Relation Between Body Composition, Systemic Inflammatory Response, and Clinical Outcomes in Patients Admitted to an Urban Teaching Hospital with COVID-19

Background COVID-19 has been associated with cases of severe respiratory illness, admissions to intensive therapy units (ITUs), and high mortality rates. Objectives The aim of the present study was to examine the relation between computed tomography- body composition (CT-BC) measurements, systemic inflammation, and clinical outcomes in those with COVID-19. Methods Patients who presented to our institution between March 17 and May 1, 2020, with a positive PCR test for COVID-19 or characteristic radiological changes, were assessed for inclusion. Data collected included general demographic details, clinicopathological variables, poGPS, NLR, CT-BC measurements, and clinical outcomes including ITU admission and 30-d mortality, of those admitted. Results Sixty-three patients met the study inclusion criteria. Forty-two patients (67%) were aged ≥70 y, 30 (47.6%) were male and 34.9% ( n = 22) had a poGPS ≥1. ITU admission was significantly associated with a high VFA ( P < 0.05). Thirty-day mortality was associated with high VFA (P < 0.05) and low SMI (P < 0.05). Conclusions Sarcopenia in the presence of obesity was associated with clinical outcomes including greater 30-d mortality.


Introduction
The WHO declared the outbreak of novel coronavirus 19 (COVID-19) a global pandemic on March 11, 2020 (1). Despite an expansion in resources for testing and contact tracing, hospital admissions and death rates within the United Kingdom remained high (2). Since first identified, COVID-19 has been associated with cases of severe respiratory illness, often requiring hospitalization and in some cases admission to an intensive therapy unit (ITU), as well as high mortality rates (3). With the potential for health services to become overwhelmed due to finite resources such as ventilators and level 3 ITU beds available and staffed, factors that aid in prognostication are essential to triage those admitted with COVID-19. This could provide an invaluable insight in the fight against the current global pandemic.
A marked systemic inflammatory response has been identified as one of the signs of severe COVID-19 (4). Recent studies have shown that severe systemic inflammation is associated with mortality in those with COVID-19, suggesting that it can have a role in determining prognosis. Furthermore, obesity, as measured by BMI (5) and visceral fat area (VFA) (6,7) derived from computed tomography (CT) image analysis, has been reported to have a detrimental impact on clinical outcomes in those with COVID-19. The relation between CT-derived measures of body composition including low skeletal muscle mass and density, systemic inflammation, and outcomes in those with cancer have previously been reported (8,9). However, to date, there have been no studies exploring the relation between systemic inflammation, CT-derived body composition (CT-BC) measurements, and clinical outcomes in those with COVID-19. Therefore, the aim of the present study was to examine the relation between CT-BC measurements, systemic inflammatory status, and clinical outcomes in those with COVID-19.

Methods
Data were collected on patients who attended the Emergency Department (ED) and Acute Assessment Unit (AAU) at Glasgow Royal Infirmary (GRI), Glasgow, United Kingdom, during the initial 7-wk period of the COVID-19 pandemic in Glasgow city (March 17, 2020 to May 1, 2020). GRI is a university teaching hospital, serving an urban population with a high burden of socioeconomic deprivation. In line with UK National Health Service (NHS) policy, this study was approved by the NHS Greater Glasgow and Clyde Caldicott guardian. The study protocol (GN20AE307) was approved by the North West England-Preston research ethics committee (20/NW/0336) and registered with clinicaltrials.gov (NCT04484545).
Patients displaying clinical signs or symptoms consistent with possible COVID-19 (as defined by Health Protection Scotland) (10), at the time of presentation to the ED and AAU, were assessed for inclusion in the study. Patients were then further analyzed to identify those with either a positive PCR test or radiological changes characteristic of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), reported on chest X-ray or CT thorax, by a board-certified radiologist. Finally, patients with confirmed COVID-19 were then assessed to identify those who had CT imaging within 3 mo of the diagnosis (see Figure 1). Eligible CT imaging required cross-sectional scanning at the level of the third lumbar vertebra. Patients whose scans were taken outwith this period were excluded from the study. Furthermore, scans with significant movement artefact or missing region of interest were not considered for inclusion.
Routine demographic details, clinical observations, hematological and biochemical laboratory results, as well as clinical outcome data were recorded. Age, sex, ethnicity, BMI, and diagnostic modality confirming COVID-19 as well as date of diagnosis were minimal inclusion criteria. Age categories were grouped to <70 y or ≥70 y. Social deprivation was defined by the Scottish Indices of Multiple Deprivation 2019 based on individuals' home postcodes. Ethnicity was classified as white or other ethnic group. Admission serum C-reactive protein (CRP), albumin, and differential blood cell counts were categorized using local reference intervals. Neutrophil:lymphocyte ratio (NLR) (11) and the perioperative Glasgow Prognostic Score (poGPS) (12) were used to assess systemic inflammation. For this study, thresholds of NLR <3, 3-5, >5 were chosen and categorized as "mild," "moderate," and "severe" systemic inflammatory response, respectively. poGPS values were grouped into "noninflamed" (i.e., poGPS = 0) and "inflamed" (i.e., poGPS = 1 or 2) cohorts. Primary outcomes measured were intensive care admission and mortality within 30 d of diagnosis with COVID-19.

