Temporal change in skeletal muscle index as a predictor of recurrence for patients with locally advanced colorectal malignancy: A retrospective cohort study

Temporal change in skeletal muscle index as a predictor of recurrence for patients with locally advanced colorectal malignancy: A retrospective cohort study | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Temporal change in skeletal muscle index as a predictor of recurrence for patients with locally advanced colorectal malignancy: A retrospective cohort study Noah B. Manz, Jordan Reed, Christopher D. Kanner, Jeremy A. Dressler This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7350138/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 07 Apr, 2026 Read the published version in Cancer Imaging → Version 1 posted 9 You are reading this latest preprint version Abstract Background: Colorectal cancer remains a leading cause of cancer-related morbidity and mortality with stage III disease carrying a substantial risk of recurrence despite curative resection. Accurate risk stratification is essential to optimize surveillance and guide adjuvant therapy. Traditional models rely heavily on pathologic features, but recent studies suggest that body composition metrics, particularly imaging-based assessments of skeletal muscle mass, may offer additional prognostic value. The skeletal muscle index, derived from routine CT imaging, has emerged as a promising surrogate marker of frailty. However, the relationship between temporal changes in SMI and cancer recurrence remains poorly defined. Methods: A retrospective cohort study was performed using single-institution data from over 500 patients aged 18 or greater who underwent resection for locally advanced stage III colorectal malignancy between years 2000 and 2020. Skeletal muscle index was measured at the third lumbar vertebral level using preoperative CT imaging from the time of initial diagnosis. For patients with recurrence, a follow-up measurement was obtained at their recurrence date, and using propensity score analysis, a patient without recurrence was selected to best match this follow-up duration. Temporal changes in skeletal muscle index were then calculated and compared with a conditional logistic regression. Receiver operating characteristics were then analyzed to identify clinically relevant thresholds of SMI decline. Results: A decrease in skeletal muscle index was independently associated with increased risk of disease recurrence with an odds ratio of 1.30 per 1-point decrease in SMI (95% CI: 1.09–1.54, p = 0.003). Receiver operating curve analysis suggests that an SMI decline of 2.5–6.0 cm²/m² within 1–5 years after resection may serve as a practical threshold for risk stratification. Conclusions: Postoperative skeletal muscle index decline is a significant, independent predictor of cancer recurrence and may serve as a clinically useful marker to personalize follow-up care and enhance risk stratification. These findings support the potential role of body composition monitoring to guide postoperative management and highlight a need for further prospective validation. Skeletal Muscle Index Sarcopenia Colorectal Cancer Recurrence Figures Figure 1 Figure 2 Figure 3 Figure 4 Background Colorectal Cancer (CRC) remains one of the most common malignancies worldwide with a substantial risk of recurrence after oncologic resection. Current 5-year recurrence rates for stage III disease are ~33% with a median recurrence time of 15.9 months. 1,2 Detection of recurrence has significant prognostic implications as patients with isolated or oligometastatic disease may be eligible for curative interventions associated with improved long-term survival. 3 As such, identifying which patients are at high risk of recurrence is critical for optimizing postoperative surveillance strategies. Traditional risk stratification models focus on pathologic features including tumor stage, margin status and lymph node involvement 4-6 , but previous studies have demonstrated a correlation between frailty and postoperative outcomes for patients who require oncologic resection. 7-15 In these studies, researchers employ various methods to assess patient frailty; however, no single clinical frailty score or surrogate marker has yet demonstrated clear superiority- an important limitation highlighted in a recent meta-analysis. 29 Existing tools such as the Fried Frailty Phenotype and the Rockwood Clinical Frailty Scale rely on subjective evaluations which are often highly variable and susceptible to individual bias and error. 16-17 Therefore, imaging-based sarcopenia assessment is increasingly seen as a viable solution to integrate frailty measurement into oncologic workflows while avoiding some of the limitations of clinical assessments. In this approach, a skeletal muscle area (SMA) measurement can be taken at the mid-level of the third lumbar vertebral body (L3) on axial computed tomography (CT) scans due to its strong correlation with whole-body muscle mass. 18 The skeletal muscle index (SMI) heuristic can then be generated by normalizing the SMA by the square of the patient’s height, and the progression of these SMI values over time can be tracked for postoperative surveillance. While sex-specific thresholds for SMI have been proposed in literature to define sarcopenia, optimal cut-off values may vary due to cancer type, patient population and other factors. 19 Furthermore, the body of longitudinal studies comparing temporal changes in SMI to long-term oncologic outcomes remains quite limited compared to the large number of cross sectional papers correlating outcomes to a single SMI datum. 30-31 As such, additional research is needed to validate a longitudinal approach to SMI surveillance and establish clinically meaningful cutoffs to guide the intensity of postoperative monitoring and adjuvant treatment. It is hypothesized that long-term decreases in SMI will be positively correlated with disease recurrence. Therefore, our study sought to better understand the time-dependent relationship between changes in skeletal muscle index and postoperative recurrence risk in patients with locally advanced colorectal malignancy. Methods We performed a single-institution (University of Vermont Medical Center) retrospective analysis that included patients aged 18 years or greater who underwent oncologic resection for locally advanced stage III colorectal malignancy between years 2000 and 2020. This study was approved by the University of Vermont Institutional Review Board (IRB #00002150) and was conducted in compliance with the Health Insurance Portability and Accountability Act (HIPAA). Researchers utilized the institution’s digital imaging PACS software to measure psoas, paraspinal and abdominal muscle cross sectional area for included patients at their date of initial diagnosis. Muscle regions were isolated by applying and manually refining a [-29, 150] Hounsfield Unit filter at the mid-vertebral body of the third lumbar vertebrae on axial CT scan (reference Figure 1). 20-23 Skeletal muscle indices were subsequently calculated as the quotient of this area and the square of the patient’s height. 19 An electronic medical record search was then performed to obtain demographic information, post-operative clinical outcomes data and confounding variables for included patients. The cohort was then split according to cancer recurrence status between 1 and 5 years following the initial diagnosis. For patients with recurrence, a follow-up SMI measurement was obtained at their recurrence date. For patients without recurrence, a comprehensive list of their CT study dates was collected, and the Python Propensity Score Matching (PSM) library (version 0.3.13) was then used to pair patients with and without recurrence. 24 The list of available CT study dates for patients in the non-recurrence category allowed for the duration between the first and second SMI measurements to be matched in addition to other specified confounders. Specifically, this was done by creating n copies of each non-recurrence patient in the database, populated with one of n unique CT study dates. Different follow-up durations could then be passed as a direct PSM covariate, and based on the indicated best-match, follow-up SMI measurements for non-recurrence patients could be obtained by reverse indexing the non-recurrence ID with its corresponding follow-up date. With both initial and follow-up SMI measurements for patients in the recurrence and non-recurrence categories, a temporal change in SMI could be calculated. The primary outcome measure of the study was the difference in these SMI changes between paired patients which were evaluated in a conditional logistic regression model using the Python Statsmodels package (version 0.14.4). The standard significance level of 0.05 was used for all hypothesis testing. Results A total of 519 patients were identified in the UVMMC database as undergoing surgical resection for colorectal malignancy between 2000 and 2020. Demographics from this sample population are presented in Table 1 . Chart review identified a subset of 292 patients with AJCC stage III colorectal cancer for whom additional confounding variables were collected. To control for factors that could independently influence recurrence or imaging frequency, we excluded a priori patients with concurrent malignancies, long-term steroid use, or significant comorbidity burden defined as having three or more conditions each with a Charlson Comorbidity Index weight of 1 (e.g., CHF, PAD, COPD, prior MI). Patients who did not undergo surgical resection were also excluded as surgery is central to curative treatment. After exclusions, 240 patients remained: 86 experienced recurrence between 1 and 5 years following resection (average 2.3 years), and 154 did not. An inclusion/exclusion criteria flowchart is shown in Fig. 2 . For patients who did not experience recurrence, all archived CT study dates were obtained averaging 7 plausible scan dates per patient. Propensity scoring matched 36 patients using the standard caliper width of 0.2 standard deviations, and good balance (effect size < 0.2) was obtained for all covariates. A sensitivity analysis evaluated the impact of varying caliper widths on the matching process with minimal observed variation in match rates. At a caliper width of 0.5, only 37 patients were matched, and a caliper of 0.2 was determined to be the narrowest possible width while maintaining adequate sample size. Formal power analysis was not conducted as the use of propensity score matching, by design, determines the final sample size based on covariate overlap and match quality. The reductions in effect size of the covariates before and after matching are outlined in Table 2 and suggested overall that the matches were well balanced. 