[18F]FDG PET/CT in Short-Term Complications of COVID-19: Metabolic Markers of Persistent Inflammation and Impaired Respiratory Function

SARS-CoV-2 virus infects organs other than the lung, such as mediastinal lymph nodes, spleen, and liver, but, to date, metabolic imaging studies obtained in short-term follow-ups of patients hospitalized with severe COVID-19 infection are rare. Our objective was to evaluate the usefulness of [18F]FDG-PET/CT in the short-term follow-up of patients admitted for COVID-19 pneumonia and to explore the association of the findings with clinical prognostic markers. The prospective study included 20 patients with COVID-19 pneumonia (November 2020–March 2021). Clinical and laboratory test findings were gathered at admission, 48–72 h post-admission, and 2–3 months post-discharge, when [18F]FDG-PET/CT and respiratory function tests were performed. Lung volumes, spirometry, lung diffusion capacity for carbon monoxide (DLCO), and respiratory muscle strength were measured. Volumetric [18F]FDG-PET/CT results were correlated with laboratory and respiratory parameters. Eleven [18F]FDG-PET/CT (55%) were positive, with hypermetabolic mediastinal lymphadenopathy in 90.9%. Mediastinal lesion’s SUVpeak was correlated with white cells’ count. Eleven (55%) patients had impaired respiratory function, including reduced DLCO (35%). SUVpeak was correlated with %predicted-DLCO. TLG was negatively correlated with %predicted-DLCO and TLC. In the short-term follow-up of patients hospitalized for COVID-19 pneumonia, [18F]FDG-PET/CT findings revealed significant detectable inflammation in lungs and mediastinal lymph nodes that correlated with pulmonary function impairment in more than half of the patients.


Introduction
There is growing interest in the diagnosis, prognosis, and optimal clinical management of the sequelae of acute COVID-19 infection.
In the acute phase of infection, the epidemiology, clinical characteristics, results of standard clinical laboratory tests, lung CT appearance, treatment strategies, and outcomes in patients with COVID-19 have been reported in previous studies [1]. Imaging techniques, especially high-resolution computed tomography (HRCT), have demonstrated a relevant diagnostic role [2], and multiple studies have been published on radiological findings in patients with COVID-19 pneumonia, especially during the acute phase and, more recently, over the short and medium terms [2,3].
The SARS-CoV-2 virus has been shown to infect organs other than the lung, such as the mediastinal lymph nodes, spleen, and liver, quantitative case studies in patients with COVID-19 are rare [3,4]. Such information can be obtained through the use of [ 18 F]-2-Fluoro-2-Deoxy-Glucose ([ 18 F]FDG) positron emission tomography/computed tomography (PET/CT), which is commonly used to assess inflammatory and infectious lung diseases [5].
The complementary functional information provided by [ 18 F]FDG-PET/CT, which has been shown to be useful for diagnosing inflammatory and infectious lung diseases, estimating their severity, monitoring their evolution, and evaluating therapeutic response [4,5], can help elucidate the pathophysiological mechanisms of COVID-19. The value of [ 18 F]FDG-PET/CT has been reported in patients with respiratory infections caused by other coronaviruses, such as MERS-CoV and SARS-CoV [6,7], as well as in patients with acute COVID-19 infection [4,8].
The [ 18 F]FDG-PET/CT studies of asymptomatic cancer patients described the incidental detection of interstitial pneumonia compatible with possible acute SARS-CoV-2 infection [7], and researchers have begun to examine the potential role of [ 18 F]FDG-PET/CT in its diagnosis and treatment [8]. As well as visual interpretation by an experienced specialist, [ 18 F]FDG-PET/CT also offers a semiquantitative approach to glycemic metabolism and, therefore, the intensity of inflammatory activity. Besides the standardized uptake value (SUV), recent studies in oncology have yielded additional parameters such as the metabolic tumor volume (MTV) and total lesion glycolysis (TLG) [9], which could be used to estimate inflammatory activity in lungs or extrapulmonary organs, especially lymph nodes. Studies of noncritical hospitalized patients have highlighted the possible relevance of lymph node hypermetabolism, quantified by the maximum SUV (SUVmax) in PET images, proposing that the highest SUVmax values for lesions and lymph nodes may indicate an increased severity of the infection and may predict a poor prognosis [3,4].
With this background, we hypothesized that [ 18 F]FDG-PET/CT could be useful to characterize pulmonary sequelae of COVID-19 infection. The objective of this study was to evaluate the usefulness of [ 18 F]FDG-PET/CT in the short-term follow-up of patients admitted for COVID-19 pneumonia and to explore the association of findings with clinical prognostic markers

