Implications of “not me” drugs for health systems: lessons from age related macular degeneration

Pharmacology of new drugs

Most cases of severe vision loss in AMD are caused by the proliferation of abnormal blood vessels beneath and within the retina.2 ^w2 Both ranibizumab and bevacizumab are monoclonal antibodies directed against vascular endothelial growth factor, which has a key role in the growth of these pathological vessels.3 ^w7-w10 They are the first treatments to consistently and substantially improve vision in patients with AMD.4 5 6 7

While ranibizumab was developed and approved for intravitreal injection to treat AMD,^{w7 w9} bevacizumab has been approved only for intravenous use as a cancer therapy.^w11 Ranibizumab is a smaller molecule than bevacizumab and consists of only the antigen binding fragment of the parent antibody. It was hypothesised that ranibizumab would penetrate the retina better and cause less inflammation than bevacizumab.^w7-w9 However, before phase III trials of ranibizumab were completed, physicians began using bevacizumab off-label for AMD, reasoning that the molecular target was the same for both drugs.^w12 Subsequently, the obvious efficacy of bevacizumab resulted in its rapid adoption in many countries.8 9

Comparative efficacy and safety

Until recently, high quality data showing that bevacizumab is effective in AMD were lacking, and this information gap heavily influenced treatment decisions made by patients, clinicians, and policy makers. However, in 2011 a National Institutes of Health sponsored trial comparing the two drugs for AMD (the Comparison of Age-Related Macular Degeneration Treatment Trials (CATT)) reported that bevacizumab and ranibizumab provide similar safety and efficacy.7 Additional studies will report in the near future, including the IVAN (alternative treatments to Inhibit VEGF in Age related choroidal Neovascularisation) trial in the UK. However, these trials were not powered to detect small but clinically relevant differences in adverse outcomes such as stroke, which could result from the differing systemic absorption and half lives of the two drugs.^{w10 w13} Ongoing post-marketing observational studies will need to examine these safety concerns.

Comparative cost

The quantity of bevacizumab administered by intravitreal injection in AMD is much less than that used intravenously in cancer treatment. As a result, the current price for bevacizumab is £30-£50 (€37-€62; $50-$80) per injection.7 ^w2 In contrast, the current price of ranibizumab is about £1250 per injection.7 Since most patients require monthly injections for extended periods, the absolute difference in total expenditures is usually thousands of pounds per patient.8 10

Regulatory approval, third party payers, and treatment choice

Typically, manufacturers design, conduct, and pay for clinical trials that provide data used by regulatory agencies when assessing a drug’s efficacy, safety, and quality. If a manufacturer does not want a drug to be used for a different indication from which it was licensed for, it can discourage use by not applying for regulatory approval for that indication, limiting supply, and questioning the drug’s safety and efficacy.1 11 12 ^{w1 w17 w18} Although bevacizumab for AMD is probably the first example of a “not me” drug, the proliferation of expensive biological drugs and industry takeovers of competitors have the potential to create similar situations in which a company is motivated to discourage the use of one of its own drugs to increase sales of another.

Most patients with AMD who have to pay their own prescription costs choose bevacizumab over ranibizumab, even though it does not have regulatory approval.8 With the potential to save billions of dollars, many third party payers, both public and private, are also keen to use the cheaper drug. For example, US Medicare covers both bevacizumab and ranibizumab for AMD. Additionally, Medicare contractors and the Centers for Medicare and Medicaid Services (CMS) have recognised the costs associated with dividing bevacizumab into small quantities for intraocular injection in the fees they pay to physicians.13 14 ^w19 Most coverage decisions for Medicare are made by regional for-profit carriers, ^w20 and this seems to have provided the latitude and motivation to cover off-label bevacizumab.15 With both drugs covered by Medicare, drug copayments make ranibizumab much more expensive for Medicare patients. Consequently, bevacizumab is used in about 60% of injections for AMD in this population.9

Despite considerable use of bevacizumab, ranibizumab sales in the US still surpassed $1.8bn in 2010, which will propel ranibizumab into the single greatest drug expenditure under Medicare Part B.16 ^w21 This suggests that many US patients are either insensitive to the price of ranibizumab because Medicare covers the majority of the cost (with private insurance potentially covering copayments) or that US patients face barriers to accessing bevacizumab for AMD. In response, public leaders have recently requested a national coverage determination supporting the use of bevacizumab for AMD in order to improve access to care.^w22

In contrast to the US Medicare system, many public formularies in countries with centralised, structured drug review processes such as the UK, Canada, and Australia have been constrained by the fact that they generally only pay for drugs that have been granted regulatory approval for the covered indication.17 18 ^{w23 w24} For example, the Ontario Public Drug Programs, which pay for prescription medications for everyone in the province who is 65 years of age and older, fund ranibizumab but not bevacizumab. As a result, use of ranibizumab in Ontario is four times the rate seen in the US Medicare population.8 9

Other jurisdictions with structured third party payer drug evaluation processes have taken an alternative approach, and have bypassed these processes altogether. For instance, the Canadian province of British Columbia covers intravitreal injections of bevacizumab through the publicly funded Provincial Health Services Authority, which is not strictly subject to the usual evaluation processes for listing a drug on the formulary of medications that the province covers.^w26 While pragmatic, this approach may be difficult to replicate in other jurisdictions and may not provide for consistent and transparent decision making.

