- Email: [email protected]
Second-line single-agent versus doublet chemotherapy as salvage therapy for metastatic urothelial cancer: a systematic review and meta-analysis
reviews
Annals of Oncology
Annals of Oncology 27: 49–61, 2016 doi:10.1093/annonc/mdv509 Published online 20 October 2015
Second-line single-agent versus doublet chemotherapy as salvage therapy for metastatic urothelial cancer: a systematic review and meta-analysis D. Raggi1, †, R. Miceli1,†, G. Sonpavde2, P. Giannatempo1, L. Mariani1, M. D. Galsky3, J. Bellmunt4 & A. Necchi1* 1 Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy; 2UAB Comprehensive Cancer Center, Birmingham; 3Mount Sinai School of Medicine, Tisch Cancer Institute, New York; 4Dana-Farber Cancer Institute and Harvard Medical School, Boston, USA
Background: The efficacy and safety of a combination of chemotherapeutic agent compared with single-agent chemotherapy in the second-line setting of advanced urothelial carcinoma (UC) are unclear. We aimed to study the survival impact of single-agent compared with doublet chemotherapy as second-line chemotherapy of advanced UC. Patients and methods: Literature was searched for studies including single-agent or doublet chemotherapy in the second-line setting after platinum-based chemotherapy. Random-effects models were used to pool trial-level data according to treatment arm, including median progression-free survival (PFS), overall survival (OS), objective response rate (ORR) probability, and grade 3–4 toxicity. Univariable and multivariable analyses, including sensitivity analyses, were carried out, adjusting for the percent of patients with ECOG performance status ≥1 and hepatic metastases. Results: Forty-six arms of trials including 1910 patients were selected: 22 arms with single agent (n = 1202) and 24 arms with doublets (n = 708). The pooled ORR with single agents was 14.2% [95% confidence interval (CI) 11.1–17.9] versus 31.9% [95% CI 27.3–36.9] with doublet chemotherapy. Pooled median PFS was 2.69 and 4.05 months, respectively. The pooled median OS was 6.98 and 8.50 months, respectively. Multivariably, the odds ratio for ORR and the pooled median difference of PFS were statistically significant (P < 0.001 and P = 0.002) whereas the median difference in OS was not (P = 0.284). When including single-agent vinflunine or taxanes only, differences were significant only for ORR (P < 0.001) favoring doublet chemotherapy. No statistically significant differences in grade 3–4 toxicity were seen between the two groups. Conclusions: Despite significant improvements in ORR and PFS, doublet regimens did not extend OS compared with single agents for the second-line chemotherapy of UC. Prospective trials are necessary to elucidate the role of combination chemotherapy, with or without targeted agents, in the salvage setting. Currently, improvements in this field should be pursued considering single-agent chemotherapy as the foundation for new more active combinations. Key words: urothelial carcinoma, chemotherapy, salvage therapy, single-agent chemotherapy, combination chemotherapy
introduction Rescue of patients with metastatic urothelial cancer (UC) who have failed first-line chemotherapy or who develop a relapse early after perioperative chemotherapy is still problematic. At present, no Food and Drug Administration (FDA)-approved agent is available as a conventional option and vinflunine is the only European Medicines Agency (EMA)-approved drug for these patients [1]. Irrespective of the type of single-agent *Correspondence to: Dr Andrea Necchi, Department of Medical Oncology, Fondazione IRCCS Istituto Nazionale dei Tumori, Via G. Venezian 1, 20133 Milan, Italy.Tel: +39-022390-2402; Fax: +39-02-2390-3150; E-mail: [email protected] †
Both authors contributed equally as co-first authors.
chemotherapy that is employed as first or later salvage therapy, results are frustratingly poor and there is a desperate need for new agents. Median progression-free (PFS) and overall survival (OS) invariably approximated 3 and 6 months, respectively, and these estimates are currently applied to the design of trials of salvage therapy. Improvements in the stratification of outcome in the secondline setting have been achieved as well, and validated prognostic grouping systems are available [2–4]. The same applies to the understanding of the prognostic impact of several additional factors for clinical use. Poor tolerability of chemotherapy in these patients is a major concern and often limits the administration of potentially more active, yet more toxic combination regimens. As a result, the administration of single-agent taxanes ( paclitaxel or
© The Author 2015. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please email: [email protected].
Downloaded from http://annonc.oxfordjournals.org/ by guest on December 23, 2015
Received 12 August 2015; revised 29 September 2015; accepted 12 October 2015
reviews
patients and methods search strategy and data abstraction We carried out a systematic review and meta-analysis in accordance with the preferred reporting items for systematic reviews and meta-analyses (PRISMA) guidelines [5]. Eligible studies were searched in Medline, EMBASE, and meeting abstracts presented at congresses of the American Society of Clinical Oncology (ASCO), European Society for Medical Oncology (ESMO), American Association of Cancer Research (AACR), and Genitourinary Cancers Symposiums. The following inclusion criteria have been adopted: period of publication between 1990 and 2014, more than 10 patients enrolled, trials/studies reporting data on second-line single-agent or doublet combination chemotherapy. Principal exclusion criteria were overlapping publications, lack of relevant outcome data, studies entirely reporting on the patients treated before 1990, studies reporting on regimens combining more than two drugs, and the use of single-agent or combinations including molecularly targeted therapies. One randomized trial that allowed the crossover between the two arms was included because of the good quality of the reported results [6]. The population, intervention, comparison, and outcome (PICO) strategy was conducted and the following search string was utilized: ‘transitional cell carcinoma’/exp OR transitional AND cell AND carcinoma:ab,ti AND ‘chemotherapy’/exp OR ‘salvage therapy’/exp OR ‘single drug dose’/exp OR ‘cancer combination chemotherapy’/exp OR ‘salvage therapy’:ab,ti OR ‘combination | Raggi et al.
chemotherapy’:ab,ti OR ‘single drug dose’:ab,ti AND ‘clinical effectiveness’/exp OR ‘overall survival’/exp OR ‘progression free survival’/exp OR ‘cancer staging’/exp OR ‘toxicity’/exp. Additional queries with relevant variants and filters have been added up, integrated by the search through the ASCO portal. Search results were independently reviewed by two authors (DR and AN). Full articles were retrieved for further qualitative review.
statistical methods The primary objective was to compare treatment containing a single-agent versus a doublet combination chemotherapy; descriptive statistics were used to summarize information across all trials according to the treatment group. The primary end points were the objective response rate (ORR), PFS, and OS; the secondary end point was the incidence of acute (grade 3–4) toxicities. PFS was commonly defined as the time from starting treatment until objective tumor progression or death, OS was defined as the time from treatment start to death for any reason, with censoring at the date of last contact for alive patients. For the purposes of this study, time to progression (TTP) was used whenever it was reported instead of PFS since it was judged an acceptable estimator of PFS in the analyzed studies [i.e. overall, there were very few death events due to toxicity or reasons other than disease progression (PD)]. Random-effects models using inverse variance weighting were used to pool trial-level data (median PFS and OS, and ORR or toxicity probability) separately in each treatment arm. We always applied random-effects models because, according to the Q test and I 2 statistic [7], heterogeneity among the studies was pretty always present (Q statistic, P < 10% and/or I 2 > 25%). The comparisons between treatment arms were performed by random-effects univariable and multivariable models, with adjustment for the percentage of patients with an Eastern Cooperative Oncology Group performance status (ECOG-PS) ≥1 and with hepatic metastases; these percentages were modeled as continuous variables using 3-knots restricted cubic splines [8]. For each study, the standard error (SE) of the median OS or PFS was derived from the respective 95% confidence interval (CI), when reported, or estimated by using a linear regression model in which the response variable was the log(SE) and the covariates were the median OS or PFS (modeled with a 3-knots restricted cubic spline) and the number of patients. Publication bias was evaluated by visually inspecting funnel plots and using Egger’s test for bias. Sensitivity analyses were conducted by using different diagnostic measures like externally standardized residuals, DFFITS values, Cook’s distances, DFBETAS values, the estimates of τ 2 (estimated amount of total heterogeneity) when each study is removed in turn. The statistical analyses were carried out using the SAS® and R software (http://www.R project.org; last access 31 July 2015). All tests and CIs were two-sided; tests were considered statistically significant when the corresponding P values were below the 5% threshold.
results search results We identified a total of 1477 studies using the search criteria (EMBASE: 633; Medline: 812; EMBASE and Medline: 32),
Volume 27 | No. 1 | January 2016
Downloaded from http://annonc.oxfordjournals.org/ by guest on December 23, 2015
docetaxel) or vinflunine are the two equally available options outside of clinical trials. Nonetheless, huge heterogeneity characterizes the available results with second-line chemotherapy and the survival advantage with the use of combination regimens remains substantially unproven. Modern research strategy in UC should focus on the rational delivery of standard systemic chemotherapy and on the optimal chemobiologic combination or sequence with novel compounds. In order to allow the design of informed clinical trials, it will be of primary importance to identify which are the best results achievable with chemotherapy and, ideally, for which patients. For fit patients who may be administered combination chemotherapy, for example, the benchmark of single-agent chemotherapy might not reflect the best achievable result. Otherwise for the majority of patients for whom tolerability of treatments is a concern due to poor performance status or organ function, the single-agent benchmark may be used. For these reasons, quantifying the survival differences between single-agent and combination chemotherapy will be a compelling argument for further research. The comparison of single-agent with doublet chemotherapy as second-line chemotherapy for metastatic UC was the objective of the present systematic review and meta-analysis. We focused on the two-drug combination chemotherapy only because the administration of three or more drugs in these patients is quite rare in clinical practice. The hypothesis was that doublet chemotherapy may be recommended as the first option for selected patients given its exhibited efficacy and that drug development in this disease should separate patient categories according to the best deliverable chemotherapy.
Annals of Oncology
reviews
Annals of Oncology
1477 abstracts and articles identified through database searching following a PICO strategy
748 records excluded (abstracts, reviews, letters, editorials, short surveys)
729 published or in press articles for full text screening
1 abstract selected [29] among 113 GU-ASCO and ASCO abstracts from 2010 to 2014 screened in EMBASE and ASCO portal
729 articles screened
686 full text articles excluded (sequential therapy, combination chemotherapy with >2 agents, use of targeted drugs, £10 patients analyzed or small retrospective series, overlapping series).
44 studies (46 arms) included in the final systematic review and meta-analysis
Figure 1. Preferred reporting items for systematic reviews and meta-analysis flowchart. ASCO (GU), American Society of Clinical Oncology (Genitourinary Cancers Symposium); PICO, population, intervention, comparison, and outcome.
Volume 27 | No. 1 | January 2016
following criteria had been used to assess the response: Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1 in 3 arms, RECIST version 1.0 in 16, World Health Organization (WHO) criteria in 22, ECOG in 3, and Japanese criteria in 1 (not available in 1).
uni- and multivariable meta-analyses for response and survival outcome The summary of the outcomes in the two groups is provided in Table 3, together with the principal subgroup analyses. The pooled ORR with single-agents was 14.2% (95% CI 11.1% to 17.9%) (I 2 = 51.50%, P for heterogeneity = 0.002, with 22 arms included, Egger’s test, P = 0.153, Figure 2A; Bellmunt study was influential, likely due to the high precision of estimates [10]) versus 31.9% (95% CI 27.3% to 36.9%) (I 2 = 38.20%, P for heterogeneity = 0.030, with 24 arms included, Egger’s test, P = 0.008, Figure 2B; no influential studies were identified) with doublets. The pooled median PFS for the single-agent group was 2.69 months (95% CI 2.25–3.12) (I 2 = 90.23%, P for heterogeneity <0.001, with 18 arms included, Egger’s test, P < 0.001, Figure 2C; no influential studies were identified) and the pooled median PFS for the doublet group was 4.05 months (95% CI 3.54–4.57) (I 2 = 30.87%, P for heterogeneity = 0.017, with 15 arms included, Egger’s test, P = 0.001, Figure 2D; 9 arms of studies were influential because median PFS diverged from others [31, 35, 37, 40, 43, 47–49]). The same trend was seen for the OS end point: the pooled median OS with single agents was 6.98 months (95% CI 6.19– 7.78) (I 2 = 68.50%; P for heterogeneity <0.001, with 20 arms included, Egger’s test, P = 0.003, Figure 2E). Akaza study was influential because median OS diverged from others [25]. The pooled median OS in the doublet group was 8.50 months (95% CI 7.35–9.64) (I 2 = 70.04%, P for heterogeneity <0.001, with 23 arms included, Egger’s test, P < 0.001, Figure 2F). No influential studies were identified. Funnel plots of the OS end point for the single-agent and doublet groups are provided as supplementary Figures S1A and B, available at Annals of Oncology online. Univariable analyses are provided in Table 4: the difference of median PFS was 1.60 months (95% CI 0.68–2.52, P = 0.001) and the difference of median OS was 1.46 months (95% CI −0.27 to 3.18, P = 0.097) in favor of the doublet chemotherapy. The results of the multivariable comparative analyses are also supplied in Table 4. While response and PFS differences were confirmed as statistically significant (P < 0.001 and 0.002, respectively), OS was not (median difference 0.90 months, range −0.74 to 2.54, P = 0.284). When considering only the taxane or vinflunine among the single-agent arms of studies, ORR was consistently significantly different (P < 0.001), but neither PFS (P = 0.134) nor OS (P = 0.547) was significantly different between the single-agent and doublet chemotherapy groups. We finally ran another multivariable model adjusting for ECOGPS 2 and visceral metastases, and the comparison between the two groups favored doublet chemotherapy as follows: P < 0.001 for both ORR probability and median PFS, P = 0.080 for median OS.
