Factors associated with variability in skeletal muscle radiodensity in patients with metastatic cancer

OBJECTIVES

This study aimed to explore factors associated with skeletal muscle radiodensity (SMD) variability in patients with metastatic cancer.

METHODS

This study included 393 patients (median age 61 y, 70% women) who had computed tomography (CT) scans within 30 days of inclusion in the study. SMD was evaluated from CT by averaging the Hounsfield unit value of the total muscle area. Skeletal muscle index (SMI), visceral adipose tissue index (VATI), subcutaneous adipose tissue index (SATI), and total adipose tissue index (TATI) were also assessed by CT. Additionally, age, sex, race/skin color, disease characteristics, comorbidities, inflammatory markers, handgrip strength (HGS), and body mass index (BMI) were recorded and evaluated in the linear regression analysis to identify factors associated with SMD variability.

RESULTS

Multivariate explanatory models having SMD as an independent variable were performed and included BMI (model 1, r2 = 0.699), TATI (model 2, r2 = 0.712) or VATI and SATI (model 3, r2 = 0.706) in addition to age, race/skin color, tumor site, kidney disease, serum albumin, HGS, and SMI as dependent variables. For all models, lower SMD was associated with higher age, BMI, and adiposity measurements, kidney disease, White race/skin color, and lower serum albumin, HGS, and SMI. The primary tumor site also contributed to changes in SMD in all models, specifically those located in the gastrointestinal tract, gynecologic, and bone and connective tissue.

CONCLUSION

In this group of patients with metastatic cancer, lower SMD was associated with older age, White race/skin color, and an overall worse clinical condition.

Establishment and validation of novel nomograms to predict muscle quality in colorectal cancer patients

OBJECTIVES

The skeletal muscle mass index and skeletal muscle radiodensity have promise as specific diagnostic indicators for muscle quality. However, the difficulties in measuring low skeletal muscle mass index and low skeletal muscle radiodensity limit their use in routine clinical practice, impeding early screening and diagnosis. The objective of this study is to develop a nomogram that incorporates preoperative factors for predicting low skeletal muscle mass index and low skeletal muscle radiodensity.

METHODS

A total of 1692 colorectal cancer patients between 2015 and 2021 were included. The patients were randomly divided into a training cohort (n = 1353) and a validation cohort (n = 339). Nomogram models were calibrated using the area under the curve, calibration curves, and the Hosmer-Lemeshow test to assess their predictive ability. Finally, a decision curve was applied to assess the clinical usefulness.

RESULTS

In a prediction model for low skeletal muscle mass index, age, body mass index, and grip strength were incorporated as variables. For low skeletal muscle radiodensity, age, sex, body mass index, serum hemoglobin level, and grip strength were included as predictors. In the training cohort, the area under the curve value for low skeletal muscle mass index was 0.750 (95% CI, 0.726-0.773), whereas for low skeletal muscle radiodensity, it was 0.763 (95% CI, 0.739-0.785). The Hosmer-Lemeshow test confirmed that both models fit well in both cohorts. Decision curve analysis was applied to assess the clinical usefulness of the model.

CONCLUSIONS

The incorporation of preoperative factors into the nomogram-based prediction model represents a significant advancement in the muscle quality assessment. Its implementation has the potential to early screen patients at risk of low skeletal muscle mass index and low skeletal muscle radiodensity.

Skeletal muscle gauge prediction by a machine learning model in patients with colorectal cancer

OBJECTIVES

Skeletal muscle gauge (SMG) was recently introduced as an imaging indicator of sarcopenia. Computed tomography is essential for measuring SMG; thus, the use of SMG is limited to patients who undergo computed tomography. We aimed to develop a machine learning algorithm using clinical and inflammatory markers to predict SMG in patients with colorectal cancer.

METHODS

The least absolute shrinkage and selection operator regression model was applied for variable selection and predictive signature building in the training set. The predictive accuracy of the least absolute shrinkage and selection operator model, defined as linear predictor (LP)-SMG, was compared using the area under the receiver operating characteristic curve and decision curve analysis in the test set.

