Pharmaceutical companies face difficult decisions during drug development in designing trial designs to optimise the likelihood of market access. Expected Net Benefit of Sampling (ENBS) is a useful method to design clinical trials that efficiently reduce cost-effectiveness (CE) model uncertainty. However, most ENBS studies evaluate the value of trials from a societal perspective. Our aim was to adapt the traditional framework for ENBS to be more compatible with drug development trials from the pharmaceutical perspective.
Method:
We modify the traditional framework for conducting ENBS from a societal perspective and make it relevant to the pharmaceutical industry. We use commercial net benefit to value trials and assume that the price of the drug is variable and conditional on the trial outcomes. Value-based Pricing (VBP) is a pricing strategy where drug prices are generated from a CE model according to the cost-per-QALY threshold. We use this criterion to determine price conditional on trial data using Bayesian Updating of CE model parameters. We assume that there is a threshold price below which the company would not market the new intervention and would receive zero profits. The expected profit forecasts are weighed against the cost of the trial to estimate the expected commercial net benefit for each trial design.
A case study in which the sample size (n=100, n=500, n=1500) and trial duration (d=1, d=2, d=3) are varied in a Phase III trial for Systemic Lupus Erythematosus (SLE). For each trial design we sampled 10,000 trial outcomes, and updated CE model parameters using a Bayesian Approximation formula. VBP was estimated for each simulated trial using a SLE CE model. Expected profit of the trial is estimated by averaging across all trial samples. The expected commercial net benefit is calculated as the expected profits minus the costs of the trial.
Result:
A clinical trial with shorter follow-up (d=1) and larger sample size (n=1500) generated greatest expected commercial net benefit. Increasing the duration of follow-up had a modest impact on profit forceasts.
Conclusion:
ENBS can be adapted to value clinical trials in the pharmaceutical industry to optimise the expected commercial net benefit. However, the analyses can be very time-consuming to run for complex CE models.