Roche’s bevacizumab (Avastin) is in the news again. This has a reasonable claim to be a wonder drug, but for macular degeneration, a disease for which Roche refuses to license it. Instead Roche has tried and failed several times to have the drug recommended by NICE for lung, breast, and colorectal cancer, all at an extremely high price of around £2k per month or £24k per patient year.

NICE considered and rejected bevacizumab for metastatic colorectal cancer before in an appraisal that put the cost per QALY at £57k to £88k.

The current appraisal consultation document considers the use of bevacizumab as part of a new less restrictive oxaliplatin-containing combination of drugs for metastatic colorectal cancer. Roche has kept the drug price at the same level as 2007. This led to a cost per QALY from £105k to £108k cost which Roche then managed to reduce to £24k to £30k (depending on the combination therapy).

How did Roche get the ICER down to £24k to £30k? By a patient access scheme which has four elements: a) an upfront payment to the relevant NHS trust (undisclosed amount) for each patient, b) free bevacizumab after 12 months, c) oxaliplatin provided free by Roche and d) bevacizumab to be purchased at full list price.

Inclusion of a payment per patient is novel, amounting to a price discount. This could be used to generate any cost per QALY one wanted. This payment seems to have been a late addition to the scheme (“The Committee reviewed the revised base case ICERs provided by the company …which included an amendment to the patient access scheme of an additional upfront payment (designated by the manufacturer as commercial in confidence) of £30,000 per QALY gained (£105k without the patient access scheme) for B-XELOX compared with XELOX….”Para 4.11. XELOX refers to the drug combination, with and without Bevacizumab. £24k per QALY refers to another drug combination FOLFOX)

Patient access schemes have to be approved by the Department of Health. The department approved the scheme but expressed concerns about its complexity which it considered might limit its uptake by the NHS and suggested the appraisal committee “may wish to be mindful.”  
 
The appraisal committee considered that in the most plausible analysis bevacizumab offered no statistically significant gain in overall survival and a it only offered a gain in progression free survival of 1.4 months. It thus rejected it as a candidate for special consideration in terms of either end of life (not 3 months survival gain) or innovation (no step change).

The appraisal committee was also not convinced by Roche’s economic model which generated a cost per QALYof £30k, “the Comittee agreed that plausible adjustments to the key model inputs would increase these ICERs substantially”

Plausible adjustments included: patients were likely to continue to use bevacizumab longer than shown in the model, that the utility values for treatment were too high, that the cost of adverse events had not been included, and oxaliplatin has been mispriced.

Roche’s proposal to provide free supply of oxaliplatin, not one of its drugs, is also notable. The committee noted that Roche had priced oxaliplatin at its proprietary price not its generic price.

What can one conclude from this?

First that Roche has tried hard to get bevacizumab accepted by NICE at its current high list price. Besides colorectal cancer Roche initiated submissions to NICE for metastatic cancer of lung and breast, but its failure to submit evidence led to these appraisals being terminated in 2008. Its recent submission included a price discount nested within a patient access scheme, but the extent of the discount remains secret (commercial in confidence). Due to the inclusion of other elements in the patient access scheme it is not possible to readily deduce the price discount offered. However given that oxaliplatin is relatively low cost and that few if any patients will survive for 12 months, to reduce the cost per QALY from £105k to £30k must have involved a price reduction of around two thirds.    

Second, despite this the appraisal committee seems to have rejected bevacizumab as first line treatment largely due to the relatively small gain in progression free survival of 1.4 months, with no proof of any gain in overall survival.

Third the appraisal committee rejected the results of Roche’s cost effectiveness model on the grounds of its unduly optimistic assumptions. This was a reasoned qualitative rather than a quantative judgement, since this was a single technology assessment in which cost per QALY was based solely on the Roche model.   

Fourth, it remains to be seen if bevacizumab will qualify for the coalition government’s cancer fund. If so it could quickly exhaust the fund. NICE’s Technology Appraisal 117 put the number of new cases of metastatic colorectal cancer at 6-18k annually. At full cost of £2k per month, this would amount to £12 million to £36 million for each month that patients were treated. However if Roche continues to offer the patient access scheme with the substantial price discount, the cost could be much less.   

The bevacizumab story continues……

James Raftery is a health economist with several decades experience of the NHS. He is Professor of Health Technology Assessment at Southampton University. A keen ‘NICE-watcher’, he has provided economic input to technical assessment reports for NICE but has never been a member of any of its committees. The opinions expressed here are his personal views. Comments welcomed.