RESULTS

A total of 1094 patients with colorectal cancer were enrolled and randomly categorized into training (n = 656) and test (n = 438) sets. Low SMG was identified in 142 (21.6%) and 90 (20.5%) patients in the training and test sets, respectively. According to multivariable analysis of the test sets, LP-SMG was identified as an independent predictor of low SMG (odds ratio = 1329.431; 95% CI, 271.684-7667.996; P < .001). Its predictive performance was similar in the training and test sets (area under the receiver operating characteristic curve = 0.846 versus 0.869; P = .427). In the test set, LP-SMG had better outcomes in predicting SMG than single clinical variables, such as sex, height, weight, and hemoglobin.

CONCLUSIONS

LP-SMG had superior performance than single variables in predicting low SMG. This machine learning model can be used as a screening tool to detect sarcopenic status without using computed tomography during the treatment period.

Construction of a nomogram for predicting COVID-19 in-hospital mortality: A machine learning analysis

Background and objectives

We aim to verify the use of ML algorithms to predict patient outcome using a relatively small dataset and to create a nomogram to assess in-hospital mortality of patients with COVID-19.

Methods

A database of 200 COVID-19 patients admitted to the Clinical Hospital of State University of Campinas (UNICAMP) was used in this analysis. Patient features were divided into three categories: clinical, chest abnormalities, and body composition characteristics acquired by computerized tomography. These features were evaluated independently and combined to predict patient outcomes. To minimize performance fluctuations due to low sample number, reduce possible bias related to outliers, and evaluate the uncertainties generated by the small dataset, we developed a shuffling technique, a modified version of the Monte Carlo Cross Validation, creating several subgroups for training the algorithm and complementary testing subgroups. The following ML algorithms were tested: random forest, boosted decision trees, logistic regression, support vector machines, and neural networks. Performance was evaluated by analyzing Receiver operating characteristic (ROC) curves. The importance of each feature in the determination of the outcome predictability was also studied and a nomogram was created based on the most important features selected by the exclusion test.

Results

Among the different sets of features, clinical variables age, lymphocyte number and weight were the most valuable features for prognosis prediction. However, we observed that skeletal muscle radiodensity and presence of pleural effusion were also important for outcome determination. Integrating these independent predictors was successfully developed to accurately predict mortality in COVID-19 in hospital patients. A nomogram based on these five features was created to predict COVID-19 mortality in hospitalized patients. The area under the ROC curve was 0.86 ± 0.04.

Conclusion

ML algorithms can be reliable for the prediction of COVID-19-related in-hospital mortality, even when using a relatively small dataset. The success of ML techniques in smaller datasets broadens the applicability of these methods in several problems in the medical area. In addition, feature importance analysis allowed us to determine the most important variables for the prediction tasks resulting in a nomogram with good accuracy and clinical utility in predicting COVID-19 in-hospital mortality.

Phase angle as a marker for muscle abnormalities and function in patients with colorectal cancer

BACKGROUND AND AIMS

Considering the applicability of phase angle (PhA) as a marker of muscle mass and function, we aimed to investigate whether PhA is a predictor of muscle abnormalities and function in patients with cancer.

METHODS

In a sample of patients with colorectal cancer (CRC), PhA was obtained from measurements of resistance and reactance from bioelectrical impedance analysis. Computerized tomography imaging at the third lumbar vertebra was used to evaluate muscle abnormalities by quantifying skeletal muscle index (SMI) and skeletal muscle radiodensity (SMD). Muscle function was assessed by handgrip strength (HGS) and gait speed (GS).

RESULTS

This cross-sectional study included 190 participants (X±SD), mean age 60.5 ± 11.3 years; 57% men; 78% had cancer stages III to IV. PhA was highly correlated with SMI (r = 0.70) and moderately correlated with HGS (r = 0.54). PhA explained 48% of the SMI variability (R² = 0.485), 21% of the SMD variability (R² = 0.214), 26% of HGS (R² = 0.261) and 9.8% of GS (R² = 0.098). In the multivariate model adjusted for age, sex, body mass index, performance status, comorbidities and cancer stage, 1-degree decrease in PhA was associated with low SMI (Odds Ratio (OR) = 6.56, 95% CI: 2.90-14.86) and with low SMI and HGS combined (OR = 11.10, 95% CI: 2.61-47.25). In addition, Receiving Operating Characteristics curve analysis showed that PhA had a good diagnostic accuracy for detecting low SMI, low SMI and SMD combined, low SMD and HGS and low SMI and HGS combined (AUC = 0.81, 95% CI: 0.74-0.88; AUC = 0.88, 95% CI: 0.81-0.95; AUC = 0.80, 95% CI: 0.70-0.91; AUC = 0.82, 95% CI: 0.74-0.89; respectively).

