COVID-19 pulmonary phenotypes and longitudinal patterns in the first wave of the pandemic

Background

The long-term evolution of COVID-19 in patients hospitalized during the pandemic's first wave remains largely unexplored. This study aimed to identify COVID-19 pulmonary phenotypes and their longitudinal patterns over a 12-month follow-up.

Methods

COVID-19 patients discharged from Pisa University Hospital (Italy) between March–September 2020, were evaluated at T3, T12, and T24 months post-discharge. Assessments included spirometry, lung volumes, DLCO, and chest CT for those with persistent pneumonia signs (PS). Latent transition analysis (LTA) identified COVID-19 phenotypes and longitudinal patterns based on PS and lung function (PFTs). Risk factors for these patterns were evaluated using multinomial logistic regression.

Results

Of 307 discharged patients, 175, 136, and 33 were followed-up at T3, T12, and T24, respectively. At T12, 21.6 % had impaired DLCO, 4.4 % a restrictive ventilatory pattern, and 31,6 % still had PS, persisting until T24. LTA identified three cross-sectional phenotypes at both T3 and T12 (no PS with normal PFTs; PS with normal PFTs; PS with impaired PFTs), and four longitudinal patterns from T3 to T12: persistence of no PS with normal PFTs (47.9 %); resolution of both PS and PFTs (15.4 %); persistent PS (36.7 %), either with (11 %) or without (25.7 %) impaired PFTs. The last two patterns correlated significantly with longer hospitalization, more comorbidities, and severe COVID-19.

Conclusions

In our cohort of COVID-19 patients hospitalized during the pandemic's first wave, we observed distinct pulmonary phenotypes and longitudinal recovery patterns. More comorbidities and severe acute disease correlated with worse progression up to 24 months, suggesting long-term monitoring for such patients.