25 Table 1 Demographics of the sampled patient population. ICD-O-3 = International Classification of Diseases for Oncology, Third Edition. AJCC = American Joint Committee on Cancer. Demographic Value/Distribution Male/Female Ratio 308/211 (59%/41%) Age at diagnosis (mean, stddev, range) in years All patients (65.8, 13.7, 32–95), Male (64.4, 13.3, 32–91), Female (67.9, 14.1, 32–95) Primary Site C209 Rectum, NOS ( 24% ), C180 Cecum ( 15% ), C182 Colon, ascending ( 14% ), C187 Colon, sigmoid ( 11% ), C184 Colon, transverse ( 6% ), C187 Sigmoid, NOS ( 4% ), C199 Rectosigmoid, NOS ( 4% ), C180 Ileocecal valve ( 3% ), C186 Colon, descending ( 3% ), C183 Colon, hepatic flexure ( 3% ), C199 Rectosigmoid colon ( 3% ), C185 Colon, splenic flexure ( 3% ), C199 Rectosigmoid junction ( 3% ), C188 Colon, overlapping lesion ( 1% ), C180 Ileocecal junction ( 1% ), C189 Colon, NOS ( 1% ), C182 Colon, right ( 1% ) Histo/Behavior ICD-O-3 81403 Adenocarcinoma, NOS ( 64% ), 84803 Mucinous adenocarcinoma ( 10% ), 82633 Adenocarcinoma in tubulovillous adenoma ( 9% ), 82103 Adenocarcinoma in tubular adenoma ( 7% ), 85103 Medullary adenocarcinoma ( 2% ), 82553 Adenocarcinoma with mixed subtypes ( 1% ), 82613 Adenocarcinoma in villous adenoma ( 1% ), 84903 Signet ring cell adenocarcinoma ( 1% ), 82103 Adenocarcinoma in adenomatous polyp ( 1% ), 85103 Medullary carcinoma, NOS ( 1% ), 82103 Adenocarcinoma in a polyp, NOS ( 1% ) Best AJCC Stage IIIB (35%) , IIA (32%) , IIIC (16%) , IIB (7%) , IIIA (6%) , IIC (4%) Table 2 Effect size of covariates before and after PSM matching. Variable Effect Size (Before) Effect Size (After) Scan Follow-Up Duration 0.433 0.009 Smoking 0.686 0.117 Chemotherapy 0.316 0.131 Radiation 0.020 0.000 Histo/Behavior ICD-O-3 0.463 0.131 Best AJCC Stage 1.198 0.123 Primary Cancer Site 0.645 0.155 Diabetes Mellitus 0.292 0.181 The CT imaging for matched recurrence and non-recurrence patients was evaluated and SMI data was collected and made publicly available in a Mendeley Data repository. 26 The temporal change in SMI was then calculated for each patient where positive changes indicated net muscle gain over time and negative changes indicated net muscle loss. A scatter plot of the change in SMI versus the duration between initial and final scan dates is shown in Fig. 3 a along with underlying distributions. The change in SMI data exhibited high central tendency (mean = -0.89 cm 2 m − 2 , stddev = 6.60 cm 2 m − 2 ) while the scan durations were more diffusely distributed. The individual distributions of SMI changes as a function of recurrence status were plotted in Fig. 3 b with lines connecting the paired observations. The relationship between SMI change and cancer recurrence was assessed using a conditional logistic regression model that accounted for matched pair IDs with results shown in Table 3 . Table 3 Conditional logistic regression on matched participants. Stratification variables generated from the matching condition are removed for readability. Variable Coefficient Std. Error z-value p-value 95% CI Lower 95% CI Upper SMI Change -0.2589 0.089 -2.922 0.003 -0.433 -0.085 Intercept 0.7689 1.448 0.531 0.595 -2.069 3.607 The logistic model indicated that each 1-point increase in SMI change was associated with a 23% decrease in the odds of recurrence ( OR = 0.77, 95% CI: 0.65–0.92, p = 0.003). Alternatively, a 1-point decrease in SMI change was associated with a 30% increase in the odds of recurrence ( OR = 1.30) which are findings consistent with previous literature. 7 – 15 A likelihood ratio test comparing the conditional logistic model to a null model yielded a non-significant result ( p ~ 1.0), reflecting the quality of the initial propensity score matching which produced tightly matched pairs with limited residual variability. However, the association between SMI change and recurrence remained significant even after accounting for this matched structure suggesting that within matched pairs, differences in SMI change were consistently associated with recurrence status independent of other matched variables. Visually, this within-pair association is shown in Fig. 3 c which displays the distribution of SMI change differences across matched pairs. To complement the regression, a Wilcoxon signed-rank test applied to this distribution yielded W = 72.0 ( p = 0.014), further supporting a meaningful difference between SMI change and recurrence status consistent with the regression results. Given these results, the null hypothesis was rejected given the standard significance level of 5%. Discussion Based on this data, it may be possible to use SMI as a surveillance tool for patients who recently underwent oncologic resection to indicate potential cancer relapse. By using predicted probabilities from the logistic model, a receiver operating characteristic curve was generated to assess predictive performance. This curve is shown in Fig. 4 in addition to a bootstrap simulation of ROC curves calculated from randomly resampling from the distributions in Fig. 3 b. The purpose of the bootstrap was to validate the stability of the ROC curve generated from the logistic model, particularly in the event of an initial random sampling error. 27 The significant proportion of bootstrapped ROC curves (depicted as an overlapping density of thin gray lines) that performed better than chance suggests that the model’s discriminative ability is robust. For such a test, sensitivity is typically prioritized, and a true positive rate of approximately 90% is commonly targeted in established oncology screenings. 28 – 29 In the ROC generated by the logistic model, the optimal classification threshold based on the Youden J statistic (J = 0.60) was a predicted probability of 0.53. At this threshold, the model achieved a sensitivity of 80% and a specificity of 80% corresponding to an SMI change of 2.5 cm²/m² on average. While this threshold provides a balanced tradeoff between sensitivity and specificity, it may be suboptimal for a surveillance context where higher sensitivity is paramount. Arbitrarily setting the sensitivity to 90% as a target value, a specificity of 40% can be obtained with a threshold of 0.31 corresponding to an SMI change of 6.0 cm²/m². Although this results in a higher false-positive rate, it maximizes the model’s ability to detect potential relapses and may serve as a clinically meaningful cutoff for early identification of high-risk individuals. Notably, the inflection point between net gain or loss of skeletal muscle (i.e SMI change = 0 cm²/m²) occurred at a threshold value of 0.68 where the model returned a sensitivity of 41% and specificity of 84%. This threshold may be more appropriate for confirming absence of recurrence in lower-risk patients, but the low sensitivity ultimately limits utility in primary surveillance. Taken together, these results suggest that an SMI decline in the range of 2.5–6.0 cm²/m² within 1–5 years following resection may represent a practical and clinically actionable threshold for increased surveillance. While prospective validation is warranted, such a cutoff could help identify patients who would benefit from intensified follow-up, imaging, or early intervention strategies, particularly if commensurate with additional surveillance tests. While a standalone cutoff may be sufficient for binary risk classification, the more robust approach involves building comprehensive models that incorporate a larger set of validated pathologic risk factors. However, expanding the input space of such models introduces challenges, particularly the need for increasingly large datasets to support reliable parameter estimation. The temporal nature of SMI change lends itself to Bayesian updating where new, independently associated risk factors can be considered enabling posterior probabilities to reflect additional evidence without requiring the construction of a single unified model. 33 This is particularly useful when different datasets contain complementary but non-overlapping features making joint model training impossible. In practical terms, this means the prior probability of recurrence (obtained from an established model) can be updated using a likelihood ratio derived from SMI change. We assume conditional independence between SMI and the prior probability given recurrence status, which is reasonable since SMI is measured independently of the pathologic features often used in existing risk models. Therefore, combining models in this manner does not introduce artificial bias. One attractive model for a prior was developed by Zafar et al at MD Anderson Cancer Center and uses a Cox regression to estimate 5-year recurrence probabilities for colorectal malignancy. 34 It considers risk factors such as age, sex, tumor histology, site, margin status and stage, but excludes any weight-based covariates which would be highly correlated with SMI change. This is best practice for ensuring conditional independence of covariates without access to the previous model’s underlying coefficients. Formally, this model outputs the probability of recurrence given the data θ, P(recurrence|θ), which is easily converted to odds via Eq. 1. This conversion enables use of the more convenient form of Bayes Theorem in odds form. 33 Then using the logistic regression coefficients from Table 3 the log-odds of recurrence based on SMI change may be calculated. The likelihood ratio is the quotient of these odds and the marginal odds of recurrence from the UVMMC dataset as shown in Eq. 2 along with a numerical approximation for convenience. The posterior odds of recurrence given the data θ and ΔSMI, O(recurrence|{θ, ΔSMI}), are then calculated as the product of the prior and the likelihood ratio as shown in Eq. 3. These odds may ultimately be converted back to a recurrence probability using Eq. 1. Eq1: \(\:O\left(A\right)=\frac{P\left(A\right)}{1-P\left(A\right)},\:P\left(A\right)=\frac{O\left(A\right)}{1+O\left(A\right)}\) Eq2: \(\:LR\left(Recurrence|\varDelta\:SMI\right)=\frac{O\left(Recurrence|\varDelta\:SMI\right)}{O\left(Recurrence\right)}=\frac{exp\left({\beta\:}_{0}+{\beta\:}_{1}\varDelta\:SMI\right)}{O\left(Recurrence\right)}\approx\:6{e}^{-0.26*\varDelta\:SMI}\) Eq3: \(\:O\left(Recurrence\left|\right\{\varDelta\:SMI,\theta\:\}\right)=O\left(Recurrence|\theta\:\right)\:*LR\left(Recurrence|\varDelta\:SMI\right)\:\) While Bayesian updating provides a coherent framework for incorporating new information, it does not constitute a fully specified or causally rigorous model. Nonetheless, it is well suited to situations where additional data such as changes in SMI emerge after the initial risk assessment. The advantage of this approach is that it offers an actionable way to begin incorporating SMI change into oncologic workflows, rather than simply being aware that a risk-relationship exists. The other advantage of the likelihood ratio is to compute the threshold for unity at which no extra information is added to the prior, or ~ 6.9 cm²/m² corresponding to LR = 1. This is consistent with the results from the receiver operating characteristic and suggests that SMI change overall may be more apt for detecting high-risk individuals below the threshold than for confidently ruling out recurrence in low-risk individuals above it. Ideally, Bayesian updating would be assessed by evaluating the posterior against known outcomes within a dataset fully spanning the input spaces of both the prior and the updating variable. In practice however, such comprehensive datasets are often unavailable, and while strict adherence to a unified likelihood framework may be statistically preferable, it is not always practical particularly in clinical research where data heterogeneity is common. This method strikes a balance between