Body composition analysis
Each CT image was individually analyzed using ImageJ-a free to download, Java-based program developed by NIH (NIH ImageJ version 1.47; http://rsbweb.nih.gov/ij/) shown to provide reliable measurements (13). Body composition measurements derived from the CT image slice at L3 included total fat area (TFA), visceral fat area (VFA), and skeletal muscle area (SMA). Attenuation thresholds were from −190 to +30 Hounsfield units (HU) for fat and −29 to +150 HU for muscle. The TFA was quantified by depicting the outer contours of the abdominal wall, compared with the inner contour of the psoas and abdominal wall muscles for VFA. Similarly, SMA was measured by manually delineating muscle areas including the quadratus lumborum, psoas, rectus abdominus, and erector spinae muscles, and the internal transverse and external oblique muscle groups. Skeletal muscle radiodensity (SMD) was calculated (in Hounsfield units) as the mean of the measured muscle area used to calculate SMI. Subcutaneous fat area (SFA) was calculated by subtraction of the VFA from TFA. SFA and SMA measurements were then normalized by division of the patient's height in meters squared to generate a subcutaneous fat index (SFI: centimeters squared/meters squared) and skeletal muscle index (SMI: centimeters squared/meters squared). These indices were then compared with established thresholds for body composition status (see Table 1). Missing data were excluded from analysis on a variable-byvariable basis. Two-tailed P values <0.05 were considered statistically significant. Statistical analysis was performed using SPSS software version 25.0 (SPSS Inc.).
Of the patients included, 3 (5%) had an ITU admission. Two patients were admitted directly to ITU from the ED, with 1 requiring escalation to a level 3 bed from ward-level care during admission. ITU admission was significantly associated with a high VFA (P < 0.05; Table 3). Thirty-day mortality was associated with high VFA (P < 0.05) and low SMI (P < 0.05; see Tables 3 and 5, respectively).

Discussion
To our knowledge, this is the first study to explore the relation between CT-BC measurements, systemic inflammation, and outcomes in patients with COVID-19. The patients included were mainly elderly, were of white ethnicity, were systemically inflamed, overweight with subcutaneous and visceral obesity, and had sarcopenia using standard thresholds. Furthermore, sarcopenia in the presence of obesity was associated with clinical outcomes including greater 30-d mortality. Therefore, it would appear that body composition could have an important role in predicting clinical outcome in patients presenting with COVID-19. Further large-scale studies are warranted to establish the prognostic role of body composition in these patients.
Numerous studies have suggested that obesity, as measured by BMI, is associated with poorer outcomes in patients with COVID-19 (5,17). However, BMI reflects both fat and muscle mass in the body and therefore it is not clear whether such increased risk is due to high fat mass, low muscle mass, or both. In the present study visceral obesity appeared to be associated with a lower 30-d mortality whereas sarcopenia was associated with a higher 30-d mortality. The basis of this divergence of body composition components and clinical outcome is not clear. However, a low muscle mass against a background of an acute (18) or chronic inflammatory state has long been recognized to be associated with poor clinical outcomes (19). Irrespective, it will be important to carry out further body composition studies in patients with COVID- 19 . Sarcopenia has been shown to be prevalent in the elderly population as well as those with cancer (20,21). The prevalence of a low SMI in this COVID-19 cohort was ∼50% when those with cancer were excluded. If we compare this with cohorts of patients with curative colorectal and advanced lung cancer, similar levels of prevalence of a low SMI are observed (8,9). This would suggest that sarcopenia is endemic and not   (22). From the present results it can be speculated that COVID-19 patients with sarcopenia will benefit most from such therapeutic agents. Systemic inflammation has been shown to be associated with poor outcomes in patients with COVID-19 (23). In addition, several studies have shown the negative impact of an elevated NLR on those with COVID-19 (24,25). The poGPS is a validated score that is independently associated with infective complications and 30-d mortality in patients undergoing surgery (12). This score was chosen due to the significant degree of inflammation exhibited by those with COVID-19. A similar prevalence of systemic inflammation, as measured by CRP ≥80, poGPS ≥1, and NLR to >5, was observed in the present cohort (49%, 36%, and 62%, respectively), and across the entire cohort from which the patients in this study were identified (51%, 25%, and 55%, respectively) (23). Furthermore, when compared with COVID-19 cohorts from further afield, such as the Far East, such systemic inflammation was also prevalent (26). Therefore, activation of the host systemic inflammatory response is a consistent feature of this disease. From the present results it may be speculated that the prognostic value and treatment of the systemic inflammatory response will be greatest in those COVID-19 patients with sarcopenia.
There are a number of limitations of this present study. Importantly, this study is a single-center study with a small sample size and therefore subject to sample bias. Although the present study has a small sample size, it is important to highlight that not all patients with COVID-19 undergo routine CT imaging in the United Kingdom. Within the literature there is a single study with a larger cohort than ours (27). However, they used a nonstandardized methodology for the calculation of SMI (27). Two other smaller studies assessed the relation between VFA and clinical outcomes in those with COVID-19 (6,7). To our knowledge, the present study has the largest cohort to date exploring the relation between CT body composition measurements, systemic inflammation, and clinical outcomes in patients with COVID-19. Therefore, the    present cohort provides a novel insight into the relation of body composition and systemic inflammation in those with COVID-19. Furthermore, although it is possible that the relation of SMI with mortality was an age-related factor, when patients older than 65 y were excluded from the univariate analysis, the association between SMI and 30-d mortality remained significant (n = 21, P = 0.028). A larger cohort of patients will be required to address this point in detail.
In summary, sarcopenia in the presence of obesity was associated with clinical outcomes including greater 30-d mortality. Therefore, it would appear that body composition can have an important role in predicting clinical outcome in patients presenting with COVID-19.