are original and potentially useful for clinical practice and future research. Apart from the depiction of practice patterns across countries with different health and insurance systems, the RISC database enabled the study of so far unexplored subjects, such as the comparison between cisplatin and carboplatin, as well as the value of published criteria of eligibility-for-cisplatin and the impact of deviations from these criteria on outcome. Given that a randomized study, directly comparing cisplatin with carboplatin is unlikely, this analysis provides the highest possible level of evidence to this question, while it also provides useful benchmarks of the outcome of the various subgroups of patients with aUTC. Patients who received chemotherapy lived significantly longer than those who did not. This OS benefit was confirmed in a multivariate analysis, which, however, is limited by the lack of PS data for untreated patients. Although the reasons for not offering chemotherapy are unclear, our analysis suggests that a combination of frailty as well as experience of the center in the treatment of aUTC may be the most relevant underlying causes. Nevertheless, other reasons, due to variations in referral patterns, which cannot be addressed by this study, cannot be excluded. Cisplatin administration was associated with OS benefit. An inherent bias exists in this analysis due to distinct patient profiles of better PS and fewer distant metastases in the cisplatin group. To overcome these imbalances, we used multivariate analyses, which showed that this benefit was independent of other prognostic factors. The OS benefit was restricted to the eligible-for-cisplatin population, confirming the validity of this feature as a selection criterion for cisplatin-based chemotherapy. Eligible patients treated with cisplatin had a median OS of 20.3 months, which compares favorably with those reported in randomized trials [14–18] and suggests that this therapy will remain a valuable option in spite of the important emerging changes in the treatment paradigm in aUTC. In contrast, eligible patients not treated with cisplatin had a significantly shorter median OS. This result underlines the importance of adhering to Galsky’s criteria and promptly identifying and correcting potentially reversible causes of ineligibility, such as recently established renal impairment due to urinary tract obstruction by primary tumor, in order to ensure optimal outcome. We found that 26% of patients not receiving cisplatin were eligible for this agent. This deviation was mainly due to advanced age and comorbidities, especially hypertension, diabetes and COPD. This practice may not be justified as our analysis showed that patients with these comorbidities derived the same benefit from cisplatin administration as their counterparts. These findings set clear targets to improve our management of advanced UTC. Strong emphasis on the eligibility criteria should be considered by Urology and Oncology Scientific and Professional Societies as well as academic institutions and health care providers at national and international level. Referral to centers with experience in treating UTC will improve utilization of cisplatin-based chemotherapy. It should be noted that patients included in this analysis were treated before the publication of Galsky et al. [9]. It could be speculated that adherence to eligibility criteria has been improved since. Nevertheless, recent observational studies suggested that considerable deviations still exist in everyday practice [10]. Interestingly, perioperative chemotherapy was associated with less frequent use of first-line chemotherapy as well as of cisplatin-based chemotherapy in otherwise eligible patients. This suggests that prior exposure to systemic chemotherapy may represent a dissuasive factor for offering similar therapy at progression. Although our results suggested similar benefit by cisplatin regardless of previous perioperative therapy, further analyses, i.e. according to the time from perioperative chemotherapy to progression or the type of chemotherapy, were not possible due to the small number of patients receiving perioperative therapy. Further study is warranted to