Despite 15 approvals for oncologics by the FDA in 2011, only one had the effect of drug on cancer-related symptoms included in the label.
This is the main observation from an article this month in the New England Journal of Medicine, which laments the lack of inclusion of patient-orientated outcome measures in oncologic clinical trials, and calls for a more structured approach to studying, and labelling, these measures. In this regard, oncologic labels are at odds with other therapeutic areas: a review examining approvals from 2006 to 2010 identified that 24% of all approvals had patient-reported outcome (PRO) claims, with the vast majority (86%) being for symptoms. However, it could be argued that 24% still represents a poor performance: the figure for oncology is very poor indeed.
In many respects, these observations are shocking, although not unexpected. It was not until 4 years ago that the FDA provided concrete guidance on PROs,3 and naturally, it may take some time for industry to implement, and of course in some cases develop and validate, instruments in mid- and late-phase clinical studies. In his NEJM article, Ethan Basch rightly highlights the need for both industry and the FDA to work harder on ensuring patient-centricity in label development. Industry has long complained that PRO development is lengthy and risky, and that guidance offered by the FDA was too vague to provide a decent surety of success. Further, the rigidity of views on primary and secondary endpoint needs for registration, and the impact this has on powering for PROs within studies, has hindered their inclusion, let alone development.
Whilst these are valid criticisms, IDEA Pharma believes that industry has not been either quick or forceful enough in pursuing PRO development. Basch is right in highlighting the need for early decisions on PRO inclusion – typically in phase II – to ensure parallel development of PRO and drug. These early judgement calls are essential if PROs are to become the norm, not the exception, in both oncology and other therapeutic areas. Too often such development has been classed as ‘too risky’, without considering the downstream risks of not developing. Basch is also right, we believe, in suggesting a more formal approach to engagement between industry, the regulator, and patients in developing and utilising PROs. It could be argued that even more stridency is needed, with regulators formalising their inclusion in registrational programmes, although whether they would be mandatory to achieve registration is more debatable. Put this way, it remains odd that amelioration of symptoms is not a necessity for consideration in studies designed to examine the effects of oncologics on terminal, metastatic disease.
If greater, and more formal, collaboration between industry and the FDA is to be encouraged on PROs, industry pro-activity would seem essential, not optional. In a world where patients, payers and regulators are increasingly defining efficacy, effectiveness and value on the basis of such measures, industry can ill afford to be in the passenger seat in driving their development and use. This recent NEJM article rightly highlights this important issue, and provides further impetus, should it be needed, for industry to consider PROs at the core, not the periphery, of their development objectives.