Y.S. Ho

Predictors of a positive Cancer Drug fund decision

OBJECTIVE Since 2011, the United Kingdom has set aside £200 million per year through the Cancer Drug Fund (CDF) to pay for oncology treatments not reviewed or approved by NICE. The CDF reviews pivotal trial data and scores drugs on disease-free/progression-free survival (PFS), overall survival (OS), quality of life (QoL), safety, unmet need, and strength of evidence. The score determines if the drug will be included on the CDF priority list. Due to the rising costs of cancer treatment and overspending in the CDF budget, the CDF added cost as a parameter in November 2014. This necessitated a re-review of all drugs on the priority list and prompted stakeholders to question whether cost would become the most influential parameter in the CDF decisionmaking process. This analysis attempts to determine if the importance of OS and PFS has diminished now that cost has been added to the evaluation. As manufacturers develop novel life-saving therapies, is cost more important than OS and PFS in the CDF evaluation of drugs?

Negative Reimbursement Consequences From Trial Design Choices

PRM218 Negative ReiMbuRseMeNt CoNsequeNCes FRoM tRial DesigN ChoiCes Jaksa A., Westbrook L., Daniel K., Ho Y.S. Context Matters, Inc., New York, NY, USA Objectives: Health Technology Assessments (HTA) use clinical trial data to determine comparative efficacy and cost-effectiveness; thus study design plays a roll in market access. The objective is to determine how often reimbursement decisions cite trial design defects. MethOds: We analyzed 1,702 HTAs from CADTH, G-BA, HAS, NICE, PBAC, and SMC. We examined the clinical assessment rationale for the decision and the reimbursement decisions. An explicit trial design defect was defined as a clinical assessment of “inappropriate comparator” or “inappropriate patient population.” Clinical assessments of lower, uncertain, or unknown efficacy or a clinical determination of insufficient or lack of evidence were defined as potential trial design defects. Results: Reviews that cited trial defects resulted in significantly more negative reimbursement decisions (6.6%) than positive reimbursement decisions (0.4%; p< .001). This pattern held true for each individual agency examined. G-BA was the agency most likely to cite an explicit trial deficit (39%), while HAS was the least likely (0.6%). In addition, significantly more reviews that cited a potential trial defect resulted in negative reimbursement decisions (44%) than positive reimbursement decisions (5.5%; p< .001). This also held true for each agency. Again, G-BA was most likely to note a potential trial defect in reviews (46%) while SMC was least likely (6.4%). Among disease conditions with more than 10 reviews, explicitly cited trial defects were mostly frequently seen in Cystic Fibrosis and Parkinson’s Disease (15% and 13%, respectively). Potential trial defects were most frequently cited in Atrial Fibrillation and Depression reviews (42% and 39% respectively). cOnclusiOns: Explicit and potential trial design issues have negative consequences for reimbursement outcomes. Negative decisions are more likely than positive decisions to cite trial design issues. G-BA is more inclined than other agencies to cite these trial design issues when issuing their reimbursement decisions. Manufacturers should consider market-access outcomes when designing clinical trials.

Number of Submissions Needed to Reach a Positive Reimbursement Decision from SMC, CADTH, and PBAC

Positive reimbursement decisions are frequently achieved only after multiple submissions. Multiple submissions can delay patient access to necessary therapies and be costly for the manufacturer. This study analyzes the number of submissions needed to gain a positive decision. For those drugs that were not recommended on the first submission but received a positive decision upon resubmission, this analysis determines the lag time between the first submission and a positive decision.

The Relationship Between Submitted Line of Therapy and Reimbursement Decisions in Solid State Tumor Cancer Drugs

In Germany, IQWIG has proposed to use correlations between surrogate endpoints and patient relevant outcomes to determine value of cancer drugs. Due to increased cost pressure on payers, such models are likely to inspire novel reimbursement schemes for other branded products. ConClusions: Cancer drug reimbursement models are setting new benchmark for payers to manage access and control costs. These models have significant implications for other expensive branded products.