Patients with COVID-19 have an increased risk for microvascular lung thrombosis. In order to evaluate the type and prevalence of perfusion defects, we performed a longitudinal analysis of combined perfusion single-photon emission and low-dose computed tomography (Q-SPECT/CT scan) in patients with COVID-19 pneumonia.

Consecutive patients with severe COVID-19 (B.1.1.7 variant SARS-CoV-2) and respiratory insufficiency underwent chest Q-SPECT/CT during hospitalization, and 3 months after discharge. At follow-up (FU), Q-SPECT/CT were analyzed and compared with pulmonary function tests (PFT), blood analysis (CRP, D-dimers, ferritin), modified Medical Research Council (mMRC) dyspnea scale, and high-resolution CT scans (HRCT). Patients with one or more segmental perfusion defects outside the area of inflammation (PDOI) were treated with anticoagulation until FU.

At baseline, PDOI were found in 50 of 105 patients (47.6 %). At FU, Q-SPECT/CT scans had improved significantly (p < 0.001) and PDOI were recorded in 14 of 77 (18.2 %) patients. There was a significant correlation between mMRC score and the number of segmental perfusion defects (r = 0.511, p < 0.001), and a weaker correlation with DLCO (r = −0.333, p = 0.002) and KCO (r = −0.373, p = 0.001) at FU. Neither corticosteroid therapy nor HRCT results showed an influence on Q-SPECT/CT changes (p = 0.94, p = 0.74). CRP, D-Dimers and ferritin improved but did not show any association with the FU Q-SPECT/CT results (p = 0.08).

Segmental mismatched perfusion defects are common in severe COVID-19 and are correlated with the degree of dyspnea. Longitudinal analyses of Q-SPECT/CT scans in severe COVID-19 may help understand possible mechanisms of long COVID and prolonged dyspnea.