CONCLUSIONS

PhA was a predictor of muscle abnormalities and function and had a good diagnostic accuracy for detecting low muscle mass, low muscle mass and radiodensity, low muscle radiodensity and strength, and low muscle mass and strength in patients with CRC.

Body composition and its association with fatigue in the first 2 years after colorectal cancer diagnosis

Purpose

Persistent fatigue among colorectal cancer (CRC) patients might be associated with unfavorable body composition, but data are sparse and inconsistent. We studied how skeletal muscle index (SMI), skeletal muscle radiodensity (SMR), visceral adipose tissue (VAT), and subcutaneous adipose tissue (SAT) at diagnosis are associated with fatigue up to 24 months post-diagnosis in stage I–III CRC patients.

Methods

SMI, SMR, VAT, and SAT were assessed among 646 CRC patients using pre-treatment computed tomography images. Fatigue at diagnosis, at 6, and 24 months post-diagnosis was assessed using the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire. The association of SMI, SMR, VAT, and SAT with fatigue (yes/no) was assessed using confounder-adjusted restricted cubic spline analyses.

Results

Prevalence of fatigue at diagnosis was 18%, at 6 months 25%, and at 24 months 12%. At diagnosis, a significant (p = 0.01) non-linear association of higher levels of SAT with higher prevalence of fatigue was observed. Lower levels of SMR were linearly associated with higher prevalence of fatigue at 6 months post-diagnosis (overall association p = 0.02). None of the body composition parameters were significantly associated with fatigue at 24 months.

Conclusion

Having more SAT was associated with more fatigue at diagnosis, while low levels of SMR were associated with more fatigue at 6 months post-diagnosis.

Implications for Cancer Survivors

Our results suggest that it may be interesting to investigate whether interventions that aim to increase SMR around the time of diagnosis may help to lower fatigue. However, more knowledge is needed to understand the mechanisms behind the association of SMR with fatigue.

Electronic supplementary material

The online version of this article (10.1007/s11764-020-00953-0) contains supplementary material, which is available to authorized users.

Associations between muscle measures, survival, and toxicity in patients with limited stage small cell lung cancer

BACKGROUND

Standard treatment for patients with limited stage small cell lung cancer (LS SCLC) is concurrent platinum-etoposide chemotherapy and thoracic radiotherapy (TRT). Up to 30% of patients are cured, but severe toxicity is common, and we are not able to identify those who are cured or those who experience severe toxicity before chemoradiotherapy commences. Studies of other cancer patients show that low muscle mass and muscle radiodensity are associated with inferior survival and that a high drug dose per kilogram lean body mass (LBM) is associated with more toxicity, but this has not been investigated in LS SCLC. We analysed patients from a randomized trial comparing two schedules of TRT (n = 157) to investigate the prognostic and predictive role of these muscle measures in LS SCLC.

METHODS

Patients from a trial comparing once daily hypofractionated with twice daily hyperfractionated TRT were analysed. The skeletal muscle index (SMI), skeletal muscle radiodensity (SMD), and LBM were assessed from baseline computed tomography scans at the L3 level using the SliceOMatic software.

