Lucentis versus avastin: needs must or devil drives?
Every so often an issue pops up its head that we are glad not to have to deal with. One such involves a new treatment for age-related macular degeneration.
New monoclonal antibodies against vascular endothelial growth factor (VGEF) have been shown to reduce choroidal neovascularisation of ARMD, because VGEF helps promote endothelial cell growth and increases vascular permeability. The antibodies usually have to be given by intravitreal injection, usually at intervals of a month or so. The result is prevention of vision loss and improved mean visual acuity [1].
In a large (716 patients), long (two year), randomised comparison of intravitreal ranibizumab 0.3 mg and 0.5 mg versus sham injections, significantly more patients gained 15 or more letters with ranibizumab (Figure 1). The NNT for a gain of 15 letters over two years was about 3 for 0.5 mg and 4 for 0.3 mg.
Figure 1: Percent of patients with age-related macular degeneration gaining at least 15 letters over 24 months of treatment
The problem
There are two sources of anti-VGEF monoclonal antibodies. One is ranibizumab (Lucentis), which is licensed for intravitreal injection in ARMD. The other is bevacizumab (Avastin), which is licensed for intravenous infusion in metastatic colorectal cancer. Avastin contains anti-VGEF monoclonals at much higher dose than for intravitreal injection, so many intravitreal doses could be made from a single intravenous dose. In consequence, while ranibizumab (Lucentis) costs $2,000 per intravitreal injection, bevacizumab (Avastin) costs $50 for an equivalent dose.
What makes them different, as the manufacturer would tell us, includes the following:
- Avastin contains no preservatives, so there could be problems in keeping it sterile when split into small quantities required for retinal treatment.
- No pre-clinical trial toxicity data exists for use of Avastin in retinal therapy.
- The half-life of Avastin is different from Lucentis, in that it clears from the system 100 times slower. This is important for cancer use, but remaining in the retina for that length of time could be harmful.
- Lucentis binds more strongly to the VGEF protein than Avastin. It is this binding that blocks the protein from developing blood vessel growth into the retina (neovascularization).
- Avastin contains full-length antibodies, which can cause inflammation. The antibody fragments in Lucentis are 1/3 the size of Avastin antibodies, so they are capable of better penetration through the retinal layers.
- Manufacturing standards differ for cancer and ophthalmic drugs. Particulate matter must be very low in drugs used in the eye, and Avastin is not manufactured with that in mind.
It is also the case that there are some small case series, but no randomised trial evidence exists for benefit from bevacizumab (Avastin), nor much at all for harm, especially rare but serious harm.
The costs
These have been modelled in detail [2] to compare the costs and impact of these alternatives. Basically, to hit the NICE threshold of £30,000 per QALY, bevacizumab (Avastin) would have to be half at least as effective as ranibizumab (Lucentis). If the price of ranibizumab (Lucentis) was reduced by 75%, bevacizumab would have to be 5% more effective to achieve this threshold.
Based on about 25,000 new cases of neovascular ARMD a year in the UK, treatment with ranibizumab (Lucentis) would cost £300 million a year. Substituting with bevacizumab (Avastin) would save £292 million.
Comment
Both drugs come from the same company. That company would argue that the price reflects the cost of development and production. The problem for authorities like NICE, which has to make some difficult decisions over this, is that it has in the past argued against off-licence use of drugs. It would be a difficult philosophical position to change.
On the one hand we have reasonably good evidence for effect and safety over a long period, but with a high cost. On the other we have no good evidence of effect or safety, but at low and affordable cost. It makes for a difficult decision. A sensible resolution is required, perhaps a third way?
References:
- PJ Rosenfeld. Ranibizumab for neovascular age-related macular degeneration. New England Journal of Medicine 2006 355: 1419-1431.
- J Raftery et al. Ranibizumab (lucentis) versus bevacizumab (avastin): modelling cost effectiveness. British Journal of Ophthalmology 2007 DOI 10.1136/bjo.2007.116616