In the UK, as in the US and Canada, coverage and use of bevacizumab varies by jurisdiction, with primary care trusts formulating their own policies.^{w27 w28} Hence, the UK Department of Health, through the National Institute for Health and Clinical Excellence (NICE), has begun working towards formally evaluating bevacizumab for AMD.19 ^w29 This has required substantial new collaborative efforts and the development of new processes. In particular, NICE has led consultations to evaluate the feasibility of formally appraising bevacizumab for AMD. This consultation was carried out at the request of the UK Department of Health and involved representation from government, primary care trust administration, industry, doctors, and patients. The Department of Health is expected to ask NICE to proceed with a formal assessment after the results of the IVAN trial are published. The outcomes of these novel policy directions should serve as an important learning opportunity for jurisdictions with centralised formularies and may help guide US Medicare national coverage determination processes as well.

Rational drug choices

While economic evaluations have supported the adoption of bevacizumab for AMD,^w30 regulators have been challenged by the unprecedented unwillingness of the drug’s manufacturer to submit an application for its use in AMD. However, third party payer coverage of bevacizumab, not regulatory approval, will be the primary factor determining treatment, as shown by its widespread use in some jurisdictions. Further, major moves to facilitate access to bevacizumab are taking place among payers and cost effectiveness evaluators including the CMS, NICE, and Canadian provincial drug programmes. In response, Novartis launched legal action this month in the UK aimed at forcing hospitals to use ranibizumab.20

Several other factors may also influence the decision to choose bevacizumab or ranibizumab. Although CATT has shown bevacizumab and ranibizumab to be of similar efficacy and safety in a trial setting, numerous instances of infection have been reported in general use, probably because of suboptimal infection control practices during the repackaging of bevacizumab for intravitreal injection.21 This recently prompted the US Veterans Administration healthcare system to suspend coverage of bevacizumab for AMD.^w31

Fear of litigation from patients with rare but serious adverse events,^{w32 w33} and fear of legal action by industry, as has occurred in Germany,^w17 may also provide disincentives to the use of bevacizumab. These fears are especially influential in systems with third party payers, where neither the patient nor the doctor is responsible for costs but they both bear the risks of adverse outcomes and litigation.

Financial considerations may also influence doctors’ choices. Genentech has paid US doctors who give large numbers of ranibizumab injections.22 Furthermore, Medicare part B reimbursement to physicians for drugs is directly proportional to the price of the drug used, and physicians are able to retain a portion of this reimbursement as income.10 ^w34 Without this unintended incentive to use ranibizumab, US Medicare policy would likely be even more effective in promoting the use of bevacizumab.

Lessons and future directions

The use of public funds to support comparative trials of bevacizumab and ranibizumab, such as the CATT and IVAN trials, was predicated on the notion that the less expensive option would be preferred if the drugs were of equivalent efficacy. Despite the results of the CATT study, ranibizumab remains the dominant therapy in jurisdictions such as Ontario, where it is fully reimbursed by public drug programmes, and maintains a surprisingly large market share in others, such as the US Medicare population, even though patients are at least partially exposed to price. Clearly, appropriate authorities, policies, and processes are needed to ensure that bevacizumab is available, that there is a structure in place that ensures accountability for drug safety and quality, and that patients are making fully informed choices.

Support for the use of bevacizumab from professional organisations such as the American Academy of Ophthalmology, the Royal College of Ophthalmologists, and the Canadian Ophthalmological Society would provide doctors with a clear definition of the standard of care, and might help to allay fears of legal liability arising from off label use of bevacizumab.^w35 In many instances, uptake of bevacizumab for AMD will also require health technology assessors such as NICE and the CMS to validate its use.

Another barrier to use in North America and the UK, is that both drug and physician regulatory bodies have policies that discourage the use of off-label and off-licence therapies. These well intentioned policies were created because evidence supporting most off-label drug uses is lacking. By contrast, the level of evidence supporting the use of bevacizumab in AMD is of the highest quality.6 7 Hence, such policies should be refined to allow for a case by case evaluation of evidence.

The process of designing appropriate, precedent setting responses to the bevacizumab-ranibizumab conundrum could be made more comprehensive and efficient through international communication among health technology assessors, as well as among drug and physician regulatory bodies. With enhanced recognition that drug policies have influenced the uptake and costs of bevacizumab and ranibizumab in countries such as the UK, the US, and Canada, improved sharing and dissemination of detailed information on new policies and their effects could guide evidence informed healthcare internationally.