Patients
This prospective, longitudinal, observational study enrolled consecutive COVID-19 patients at their follow-up visit 1-2 months after discharge from a third-level hospital between 27 November 2020 to 1 March 2021.
Study inclusion criteria were confirmation of COVID-19 in accordance with WHO guidelines [10] by a positive RT-PCR result for nasopharyngeal swabs, hospital admission between November 2020 to March 2021 (dates of "third wave" in Spain), and findings of ground-glass opacity or consolidation on chest HRCT scan or X-ray at admission. Exclusion criteria were age under 18 years, absence of microbiological confirmation of COVID-19 infection, history or presence of pulmonary fibrosis, active or uncontrolled COVID-19 infection at the time of the [ 18 F]FDG-PET/CT study, history or suspicion of oncological disease, pregnancy, and inability to sign informed consent.
The study was approved by the local Research Ethics Committee, and written, informed consent was obtained from all participants. Personal protective equipment was available for all staff, and COVID-19 infection prevention guidelines were always rigorously followed [11].

Clinical Information and Laboratory Test Results
For all patients, data were gathered from electronic medical records, including the results of clinical and laboratory tests at admission, at 48-72 h post-admission, and at the follow-up PET/CT examination. Analytical data included complete blood count, standard blood biochemistry, acute phase reactants, coagulation status [12], and neutrophil/lymphocyte ratios (NLRs). All patients underwent RT-PCR for nucleic acid testing of SARS-CoV-2.

Respiratory Function Tests
Respiratory function tests were performed at 2-3 months after hospital discharge. Spirometry results (in mL and % predicted) were obtained for forced vital capacity (FVC), forced expiratory volume in 1 s (FEV1), and FEV1/FVC ratio. Body plethysmography was used to measure the residual volume (RV, in mL and % predicted), and total lung capacity (TLC, in mL and % predicted). The diffusing capacity of the lungs for carbon monoxide (DLCO) and the CO transfer coefficient (KCO) were expressed in absolute numbers and as % predicted. The results of the 6-min walk test (TM6M) were expressed as distance (in m) and % oxygen saturation at start and finish. Specifically trained personnel carried out functional tests using MasterScreen Body equipment (Jaeger, Hoechberg, Germany), considering reference values for the Mediterranean population and acceptability criteria established by European and Spanish regulations [13,14].

PET/CT Data Acquisition
After two consecutive negative RT-PCR test results for SARS-CoV-2 nucleic acid, confirming that patients were no longer infected, patients underwent [ 18 F]FDG-PET/CT imaging (Siemens Biograph Vision 600 PET/CT, Siemens Healthcare, Erlangen, Germany), always performed within 2-3 months after discharge from hospital. The test protocol was based on international recommendations [15]. Patients were administered intravenously with the radiopharmaceutical (3.7-4.81 MBq/kg) at rest after fasting for at least 6 h with adequate hydration as long as their capillary blood glucose level was below 6.8 mmol/L. Image acquisition (whole body in 3D) started at 50-60 min post-injection with the acquisition of a topogram (50 mA, 120 kV), followed by helical CT without contrast (170 mA, 120 kV) and the acquisition of PET images with coverage from skull base to mid-thigh.

