Forced vital capacity in pulmonary exacerbations of cystic fibrosis

Forced expiratory volume in one second (FEV₁) is routinely used to assess pulmonary exacerbations (PEx) in cystic fibrosis (CF) but underrepresents air trapping and small airways dysfunction. Forced vital capacity (FVC) may better capture these processes, yet its behavior, clinical relevance, and utility as a trial endpoint remain uncertain.

We performed a secondary analysis of STOP2, a multicenter randomized trial of antimicrobial duration in adults with CF pulmonary exacerbations (PEx). Changes in percent-predicted FVC (ppFVC) and FEV₁ (ppFEV₁) were assessed relative to pre-PEx baseline. Clinical characteristics associated with large ppFVC decline were evaluated using adjusted analyses. Propensity score matching was used to compare recovery, residual deficit, and symptom response between groups with large versus modest ppFVC decline. Associations with time-to-next-exacerbation (TTNE) were assessed using survival analyses. Variability, responsiveness, and projected clinical trial sample sizes were compared between ppFVC- and ppFEV₁-based endpoints.

Among 760 participants, ppFVC responses were heterogeneous. Larger absolute declines were associated with higher baseline FVC and elevated C-reactive protein (CRP), while larger relative declines were associated with CRP alone. Participants with large initial ppFVC deficits showed greater absolute recovery but larger residual deficits, shorter TTNE, and higher future exacerbation risk. A one-standard-deviation worse residual ppFEV₁ deficit was associated with increased re-exacerbation risk, whereas residual ppFVC deficit was weaker and not independently prognostic. Absolute recovery in either measure was not associated with TTNE. Compared with ppFEV₁, ppFVC demonstrated similar or greater variability, resulting in larger estimated sample sizes for clinical trials.

Distinct FVC response patterns occur during CF PEx but do not provide independent prognostic information beyond FEV₁. Use of ppFVC as a primary endpoint would require larger sample sizes without clear advantage.