accurately evaluate the potential of this factor as an additional criterion for cisplatin-ineligibility, especially in the context of the development of novel effective options in first-line therapy of aUTC. Fifty-seven percent of patients were ineligible-for-cisplatin. The distribution of criteria of ineligibility is similar to those of a national retrospective study [10] as well as of a recently reported phase II study [19]. This similarity supports the validity of these criteria and their applicability in the context of clinical trials. Optimal therapy for cisplatin-ineligible patients suffering from aUTC remains an unmet medical need. Importantly, cisplatin was used in 176 ineligible patients but this was not associated with an equally favorable effect as among eligible patients, which implies that cisplatin-ineligibility may be a surrogate for a more aggressive tumor behavior and not merely for unfitness for this particular agent. Currently, carboplatin-based chemotherapy is considered the standard for ineligible-for-cisplatin patients [20]. Our results support this recommendation, since carboplatin-based chemotherapy produced a longer median OS than nonplatinum therapy. Nevertheless, the median OS of 11.6 months emphasizes the need for developing more effective choices for this population. In this context, the median OS of 15.9 months recently reported for the programmed death ligand 1 inhibitor atezolizumab in first-line treatment of ineligible-for-cisplatin patients appears particularly promising [14]. Conclusion In conclusion, all efforts should be made to treat eligible-for-cisplatin patients with aUTC with cisplatin-based chemotherapy. The field should recognize that cisplatin may be underutilized and that adherence to published criteria can help increase appropriate cisplatin use. Novel, more effective therapies for ineligible-for-cisplatin patients should be sought through clinical trials specifically focused in this population. Funding None declared. Disclosure The authors have declared no conflicts of interest. References 1 Sonpavde G, Watson D, Tourtellott M et al.   Administration of cisplatin-based chemotherapy for advanced urothelial carcinoma in the community. Clin Genitourin Cancer  2012; 10( 1): 1– 5. Google Scholar CrossRef Search ADS PubMed  2 Dash A, Galsky MD, Vickers AJ et al.   Impact of renal impairment on eligibility for adjuvant cisplatin-based chemotherapy in patients with urothelial carcinoma of the bladder. Cancer  2006; 107( 3): 506– 513. Google Scholar CrossRef Search ADS PubMed  3 Linardou H, Aravantinos G, Efstathiou E et al.   Gemcitabine and carboplatin combination as first-line treatment in elderly patients and those unfit for cisplatin-based chemotherapy with advanced bladder carcinoma: phase II study of the Hellenic Co-operative Oncology Group. Urology  2004; 64( 3): 479– 484. Google Scholar CrossRef Search ADS PubMed  4 Bamias A, Lainakis G, Kastritis E et al.   Biweekly carboplatin/gemcitabine in patients with advanced urothelial cancer who are unfit for cisplatin-based chemotherapy: report of efficacy, quality of life and geriatric assessment. Oncology  2007; 73( 5–6): 290– 297. Google Scholar CrossRef Search ADS PubMed  5 Bellmunt J, de Wit R, Albanell J et al.   A feasibility study of carboplatin with fixed dose of gemcitabine in “unfit” patients with advanced bladder cancer. Eur J Cancer  2001; 37( 17): 2212– 2215. Google Scholar CrossRef Search ADS PubMed  6 Carles J, Nogue M, Domenech M et al.   Carboplatin-gemcitabine treatment of patients with transitional cell carcinoma of the bladder and impaired renal function. Oncology  2000; 59( 1): 24– 27. Google Scholar CrossRef Search ADS PubMed  7 De Santis M, Bellmunt J, Mead G et al.   Randomized phase II/III trial assessing gemcitabine/carboplatin and methotrexate/carboplatin/vinblastine in patients with advanced urothelial cancer “unfit” for cisplatin-based chemotherapy: phase II–results of EORTC study 30986. J Clin Oncol  2009; 27( 33): 5634– 5639. Google Scholar CrossRef Search ADS PubMed  8 Bajorin DF, Dodd PM, Mazumdar M et al.   Long-term survival in metastatic transitional-cell carcinoma and prognostic factors predicting outcome of therapy. J Clin Oncol  1999; 17( 10): 3173– 3181. Google Scholar CrossRef Search ADS PubMed  9 Galsky MD, Hahn NM, Rosenberg J et al.   