PP299 - 6-MONTH MORTALITY PREDICTION BASED ON A NEW BIA EQUATION ESTIMATING CT-DERIVED MUSCLE QUANTITY IN PATIENTS WITH CANCER

Linked sessions

PP299

6-MONTH MORTALITY PREDICTION BASED ON A NEW BIA EQUATION ESTIMATING CT-DERIVED MUSCLE QUANTITY IN PATIENTS WITH CANCER

A. P. T. Fayh1,*, J. P. Costa-Pereira2, A. D. S. Rebouças3, M. C. Gonzalez4, C. M. Prado5

1Universidade Federal de Ciências da Saúde de Porto Alegre, Porto Alegre, 2Universidade Federal de Pernambuco, Recife, 3Universidade Federal do Rio Grande do Norte, Natal, 4Universidade Federal de Pelotas, Pelotas, Brazil, 5University of Alberta, Edmonton, Canada

 

Rationale: Although computed tomography (CT) is the gold-standard method to assess muscle  quantity in patients with cancer, bioelectrical impedance analysis (BIA) emerges as a potential alternative in clinical practice. We previously proposed a new BIA equation using CT as the reference method, but its clinical validity requires further exploration, particularly the predictive value for mortality among patients with cancer.

Methods: This multicenter cohort included 1,151 adult patients with cancer in Brazil (≥18 years; 52.1% females; mean age 60.2 ± 13.4 years; 44% colorectal cancer; 34.6% TNM IV). Skeletal muscle (SM) area was estimated using a previously published equation that uses raw BIA data: SM (cm2) = 100.401 – (0.329 x age) – (21.246 x sex,  male= 0; female = 1) + (0.545 x weight, kg) + (30.052 x height, m) - (0.126 x resistance, Ω) + (0.642 x reactance, Ω). The outcome was 6-month mortality. Cox regression model examined crude and adjusted associations between SM and mortality.

Results: During follow-up, 240 (20.9%) died. Patients who died were older than survivors (MD: 2 years, 95% CI -0.33 to - 4) and presented with advanced TNM stage (stage IV, 50.4%). SM values were significantly lower in non-survivors (110.7 ± 29.2 cm2 vs 119.9 ± 26.4 cm2 for survivors, p = 0.04). In multivariate Cox models, SM was independently associated with 6-month mortality after adjustment for confounders: age and TNM stage (hazard ratio [HR] 0.988; 95% confidence interval [CI]:0.982; 0.993), p < 0.001).

Conclusion: SM estimated with the equation is an independent predictor of 6-month mortality in patients with cancer, which improve its potential utilization in clinical scenarios.

Disclosure of Interest: None declared