absolute statistical rigor and the practical necessity of leveraging available information. While every effort was taken to ensure methodological rigor and minimize bias, certain limitations of the current work are present. First, the sample size that was passed into the logistic model was relatively small despite a large initial dataset, limiting the overall study power. This was a natural consequence of the strict inclusion/exclusion criteria and number of covariates that were controlled for in the propensity score matching. Furthermore, and as a direct result of the limited size of the dataset, not all covariates could be adequately controlled for, and the matching process was limited to variables for which great balance (effect size < 0.2) could be obtained. While this improved the overall quality of the logistic model for the limited covariates that were included, it is not clear what impact other comorbidities (e.g neoadjuvant vs. adjuvant treatment) would have had. Finally, data was collected from a single institution, introducing a potential selection bias. Future studies with larger, multi-center cohorts and more comprehensive data collection are warranted to validate and expand upon these findings. Conclusions Assessing frailty is a crucial component to providing optimum multidisciplinary care for patients with colorectal cancer who require surgical resection. Imaging-based sarcopenia assessment is increasingly seen as a viable solution to integrate frailty measurement into oncologic workflows, though standardized surveillance protocols have not been established. Furthermore, while previous literature has demonstrated a correlation between frailty and recurrence status, 7 – 15 the body of longitudinal studies comparing temporal changes in SMI to long-term oncologic outcomes remains quite limited. 31 – 32 Our results compile data from a single-institution study of over 500 patients with colorectal malignancy over a 20 year period to estimate threshold values for surveillance risk stratification. These data show a positive correlation between skeletal muscle loss and disease recurrence for patients with CRC where a 1 cm²/m² decline in skeletal muscle index 1–5 years after resection is associated with a 30% increase in the odds of recurrence. Based on the ROC curve generated from regression analysis, an SMI decline in the range of 2.5–6.0 cm²/m² within 1–5 years following resection may represent a practical and clinically actionable threshold for increased surveillance. These findings support the role of body composition monitoring to guide postoperative management and highlight a need for further prospective validation. Abbreviations CRC Colorectal cancer SMI Skeletal muscle index SMA Skeletal muscle area CT Computed tomography ROC Receiver operating characteristic PSM Propensity score matching AJCC American Joint Committee on Cancer UVMMC University of Vermont Medical Center Declarations Ethics Approval and Consent to Participate: This study was conducted in accordance with the Declaration of Helsinki and approved by the University of Vermont Institutional Review Board (IRB #00002150); Date of approval: December 1, 2023. Consent for Publication: Not applicable. Availability of Data and Materials: The datasets generated for this study are publicly available in the Mendeley Data repository, https://data.mendeley.com/datasets/cxyjw6v4gh/1. 26 Competing Interests: The authors declare that they have no competing interests. Funding: Funding was provided by the University of Vermont Larner College of Medicine Department of Radiology Research Fund. Authors’ Contributions: NM developed the methodology, implemented the software, performed validation and formal analysis, conducted investigation and data curation, drafted the original manuscript, and created the visualizations. JR contributed to conceptualization, methodology, investigation, resource provision, data curation, manuscript review, and project administration. CK led conceptualization, methodology, validation, investigation, resource management, data curation, manuscript review, supervision, and project administration. JD contributed to conceptualization, methodology, validation, resources, manuscript review, supervision, and project administration. All authors read and approved the final manuscript. Acknowledgements: We would like to thank Tom Ahern, PhD, for his expert guidance and contributions to the statistical methodology of this study. Authors’ Information: Not applicable. References Osterman E, Glimelius B. Recurrence Risk After Up-to-Date Colon Cancer Staging, Surgery, and Pathology: Analysis of the Entire Swedish Population. Dis Colon Rectum. 2018;61(9):1016–25. 10.1097/dcr.0000000000001158 . Nors J, Iversen LH, Erichsen R, Gotschalck KA, Andersen CL. Incidence of Recurrence and Time to Recurrence in Stage I to III Colorectal Cancer. JAMA Oncol. 2024;10(1):54. 10.1001/jamaoncol.2023.5098 . Zucchelli G, Moretto R, Rossini D, et al. Oligometastatic colorectal cancer: Prognostic implications of tumor load, role of locoregional treatments, and of first-line therapy intensification—A pooled analysis of TRIBE and TRIBE2 studies by GONO. JCO. 2020;38(4suppl):12–12. 10.1200/jco.2020.38.4_suppl.12 . Edge SB, Compton CC. The American Joint Committee on Cancer: the 7th Edition of the AJCC Cancer Staging Manual and the Future of TNM. Ann Surg Oncol. 2010;17(6):1471–4. 10.1245/s10434-010-0985-4 . Nagtegaal ID, Quirke P, Schmoll H-J. Has the new TNM classification for colorectal cancer improved care? Nat Rev Clin Oncol. 2011;9(2):119–23. 10.1038/nrclinonc.2011.157 . Ong MLH, Schofield JB. Assessment of lymph node involvement in colorectal cancer. WJGS. 2016;8(3):179. 10.4240/wjgs.v8.i3.179 . Gartrell R, Qiao J, Kiss N, et al. Can sarcopenia predict survival in locally advanced rectal cancer patients? ANZ J Surg. 2023;93(9):2166–71. 10.1111/ans.18512 . Hopkins JJ, Reif RL, Bigam DL, Baracos VE, Eurich DT, Sawyer MB. The Impact of Muscle and Adipose Tissue on Long-term Survival in Patients With Stage I to III Colorectal Cancer. Dis Colon Rectum. 2019;62(5):549–60. 10.1097/dcr.0000000000001352 . Horie K, Matsuda T, Yamashita K, et al. Sarcopenia assessed by skeletal muscle mass volume is a prognostic factor for oncological outcomes of rectal cancer patients undergoing neoadjuvant chemoradiotherapy followed by surgery. Eur J Surg Oncol. 2022;48(4):850–6. 10.1016/j.ejso.2021.10.018 . Portale G, Zuin M, Spolverato YC, et al. Prognostic effect of sarcopenia in patients undergoing laparoscopic rectal cancer resection. ANZ J Surg. 2023;93(6):1631–7. 10.1111/ans.18269 . Saur NM, Davis BR, Montroni I, et al. The American Society of Colon and Rectal Surgeons Clinical Practice Guidelines for the Perioperative Evaluation and Management of Frailty Among Older Adults Undergoing Colorectal Surgery. Dis Colon Rectum. 2022;65(4):473–88. 10.1097/dcr.0000000000002410 . Xiao J, Caan BJ, Cespedes Feliciano EM, et al. Association of Low Muscle Mass and Low Muscle Radiodensity With Morbidity and Mortality for Colon Cancer Surgery. JAMA Surg. 2020;155(10):942. 10.1001/jamasurg.2020.2497 . Benedek Z, Coroș MF. The impact of sarcopenia on the postoperative outcome in colorectal cancer surgery. Med Pharm Rep. 2023;96(1):20–7. 10.15386/mpr-2483 . Fleming CA, O’Connell EP, Kavanagh RG, et al. Body Composition, Inflammation, and 5-Year Outcomes in Colon Cancer. JAMA Netw Open. 2021;4(8):e2115274. 10.1001/jamanetworkopen.2021.15274 . Sibia US, Badve SB, Istl AC, Klune JR, Riker AI. Impact of Frailty Upon Surgical Decision-Making for Left-Sided Colon Cancer. TOJ. 2023;23(2):120–8. 10.31486/toj.22.0120 . Fried LP, Tangen CM, Walston J et al. Frailty in Older Adults: Evidence for a Phenotype. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences. 2001;56(3):M146–57. 10.1093/gerona/56.3.m146 Rockwood K. A global clinical measure of fitness and frailty in elderly people. Can Med Assoc J. 2005;173(5):489–95. 10.1503/cmaj.050051 . n den Broeck J, Sealy MJ, Brussaard C, Kooijman J, Jager-Wittenaar H, Scafoglieri A. The correlation of muscle quantity and quality between all vertebra levels and level L3, measured with CT: An exploratory study. Front Nutr. 2023;10. 10.3389/fnut.2023.1148809 . Freire PP, Fernandez GJ, de Moraes D, et al. The expression landscape of cachexia-inducing factors in human cancers. J cachexia sarcopenia muscle. 2020;11(4):947–61. 10.1002/jcsm.12565 . Derstine BA, Holcombe SA, Ross BE, Wang NC, Su GL, Wang SC. Skeletal muscle cutoff values for sarcopenia diagnosis using T10 to L5 measurements in a healthy US population. Sci Rep. 2018;8(1). 10.1038/s41598-018-29825-5 . Aubrey J, Esfandiari N, Baracos VE, et al. Measurement of skeletal muscle radiation attenuation and basis of its biological variation. Acta Physiol. 2014;210(3):489–97. 10.1111/apha.12224 . Derstine BA, Holcombe SA, Goulson RL, et al. Quantifying Sarcopenia Reference Values Using Lumbar and Thoracic Muscle Areas in a Healthy Population. J Nutr health aging. 2018;22(1):180–5. 10.1007/s12603-017-0983-3 . n der Werf A, Dekker IM, Meijerink MR, de van der Wierdsma NJ, Langius JAE. Skeletal muscle analyses: agreement between non-contrast and contrast CT scan measurements of skeletal muscle area and mean muscle attenuation. Clin Physio Funct Imaging. 2017;38(3):366–72. 10.1111/cpf.12422 . Kline A, Luo Y, PsmPy:. A Package for Retrospective Cohort Matching in Python. In: 2022 44th Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC). IEEE; 2022:1354–1357. 10.1109/embc48229.2022.9871333 Austin PC. An Introduction to Propensity Score Methods for Reducing the Effects of Confounding in Observational Studies. Multivar Behav Res. 2011;46(3):399–424. 10.1080/00273171.2011.568786 . Manz N, Reed J, Kanner C, Dressler J. Temporal changes in skeletal muscle index dataset for colorectal malignancy recurrence study. Mendeley Data. 2025;V1. 10.17632/cxyjw6v4gh.1 . Gu J, Ghosal S, Roy A. Bayesian bootstrap estimation of ROC curve. Stat Med. 2008;27(26):5407–20. 10.1002/sim.3366 . Gupta S. Screening for Colorectal Cancer. Hematol Oncol Clin N Am. 2022;36(3):393–414. 10.1016/j.hoc.2022.02.001 . Bhamani A, Creamer A, Verghese P, et al. Low-dose CT for lung cancer screening in a high-risk population (SUMMIT): a prospective, longitudinal cohort study. Lancet Oncol. 2025;26(5):609–19. 10.1016/s1470-2045(25)00082-8 . Han J, Zhang Q, Lan J, Yu F, Liu J. Frailty worsens long-term survival in patients with colorectal cancer: a systematic review and meta-analysis. Front Oncol. 2024;14. 10.3389/fonc.2024.1326292 . Hopkins JJ, Reif R, Bigam D, Baracos VE, Eurich DT, Sawyer MM. Change in Skeletal Muscle Following Resection of Stage I–III Colorectal Cancer is Predictive of Poor Survival: A Cohort Study. World j surg. 2019;43(10):2518–26. 10.1007/s00268-019-05054-3 . Deng C-Y, Lin Y-C, Wu JS, et al. Progressive Sarcopenia in Patients With Colorectal Cancer Predicts Survival. Am J Roentgenol. 2018;210(3):526–32. 10.2214/ajr.17.18020 . Spiegelhalter DJ, Abrams KR, Myles JP. Bayesian Approaches to Clinical Trials and Health-Care Evaluation. Published online Dec. 2003;9. 10.1002/0470092602 . Zafar SN, Hu C-Y, Snyder RA, et al. Predicting Risk of Recurrence After Colorectal Cancer Surgery in the United States: An Analysis of a Special Commission on Cancer National Study. Ann Surg Oncol. 