Treatment of patients with metastatic urothelial cancer “unfit” for Cisplatin-based chemotherapy. J Clin Oncol  2011; 29( 17): 2432– 2438. Google Scholar CrossRef Search ADS PubMed  10 Bamias A, Peroukidis S, Stamatopoulou S et al.   Utilization of Systemic Chemotherapy in Advanced Urothelial Cancer: a Retrospective Collaborative Study by the Hellenic Genitourinary Cancer Group (HGUCG). Clin Genitourin Cancer  2016; 14( 2): e153– e159. Google Scholar CrossRef Search ADS PubMed  11 Necchi A, Sonpavde G, Lo Vullo S et al.   RISC Investigators. Nomogram-based prediction of overall survival in patients with metastatic urothelial carcinoma receiving first-line platinum-based chemotherapy: Retrospective International Study of Invasive/Advanced Cancer of the Urothelium (RISC). Eur Urol  2017; 71( 2): 281– 289. Google Scholar CrossRef Search ADS PubMed  12 Ramos JD, Casey MF, Bamias A et al.   Retrospective International Study of Cancers of the Urothelium (RISC) Investigators. The Khorana Score in predicting venous thromboembolism for patients with metastatic urothelial carcinoma and variant histology treated with chemotherapy. Clin Appl Thromb Hemost  2017; 23( 7): 755– 760. Google Scholar CrossRef Search ADS PubMed  13 Ramos JD, Casey MF, Crabb SJ et al.   RISC Investigators. Venous thromboembolism in metastatic urothelial carcinoma or variant histologies: incidence, associative factors, and effect on survival. Cancer Med  2017; 6( 1): 186– 194. Google Scholar CrossRef Search ADS PubMed  14 Sternberg CN, de Mulder PH, Schornagel JH et al.   Randomized phase III trial of high-dose-intensity methotrexate, vinblastine, doxorubicin, and cisplatin (MVAC) chemotherapy and recombinant human granulocyte colony-stimulating factor versus classic MVAC in advanced urothelial tract tumors: European Organization for Research and Treatment of Cancer Protocol no. 30924. J Clin Oncol  2001; 19: 2638– 2646. Google Scholar CrossRef Search ADS PubMed  15 von der Maase H, Hansen SW, Roberts JT et al.   Gemcitabine and cisplatin versus methotrexate, vinblastine, doxorubicin, and cisplatin in advanced or metastatic bladder cancer: results of a large, randomized, multinational, multicenter, phase III study. J Clin Oncol  2000; 18( 17): 3068– 3077. Google Scholar CrossRef Search ADS PubMed  16 Bamias A, Aravantinos G, Deliveliotis C et al.   Docetaxel and cisplatin with granulocyte colony-stimulating factor (G-CSF) versus MVAC with G-CSF in advanced urothelial carcinoma: a multicenter, randomized, phase III study from the Hellenic Cooperative Oncology Group. J Clin Oncol  2004; 22: 220– 228. Google Scholar CrossRef Search ADS PubMed  17 Bamias A, Dafni U, Karadimou A et al.   Prospective, open-label, randomized, phase III study of two dose-dense regimens MVAC versus gemcitabine/cisplatin in patients with inoperable, metastatic or relapsed urothelial cancer: a Hellenic Cooperative Oncology Group study (HE 16/03). Ann Oncol  2013; 24( 4): 1011– 1017. Google Scholar CrossRef Search ADS PubMed  18 Bellmunt J, von der Maase H, Mead GM et al.   Randomized phase III study comparing paclitaxel/cisplatin/gemcitabine and gemcitabine/cisplatin in patients with locally advanced or metastatic urothelial cancer without prior systemic therapy: EORTC Intergroup Study 30987. J Clin Oncol  2012; 30( 10): 1107– 1113. Google Scholar CrossRef Search ADS PubMed  19 Balar AV, Galsky MD, Rosenberg JE et al.   Atezolizumab as first-line treatment in cisplatin-ineligible patients with locally advanced and metastatic urothelial carcinoma: a single-arm, multicentre, phase 2 trial. Lancet  2017; 389( 10064): 67– 76. Google Scholar CrossRef Search ADS PubMed  20 Bellmunt J, Orsola A, Leow JJ et al.   Bladder cancer: ESMO Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol  2014; 25(Suppl 3): iii40– iii48. Google Scholar CrossRef Search ADS PubMed  © The Author 2017. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For Permissions, please email: journals.permissions@oup.com.

Journal

Annals of OncologyOxford University Press

Published: Feb 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off