Predicting clinical efficacy of lung gene delivery in people with cystic fibrosis: which in vitro and in vivo models best predict outcomes in human lungs?

INTRODUCTION: Despite over three-decades of work, no inhalable nucleic acid delivery system for cystic fibrosis (CF) has advanced beyond Phase II due to a lack of congruence between positive outcomes from in vitro and preclinical work, and lung function improvements in people with CF.

AREAS COVERED: We review publications covering in vitro and in vivo animal models for CF research from databases including Web of Science and PubMed. Data were used to evaluate the performance of inhaled nucleic acid delivery systems for CF and examine why positive outcomes in these models fail to predict clinical responses.

EXPERT OPINION: The field should recalibrate expectations regarding the proportion of epithelial cell correction required for meaningful clinical improvement since the threshold for sustained improvement in lung function in humans may be substantially higher. No single model can predict clinical responses. Translation robustness depends on integrating multiple systems that collectively interrogate delivery, persistence, safety and functional restoration of lung function in clinically meaningful settings. Appropriate in vitro air-liquid interface models that accurately recapitulate the CF airways are important starting points, followed by delivery and functional evaluation in CF animal models that closely resemble human lung physiology and disease phenotype, including pigs and ferrets.