2020;27(8):2740–9. 10.1245/s10434-020-08238-7 . Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 07 Apr, 2026 Read the published version in Cancer Imaging → Version 1 posted Editorial decision: Revision requested 09 Dec, 2025 Reviews received at journal 03 Nov, 2025 Reviewers agreed at journal 08 Oct, 2025 Reviews received at journal 17 Sep, 2025 Reviewers agreed at journal 08 Sep, 2025 Reviewers invited by journal 02 Sep, 2025 Editor assigned by journal 13 Aug, 2025 Submission checks completed at journal 13 Aug, 2025 First submitted to journal 11 Aug, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7350138","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":510143252,"identity":"42c0f55b-5fbb-4c42-899a-65805871aad9","order_by":0,"name":"Noah B. Manz","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA4UlEQVRIiWNgGAWjYHACNgjF3sDAwAMVkiBOC88BkrVIJBCpxZy9+dkDxhy7PPmZrxMfvGE4nLi2gfngbR48Wix7jpkbMG5LLmacnbvZcA5Qy7YDbMnW+LQY3Egwk2DcxpzYLJ27TZqH4bCx2QEeM2n8WtK/AbXUJ7ZJnt3+G6KF/xsBLTkgWw4n9kjwbmMGapED2sKGX8uZM2USiduOJ87gyd0sOccgXc7sMJux5Rx8Wo63b5P4uK06cX772Y0f3lRY85gdb3544w0eLWCQgDABiJkJKR8Fo2AUjIJRQBAAAM3ySB+oHti8AAAAAElFTkSuQmCC","orcid":"","institution":"University of Vermont","correspondingAuthor":true,"prefix":"","firstName":"Noah","middleName":"B.","lastName":"Manz","suffix":""},{"id":510143254,"identity":"e95432b5-d189-4f28-87fc-390354ceac9b","order_by":1,"name":"Jordan Reed","email":"","orcid":"","institution":"University of Vermont Medical Center","correspondingAuthor":false,"prefix":"","firstName":"Jordan","middleName":"","lastName":"Reed","suffix":""},{"id":510143255,"identity":"b9b0e848-4c9f-4e90-9262-fd0dbc96aece","order_by":2,"name":"Christopher D. Kanner","email":"","orcid":"","institution":"University of Vermont Medical Center","correspondingAuthor":false,"prefix":"","firstName":"Christopher","middleName":"D.","lastName":"Kanner","suffix":""},{"id":510143257,"identity":"333b4015-f286-475f-9996-09d28ff5f9e3","order_by":3,"name":"Jeremy A. Dressler","email":"","orcid":"","institution":"University of Vermont Medical Center","correspondingAuthor":false,"prefix":"","firstName":"Jeremy","middleName":"A.","lastName":"Dressler","suffix":""}],"badges":[],"createdAt":"2025-08-12 01:08:12","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7350138/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7350138/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s40644-026-01025-9","type":"published","date":"2026-04-07T15:57:55+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":90961791,"identity":"8e542cef-7b17-4699-b88e-92ab5ec74b4a","added_by":"auto","created_at":"2025-09-10 05:14:07","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":722343,"visible":true,"origin":"","legend":"\u003cp\u003eAxial non-contrast computed tomography slice at the mid-level of the third lumbar vertebral body (L3) with a [-29, 150] HU filter manually refined to isolate skeletal muscle including psoas, paraspinal muscles, obliques and rectus abdominis (shown in red). Top left = Sagittal, bottom left = coronal, right = axial plane.\u003c/p\u003e","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7350138/v1/8941d1760e53ce170d7feb8e.jpeg"},{"id":90961792,"identity":"8b50b900-248b-4581-a24c-601e0620bda2","added_by":"auto","created_at":"2025-09-10 05:14:07","extension":"jpeg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":234366,"visible":true,"origin":"","legend":"\u003cp\u003eFlowchart illustrating the inclusion and exclusion process for study participants.\u003c/p\u003e","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7350138/v1/c5879d7e25a7814fb4f49a5f.jpeg"},{"id":90961807,"identity":"e189c778-ddc7-437c-b8d1-a8e648c1ee30","added_by":"auto","created_at":"2025-09-10 05:14:07","extension":"jpeg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":338954,"visible":true,"origin":"","legend":"\u003cp\u003e(a) Scatter plot of all SMI change measurements plotted against duration between initial and follow-up CT scan with distributions. (b) Distributions of SMI change split by recurrence and non-recurrence status with lines connecting paired participants from the propensity score analysis. (c) Differences in SMI changes between matched patients to evaluate the null hypothesis.\u003c/p\u003e","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7350138/v1/b7b78b67d9ca4a06cea015fb.jpeg"},{"id":91148958,"identity":"ffc651cd-4b2a-43e0-a804-a9fc926a4a59","added_by":"auto","created_at":"2025-09-12 06:46:18","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":766015,"visible":true,"origin":"","legend":"\u003cp\u003eAverage receiver operating characteristic curve of a recurrence status classification test based on the logistic regression model. A 10,000 sample bootstrap of ROC curves was also generated from the distributions in Figure 3b to assess for model stability in the setting of random sampling error. Individual simulated ROC curves are depicted as an overlapping density of thin gray lines.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7350138/v1/6f365a16b7e8f8e4d96168d9.png"},{"id":106809446,"identity":"df0f5cb7-667c-4659-9a8f-b61cbd8efba0","added_by":"auto","created_at":"2026-04-13 16:10:58","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2756901,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7350138/v1/e9a1e119-4689-4130-9984-d0ab6f2ff30c.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Temporal change in skeletal muscle index as a predictor of recurrence for patients with locally advanced colorectal malignancy: A retrospective cohort study","fulltext":[{"header":"Background","content":"\u003cp\u003eColorectal Cancer (CRC) remains one of the most common malignancies worldwide with a substantial risk of recurrence after oncologic resection. Current 5-year recurrence rates for stage III disease are ~33% with a median recurrence time of 15.9 months.\u003csup\u003e1,2\u003c/sup\u003e Detection of recurrence has significant prognostic implications as patients with isolated or oligometastatic disease may be eligible for curative interventions associated with improved long-term survival.\u003csup\u003e3\u003c/sup\u003e As such, identifying which patients are at high risk of recurrence is critical for optimizing postoperative surveillance strategies. Traditional risk stratification models focus on pathologic features including tumor stage, margin status and lymph node involvement\u003csup\u003e4-6\u003c/sup\u003e, but previous studies have demonstrated a correlation between frailty and postoperative outcomes\u0026nbsp;for patients who require oncologic resection.\u003csup\u003e7-15\u003c/sup\u003e In these studies, researchers employ various methods to assess patient frailty; however, no single clinical frailty score or surrogate marker has yet demonstrated clear superiority- an important limitation highlighted in a recent meta-analysis.\u003csup\u003e29\u003c/sup\u003e Existing tools such as the Fried Frailty Phenotype and the Rockwood Clinical Frailty Scale rely on subjective evaluations which are often highly variable and susceptible to individual bias and error.\u003csup\u003e16-17\u003c/sup\u003e Therefore, imaging-based sarcopenia assessment is increasingly seen as a viable solution to integrate frailty measurement into oncologic workflows while avoiding some of the limitations of clinical assessments. In this approach, a skeletal muscle area (SMA) measurement can be taken at the mid-level of the third lumbar vertebral body (L3) on axial computed tomography (CT) scans due to its strong correlation with whole-body muscle mass.\u003csup\u003e18\u003c/sup\u003e The skeletal muscle index (SMI) heuristic can then be generated by normalizing the SMA by the square of the patient’s height, and the progression of these SMI values over time can be tracked for postoperative surveillance. While sex-specific thresholds for SMI have been proposed in literature to define sarcopenia, optimal cut-off values may vary due to cancer type, patient population and other factors.\u003csup\u003e19\u003c/sup\u003e Furthermore, the body of longitudinal studies comparing temporal changes in SMI to long-term oncologic outcomes remains quite limited compared to the large number of cross sectional papers correlating outcomes to a single SMI datum.\u003csup\u003e30-31\u003c/sup\u003e As such, additional research is needed to validate a longitudinal approach to SMI surveillance and establish clinically meaningful cutoffs to guide the intensity of postoperative monitoring and adjuvant treatment. It is hypothesized that long-term decreases in SMI will be positively correlated with disease recurrence. Therefore, our study sought to better understand the time-dependent relationship between changes in skeletal muscle index and postoperative recurrence risk in patients with locally advanced colorectal malignancy.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eWe performed a single-institution (University of Vermont Medical Center) retrospective analysis that included patients aged 18 years or greater who underwent oncologic resection for locally advanced stage III colorectal malignancy between years 2000 and 2020. This study was approved by the University of Vermont Institutional Review Board (IRB #00002150) and was conducted in compliance with the Health Insurance Portability and Accountability Act (HIPAA). Researchers utilized the institution’s digital imaging PACS software to measure psoas, paraspinal and abdominal muscle cross sectional area for included patients at their date of initial diagnosis. Muscle regions were isolated by applying and manually refining a [-29, 150] Hounsfield Unit filter at the mid-vertebral body of the third lumbar vertebrae on axial CT scan (reference Figure 1).\u003csup\u003e20-23\u003c/sup\u003e Skeletal muscle indices were subsequently calculated as the quotient of this area and the square of the patient’s height.\u003csup\u003e19\u003c/sup\u003e An electronic medical record search was then performed to obtain demographic information, post-operative clinical outcomes data and confounding variables for included patients. The cohort was then split according to cancer recurrence status between 1 and 5 years following the initial diagnosis. For patients with recurrence, a follow-up SMI measurement was obtained at their recurrence date. For patients without recurrence, a comprehensive list of their CT study dates was collected, and the Python Propensity Score Matching (PSM) library (version 0.3.13) was then used to pair patients with and without recurrence.\u003csup\u003e24\u003c/sup\u003e The list of available CT study dates for patients in the non-recurrence category allowed for the duration between the first and second SMI measurements to be matched in addition to other specified confounders. Specifically, this was done by creating n copies of each non-recurrence patient in the database, populated with one of n unique CT study dates. Different follow-up durations could then be passed as a direct PSM covariate, and based on the indicated best-match, follow-up SMI measurements for non-recurrence patients could be obtained by reverse indexing the non-recurrence ID with its corresponding follow-up date. With both initial and follow-up SMI measurements for patients in the recurrence and non-recurrence categories, a temporal change in SMI could be calculated. The primary outcome measure of the study was the difference in these SMI changes between paired patients which were evaluated in a conditional logistic regression model using the Python Statsmodels package (version 0.14.4). The standard significance level of 0.05 was used for all hypothesis testing.