Intravitreal ranibizumab versus aflibercept versus bevacizumab for macular oedema due to central retinal vein occlusion: the LEAVO non-inferiority three-arm RCT.

Abstract

BACKGROUND\nLicensed ranibizumab (0.5\u2009mg/0.05\u2009ml Lucentis®; Novartis International AG, Basel, Switzerland) and aflibercept (2\u2009mg/0.05\u2009ml Eylea®; Bayer AG, Leverkusen, Germany) and unlicensed bevacizumab (1.25\u2009mg/0.05\u2009ml Avastin®; F. Hoffmann-La Roche AG, Basel, Switzerland) are used to treat macula oedema due to central retinal vein occlusion, but their relative clinical effectiveness, cost-effectiveness and impact on the UK NHS and Personal Social Services have never been directly compared over the typical disease treatment period.\n\n\nOBJECTIVE\nThe objective was to compare the clinical effectiveness and cost-effectiveness of three intravitreal antivascular endothelial growth factor agents for the management of macula oedema due to central retinal vein occlusion.\n\n\nDESIGN\nThis was a three-arm, double-masked, randomised controlled non-inferiority trial.\n\n\nSETTING\nThe trial was set in 44 UK NHS ophthalmology departments, between 2014 and 2018.\n\n\nPARTICIPANTS\nA total of 463 patients with visual impairment due to macula oedema secondary to central retinal vein occlusion were included in the trial.\n\n\nINTERVENTIONS\nThe participants were treated with repeated intravitreal injections of ranibizumab (n\u2009=\u2009155), aflibercept (n\u2009=\u2009154) or bevacizumab (n\u2009=\u2009154).\n\n\nMAIN OUTCOME MEASURES\nThe primary outcome was an increase in the best corrected visual acuity letter score from baseline to 100 weeks in the trial eye. The null hypothesis that aflibercept and bevacizumab are each inferior to ranibizumab was tested with a non-inferiority margin of -5 visual acuity letters over 100 weeks. Secondary outcomes included additional visual acuity, and imaging outcomes, Visual Function Questionnaire-25, EuroQol-5 Dimensions with and without a vision bolt-on, and drug side effects. Cost-effectiveness was estimated using treatment costs and Visual Function Questionnaire-Utility Index to measure quality-adjusted life-years.\n\n\nRESULTS\nThe adjusted mean changes at 100 weeks in the best corrected visual acuity letter scores were as follows - ranibizumab, 12.5 letters (standard deviation 21.1 letters); aflibercept, 15.1 letters (standard deviation 18.7 letters); and bevacizumab, 9.8 letters (standard deviation 21.4 letters). Aflibercept was non-inferior to ranibizumab in the intention-to-treat population (adjusted mean best corrected visual acuity difference 2.23 letters, 95% confidence interval -2.17 to 6.63 letters; p\u2009=\u20090.0006), but not superior. The study was unable to demonstrate that bevacizumab was non-inferior to ranibizumab in the intention-to-treat population (adjusted mean best corrected visual acuity difference -1.73 letters, 95% confidence interval -6.12 to 2.67 letters; p\u2009=\u20090.071). A post hoc analysis was unable to demonstrate that bevacizumab was non-inferior to aflibercept in the intention-to-treat population (adjusted mean best corrected visual acuity difference was -3.96 letters, 95% confidence interval -8.34 to 0.42 letters; p\u2009=\u20090.32). All per-protocol population results were the same. Fewer injections were required with aflibercept (10.0) than with ranibizumab (11.8) (difference in means -1.8, 95% confidence interval -2.9 to -0.8). A post hoc analysis showed that more bevacizumab than aflibercept injections were required (difference in means 1.6, 95% confidence interval 0.5 to 2.7). There were no new safety concerns. The model- and trial-based cost-effectiveness analyses estimated that bevacizumab was the most cost-effective treatment at a threshold of £20,000-30,000 per quality-adjusted life-year.\n\n\nLIMITATIONS\nThe comparison of aflibercept and bevacizumab was a post hoc analysis.\n\n\nCONCLUSION\nThe study showed aflibercept to be non-inferior to ranibizumab. However, the possibility that bevacizumab is worse than ranibizumab and aflibercept by 5 visual acuity letters cannot be ruled out. Bevacizumab is an economically attractive treatment alternative and would lead to substantial cost savings to the NHS and other health-care systems. However, uncertainty about its relative effectiveness should be discussed comprehensively with patients, their representatives and funders before treatment is considered.\n\n\nFUTURE WORK\nTo obtain extensive patient feedback and discuss with all stakeholders future bevacizumab NHS use.\n\n\nTRIAL REGISTRATION\nCurrent Controlled Trials ISRCTN13623634.\n\n\nFUNDING\nThis project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 25, No. 38. See the NIHR Journals Library website for further project information.