Methods

Parallel RCT

Randomisation ratio: 1:1

Superiority design

Participants

Inclusion criteria:

• Histologically or cytologically confirmed urothelial carcinoma of the renal pelvis, ureter, bladder, or urethra that showed predominantly transitional‐cell features on histologic testing.

• Progression or recurrence of urothelial cancer following a first‐line platinum‐containing regimen (e.g. cisplatin, carboplatin) for metastatic or inoperable locally advanced disease; or adjuvant platinum‐based therapy following cystectomy for localized muscle‐invasive urothelial cancer with recurrence/progression <=12 months following completion of therapy; or neoadjuvant platinum‐containing therapy prior to cystectomy for localized muscle‐invasive urothelial cancer with recurrence <=12 months following completion of therapy.

• No more than 2 prior lines of systemic chemotherapy for metastatic urothelial cancer.

• Able to provide tissue for biomarker analysis from an archival tissue sample or newly obtained core or excisional biopsy of a tumour lesion not previously irradiated.

• Measurable disease.

• ECOG performance status of 0, 1, or 2.

• Adequate organ function.

• Female participants of childbearing potential have a negative urine or serum pregnancy test and willing to use 2 acceptable methods of birth control or abstain from heterosexual activity for the course of the study through 120 days after the last dose of pembrolizumab or 180 days after the last dose of paclitaxel, docetaxel, or vinflunine; or are surgically sterile.

• Male participants must be willing to use an adequate method of contraception starting with the first dose of study medication through 120 days after the last dose of pembrolizumab or 180 days after the last dose of paclitaxel, docetaxel, or vinflunine.

Exclusion criteria:

• Urothelial cancer that is suitable for local therapy administered with curative intent.

• Currently participating in or has participated in a study of an investigational agent or using an investigational device within 4 weeks prior to the first dose of trial medication.

• Diagnosis of immunodeficiency or receiving systemic steroid therapy or any other form of immunosuppressive therapy within 7 days prior to the first dose of study medication.

• Anti‐cancer mAb within 4 weeks prior to study Day 1.

• Not recovered from adverse events due to agents administered more than 4 weeks earlier, prior chemotherapy, targeted small molecule therapy, or radiation therapy within 2 weeks of study Day 1.

• Not recovered from adverse events due to a previously administered agent or prior therapy with all choices of active comparator.

• Known additional malignancy that is progressing or requires active treatment with the exception of basal cell carcinoma of the skin, squamous cell carcinoma of the skin that has undergone potentially curative therapy, or in situ cancer; or prostate cancer that was identified incidentally following cystoprostatectomy for bladder cancer that is Stage T2N0M0 or lower, Gleason score<= 6, or PSA undetectable.

• Known active CNS metastases and/or carcinomatous meningitis.

• Active autoimmune disease requiring systemic treatment within the past 3 months or a documented history of clinically severe autoimmune disease, or a syndrome that requires systemic or immunosuppressive agents, active cardiac disease, evidence of interstitial lung disease or active non‐infectious pneumonitis, or active infection requiring systemic therapy.

• History of severe hypersensitivity reaction to paclitaxel, docetaxel, or to other drugs formulated with polysorbate 80 or polyoxyethylated castor oil, or to vinflunine or other vinca alkaloids.

• Requires ongoing therapy with a medication that is a strong inhibitor or inducer of the cytochrome 3A4 (CYP3A4) enzymes.

• Pregnant, breastfeeding, or expecting to conceive or father children within the projected duration of the trial, starting with the screening visit through 120 days after the last dose of pembrolizumab or 180 days after the last dose of paclitaxel, docetaxel, or vinflunine.

• Prior therapy with an anti‐PD‐1 or anti‐PD‐ligand 1 agent, or with an agent directed to another co‐inhibitory T‐cell receptor.

• HIV or active hepatitis B or hepatitis C.

• Received a live virus vaccine within 30 days of planned start of trial treatment.