incidence of acute toxicities in the two groups A summary of grade 1–2 and grade 3–4 hematologic toxicity and of grade 3–4 peripheral neuropathy and nephrotoxicity is
doi:10.1093/annonc/mdv509 |
Downloaded from http://annonc.oxfordjournals.org/ by guest on December 23, 2015
published between 1995 and 2014. Figure 1 outlines the selection process and reasons for study exclusion. Table 1 presents the principal study characteristics and outcomes [6, 9–51]. A total of 44 studies accounting for 46 arms of trials (22 with single agent and 24 with doublets) were selected for the metaanalysis, all of them performed in the second-line setting. A total of 1910 patients were selected: 1202 had received a singleagent, including 486 vinflunine and 201 paclitaxel or docetaxel, and 708 had received a doublet chemotherapy, including cis- or carboplatin in 191, or a taxane in 602 patients. A summary of the study characteristics is provided in Table 2. There were four randomized trials, two in both groups, for both of which both arms were included in the doublet chemotherapy group and only one arm for both in the single-agent group. The median sample size of the trials was 45 (range 13–253) for single-agent and 32 (range 11–48) for doublet chemotherapy. Among the single-agent arms of trials, there were three (13.6%) with vinflunine and five (22.7%) with paclitaxel or docetaxel. Among the doublet chemotherapy arms, 20 (83.3%) were taxanecontaining and 4 (16.7%) carboplatin-containing regimens. Only two studies consisted of a cisplatin-based doublet. Patient-level characteristics were equally distributed between the two groups, except for a nonstatistically significant higher proportion of patients with ECOG-PS 2 in the doublet chemotherapy group [14% (range 0–23.8) versus 8.5% (range 0–57), Wilcoxon test, P = 0.675]. The pooled median follow-up was 17 months (range 1.8–21.5 months) in the single-agent group and 20.4 months (range 9–48 months) in the doublet chemotherapy group. The
Year
ECOG-PS 1 (%)
ECOG-PS 2 (%)
Prior CDDP (%)
Prior CBDCA (%)
Visceral metastases, N (%)
Hepatic metastases, N (%)
Treatment
N patients (evaluable for response)
Criteria of response assessment
Median age (years)
Single-agent chemotherapy Beer et al. [9] Bellmunt et al. [10] Culine et al. [11] Galsky et al. [12] Dreicer et al. [13] Joly et al. [14] Lorusso et al. [15] McCaffrey et al. [16] Sweeney et al. [17] Vaughn et al. [18] Vaughn et al. [19] Witte et al. [20] Winquist et al. [21] Witte et al. [22] Ko et al. [23] Lee et al. [24] Akaza et al. [25] Ponzato et al. [26] Papamichael et al. [27] Sengelov et al. [28] Choueiri et al. [6] Loriot et al. [29]
2008 2009 2006 2007 2007 2009 1998 1997 2006 2002 2009 1997 2005 1998 2013 2012 2007 1997 1997 1995 2012 2012
60 72 45 54 62 82 40 100 34 39 100 71 55 100 50 48 27.3 NA 50 47 52.8 NA
10 0 0 0 7 18 57 0 4.3 13 0 29 20 0 16 14 4.5 NA 28 26 NA 0
NA 64.8 NA 31 NA NA 100 NA NA 45 NA 94 100 NA 69 NA 95 70 64 90 NA NA
NA 29.6 NA 69 NA NA 0 NA NA 29 NA NA 0 NA NA NA 0 30 0 0 NA NA
72 74 NA 61 NA 58 NA NA 42.5 77 NA NA NA 96 73 62 NA NA 71 61 76.4 NA
NA NA NA NA 26 38 23 27 NA 45 50.3 27 25 32 31 30 13.6 15 43 57 37.5 NA
CPT-11 VFL VFL PEM IXB T G TXT PEM T VFL I O TOP NabT Genexol-PM G I T MMC TXT PTX
45 (40) 253 (253) 58 (51) 13 (12) 45 (42) 50 (45) 35 (31) 31 (30) 47 (47) 31 (31) 175 (151) 60 (56) 20 (18) 46 (44) 48 (47) 37 (34) 46 (44) 20 (20) 14 (14) 23 (19) 75 (72) 30 (30)
RECIST 1.0 RECIST 1.0 WHO RECIST 1.0 RECIST 1.0 RECIST 1.0 WHO WHO WHO WHO WHO ECOG WHO ECOG RECIST 1.0 RECIST 1.0 JUA/JSP WHO WHO WHO RECIST 1.0 NA
64.4 NA 63 69 63 64 64 61 64 66 66 67 64 62 66 57 65 68 68 61 NA NA
Author
Year
Median treatment duration (months) 2 NA 3 2 2.2 2 1.6 2.1 2.1 2.1 2.3
Median N cycles
Median FUP (months)
ORR: %
ORR 95% CI
PFS: median (months)
PFS: CI
OS: median (months)
OS 95% CI
3 NA 4 3 3 2 1.6 3 3 3 3
33 21.5 20 5.8 20 NA NA NA 9.2 NA 11.9
5 8.6 18 8 11.9 9 22.5 13.3 27.7 10 14.6
1.0–17.0 NA 8.4–30.9 NA 5.3–26.5 2.0–21.0 8.0–37.0 3.8–30.7 15.6–42.6 0–20 9.4–21.2
2.1 3 3 NA 2.7 3.2c 3.8c NA 2.9c 2.2c 2.8
1.8–2.3 NA 2.4–3.8 NA NA 0.2–7.5d NA NA 1.7–4.6 NA 2.6–3.8
5.4 6.9 6.6 NA 8 6.9 5 9 9.6 7.2 8.2
3.4–7.1 5.7–8.0 4.8–7.6 NA NA 0.36–18.3d 2–21d 6–12 5.1–14.6 NA 6.8–9.6
2008 2009 2006 2007 2007 2009 1998 1997 2006 2002 2009
Downloaded from http://annonc.oxfordjournals.org/ by guest on December 23, 2015
Beer et al. [9] Bellmunt et al. [10] Culine et al. [11] Galsky et al. [12] Dreicer et al. [13] Joly et al. [14] Lorusso et al. [15] McCaffrey et al. [16] Sweeney et al. [17] Vaughn et al. [18] Vaughn et al. [19]
Annals of Oncology
Volume 27 | No. 1 | January 2016
Author
reviews
| Raggi et al.
Table 1. Patient, disease, and treatment characteristics of included studies
Annals of Oncology
NA 1.5 1.5 2.53 3 3 2.3 NA NA 1.5 2
NA 2 2 6 4 3 3 NA 3 2 2
16 NA 20 25 10 18 NA 1.8 NA 7.1 NA
20 5 9.1 27.7 21 25 5 7 5 11.1 16.6
10.0–32.0 NA 2.9–25.5 17.3–44.4 7.0–34.0 13.2–40.3 NA 2.0–12.0 NA NA NA
2.4c 1.4c 1.5c 6 2.7c 3.1 6c NA NA 1.58 4
NA 0.6–4.8d NA 3.9–8.5 0.9–4.6 2–4.1 1–8d NA NA NA 2.1–4.5
5.3 6.9 6.7 10.8 6.5 12.6 8 NA 4 7.03 9.3
NA 0.3–24.2d NA 5.8–16.9 5–8 7.8–14.4 2–12d NA 1.3–15.9d NA 5.6–13.2
Author
Year
ECOG-PS 1 (%)
ECOG-PS 2 (%)
Prior CDDP (%)
27 100 45 NA 28 NA NA 27 60 NA NA NA
38 0 0 NA 22 NA NA 32 0 NA NA 23
NA 100 100 NA 100 100 100 95 71 98 42 95
Visceral metastases, N (%) 88 78 66 NA 100 42 29 63 69 NA 61 64
Hepatic metastases, N (%) 21 43 15.2 50 50 NA NA 36 NA 7 19 NA
Treatment
2001 2007 2011 2004 2009 2010 2010 2001 2011 2001 1999 2005
Prior CBDCA (%) NA 0 0 NA 0 0 0 5 29 2 0 NA
GI GI PG FOLFOX-4 T-CTX TG-1a TG-2a TXT-I EPI-T TG TI TC
N patients (evaluable for response) 34 (34) 23 (23) 33 (33) 16 (16) 32 (32) 48 (40) 48 (41) 22 (20) 35 (35) 41 (40) 26 (13) 44 (44)
Criteria of response assessment WHO WHO RECIST 1.1 WHO RECIST 1.0 RECIST 1.0 RECIST 1.0 WHO RECIST 1.0 WHO WHO WHO
Median age (years) 61 66 66 63 66 63.9 65.1 61 64 62 66 64.6
2008 2009 2011 2007 2007 2003 2006 2006 2007 2004 2014 2009
40 100 45.8 100 87 66 NA NA 89 36 57.1 18
20 0 20.8 0 13 14 NA NA 4 28 23.8 0
100 100 100 100 100 35 100 100 100 NA NA NA
0 0 0 0 0 65 0 0 0 NA NA NA
NA NA 63 NA 71 NA NA NA 75 65 86.7 73
30 9 17 33 39 21 14 8 29 NA 28.6 27
TG TG TG TC TC TXT-G TG-3wb TG-2wb TP TC T-CTX TXT-O
20 (20) 33 (30) 24 (24) 18 (18) 35 (31) 31 (29) 14 (14) 13 (13) 28 (24) 33 (33) 46 (42) 11 (11)
RECIST 1.0 RECIST 1.0 RECIST 1.1 WHO RECIST 1.0 ECOG WHO WHO WHO WHO RECIST 1.1 RECIST 1.0
62.9 66.1 64.5 62 67 64 62 4.5 61 70 65 65
Pectasides et al. [30] Lin et al. [31] Gondo et al. [32] Di Lorenzo et al. [33] Di Lorenzo et al. [34] Albers et al. [35]
doi:10.1093/annonc/mdv509 |
Krege et al. [36] Rozzi et al. [37] Sternberg et al. [38] Sweeney et al. [39] Vaishampayan et al. [40] Kanai et al. [41] Suyama et al. [42] Ikeda et al. [43] Soga et al. [44] Kouno et al. [45] Dreicer et al. [46] Fechner et al. [47] Uhm et al. [48] Friedland et al. [49] Park and Lee [50] Srinivas and Harshman [51]
Continued
reviews
1997 2005 1998 2013 2012 2007 1997 1997 1995 2012 2012
Downloaded from http://annonc.oxfordjournals.org/ by guest on December 23, 2015
Volume 27 | No. 1 | January 2016
Witte et al. [20] Winquist et al. [21] Witte et al. [22] Ko et al. [23] Lee et al. [24] Akaza et al. [25] Ponzato et al. [26] Papamichael et al. [27] Sengelov et al. [28] Choueiri et al. [6] Loriot et al. [29] Doublet chemotherapy
Author
Year
ECOG-PS 1 (%)
ECOG-PS 2 (%)
Prior CDDP (%)
Prior CBDCA (%)
Visceral metastases, N (%)
Hepatic metastases, N (%)
Treatment
N patients (evaluable for response)
Criteria of response assessment
Author
Year
Median N cycles
Median FUP (months)
ORR: %
ORR 95% CI
PFS: median (months)
PFS: CI
OS: median (months)
OS 95% CI
2001 2007 2011 2004 2009 2010 2010 2001 2011 2001 1999 2005
Median treatment duration (months) 3 3 3 NA 4.5 NA NA 1.5 3 4 2 2.45
4 3 4 NA 6.2 3 4 2 4 6 3 3.5
24 18 30 NA 9 40 40 NA NA NA 48 24
21 22 39.4 19 31 37.5 41.4 25 29 60 15.4 16
9.0–38.0 5.0–39.0 NA NA 17.0–45.0 NA NA NA NA 45.0–75.0 2.0–45.0 7.0–30.0
4c 3.5 NA NA 5c 4 3.1 NA 7.6c NA NA 4
0.52–21.6d 2.8–4.2 NA NA NA NA NA NA 3.2–10.7 NA NA 3–5
9 4.8 10.5 4 8 7.8 8 NA 12.6 14.4 8 6
0.5–28d 2.9–6.6 3–29.9d 2–8d 4–14 4.2–11.4 4.9–11.1 NA 4.6–18.8 2–43d 1–53d 5–8
2008 2009 2011 2007 2007 2003 2006 2006 2007 2004 2014 2009
5.3 3.5 3 3 NA NA 4.5 9 3.5 1.5 1.5 1.5
7.7 5 4 4 10 NA 6 11 5 2 2 2
11.9 15 20.4 20 7.8 NA 11 38 16.4 48 24 45
30 33.3 42 33 32.3 17.2 50 NA 36 24 33.3 18
10.0–50.0 19.2–51.2 NA NA 15.8–48.7 7.0–33.0 NA 6c 9.0–41.0 10.0–39.0 19.0–47.6 NA
NA NA 6.1 4 3.7 NA 11c 1–15d 6.2c 3.6c 3c NA
NA NA 0.5–23.9d NA NA NA 3–41d 9 3.9–8.5 1–29d 1.7–4.3 NA
11.5 11.3 12.4 11 7.9 7.7 13 0–16d 10.3 10.3 6.3 7
2–22d 7.2–13.6 0.5–30.2d NA NA NA 5.46d
Pectasides et al. [30] Lin et al. [31] Gondo et al. [32] Di Lorenzo et al. [33] Di Lorenzo et al. [34] Albers et al. [35] Krege et al. [36] Rozzi et al. [37] Sternberg et al. [38] Sweeney et al. [39] Vaishampayan et al. [40] Kanai et al. [41] Suyama et al. [42] Ikeda et al. [43] Soga et al. [44] Kouno et al. [45] Dreicer et al. [46] Fechner et al. [47]
6.1–14.1 1–33d 4.6–8 2–48d
a
Short-term (1) versus prolonged (2) course of chemotherapy. Two- versus 3-weekly schedule. c Numbers are indicating the time to progression. d Numbers are indicating the range instead of the confidence interval. CBDCA, carboplatin; CDDP, cisplatin; CI, confidence interval; CPT-11, irinotecan; CTX, cyclophosphamide; ECOG-PS, Eastern Cooperative Oncology Group Performance Status; EPI, epirubicin; FOLFOX, fluorouracil, leucovorin, oxaliplatin; FUP, follow-up; GI, gemcitabine, ifosfamide; IXB, ixabepilone; MMC; mitomycin C; NA, not available; NabT, nab-paclitaxel; O, oxaliplatin; ORR, objective response rate; PEM, pemetrexed; PFS, progression-free survival; PG, cisplatin, gemcitabine; PTX, pralatrexate; OS, overall survival; RECIST, Response Evaluation Criteria in Solid Tumors; TC, paclitaxel, carboplatin; TG, paclitaxel, gemcitabine; TI, paclitaxel, ifosfamide; TOP, topotecan; TP; paclitaxel, cisplatin; TTP, time-to-progression; TXT, docetaxel; VFL, vinflunine; WHO, World Health Organization. b
Annals of Oncology
Downloaded from http://annonc.oxfordjournals.org/ by guest on December 23, 2015
Volume 27 | No. 1 | January 2016
Uhm et al. [48] Friedland et al. [49] Park and Lee [50] Srinivas and Harshman [51]
Median age (years)
reviews
| Raggi et al.
Table 1. Continued
reviews
Annals of Oncology
groups in regard to the grade 3–4 toxicities. Univariably, there were no statistically significant differences between the two groups for any type of toxicity. Results were confirmed after adjusting the P values for ECOG-PS 2 and visceral metastases (data not shown). Three studies reported on drug-related death events, two with single-agents [Dreicer et al. (n = 1), Winquist et al. (n = 2)] [13, 21] and one with doublets [Albers et al. (n = 2)] [35]. We were not able to find any data regarding longterm side-effects for any treatment administered.
detailed for each trial in supplementary Table S1, available at Annals of Oncology online. Table 5 presents the univariable comparison between single-agent and doublet chemotherapy
Table 2. Summary of the characteristics of single-agent and doublet chemotherapy arms of studies Arms with singleagent (N = 22)
discussion 7 (31.8) 4 (18.2) 8 (36.4) 3 (13.6)
4 (16.7) 6 (25.0) 13 (54.2) 1 (4.2)
2 (9.1) 20 (90.9)
4 (16.7) 20 (83.3)
– – 3 (13.6) 2 (9.0) 3 (13.6) 45 (13–253)
2 (9.1) 4 (16.7) 17 (70.8) 3 (12.5) – 32 (11–48)
64 (57–69)
64 (61–70)
69.1 (30.4–97.1) 2 (9.0)
65.7 (18.2–100) 8 (33.3)
8.5 (0–57) 2 (9.0)
14 (0–23.8) 7 (29.2)
71.5 (61–96) 10 (45.4)
69 (35–100) 9 (37.5)
30.5 (13.6–50.3) 6 (27.2)
In this trial-level meta-analysis of second-line chemotherapy for UC, unique to the authors’ knowledge and resembling a huge number of patients, we found statistically significant differences in ORR and median PFS favoring the use of two-drug regimens over single-agents. The difference of median OS between the two groups was trending toward an improvement with the use of doublet chemotherapy regimens, but this was not statistically significant. However, when analyzing only the single-agent arms with vinflunine and taxanes ( paclitaxel or docetaxel), which are the most utilized agents or the standard of care (i.e. vinflunine, but only in Europe) outside of clinical trials, the significant difference for ORR was maintained but it was lost for PFS. First of all, the inherent limitations in running such metaanalyses are to be accounted for when interpreting the results, and the analyses should be taken as exploratory. The majority of included studies were actually single-arm trials, often with a small sample size. Most importantly, the heterogeneity in reporting the outcomes across trials (e.g. median with 95% CI in some cases, median with ranges in others, PFS in some cases and TTP in others, as outlined in Table 1) did represent the most challenging issue when pooling the data together. Available literature is limited in regard to performing this type of meta-analyses and principal references have been mostly supplied herein. This is the reason that we have included sensitivity analyses with the aim to identify possible influential studies/treatments that might have accounted for major biases for each end point. The outcomes were consistent throughout these analyses and only one influential arm has been reported for OS. As a further limitation, we were unable to find information on baseline hemoglobin,
27.8 (10–50) 6 (25)
a
Numbers are accounting for the study arms. ECOG-PS, Eastern Cooperative Oncology Group performance status.
Table 3. Subgroup analyses of studies investigating single-agent or doublet second-line chemotherapy in advanced urothelial cancer Study selection
Single-agent chemotherapy Vinflunine Paclitaxel or docetaxel Doublet chemotherapy Doublet with cisplatin Doublet without cisplatin Doublet with carboplatin
ORR No. of evaluable arms of studies
Probability % (95% CI)
PFS No. of evaluable arms of studies
Median PFS (95% CI)
OS No. of evaluable arms of studies
Median OS (95% CI)
22 3 5 24 2 22 4
14.2 (11.1–17.9) 11.7 (6.2–20.9) 10.5 (6.9–15.8) 31.9 (27.3–36.9) 40.4 (28.5–53.5) 30.9 (26.1–36.3) 25.4 (17.9–34.7)
18 3 3 15 1 14 4
2.69 (2.25–3.12) 2.92 (2.55–3.29) 2.20 (1.36–3.04) 4.05 (3.54–4.57) 6.20 (3.95–8.45) 3.79 (3.40–4.17) 3.86 (3.20–4.51)
20 3 4 23 2 21 4
6.98 (6.19–7.78) 7.20 (6.30–8.10) 7.35 (6.16–8.55) 8.50 (7.35–9.64) 10.39 (7.53–13.26) 8.35 (7.15–9.55) 8.14 (5.76–10.52)
CI, confidence interval; ORR, objective response rate; PFS, progression-free survival; OS, overall survival.
