Banca de DEFESA: AMANDA DE SOUSA REBOUÇAS

Uma banca de DEFESA de DOUTORADO foi cadastrada pelo programa.
STUDENT : AMANDA DE SOUSA REBOUÇAS
DATE: 25/03/2026
TIME: 14:00
LOCAL: REMOTA
TITLE:

BIOELECTRIC IMPEDANCE ANALYSIS IN CANCER PATIENTS: PROPOSAL OF AN EQUATION FOR ESTIMATING SKELETAL MUSCLE AREA AND PREDICTING PHASE ANGLE FOR MORTALITY

KEY WORDS:

bioelectrical impedance; cancer; cancer imaging tests, risk factors; phase angle; prognosis.

PAGES: 58
BIG AREA: Ciências da Saúde
AREA: Nutrição
SUMMARY:

The decrease in skeletal muscle mass is a condition that affects cancer patients, requiring methods of body composition assessment that enable safe and accurate nutritional diagnosis. Thus, computed tomography is an opportunistic technique that allows this assessment to be performed in cancer patients. However, in clinical practice, its routine applicability is limited, with bioelectrical impedance analysis (BIA) being an accessible and non-invasive alternative, although it depends on predictive equations validated for specific populations. In addition, the phase angle (PA), derived from BIA, is considered a prognostic marker in cancer patients, although it still lacks well-established cut-off points. Therefore, the objective of this study was to explore bioelectrical data derived from BIA for estimating muscle mass and predicting mortality in cancer patients. The thesis consists of products from two studies. A four-pole model was used to evaluate BIA, which provided resistance (R) and reactance (Xc) values. The first study aimed to develop a predictive equation based on BIA to estimate skeletal muscle area, using CT as the reference method. This is a bicentric cross-sectional study involving 211 patients. The equation was developed using a linear regression model, maintaining the variables that best correlate with the skeletal muscle area of the CT. Bland-Altman plots and the bootstrap resampling method were used for validation. Lins' coefficient of concordance (CCC), mean square error, and mean absolute error were calculated before and after resampling. The proposed equation included sex, age, weight, height, resistance, and reactance. The second study aimed to determine the AF cut-off points associated with short-term mortality at different stages of cancer. It was a multicenter cohort study with 1,121 adult cancer patients in Brazil. The primary outcome was 6-month mortality. The predictive accuracy of FA was estimated using ROC curves, and cut-off points were estimated using Youden's index. Cox regression models examined the crude and adjusted associations between FA and mortality at different TNM stages (I-II, III, IV). As for the results of the first study, the model explained more than 85% of the variability in skeletal muscle area (adjusted R² = 0.86), with a mean square error of 10.37 cm² and a mean absolute error of 8.28 cm². The skeletal muscle area of BIA showed a high correlation with that of CT (ρ = 0.93, p < 0.001). Bland-Altman plots and Cox's correlation coefficient demonstrated moderate agreement between the methods (0.92). The proposed bioimpedance equation showed potential for predicting skeletal muscle area when compared to the reference standard. The second study noted that FA values were significantly lower in non-surviving patients (mean difference: -1.02° in men; -1.21° in women). ROC analyses demonstrated reasonable predictive performance, with optimal thresholds of ≤ 5.43° for men (AUROC 0.74) and ≤ 4.22° for women (AUROC 0.77). At all stages of the disease, AF consistently predicted mortality, with greater accuracy at stage IV (AUROC 0.75; criterion ≤ 4.72°). In multivariate Cox models, AF remained independently associated with mortality at 6 months after adjusting for confounding factors. Therefore, in this study, it was observed that values below 5.5° can be used for risk stratification and to aid in decision-making in cancer patients.

COMMITTEE MEMBERS:
Externa à Instituição - FERNANDA MICHIELIN BUSNELLO - UFCSPA
Presidente - 1879430 - ANA PAULA TRUSSARDI FAYH
Externo à Instituição - POLIANA COELHO CABRAL - UFPE
Externa à Instituição - SARA MARIA MOREIRA LIMA VERDE - UECE
Externa ao Programa - 1452705 - SEVERINA CARLA VIEIRA CUNHA LIMA - UFRN