DLco simulations for qualification of DLco devices globally for inhaled levodopa clinical trials

Abstract

Previously published studies describe a 25% DLco device failure rate on the first DLco simulator test then decreased failure rates over time (Jensen et al, Eur Respir J 2009; 33: 1-7).

The study aim evaluates current DLco device failure rates at initial assessment and throughout 3 concurrent phase 3 clinical trials of CVT-301, inhaled levadopa. A pre-qualification process along with on-site training may decrease initial and subsequent DLco device failures during clinical trials.

Methods: Pre-qualification forms were sent to sites to collect equipment type and purchase date. Study requirements stipulated 5 DLco device manufacturers at less than 8 years of age. 114 DLco devices were evaluated for accuracy prior to subject testing using the DLco simulator (Hans Rudolph, Kansas City, MO) during on-site training. Patient testing was simulated with 3 inspired volumes (3L, 4L, and 5L) and 2 gas ranges (Mid and High). 108 DLco devices were evaluated with simulations every 12-weeks from October 2014 to April 2018. Results were sent to TechEd for centralized calculation and review. Tests were acceptable if the DLco values were within ±10% of simulator expected target values. Only DLco devices within control limits were used for subject testing.

Results: Initial DLco device failures occurred in 13(11.4%) of the 114 devices. The failure sources were inaccurate medical gas (46%), CO analyzers (46%) and software (8%). Post-training, 117 of 919 DLco simulations failed (12.7%), 56(51.9%) devices had zero failures, and 31(28.7%) had greater than one failure.

Conclusion: Pre-qualification selection and on-site training led to fewer initial DLco device failures than previously reported in the literature.