Volume 27 | No. 1 | January 2016
doi:10.1093/annonc/mdv509 |
Downloaded from http://annonc.oxfordjournals.org/ by guest on December 23, 2015
Study-level characteristica Publication year, N (%) 1995–2001 2002–2006 2007–2011 2012–2014 Type of study arms, N (%) Randomized Single-group Treatment, N (%) Cisplatin Carboplatin Paclitaxel Docetaxel Vinflunine Sample size (median, range) Patient-level characteristic Age (years; median, range) ECOG-PS ≥1 % (median, range) Not reported (N arms, %) ECOG-PS = 2 % (median, range) Not reported (N arms, %) Visceral metastases % (median, range) Not reported (N arms, %) Hepatic metastases % (median, range) Not reported (N arms, %)
Arms with doublet (N = 24)
reviews
Annals of Oncology
A 0.25 [0.14, 0.40] 0.05 [0.01, 0.18] 0.06 [0.04, 0.10] 0.11 [0.06, 0.21] 0.18 [0.09, 0.31] 0.12 [0.05, 0.26] 0.08 [0.01, 0.41] 0.09 [0.03, 0.21] 0.28 [0.17, 0.42] 0.21 [0.10, 0.37] 0.17 [0.07, 0.34] 0.23 [0.11, 0.40] 0.13 [0.05, 0.31] 0.07 [0.01, 0.37] 0.05 [0.01, 0.28] 0.05 [0.01, 0.29] 0.28 [0.17, 0.42] 0.10 [0.03, 0.26] 0.15 [0.10, 0.21] 0.06 [0.01, 0.31] 0.20 [0.11, 0.32] 0.09 [0.03, 0.22]
RE model
0.14 [0.11, 0.18] 0.00
0.10
0.20 0.30 0.40 Response probability
0.50
B Albers P, 2010 Albers P, 2010.1 Di Lorenzo G, 2004 Di Lorenzo G, 2009 Dreicer R, 2003 Fechner G, 2006 Fechner G, 2006.1 Friedland DM, 2004 Gondo T, 2011 Ikeda M, 2011 Kanai K, 2008 Kouno T, 2007 Krege S, 2001 Lin C-C, 2007 Park JH, 2014 Pectasides D, 2001 Rozzi A, 2011 Soga N, 2007 Srinivas S, 2009 Sternberg C, 2001 Suyama T, 2009 Sweeney CJ, 1999 Uhm JE, 2007 Vaishampayan UN, 2005
0.37 [0.24, 0.53] 0.41 [0.28, 0.57] 0.19 [0.06, 0.45] 0.31 [0.18, 0.49] 0.17 [0.07, 0.35] 0.50 [0.26, 0.74] 0.38 [0.17, 0.66] 0.24 [0.13, 0.42] 0.39 [0.24, 0.57] 0.42 [0.24, 0.62] 0.30 [0.14, 0.53] 0.32 [0.18, 0.50] 0.25 [0.11, 0.48] 0.22 [0.09, 0.43] 0.33 [0.21, 0.49] 0.21 [0.10, 0.37] 0.29 [0.16, 0.45] 0.33 [0.16, 0.57] 0.09 [0.01, 0.44] 0.60 [0.44, 0.74] 0.33 [0.19, 0.52] 0.15 [0.04, 0.45] 0.42 [0.24, 0.62] 0.16 [0.08, 0.30]
RE model
0.32 [0.27, 0.37] 0.00
0.20 0.40 0.60 Response probability
0.80
Figure 2. Forest plots: (A) objective response rate, single-agent studies; (B) objective response rate, doublet chemotherapy studies; (C) progression-free survival, single-agent studies; (D) progression-free survival, doublet chemotherapy studies; (E) overall survival, single-agent studies; (F) overall survival, doublet chemotherapy studies.
| Raggi et al.
Volume 27 | No. 1 | January 2016
Downloaded from http://annonc.oxfordjournals.org/ by guest on December 23, 2015
Akaza H, 2007 Beer T M, 2008 Bellmunt J, 2009 Choueiri TK, 2012 Culine S, 2006 Dreicer R, 2007 Galsky M.D., 2007 Joly F, 2009 Ko YJ, 2013 Lee J-L, 2011 Loriot Y, 2012 Lorusso V, 1998 McCaffrey JA, 1997 Papamichael D, 1997 Ponzato P, 1997 Sengelov L, 1995 Sweeney CJ, 2006 Vaughn DJ, 2002 Vaughn DJ, 2009 Winquist E, 2005 Witte RS, 1997 Witte RS, 1998
reviews
Annals of Oncology
C 3.10 [2.07, 4.13]
Beer T M, 2008
2.10 [1.86, 2.34]
Bellmunt J, 2009
3.00 [2.33, 3.67]
Choueiri TK, 2012
1.58 [1.29, 1.87]
Culine S, 2006
3.00 [2.31, 3.69]
Dreicer R, 2007
2.70 [1.95, 3.45]
Joly F, 2009
3.20 [2.12, 4.28]
Joung Ko Y, 2013
6.00 [3.75, 8.25]
Lee J-L, 2011
2.70 [0.89, 4.51]
Loriot Y, 2012
4.00 [2.82, 5.18]
Lorusso V, 1998
3.80 [2.24, 5.36]
Ponzato P, 1997
6.00 [3.32, 8.68]
Sweeney CJ, 2006
2.90 [1.48, 4.32]
Vaughn DJ, 2002
2.20 [1.69, 2.71]
Vaughn DJ, 2009
2.80 [2.21, 3.39]
Winquist E, 2005
1.40 [1.13, 1.67]
Witte RS, 1997
2.40 [1.82, 2.98]
Witte RS, 1998
1.50 [1.22, 1.78]
RE model
2.69 [2.25, 3.12]
0.00
2.00
4.00 6.00 Median PFS
8.00
10.00
D Albers P, 2010
4.00 [0.08, 7.92]
Albers P, 2010.1
3.10 [1.97, 4.23]
Di Lorenzo G, 2009
5.00 [2.31, 7.69]
Fechner G, 2006
11.00 [6.25, 15.75]
Fechner G, 2006.1
6.00 [3.29, 8.71]
Friedland DM, 2004
3.60 [2.18, 5.02]
Ikeda M, 2011
6.10 [3.41, 8.79]
Kouno T, 2007
3.70 [2.21, 5.19]
Lin C-C, 2007
3.50 [2.81, 4.19]
Park JH, 2014
3.00 [1.73, 4.27]
Pectasides D, 2001
4.00 [2.29, 5.71]
Rozzi A, 2011
7.60 [3.93, 11.27]
Soga N, 2007
4.00 [2.24, 5.76]
Uhm JE, 2007
6.20 [3.95, 8.45]
Vaishampayan UN, 2005
4.00 [3.02, 4.98]
RE model
4.05 [3.54, 4.57]
0.00
5.00
10.00 Median PFS
15.00
20.00
Fig. 2 Continued
albumin levels, and the time from prior chemotherapy. Each one of these factors has a recognized role in the salvage setting [2–4]. Although multivariable analyses were adjusted for ECOGPS and hepatic metastases only, it is unlikely that the present
Volume 27 | No. 1 | January 2016
findings would have been modified by adjusting for additional confounders. Of course, we could not exclude that other confounders could have impacted the results beyond what could have been adjusted by the regression model. For example, the
doi:10.1093/annonc/mdv509 |
Downloaded from http://annonc.oxfordjournals.org/ by guest on December 23, 2015
Akaza H, 2007
reviews
Annals of Oncology
E
RE model
6.98 [6.19, 7.78] 0.00
5.00
10.00 Median OS
15.00
20.00
F Albers P, 2010 Albers P, 2010.1 Di Lorenzo G, 2004 Di Lorenzo G, 2009 Dreicer R, 2003 Fechner G, 2006 Fechner G, 2006.1 Friedland DM, 2004 Gondo T, 2011 Ikeda M, 2011 Kanai K, 2008 Kouno T, 2007 Lin C-C, 2007 Park JH, 2014 Pectasides D, 2001 Rozzi A, 2011 Soga N, 2007 Srinivas S, 2009 Sternberg C, 2001 Suyama T, 2009 Sweeney CJ, 1999 Uhm JE, 2007 Vaishampayan UN, 2005
7.80 [4.27, 11.33] 8.00 [4.96, 11.04] 4.00 [2.85, 5.15] 8.00 [3.10, 12.90] 7.70 [4.80, 10.60] 13.00 [7.74, 18.26] 9.00 [5.08, 12.92] 10.30 [6.16, 14.44] 10.50 [6.29, 14.71] 12.40 [7.51, 17.29] 11.50 [6.75, 16.25] 7.90 [4.93, 10.87] 4.80 [2.99, 6.61] 6.30 [4.63, 7.97] 9.00 [5.41, 12.59] 12.60 [5.64, 19.56] 11.00 [6.36, 15.64] 7.00 [4.31, 9.69] 14.40 [9.37, 19.43] 11.30 [8.16, 14.44] 8.00 [4.85, 11.15] 10.30 [6.38, 14.22] 6.00 [4.53, 7.47]
RE model
8.50 [7.35, 9.64] 0.00
5.00
10.00 Median OS
15.00
20.00
Fig. 2 Continued
disproportionate representation of ECOG-PS 2 patients, being more frequent in the doublet chemotherapy group, was a concern and it was somewhat surprising, although an exploratory
| Raggi et al.
multivariable analysis that included ECOG-PS 2 was made and results did not change substantially (OS difference still not significant).
Volume 27 | No. 1 | January 2016
Downloaded from http://annonc.oxfordjournals.org/ by guest on December 23, 2015
12.60 [9.42, 15.78] 5.40 [3.59, 7.21] 6.90 [5.77, 8.03] 7.03 [4.97, 9.09] 6.60 [5.23, 7.97] 8.00 [5.09, 10.91] 6.90 [4.68, 9.12] 10.80 [5.36, 16.24] 6.50 [5.03, 7.97] 9.30 [5.58, 13.02] 5.00 [3.59, 6.41] 9.00 [6.06, 11.94] 8.00 [4.77, 11.23] 4.00 [2.88, 5.12] 9.60 [4.95, 14.25] 7.20 [4.61, 9.79] 8.20 [6.83, 9.57] 6.90 [4.38, 9.42] 5.30 [3.92, 6.68] 6.70 [4.55, 8.85]
Akaza H, 2007 Beer T M, 2008 Bellmunt J, 2009 Choueiri TK, 2012 Culine S, 2006 Dreicer R, 2007 Joly F, 2009 Ko YJ, 2013 Lee J-L, 2011 Loriot Y, 2012 Lorusso V, 1998 McCaffrey JA, 1997 Ponzato P, 1997 Sengelov L, 1995 Sweeney CJ, 2006 Vaughn DJ, 2002 Vaughn DJ, 2009 Winquist E, 2005 Witte RS, 1997 Witte RS, 1998
reviews
Annals of Oncology
Table 4. Pooled estimates from the univariable and multivariable analyses
Univariable analyses Doublet versus single-agent Multivariable analyses Doublet versus single-agent Doublet versus single-agent paclitaxel or docetaxel or vinflunine Doublet cisplatin or carboplatin versus other doublets
ORR Odds ratio (95% CI) P value
PFS median, difference (95% CI) P value
OS median, difference (95% CI) P value
2.78 (1.95 to 3.94) P < 0.001
1.60 (0.68 to 2.52) P = 0.001
1.46 (−0.27 to 3.18) P = 0.097
2.53 (1.79 to 3.58) P < 0.001 3.20 (1.94 to 5.28) P < 0.001 1.76 (0.98 to 3.15) P = 0.058
1.60 (0.57 to 2.63) P = 0.002 1.10 (−0.34 to 2.55) P = 0.134 0.78 (−1.46 to 2.02) P = 0.495
0.90 (−0.74 to 2.54) P = 0.284 −0.64 (−2.72 to 1.44) P = 0.547 2.34 (−0.36 to 5.05) P = 0.089
CI, confidence interval; ORR, objective response rate; PFS, progression-free survival; OS, overall survival.
Anemia Neutropenia Thrombocytopenia Peripheral neuropathy Nephrotoxicity
Single-agent Number of evaluable arms of studies
Probability % (95% CI)
Doublet Number of evaluable arms of studies
Probability % (95% CI)
P value*
22 22 22 15 13
9.7 (5.6–16.1) 11.6 (5.5–23.3) 7.5 (4.3–12.6) 3.6 (1.9–6.7) 3.0 (1.3–6.7)
24 24 24 16 13
14.5 (10.1–20.3) 16.8 (9.0–29.3) 8.9 (5.5–14.1) 5.5 (3.0–9.7) 3.1 (1.6–5.8)
0.438 0.531 0.781 0.431 0.699
*Univariable model. CI, confidence interval.
It is plausible that a rational administration of doublet chemotherapy to ECOG-PS 0 patients and of single-agents to ECOGPS 1–2 or at least PS 2 patients, or a larger dataset of randomized trials would have led to more clinically meaningful and statistically significant differences in survival between the two groups. Taken together, these results might help guiding the design of the next clinical trials in the salvage setting. We have provided the precise pooled estimates of median PFS and OS to be utilized when determining the advantage from adding or comparing a new drug to standard chemotherapy. Although the comparison of median OS was not statistically significant in our study, these benchmark estimates may be considered separately in randomized trials with single- or two-drug combination in the standard therapy arm. This separation might best apply to set the null hypothesis of phase II trials, for which the ORR and PFS improvement (which were both statistically significantly different here between the two groups) are the end points. Such benchmark estimates are as follows: median PFS of 2.69 months for the single-agents and 4.05 months for the doublet chemotherapy, and median OS of 6.98 months for single-agents and 8.50 months for doublet chemotherapy. Moving outside of clinical trials, some recommendations could be made for clinical practice on the basis of the present
Volume 27 | No. 1 | January 2016
findings. The outcomes with the use of paclitaxel, docetaxel, or vinflunine are unlikely to be significantly improved by adding a second chemotherapeutic drug even when it is feasible. Furthermore, the promising results with the use of the nanoparticle albumin-bound (nab)-paclitaxel lend further confidence to the single-agent strategy. Nab-paclitaxel was investigated as second-line chemotherapy in a Canadian multicenter, singlearm, phase II study [23]. The overall response rate was 27.7% and, most importantly, median OS was 10.8 months despite a high proportion of patients with baseline poor prognostic features. Yet investigating nab-paclitaxel in combination regimens (i.e. with gemcitabine/carboplatin chemotherapy) exhibited significant myelosuppression in the neoadjuvant setting which needs to be carefully addressed if evaluating the same combination in the second line [52]. Major tolerability issues also characterized the efforts of combining vinflunine or pemetrexed with doublet chemotherapy [53], and consequently investigations of these compounds are mainly conceivable in sequential or maintenance strategies. Substantial uncertainties still lie regarding the use of doublet chemotherapy regimens in the overall population of patients failing platinum-based chemotherapy, but certainly the patient selection is the key to investigate combination chemotherapy further. A few additional studies have been published in
doi:10.1093/annonc/mdv509 |
Downloaded from http://annonc.oxfordjournals.org/ by guest on December 23, 2015
Table 5. Summary of the incidence of grade 3–4 adverse events between single-agent and doublet chemotherapy groups
reviews
| Raggi et al.
significant improvement of PFS and the trending-to-significance improvement of OS are suggesting a potentially meaningful benefit from combining agents with proven activity in UC. In addition to the development of novel compounds such as PD1 and PD-L1 inhibitors as single agents, the prospective evaluation of tolerable combinations of chemotherapeutic drugs as well as chemobiologic combinations is rational.
disclosure The authors have declared no conflicts of interest.
