Incidental Findings Suggestive of COVID-19 Pneumonia in Oncologic Patients Undergoing (18)F-FDG PET/CT Studies: Association Between Metabolic and Structural Lung Changes

Although the novel coronavirus disease 2019 (COVID-19) can present as nonspecific clinical forms, subclinical cases represent an important route of transmission and a significant source of mortality, mainly in high-risk subpopulations such as cancer patients. A deeper knowledge of the metabolic shift in cells infected with severe acute respiratory syndrome coronavirus 2 could provide new insights about its pathogenic and host response and help to diagnose pulmonary involvement. We explored the potential added diagnostic value of (18)F-FDG PET/CT scans in asymptomatic cancer patients with suspected COVID-19 pneumonia by investigating the association between metabolic and structural changes in the lung parenchyma. Methods: (18)F-FDG PET/CT studies acquired between February 19 and May 29, 2020, were reviewed to identify those cancer patients with incidental findings suggestive of COVID-19 pneumonia. PET studies were interpreted through qualitative (visual) and semiquantitative (measurement of SUV(max)) analysis evaluating lung findings. Several characteristic signs of COVID-19 pneumonia on CT were described as COVID-19 Reporting and Data System (CO-RADS) categories (1–6). After comparing the SUV(max) of pulmonary infiltrates among different CO-RADS categories, we explored the best potential cutoffs for pulmonary SUV(max) against CO-RADS categories as the gold standard result to eliminate the possibility that the diagnosis of COVID-19 pneumonia exists. Results: On multimodal PET/CT imaging, CT signs classified as CO-RADS category 5 or 6 were found in 16 of 41 (39%) oncologic patients. SUV(max) was higher in patients with categories 5 and 6 than in patients with category 4 (6.17 ± 0.82 vs. 3.78 ± 0.50, P = 0.04) or categories 2 and 3 (3.59 ± 0.41, P = 0.01). A specificity of 93.8% (95% CI, 71.7%–99.7%) and an accuracy of 92.9% were obtained when combining a CO-RADS score of 5 or 6 with an SUV(max) of 2.45 in pulmonary infiltrates. Conclusion: In asymptomatic cancer patients, the metabolic activity in lung infiltrates is closely associated with several combined tomographic changes characteristic of COVID-19 pneumonia. Multimodal (18)F-FDG PET/CT imaging could provide additional information during early diagnosis in selected predisposed patients during the pandemic. The prognostic implications of simultaneous radiologic and molecular findings in cancer patients and other subpopulations at high risk for COVID-19 pneumonia deserve further evaluation in prospective research.