Rini 2010

Methods	Study design: 2‐arm, parallel‐group RCT. Study dates: randomization from October 2003 to July 2005, data cutoff March 2009, median follow‐up among surviving participants 46.2 months. Setting: multicentre, international. Countries: Canada, US.
Participants	Inclusion criteria: people with mRCC; clear‐cell histological component confirmed by local pathology review; no prior systemic therapy for RCC; KPS ≥ 70%; aged ≥ 18 years; adequate bone marrow, hepatic and renal function; serum creatinine ≤ 1.5 times ULN. Exclusion criteria: central nervous system metastases; NYHA class II to IV heart failure; bleeding within 6 months; blood pressure that could not be controlled < 160/90 mmHg with medication; history of venous thrombosis within 1 year or arterial thrombosis within 6 months or who required ongoing therapeutic anticoagulation; uncontrolled thyroid function; pregnancy; requirement for systemic corticosteroids greater than physiological replacement doses or delayed healing wounds, ulcers or bone fractures. Sample size:732. Age (years, median with IQR): group 1: 61 (56 to 70); group 0: 62 (55 to 70). Sex (M/F, %): group 1: 73/27; group 0: 66/34. Prognostic factors: performance status (ECOG Performance Status 0/1/2, %): 62/37/1; prior nephrectomy (%): 85; prior systemic therapies (%): 0 (as per inclusion criteria); risk prognosis (MSKCC) risk score poor/intermediate/high (%): 10/64/26.
Interventions	Group 1 (n = 363): IFN‐α IFN‐α‐2a (Intron; Schering‐Plough, Kenilworth, NJ), provided by the NCI Cancer Therapy Evaluation Program, 9 MU SC 3 times/week (non‐consecutive days). Dose reduction to 6 MU and 3 MU if IFN‐related toxicity present. Group 0 (n = 369): IFN‐α + bevacizumab IFN‐α‐2a (Intron; Schering‐Plough, Kenilworth, NJ), provided by the NCI Cancer Therapy Evaluation Program, 9 MU SC 3 times/week (non‐consecutive days). Dose reduction to 6 MU and 3 MU if IFN‐related toxicity present. Bevacizumab (provided by the NCI Cancer Therapy Evaluation Program) 10 mg/kg IV every 2 weeks. Cointerventions (standard‐of‐care according to local practice): not specified.
Outcomes	OS (primary outcome) How measured: time from registration to death from any cause. Time points measured: during treatment and follow‐up. Time points reported: Kaplan‐Meier curves over up to 60 months, median OS. Subgroups: nephrectomy, MSKCC, liver metastases, age, gender. AEs, grade ≥ 3 (secondary outcome) How measured: ongoing documentation of AEs (CTCAE v.3.0). Time points measured: baseline, every 12 weeks. Time points reported: frequency of participants with AEs grade ≥ 3, deaths due to AEs, treatment‐related AEs to March 2009. Subgroups: not reported. QoL not evaluated PFS (secondary outcome) How measured: time between randomization and date of progression or death, investigator assessment of x‐rays. Time points measured: baseline, every 12 weeks. Time points reported: Kaplan‐Meier survival curves for up to 60 months, median PFS. Subgroups: number of adverse risk factors. Tumour remission (secondary outcome) How measured: investigator assessment of x‐rays, RECIST criteria. Time points measured: baseline, every 12 weeks. Time points reported: overall response rate. Subgroups: not reported.
Funding sources	Supported in part by National Cancer Institute to the Cancer and Leukemia Group B (CALGB) (Grant No. CA31946, CA33601) and by National Cancer Institute (Grants No. CA60138, CA41287, CA47642, CA45808, CA77440, CA14985, CA77202).
Declarations of interest	BIR: consultant or advisory role: Genentech. WMS: consultant or advisory role: Genentech; research funding: Genentech. JP: consultant or advisory role: Genentech; honoraria: Genentech. JD: consultant or advisory role: Genentech and Novartis; honoraria: Pfizer, Novartis; research funding: Novartis, Genentech, Pfizer. DAV: research funding: Genentech.
Notes	Registration: NCT00072046 (CALGB 90206). Results on PFS and overall response rate published in Rini 2008, no cross‐over was permitted for participants randomly assigned to IFN‐α monotherapy, a substantial percentage of participants in both arms received systemic anticancer therapy subsequent to progression (62% of participants on IFN‐ monotherapy and 54% of participants on bevacizumab + IFN‐α) (mostly with sunitinib or sorafenib).
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Stratified random block design, stratified by nephrectomy status and number of adverse prognostic factors.
Allocation concealment (selection bias)	Unclear risk	No information provided.
Blinding of participants and personnel (performance bias) Subjective outcomes	High risk	No blinding.
Blinding of participants and personnel (performance bias) Objective outcomes	Low risk	No performance bias on OS assumed.
Blinding of outcome assessment (detection bias) Subjective outcomes	High risk	Non‐blinded trial and no independent review of x‐rays could potentially have contributed to the improved PFS and overall response rate.
Blinding of outcome assessment (detection bias) Objective outcomes	Low risk	No detection bias on OS assumed.
Incomplete outcome data (attrition bias) (OS and PFS)	Low risk	No attrition bias assumed due to similar censoring, 657/732 (90%) participants experienced progression or death.
Incomplete outcome data (attrition bias) (safety)	Low risk	Based on all participants who were eligible for evaluation for toxicity (362/363 from the intervention and 347/369 from the control group, reasons not reported).
Incomplete outcome data (attrition bias) (tumour remission)	Unclear risk	No data reported.
Incomplete outcome data (attrition bias) (QoL)	Unclear risk	Not evaluated.
Selective reporting (reporting bias)	Low risk	No differences in outcomes to the protocol.
Other bias	High risk	Treatment with second‐line systemic anticancer therapy subsequent to progression (62% of participants on IFN‐α monotherapy and 54% of participants on bevacizumab + IFN‐α) might bias OS.