CDER-Approved NDA for EVRYSDI™ (risdiplam)
On August 7, 2020, the FDA approved Genentech’s EVRYSDI™ (risdiplam) “for the treatment of spinal muscular atrophy (SMA) in patients 2 months of age and older.” Of note: Genentech was granted Fast Track, Orphan Drug, and Priority Review designations.
The results of part two of the phase 2/3 randomized controlled trial (RCT), SUNFISH (NCT02908685; n=231), provided substantial evidence of the effectiveness of risdiplam in pediatric and adult patients with Type 2 and Type 3 SMA. The results of part one of the phase 2/3 open-label, single-arm, externally controlled trial, FIREFISH (NCT02913482; n=62), provided substantial evidence of the effectiveness of risdiplam in infants with Type 1 SMA.
Intent of the RWE studies
The applicant used multiple natural history studies to serve as an external control for the experimental arm of FIREFISH. The primary objective of these studies was to demonstrate that patients with Type 1 SMA are never able to sit without support. The secondary objective was to show the rates of ventilator-free survival. The applicant also used natural history studies as context for CHOP-INTEND baseline scores.
Protocols for RWE generation
In pre-submission meetings, the FDA repeatedly stated its preference for a placebo-controlled trial, citing the following concerns with a single-arm trial in patients with Type 1 SMA: “many possible sources of bias with such a design (and small sample size), such as lack of a concurrent control, differences in study populations, change in standard of care, etc.” The FDA eventually agreed to the single-arm trial if the threshold for a finding of efficacy was that five out of 40 patients would be able to sit without support and that the sponsor “provide data that supports the assertion that untreated children essentially never would be expected to reach this milestone as defined by the protocol.”
For the primary endpoint of sitting without support, the applicant submitted five studies: Finkel et al. 2014a; Finkel et al. 2015; Faravelli et al. 2015; De Sanctis et al. 2016; and Kolb et al. 2017. The primary objective of these studies was to demonstrate that patients with Type 1 SMA are never able to sit without support.
For the secondary endpoints of survival and ventilation-free survival, the applicant submitted an additional five studies: Finkel et al. 2014, Kolb et al. 2017, Oskoui et al. 2007, Rudnik-Schoeneborn et al. 2009, and one other study. These studies showed that the rate of ventilation-free survival in Type I SMA at 18 months of age ranged from nine to 51 percent.
Outcome of the RWE submissions
When discussing the primary endpoint of the FIREFISH experimental arm compared to the natural history studies, the FDA’s biometrics review concluded “the evidence from Firefish, though impressive on face compared to the reported natural history, is not well controlled.” However, the clinical reviewer came to a different conclusion, stating it “considers the external natural history control as sufficient to describe the Firefish study as ‘well-controlled’ because no Type 1 SMA patients would be expected to achieve sitting without support in the natural history of the disease.” The FDA also noted that the threshold of five patients sitting without support was previously agreed upon by the FDA and sponsor before the study initiation, and stated that such an endpoint was “well-defined with a low potential for bias.”
For the secondary endpoint of ventilation-free survival, 94 percent of the higher-dose cohort in FIREFISH was alive at month 12. The FDA concluded that “the Firefish Part 1 study result indicates reduced mortality and permanent ventilation in the risdiplam group compared to the natural history of Type 1 SMA.”
The FIREFISH study showed a four point increase in CHOP-INTEND scores in 88 percent of the higher dose cohort. An improvement of that magnitude was never seen in Darras or Kolb. 53 percent of the participants in this cohort also achieved head control, compared to 39 percent achieving partial head control in De Sanctis et al. 2016.