RESULTS

Images at the L3 level were available for 122 patients (77.7%). Median age was 64 years, 18% had performance status 2, and 38% had stage III. Grade 3-4 toxicity was observed in 89%, and 5% died from treatment-related side effects. Overall, the median overall survival was 23 months, and the 5 year survival was 25%. Median LBM was 45.2 (range: 16-65) kg, the median SMI 44.8 (range: 29-77) cm² /m² , and the median SMD 39.3 (range 16-62) HU. There were no significant associations between survival and any of the muscle measures in the univariable analyses (SMI: P = 0.906, SMD: P = 0.829) or in multivariable analyses adjusting for baseline characteristics (SMI: P = 0.836, SMD: P = 0.260). A higher cisplatin dose per kilogram LBM in the first course significantly increased the risk of grade 3-4 haematological toxicity (P = 0.011) and neutropenic infections (P = 0.012).

CONCLUSIONS

Patients who received a high dose of cisplatin per kilogram LBM had more haematological toxicity and neutropenic infections than other patients. None of the muscle measures were independent prognostic factors for survival in our cohort of LS SCLC patients who underwent standard chemoradiotherapy.

Percentiles for body composition parameters based on computed tomography in patients with endometrial cancer

OBJECTIVES

The aim of this study was to provide the percentiles of distribution of body composition parameters according to cancer staging and body mass index (BMI) stratum, as well as to identify the contribution of age, BMI, and cancer staging in the variation of the different parameters of body composition in a population of patients with endometrial cancer.

METHODS

We enrolled 545 patients who had pretreatment computed tomography images, which were used to assess total skeletal muscle (SM); low- and high-radiodensity SM; visceral, subcutaneous, and intramuscular adipose tissue; and mean skeletal muscle radiodensity (SMD). All the body composition parameters were normalized by the square of the stature. They were then presented on average and at the 5th, 50th and 95th percentiles. The correlation of these parameters with age, BMI, and cancer stage was tested, and then a multiple linear regression analysis was performed. P ≤ 0.05 was accepted as statistically significant.

RESULTS

BMI was associated with body fat parameters and low-radiodensity SM index; cancer stage was associated with SM index, mean SMD, and high-radiodensity SM index.

CONCLUSION

This study provides age, stage, and BMI specific percentiles for body composition parameters, which allowed an in-depth interpretation of how such body compartments, especially the low/high SM sub-ranges, varies according to these stratification variables.

Skeletal muscle density is an independent predictor of diffuse large B-cell lymphoma outcomes treated with rituximab-based chemoimmunotherapy

BACKGROUND

While much cancer research focuses on tumours and their microenvironment, malignancies cause widespread physiologic changes. Cancer and treatment-related sarcopenia, measured with quantitative imaging or as a decrease in overall body mass, are indicative of poor prognosis in elderly diffuse large B-cell lymphoma (DLBCL) patients, skeletal muscle radiodensity (SMD) may be a better prognostic marker. SMD, a measure of muscle radiation attenuation on CT imaging, is more prognostic than sarcopenia or International Prognostic Index (IPI) scores in follicular lymphoma and multiple solid organ malignancies. Low SMD appears to correlate with fat accumulation in muscle and is associated with inflammation. This study set out to examine SMD's prognostic ability in DLBCL.

METHODS

All DLBCL patients treated with rituximab-containing therapy between 2004 and 2009 were compared to determine SMD's prognostic ability in this single centre, retrospective study. Pre-treatment CT scans were used to measure SMD and muscle cross-sectional area. Primary endpoints included progression free (PFS) and overall survival (OS) while objective response rates (ORR) were secondary.

RESULTS

Of 224 evaluable patients, 116 were identified as having low SMD. Low SMD predicted poorer 5 year PFS, 60 vs. 81% (p = 0.001) and OS, 58 vs. 86% (p < 0.0001). SMD's prognostic ability retained significance in multivariate analysis taking into consideration the Revised International Prognostic Index (R-IPI) and sex. Although high SMD was not predictive of ORR (95.4 vs. 91.4%, p = 0.17), it was strongly associated with radiographic complete response (85 vs. 66%, p = 0.0007). Contrary to previous findings, sarcopenia did not predict for poorer OS but suggested improved OS in elderly DLBCL patients (HR 0.38, p = 0.01).

CONCLUSIONS

SMD is a novel prognostic (and potentially treatment predictive) marker independent of R-IPI in DLBCL. It presents an inexpensive yet complementary assessment to R-IPI for prognosticating DLBCL outcomes.