Objective:

To assess if CREXONT^® (IPX203) improves absorption by evaluating whether patients with poor IR CD-LD response show greater clinical improvement after switching to CREXONT.

Background:

CREXONT is the first extended-release (ER) CD-LD formulation designed to optimize absorption in the proximal small intestine. The new formulation intended primarily to extend dose duration, could also mitigate levodopa absorption challenges in some Parkinson’s disease patients. Such patients may be initiated on doses too high by the current conversion algorithm, increasing risk of dopaminergic adverse events. This analysis describes the range of responses to formulation change observed in the RISE-PD trial and highlights those requiring dose reductions when converting to CREXONT.

Design/Methods:

During the open-label conversion phase of RISE-PD, all participants were converted from IR CD-LD to CREXONT using a predefined conversion algorithm. We modeled the effects of formulation change on dose-response (δ), defined as a patient’s daily “Good On” time(h) per their total daily levodopa dose (TDD-LD). Gaussian Mixture Models (GMMs) were used to isolate putative populations based on δ values.

Results:

After conversion to CREXONT, GMMs isolated two main response types: A main population ([MP], 93% of subjects) of lower δ values (n=470, mean=0.88h), and a smaller population (7%) of “super-responders” (SR) whose d values indicated increased dose-response benefits associated with conversion (n=35, mean=5.08h). Compared to MP, while on IR CD-LD, SR were on higher TDD (1003mg [SR] vs 855mg [MP], p=0.05) and had lower dose-responses (0.99h/100mg [SR] vs 1.45h/100mg [MP], p=0.0007). On average, after using the predefined IR CD-LD to CREXONT conversion algorithm, SR needed reductions from their CREXONT starting dose (-213mg of TDD), while MP needed an increase (+144mg of TDD, p=1.1x10^-6).

Conclusions:

Patients with short response durations on high doses of IR CD-LD are most likely to have enhanced response to CREXONT and may require lower doses than suggested by the dosing algorithm.