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eA total of 519 patients were identified in the UVMMC database as undergoing surgical resection for colorectal malignancy between 2000 and 2020. Demographics from this sample population are presented in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. Chart review identified a subset of 292 patients with AJCC stage III colorectal cancer for whom additional confounding variables were collected. To control for factors that could independently influence recurrence or imaging frequency, we excluded a priori patients with concurrent malignancies, long-term steroid use, or significant comorbidity burden defined as having three or more conditions each with a Charlson Comorbidity Index weight of 1 (e.g., CHF, PAD, COPD, prior MI). Patients who did not undergo surgical resection were also excluded as surgery is central to curative treatment. After exclusions, 240 patients remained: 86 experienced recurrence between 1 and 5 years following resection (average 2.3 years), and 154 did not. An inclusion/exclusion criteria flowchart is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. For patients who did not experience recurrence, all archived CT study dates were obtained averaging 7 plausible scan dates per patient. Propensity scoring matched 36 patients using the standard caliper width of 0.2 standard deviations, and good balance (effect size\u0026thinsp;\u0026lt;\u0026thinsp;0.2) was obtained for all covariates. A sensitivity analysis evaluated the impact of varying caliper widths on the matching process with minimal observed variation in match rates. At a caliper width of 0.5, only 37 patients were matched, and a caliper of 0.2 was determined to be the narrowest possible width while maintaining adequate sample size. Formal power analysis was not conducted as the use of propensity score matching, by design, determines the final sample size based on covariate overlap and match quality. The reductions in effect size of the covariates before and after matching are outlined in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e and suggested overall that the matches were well balanced.\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eDemographics of the sampled patient population. ICD-O-3\u0026thinsp;=\u0026thinsp;International Classification of Diseases for Oncology, Third Edition. AJCC\u0026thinsp;=\u0026thinsp;American Joint Committee on Cancer.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"2\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDemographic\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eValue/Distribution\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMale/Female Ratio\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e308/211 (59%/41%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge at diagnosis (mean, stddev, range) in years\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eAll patients (65.8, 13.7, 32\u0026ndash;95), Male (64.4, 13.3, 32\u0026ndash;91), Female (67.9, 14.1, 32\u0026ndash;95)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePrimary Site\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eC209 Rectum, NOS (\u003cem\u003e24%\u003c/em\u003e), C180 Cecum (\u003cem\u003e15%\u003c/em\u003e), C182 Colon, ascending (\u003cem\u003e14%\u003c/em\u003e), C187 Colon, sigmoid (\u003cem\u003e11%\u003c/em\u003e), C184 Colon, transverse (\u003cem\u003e6%\u003c/em\u003e), C187 Sigmoid, NOS (\u003cem\u003e4%\u003c/em\u003e), C199 Rectosigmoid, NOS (\u003cem\u003e4%\u003c/em\u003e), C180 Ileocecal valve (\u003cem\u003e3%\u003c/em\u003e), C186 Colon, descending (\u003cem\u003e3%\u003c/em\u003e), C183 Colon, hepatic flexure (\u003cem\u003e3%\u003c/em\u003e), C199 Rectosigmoid colon (\u003cem\u003e3%\u003c/em\u003e), C185 Colon, splenic flexure (\u003cem\u003e3%\u003c/em\u003e), C199 Rectosigmoid junction (\u003cem\u003e3%\u003c/em\u003e), C188 Colon, overlapping lesion (\u003cem\u003e1%\u003c/em\u003e), C180 Ileocecal junction (\u003cem\u003e1%\u003c/em\u003e), C189 Colon, NOS (\u003cem\u003e1%\u003c/em\u003e), C182 Colon, right (\u003cem\u003e1%\u003c/em\u003e)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHisto/Behavior ICD-O-3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e81403 Adenocarcinoma, NOS (\u003cem\u003e64%\u003c/em\u003e), 84803 Mucinous adenocarcinoma (\u003cem\u003e10%\u003c/em\u003e), 82633 Adenocarcinoma in tubulovillous adenoma (\u003cem\u003e9%\u003c/em\u003e), 82103 Adenocarcinoma in tubular adenoma (\u003cem\u003e7%\u003c/em\u003e), 85103 Medullary adenocarcinoma (\u003cem\u003e2%\u003c/em\u003e), 82553 Adenocarcinoma with mixed subtypes (\u003cem\u003e1%\u003c/em\u003e), 82613 Adenocarcinoma in villous adenoma (\u003cem\u003e1%\u003c/em\u003e), 84903 Signet ring cell adenocarcinoma (\u003cem\u003e1%\u003c/em\u003e), 82103 Adenocarcinoma in adenomatous polyp (\u003cem\u003e1%\u003c/em\u003e), 85103 Medullary carcinoma, NOS (\u003cem\u003e1%\u003c/em\u003e), 82103 Adenocarcinoma in a polyp, NOS (\u003cem\u003e1%\u003c/em\u003e)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBest AJCC Stage\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eIIIB \u003cem\u003e(35%)\u003c/em\u003e, IIA \u003cem\u003e(32%)\u003c/em\u003e, IIIC \u003cem\u003e(16%)\u003c/em\u003e, IIB \u003cem\u003e(7%)\u003c/em\u003e, IIIA \u003cem\u003e(6%)\u003c/em\u003e, IIC \u003cem\u003e(4%)\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eEffect size of covariates before and after PSM matching.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"3\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eVariable\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eEffect Size (Before)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eEffect Size (After)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eScan Follow-Up Duration\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.433\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.009\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSmoking\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.686\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.117\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eChemotherapy\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.316\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.131\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRadiation\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.020\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.000\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHisto/Behavior ICD-O-3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.463\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.131\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBest AJCC Stage\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.198\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.123\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePrimary Cancer Site\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.645\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.155\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDiabetes Mellitus\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.292\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.181\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eThe CT imaging for matched recurrence and non-recurrence patients was evaluated and SMI data was collected and made publicly available in a Mendeley Data repository.\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e The temporal change in SMI was then calculated for each patient where positive changes indicated net muscle gain over time and negative changes indicated net muscle loss. A scatter plot of the change in SMI versus the duration between initial and final scan dates is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea along with underlying distributions. The change in SMI data exhibited high central tendency (mean = -0.89 cm\u003csup\u003e2\u003c/sup\u003em\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e, stddev\u0026thinsp;=\u0026thinsp;6.60 cm\u003csup\u003e2\u003c/sup\u003em\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e) while the scan durations were more diffusely distributed. The individual distributions of SMI changes as a function of recurrence status were plotted in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb with lines connecting the paired observations. The relationship between SMI change and cancer recurrence was assessed using a conditional logistic regression model that accounted for matched pair IDs with results shown in Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eConditional logistic regression on matched participants. Stratification variables generated from the matching condition are removed for readability.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"7\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eVariable\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eCoefficient\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eStd. Error\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003ez-value\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003ep-value\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003e95% CI Lower\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e\u003cp\u003e95% CI Upper\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSMI Change\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e-0.2589\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.089\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e-2.922\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.003\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e-0.433\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e-0.085\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eIntercept\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.7689\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e1.448\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.531\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.595\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e-2.069\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e3.607\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eThe logistic model indicated that each 1-point increase in SMI change was associated with a 23% decrease in the odds of recurrence (\u003cem\u003eOR\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.77, 95% CI: 0.65\u0026ndash;0.92, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.003). Alternatively, a 1-point decrease in SMI change was associated with a 30% increase in the odds of recurrence (\u003cem\u003eOR\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1.30) which are findings consistent with previous literature.