Single arm trials (SATs) are used to accelerate patient access to innovative medicines or when large randomised control trials (RCTs) are not feasible or ethical. This is particularly true for diseases with small populations or diseases with no reasonable comparator treatment, such as orphan diseases and diseases with severely limited treatment options in oncology. Marketing authorisation based on SAT data is a key strategy for accelerating patient access to innovative medicines in oncology. The objective of this research was to understand key payer concerns of the value of SAT data and to recommend strategies which may maximise future value of oncology products with evidence from SATs. An in-depth review was performed on assessments for six products (ofatumumab, vismodegib, ponatinib, ibrutinib, idelalisib and ceritinib) by nine agencies (NICE, SMC, AWMSG, HAS, G-BA, pCODR, PBAC, ZIN and TLV) from January 2011 to December 2015. An in-depth analysis of 31 assessments was performed. In most cases payers challenged the methods used for indirect comparison and highlighted the limited amount of evidence presented. There was no clear preference for the type of indirect comparison methodology preferred by payers. Overall response rate (ORR) was used as the primary endpoint in all six studies. Though payers considered overall survival to be the most relevant endpoint they appeared to also accept ORR. Immaturity of data within SAT trials was also highlighted as a key concern by payers due to a lack of validity of extrapolated data. However, despite these issues being raised, the majority (20/31) of submissions based on SAT data received positive recommendations. From our analysis it appears that payers are willing to accept a greater level of uncertainty in clinical evidence when there is a significant unmet need.

Upon disease progression or early termination of a trial, cancer patients in the control arm of clinical studies commonly switch to the experimental drug or receive other post-study treatments. In many oncology studies survival data is far from being complete with many patients still alive by the end of follow-up. There is no universally accepted definition of the cut-off for immature overall survival (OS).The objective of this study was to assess the methods agencies accept to adjust for treatment switching and for extrapolation of immature OS data. An in-depth review was performed on assessments for 5 products with immature overall survival data and where treatment switching occurred (pembrolizumab, dabrafenib, crizotinib, enzalutamide and vemurafenib) by 7 agencies (NICE, HAS, G-BA, IQWiG, pCODR, PBAC and TLV) from January 2011 to December 2015. 36 HTA appraisals were selected for in-depth analysis. In 9 appraisals methods to adjust for treatment switching were reported (NICE=5, TLV=2, PBAC=1, G-BA=1). NICE reviews and, if appropriately justified, will accept the implementation of switching adjustment. At the other end, a recent case example among G-BA submissions suggests that they do not accept the use of treatment switching methods. PBAC, TLV and pCODR have accepted adjustment methods in certain cases. However, no evidence of such adjustments were identified in the selected HAS reports. In the majority of cases NICE disagreed with selection of extrapolation method used by the manufacturer. Other agencies provide less details but in the examples found appeared critical. Payers highlighted the uncertainty of immature data results and that extrapolation results are questionable, invalid or overestimate potential benefits. NICE will review complex treatment switching methodologies while at other agencies it is not common practice as yet.