"No!" says NICE, to Sprycel, Tasigna

The UK cost-effectiveness watchdog NICE today delivered a resounding "no" to the use on the National Health Service of Bristol's dasatinib (Sprycel) and Novartis' nilotinib (Tasigna) in chronic myeloid leukemia patients intolerant to imatinib (Glivec).

"The evidence available to support [the clinical effectiveness] of dasatinib and nilotinib was very poor," declared Professor Peter Littlejohns, clinical and public health director at NICE. "The drugs' cost is also very high," he added, in a press release announcing the latest draft guidance.

Sprycel costs about £30,477 per year, and nilotinib about £31,711, according to appraisal documents on NICE's website. And the drugs are taken for several years, with no evidence-based 'cut-off' point currently in use.

It doesn't even look as if the drugs came close, in other words. And Bristol and Novartis can't even consider one of the loopholes now available to companies, the end-of-life guidance issued in late-2008, which permits a somewhat higher cost-per-QALY (quality-adjusted life year) than usual for drugs that extend life in niche yet terminal diseases. (This, you will recall, is what allowed Celgene's multiple myeloma drug Revlimid to slip past the agency.)

The available evidence on the drugs' extension of life--typically required to be of at least three months--"is too weak", declares the NICE PR in yet another blow to the products' manufacturers. That the drugs, both second-generation tyrosine kinase inhibitors, offer such an extension, documents declare, "is plausible, but definitely not proven."

But at the end of this rather damning announcement came an olive branch. "It would be heartening to hear that pharmaceutical company manufacturers are prepared to share some of the very high cost of these drugs with the NHS," suggested Littlejohns.

Now if that isn't a call for a cost-share (or should we say 'patient-access') scheme, then I don't know what is. Recall that such schemes have allowed NICE to green-light a good handful of expensive drugs that likely would not have otherwise made the cut--including most recently UCB's RA drug certolizumab (Cimzia). (Interestingly, although Celgene also put forward such a plan for Revlimid, this wasn't what tipped the decision in its favor.)

So we understand NICE's call for companies to make an effort on the cost-share front--indeed, the agency's CEO Andrew Dillon has told us clearly that he'd prefer if manufacturers simply submitted such schemes up front rather than waiting for a rejection in order to fish one out.

But is Littlejohns implying that a cost-share proposal would simply eliminate all the problems that the appraisal committee identified in the submission, around trial data and design? These seemed considerable: no studies submitted assessed either drug against relevant comparator; trials were 'heterogenous in terms of design, population, implementation and analysis'.

We put this question to NICE. Their reply:

Although there is some evidence to suggest that dasatinib and nilotinib could be considered clinically effective in cases of chronic myeloid leukaemia (CML), the quality of that evidence was extremely poor. This, coupled with the very high cost of the drugs, meant that the independent appraisal committee could not recommend them as an appropriate use of NHS resources.

During the public consultation on the draft recommendations manufacturers will have the opportunity to propose a patient access scheme, to make it easier for the NHS to afford expensive new treatments. We would be happy to look at such a scheme.

The answer is still not entirely clear (to me anyway; and I'm pushing for further clarification). But it sure looks as if patient access schemes will trump poor data.

If that's true, we're not sure that will do anyone any good--the NHS (paying, if a reduced price, for drugs that aren't effective), companies (forced to submit access schemes above all else), or patients (potentially receiving an ineffective drug and, as a group, perhaps not getting something else as a result).

We hope, then, that we're wrong.

image by flikrer greenchartreuse used under a creative commons license