\u003csup\u003e\u003cspan additionalcitationids=\"CR8 CR9 CR10 CR11 CR12 CR13 CR14\" citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e A likelihood ratio test comparing the conditional logistic model to a null model yielded a non-significant result (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;~\u0026thinsp;1.0), reflecting the quality of the initial propensity score matching which produced tightly matched pairs with limited residual variability. However, the association between SMI change and recurrence remained significant even after accounting for this matched structure suggesting that within matched pairs, differences in SMI change were consistently associated with recurrence status independent of other matched variables. Visually, this within-pair association is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec which displays the distribution of SMI change differences across matched pairs. To complement the regression, a Wilcoxon signed-rank test applied to this distribution yielded W\u0026thinsp;=\u0026thinsp;72.0 (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.014), further supporting a meaningful difference between SMI change and recurrence status consistent with the regression results. Given these results, the null hypothesis was rejected given the standard significance level of 5%.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eBased on this data, it may be possible to use SMI as a surveillance tool for patients who recently underwent oncologic resection to indicate potential cancer relapse. By using predicted probabilities from the logistic model, a receiver operating characteristic curve was generated to assess predictive performance. This curve is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e in addition to a bootstrap simulation of ROC curves calculated from randomly resampling from the distributions in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb. The purpose of the bootstrap was to validate the stability of the ROC curve generated from the logistic model, particularly in the event of an initial random sampling error.\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e The significant proportion of bootstrapped ROC curves (depicted as an overlapping density of thin gray lines) that performed better than chance suggests that the model\u0026rsquo;s discriminative ability is robust.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eFor such a test, sensitivity is typically prioritized, and a true positive rate of approximately 90% is commonly targeted in established oncology screenings.\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e In the ROC generated by the logistic model, the optimal classification threshold based on the Youden J statistic (J\u0026thinsp;=\u0026thinsp;0.60) was a predicted probability of 0.53. At this threshold, the model achieved a sensitivity of 80% and a specificity of 80% corresponding to an SMI change of 2.5 cm\u0026sup2;/m\u0026sup2; on average. While this threshold provides a balanced tradeoff between sensitivity and specificity, it may be suboptimal for a surveillance context where higher sensitivity is paramount. Arbitrarily setting the sensitivity to 90% as a target value, a specificity of 40% can be obtained with a threshold of 0.31 corresponding to an SMI change of 6.0 cm\u0026sup2;/m\u0026sup2;. Although this results in a higher false-positive rate, it maximizes the model\u0026rsquo;s ability to detect potential relapses and may serve as a clinically meaningful cutoff for early identification of high-risk individuals. Notably, the inflection point between net gain or loss of skeletal muscle (i.e SMI change\u0026thinsp;=\u0026thinsp;0 cm\u0026sup2;/m\u0026sup2;) occurred at a threshold value of 0.68 where the model returned a sensitivity of 41% and specificity of 84%. This threshold may be more appropriate for confirming absence of recurrence in lower-risk patients, but the low sensitivity ultimately limits utility in primary surveillance. Taken together, these results suggest that an SMI decline in the range of 2.5\u0026ndash;6.0 cm\u0026sup2;/m\u0026sup2; within 1\u0026ndash;5 years following resection may represent a practical and clinically actionable threshold for increased surveillance. While prospective validation is warranted, such a cutoff could help identify patients who would benefit from intensified follow-up, imaging, or early intervention strategies, particularly if commensurate with additional surveillance tests.\u003c/p\u003e\u003cp\u003eWhile a standalone cutoff may be sufficient for binary risk classification, the more robust approach involves building comprehensive models that incorporate a larger set of validated pathologic risk factors. However, expanding the input space of such models introduces challenges, particularly the need for increasingly large datasets to support reliable parameter estimation. The temporal nature of SMI change lends itself to Bayesian updating where new, independently associated risk factors can be considered enabling posterior probabilities to reflect additional evidence without requiring the construction of a single unified model.\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e This is particularly useful when different datasets contain complementary but non-overlapping features making joint model training impossible. In practical terms, this means the prior probability of recurrence (obtained from an established model) can be updated using a likelihood ratio derived from SMI change. We assume conditional independence between SMI and the prior probability given recurrence status, which is reasonable since SMI is measured independently of the pathologic features often used in existing risk models. Therefore, combining models in this manner does not introduce artificial bias.\u003c/p\u003e\u003cp\u003eOne attractive model for a prior was developed by Zafar et al at MD Anderson Cancer Center and uses a Cox regression to estimate 5-year recurrence probabilities for colorectal malignancy.\u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e It considers risk factors such as age, sex, tumor histology, site, margin status and stage, but excludes any weight-based covariates which would be highly correlated with SMI change. This is best practice for ensuring conditional independence of covariates without access to the previous model\u0026rsquo;s underlying coefficients. Formally, this model outputs the probability of recurrence given the data θ, P(recurrence|θ), which is easily converted to odds via Eq.\u0026nbsp;1. This conversion enables use of the more convenient form of Bayes Theorem in odds form.\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e Then using the logistic regression coefficients from Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e the log-odds of recurrence based on SMI change may be calculated. The likelihood ratio is the quotient of these odds and the marginal odds of recurrence from the UVMMC dataset as shown in Eq.\u0026nbsp;2 along with a numerical approximation for convenience. The posterior odds of recurrence given the data θ \u003cem\u003eand\u003c/em\u003e ΔSMI, O(recurrence|{θ, ΔSMI}), are then calculated as the product of the prior and the likelihood ratio as shown in Eq.\u0026nbsp;3. These odds may ultimately be converted back to a recurrence probability using Eq.\u0026nbsp;1.\u003c/p\u003e\u003cp\u003eEq1: \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:O\\left(A\\right)=\\frac{P\\left(A\\right)}{1-P\\left(A\\right)},\\:P\\left(A\\right)=\\frac{O\\left(A\\right)}{1+O\\left(A\\right)}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp\u003eEq2: \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:LR\\left(Recurrence|\\varDelta\\:SMI\\right)=\\frac{O\\left(Recurrence|\\varDelta\\:SMI\\right)}{O\\left(Recurrence\\right)}=\\frac{exp\\left({\\beta\\:}_{0}+{\\beta\\:}_{1}\\varDelta\\:SMI\\right)}{O\\left(Recurrence\\right)}\\approx\\:6{e}^{-0.26*\\varDelta\\:SMI}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp\u003eEq3: \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:O\\left(Recurrence\\left|\\right\\{\\varDelta\\:SMI,\\theta\\:\\}\\right)=O\\left(Recurrence|\\theta\\:\\right)\\:*LR\\left(Recurrence|\\varDelta\\:SMI\\right)\\:\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\u003cp\u003eWhile Bayesian updating provides a coherent framework for incorporating new information, it does not constitute a fully specified or causally rigorous model. Nonetheless, it is well suited to situations where additional data such as changes in SMI emerge after the initial risk assessment. The advantage of this approach is that it offers an actionable way to begin incorporating SMI change into oncologic workflows, rather than simply being aware that a risk-relationship exists. The other advantage of the likelihood ratio is to compute the threshold for unity at which no extra information is added to the prior, or ~\u0026thinsp;6.9 cm\u0026sup2;/m\u0026sup2; corresponding to LR\u0026thinsp;=\u0026thinsp;1. This is consistent with the results from the receiver operating characteristic and suggests that SMI change overall may be more apt for detecting high-risk individuals below the threshold than for confidently ruling out recurrence in low-risk individuals above it. Ideally, Bayesian updating would be assessed by evaluating the posterior against known outcomes within a dataset fully spanning the input spaces of both the prior and the updating variable. In practice however, such comprehensive datasets are often unavailable, and while strict adherence to a unified likelihood framework may be statistically preferable, it is not always practical particularly in clinical research where data heterogeneity is common. This method strikes a balance between absolute statistical rigor and the practical necessity of leveraging available information.\u003c/p\u003e\u003cp\u003eWhile every effort was taken to ensure methodological rigor and minimize bias, certain limitations of the current work are present. First, the sample size that was passed into the logistic model was relatively small despite a large initial dataset, limiting the overall study power. This was a natural consequence of the strict inclusion/exclusion criteria and number of covariates that were controlled for in the propensity score matching. Furthermore, and as a direct result of the limited size of the dataset, not all covariates could be adequately controlled for, and the matching process was limited to variables for which great balance (effect size\u0026thinsp;\u0026lt;\u0026thinsp;0.2) could be obtained. While this improved the overall quality of the logistic model for the limited covariates that were included, it is not clear what impact other comorbidities (e.g neoadjuvant vs. adjuvant treatment) would have had. Finally, data was collected from a single institution, introducing a potential selection bias. Future studies with larger, multi-center cohorts and more comprehensive data collection are warranted to validate and expand upon these findings.