Here's a fun paper for anyone interested in today's news story on Lucentis.
First, the background story. There is a very expensive drug called ranibizumab (trade name "Lucentis") used to treat an eye condition called wet AMD. It costs a huge amount of money, for a tiny dose. This expensive drug can be substituted with a very similar cancer drug called bevacizumab (Avastin). Because the amounts required for an eye injection are so low, substituting a tiny dose of bevacizumab makes the treatment much cheaper: like about £60 a pop, because you can get so many tiny eye injections out of one dose of a cancer drug, even if it's really expensive. The drug company don't like this cost-saving practice, and are trying to get a judicial review saying that this use of bevacizumab is unacceptable in the NHS.
There are arguments about whether one treatment is better than the other. But how much would it cost for the NHS to do what the drug company want, and pay for proper Lucentis every time? This paper, published last year, does exactly that calculation (amusingly I see my dad is last author):
They looked at annual rates of injection for wet AMD, which have been rising rapidly, and in the discussion section they calculate the likely impact for the NHS:
"Our data demonstrate that the annual number of intravitreal
injection episodes has increased over 150-fold in England from
203 in 1989 to 30 458 in 2008. It is presumed that the majority
relate to the licensed form of treatment, ranibizumab. If all
injections in 2008 were indeed ranibizumab, the cost to the NHS
of the drug alone for that year would have been over £3 million.
Using recent estimates that over 25 000 cases of neovascular
AMD might be eligible for anti-VEGF therapy each year in the
UK, that the combined cost of a hospital assessment and
injection is £1750 and that a typical treatment course involves
14 injections over 2 years, the cost to the NHS of ranibizumab
treatment for these patients would be over £600 million per year
At £600 million a year - six hundred million pounds a year - you can now see why Novartis are fighting so hard to force the NHS to use Lucentis.
On the broader arguments made by the drug company, I would point out two further things. Firstly, the claim that they need lots of money to develop new drugs is undermined by the fact that they spend twice as much on marketing and advertising as they do on research and development. Secondly, this sudden principled stance against "off license prescribing" is undermined by the fact that the industry as a whole promotes such off license use widely when it suits them (eg a
$400m fine for Novartis for doing so, via the comments below). Lots more on both of these points in my book on big pharma, out soon!
Anyway, here's the abstract of the paper, and a link to the full text:
Trends over time and geographical variation in rates of intravitreal injections in England.
Br J Ophthalmol. 2012 Mar;96(3):413-8. Epub 2011 Aug 28.
Keenan TD, Wotton CJ, Goldacre MJ.
AIMS:
The recent emergence of antivascular endothelial growth factor (anti-VEGF) drugs has led to increased numbers of patients undergoing intravitreal injection for age-related macular degeneration (AMD). The aims of this study were to report on trends over time and geographical variation in intravitreal injection rates in England, and consider the implications for publicly funded health services of introducing new and expensive treatments.
METHODS:
Hospital episode statistics were analysed for annual treatment rates of intravitreal injection between the NHS financial years of 1989/1990 and 2008/1999.
RESULTS:
Annual injection rates increased from 0.4 episodes (95% CI 0.37 to 0.49) per 100,000 population in 1989/1990 to 10.7 (10.4-11.0) in 2006/2007. Rates then rose exponentially to 59.5 (58.8-60.2) in 2008/2009, with increasing use of multiple injections per person. The largest growth in injection rates was found in older people, and for AMD. Numbers of treatment episodes increased from 203 (1989/1990) to 30,458 (2008/2009). Geographical analysis showed a very wide variation across local authority areas in injection rates, from 0.9 (0.2-2.2) to 42.2 (38.9-45.7) people per 100,000 population in 2005-2008.
CONCLUSION:
Rates of intravitreal injection increased exponentially from 2006/2007. This followed the US Food and Drug Association licensing of ranibizumab for the treatment of neovascular AMD (2006), and its recommendation by National Institute for Health and Clinical Excellence (2008). This study demonstrates some of the major issues which arise with the emergence of expensive new treatments, including speed and cost of adoption, geographical variation in access, and implications for licensing, commissioning and health financing in an ageing society.
I like stats. I'm a doctor. I write about dodgy scientific claims in my spare time. 
