Association of Skeletal Muscle, Body Fat and Frailty With Prolonged Length of Hospital Stay Following Colectomy for Adenocarcinoma: A Pilot Case-Control Study.

1
Introduction
Frailty and sarcopenia are known contributors of poor surgical outcomes [1]. Although muscle mass naturally decreases with age, patients with cancer are at an increased risk of accelerated muscle loss and physical function [2]. Several tools have been developed for identifying frailty and sarcopenia for predicting post‐operative complications, morbidity, and mortality; however, none have been used to identify patients at risk of prolonged hospital length of stay (LOS) following oncologic resection. Prolonged hospitalization increases healthcare costs and further impairs mobility for patients following surgery [3]. The primary aim of this pilot study is to examine the association of body composition on LOS following colectomy for colonic adenocarcinoma. Secondly, this study examined association of select patient demographic factors on prolonged LOS.

2
Methods
A case‐control pilot study was designed to compare body composition and frailty to LOS and determine feasibility. A single institutional database was queried to identify adult patients who had undergone a colon resection for adenocarcinoma. A simple random sample was selected to include 50 patients in each arm based on prolonged LOS, defined as greater than or equal to the 75th percentile for the cohort (7 days) [4]. Frailty was determined by the 5‐factor modified frailty index (mFI‐5) and pre‐operative weight loss (Yes vs No) [2]. Multiple studies have validated the use of the third lumbar vertebral segment (L3) on computerized tomography (CT) scan to calculate muscle mass cross‐sectional area and normalize segmental cross‐sectional muscle area to the patient's height [5]. The L3 segment on CT scans was analyzed with two methods in this study. ODIASP [6], an open‐label artificial intelligence platform, was utilized to calculate cross‐sectional muscle area (CSMA) and skeletal muscle index (SMI). The software automatically calculates CSMA at the L3 segment. SMI is then calculated by dividing CSMA by the patient's height (cm2/m2) [5].
3D‐Slicer, a free image processing software, was used to analyze skeletal muscle and fat mass ratios. Using the “Segment Cross‐Sectional Area” module, the L3 segment was manually identified and the software produced cross‐sectional area values of skeletal muscle, torso fat (TF), subcutaneous fat (SQF), and intramuscular fat (IMF), recorded in mm2. Just as the widely accepted body mass index and SMI normalize fat and muscle values to the patient's height, this study generated a total fat index, which was calculated by dividing the sum of TF, SQF, and IMF by the patient's height in m2.
Skeletal muscle and body fat have each independently proven to have an impact on post‐operative outcomes. In order to identify a potential relationship between different types of fat and skeletal muscle, several ratios were calculated by dividing the cross‐sectional area of TF, SQF, IMF by the cross‐sectional area of skeletal muscle obtained through the 3D Slicer software. The Chi‐square test and binary logistic regression analysis were used to determine associations between covariates and outcome.

3
Results
Of the 100 patients selected, 24 were excluded due to metastatic disease at diagnosis, receipt of neoadjuvant therapy, or missing radiographic data. After exclusions, the prolonged LOS group included 40 patients, and the control group included 36 patients. Groups did not differ in age, sex, body mass index, and socioeconomic status, as determined by the County Deprivation Index [5] (Table 1). Patient selection through a simple random sample was intended to reduce selection bias and ensure appropriate representation of all patients undergoing colonic resection of any type. Weight loss and mFI‐5 were not associated with prolonged LOS (p = 0.060 and p = 0.469). CSMA (OR: 0.994, 95% CI: 0.989–0.999, p = 0.029) and SMI (OR: 0.981, 95% CI: 0.965–0.996, p = 0.015) were associated with decreased LOS. Ratios of skeletal muscle to body fat were not associated with LOS (p = 0.195, p = 0.237, p = 0.161) (Table 2).

4
Discussion
The results of this study suggest that skeletal muscle mass may be associated with prolonged LOS. Previous studies have shown that sarcopenia is correlated with increased mortality and morbidity, increased rates of readmission, and increased rates of non‐home discharge, particularly in patients recovering from oncologic surgery [7]. The present study warrants further investigation into a relationship between CSMA, SMI, and LOS.
While some body composition analysis studies have shown a relationship between body fat and post‐operative outcomes, these results did not hold true in the current study [8]. Body fat indices and ratios of body fat to skeletal muscle were not associated with LOS, which may suggest that skeletal muscle is a better predictor of post‐operative recovery when compared to body fat. With further studies, this may guide targeted prehabilitation measures which emphasize increasing muscle mass rather than decreasing body fat.

5
Conclusions
The results from this pilot study warrant further investigation in a large, observational cohort study examining the associations of frailty measures and prolonged LOS. These findings have the potential to provide better recommendations for prehabilitation and help to guide in post‐operative discussions with patients and their families.

Author Contributions

Annmarie Butare: conceptualization, methodology, investigation, writing – original draft, writing – review and editing, data curation, resources. Justin Drake: conceptualization, writing – review and editing, supervision. Michael D. Honaker: conceptualization, writing – review and editing, methodology, validation, formal analysis, supervision.

Funding
The authors have nothing to report.

Ethics Statement
This study was reviewed by the University IRB and considered exempt because information has been recorded by the investigator in such a manner that the identity of the human subjects cannot readily be ascertained directly or through identifiers linked to the subjects, the investigator does not contact the subjects, and the investigator will not re‐identify subjects. IRB Approval Number: UMCIRB 24‐002006.

Consent
The authors have nothing to report. This project was approved by the university IRB and was deemed exempt.

Conflicts of Interest
The authors declare no conflicts of interest.