references 1. Sonpavde G, Jones BS, Bellmunt J et al. Future directions and targeted therapies in bladder cancer. Hematol Oncol Clin North Am 2015; 29: 361–376. 2. Bellmunt J, Choueiri TK, Fougeray R et al. Prognostic factors in patients with advanced transitional cell carcinoma of the urothelial tract experiencing failure with platinum-containing regimens. J Clin Oncol 2010; 28: 1850–1855. 3. Sonpavde G, Pond GR, Fougeray R et al. Time from prior chemotherapy enhances prognostic risk grouping in the second-line setting of advanced urothelial carcinoma: a retrospective analysis of pooled, prospective phase 2 trials. Eur Urol 2013; 63: 717–723. 4. Sonpavde G, Pond GR, Rosenberg JE et al. Improved 5-factor classification of patients receiving salvage systemic therapy for advanced urothelial carcinoma. J Urol 2015 Aug 17 [epub ahead of print], doi: 10.1016/j.juro.2015.07.111. 5. Liberati A, Altman DG, Tetzlaff J et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLoS Med 2009; 6: e1000100. 6. Choueiri TK, Ross RW, Jacobus S et al. Double-blind, randomized trial of docetaxel plus vandetanib versus docetaxel plus placebo in platinum-pretreated metastatic urothelial cancer. J Clin Oncol 2012; 30: 507–512. 7. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ 2003; 327: 557–560. 8. Durrleman S, Simon R. Flexible regression models with cubic splines. Stat Med 1989; 8: 551–556. 9. Beer TM, Goldman B, Nichols CR et al. Southwest Oncology Group phase II study of irinotecan in patients with advanced transitional cell carcinoma of the urothelium that progressed after platinum-based chemotherapy. Clin Genitourin Cancer 2008; 6: 36–39. 10. Bellmunt J, Theodore C, Demkov T et al. Phase III trial of vinflunine plus best supportive care compared with best supportive care alone after a platinumcontaining regimen in patients with advanced transitional cell carcinoma of the urothelial tract. J Clin Oncol 2009; 27: 4454–4461. 11. Culine S, Theodore C, De Santis M et al. A phase 2 study of vinflunine in bladder cancer patients progressing after first-line platinum-containing regimen. Br J Cancer 2006; 94: 1395–1401. 12. Galsky MD, Mironov S, Iasonos A et al. Phase II trial of pemetrexed as second-line therapy in patients with metastatic urothelial carcinoma. Invest New Drugs 2007; 25: 265–270. 13. Dreicer R, Li S, Manola J et al. Phase 2 trial of epothilone B analog BMS-247550 (ixabepilone) in advanced carcinoma of the urothelium (E-3800): a trial of the Eastern Cooperative Oncology Group. Cancer 2007; 110: 759–763. 14. Joly F, Houédé N, Noal S et al. Do patients with advanced urothelial carcinoma benefit from weekly paclitaxel chemotherapy? A GETUG phase II study. Clin Genitourin Cancer 2009; 7: E28–E33. 15. Lorusso V, Pollera CF, Antimi M et al. A phase II study of gemcitabine in patients with transitional cell carcinoma of the urinary tract previously treated with platinum. Italian Co-operative Group on Bladder Cancer. Eur J Cancer 1998; 34: 1208–1212. 16. McCaffrey JA, Hilton S, Mazumdar M et al. Phase II trial of docetaxel in patients with advanced or metastatic transitional-cell carcinoma. J Clin Oncol 1997; 15: 1853–1857.
Volume 27 | No. 1 | January 2016
Downloaded from http://annonc.oxfordjournals.org/ by guest on December 23, 2015
2015 and were not included in this analysis, like the mature results with the use of pemetrexed in the salvage setting or the results of standard arms of randomized studies with new drugs [54, 55]. Unfortunately, we could not identify any patient or treatment characteristics that may portend a clinical benefit from a certain treatment, due to the inherent limitations of the trial-level meta-analysis. It is possible that, at least for a small proportion of patients, the use of double or even triple combination chemotherapy may result in a significant OS advantage compared with what is expected from the single-agents. Actually, this OS advantage has emerged from a retrospective individual patient-level analysis of 370 multicentric patients. Sonpavde et al. reported incrementally significant improvements in OS with the addition of one or two agents to taxanes compared with taxanes alone, although the results are pending a validation through a prospective clinical trial [56]. This apparent superiority of taxane-based combination chemotherapy compared with single-agent taxanes was partially confirmed here, in particular for OR and PFS, but not for OS. Taxanes were largely represented in more than 80% of doublet chemotherapy arms. As single agents, the efficacy was clearly overlapping to that of other regimens (as provided in Table 3), and this is an indirect evidence corroborating the use of either taxanes or vinflunine as equally active single agents in clinical practice. The present meta-analysis focused on two-drug combinations only, because the use of triple chemotherapy regimens is rarely possible in clinical practice and very few studies are available. Indeed, the possibility to combine two or more chemotherapeutic agents in the salvage therapy for UC may be limited due to poor performance status or comorbidities in most cases. Although we did not find significant differences in the incidence of severe toxicities between the two groups, even after adjusting for some important clinical factors, the critical limitation results from analyzing patients who have been included in clinical trials, who may be quite different from the overall population of advanced and relapsing UC. Looking at the next steps, the advent of the immune checkpoint inhibitors targeting the programmed death-1/ligand-1 (PD1/PD-L1) is revolutionizing the salvage setting and several active compounds are shedding light on the horizon and will likely be granted FDA and EMA approval for this indication [57–59]. Pivotal phase III trials are ongoing with these drugs having single-agent chemotherapy in the comparator arm in each case. The same consideration may apply to other promising targeted agents that are in development for molecularly defined cohort of patients [1]. Of course, the association that we will observe between the clinical and biological characteristics of eligible patients will impact the design of the next clinical trials. The possibility to have potentially active compounds in the salvage setting may lead to an extended OS for at least biologically defined patients. For these patients, the advantage may even be greater with the use of chemobiologic doublets compared with the benchmark of single-agent chemotherapy, although this notion remains to be proved. In conclusion, in the present meta-analysis comparing singleagent to doublet second-line chemotherapy for UC, we identified a better activity of the latter in terms of ORR and PFS, but we did not find any statistically significant difference in OS. The general recommendation is to continue administering singleagent taxanes or vinflunine outside of clinical trials, although a
Annals of Oncology
Annals of Oncology
Volume 27 | No. 1 | January 2016
38. Sternberg CN, Calabrò F, Pizzocaro G et al. Chemotherapy with an every-2-week regimen of gemcitabine and paclitaxel in patients with transitional cell carcinoma who have received prior cisplatin-based therapy. Cancer 2001; 92: 2993–2998. 39. Sweeney CJ, Williams SD, Finch DE et al. A phase II study of paclitaxel and ifosfamide for patients with advanced refractory carcinoma of the urothelium. Cancer 1999; 86: 514–518. 40. Vaishampayan UN, Faulkner JR, Small EJ et al. Phase II trial of carboplatin and paclitaxel in cisplatin-pretreated advanced transitional cell carcinoma: a Southwest Oncology Group Study. Cancer 2005; 104: 1627–1632. 41. Kanai K, Kikuchi E, Ohigashi T et al. Gemcitabine and paclitaxel chemotherapy for advanced urothelial carcinoma in patients who have received prior cisplatin-based chemotherapy. Int J Clin Oncol 2008; 13: 510–514. 42. Suyama T, Ueda T, Fukasawa S et al. Combination of gemcitabine and paclitaxel as second-line chemotherapy for advanced urothelial carcinoma. Jpn J Clin Oncol 2009; 39: 244–250. 43. Ikeda M, Matsumoto K, Tabata K et al. Combination of gemcitabine and paclitaxel is a favorable option for patients with advanced or metastatic urothelial carcinoma previously treated with cisplatin-based chemotherapy. Jpn J Clin Oncol 2011; 41: 1214–1220. 44. Soga N, Onishi T, Arima K, Sugimura Y. Paclitaxel carboplatin chemotherapy as a second-line chemotherapy for advanced platinum resistant urothelial cancer in Japanese cases. Int J Urol 2007; 14: 828–832. 45. Kouno T, Ando M, Yonemori K et al. Weekly paclitaxel and carboplatin against advanced transitional cell cancer after failure of a platinum-based regimen. Eur Urol 2007; 52: 1115–1122. 46. Dreicer R, Manola J, Schneider DJ et al. Phase II trial of gemcitabine and docetaxel in patients with advanced carcinoma of the urothelium: a trial of the Eastern Cooperative Oncology Group. Cancer 2003; 97: 2743–2747. 47. Fechner G, Siener R, Reimann M et al. Randomised phase II trial of gemcitabine and paclitaxel second-line chemotherapy in patients with transitional cell carcinoma (AUO Trial AB 20/99). Int J Clin Pract 2006; 60: 27–31. 48. Uhm JE, Lim HY, Kim WS et al. Paclitaxel with cisplatin as salvage treatment for patients with previously treated advanced transitional cell carcinoma of the urothelial tract. Neoplasia 2007; 9: 18–22. 49. Friedland DM, Dakhil S, Hollen C et al. A phase II evaluation of weekly paclitaxel plus carboplatin in advanced urothelial cancer. Cancer Invest 2004; 22: 374–382. 50. Park JH, Lee JL. Intravenous 3-weekly paclitaxel and metronomic oral cyclophosphamide in patients with advanced urothelial cancer previously treated with gemcitabine and platinum. Cancer Chemother Pharmacol 2015; 75: 247–254. 51. Srinivas S, Harshman LC. A phase II study of docetaxel and oxaliplatin for secondline treatment of urothelial carcinoma. Chemotherapy 2009; 55: 321–326. 52. Grivas PD, Hussain M, Hafez K et al. A phase II trial of neoadjuvant nab-paclitaxel, carboplatin, and gemcitabine (ACaG) in patients with locally advanced carcinoma of the bladder. Urology 2013; 82: 111–117. 53. Hutson TE, Vukelja S, Atienza D et al. Phase I study of a 3-drug regimen of gemcitabine/cisplatin/pemetrexed in patients with metastatic transitional cell carcinoma of the urothelium. Invest New Drugs 2008; 26: 151–158. 54. Bambury RM, Benjamin DJ, Chaim JL et al. The safety and efficacy of single-agent pemetrexed in platinum resistant advanced urothelial carcinoma: a large single institution experience. Oncologist 2015; 20: 508–575. 55. Petrylak DP, Tagawa S, Kohli M et al. Interim results of a randomized phase 2 study of docetaxel with ramucirumab versus docetaxel in second-line advanced or metastatic urothelial carcinoma. J Clin Oncol 2015; 33(suppl 7): 295a. 56. Sonpavde G, Pond GR, Choueiri TK et al. Single agent taxane versus taxane containing combination chemotherapy as salvage therapy for advanced urothelial carcinoma. Eur Urol 2015 Aug 8 [epub ahead of print], doi: 10.1016/j.eururo.2015.07.042. 57. Powles T, Eder JP, Fine GD et al. MPDL3280A (anti-PD-L1) treatment leads to clinical activity in metastatic bladder cancer. Nature 2014; 515: 558–562. 58. Plimack ER, Bellmunt J, Gupta S et al. Pembrolizumab (MK-3475) for advanced urothelial cancer: updated results and biomarker analysis from KEYNOTE-012. J Clin Oncol 2015; 33(suppl): abstr 4502. 59. Petrylak DP, Powles T, Bellmunt J et al. A phase is study of MPDL3280A (antiPDL1): updated response and survival data in urothelial bladder cancer (UBC). J Clin Oncol 2015; 33(suppl): abstr 4501.
doi:10.1093/annonc/mdv509 |
Downloaded from http://annonc.oxfordjournals.org/ by guest on December 23, 2015
17. Sweeney CJ, Roth BJ, Kabbinavar FF et al. Phase II study of pemetrexed for second-line treatment of transitional cell cancer of the urothelium. J Clin Oncol 2006; 24: 3451–3457. 18. Vaughn DJ, Broome CM, Hussain M et al. Phase II trial of weekly paclitaxel in patients with previously treated advanced urothelial cancer. J Clin Oncol 2002; 20: 937–940. 19. Vaughn DJ, Srinivas S, Stadler WM et al. Vinflunine in platinum-pretreated patients with locally advanced or metastatic urothelial carcinoma: results of a large phase 2 study. Cancer 2009; 115: 4110–4117. 20. Witte RS, Elson P, Bono B et al. Eastern Cooperative Oncology Group phase II trial of ifosfamide in the treatment of previously treated advanced urothelial carcinoma. J Clin Oncol 1997; 15: 589–593. 21. Winquist E, Vokes E, Moore MJ et al. A phase II study of oxaliplatin in urothelial cancer. Urol Oncol 2005; 23: 150–154. 22. Witte RS, Manola J, Burch PA et al. Topotecan in previously treated advanced urothelial carcinoma: an ECOG phase II trial. Invest New Drugs 1998; 16: 191–195. 23. Ko YJ, Canil CM, Mukherjee SD et al. Nanoparticle albumin-bound paclitaxel for second-line treatment of metastatic urothelial carcinoma: a single group, multicentre, phase 2 study. Lancet Oncol 2013; 14: 769–776. 24. Lee JL, Ahn JH, Park SH et al. Phase II study of a cremophor-free, polymeric micelle formulation of paclitaxel for patients with advanced urothelial cancer previously treated with gemcitabine and platinum. Invest New Drugs 2012; 30: 1984–1990. 25. Akaza H, Naito S, Usami M et al. Efficacy and safety of gemcitabine monotherapy in patients with transitional cell carcinoma after cisplatin containing therapy: a Japanese experience. Jpn J Clin Oncol 2007; 37: 201–206. 26. Ponzato P, Vigani A, Pensa F et al. Second line chemotherapy with ifosfamide as outpatient treatment for advanced bladder cancer. Am J Clin Oncol 1997; 20: 519–521. 27. Papamichael D, Gallagher CJ, Oliver RT et al. Phase II study of paclitaxel in pretreated patients with locally advanced/metastatic cancer of the bladder and ureter. Br J Cancer 1997; 75: 606–607. 28. Sengelov L, Nielsen OS, Kamby C, von der Maase H. Systemic mitomycin C as secondline treatment for metastatic urothelial cancer. Acta Oncol 1995; 34: 978–979. 29. Loriot Y, Fizazi K, Carles J et al. Results of a phase II study of pralatrexate in patients with advanced/metastatic relapsed transitional cell carcinoma of the urinary bladder. J Clin Oncol 2012; 30(suppl): abstr 4574. 30. Pectasides D, Aravantinos G, Kalofonos H et al. Combination chemotherapy with gemcitabine and ifosfamide as second-line treatment in metastatic urothelial cancer. A phase II trial conducted by the Hellenic Cooperative Oncology Group. Ann Oncol 2001; 12: 1417–1422. 31. Lin CC, Hsu CH, Huang CY et al. Gemcitabine and ifosfamide as second-line treatment for cisplatin-refractory metastatic urothelial carcinoma: a phase II study. Anticancer Drugs 2007; 18: 487–491. 32. Gondo T, Ohori M, Hamada R et al. The efficacy and safety of gemcitabine plus cisplatin regimen for patients with advanced urothelial carcinoma after failure of M-VAC regimen. Int J Clin Oncol 2011; 16: 345–351. 33. Di Lorenzo G, Autorino R, Giordano A et al. FOLFOX-4 in pre-treated patients with advanced transitional cell carcinoma of the bladder. Jpn J Clin Oncol 2004; 34: 747–750. 34. Di Lorenzo G, Montesarchio V, Autorino R et al. Phase 1/2 study of intravenous paclitaxel and oral cyclophosphamide in pretreated metastatic urothelial bladder cancer patients. Cancer 2009; 115: 517–523. 35. Albers P, Park SI, Niegisch G et al. Randomized phase III trial of 2nd line gemcitabine and paclitaxel chemotherapy in patients with advanced bladder cancer: short-term versus prolonged treatment [German Association of Urological Oncology (AUO) trial AB 20/99]. Ann Oncol 2011; 22: 288–294. 36. Krege S, Rembrink V, Bogermann C et al. Docetaxel and ifosfamide as second line treatment for patients with advanced and metastatic urothelial cancer after failure of platinum chemotherapy: a phase 2 study. J Urol 2001; 165: 67–71. 37. Rozzi A, Salerno M, Bordin F et al. Weekly regimen of epirubicin and paclitaxel as second-line chemotherapy in patients with metastatic transitional cell carcinoma of urothelial tract: results of a phase II study. Med Oncol 2011; 28(suppl 1): S426–S432.
