Intramuscular Adipose Tissue At Level Th12 Is Associated With Survival In Covid-19

Body composition refers to the amount and distribution of skeletal muscle, adipose tissue and bone in the human body. Sarcopenia, as an example of abnormal body composition, is defined as a significant loss of skeletal muscle mass (muscle wasting) and muscle strength and infiltration of muscle by fat and connective tissue.1, 2 Body composition has been studied using a single computed tomography (CT) slice, which is considered the reference standard for quantitative body composition studies.3, 4 Abnormal body composition, and in particular sarcopenia, has been associated with survival in patients with cancer5 or an increased cardiometabolic risk.6 Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has caused a pandemic of coronavirus disease 19 (COVID-19). In COVID-19 patients, age is associated with lower survival,7 and a higher body mass index (BMI) is associated with invasive ventilation.8 Regarding abnormal body composition, an increase of visceral fat has been associated with ICU admission of COVID-19 patients.9 Effects of abnormal body composition on survival are currently unknown. We hypothesized that abnormal body composition, as measured on standard chest CT images, is associated with lower survival in COVID-19 patients. The aim of our study was to examine the association between body composition measures and survival in COVID-19 patients. We prospectively included consecutive patients admitted to the Amsterdam University Medical Centers, location Academic Medical Center, between March 2020 and June 2020. Inclusion criteria were a polymerase chain reaction-confirmed COVID-19 infection, age ≥18 years, need for hospitalization, availability of a CT scan of the chest and availability of clinical outcome data. This study has been conducted in accordance with the ethical principles set out in the declaration of Helsinki and all participants provided written informed consent, if applicable. Ethics approval was obtained from the Amsterdam UMC Biobank Committee (202_065#A202029).10 We collected the following demographic and clinical variables: age, sex, length, weight and BMI and survival status at Day 21. CT images were obtained using standard multi-slice CT scanners and a clinical non-contrast enhanced low-dose CT chest protocol. The first scan, at the day of admission, was used if more scans were acquired. Body composition is typically estimated at the level of vertebra L3/L4.11 As these levels were not available on standard CT chest examinations, measurements were performed at level Th12, in the cross-sectional slice that showed both transverse processes. Images were anonymized and stored in 512 × 512 matrix, 16-bit Digital Imaging and Communications in Medicine (DICOM) format. Muscle segmentation was performed using manual outlining and semi-automated thresholding using the Horos DICOM viewer (version 3.3.6, www.horosproject.org) by an experienced operator. In all examinations, skeletal muscle and subcutaneous adipose tissue (SAT) were segmented manually, carefully excluding bone, cartilage and intra-abdominal/thoracic tissues (Figure 1A–D). For segmentation, previously used thresholds in Hounsfield units (HU) were applied: −29 to +150 HU for muscle and −190 to −30 HU for SAT.12 Cross-sectional area (CSA; cm2) and mean radiodensity (HU) of muscle and SAT were calculated. As an indicator for fatty muscle degeneration, the CSA (cm2) of intramuscular adipose tissue (IMAT) was determined by measuring fat pixels (below −29 HU) within the muscle contours. In order to correct for body size, the skeletal muscle index (SMI) was computed (cm2/m2) by dividing cross-sectional muscle area (cm2) by squared patient length (m). The same correction for body size was applied to CSA of IMAT and SAT, creating IMAT and SAT indices (cm2/m2). Descriptive statistics were reported as percentages, means and standard deviations or, when appropriate, median and interquartile range (IQR). We performed logarithmic transformation on non-normal distributed data and used Fisher exact test, independent t-test and Mann–Whitney U test, where appropriate. All tests assumed a two-tailed probability and a P-value of <0.05 indicated statistically significant difference. Kaplan–Meier survival analysis was performed with groups based on median split and the log-rank test. We used a multivariate Cox proportional hazards model with, age, BMI, muscle density and IMAT index (Enter method) as explanatory variables. We included 215 of the eligible 278 COVID-19 patients (see Supporting Information, Data S2). Eighty-six patients (40.0%) were female; mean age at hospital admission was 61.1 (SD 