PET/CT Image Interpretation
The [ 18 F]FDG-PET/CT and chest CT images were independently analyzed by two nuclear medicine physicians (E.M.T.I. and M.G.M.) with a great deal of experience in the interpretation of cardiothoracic images, using syngo.via version VB40B software (Siemens Healthcare, Erlangen, Germany). They were blinded to the biological and clinical data of patients. Discrepancies in interpretations were resolved by consensus with a third expert nuclear medicine physician (A.R.F.).
The [ 18 F]PET/CT data were transferred to a computer workstation (syngo.via) for the co-registration of PET and CT images. Regions of interest (ROIs) were drawn on CT images of lungs around areas with evident loss of aeration and adjacent areas of normal appearance. ROIs were also drawn on CT images of mediastinal lymph nodes. The ROIs drawn on the CT images of each patient were transferred to the co-registered PET images and the amount of [ 18 F]FDG pathological uptake was calculated for each ROI, determining maximum, peak, and minimum SUVs, normalized by body weight (SUVmax, SUVpeak, and SUVmin, respectively) and lean body mass (SUL); metabolic tumoral volume (MTV; volume of pixels in the ROI with SUVmax >40%); and total lesion glycolysis (TLG; MTV multiplied by SUVmean).

Chest CT and X-ray Image Interpretation
Upon their diagnosis, all patients underwent chest X-ray in posterior-anterior and lateral projections, reported by specialist radiologists according to current recommenda-  [16,17]. They characterized the density (alveolar, ground glass, or mixed), distribution (central, peripheral, or diffuse), location (unilateral or bilateral), and extent (unilobar or multilobar).

Statistical Analysis
All measurements for each participant were independently conducted by two nuclear medicine physicians, considering the mean value in statistical analyses. Absolute numbers and percentages were calculated for categorical variables and means with standard deviation (SD) for continuous variables. For comparisons of quantitative data between the positive and negative PET groups, the Student's t-test was applied when the distribution was normal and the Mann-Whitney U test when it was not. Associations with categorical variables were evaluated by constructing contingency tables, applying the chi-square test for individual comparisons and Fisher's exact test for multiple comparisons. Volumetric [ 18 F]FDG-PET/CT results were correlated with laboratory test results and respiratory function parameters by using Spearman's rank correlation coefficient. IBM SPSS version 15.0 (IBM Corp, Armonk, NY, USA) and R software were used for statistical analyses. A p ≤ 0.05 was considered significant in all tests.

Results
The study included 20 patients (60% males) with a mean age of 55.85 ± 9.28 years admitted for pneumonia and/or respiratory failure between 27 November 2020 and 1 March 2021 (during the "third wave" of COVID-19 in Spain). The mean hospital stay was 16.70 ± 11.99 days. Table 1 summarizes the baseline characteristics of the patients. Selective inhibitors of pro-inflammatory cytokines 6 (30) Continuous variables are presented as means ± standard deviation (SD) and categorical variables as frequencies (percentages). ARDS: acute respiratory distress syndrome; AST: aspartate aminotransferase; ALT: alanine transaminase; NLR: neutrophil/lymphocyte ratio; PCT: procalcitonin; NT-proBNP: N terminal pro-B-type natriuretic peptide; LDH: lactate dehydrogenase; ICU: intensive care unit.

Discussion
In this study, [ 18 F]FDG-PET/CT was used to measure the metabolism of lungs and other organs in the short-medium follow-up of patients admitted to hospital for pneumonia or respiratory failure due to COVID-19 infection. Despite testing negative for the infection in two successive RT-PCR tests of nasopharyngeal swabs, more than half of the patients showed increased metabolic activity (i.e., persistent inflammation) on [ 18 F]FDG-PET/CT images in lung tissue of normal appearance and in mediastinal lymph nodes. To our best knowledge, [ 18 F]FDG-PET/CT has not previously been used to detect residual inflammatory processes after COVID-19 infection. These findings contribute evidence on

