Pcn310 - survival in cancer clinical trials: a comparison of ema approvals and nice recommendations.

Davis et al. (BMJ 2017;359:j4530) recently carried out a review of European Medicines Agency (EMA) approvals for cancer drugs which concluded that in only 23 of 68 (34%) indications was there statistically significant prolongation of overall survival (OS). This study reviewed NICE Technology Appraisal (TA) guidance related to the drugs and indications discussed in Davis et al. The available TA guidance for each indication analysed in the Davis et al. paper was reviewed. The date, status of guidance (published/suspended) and final recommendations were extracted. NICE appraisal data were available for 63 of the 68 original indications. Of the 23 drugs that showed an OS benefit in the Davis et al. paper, NICE made a positive recommendation for 11/23 (48%); a negative recommendation for 8/23 (35%); 2/23 (9%) did not proceed (due to non-submission/scope discontinuation) and data for 2/23 (9%) were not available. All of the negative decisions were because the treatment was not cost-effective. Of the 45 drugs that did not show an OS benefit in the Davis et al analysis, NICE made a positive recommendation for 16/45 (36%); a negative recommendation for 15/45 (33%); 11/45 (24%) did not proceed (due to non-submission/scope discontinuation) and data for 3/45 (7%) were not available. Reasons for a positive recommendation when no OS benefit was shown included an improvement in PFS; commercial arrangements; restriction to a selected subgroup or assessing a different or indirect comparator. Although OS is an important consideration in cancer drug appraisals, NICE committees took additional factors into account when making their recommendations. Additionally, regulatory reviews only consider trial data which may include a comparator which is not relevant to clinical practice in England. In addition to quality of life considerations, analyses of progression-free survival and indirect treatment comparisons are important factors to consider in decision-making.