To our knowledge, this is the largest study to investigate patient reported features of cough hypersensitivity in ILD. Furthermore, we are the first to report a relationship between the number of cough hypersensitivity features and health status impact. Qualitative analysis lent weight to the findings that cough severely impacts the quality of life in patients with ILD.

Here, participants with ILD reported many features consistent with cough hypersensitivity (CH), thus suggesting that CH may be a phenotype of cough in ILD and play a mechanistic role [1]. Whilst few studies have investigated CH in ILD comprehensively, multiple cough triggers and heightened cough reflex sensitivity have been reported in IPF, CT-ILD, and sarcoidosis [7, 17]. Neurally mediated CH is hypothesised as a key mechanism in patients with CC that is refractory to treatment [1, 2]. CH in ILD remains enigmatic, and may involve both peripheral and central neural pathways. Inflammation and architectural distortion by fibrosis may sensitise peripheral airway chemoreceptors and mechanoreceptors [7]. However, a peripherally acting P2X2/3 receptor antagonist (Gefapixant) lacked efficacy in IPF associated cough despite positive RCTs in refractory chronic cough [18]. Conversely, recent multi-centre trials with opioids (morphine and nalbuphine) reported efficacy in IPF-associated cough [14, 15]. Taken together, the neuropathophysiology of chronic cough in ILD may be more centrally located. Indeed, recent functional neuroimaging and physiological studies have identified central mechanisms of cough in refractory chronic cough [19].

Several studies have detailed the impact of cough in ILD [9, 10], including an association between cough and disease severity [8, 11, 12]. We report that self-reported CH features were associated with the impact on quality of life. Whilst our data does not indicate causality, future development of CH-targeted therapy may have a direct impact on quality of life in ILD-associated chronic cough.

Our study has certain limitations. Selection bias may be present due to the recruitment of participants through a patient-driven charity, and the use of an online questionnaire could have excluded under-served cohorts who are not digitally capable or who do not understand English. The lists of cough hypersensitivity features and impacts were not exhaustive to avoid response fatigue; these could be developed for further study with patient input. Furthermore, the lack of validated patient reported outcome measures impedes interpretation and direct comparison with previous studies. Future studies could include use of standardised hypersensitivity questionnaires such as the Hull Cough Reflux Questionnaire (HARQ) or Cough Hypersensitivity Questionnaire (CHQ). The questionnaire did not assess the subjective severity of cough and, as it was anonymous, objective measures of cough or ILD severity were unavailable. Comorbidities associated with CC (e.g. gastroesophageal reflux disease [GORD], asthma, rhinitis, bronchiectasis) are common in ILD, and may contribute to cough hypersensitivity [11, 20]. This was not assessed in our study; however, empiric treatment of these conditions, such as PPI for acid-GORD, does not benefit cough [10, 11]. Finally, as with all questionnaires, the presence of recall bias cannot be excluded, such as with the self-reported underlying ILD diagnosis.