Discussion
In this study, [ 18 F]FDG-PET/CT was used to measure the metabolism of lungs and other organs in the short-medium follow-up of patients admitted to hospital for pneumonia or respiratory failure due to COVID-19 infection. Despite testing negative for the infection in two successive RT-PCR tests of nasopharyngeal swabs, more than half of the patients showed increased metabolic activity (i.e., persistent inflammation) on [ 18 F]FDG-PET/CT images in lung tissue of normal appearance and in mediastinal lymph nodes. To our best knowledge, [ 18 F]FDG-PET/CT has not previously been used to detect residual inflammatory processes after COVID-19 infection. These findings contribute evidence on the pathophysiological processes in patients who survive hospital admission for COVID-19 pneumonia.
The [ 18 F]FDG-PET/CT has been employed in patients with influenza A, aspiration pneumonia, and organized pneumonia to assess the extent and severity of the disease, to follow its course, and to evaluate the response to therapy [5,18,19].  [22], and Scarlattei et al. reported that this metabolic activity remained high many weeks after the disappearance of symptoms and a negative RT-PCR test result [23]. The present results are in line with the above findings and contribute novel data on increased metabolic activity in lung tissue of normal appearance and in mediastinal lymph nodes of normal size. In this context, Xu et al. described lymphocyte-dominated interstitial mononuclear inflammatory infiltrates in both lungs of a patient with COVID-19 and reported that substantial inflammation may persist in the lungs after the disappearance of the infection [24]. The elevated [ 18 F]FDG uptake would reflect increased glycolytic activity due to infiltration and inflammation of the lung, even in normally aerated areas that show no morphological alterations on CT images, demonstrating the greater capacity of [ 18 F]FDG-PET/CT to detect inflamed lung areas in comparison to CT alone [8,22], which may persist long after the disappearance of COVID-19 infection. The possible duration of the post-COVID-19 inflammatory response in lungs and extrapulmonary sites has yet to be established and warrants further research. At 2-3 months post-discharge, patients with elevated chest [ 18 F]FDG uptake were older and characterized by a higher Charlson index, more frequent fatigue and respiratory distress, and lower hemoglobin and lymphocyte counts in comparison to those with normal [ 18 F]FDG uptake. The SUVpeak of the target lesion and pulmonary TLG were significantly correlated with acute phase reactants and white blood cell counts at admission, during the hospital stay, and at 2-3 months post-discharge. Although there is a lack of similar studies in severely ill COVID-19 survivors for comparison with these results, they are consistent with previous findings on risk factors for more severe infection, including old age, underlying comorbidities [12,25], and similar changes in white blood cell counts, lymphocyte counts, procalcitonin and CRP levels, and NLR [26,27]. The [ 18 F]FDG-PET/CT findings were correlated with the NLR in all studied phases of COVID-19 disease. The persistence over time of increased [ 18 F]FDG uptake intensity may reflect a more severe acute phase of the disease.
The lung appears to be the most frequently involved organ in COVID-19, with reports of diffuse alveolar epithelium destruction, capillary damage/bleeding, hyaline membrane formation, alveolar septal fibrous proliferation, and/or pulmonary consolidation, among others [12,24]. Long-term follow-up studies of survivors of other coronavirus infections found that respiratory function limitations frequently last for months or even years, including impaired DLCO (in 15.5-43.6% of patients) and decreased TLC (5.2-10.9%) [28][29][30]. Various authors have addressed short-and medium-term respiratory function outcomes in survivors of COVID-19 infection, usually at hospital discharge [31,32]. In a study at 2-3 months post-discharge of 55 COVID-19 survivors who had not required mechanical ventilation, Zhao et al. described residual pulmonary function in 14 patients (25.45%), mainly impaired DLCO (in 13.6%) [33]. In a study at 6 weeks post-discharge of 124 COVID-19 survivors, van den Borst et al. [34] described an improvement in radiological images for almost all patients (99%) but observed residual lung parenchymal alterations in 91% of the patients and reduced lung diffusion capacity in 42%. Likewise, in their study at 3 months post-discharge of 76 healthcare workers who recovered from COVID-19, Liang et al. reported normal FEV1, FVC, FEV1/FVC, TLC, and DLCO values (>80% predicted) in 82% of the patients but the persistence of mild pulmonary function abnormalities in 42% [35]. The proportion of the present patients with impaired pulmonary function at 2-3 months was in line with previous findings on the short-to medium-term effects of COVID-19 infection [33,34].
