Changes in computed tomography derived muscle and pulmonary tissue in patients hospitalized for COVID-19: A 12-month follow-up study.

BACKGROUND

Prolonged muscle loss and persistent pulmonary radiological manifestations have been observed among previously hospitalized COVID-19 patients. This study aimed to investigate the patterns of change and the association between computed tomography (CT)-derived skeletal muscle cross-sectional area (CSA) and the percentage affected pulmonary parenchymal tissue in previously hospitalized COVID-19 patients during the first-year post-discharge.

METHODS

This single-centre retrospective cohort study included hospitalized COVID-19 patients between February and December 2020. Patients underwent routine chest CT scans at hospital admission, during outpatient clinic visits at three months, and if indicated, twelve months post-discharge. Pectoralis muscle CSA was demarcated manually at the level of the fourth thoracic vertebra and skeletal muscle at the first lumbar vertebra level (L1). A deep learning algorithm was applied to compute the percentage of total affected pulmonary parenchymal tissue.

RESULTS

Chest CT scans from 211 patients (mean age 66 ± 12 years, 61% male) were available for analysis between baseline and three months follow-up. For the twelve-month follow-up analyses, scans from 100 patients were available. From baseline to three months, a decrease in mean pectoralis (-1.3 ± 0.5 cm2 (-1.1 ± 1.4%); p= 0.011) and L1 muscle CSA (-3.2 ± 1.5 cm2 (-2.2 ± 1.5%); p< 0.001) was observed. Between three- and twelve months, mean pectoralis and L1 muscle CSA increased (2.3 ± 0.6 cm2 (7.2 ± 1.8%) and 10.2 ± 3.0 cm2 (12.3 ± 3.3%), respectively; p< 0.01), with levels returning to baseline at twelve months. Percentage of affected parenchymal tissue reduced between baseline-three months (-18.4 ± 1.0%; p< 0.001), baseline-twelve months (-21.6 ± 1.3%; p< 0.001), and three- and twelve months follow-up (-1.7 ± 0.3%; p< 0.001), with near-complete resolution by three and twelve months. No significant associations were found between decrements in pectoralis and L1 muscle CSA and improvements in percentage affected parenchymal tissue between baseline and three months. In addition, increases in pectoralis and L1 muscle CSA were not associated with improvements in percentage affected parenchymal tissue between baseline and twelve months, nor between three- and twelve months follow-up.

CONCLUSIONS

Decreases and increases in CT-derived skeletal muscle CSA occurred independently of improvements in pulmonary parenchymal abnormalities in previously hospitalized COVID-19 patients during the first-year post-discharge.