reviews
Annals of Oncology
Annals of Oncology 27: 49–61, 2016 doi:10.1093/annonc/mdv509 Published online 20 October 2015
Second-line single-agent versus doublet chemotherapy as salvage therapy for metastatic urothelial cancer: a systematic review and meta-analysis D. Raggi1, †, R. Miceli1,†, G. Sonpavde2, P. Giannatempo1, L. Mariani1, M. D. Galsky3, J. Bellmunt4 & A. Necchi1* 1 Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy; 2UAB Comprehensive Cancer Center, Birmingham; 3Mount Sinai School of Medicine, Tisch Cancer Institute, New York; 4Dana-Farber Cancer Institute and Harvard Medical School, Boston, USA
Background: The efficacy and safety of a combination of chemotherapeutic agent compared with single-agent chemotherapy in the second-line setting of advanced urothelial carcinoma (UC) are unclear. We aimed to study the survival impact of single-agent compared with doublet chemotherapy as second-line chemotherapy of advanced UC. Patients and methods: Literature was searched for studies including single-agent or doublet chemotherapy in the second-line setting after platinum-based chemotherapy. Random-effects models were used to pool trial-level data according to treatment arm, including median progression-free survival (PFS), overall survival (OS), objective response rate (ORR) probability, and grade 3–4 toxicity. Univariable and multivariable analyses, including sensitivity analyses, were carried out, adjusting for the percent of patients with ECOG performance status ≥1 and hepatic metastases. Results: Forty-six arms of trials including 1910 patients were selected: 22 arms with single agent (n = 1202) and 24 arms with doublets (n = 708). The pooled ORR with single agents was 14.2% [95% confidence interval (CI) 11.1–17.9] versus 31.9% [95% CI 27.3–36.9] with doublet chemotherapy. Pooled median PFS was 2.69 and 4.05 months, respectively. The pooled median OS was 6.98 and 8.50 months, respectively. Multivariably, the odds ratio for ORR and the pooled median difference of PFS were statistically significant (P < 0.001 and P = 0.002) whereas the median difference in OS was not (P = 0.284). When including single-agent vinflunine or taxanes only, differences were significant only for ORR (P < 0.001) favoring doublet chemotherapy. No statistically significant differences in grade 3–4 toxicity were seen between the two groups. Conclusions: Despite significant improvements in ORR and PFS, doublet regimens did not extend OS compared with single agents for the second-line chemotherapy of UC. Prospective trials are necessary to elucidate the role of combination chemotherapy, with or without targeted agents, in the salvage setting. Currently, improvements in this field should be pursued considering single-agent chemotherapy as the foundation for new more active combinations. Key words: urothelial carcinoma, chemotherapy, salvage therapy, single-agent chemotherapy, combination chemotherapy
introduction Rescue of patients with metastatic urothelial cancer (UC) who have failed first-line chemotherapy or who develop a relapse early after perioperative chemotherapy is still problematic. At present, no Food and Drug Administration (FDA)-approved agent is available as a conventional option and vinflunine is the only European Medicines Agency (EMA)-approved drug for these patients [1]. Irrespective of the type of single-agent *Correspondence to: Dr Andrea Necchi, Department of Medical Oncology, Fondazione IRCCS Istituto Nazionale dei Tumori, Via G. Venezian 1, 20133 Milan, Italy.Tel: +39-022390-2402; Fax: +39-02-2390-3150; E-mail: [email protected] †
Both authors contributed equally as co-first authors.
chemotherapy that is employed as first or later salvage therapy, results are frustratingly poor and there is a desperate need for new agents. Median progression-free (PFS) and overall survival (OS) invariably approximated 3 and 6 months, respectively, and these estimates are currently applied to the design of trials of salvage therapy. Improvements in the stratification of outcome in the secondline setting have been achieved as well, and validated prognostic grouping systems are available [2–4]. The same applies to the understanding of the prognostic impact of several additional factors for clinical use. Poor tolerability of chemotherapy in these patients is a major concern and often limits the administration of potentially more active, yet more toxic combination regimens. As a result, the administration of single-agent taxanes ( paclitaxel or
© The Author 2015. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please email: [email protected].
Downloaded from http://annonc.oxfordjournals.org/ by guest on December 23, 2015
Received 12 August 2015; revised 29 September 2015; accepted 12 October 2015
reviews
patients and methods search strategy and data abstraction We carried out a systematic review and meta-analysis in accordance with the preferred reporting items for systematic reviews and meta-analyses (PRISMA) guidelines [5]. Eligible studies were searched in Medline, EMBASE, and meeting abstracts presented at congresses of the American Society of Clinical Oncology (ASCO), European Society for Medical Oncology (ESMO), American Association of Cancer Research (AACR), and Genitourinary Cancers Symposiums. The following inclusion criteria have been adopted: period of publication between 1990 and 2014, more than 10 patients enrolled, trials/studies reporting data on second-line single-agent or doublet combination chemotherapy. Principal exclusion criteria were overlapping publications, lack of relevant outcome data, studies entirely reporting on the patients treated before 1990, studies reporting on regimens combining more than two drugs, and the use of single-agent or combinations including molecularly targeted therapies. One randomized trial that allowed the crossover between the two arms was included because of the good quality of the reported results [6]. The population, intervention, comparison, and outcome (PICO) strategy was conducted and the following search string was utilized: ‘transitional cell carcinoma’/exp OR transitional AND cell AND carcinoma:ab,ti AND ‘chemotherapy’/exp OR ‘salvage therapy’/exp OR ‘single drug dose’/exp OR ‘cancer combination chemotherapy’/exp OR ‘salvage therapy’:ab,ti OR ‘combination | Raggi et al.
chemotherapy’:ab,ti OR ‘single drug dose’:ab,ti AND ‘clinical effectiveness’/exp OR ‘overall survival’/exp OR ‘progression free survival’/exp OR ‘cancer staging’/exp OR ‘toxicity’/exp. Additional queries with relevant variants and filters have been added up, integrated by the search through the ASCO portal. Search results were independently reviewed by two authors (DR and AN). Full articles were retrieved for further qualitative review.
statistical methods The primary objective was to compare treatment containing a single-agent versus a doublet combination chemotherapy; descriptive statistics were used to summarize information across all trials according to the treatment group. The primary end points were the objective response rate (ORR), PFS, and OS; the secondary end point was the incidence of acute (grade 3–4) toxicities. PFS was commonly defined as the time from starting treatment until objective tumor progression or death, OS was defined as the time from treatment start to death for any reason, with censoring at the date of last contact for alive patients. For the purposes of this study, time to progression (TTP) was used whenever it was reported instead of PFS since it was judged an acceptable estimator of PFS in the analyzed studies [i.e. overall, there were very few death events due to toxicity or reasons other than disease progression (PD)]. Random-effects models using inverse variance weighting were used to pool trial-level data (median PFS and OS, and ORR or toxicity probability) separately in each treatment arm. We always applied random-effects models because, according to the Q test and I 2 statistic [7], heterogeneity among the studies was pretty always present (Q statistic, P < 10% and/or I 2 > 25%). The comparisons between treatment arms were performed by random-effects univariable and multivariable models, with adjustment for the percentage of patients with an Eastern Cooperative Oncology Group performance status (ECOG-PS) ≥1 and with hepatic metastases; these percentages were modeled as continuous variables using 3-knots restricted cubic splines [8]. For each study, the standard error (SE) of the median OS or PFS was derived from the respective 95% confidence interval (CI), when reported, or estimated by using a linear regression model in which the response variable was the log(SE) and the covariates were the median OS or PFS (modeled with a 3-knots restricted cubic spline) and the number of patients. Publication bias was evaluated by visually inspecting funnel plots and using Egger’s test for bias. Sensitivity analyses were conducted by using different diagnostic measures like externally standardized residuals, DFFITS values, Cook’s distances, DFBETAS values, the estimates of τ 2 (estimated amount of total heterogeneity) when each study is removed in turn. The statistical analyses were carried out using the SAS® and R software (http://www.R project.org; last access 31 July 2015). All tests and CIs were two-sided; tests were considered statistically significant when the corresponding P values were below the 5% threshold.
results search results We identified a total of 1477 studies using the search criteria (EMBASE: 633; Medline: 812; EMBASE and Medline: 32),
Volume 27 | No. 1 | January 2016
Downloaded from http://annonc.oxfordjournals.org/ by guest on December 23, 2015
docetaxel) or vinflunine are the two equally available options outside of clinical trials. Nonetheless, huge heterogeneity characterizes the available results with second-line chemotherapy and the survival advantage with the use of combination regimens remains substantially unproven. Modern research strategy in UC should focus on the rational delivery of standard systemic chemotherapy and on the optimal chemobiologic combination or sequence with novel compounds. In order to allow the design of informed clinical trials, it will be of primary importance to identify which are the best results achievable with chemotherapy and, ideally, for which patients. For fit patients who may be administered combination chemotherapy, for example, the benchmark of single-agent chemotherapy might not reflect the best achievable result. Otherwise for the majority of patients for whom tolerability of treatments is a concern due to poor performance status or organ function, the single-agent benchmark may be used. For these reasons, quantifying the survival differences between single-agent and combination chemotherapy will be a compelling argument for further research. The comparison of single-agent with doublet chemotherapy as second-line chemotherapy for metastatic UC was the objective of the present systematic review and meta-analysis. We focused on the two-drug combination chemotherapy only because the administration of three or more drugs in these patients is quite rare in clinical practice. The hypothesis was that doublet chemotherapy may be recommended as the first option for selected patients given its exhibited efficacy and that drug development in this disease should separate patient categories according to the best deliverable chemotherapy.
Annals of Oncology
reviews
Annals of Oncology
1477 abstracts and articles identified through database searching following a PICO strategy
748 records excluded (abstracts, reviews, letters, editorials, short surveys)
729 published or in press articles for full text screening
1 abstract selected [29] among 113 GU-ASCO and ASCO abstracts from 2010 to 2014 screened in EMBASE and ASCO portal
729 articles screened
686 full text articles excluded (sequential therapy, combination chemotherapy with >2 agents, use of targeted drugs, £10 patients analyzed or small retrospective series, overlapping series).
44 studies (46 arms) included in the final systematic review and meta-analysis
Figure 1. Preferred reporting items for systematic reviews and meta-analysis flowchart. ASCO (GU), American Society of Clinical Oncology (Genitourinary Cancers Symposium); PICO, population, intervention, comparison, and outcome.
Volume 27 | No. 1 | January 2016
following criteria had been used to assess the response: Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1 in 3 arms, RECIST version 1.0 in 16, World Health Organization (WHO) criteria in 22, ECOG in 3, and Japanese criteria in 1 (not available in 1).
uni- and multivariable meta-analyses for response and survival outcome The summary of the outcomes in the two groups is provided in Table 3, together with the principal subgroup analyses. The pooled ORR with single-agents was 14.2% (95% CI 11.1% to 17.9%) (I 2 = 51.50%, P for heterogeneity = 0.002, with 22 arms included, Egger’s test, P = 0.153, Figure 2A; Bellmunt study was influential, likely due to the high precision of estimates [10]) versus 31.9% (95% CI 27.3% to 36.9%) (I 2 = 38.20%, P for heterogeneity = 0.030, with 24 arms included, Egger’s test, P = 0.008, Figure 2B; no influential studies were identified) with doublets. The pooled median PFS for the single-agent group was 2.69 months (95% CI 2.25–3.12) (I 2 = 90.23%, P for heterogeneity <0.001, with 18 arms included, Egger’s test, P < 0.001, Figure 2C; no influential studies were identified) and the pooled median PFS for the doublet group was 4.05 months (95% CI 3.54–4.57) (I 2 = 30.87%, P for heterogeneity = 0.017, with 15 arms included, Egger’s test, P = 0.001, Figure 2D; 9 arms of studies were influential because median PFS diverged from others [31, 35, 37, 40, 43, 47–49]). The same trend was seen for the OS end point: the pooled median OS with single agents was 6.98 months (95% CI 6.19– 7.78) (I 2 = 68.50%; P for heterogeneity <0.001, with 20 arms included, Egger’s test, P = 0.003, Figure 2E). Akaza study was influential because median OS diverged from others [25]. The pooled median OS in the doublet group was 8.50 months (95% CI 7.35–9.64) (I 2 = 70.04%, P for heterogeneity <0.001, with 23 arms included, Egger’s test, P < 0.001, Figure 2F). No influential studies were identified. Funnel plots of the OS end point for the single-agent and doublet groups are provided as supplementary Figures S1A and B, available at Annals of Oncology online. Univariable analyses are provided in Table 4: the difference of median PFS was 1.60 months (95% CI 0.68–2.52, P = 0.001) and the difference of median OS was 1.46 months (95% CI −0.27 to 3.18, P = 0.097) in favor of the doublet chemotherapy. The results of the multivariable comparative analyses are also supplied in Table 4. While response and PFS differences were confirmed as statistically significant (P < 0.001 and 0.002, respectively), OS was not (median difference 0.90 months, range −0.74 to 2.54, P = 0.284). When considering only the taxane or vinflunine among the single-agent arms of studies, ORR was consistently significantly different (P < 0.001), but neither PFS (P = 0.134) nor OS (P = 0.547) was significantly different between the single-agent and doublet chemotherapy groups. We finally ran another multivariable model adjusting for ECOGPS 2 and visceral metastases, and the comparison between the two groups favored doublet chemotherapy as follows: P < 0.001 for both ORR probability and median PFS, P = 0.080 for median OS.
