Presentation (CTI)

OBJECTIVES: To estimate the annual decline of ppFEV₁ in patients receiving CFTR modulator treatments, a novel approach was developed due to the lack of robust long-term comparative data and potential COVID-19 pandemic confounding effects on existing single-arm trial data.
METHODS: Two studies were identified through bibliographic searching of a relevant systematic literature review (SLR) that used UK CF registry data of people treated with ivacaftor (IVA) between 2008 and 2016. Using data reported on the causal treatment effect of IVA on the long-term rate of ppFEV₁ decline, a relative reduction in ppFEV₁ decline was estimated for people treated with IVA compared to ECM (38%, 95% CI: -12% to 87%). This was assumed to be a lower bound for the long-term effectiveness of elexacafor/tezacaftor/ivacaftor (ELX/TEZ/IVA). A network meta-analysis (NMA) was used to calculate the ratio of the ELX/TEZ/IVA to IVA acute treatment effect. This was applied to the IVA relative reduction, to estimate the treatment effect of ELX/TEZ/IVA. For tezacaftor/ivacaftor (TEZ/IVA), the estimated relative reduction for ELX/TEZ/IVA was scaled by the ratio of the TEZ/IVA to ELX/TEZ/IVA acute treatment effect in the NMA. Based on an SLR, there was no robust evidence of a long-term reduction in rate of decline of ppFEV₁ for lumacaftor/ivacaftor (LUM/IVA) compared to ECM.
RESULTS: The relative rate of decline for the CFTR modulator therapies was estimated as: ELX/TEZ/IVA (61%, 95% CI: 11.7% to 110%), TEZ/IVA (17.2%, 95% CI: -32.0% to 66.4%) and LUM/IVA (0%, 95% CI: -49.2% to 49.2%). These estimates rely on the assumption of a similar relative difference in acute and long-term effectiveness between IVA, for which robust long-term data were available, and ELX/TEZ/IVA, TEZ/IVA and LUM/IVA.
CONCLUSIONS: Modelling long-term ppFEV₁ changes following CFTR modulators remains challenging. Future modelling may prefer to use the directly observed rate/shape of ppFEV₁, or as a calibration, as longer follow-up data emerges.