(510) Prevalence and Clinical Impact of Respiratory Viral Infections in the STOP2 Study of Cystic Fibrosis Pulmonary Exacerbations

Background: Respiratory symptoms are hallmarks of both pulmonary exacerbations (PEx) and viral infections. Although viral respiratory infection rates are similar in persons with cystic fibrosis (pwCF) and the general population, the association between PEx and viral infection is unclear. The Standardized Treatment of Pulmonary Exacerbations II (STOP2) study was a multi-center, randomized clinical trial among pwCF ≥18 years of age to evaluate response to different IV antibiotic durations for PEx. Sputum collected during study visits provided an opportunity to evaluate respiratory viral infection prevalence and clinical impact in the context of PEx.

Methods: Sputum, when available, was collected on Day -3-1 of therapy (Visit 1), Day 7-10 (Visit 2), and 2-weeks following treatment completion (Visit 3) and shipped to the University of Michigan. Samples were tested for 17 respiratory viruses using multiplex-PCR (FilmArray® Respiratory Panel 2, BioFire Diagnostics, LLC). Differences in lung function (percent predicted forced expiratory volume in one second; ppFEV₁) and symptom (Chronic Respiratory Infection Symptom Score; CRISS) changes across visits were compared as a function of viral detection by ANCOVA. Demographic variables were assessed by logistic or linear regression.

Results: 1323 samples from 690 STOP2 participants were analyzed. One or more viruses were detected in 393 samples (30%) from 269 participants (39%). Participants who were virus-positive at Visit 1 had significantly greater mean ppFEV₁ drop from the preceding 6-months (5.8±1.3 v. 3.6±0.9, p=0.03), worse CRISS scores (53.8±2.1 v. 51.1±1.0, p=0.03) and greater CFTR modulator use (47% vs. 34%, p< 0.0005). 68% of virus-positive participants had rhino/enterovirus (RV). Compared to other virus-positive participants, RV participants had fewer PEx in the prior year (53% v. 66%, p=0.03).

Adjusted mean ppFEV₁, CRISS and weight changes from Visit 1 to 3 were similar between virus negative and positive participants, except for greater mean ppFEV₁ change from Visit 2 to 3 in participants virus-positive at Visit 1 (0.12±0.9 v. –1.4±1.0, p=0.04). Participants co-infected with ≥2 viruses had higher mean ppFEV₁ change from Visit 1 to 3 (10.8±4.1, p=0.04) and from Visit 2 to 3 (1.2±2.3, p=0.04) and greater CRISS improvement from Visit 1 to 2 (–20.7±5.2, p=0.03) compared to those with one or no virus. No significant differences in mean ppFEV₁, CRISS or weight changes across visits were observed with RV infection compared to other viruses. Amongst participants with samples available for all three visits (n=176), concordance in viral detection across visits was observed in 69% of participants. Subjects with negative samples across all visits had lower CRISS score change from Visit 1 to 2 (– 13.3±2.4, p=0.02) and Visit 1 to 3 (–14.1±2.4, p=0.04) when compared to three positive samples (–18.7±4.4, 21.2±4.7) and one or two positive samples (–17.9±3.4, –21.4±3.7).

Conclusions: Over one-third of STOP2 participants tested positive for a respiratory virus. Viral detection, particularly co-infection with multiple viruses, was associated with modest differences in both baseline presentation and response to therapy. Further evaluation of long-term outcomes, particularly the impact of modulator therapy given the association with viral-positive participants, is required to better define prevalence and clinical impact of viral infections in pwCF.

Acknowledgements: This work was supported by CFF and BioFire Diagnostics with acknowledgement of TDN sites and CF patients.

References: