Computed Tomography-defined Sarcopenia and Performance of Gfr Estimating Equations in Patients With Cancer
Year: 2024
Type: conference-abstract
Abstract: Introduction: Estimated glomerular filtration rate (eGFR) equations based on serum creatinine (Scr)(eGFRcr) combined or not with serum cystatin C (Scys) present worse accuracy in patients with low body mass index (BMI). This analysis aims to evaluate the impact of sarcopenia, defined by computed tomography (CT), on the performance of the CKD-EPI equations without race based on Scr and Scys in patients with cancer. Methods: We included adult patients undergoing treatment for solid tumors between May 2017 and October 2017 and an abdominal CT scan within 90 days of measured GFR (mGFR) using the plasma clearance of 51Cr-EDTA. Body composition on an axial image at the third lumbar vertebral body level was performed using a previously validated machine-learning pipeline. Skeletal muscle index was defined as cross-sectional skeletal muscle area [cm2] divided by the square of the patient‘s height [m2]. Sarcopenia was defined using independently established cutoffs of skeletal muscle index (< 39 cm2/m2 for women, < 55 cm2/m2 for men). Bias was calculated as the median value of (mGFR-eGFR). 1-P30 is the percentage of estimates that differed by more than 30% from the measured GFR. Results: A group of 465 patients was included for analysis. Patients were 58±14 years, 50% male. The most common cancer sites were gastrointestinal (39.8%), breast (15.9%), and male genital (9.9%); 20% had metastasis. CT-based sarcopenia was observed in 33.8% of patients in the entire cohort and in 23.4% of those with BMI ≥ 25 kg/m2. Mean (SD) mGFR was 78.0 (22.5) ml/min/1.73 m2. In patients with sarcopenia, eGFRcr largely overestimated mGFR (median bias: -14.4 [-17.8, -13.2] ml/min/1.73 m2) and presented poor accuracy (1-P30: 42.0 [34.4, 49.6] %); eGFRcys underestimated mGFR (median bias: 4.1 [2.5, 6.5] ml/min/1.73 m2)and presented better accuracy (1-P30: 20.4 [14.0, 26.8] %) compared to eGFRcr. eGFRcrcys was the most accurate (1-P30: 14.0 [8.3, 19.1] %). Even in patients with high BMI, those with sarcopenia had large eGFRcr overestimation (median bias: -14.7 [-18.1, -11.6] ml/min/1.73 m2), and poor accuracy (1-P30: 41.7 [29.2, 52.8] %). Conclusion: This is the first study to demonstrate large bias and poor accuracy for eGFRcr and eGFRcrcys in patients with CT-defined sarcopenia, demanding other instruments such as Scys or mGFR to assess GFR more accurately in this population. Overweight and obese patients might benefit from body composition analysis to unveil underlying sarcopenia.
Language: English
Subfield: Physiology
Field: Medicine
Domain: Health Sciences
Open Access status: hybrid
OpenAlex ID: W4415776810