incidence of acute toxicities in the two groups A summary of grade 1–2 and grade 3–4 hematologic toxicity and of grade 3–4 peripheral neuropathy and nephrotoxicity is
doi:10.1093/annonc/mdv509 |
Downloaded from http://annonc.oxfordjournals.org/ by guest on December 23, 2015
published between 1995 and 2014. Figure 1 outlines the selection process and reasons for study exclusion. Table 1 presents the principal study characteristics and outcomes [6, 9–51]. A total of 44 studies accounting for 46 arms of trials (22 with single agent and 24 with doublets) were selected for the metaanalysis, all of them performed in the second-line setting. A total of 1910 patients were selected: 1202 had received a singleagent, including 486 vinflunine and 201 paclitaxel or docetaxel, and 708 had received a doublet chemotherapy, including cis- or carboplatin in 191, or a taxane in 602 patients. A summary of the study characteristics is provided in Table 2. There were four randomized trials, two in both groups, for both of which both arms were included in the doublet chemotherapy group and only one arm for both in the single-agent group. The median sample size of the trials was 45 (range 13–253) for single-agent and 32 (range 11–48) for doublet chemotherapy. Among the single-agent arms of trials, there were three (13.6%) with vinflunine and five (22.7%) with paclitaxel or docetaxel. Among the doublet chemotherapy arms, 20 (83.3%) were taxanecontaining and 4 (16.7%) carboplatin-containing regimens. Only two studies consisted of a cisplatin-based doublet. Patient-level characteristics were equally distributed between the two groups, except for a nonstatistically significant higher proportion of patients with ECOG-PS 2 in the doublet chemotherapy group [14% (range 0–23.8) versus 8.5% (range 0–57), Wilcoxon test, P = 0.675]. The pooled median follow-up was 17 months (range 1.8–21.5 months) in the single-agent group and 20.4 months (range 9–48 months) in the doublet chemotherapy group. The
Year
ECOG-PS 1 (%)
ECOG-PS 2 (%)
Prior CDDP (%)
Prior CBDCA (%)
Visceral metastases, N (%)
Hepatic metastases, N (%)
Treatment
N patients (evaluable for response)
Criteria of response assessment
Median age (years)
Single-agent chemotherapy Beer et al. [9] Bellmunt et al. [10] Culine et al. [11] Galsky et al. [12] Dreicer et al. [13] Joly et al. [14] Lorusso et al. [15] McCaffrey et al. [16] Sweeney et al. [17] Vaughn et al. [18] Vaughn et al. [19] Witte et al. [20] Winquist et al. [21] Witte et al. [22] Ko et al. [23] Lee et al. [24] Akaza et al. [25] Ponzato et al. [26] Papamichael et al. [27] Sengelov et al. [28] Choueiri et al. [6] Loriot et al. [29]
2008 2009 2006 2007 2007 2009 1998 1997 2006 2002 2009 1997 2005 1998 2013 2012 2007 1997 1997 1995 2012 2012
60 72 45 54 62 82 40 100 34 39 100 71 55 100 50 48 27.3 NA 50 47 52.8 NA
10 0 0 0 7 18 57 0 4.3 13 0 29 20 0 16 14 4.5 NA 28 26 NA 0
NA 64.8 NA 31 NA NA 100 NA NA 45 NA 94 100 NA 69 NA 95 70 64 90 NA NA
NA 29.6 NA 69 NA NA 0 NA NA 29 NA NA 0 NA NA NA 0 30 0 0 NA NA
72 74 NA 61 NA 58 NA NA 42.5 77 NA NA NA 96 73 62 NA NA 71 61 76.4 NA
NA NA NA NA 26 38 23 27 NA 45 50.3 27 25 32 31 30 13.6 15 43 57 37.5 NA
CPT-11 VFL VFL PEM IXB T G TXT PEM T VFL I O TOP NabT Genexol-PM G I T MMC TXT PTX
45 (40) 253 (253) 58 (51) 13 (12) 45 (42) 50 (45) 35 (31) 31 (30) 47 (47) 31 (31) 175 (151) 60 (56) 20 (18) 46 (44) 48 (47) 37 (34) 46 (44) 20 (20) 14 (14) 23 (19) 75 (72) 30 (30)
RECIST 1.0 RECIST 1.0 WHO RECIST 1.0 RECIST 1.0 RECIST 1.0 WHO WHO WHO WHO WHO ECOG WHO ECOG RECIST 1.0 RECIST 1.0 JUA/JSP WHO WHO WHO RECIST 1.0 NA
64.4 NA 63 69 63 64 64 61 64 66 66 67 64 62 66 57 65 68 68 61 NA NA
Author
Year
Median treatment duration (months) 2 NA 3 2 2.2 2 1.6 2.1 2.1 2.1 2.3
Median N cycles
Median FUP (months)
ORR: %
ORR 95% CI
PFS: median (months)
PFS: CI
OS: median (months)
OS 95% CI
3 NA 4 3 3 2 1.6 3 3 3 3
33 21.5 20 5.8 20 NA NA NA 9.2 NA 11.9
5 8.6 18 8 11.9 9 22.5 13.3 27.7 10 14.6
1.0–17.0 NA 8.4–30.9 NA 5.3–26.5 2.0–21.0 8.0–37.0 3.8–30.7 15.6–42.6 0–20 9.4–21.2
2.1 3 3 NA 2.7 3.2c 3.8c NA 2.9c 2.2c 2.8
1.8–2.3 NA 2.4–3.8 NA NA 0.2–7.5d NA NA 1.7–4.6 NA 2.6–3.8
5.4 6.9 6.6 NA 8 6.9 5 9 9.6 7.2 8.2
3.4–7.1 5.7–8.0 4.8–7.6 NA NA 0.36–18.3d 2–21d 6–12 5.1–14.6 NA 6.8–9.6
2008 2009 2006 2007 2007 2009 1998 1997 2006 2002 2009
Downloaded from http://annonc.oxfordjournals.org/ by guest on December 23, 2015
Beer et al. [9] Bellmunt et al. [10] Culine et al. [11] Galsky et al. [12] Dreicer et al. [13] Joly et al. [14] Lorusso et al. [15] McCaffrey et al. [16] Sweeney et al. [17] Vaughn et al. [18] Vaughn et al. [19]
Annals of Oncology
Volume 27 | No. 1 | January 2016
Author
reviews
| Raggi et al.
Table 1. Patient, disease, and treatment characteristics of included studies
Annals of Oncology
NA 1.5 1.5 2.53 3 3 2.3 NA NA 1.5 2
NA 2 2 6 4 3 3 NA 3 2 2
16 NA 20 25 10 18 NA 1.8 NA 7.1 NA
20 5 9.1 27.7 21 25 5 7 5 11.1 16.6
10.0–32.0 NA 2.9–25.5 17.3–44.4 7.0–34.0 13.2–40.3 NA 2.0–12.0 NA NA NA
2.4c 1.4c 1.5c 6 2.7c 3.1 6c NA NA 1.58 4
NA 0.6–4.8d NA 3.9–8.5 0.9–4.6 2–4.1 1–8d NA NA NA 2.1–4.5
5.3 6.9 6.7 10.8 6.5 12.6 8 NA 4 7.03 9.3
NA 0.3–24.2d NA 5.8–16.9 5–8 7.8–14.4 2–12d NA 1.3–15.9d NA 5.6–13.2
Author
Year
ECOG-PS 1 (%)
ECOG-PS 2 (%)
Prior CDDP (%)
27 100 45 NA 28 NA NA 27 60 NA NA NA
38 0 0 NA 22 NA NA 32 0 NA NA 23
NA 100 100 NA 100 100 100 95 71 98 42 95
Visceral metastases, N (%) 88 78 66 NA 100 42 29 63 69 NA 61 64
Hepatic metastases, N (%) 21 43 15.2 50 50 NA NA 36 NA 7 19 NA
Treatment
2001 2007 2011 2004 2009 2010 2010 2001 2011 2001 1999 2005
Prior CBDCA (%) NA 0 0 NA 0 0 0 5 29 2 0 NA
GI GI PG FOLFOX-4 T-CTX TG-1a TG-2a TXT-I EPI-T TG TI TC
N patients (evaluable for response) 34 (34) 23 (23) 33 (33) 16 (16) 32 (32) 48 (40) 48 (41) 22 (20) 35 (35) 41 (40) 26 (13) 44 (44)
Criteria of response assessment WHO WHO RECIST 1.1 WHO RECIST 1.0 RECIST 1.0 RECIST 1.0 WHO RECIST 1.0 WHO WHO WHO
Median age (years) 61 66 66 63 66 63.9 65.1 61 64 62 66 64.6
2008 2009 2011 2007 2007 2003 2006 2006 2007 2004 2014 2009
40 100 45.8 100 87 66 NA NA 89 36 57.1 18
20 0 20.8 0 13 14 NA NA 4 28 23.8 0
100 100 100 100 100 35 100 100 100 NA NA NA
0 0 0 0 0 65 0 0 0 NA NA NA
NA NA 63 NA 71 NA NA NA 75 65 86.7 73
30 9 17 33 39 21 14 8 29 NA 28.6 27
TG TG TG TC TC TXT-G TG-3wb TG-2wb TP TC T-CTX TXT-O
20 (20) 33 (30) 24 (24) 18 (18) 35 (31) 31 (29) 14 (14) 13 (13) 28 (24) 33 (33) 46 (42) 11 (11)
RECIST 1.0 RECIST 1.0 RECIST 1.1 WHO RECIST 1.0 ECOG WHO WHO WHO WHO RECIST 1.1 RECIST 1.0
62.9 66.1 64.5 62 67 64 62 4.5 61 70 65 65
Pectasides et al. [30] Lin et al. [31] Gondo et al. [32] Di Lorenzo et al. [33] Di Lorenzo et al. [34] Albers et al. [35]
doi:10.1093/annonc/mdv509 |
Krege et al. [36] Rozzi et al. [37] Sternberg et al. [38] Sweeney et al. [39] Vaishampayan et al. [40] Kanai et al. [41] Suyama et al. [42] Ikeda et al. [43] Soga et al. [44] Kouno et al. [45] Dreicer et al. [46] Fechner et al. [47] Uhm et al. [48] Friedland et al. [49] Park and Lee [50] Srinivas and Harshman [51]
Continued
reviews
1997 2005 1998 2013 2012 2007 1997 1997 1995 2012 2012
Downloaded from http://annonc.oxfordjournals.org/ by guest on December 23, 2015
Volume 27 | No. 1 | January 2016
Witte et al. [20] Winquist et al. [21] Witte et al. [22] Ko et al. [23] Lee et al. [24] Akaza et al. [25] Ponzato et al. [26] Papamichael et al. [27] Sengelov et al. [28] Choueiri et al. [6] Loriot et al. [29] Doublet chemotherapy
Author
Year
ECOG-PS 1 (%)
ECOG-PS 2 (%)
Prior CDDP (%)
Prior CBDCA (%)
Visceral metastases, N (%)
Hepatic metastases, N (%)
Treatment
N patients (evaluable for response)
Criteria of response assessment
Author
Year
Median N cycles
Median FUP (months)
ORR: %
ORR 95% CI
PFS: median (months)
PFS: CI
OS: median (months)
OS 95% CI
2001 2007 2011 2004 2009 2010 2010 2001 2011 2001 1999 2005
Median treatment duration (months) 3 3 3 NA 4.5 NA NA 1.5 3 4 2 2.45
4 3 4 NA 6.2 3 4 2 4 6 3 3.5
24 18 30 NA 9 40 40 NA NA NA 48 24
21 22 39.4 19 31 37.5 41.4 25 29 60 15.4 16
9.0–38.0 5.0–39.0 NA NA 17.0–45.0 NA NA NA NA 45.0–75.0 2.0–45.0 7.0–30.0
4c 3.5 NA NA 5c 4 3.1 NA 7.6c NA NA 4
0.52–21.6d 2.8–4.2 NA NA NA NA NA NA 3.2–10.7 NA NA 3–5
9 4.8 10.5 4 8 7.8 8 NA 12.6 14.4 8 6
0.5–28d 2.9–6.6 3–29.9d 2–8d 4–14 4.2–11.4 4.9–11.1 NA 4.6–18.8 2–43d 1–53d 5–8
2008 2009 2011 2007 2007 2003 2006 2006 2007 2004 2014 2009
5.3 3.5 3 3 NA NA 4.5 9 3.5 1.5 1.5 1.5
7.7 5 4 4 10 NA 6 11 5 2 2 2
11.9 15 20.4 20 7.8 NA 11 38 16.4 48 24 45
30 33.3 42 33 32.3 17.2 50 NA 36 24 33.3 18
10.0–50.0 19.2–51.2 NA NA 15.8–48.7 7.0–33.0 NA 6c 9.0–41.0 10.0–39.0 19.0–47.6 NA
NA NA 6.1 4 3.7 NA 11c 1–15d 6.2c 3.6c 3c NA
NA NA 0.5–23.9d NA NA NA 3–41d 9 3.9–8.5 1–29d 1.7–4.3 NA
11.5 11.3 12.4 11 7.9 7.7 13 0–16d 10.3 10.3 6.3 7
2–22d 7.2–13.6 0.5–30.2d NA NA NA 5.46d
Pectasides et al. [30] Lin et al. [31] Gondo et al. [32] Di Lorenzo et al. [33] Di Lorenzo et al. [34] Albers et al. [35] Krege et al. [36] Rozzi et al. [37] Sternberg et al. [38] Sweeney et al. [39] Vaishampayan et al. [40] Kanai et al. [41] Suyama et al. [42] Ikeda et al. [43] Soga et al. [44] Kouno et al. [45] Dreicer et al. [46] Fechner et al. [47]
6.1–14.1 1–33d 4.6–8 2–48d
a
Short-term (1) versus prolonged (2) course of chemotherapy. Two- versus 3-weekly schedule. c Numbers are indicating the time to progression. d Numbers are indicating the range instead of the confidence interval. CBDCA, carboplatin; CDDP, cisplatin; CI, confidence interval; CPT-11, irinotecan; CTX, cyclophosphamide; ECOG-PS, Eastern Cooperative Oncology Group Performance Status; EPI, epirubicin; FOLFOX, fluorouracil, leucovorin, oxaliplatin; FUP, follow-up; GI, gemcitabine, ifosfamide; IXB, ixabepilone; MMC; mitomycin C; NA, not available; NabT, nab-paclitaxel; O, oxaliplatin; ORR, objective response rate; PEM, pemetrexed; PFS, progression-free survival; PG, cisplatin, gemcitabine; PTX, pralatrexate; OS, overall survival; RECIST, Response Evaluation Criteria in Solid Tumors; TC, paclitaxel, carboplatin; TG, paclitaxel, gemcitabine; TI, paclitaxel, ifosfamide; TOP, topotecan; TP; paclitaxel, cisplatin; TTP, time-to-progression; TXT, docetaxel; VFL, vinflunine; WHO, World Health Organization. b
Annals of Oncology
Downloaded from http://annonc.oxfordjournals.org/ by guest on December 23, 2015
Volume 27 | No. 1 | January 2016
Uhm et al. [48] Friedland et al. [49] Park and Lee [50] Srinivas and Harshman [51]
Median age (years)
reviews
| Raggi et al.
Table 1. Continued
reviews
Annals of Oncology
groups in regard to the grade 3–4 toxicities. Univariably, there were no statistically significant differences between the two groups for any type of toxicity. Results were confirmed after adjusting the P values for ECOG-PS 2 and visceral metastases (data not shown). Three studies reported on drug-related death events, two with single-agents [Dreicer et al. (n = 1), Winquist et al. (n = 2)] [13, 21] and one with doublets [Albers et al. (n = 2)] [35]. We were not able to find any data regarding longterm side-effects for any treatment administered.
detailed for each trial in supplementary Table S1, available at Annals of Oncology online. Table 5 presents the univariable comparison between single-agent and doublet chemotherapy
Table 2. Summary of the characteristics of single-agent and doublet chemotherapy arms of studies Arms with singleagent (N = 22)
discussion 7 (31.8) 4 (18.2) 8 (36.4) 3 (13.6)
4 (16.7) 6 (25.0) 13 (54.2) 1 (4.2)
2 (9.1) 20 (90.9)
4 (16.7) 20 (83.3)
– – 3 (13.6) 2 (9.0) 3 (13.6) 45 (13–253)
2 (9.1) 4 (16.7) 17 (70.8) 3 (12.5) – 32 (11–48)
64 (57–69)
64 (61–70)
69.1 (30.4–97.1) 2 (9.0)
65.7 (18.2–100) 8 (33.3)
8.5 (0–57) 2 (9.0)
14 (0–23.8) 7 (29.2)
71.5 (61–96) 10 (45.4)
69 (35–100) 9 (37.5)
30.5 (13.6–50.3) 6 (27.2)
In this trial-level meta-analysis of second-line chemotherapy for UC, unique to the authors’ knowledge and resembling a huge number of patients, we found statistically significant differences in ORR and median PFS favoring the use of two-drug regimens over single-agents. The difference of median OS between the two groups was trending toward an improvement with the use of doublet chemotherapy regimens, but this was not statistically significant. However, when analyzing only the single-agent arms with vinflunine and taxanes ( paclitaxel or docetaxel), which are the most utilized agents or the standard of care (i.e. vinflunine, but only in Europe) outside of clinical trials, the significant difference for ORR was maintained but it was lost for PFS. First of all, the inherent limitations in running such metaanalyses are to be accounted for when interpreting the results, and the analyses should be taken as exploratory. The majority of included studies were actually single-arm trials, often with a small sample size. Most importantly, the heterogeneity in reporting the outcomes across trials (e.g. median with 95% CI in some cases, median with ranges in others, PFS in some cases and TTP in others, as outlined in Table 1) did represent the most challenging issue when pooling the data together. Available literature is limited in regard to performing this type of meta-analyses and principal references have been mostly supplied herein. This is the reason that we have included sensitivity analyses with the aim to identify possible influential studies/treatments that might have accounted for major biases for each end point. The outcomes were consistent throughout these analyses and only one influential arm has been reported for OS. As a further limitation, we were unable to find information on baseline hemoglobin,
27.8 (10–50) 6 (25)
a
Numbers are accounting for the study arms. ECOG-PS, Eastern Cooperative Oncology Group performance status.
Table 3. Subgroup analyses of studies investigating single-agent or doublet second-line chemotherapy in advanced urothelial cancer Study selection
Single-agent chemotherapy Vinflunine Paclitaxel or docetaxel Doublet chemotherapy Doublet with cisplatin Doublet without cisplatin Doublet with carboplatin
ORR No. of evaluable arms of studies
Probability % (95% CI)
PFS No. of evaluable arms of studies
Median PFS (95% CI)
OS No. of evaluable arms of studies
Median OS (95% CI)
22 3 5 24 2 22 4
14.2 (11.1–17.9) 11.7 (6.2–20.9) 10.5 (6.9–15.8) 31.9 (27.3–36.9) 40.4 (28.5–53.5) 30.9 (26.1–36.3) 25.4 (17.9–34.7)
18 3 3 15 1 14 4
2.69 (2.25–3.12) 2.92 (2.55–3.29) 2.20 (1.36–3.04) 4.05 (3.54–4.57) 6.20 (3.95–8.45) 3.79 (3.40–4.17) 3.86 (3.20–4.51)
20 3 4 23 2 21 4
6.98 (6.19–7.78) 7.20 (6.30–8.10) 7.35 (6.16–8.55) 8.50 (7.35–9.64) 10.39 (7.53–13.26) 8.35 (7.15–9.55) 8.14 (5.76–10.52)
CI, confidence interval; ORR, objective response rate; PFS, progression-free survival; OS, overall survival.