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eAssessing frailty is a crucial component to providing optimum multidisciplinary care for patients with colorectal cancer who require surgical resection. Imaging-based sarcopenia assessment is increasingly seen as a viable solution to integrate frailty measurement into oncologic workflows, though standardized surveillance protocols have not been established. Furthermore, while previous literature has demonstrated a correlation between frailty and recurrence status,\u003csup\u003e\u003cspan additionalcitationids=\"CR8 CR9 CR10 CR11 CR12 CR13 CR14\" citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e the body of longitudinal studies comparing temporal changes in SMI to long-term oncologic outcomes remains quite limited.\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e Our results compile data from a single-institution study of over 500 patients with colorectal malignancy over a 20 year period to estimate threshold values for surveillance risk stratification. These data show a positive correlation between skeletal muscle loss and disease recurrence for patients with CRC where a 1 cm\u0026sup2;/m\u0026sup2; decline in skeletal muscle index 1\u0026ndash;5 years after resection is associated with a 30% increase in the odds of recurrence. Based on the ROC curve generated from regression analysis, an SMI decline in the range of 2.5\u0026ndash;6.0 cm\u0026sup2;/m\u0026sup2; within 1\u0026ndash;5 years following resection may represent a practical and clinically actionable threshold for increased surveillance. These findings support the role of body composition monitoring to guide postoperative management and highlight a need for further prospective validation.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cdiv class=\"DefinitionList\"\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eCRC\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eColorectal cancer\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eSMI\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eSkeletal muscle index\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eSMA\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eSkeletal muscle area\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eCT\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eComputed tomography\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eROC\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eReceiver operating characteristic\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003ePSM\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003ePropensity score matching\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eAJCC\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eAmerican Joint Committee on Cancer\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eUVMMC\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eUniversity of Vermont Medical Center\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics Approval and Consent to Participate:\u0026nbsp;\u003c/strong\u003eThis study was conducted in accordance with the Declaration of Helsinki and approved by the University of Vermont Institutional Review Board (IRB #00002150); Date of approval: December 1, 2023.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for Publication:\u0026nbsp;\u003c/strong\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of Data and Materials:\u003c/strong\u003e The datasets generated for this study are publicly available in the Mendeley Data repository, https://data.mendeley.com/datasets/cxyjw6v4gh/1.\u003csup\u003e26\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting Interests:\u003c/strong\u003e The authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding:\u003c/strong\u003e Funding was provided by the University of Vermont Larner College of Medicine Department of Radiology Research Fund.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026rsquo; Contributions:\u0026nbsp;\u003c/strong\u003eNM developed the methodology, implemented the software, performed validation and formal analysis, conducted investigation and data curation, drafted the original manuscript, and created the visualizations. JR contributed to conceptualization, methodology, investigation, resource provision, data curation, manuscript review, and project administration. CK led conceptualization, methodology, validation, investigation, resource management, data curation, manuscript review, supervision, and project administration. JD contributed to conceptualization, methodology, validation, resources, manuscript review, supervision, and project administration. All authors read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements:\u0026nbsp;\u003c/strong\u003eWe would like to thank Tom Ahern, PhD, for his expert guidance and contributions to the statistical methodology of this study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026rsquo; Information:\u003c/strong\u003e Not applicable.\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eOsterman E, Glimelius B. Recurrence Risk After Up-to-Date Colon Cancer Staging, Surgery, and Pathology: Analysis of the Entire Swedish Population. Dis Colon Rectum. 2018;61(9):1016\u0026ndash;25. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/dcr.0000000000001158\u003c/span\u003e\u003cspan address=\"10.1097/dcr.0000000000001158\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNors J, Iversen LH, Erichsen R, Gotschalck KA, Andersen CL. Incidence of Recurrence and Time to Recurrence in Stage I to III Colorectal Cancer. JAMA Oncol. 2024;10(1):54. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1001/jamaoncol.2023.5098\u003c/span\u003e\u003cspan address=\"10.1001/jamaoncol.2023.5098\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZucchelli G, Moretto R, Rossini D, et al. Oligometastatic colorectal cancer: Prognostic implications of tumor load, role of locoregional treatments, and of first-line therapy intensification\u0026mdash;A pooled analysis of TRIBE and TRIBE2 studies by GONO. JCO. 2020;38(4suppl):12\u0026ndash;12. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1200/jco.2020.38.4_suppl.12\u003c/span\u003e\u003cspan address=\"10.1200/jco.2020.38.4_suppl.12\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eEdge SB, Compton CC. The American Joint Committee on Cancer: the 7th Edition of the AJCC Cancer Staging Manual and the Future of TNM. Ann Surg Oncol. 2010;17(6):1471\u0026ndash;4. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1245/s10434-010-0985-4\u003c/span\u003e\u003cspan address=\"10.1245/s10434-010-0985-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNagtegaal ID, Quirke P, Schmoll H-J. Has the new TNM classification for colorectal cancer improved care? Nat Rev Clin Oncol. 2011;9(2):119\u0026ndash;23. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/nrclinonc.2011.157\u003c/span\u003e\u003cspan address=\"10.1038/nrclinonc.2011.157\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eOng MLH, Schofield JB. Assessment of lymph node involvement in colorectal cancer. WJGS. 2016;8(3):179. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.4240/wjgs.v8.i3.179\u003c/span\u003e\u003cspan address=\"10.4240/wjgs.v8.i3.179\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGartrell R, Qiao J, Kiss N, et al. Can sarcopenia predict survival in locally advanced rectal cancer patients? ANZ J Surg. 2023;93(9):2166\u0026ndash;71. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/ans.18512\u003c/span\u003e\u003cspan address=\"10.1111/ans.18512\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHopkins JJ, Reif RL, Bigam DL, Baracos VE, Eurich DT, Sawyer MB. The Impact of Muscle and Adipose Tissue on Long-term Survival in Patients With Stage I to III Colorectal Cancer. Dis Colon Rectum. 2019;62(5):549\u0026ndash;60. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/dcr.0000000000001352\u003c/span\u003e\u003cspan address=\"10.1097/dcr.0000000000001352\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHorie K, Matsuda T, Yamashita K, et al. Sarcopenia assessed by skeletal muscle mass volume is a prognostic factor for oncological outcomes of rectal cancer patients undergoing neoadjuvant chemoradiotherapy followed by surgery. Eur J Surg Oncol. 2022;48(4):850\u0026ndash;6. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.ejso.2021.10.018\u003c/span\u003e\u003cspan address=\"10.1016/j.ejso.2021.10.018\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePortale G, Zuin M, Spolverato YC, et al. Prognostic effect of sarcopenia in patients undergoing laparoscopic rectal cancer resection. ANZ J Surg. 2023;93(6):1631\u0026ndash;7. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/ans.18269\u003c/span\u003e\u003cspan address=\"10.1111/ans.18269\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSaur NM, Davis BR, Montroni I, et al. The American Society of Colon and Rectal Surgeons Clinical Practice Guidelines for the Perioperative Evaluation and Management of Frailty Among Older Adults Undergoing Colorectal Surgery. Dis Colon Rectum. 2022;65(4):473\u0026ndash;88. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/dcr.0000000000002410\u003c/span\u003e\u003cspan address=\"10.1097/dcr.0000000000002410\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eXiao J, Caan BJ, Cespedes Feliciano EM, et al. Association of Low Muscle Mass and Low Muscle Radiodensity With Morbidity and Mortality for Colon Cancer Surgery. JAMA Surg. 2020;155(10):942. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1001/jamasurg.2020.2497\u003c/span\u003e\u003cspan address=\"10.1001/jamasurg.2020.2497\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBenedek Z, Coroș MF. The impact of sarcopenia on the postoperative outcome in colorectal cancer surgery. Med Pharm Rep. 2023;96(1):20\u0026ndash;7. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.15386/mpr-2483\u003c/span\u003e\u003cspan address=\"10.15386/mpr-2483\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFleming CA, O\u0026rsquo;Connell EP, Kavanagh RG, et al. Body Composition, Inflammation, and 5-Year Outcomes in Colon Cancer. JAMA Netw Open. 2021;4(8):e2115274. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1001/jamanetworkopen.2021.15274\u003c/span\u003e\u003cspan address=\"10.1001/jamanetworkopen.2021.15274\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSibia US, Badve SB, Istl AC, Klune JR, Riker AI. Impact of Frailty Upon Surgical Decision-Making for Left-Sided Colon Cancer. TOJ. 2023;23(2):120\u0026ndash;8. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.31486/toj.22.0120\u003c/span\u003e\u003cspan address=\"10.31486/toj.22.0120\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFried LP, Tangen CM, Walston J et al. Frailty in Older Adults: Evidence for a Phenotype. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences. 2001;56(3):M146\u0026ndash;57. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/gerona/56.3.m146\u003c/span\u003e\u003cspan address=\"10.1093/gerona/56.3.m146\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRockwood K. A global clinical measure of fitness and frailty in elderly people. Can Med Assoc J. 2005;173(5):489\u0026ndash;95. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1503/cmaj.050051\u003c/span\u003e\u003cspan address=\"10.1503/cmaj.050051\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003en den Broeck J, Sealy MJ, Brussaard C, Kooijman J, Jager-Wittenaar H, Scafoglieri A. The correlation of muscle quantity and quality between all vertebra levels and level L3, measured with CT: An exploratory study. Front Nutr. 2023;10. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fnut.2023.1148809\u003c/span\u003e\u003cspan address=\"10.3389/fnut.2023.1148809\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFreire PP, Fernandez GJ, de Moraes D, et al. The expression landscape of cachexia-inducing factors in human cancers. J cachexia sarcopenia muscle. 2020;11(4):947\u0026ndash;61. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/jcsm.12565\u003c/span\u003e\u003cspan address=\"10.1002/jcsm.12565\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDerstine BA, Holcombe SA, Ross BE, Wang NC, Su GL, Wang SC. Skeletal muscle cutoff values for sarcopenia diagnosis using T10 to L5 measurements in a healthy US population. Sci Rep. 2018;8(1). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41598-018-29825-5\u003c/span\u003e\u003cspan address=\"10.1038/s41598-018-29825-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAubrey J, Esfandiari N, Baracos VE, et al. Measurement of skeletal muscle radiation attenuation and basis of its biological variation. Acta Physiol. 2014;210(3):489\u0026ndash;97. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/apha.12224\u003c/span\u003e\u003cspan address=\"10.1111/apha.12224\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDerstine BA, Holcombe SA, Goulson RL, et al. Quantifying Sarcopenia Reference Values Using Lumbar and Thoracic Muscle Areas in a Healthy Population. J Nutr health aging. 2018;22(1):180\u0026ndash;5. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s12603-017-0983-3\u003c/span\u003e\u003cspan address=\"10.1007/s12603-017-0983-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003en der Werf A, Dekker IM, Meijerink MR, de van der Wierdsma NJ, Langius JAE. Skeletal muscle analyses: agreement between non-contrast and contrast \u0026lt; scp \u0026gt;CT scan measurements of skeletal muscle area and mean muscle attenuation. Clin Physio Funct Imaging. 2017;38(3):366\u0026ndash;72. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/cpf.12422\u003c/span\u003e\u003cspan address=\"10.1111/cpf.12422\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKline A, Luo Y, PsmPy:. A Package for Retrospective Cohort Matching in Python. In: 2022 44th Annual International Conference of the IEEE Engineering in Medicine \u0026amp; Biology Society (EMBC). IEEE; 2022:1354\u0026ndash;1357. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1109/embc48229.2022.9871333\u003c/span\u003e\u003cspan address=\"10.1109/embc48229.2022.9871333\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAustin PC. An Introduction to Propensity Score Methods for Reducing the Effects of Confounding in Observational Studies. Multivar Behav Res. 2011;46(3):399\u0026ndash;424. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1080/00273171.2011.568786\u003c/span\u003e\u003cspan address=\"10.1080/00273171.2011.568786\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eManz N, Reed J, Kanner C, Dressler J. Temporal changes in skeletal muscle index dataset for colorectal malignancy recurrence study. Mendeley Data. 2025;V1. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.17632/cxyjw6v4gh.1\u003c/span\u003e\u003cspan address=\"10.17632/cxyjw6v4gh.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGu J, Ghosal S, Roy A. Bayesian bootstrap estimation of ROC curve. Stat Med. 2008;27(26):5407\u0026ndash;20. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/sim.3366\u003c/span\u003e\u003cspan address=\"10.1002/sim.3366\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGupta S. Screening for Colorectal Cancer. Hematol Oncol Clin N Am. 2022;36(3):393\u0026ndash;414. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.hoc.2022.02.001\u003c/span\u003e\u003cspan address=\"10.1016/j.hoc.2022.02.001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBhamani A, Creamer A, Verghese P, et al. Low-dose CT for lung cancer screening in a high-risk population (SUMMIT): a prospective, longitudinal cohort study. Lancet Oncol. 2025;26(5):609\u0026ndash;19. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/s1470-2045(25)00082-8\u003c/span\u003e\u003cspan address=\"10.1016/s1470-2045(25)00082-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHan J, Zhang Q, Lan J, Yu F, Liu J. Frailty worsens long-term survival in patients with colorectal cancer: a systematic review and meta-analysis. Front Oncol. 2024;14. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fonc.2024.1326292\u003c/span\u003e\u003cspan address=\"10.3389/fonc.2024.1326292\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHopkins JJ, Reif R, Bigam D, Baracos VE, Eurich DT, Sawyer MM. Change in Skeletal Muscle Following Resection of Stage I\u0026ndash;III Colorectal Cancer is Predictive of Poor Survival: A Cohort Study. World j surg. 2019;43(10):2518\u0026ndash;26. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00268-019-05054-3\u003c/span\u003e\u003cspan address=\"10.1007/s00268-019-05054-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDeng C-Y, Lin Y-C, Wu JS, et al. Progressive Sarcopenia in Patients With Colorectal Cancer Predicts Survival. Am J Roentgenol. 2018;210(3):526\u0026ndash;32. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2214/ajr.17.18020\u003c/span\u003e\u003cspan address=\"10.2214/ajr.17.18020\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSpiegelhalter DJ, Abrams KR, Myles JP. Bayesian Approaches to Clinical Trials and Health-Care Evaluation. Published online Dec. 2003;9. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/0470092602\u003c/span\u003e\u003cspan address=\"10.1002/0470092602\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZafar SN, Hu C-Y, Snyder RA, et al. Predicting Risk of Recurrence After Colorectal Cancer Surgery in the United States: An Analysis of a Special Commission on Cancer National Study. Ann Surg Oncol. 2020;27(8):2740\u0026ndash;9. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1245/s10434-020-08238-7\u003c/span\u003e\u003cspan address=\"10.1245/s10434-020-08238-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"cancer-imaging","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"caig","sideBox":"Learn more about [Cancer Imaging](https://cancerimagingjournal.biomedcentral.com/)","snPcode":"40644","submissionUrl":"https://submission.nature.com/new-submission/40644/3","title":"Cancer Imaging","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Skeletal Muscle Index, Sarcopenia, Colorectal Cancer Recurrence","lastPublishedDoi":"10.21203/rs.3.rs-7350138/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7350138/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground: \u003c/strong\u003eColorectal cancer remains a leading cause of cancer-related morbidity and mortality with stage III disease carrying a substantial risk of recurrence despite curative resection. Accurate risk stratification is essential to optimize surveillance and guide adjuvant therapy. Traditional models rely heavily on pathologic features, but recent studies suggest that body composition metrics, particularly imaging-based assessments of skeletal muscle mass, may offer additional prognostic value. The skeletal muscle index, derived from routine CT imaging, has emerged as a promising surrogate marker of frailty. However, the relationship between temporal changes in SMI and cancer recurrence remains poorly defined.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods:\u003c/strong\u003e A retrospective cohort study was performed using single-institution data from over 500 patients aged 18 or greater who underwent resection for locally advanced stage III colorectal malignancy between years 2000 and 2020. Skeletal muscle index was measured at the third lumbar vertebral level using preoperative CT imaging from the time of initial diagnosis. For patients with recurrence, a follow-up measurement was obtained at their recurrence date, and using propensity score analysis, a patient without recurrence was selected to best match this follow-up duration. Temporal changes in skeletal muscle index were then calculated and compared with a conditional logistic regression. Receiver operating characteristics were then analyzed to identify clinically relevant thresholds of SMI decline.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults:\u003c/strong\u003e A decrease in skeletal muscle index was independently associated with increased risk of disease recurrence with an odds ratio of 1.30 per 1-point decrease in SMI (95% CI: 1.09–1.54, \u003cem\u003ep\u003c/em\u003e = 0.003). Receiver operating curve analysis suggests that an SMI decline of 2.5–6.0 cm²/m² within 1–5 years after resection may serve as a practical threshold for risk stratification.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusions:\u003c/strong\u003ePostoperative skeletal muscle index decline is a significant, independent predictor of cancer recurrence and may serve as a clinically useful marker to personalize follow-up care and enhance risk stratification. These findings support the potential role of body composition monitoring to guide postoperative management and highlight a need for further prospective validation.\u003c/p\u003e","manuscriptTitle":"Temporal change in skeletal muscle index as a predictor of recurrence for patients with locally advanced colorectal malignancy: A retrospective cohort study","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-09-10 05:14:02","doi":"10.21203/rs.3.rs-7350138/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-12-09T14:14:38+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-11-03T10:18:03+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"299412542868710115761296131805517374644","date":"2025-10-08T05:59:52+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-09-17T12:26:31+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"337018062195209847091429201390952766631","date":"2025-09-08T12:07:19+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-09-02T09:11:39+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-08-13T12:41:42+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-08-13T10:19:08+00:00","index":"","fulltext":""},{"type":"submitted","content":"Cancer Imaging","date":"2025-08-12T00:56:32+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"cancer-imaging","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"caig","sideBox":"Learn more about [Cancer Imaging](https://cancerimagingjournal.biomedcentral.com/)","snPcode":"40644","submissionUrl":"https://submission.nature.com/new-submission/40644/3","title":"Cancer Imaging","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"4809e3f5-682d-4c2f-8008-c2eb9767ab2a","owner":[],"postedDate":"September 10th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2026-04-13T16:06:52+00:00","versionOfRecord":{"articleIdentity":"rs-7350138","link":"https://doi.org/10.1186/s40644-026-01025-9","journal":{"identity":"cancer-imaging","isVorOnly":false,"title":"Cancer Imaging"},"publishedOn":"2026-04-07 15:57:55","publishedOnDateReadable":"April 7th, 2026"},"versionCreatedAt":"2025-09-10 05:14:02","video":"","vorDoi":"10.1186/s40644-026-01025-9","vorDoiUrl":"https://doi.org/10.1186/s40644-026-01025-9","workflowStages":[]},"version":"v1","identity":"rs-7350138","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7350138","identity":"rs-7350138","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}