Characteristics:

• Total number randomly assigned: 542.

• Baseline imbalances: no.

• Mean (range) age: pembrolizumab 67 (29‐88) years, chemotherapy 65 (26‐84) years.

• Number of patients with ECOG 0: pembrolizumab 119, chemotherapy 106.

• Number of patients with ECOG 1: pembrolizumab 143, chemotherapy 158.

• Number of patients with ECOG ≥2: pembrolizumab 2, chemotherapy 4.

• Number of patients with liver metastases: pembrolizumab 91, chemotherapy 95.

• Number of patients with visceral metastases: pembrolizumab 240, chemotherapy 233.

• Male sex: pembrolizumab 200 (74.1%), chemotherapy 202 (74.3%).

Interventions

Number of study centres: 748 patients were screened for enrolment in 120 sites in 29 countries (Australia, Austria, Belgium, Canada, Chile, Denmark, France, Germany, Hungary, Ireland, Israel, Italy, Japan, Republic of Korea, Netherlands, New Zealand, Norway, Peru, Poland, Portugal, Puerto Rico, Romania, Singapore, Spain, Sweden, Taiwan, Turkey, United Kingdom, United States).

Run‐in period: 11/2014‐11/2015.

Extension period: no.

Intervention: pembrolizumab 200 mg IV on Day 1 Q3W; 270 randomised patients, 266 patients received treatment.

Comparison: paclitaxel 175 mg/m² IV or docetaxel 75 mg/m² IV or vinflunine 320 mg/m² IV, on Day 1 Q3W, 272 randomised patients, 255 received treatment, 84 patients received docetaxel, 84 received paclitaxel, and 87 received vinflunine.

Other co‐interventions for both groups: no.

Outcomes

Primary outcome measures

Overall survival:

• time points measured: up to 30 months

• time point reported: median duration of follow‐up was 14.1 months (range 9.9 to 22.1)

• outcome definition: overall survival was defined as the time from randomisation to death from any cause

• subgroups (of interest): performance status (ECOG 0 or 1 versus ≥ 2); time since last chemotherapy administration (< three months versus ≥ three months); degree of pretreatment; PDL‐1 tumour expression status (positive versus negative; 1% cutoff)

Progression‐free survival:

• time points measured: up to 30 months

• time point reported: median duration of follow‐up was 14.1 months (range 9.9 to 22.1)

• outcome definition: progression‐free survival was defined as the time from randomisation to disease progression or death from any cause, per RECIST 1.1

• subgroups: none (subgroup of participants with PD‐L1 positive score (10% cutoff) not relevant for this review)

Secondary outcome measures

Objective response rate:

• time points measured: up to 30 months

• time point reported: median duration of follow‐up was 14.1 months (range 9.9 to 22.1)

• outcome definition: per RECIST 1.1; complete/partial response

• subgroups: none

Progression‐free survival per modified RECIST 1.1:

• time points measured: up to 30 months

• time point reported: median duration of follow‐up was 14.1 months (range 9.9 to 22.1)

• outcome definition: per modified RECIST 1.1

• subgroups: none

Objective response reaction per modified RECIST 1.1:

• time points measured: up to 30 months

• time point reported: median duration of follow‐up was 14.1 months (range 9.9 to 22.1)

• outcome definition: per modified RECIST 1.1

• subgroups: none

Adverse event (time frame: up to 31 months):

• time points measured: up to 30 months

• time point reported: median duration of follow‐up was 14.1 months (range 9.9 to 22.1)

• subgroups: none

Discontinuation of study drug due to an adverse event:

• time points measured: up to 30 months

• time point reported: median duration of follow‐up was 14.1 months (range 9.9 to 22.1)

• subgroups: none

Quality of life:

• was not listed as predefined outcome in ClinicalTrials.gov registry (NCT02256436)

Funding sources

Supported by Merck (Kenilworth, NJ)

Declarations of interest

• Dr Bellmunt reports receiving consulting fees from Merck and Genentech.