Volume 27 | No. 1 | January 2016
doi:10.1093/annonc/mdv509 |
Downloaded from http://annonc.oxfordjournals.org/ by guest on December 23, 2015
Study-level characteristica Publication year, N (%) 1995–2001 2002–2006 2007–2011 2012–2014 Type of study arms, N (%) Randomized Single-group Treatment, N (%) Cisplatin Carboplatin Paclitaxel Docetaxel Vinflunine Sample size (median, range) Patient-level characteristic Age (years; median, range) ECOG-PS ≥1 % (median, range) Not reported (N arms, %) ECOG-PS = 2 % (median, range) Not reported (N arms, %) Visceral metastases % (median, range) Not reported (N arms, %) Hepatic metastases % (median, range) Not reported (N arms, %)
Arms with doublet (N = 24)
reviews
Annals of Oncology
A 0.25 [0.14, 0.40] 0.05 [0.01, 0.18] 0.06 [0.04, 0.10] 0.11 [0.06, 0.21] 0.18 [0.09, 0.31] 0.12 [0.05, 0.26] 0.08 [0.01, 0.41] 0.09 [0.03, 0.21] 0.28 [0.17, 0.42] 0.21 [0.10, 0.37] 0.17 [0.07, 0.34] 0.23 [0.11, 0.40] 0.13 [0.05, 0.31] 0.07 [0.01, 0.37] 0.05 [0.01, 0.28] 0.05 [0.01, 0.29] 0.28 [0.17, 0.42] 0.10 [0.03, 0.26] 0.15 [0.10, 0.21] 0.06 [0.01, 0.31] 0.20 [0.11, 0.32] 0.09 [0.03, 0.22]
RE model
0.14 [0.11, 0.18] 0.00
0.10
0.20 0.30 0.40 Response probability
0.50
B Albers P, 2010 Albers P, 2010.1 Di Lorenzo G, 2004 Di Lorenzo G, 2009 Dreicer R, 2003 Fechner G, 2006 Fechner G, 2006.1 Friedland DM, 2004 Gondo T, 2011 Ikeda M, 2011 Kanai K, 2008 Kouno T, 2007 Krege S, 2001 Lin C-C, 2007 Park JH, 2014 Pectasides D, 2001 Rozzi A, 2011 Soga N, 2007 Srinivas S, 2009 Sternberg C, 2001 Suyama T, 2009 Sweeney CJ, 1999 Uhm JE, 2007 Vaishampayan UN, 2005
0.37 [0.24, 0.53] 0.41 [0.28, 0.57] 0.19 [0.06, 0.45] 0.31 [0.18, 0.49] 0.17 [0.07, 0.35] 0.50 [0.26, 0.74] 0.38 [0.17, 0.66] 0.24 [0.13, 0.42] 0.39 [0.24, 0.57] 0.42 [0.24, 0.62] 0.30 [0.14, 0.53] 0.32 [0.18, 0.50] 0.25 [0.11, 0.48] 0.22 [0.09, 0.43] 0.33 [0.21, 0.49] 0.21 [0.10, 0.37] 0.29 [0.16, 0.45] 0.33 [0.16, 0.57] 0.09 [0.01, 0.44] 0.60 [0.44, 0.74] 0.33 [0.19, 0.52] 0.15 [0.04, 0.45] 0.42 [0.24, 0.62] 0.16 [0.08, 0.30]
RE model
0.32 [0.27, 0.37] 0.00
0.20 0.40 0.60 Response probability
0.80
Figure 2. Forest plots: (A) objective response rate, single-agent studies; (B) objective response rate, doublet chemotherapy studies; (C) progression-free survival, single-agent studies; (D) progression-free survival, doublet chemotherapy studies; (E) overall survival, single-agent studies; (F) overall survival, doublet chemotherapy studies.
| Raggi et al.
Volume 27 | No. 1 | January 2016
Downloaded from http://annonc.oxfordjournals.org/ by guest on December 23, 2015
Akaza H, 2007 Beer T M, 2008 Bellmunt J, 2009 Choueiri TK, 2012 Culine S, 2006 Dreicer R, 2007 Galsky M.D., 2007 Joly F, 2009 Ko YJ, 2013 Lee J-L, 2011 Loriot Y, 2012 Lorusso V, 1998 McCaffrey JA, 1997 Papamichael D, 1997 Ponzato P, 1997 Sengelov L, 1995 Sweeney CJ, 2006 Vaughn DJ, 2002 Vaughn DJ, 2009 Winquist E, 2005 Witte RS, 1997 Witte RS, 1998
reviews
Annals of Oncology
C 3.10 [2.07, 4.13]
Beer T M, 2008
2.10 [1.86, 2.34]
Bellmunt J, 2009
3.00 [2.33, 3.67]
Choueiri TK, 2012
1.58 [1.29, 1.87]
Culine S, 2006
3.00 [2.31, 3.69]
Dreicer R, 2007
2.70 [1.95, 3.45]
Joly F, 2009
3.20 [2.12, 4.28]
Joung Ko Y, 2013
6.00 [3.75, 8.25]
Lee J-L, 2011
2.70 [0.89, 4.51]
Loriot Y, 2012
4.00 [2.82, 5.18]
Lorusso V, 1998
3.80 [2.24, 5.36]
Ponzato P, 1997
6.00 [3.32, 8.68]
Sweeney CJ, 2006
2.90 [1.48, 4.32]
Vaughn DJ, 2002
2.20 [1.69, 2.71]
Vaughn DJ, 2009
2.80 [2.21, 3.39]
Winquist E, 2005
1.40 [1.13, 1.67]
Witte RS, 1997
2.40 [1.82, 2.98]
Witte RS, 1998
1.50 [1.22, 1.78]
RE model
2.69 [2.25, 3.12]
0.00
2.00
4.00 6.00 Median PFS
8.00
10.00
D Albers P, 2010
4.00 [0.08, 7.92]
Albers P, 2010.1
3.10 [1.97, 4.23]
Di Lorenzo G, 2009
5.00 [2.31, 7.69]
Fechner G, 2006
11.00 [6.25, 15.75]
Fechner G, 2006.1
6.00 [3.29, 8.71]
Friedland DM, 2004
3.60 [2.18, 5.02]
Ikeda M, 2011
6.10 [3.41, 8.79]
Kouno T, 2007
3.70 [2.21, 5.19]
Lin C-C, 2007
3.50 [2.81, 4.19]
Park JH, 2014
3.00 [1.73, 4.27]
Pectasides D, 2001
4.00 [2.29, 5.71]
Rozzi A, 2011
7.60 [3.93, 11.27]
Soga N, 2007
4.00 [2.24, 5.76]
Uhm JE, 2007
6.20 [3.95, 8.45]
Vaishampayan UN, 2005
4.00 [3.02, 4.98]
RE model
4.05 [3.54, 4.57]
0.00
5.00
10.00 Median PFS
15.00
20.00
Fig. 2 Continued
albumin levels, and the time from prior chemotherapy. Each one of these factors has a recognized role in the salvage setting [2–4]. Although multivariable analyses were adjusted for ECOGPS and hepatic metastases only, it is unlikely that the present
Volume 27 | No. 1 | January 2016
findings would have been modified by adjusting for additional confounders. Of course, we could not exclude that other confounders could have impacted the results beyond what could have been adjusted by the regression model. For example, the
doi:10.1093/annonc/mdv509 |
Downloaded from http://annonc.oxfordjournals.org/ by guest on December 23, 2015
Akaza H, 2007
reviews
Annals of Oncology
E
RE model
6.98 [6.19, 7.78] 0.00
5.00
10.00 Median OS
15.00
20.00
F Albers P, 2010 Albers P, 2010.1 Di Lorenzo G, 2004 Di Lorenzo G, 2009 Dreicer R, 2003 Fechner G, 2006 Fechner G, 2006.1 Friedland DM, 2004 Gondo T, 2011 Ikeda M, 2011 Kanai K, 2008 Kouno T, 2007 Lin C-C, 2007 Park JH, 2014 Pectasides D, 2001 Rozzi A, 2011 Soga N, 2007 Srinivas S, 2009 Sternberg C, 2001 Suyama T, 2009 Sweeney CJ, 1999 Uhm JE, 2007 Vaishampayan UN, 2005
7.80 [4.27, 11.33] 8.00 [4.96, 11.04] 4.00 [2.85, 5.15] 8.00 [3.10, 12.90] 7.70 [4.80, 10.60] 13.00 [7.74, 18.26] 9.00 [5.08, 12.92] 10.30 [6.16, 14.44] 10.50 [6.29, 14.71] 12.40 [7.51, 17.29] 11.50 [6.75, 16.25] 7.90 [4.93, 10.87] 4.80 [2.99, 6.61] 6.30 [4.63, 7.97] 9.00 [5.41, 12.59] 12.60 [5.64, 19.56] 11.00 [6.36, 15.64] 7.00 [4.31, 9.69] 14.40 [9.37, 19.43] 11.30 [8.16, 14.44] 8.00 [4.85, 11.15] 10.30 [6.38, 14.22] 6.00 [4.53, 7.47]
RE model
8.50 [7.35, 9.64] 0.00
5.00
10.00 Median OS
15.00
20.00
Fig. 2 Continued
disproportionate representation of ECOG-PS 2 patients, being more frequent in the doublet chemotherapy group, was a concern and it was somewhat surprising, although an exploratory
| Raggi et al.
multivariable analysis that included ECOG-PS 2 was made and results did not change substantially (OS difference still not significant).
Volume 27 | No. 1 | January 2016
Downloaded from http://annonc.oxfordjournals.org/ by guest on December 23, 2015
12.60 [9.42, 15.78] 5.40 [3.59, 7.21] 6.90 [5.77, 8.03] 7.03 [4.97, 9.09] 6.60 [5.23, 7.97] 8.00 [5.09, 10.91] 6.90 [4.68, 9.12] 10.80 [5.36, 16.24] 6.50 [5.03, 7.97] 9.30 [5.58, 13.02] 5.00 [3.59, 6.41] 9.00 [6.06, 11.94] 8.00 [4.77, 11.23] 4.00 [2.88, 5.12] 9.60 [4.95, 14.25] 7.20 [4.61, 9.79] 8.20 [6.83, 9.57] 6.90 [4.38, 9.42] 5.30 [3.92, 6.68] 6.70 [4.55, 8.85]
Akaza H, 2007 Beer T M, 2008 Bellmunt J, 2009 Choueiri TK, 2012 Culine S, 2006 Dreicer R, 2007 Joly F, 2009 Ko YJ, 2013 Lee J-L, 2011 Loriot Y, 2012 Lorusso V, 1998 McCaffrey JA, 1997 Ponzato P, 1997 Sengelov L, 1995 Sweeney CJ, 2006 Vaughn DJ, 2002 Vaughn DJ, 2009 Winquist E, 2005 Witte RS, 1997 Witte RS, 1998
reviews
Annals of Oncology
Table 4. Pooled estimates from the univariable and multivariable analyses
Univariable analyses Doublet versus single-agent Multivariable analyses Doublet versus single-agent Doublet versus single-agent paclitaxel or docetaxel or vinflunine Doublet cisplatin or carboplatin versus other doublets
ORR Odds ratio (95% CI) P value
PFS median, difference (95% CI) P value
OS median, difference (95% CI) P value
2.78 (1.95 to 3.94) P < 0.001
1.60 (0.68 to 2.52) P = 0.001
1.46 (−0.27 to 3.18) P = 0.097
2.53 (1.79 to 3.58) P < 0.001 3.20 (1.94 to 5.28) P < 0.001 1.76 (0.98 to 3.15) P = 0.058
1.60 (0.57 to 2.63) P = 0.002 1.10 (−0.34 to 2.55) P = 0.134 0.78 (−1.46 to 2.02) P = 0.495
0.90 (−0.74 to 2.54) P = 0.284 −0.64 (−2.72 to 1.44) P = 0.547 2.34 (−0.36 to 5.05) P = 0.089
CI, confidence interval; ORR, objective response rate; PFS, progression-free survival; OS, overall survival.
Anemia Neutropenia Thrombocytopenia Peripheral neuropathy Nephrotoxicity
Single-agent Number of evaluable arms of studies
Probability % (95% CI)
Doublet Number of evaluable arms of studies
Probability % (95% CI)
P value*
22 22 22 15 13
9.7 (5.6–16.1) 11.6 (5.5–23.3) 7.5 (4.3–12.6) 3.6 (1.9–6.7) 3.0 (1.3–6.7)
24 24 24 16 13
14.5 (10.1–20.3) 16.8 (9.0–29.3) 8.9 (5.5–14.1) 5.5 (3.0–9.7) 3.1 (1.6–5.8)
0.438 0.531 0.781 0.431 0.699
*Univariable model. CI, confidence interval.
It is plausible that a rational administration of doublet chemotherapy to ECOG-PS 0 patients and of single-agents to ECOGPS 1–2 or at least PS 2 patients, or a larger dataset of randomized trials would have led to more clinically meaningful and statistically significant differences in survival between the two groups. Taken together, these results might help guiding the design of the next clinical trials in the salvage setting. We have provided the precise pooled estimates of median PFS and OS to be utilized when determining the advantage from adding or comparing a new drug to standard chemotherapy. Although the comparison of median OS was not statistically significant in our study, these benchmark estimates may be considered separately in randomized trials with single- or two-drug combination in the standard therapy arm. This separation might best apply to set the null hypothesis of phase II trials, for which the ORR and PFS improvement (which were both statistically significantly different here between the two groups) are the end points. Such benchmark estimates are as follows: median PFS of 2.69 months for the single-agents and 4.05 months for the doublet chemotherapy, and median OS of 6.98 months for single-agents and 8.50 months for doublet chemotherapy. Moving outside of clinical trials, some recommendations could be made for clinical practice on the basis of the present
Volume 27 | No. 1 | January 2016
findings. The outcomes with the use of paclitaxel, docetaxel, or vinflunine are unlikely to be significantly improved by adding a second chemotherapeutic drug even when it is feasible. Furthermore, the promising results with the use of the nanoparticle albumin-bound (nab)-paclitaxel lend further confidence to the single-agent strategy. Nab-paclitaxel was investigated as second-line chemotherapy in a Canadian multicenter, singlearm, phase II study [23]. The overall response rate was 27.7% and, most importantly, median OS was 10.8 months despite a high proportion of patients with baseline poor prognostic features. Yet investigating nab-paclitaxel in combination regimens (i.e. with gemcitabine/carboplatin chemotherapy) exhibited significant myelosuppression in the neoadjuvant setting which needs to be carefully addressed if evaluating the same combination in the second line [52]. Major tolerability issues also characterized the efforts of combining vinflunine or pemetrexed with doublet chemotherapy [53], and consequently investigations of these compounds are mainly conceivable in sequential or maintenance strategies. Substantial uncertainties still lie regarding the use of doublet chemotherapy regimens in the overall population of patients failing platinum-based chemotherapy, but certainly the patient selection is the key to investigate combination chemotherapy further. A few additional studies have been published in
doi:10.1093/annonc/mdv509 |
Downloaded from http://annonc.oxfordjournals.org/ by guest on December 23, 2015
Table 5. Summary of the incidence of grade 3–4 adverse events between single-agent and doublet chemotherapy groups
reviews
| Raggi et al.
significant improvement of PFS and the trending-to-significance improvement of OS are suggesting a potentially meaningful benefit from combining agents with proven activity in UC. In addition to the development of novel compounds such as PD1 and PD-L1 inhibitors as single agents, the prospective evaluation of tolerable combinations of chemotherapeutic drugs as well as chemobiologic combinations is rational.
disclosure The authors have declared no conflicts of interest.
