Lenvatinib (1) – Lenvima®

• Clinical trials
  • SELECT is a randomised, placebo-controlled, double-blind Phase III trial designed to assess the efficacy of lenvatinib (24 mg orally) compared with placebo, in terms of progression-free survival (PFS) in 392 patients with radioiodine-refractory differentiated thyroid carcinoma and radiologically proven disease progression within the last 12 months.

adult patients with progressive, locally advanced or metastatic differentiated (papillary/follicular/Hürthle cell) thyroid carcinoma (DTC) that has not responded to radioiodine therapy (RAI)

• mortality
  • The secondary endpoint of the SELECT trial, overall survival, was defined as the time from randomisation to death from any cause.
  • Neither at the time of the primary analysis of overall survival as planned in the study protocol (1st data cut-off: 15 November 2013) nor at the time of the post-hoc analysis requested by the EMA (2nd data cut-off: 15 June 2014) was there a statistically significant difference between the lenvatinib and placebo arms.
  • Patients in the placebo arm had the option of switching to the lenvatinib arm (crossover) following disease progression. By the first data cut-off, approximately 83 per cent of patients in the placebo arm had already switched to the lenvatinib arm; by the second data cut-off, this figure had risen to approximately 88 per cent.
  • To increase the reliability of the results, adjusted analyses for the overall survival endpoint were carried out using the Rank Preserving Structural Failure Time Model (RPSFTM). The adjusted analysis also showed no statistically significant advantage of lenvatinib at the first data cut-off. At the second data cut-off, a statistically significant effect in favour of lenvatinib was observed (hazard ratio: 0.53; 95% CI: [0.34; 0.82], p = 0.0051).
  • With regard to an adjustment method for a crossover trial, it cannot be concluded that there is currently a scientific consensus on which specific method might be more suitable in individual cases. The application of the method, depending on whether the basic assumptions are met, must be described and justified in the dossier. Sufficient evidence has not been provided to demonstrate that the basic assumption of a treatment effect independent of the start of treatment in both the active and placebo arms has been met.
  • The EMA has already provided an indication of the limitations of the RPSFTM in this case in the European Public Assessment Report (EPAR) for lenvatinib and has assessed the analyses as merely supportive.
• Morbidity – Progression-free survival (PFS)
  • The primary endpoint of the SELECT trial, progression-free survival, was defined as the time from randomisation to the occurrence of documented disease progression or death, whichever occurred first.
  • A statistically significant difference was observed in favour of lenvatinib compared with placebo (hazard ratio: 0.21; 95% CI: [0.16; 0.28]; p < 0.0001) with a median progression-free survival of 18.3 months in the lenvatinib arm and 3.6 months in the placebo arm (absolute difference of 14.7 months).
  • The PFS endpoint is a composite endpoint comprising endpoints from the mortality and morbidity categories. The ‘mortality’ component of the endpoint is already assessed as a standalone endpoint via the secondary endpoint of overall survival. Furthermore, the ‘disease progression’ component of morbidity in the PFS assessment was not symptom-based, but was determined exclusively by means of imaging procedures. Taking the above aspects into account, there are differing views within the G-BA regarding the clinical relevance of the PFS endpoint to patients. This does not affect the overall conclusion regarding the extent of the additional benefit.
• quality of life
  • No data on quality of life are available. Consequently, no conclusion regarding the extent of the additional benefit can be drawn with respect to quality of life.
• Side effects
  • Adverse events (AEs) were documented until they resolved or up to 30 days after the patient’s last visit.
  • The analysis based on the data cut-off date of 15 November 2013 was carried out using the safety population, comprising all patients who had received at least one dose of the study medication (N = 392). To address the imbalance regarding the differing treatment durations in the two treatment groups (median treatment duration of 13.8 months on lenvatinib versus approximately 3.9 months on placebo), the dossier presents not onlytime-adjusted (time-to-event) analyses were presented in the dossier (data cut-off date: 15 March 2014).
  • For the time-adjusted analyses, the SELECT trial showed a statistically significant disadvantage for lenvatinib compared with placebo across all safety endpoints (AE, AE of CTCAE grades 3 and 4, serious AE and AE leading to therapy discontinuation). The hazard ratios range from 2.63 [95% CI: 2.09; 3.32] for AEs to 5.41 [95% CI: 3.84; 7.64] for AEs of CTCAE Grade 3 and 4.
  • In addition to the results of the primary analysis presented above, the non-time-adjusted results for the safety endpoints are shown. A significant adverse effect of lenvatinib is also evident here (for both data cuts), although the differing follow-up periods in the two treatment groups must be taken into account when interpreting these results.
  • As the SELECT trial compared lenvatinib with placebo, a higher risk of harm is to be expected. However, these results regarding side effects were decisive in determining the need to establish a dose/toxicity management plan to allow for daily dose adjustments. The EMA therefore recommended in the EPAR, depending on the degree of toxicity, the use of an algorithm for dose reductions, dose interruptions and, where necessary, the discontinuation of lenvatinib. In addition, the EMA called for an additional study to be conducted on the initial dose.
• Conclusion
  • Taking the available results as a whole, the G-BA has reached the following assessment of the extent of the additional benefit: there is an additional benefit, but it is non-quantifiable because the available scientific data do not currently permit a quantifiable statement on the extent of the additional benefit for patient-relevant endpoints.

Courtesy translation only, please refer to the German original.