• Dr de Wit reports receiving fees for serving on an advisory board for Eli Lilly.

• Dr Vaughn reports receiving fees for serving on a data and safety monitoring board from Astellas Pharma.

• Dr Fradet reports receiving consulting fees from Astellas Pharma, Bayer, Amgen, Merck, Roche, Sanofi, and AstraZeneca, and grant support from Astellas Pharma, Amgen, and AstraZeneca.

• Dr Lee reports receiving fees for serving on advisory boards for AstraZeneca, Astellas Pharma, Eisai, and Pfizer.

• Dr Fong reports receiving grant support from Dendreon, Bristol‐Myers Squibb, Roche/Genentech, AbbVie, and Amgen

• Dr. Climent reports receiving fees for serving on advisory boards for Roche, Pierre Fabre Laboratories, and Bristol‐Myers Squibb.

• Dr Petrylak reports receiving consulting fees from Millennium Pharmaceuticals, Dendreon, Sanofi Aventis, Roche, Bayer, Johnson & Johnson, Exelixis, Ferring Pharmaceuticals, Medivation, Pfizer, Bellicum Pharmaceuticals, and Tyme Pharmaceuticals; grant support from Merck, Celgene, Agensys, Eli Lilly, Millennium Pharmaceuticals, Dendreon, Sanofi Aventis, and Roche Laboratories; and having ownership interest or involvement in Bellicum Pharmaceuticals and Tyme Pharmaceuticals.

• Dr Choueiri reports receiving fees for serving on an advisory board for Merck.

• Dr Necchi reports receiving consulting fees from Merck Sharp & Dohme, Roche, AstraZeneca, and Bayer, and grant support from Millennium Pharmaceuticals, Amgen, Novartis Pharmaceuticals, Merck Sharp & Dohme, Roche, and AstraZeneca.

• Dr Gerritsen reports receiving fees for serving on advisory boards from Amgen, Astellas Pharma, Janssen‐Cilag, CureVac, Merck Sharp & Dohme, and Bayer; lecture fees from Bristol‐Myers Squibb, Janssen‐Cilag, and Bayer; and grant support from Astellas Pharma and Janssen‐Cilag.

• Dr Gurney reports receiving fees for serving on advisory boards for Bristol‐Myers Squibb, Astellas Pharma, and Roche, and grant support from Pfizer.

• Dr Quinn reports receiving fees for serving on advisory boards for Merck, Bristol‐Myers Squibb, Genentech, AstraZeneca, Pfizer, Novartis Pharmaceuticals, Astellas Pharma, Peloton Therapeutics, Dendreon, Exelixis, Bayer, Piramal Life Sciences, and Sanofi.

• Dr Sternberg reports receiving honoraria from OncoGeneX, Eli Lilly, Merck, and Bristol‐Myers Squibb, and grant support from Eli Lilly, Janssen Pharmaceuticals, AstraZeneca, and Roche/Genentech.

• Drs Mai, Poehlein, and Perini report being employees of and holding stock or stock options in Merck.

• Dr Poehlein reports having been employed in the past by and holding stock options in Bayer Health Care Oncology and holding a patent related to the augmentation of immune response to a cancer vaccine (U.S. patent number, 20090317407A1).

• Dr Perini reports holding pending patents related to the combination of a PD‐1 antagonist and a listeria‐based vaccine for treating prostate cancer (U.S. patient application number 20160022814 A1) and treating cancer with a combination of a PD‐1 antagonist and a vascular endothelial growth factor receptor inhibitor (U.S. patent application number 15101409).

• Dr Bajorin reports receiving lecture fees and travel support from Merck.

• No other potential conflict of interest relevant to this article was reported.

