Conference Coverage
Tildrakizumab for psoriasis scores high marks in phase III
VIENNA – Tildrakizumab, a novel biologic for moderate to severe psoriasis, showed favorable efficacy and safety outcomes in two pivotal phase III...
Controversy over how to report long-term outcomes
A hot topic among clinical trialists in dermatology concerns how to report study results. The traditional method in studies funded by pharmaceutical companies is known as the “last observation carried forward” analysis. It casts the study drug results in the most favorable possible light because, when a subject drops out of a trial for any reason, their last measured value for response to treatment is carried forward as though the patient completed the study. Thus, psoriasis patients who drop out because they couldn’t tolerate a therapy or developed a serious side effect dictating discontinuation will be scored on the basis of their last PASI response, creating a bias in favor of active treatment.
A more conservative analytic method is known as the “nonresponder imputation” analysis. By this method, a patient who drops out of a trial is automatically categorized as a treatment failure, even if the reason was that the patient moved and could no longer make visits to the study center.
The prespecified guselkumab analysis presented by Dr. Blauvelt involved nonresponder imputation through year 1 and imputation based on the reason for discontinuation in the second year. In contrast, the 2-year tildrakizumab analysis presented by Dr. Papp used the far more common last observation carried forward method.
To help the audience appreciate the importance of looking at the analytic methods used in a studies and help them understand the clinical significance of the results, Dr. Blauvelt provided a reanalysis of the 2-year guselkumab data using the last observation carried forward method. Across the board, the numbers became more favorable. For example, the PASI 75 rate of 95.7% using the prespecified nonresponder imputation analysis crept up to 96.8% under the last observation carried forward method; for comparison, the PASI 75 rates were 81%-84% in the tildrakizumab analysis.
“If you wanted to compare apples to apples with some other drugs, you would use these numbers – the as-observed analysis numbers used by most other companies with other drugs. If you wanted to determine what the true-life numbers are, they’d probably be something between the nonresponder imputation and as-observed numbers,” said to Dr. Blauvelt.
Dr. Papp was untroubled by the use of the last observation carried forward method in the particular case of the tildrakizumab long-term extension study.
“There is reason to believe the as-observed analysis doesn’t affect the integrity of the data because the dropout rate is extraordinarily low,” he said.
The guselkumab analysis was sponsored by Janssen Pharmaceutica; the tildrakizumab analysis was sponsored by Merck and by Sun Pharma. Dr. Blauvelt and Dr. Papp were paid investigators in both studies and serve as scientific advisers to virtually all companies invested in the psoriasis therapy developmental pipeline.