The most frequent respiratory sequela of COVID-19 was DLCO alteration, as reported in previous studies, which may indicate the presence of pulmonary fibrosis [12,24]. DLCO and other respiratory function parameters were negatively correlated with the lung [ 18 F]FDG uptake as quantified by TLG. Although only a small proportion of the present patients had severe airway dysfunction, the results suggest that COVID-19 produces diffuse pulmonary epithelial damage and mild congestion of the airway mediated by lymphocytedominated interstitial inflammatory infiltrates. No published data appear to be available on the association between respiratory function test results and pulmonary TLG. The majority of the present patients showed no lung lesions on CT scans at 2-3 months after discharge; however, pulmonary function was impaired in more than half of the patients with a normal lung CT scan. Hence, pulmonary function and [ 18 F]FDG-PET/CT testing is more sensitive than CT alone for identifying candidates for pulmonary rehabilitation after SARS-CoV-2 pneumonia [22].
High [ 18 F]FDG uptake may be related to increased anaerobic glycolysis caused by a cascade of reactions involving inflammatory cells [7,36]. In this way, the uptake of [ 18 F]FDG by lung lesions and lymph nodes observed in this study may be due to nonspecific immune or inflammatory activation, similar to the high [ 18 F]FDG uptake observed in lung lesions caused by the Middle East respiratory syndrome, pandemic H1N1 influenza virus, and organized pneumonia [18,19,37].
The [ 18 F]FDG-PET/CT offers a complementary approach to other imaging modalities by providing metabolic information. Although not currently recommended for the diagnosis of COVID-19 in the acute phase [8], it can yield relevant information for the diagnosis of short-and medium-term complications, including the chronic damage to the lungs and extrapulmonary sites that can follow acute infection [6,22]. However, radiologists and nuclear physicians need to develop a thorough understanding of the cellular mechanisms that underlie the pathophysiology of COVID-19 in the clinical settings of lung and extrapulmonary malignancies and inflammatory diseases in order to avoid misinterpretation of [ 18 F]FDG-PET/CT images [31].
Besides the small sample size, the main limitation of this study was the absence of a control group, hampering the possibility to detect causal relationships between the findings and COVID-19 infection. The epidemiological environment in which this study was carried out determined strict, restrictive conditions for access to hospital centers in our center and population. Evidently, the performance of [ 18 F]PET/CT in healthy collaborating patients was obviously not authorized. In addition, no test results were available for the baseline respiratory function of patients before COVID-19, although the presence of chronic lung disease was an exclusion criterion. Further research is required to fully elucidate the impact of COVID-19 on pulmonary function. In this regard, the present results cannot be extrapolated to patients with chronic lung disease. Another study limitation was the absence of a follow-up period to explore the long-term clinical relevance of the respiratory function impairment. Finally, biopsy specimens were not available for the studied organs. Nevertheless, the present findings contribute to laying the foundations for future studies with larger series on the potential role of [ 18 F]FDG-PET/CT in evaluating the sequelae of COVID-19 infection. These should have prolonged follow-up periods to explore the possible relationship between initial lung inflammation and long-term sequelae such as residual lung fibrosis and respiratory failure.

Conclusions
In conclusion, at 2-3 months after the acute phase of SARS-CoV-2 infection, almost half of the patients evidenced an impairment of pulmonary function that was correlated with [ 18 F]FDG-PET/CT findings. In addition, the increased metabolic activity observed in the lung and mediastinal lymph node was associated with clinical and laboratory markers of disease severity. The [ 18 F]FDG-PET/CT is useful to obtain novel information on the pathogenesis of COVID-19 and on the diagnostic and evaluation of short-and medium-term sequelae, contributing to their management.  Informed Consent Statement: Informed consent was obtained from all subjects involved in the study. Written informed consent has been obtained from the patient(s) to publish this paper.