• The trial was designed by academic advisors and employees of the sponsor (Merck). Data were collected by investigators and their site personnel and analysed by statisticians who were employees of the sponsor. Results were interpreted by the academic authors, by authors who were employees of the sponsor, and by other employees of the sponsor who did not fulfil all the authorship criteria as outlined by the International Committee of Medical Journal Editors. An external data and safety monitoring committee oversaw the trial and assessed efficacy and safety at the time of prespecified interim analyses that were performed by statisticians from QuintilesIMS, with funding by the sponsor.

• The first draft of the manuscript was written by the first and last authors, with input from authors who were employees of the sponsor. Assistance with manuscript preparation was provided by a medical writer employed by the sponsor.

Notes

• "The data and safety monitoring committee reviewed the results of the second interim analysis on October 18, 2016, and recommended early termination of the trial because pembrolizumab met the superiority thresholds for overall survival in the co‐primary populations".

• Manuscript was published in English.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Quote from publication: "...central randomisation as described in the protocol".

Comment: Randomization was performed adequately.

Allocation concealment (selection bias)

Low risk

Quote from publication: "Centrally...interactive voice response system/integrated web response system".

Comment: allocation concealment was performed adequately.

Blinding of participants and personnel (performance bias)
Objective outcomes

Unclear risk

Quote from publication: "...there was no blinding".

Comment: overall survival was measured and reported. It might be conceivable that even objective outcomes are influenced by lack of blinding. We finally judge that there is an unclear risk of bias.

Blinding of participants and personnel (performance bias)
Subjective outcomes

High risk

Quote from publication: "...there was no blinding".

Comment: progression‐free survival, treatment‐related mortality, discontinuation due to adverse events, and adverse events were measured and reported (quality of life was not listed as a predefined outcome in ClinicalTrials.gov registry (NCT02256436) and was not reported in the full text publication, but results were presented in abstract form at ASCO GU 2017 conference (Bellmunt 2017)). We judge that subjective outcomes are likely to be influenced by lack of blinding leading to high risk of bias. For response rate, we are uncertain if lack of blinding of participants and personnel might have influenced results, and therefore judged that there is unclear risk of bias for this outcome.

Blinding of outcome assessment (detection bias)
Objective outcome (overall survival

Low risk

Quote from publication: no applicable quote.

Comment: Not likely that outcome assessment for overall survival would be influenced by lack of blinding.

Blinding of outcome assessment (detection bias)
Subjective outcomes (all except overall survival and response rate)

High risk

Quote from publication: no applicable quote.

Comment: We judged that the assessment of these subjective outcomes is likely to be influenced by the lack of blinding.

Blinding of outcome assessment (detection bias)
Subjective outcome (treatment response (partial or complete radiological)

Low risk

Quote from publication: "Imaging data were centrally reviewed, central imaging vendor were blinded to the subject treatment, imaging results were blinded to the clinical study team".

Comment: Adequate assurance of blinding

Incomplete outcome data (attrition bias)
Oncological outcomes

Low risk

All patients randomised were included in the analysis for overall survival, progression‐free survival, and response rate.

Incomplete outcome data (attrition bias)
Quality of life

Low risk

Missing outcome data are less than 10% in both groups (randomised: 270 pembrolizumab, 272 chemotherapy; in evaluation: 266 pembrolizumab, 253 chemotherapy). We judge that this number of withdrawals is not enough to have a clinically relevant effect.

Incomplete outcome data (attrition bias)
Treatment‐related mortality, Discontinuation due to adverse events, adverse events

Low risk

Missing outcome data are less than 10% in both groups (randomised: 270 pembrolizumab, 272 chemotherapy; in evaluation: 255 pembrolizumab, 255 chemotherapy). We judge that this number of withdrawals is not enough to have a clinically relevant effect.

Selective reporting (reporting bias)

Unclear risk

Quality of life was not listed as a predefined outcome in the ClinicalTrials.gov registry (NCT02256436)

Other bias

Low risk

No other potential bias identified.