references 1. Sonpavde G, Jones BS, Bellmunt J et al. Future directions and targeted therapies in bladder cancer. Hematol Oncol Clin North Am 2015; 29: 361–376. 2. Bellmunt J, Choueiri TK, Fougeray R et al. Prognostic factors in patients with advanced transitional cell carcinoma of the urothelial tract experiencing failure with platinum-containing regimens. J Clin Oncol 2010; 28: 1850–1855. 3. Sonpavde G, Pond GR, Fougeray R et al. Time from prior chemotherapy enhances prognostic risk grouping in the second-line setting of advanced urothelial carcinoma: a retrospective analysis of pooled, prospective phase 2 trials. Eur Urol 2013; 63: 717–723. 4. Sonpavde G, Pond GR, Rosenberg JE et al. Improved 5-factor classification of patients receiving salvage systemic therapy for advanced urothelial carcinoma. J Urol 2015 Aug 17 [epub ahead of print], doi: 10.1016/j.juro.2015.07.111. 5. Liberati A, Altman DG, Tetzlaff J et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLoS Med 2009; 6: e1000100. 6. Choueiri TK, Ross RW, Jacobus S et al. Double-blind, randomized trial of docetaxel plus vandetanib versus docetaxel plus placebo in platinum-pretreated metastatic urothelial cancer. J Clin Oncol 2012; 30: 507–512. 7. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ 2003; 327: 557–560. 8. Durrleman S, Simon R. Flexible regression models with cubic splines. Stat Med 1989; 8: 551–556. 9. Beer TM, Goldman B, Nichols CR et al. Southwest Oncology Group phase II study of irinotecan in patients with advanced transitional cell carcinoma of the urothelium that progressed after platinum-based chemotherapy. Clin Genitourin Cancer 2008; 6: 36–39. 10. Bellmunt J, Theodore C, Demkov T et al. Phase III trial of vinflunine plus best supportive care compared with best supportive care alone after a platinumcontaining regimen in patients with advanced transitional cell carcinoma of the urothelial tract. J Clin Oncol 2009; 27: 4454–4461. 11. Culine S, Theodore C, De Santis M et al. A phase 2 study of vinflunine in bladder cancer patients progressing after first-line platinum-containing regimen. Br J Cancer 2006; 94: 1395–1401. 12. Galsky MD, Mironov S, Iasonos A et al. Phase II trial of pemetrexed as second-line therapy in patients with metastatic urothelial carcinoma. Invest New Drugs 2007; 25: 265–270. 13. Dreicer R, Li S, Manola J et al. Phase 2 trial of epothilone B analog BMS-247550 (ixabepilone) in advanced carcinoma of the urothelium (E-3800): a trial of the Eastern Cooperative Oncology Group. Cancer 2007; 110: 759–763. 14. Joly F, Houédé N, Noal S et al. Do patients with advanced urothelial carcinoma benefit from weekly paclitaxel chemotherapy? A GETUG phase II study. Clin Genitourin Cancer 2009; 7: E28–E33. 15. Lorusso V, Pollera CF, Antimi M et al. A phase II study of gemcitabine in patients with transitional cell carcinoma of the urinary tract previously treated with platinum. Italian Co-operative Group on Bladder Cancer. Eur J Cancer 1998; 34: 1208–1212. 16. McCaffrey JA, Hilton S, Mazumdar M et al. Phase II trial of docetaxel in patients with advanced or metastatic transitional-cell carcinoma. J Clin Oncol 1997; 15: 1853–1857.
Volume 27 | No. 1 | January 2016
Downloaded from http://annonc.oxfordjournals.org/ by guest on December 23, 2015
2015 and were not included in this analysis, like the mature results with the use of pemetrexed in the salvage setting or the results of standard arms of randomized studies with new drugs [54, 55]. Unfortunately, we could not identify any patient or treatment characteristics that may portend a clinical benefit from a certain treatment, due to the inherent limitations of the trial-level meta-analysis. It is possible that, at least for a small proportion of patients, the use of double or even triple combination chemotherapy may result in a significant OS advantage compared with what is expected from the single-agents. Actually, this OS advantage has emerged from a retrospective individual patient-level analysis of 370 multicentric patients. Sonpavde et al. reported incrementally significant improvements in OS with the addition of one or two agents to taxanes compared with taxanes alone, although the results are pending a validation through a prospective clinical trial [56]. This apparent superiority of taxane-based combination chemotherapy compared with single-agent taxanes was partially confirmed here, in particular for OR and PFS, but not for OS. Taxanes were largely represented in more than 80% of doublet chemotherapy arms. As single agents, the efficacy was clearly overlapping to that of other regimens (as provided in Table 3), and this is an indirect evidence corroborating the use of either taxanes or vinflunine as equally active single agents in clinical practice. The present meta-analysis focused on two-drug combinations only, because the use of triple chemotherapy regimens is rarely possible in clinical practice and very few studies are available. Indeed, the possibility to combine two or more chemotherapeutic agents in the salvage therapy for UC may be limited due to poor performance status or comorbidities in most cases. Although we did not find significant differences in the incidence of severe toxicities between the two groups, even after adjusting for some important clinical factors, the critical limitation results from analyzing patients who have been included in clinical trials, who may be quite different from the overall population of advanced and relapsing UC. Looking at the next steps, the advent of the immune checkpoint inhibitors targeting the programmed death-1/ligand-1 (PD1/PD-L1) is revolutionizing the salvage setting and several active compounds are shedding light on the horizon and will likely be granted FDA and EMA approval for this indication [57–59]. Pivotal phase III trials are ongoing with these drugs having single-agent chemotherapy in the comparator arm in each case. The same consideration may apply to other promising targeted agents that are in development for molecularly defined cohort of patients [1]. Of course, the association that we will observe between the clinical and biological characteristics of eligible patients will impact the design of the next clinical trials. The possibility to have potentially active compounds in the salvage setting may lead to an extended OS for at least biologically defined patients. For these patients, the advantage may even be greater with the use of chemobiologic doublets compared with the benchmark of single-agent chemotherapy, although this notion remains to be proved. In conclusion, in the present meta-analysis comparing singleagent to doublet second-line chemotherapy for UC, we identified a better activity of the latter in terms of ORR and PFS, but we did not find any statistically significant difference in OS. The general recommendation is to continue administering singleagent taxanes or vinflunine outside of clinical trials, although a
Annals of Oncology
Annals of Oncology
Volume 27 | No. 1 | January 2016
38. Sternberg CN, Calabrò F, Pizzocaro G et al. Chemotherapy with an every-2-week regimen of gemcitabine and paclitaxel in patients with transitional cell carcinoma who have received prior cisplatin-based therapy. Cancer 2001; 92: 2993–2998. 39. Sweeney CJ, Williams SD, Finch DE et al. A phase II study of paclitaxel and ifosfamide for patients with advanced refractory carcinoma of the urothelium. Cancer 1999; 86: 514–518. 40. Vaishampayan UN, Faulkner JR, Small EJ et al. Phase II trial of carboplatin and paclitaxel in cisplatin-pretreated advanced transitional cell carcinoma: a Southwest Oncology Group Study. Cancer 2005; 104: 1627–1632. 41. Kanai K, Kikuchi E, Ohigashi T et al. Gemcitabine and paclitaxel chemotherapy for advanced urothelial carcinoma in patients who have received prior cisplatin-based chemotherapy. Int J Clin Oncol 2008; 13: 510–514. 42. Suyama T, Ueda T, Fukasawa S et al. Combination of gemcitabine and paclitaxel as second-line chemotherapy for advanced urothelial carcinoma. Jpn J Clin Oncol 2009; 39: 244–250. 43. Ikeda M, Matsumoto K, Tabata K et al. Combination of gemcitabine and paclitaxel is a favorable option for patients with advanced or metastatic urothelial carcinoma previously treated with cisplatin-based chemotherapy. Jpn J Clin Oncol 2011; 41: 1214–1220. 44. Soga N, Onishi T, Arima K, Sugimura Y. Paclitaxel carboplatin chemotherapy as a second-line chemotherapy for advanced platinum resistant urothelial cancer in Japanese cases. Int J Urol 2007; 14: 828–832. 45. Kouno T, Ando M, Yonemori K et al. Weekly paclitaxel and carboplatin against advanced transitional cell cancer after failure of a platinum-based regimen. Eur Urol 2007; 52: 1115–1122. 46. Dreicer R, Manola J, Schneider DJ et al. Phase II trial of gemcitabine and docetaxel in patients with advanced carcinoma of the urothelium: a trial of the Eastern Cooperative Oncology Group. Cancer 2003; 97: 2743–2747. 47. Fechner G, Siener R, Reimann M et al. Randomised phase II trial of gemcitabine and paclitaxel second-line chemotherapy in patients with transitional cell carcinoma (AUO Trial AB 20/99). Int J Clin Pract 2006; 60: 27–31. 48. Uhm JE, Lim HY, Kim WS et al. Paclitaxel with cisplatin as salvage treatment for patients with previously treated advanced transitional cell carcinoma of the urothelial tract. Neoplasia 2007; 9: 18–22. 49. Friedland DM, Dakhil S, Hollen C et al. A phase II evaluation of weekly paclitaxel plus carboplatin in advanced urothelial cancer. Cancer Invest 2004; 22: 374–382. 50. Park JH, Lee JL. Intravenous 3-weekly paclitaxel and metronomic oral cyclophosphamide in patients with advanced urothelial cancer previously treated with gemcitabine and platinum. Cancer Chemother Pharmacol 2015; 75: 247–254. 51. Srinivas S, Harshman LC. A phase II study of docetaxel and oxaliplatin for secondline treatment of urothelial carcinoma. Chemotherapy 2009; 55: 321–326. 52. Grivas PD, Hussain M, Hafez K et al. A phase II trial of neoadjuvant nab-paclitaxel, carboplatin, and gemcitabine (ACaG) in patients with locally advanced carcinoma of the bladder. Urology 2013; 82: 111–117. 53. Hutson TE, Vukelja S, Atienza D et al. Phase I study of a 3-drug regimen of gemcitabine/cisplatin/pemetrexed in patients with metastatic transitional cell carcinoma of the urothelium. Invest New Drugs 2008; 26: 151–158. 54. Bambury RM, Benjamin DJ, Chaim JL et al. The safety and efficacy of single-agent pemetrexed in platinum resistant advanced urothelial carcinoma: a large single institution experience. Oncologist 2015; 20: 508–575. 55. Petrylak DP, Tagawa S, Kohli M et al. Interim results of a randomized phase 2 study of docetaxel with ramucirumab versus docetaxel in second-line advanced or metastatic urothelial carcinoma. J Clin Oncol 2015; 33(suppl 7): 295a. 56. Sonpavde G, Pond GR, Choueiri TK et al. Single agent taxane versus taxane containing combination chemotherapy as salvage therapy for advanced urothelial carcinoma. Eur Urol 2015 Aug 8 [epub ahead of print], doi: 10.1016/j.eururo.2015.07.042. 57. Powles T, Eder JP, Fine GD et al. MPDL3280A (anti-PD-L1) treatment leads to clinical activity in metastatic bladder cancer. Nature 2014; 515: 558–562. 58. Plimack ER, Bellmunt J, Gupta S et al. Pembrolizumab (MK-3475) for advanced urothelial cancer: updated results and biomarker analysis from KEYNOTE-012. J Clin Oncol 2015; 33(suppl): abstr 4502. 59. Petrylak DP, Powles T, Bellmunt J et al. A phase is study of MPDL3280A (antiPDL1): updated response and survival data in urothelial bladder cancer (UBC). J Clin Oncol 2015; 33(suppl): abstr 4501.
doi:10.1093/annonc/mdv509 |
Downloaded from http://annonc.oxfordjournals.org/ by guest on December 23, 2015
17. Sweeney CJ, Roth BJ, Kabbinavar FF et al. Phase II study of pemetrexed for second-line treatment of transitional cell cancer of the urothelium. J Clin Oncol 2006; 24: 3451–3457. 18. Vaughn DJ, Broome CM, Hussain M et al. Phase II trial of weekly paclitaxel in patients with previously treated advanced urothelial cancer. J Clin Oncol 2002; 20: 937–940. 19. Vaughn DJ, Srinivas S, Stadler WM et al. Vinflunine in platinum-pretreated patients with locally advanced or metastatic urothelial carcinoma: results of a large phase 2 study. Cancer 2009; 115: 4110–4117. 20. Witte RS, Elson P, Bono B et al. Eastern Cooperative Oncology Group phase II trial of ifosfamide in the treatment of previously treated advanced urothelial carcinoma. J Clin Oncol 1997; 15: 589–593. 21. Winquist E, Vokes E, Moore MJ et al. A phase II study of oxaliplatin in urothelial cancer. Urol Oncol 2005; 23: 150–154. 22. Witte RS, Manola J, Burch PA et al. Topotecan in previously treated advanced urothelial carcinoma: an ECOG phase II trial. Invest New Drugs 1998; 16: 191–195. 23. Ko YJ, Canil CM, Mukherjee SD et al. Nanoparticle albumin-bound paclitaxel for second-line treatment of metastatic urothelial carcinoma: a single group, multicentre, phase 2 study. Lancet Oncol 2013; 14: 769–776. 24. Lee JL, Ahn JH, Park SH et al. Phase II study of a cremophor-free, polymeric micelle formulation of paclitaxel for patients with advanced urothelial cancer previously treated with gemcitabine and platinum. Invest New Drugs 2012; 30: 1984–1990. 25. Akaza H, Naito S, Usami M et al. Efficacy and safety of gemcitabine monotherapy in patients with transitional cell carcinoma after cisplatin containing therapy: a Japanese experience. Jpn J Clin Oncol 2007; 37: 201–206. 26. Ponzato P, Vigani A, Pensa F et al. Second line chemotherapy with ifosfamide as outpatient treatment for advanced bladder cancer. Am J Clin Oncol 1997; 20: 519–521. 27. Papamichael D, Gallagher CJ, Oliver RT et al. Phase II study of paclitaxel in pretreated patients with locally advanced/metastatic cancer of the bladder and ureter. Br J Cancer 1997; 75: 606–607. 28. Sengelov L, Nielsen OS, Kamby C, von der Maase H. Systemic mitomycin C as secondline treatment for metastatic urothelial cancer. Acta Oncol 1995; 34: 978–979. 29. Loriot Y, Fizazi K, Carles J et al. Results of a phase II study of pralatrexate in patients with advanced/metastatic relapsed transitional cell carcinoma of the urinary bladder. J Clin Oncol 2012; 30(suppl): abstr 4574. 30. Pectasides D, Aravantinos G, Kalofonos H et al. Combination chemotherapy with gemcitabine and ifosfamide as second-line treatment in metastatic urothelial cancer. A phase II trial conducted by the Hellenic Cooperative Oncology Group. Ann Oncol 2001; 12: 1417–1422. 31. Lin CC, Hsu CH, Huang CY et al. Gemcitabine and ifosfamide as second-line treatment for cisplatin-refractory metastatic urothelial carcinoma: a phase II study. Anticancer Drugs 2007; 18: 487–491. 32. Gondo T, Ohori M, Hamada R et al. The efficacy and safety of gemcitabine plus cisplatin regimen for patients with advanced urothelial carcinoma after failure of M-VAC regimen. Int J Clin Oncol 2011; 16: 345–351. 33. Di Lorenzo G, Autorino R, Giordano A et al. FOLFOX-4 in pre-treated patients with advanced transitional cell carcinoma of the bladder. Jpn J Clin Oncol 2004; 34: 747–750. 34. Di Lorenzo G, Montesarchio V, Autorino R et al. Phase 1/2 study of intravenous paclitaxel and oral cyclophosphamide in pretreated metastatic urothelial bladder cancer patients. Cancer 2009; 115: 517–523. 35. Albers P, Park SI, Niegisch G et al. Randomized phase III trial of 2nd line gemcitabine and paclitaxel chemotherapy in patients with advanced bladder cancer: short-term versus prolonged treatment [German Association of Urological Oncology (AUO) trial AB 20/99]. Ann Oncol 2011; 22: 288–294. 36. Krege S, Rembrink V, Bogermann C et al. Docetaxel and ifosfamide as second line treatment for patients with advanced and metastatic urothelial cancer after failure of platinum chemotherapy: a phase 2 study. J Urol 2001; 165: 67–71. 37. Rozzi A, Salerno M, Bordin F et al. Weekly regimen of epirubicin and paclitaxel as second-line chemotherapy in patients with metastatic transitional cell carcinoma of urothelial tract: results of a phase II study. Med Oncol 2011; 28(suppl 1): S426–S432.
reviews