14.3) years, and mean BMI was 28.9 (SD 6.1). Fifty-eight patients (27.0%) were admitted to MCU/ICU, of whom 56 received invasive ventilation during a median of 11.5 days (IQR 7.8–17.3). In total, 192 (89.3%) patients had oxygen therapy, and 16 (7.4%) patients had non-invasive ventilation. Forty patients (18.6%) died within 21 days. Compared with non-deceased patients, patients who died were older (66.9 [SD 12.0] vs. 59.8 [SD 14.5] years; P < 0.005) and more often invasively ventilated; BMI was similar between these two groups (Table 1). Non-survivors had a larger CSA of IMAT (median 10.1 cm2 [IQR 5.0–18.0] vs. 6.2 cm2 [IQR 3.7–11.4], P < 0.01) and a larger IMAT index (median 3.6 cm2/m2 [IQR 1.6–8.1] vs. 2.1 cm2/m2 [1.2–3.9], P < 0.05) as compared with survivors (Table 1). No statistically significant differences were observed for CSA or mean radiodensity (HU) of muscle, SMI or CSA of SAT (Table 1). Figure 1E shows the Kaplan–Meier curve of the two groups divided by the median IMAT index (2.35, P < 0.05). The Cox proportional hazards model including age, BMI, muscle density and IMAT index showed an effect of IMAT index only (HR = 1.2, 95% CI 1.1–1.3, P < 0.001). See Supporting Information, Data S3, for the complete model. Our findings indicate that a larger CSA of intramuscular adipose tissue at Th12 is a risk factor for survival in COVID-19 patients. This association was independent of age, BMI and muscle density. Our finding is in line with studies in non-COVID patients, showing a relation between abnormal body composition, in particular sarcopenia, and survival in patients with malignancies or other conditions.5, 6 Our study adds to previously reported associations in COVID-19 patients between visceral fat and ICU admission.8, 9 Our data suggest that survival in COVID-19 is related to a marker of fatty muscle degeneration. At Th12, muscles of both inspiration (external intercostals) and active expiration (abdominal muscles, quadratus lumborum) contribute to optimal breathing function. We did not examine the diaphragm, pulmonary function tests or muscle biopsies. Consequently, we can only speculate on mechanisms explaining how fatty muscle degeneration at the low thoracic level leads to lower survival, which may include respiratory muscle impairment (e.g. ineffective cough leading to more progressive pulmonary disease). In addition, our findings may be of interest in relation to a description of a severe diaphragm myopathy with increased fibrosis in a post-mortem study of severely ill COVID-19 patients.13 The most frequently used level for measurements of body composition is L3, because the CSA of muscles and adipose tissue at L3 correlates well with total body volumes of skeletal muscle and adipose tissue.14 Our approach seems valid, however, as a strong correlation between CSA of muscle at Th12 and L3 has been reported.15 Our study has some limitations. As only COVID-19 patients with the availability of chest CT were included, some admitted patients with mild disease were not included. Second, only clinical data and outcome during the first 21st days were available, as many patients were transferred to other centers. Finally, not for all patients, other variables of interest (e.g. diabetes or cardiovascular disease) were present. Our findings indicate that intramuscular adipose tissue is associated with survival in COVID-19 patients. Quantification of IMAT on chest CT examinations might be a tool for risk assessment in COVID-19 patients. A.R. Viddeleer: Conceptualization, methodology, software, formal analysis, investigation, writing—original draft preparation, resources. M. Min: Formal analyses, writing—review and editing preparation. J. Raaphorst: Conceptualization, writing—review and editing preparation. L.F.M. Beenen: Resources, writing—review and editing preparation. M.J. Scheerder: Resources, writing—review and editing preparation. A.P.J. Vlaar: Writing—review and editing preparation. M. Beudel: Conceptualization, writing—review and editing preparation. R. Hemke: Conceptualization, methodology, formal analysis, investigation, writing—original draft preparation, resources, project administration management. Amsterdam UMC COVID-19 Biobank: Data acquisition. The Amsterdam UMC COVID-19 Biobank was supported by grants of the Amsterdam Corona Research Fund, Dr. C.J. Vaillant Fund, and Netherlands Organization for Health Research and Development (ZonMw; NWO-Vici-Grant [grant number 918·19·627]) to Prof. D. van de Beek. None declared. Data S1. Supplement 1. Collaborators Amsterdam UMC Covid-19 Biobank. Data S2. Supplement 2. Study Flow Chart. Data S3. Supplement 3. Cox proportional hazards model. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.