Everolimus, lutetium-177 DOTATATE and sunitinib for treating unresectable or metastatic neuroendocrine tumours with disease progression

Background: Neuroendocrine tumours (NETs) are a group of heterogeneous cancers which develop in cells in the diffuse neuroendocrine system. Patients considered here have unresectable or metastatic neuroendocrine tumours with disease progression. The interventions are everolimus, 177Lu-DOTATATE and sunitinib. The following NETs locations are considered separately: pancreatic, gastrointestinal (GI) & lung and GI (midgut only). Here, we present a systematic review of clinical effectiveness and cost-effectiveness studies, and our de novo economic analysis. We also critique submissions from the pharmaceutical companies. Methods: We systematically reviewed the effectiveness literature on advanced, progressive NETs. We wrote a survival partition cohort-based economic evaluation in Microsoft Excel. This comprised three health states: progression-free survival, progressed disease, and death. The perspective was that of the UK NHS & Personal Social Services. Results: Three RCTs, RADIANT-3 (pancreatic NETs: everolimus vs. BSC), A6181111 (pancreatic NETs: sunitinib vs. BSC) and RADIANT-4 (GI and lung NETs: everolimus vs. BSC), met the inclusion criteria in our clinical effectiveness systematic review. The risk of bias was low. Whilst the NETTER-1 RCT, of 177Lu-DOTATATE + Octreotide 30mg vs Octreotide 60mg was excluded from our review, we nonetheless present the results of this trial, as it informs our estimate of cost-effectiveness of 177Lu-DOTATATE. The pancreatic NETs trials consistently found that the interventions improved PFS and OS versus BSC. Our indirect comparison in pancreatic NETs found no significant difference in PFS between everolimus and sunitinib. Estimates of OS gain were confounded due to high rates of treatment switching from BSC to sunitinib or everolimus. The companies used a statistical technique to adjust for this switching. After adjustment, our indirect comparison suggests a lower, but non-significant, hazard of death with sunitinib compared to everolimus. In GI and lung NETs, everolimus significantly improved PFS compared to BSC, and a non-significant trend in improved OS compared to BSC. Adverse events were more commonly reported following treatment with interventions compared to placebo. Novartis compared the cost-effectiveness of everolimus vs. sunitinib in pancreatic NETs and everolimus vs. BSC in GI and Lung NETs. AAA Ltd compared the cost-effectiveness of 177Lu-DOTATATE plus octreotide 30mg vs. sunitinib vs. everolimus for pancreatic NETs and 177Lu-DOTATATE plus octreotide 30 mg vs. everolimus for GI NETs. Pfizer did not submit an economic evaluation of sunitinib. In our base case for pancreatic NETs, assuming list prices, we estimate incremental cost-effectiveness ratios (ICERs) for everolimus vs. BSC of £45,493 per QALY and sunitinib vs. BSC of £20,717 per QALY. These ICERs increase substantially without the adjustment for treatment switching. For GI and lung NETs, we estimate the ICER for everolimus vs. BSC of £44,557 per QALY. For GI (midgut), the ICERs were: everolimus vs. BSC £199,233 per QALY and in a scenario analysis, 177Lu-DOTATATE vs. BSC £62,158 per QALY. We judge that no treatment meets NICE’s End of Life criteria, although we cannot rule out that sunitinib in A6181111 does. Conclusions: Given NICE’s current stated range for the cost-effectiveness threshold of £20,000 to £30,000 per QALY, based on list prices, only sunitinib might be considered good value for money in England and Wales.