Article Text
PDF
Abstract
Background Cough challenges are utilised in clinical research to help determine efficacy of treatments for cough. Recently adenosine triphosphate (ATP) has gained interest as a potential tussive agent. Previous research suggests such challenges are reproducible in healthy individuals, but little is known about their reproducibility in chronic cough patients, particularly the use of ATP as a tussive agent. This study aims to clarify if ATP is a reproducible tussive agent in chronic cough patients.
Method Data was collected on subjects undergoing cough challenges in the clinical trials unit. Subjects performed tussive challenges with four agents (capsaicin, ATP, citric acid and fog) in a randomly allocated order (visit 1); C2 and C5 were noted. This test was then repeated a week later (visit 2). Intra-patient variability was analysed using the Bland-Altman method for each tussive agent, presented as mean difference (95% limits of agreement). Inter-patient variability was analysed using paired t-tests. Pearson’s correlation coefficient (r) between ATP and other agents was calculated.
Results 26 subjects were recruited; 21 with chronic cough. Average age of 57.4 ± 12.2, mean BMI of 26.8 ± 5.7, with an 85% female predominance. ATP showed a strong correlation with citric acid (r = 0.76, p < 0.001) and capsaicin (r = 0.66, p < 0.001). Bland-Altman analysis at C5 showed a 95% limit of agreement to be more than a two log dose difference except for fog: citric acid −0.07 (−1.7 to 1.5), capsaicin −0.1 (−1.5 to 1.2), ATP −0.2 (−2.5 to 2.1) (Figure 1), fog −0.01 (−0.3 to 0.3). Comparing visit 1 and visit 2 for each tussive agent showed no significant difference (p-value of 0.58, 0.80, 0.90, and 0.80 for ATP, fog, capsaicin and citric acid, respectively).
Conclusion ATP shows a strong correlation with other agents currently being utilised in cough challenges. This suggests direct cell surface activation similar to other agents. Comparing challenges in chronic cough patients shows that whilst inter-patient variability is low, intra-patient variability is high. Therefore, in contrast to healthy subjects, cough challenges are unlikely to be a useful measure of determining individual improvement in chronic cough patients.
Bland-Altman analysis of the 95% limit of agreement for ATP at C5