The application was for Regorafenib ( Stivarga ) to remain in the Cancer Drugs Fund ( CDF ) as treatment for patients with non-resectable or metastatic gastrointestinal stromal tumours ( GISTs ) who were either Imatinib-refractory or Imatinib-intolerant but also who had progressed on Sunitinib for treatment of their GISTs.
The CDF panel was aware that this indication for Regorafenib has not been appraised by NICE ( National Institute for Health and Care Excellence ). It was also aware that its licensed indication in patients with GISTs was in those patients who had progressed on or are intolerant to prior treatment with Imatinib and Sunitinib.
There was one randomised trial ( GRID study ) submitted by the manufacturer ( Bayer ) as the main evidence base related to this indication. This randomised 199 patients with progressive non-resectable or metastatic GISTs previously treated with Imatinib ( either refractory to it or intolerant of it ) and progressing on Sunitinib to receive Regorafenib plus best supportive care ( BSC ) or placebo plus BSC.
The primary end point was independently-assessed progression free survival ( PFS ).
Patients receiving placebo who experienced disease progression were offered cross over to open label Regorafenib ( 85% did so ).
Patients receiving Regorafenib who had progressive disease and for whom in the investigator’s opinion, treatment with Regorafenib was providing clinical benefit, were offered the option to continue with open label Regorafenib ( at time of analysis, 31% had done so ).
Independently-assessed median progression-free survival ( PFS ) was significantly prolonged with Regorafenib ( 4.8 vs 0.9 months, difference [ delta ] 3.9 months, hazard ratio [ HR ] 0.27, 95% confidence interval ( CI ) 0.19-0.39, p less than 0.0001 ), respectively.
Investigator-assessed progression-free survival was 7.4 vs 1.7 months ( HR=0.22, 95% CI 0.14-0.35, p less than 0.0001 ).
Follow-up was short in the primary analysis as the CDF Panel observed that few patients were at risk at or beyond 8 months of follow-up.
Median overall survival ( OS ) durations in the primary analysis could not be estimated due to the degree of censored data ( approximately 75% censoring rate ).
Overall survival was not significantly different between the two arms ( HR=0.77, 95% CI 0.42-1.41, p=0.20 ).
Toxicity was assessed by the National Cancer Institute Common Terminology Criteria for Adverse Events and was increased in the Regorafenib arm.
Grade 3 and 4 summated toxicities were observed in 60% vs 10%, respectively. In terms of specific grade 3 and 4 toxicities, hand-foot skin reactions occurred in 20% vs 0% and hypertension in 24% vs 3%, respectively.
As regards all grade toxicities, diarrhoea was observed in 40% vs 5%, fatigue 39% vs 27%, oral mucositis in 38% vs 8%, alopecia 24% vs 2%, anorexia 21% vs 8% and skin rashes 18% vs 3%, respectively.
Dose modifications occurred in 72 vs 26%, respectively.
There was collection of quality of life ( QOL ) data in this study ( EQ5D ) and utilities were similar when comparing both arms for both the pre- and post-progression periods.
Since progression-free survival was greater with Regorafenib, the CDF Panel considered stable QOL to have been preserved for longer in the Regorafenib arm.
The median duration of treatment in the Regorafenib arm was 22.9 weeks ( 5.3 months ). The Panel noted that the duration of treatment might be considerably longer than this if patients continued on Regorafenib after disease progression as had occurred in the GRID study.
The CDF Panel noted that the manufacturer also submitted two single arm phase 2 studies of Regorafenib in GIST patients previously treated with Imatinib and Sunitinib.
The investigator-assessed progression-free survival durations were about 10 months in both studies and the median treatment durations were 8 and 11 months.
Overall survival data were immature.
The CDF Panel observed that the European Public Assessment Report ( EPAR ) contained an updated overall survival analysis which was based on 139 events.
The median overall survival was 17.4 months in both arms of the GRID study ( HR=0.85, 95% CI 0.60-1.20, p=0.18 ).
The manufacturer submitted an analysis of overall survival in the GRID study which used two different methods to adjust for cross over, the Rank Preserving Structural Failure Time ( RPSFT ) model and the Iterative Parameter Estimation ( IPE ) algorithm. The results were academic in confidence.
The EPAR documented that use of the RPSFT model resulted in an overall survival analysis with a HR of 0.54 ( 95% CI 0.29-1.01, p value from a one-sided logrank test was 0.02 ) and that of the IPE algorithm an overall survival HR of 0.57 ( 95% CI 0.30-1.06, p value from a one-sided log rank test was 0.03 ).
The Panel noted that such cross over methodologies created uncertainties as different methods led to different results and there were other methods too than just the RPSFT and IPE. Such analyses in view of their complexity and controversy and the choice of one method rather than another always required a minimum of peer review examination and publication before assessment by the CDF Panel.
The EPAR stated that the median overall survival was 17.4 in both arms of the GRID study.
The manufacturer indicated that the median overall surivival durations for GIST patients failing Imatinib and Sunitinib were 8-9 months in other studies.
The CDF Panel noted that the EPAR contained a table of duration of median overall survival durations in 7 published studies in patients with GISTS after treatment with Imatinib and Sunitinib, these studies not investigating the use of Regorafenib. The range of median overall survival was between 8 and 19 months.
The CDF Panel considered these median overall survival durations in these other studies but was very wary of the results of indirect comparisons between potentially heterogeneous populations.
The only peer-reviewed and published data it had to assess survival was the overall survival data reported in the GRID trial publication: follow-up was immature as there were few patients at risk beyond 8 months and no difference in overall survival was seen.
There was unpublished overall survival data as documented in the EPAR and this showed no difference in overall survival.
The CDF Panel recognised the bias introduced by the crossover design of the GRID study but concluded that it could not be sure of the degree of any survival advantage with the use of Regorafenib.
The CDF panel assessed whether Regorafenib met the CDF criteria for the definition of unmet need. It observed that there were other active drugs in the treatment of GISTs ( Imatinib and Sunitinib ) and thus Regorafenib was not the only active drug in GISTs.
It knew that Regorafenib was the only drug licensed for use in GISTs after Imatinib and Suntinib but the CDF SOP only allowed the recognition of unmet need when a drug was the only proven therapy for that particular cancer.
The CDF Panel therefore concluded that Regorafenib in GISTs did not meet this CDF definition of unmet need. The Panel also did not consider that Regorafenib represented a step change in the treatment of GISTs as it did not have any proven survival advantage associated with its use.
The CDF Panel’s scores for Regorafenib in GISTs previously failing Imatinib and Sunitinib were as follows: 3 for progression-free survival ( difference in PFS of 3.9 months ); 0 for overall survival ( no proven OS benefit ); 1 for quality of life ( maintained QOL ); minus 1 for toxicity; 0 for unmet need.
This resulted in a total of 3B.
The CDF Panel considered the clinical benefits of Regorafenib after two previous tyrosine kinase inhibitors in non-resectable or metastatic gastrointestinal stromal tumours to offer sufficient clinical benefit to potentially remain included in the CDF.
It noted that the patient numbers treated with Regorafenib for GISTs were small yet Regorafenib had a European approved indication in the treatment of bowel cancer, a much more common cancer ( CDF Panel appreciated that Regorafenib in bowel cancer is not approved by NICE nor by the CDF ). ( Xagena )
Source: Cancer Drugs Fund ( UK ), 2015