ESTRO 2024 - Abstract Book

S2441

Clinical - Urology

ESTRO 2024

Segmentations were visually inspected and failures removed. Skeletal muscle index (SMI) was then calculated by dividing the skeletal muscle cross-sectional area by patient height squared.

Univariable and multivariable linear regression models analysed relationships with acute and late standardised total average toxicity (STAT) scores [6] as well as two individual toxicity end points. Acute toxicity was defined as max reported toxicity within 90 days of the end of radiotherapy and late toxicity defined as max reported after 90 days; rectal, bladder and bowel toxicity endpoints were included. Multivariable analyses included: age, diabetes, previous surgery, length of hormone therapy and prescribed radiotherapy dose (converted to biologically equivalent dose), as known prognostic factors for toxicity. In addition, Gleason score and baseline PSA were included as more aggressive disease may drive muscle loss through cachectic processes. Finally, to further understand overall body composition, intra-muscular adipose tissue area (identified by applying a threshold inside the skeletal muscle mask of -190 to -30 HU) and BMI were included in the multivariable models with the interaction between SMI and BMI investigated. Due to limited field of view of the radiotherapy planning CT scans, L3 was only visible on 973 images. Of these, 885 (91%) passed visual inspection and were included in the analysis. 635 individuals received hormone therapy, with median duration of two-years; 249 had undergone prostatectomy. Median SMI was 50.5cm2 /m^2 (range 31.6 79.5cm2/m^2), figure 1a and median BMI 27 kg/m^2 (17.4-43 kg/m^2), figure 1b, with the majority classified as overweight (54%) or obese (21%) (classified with the world health organisation definition). There was no significant association with SMI and STAT for acute toxicity. SMI was significantly associated with worse STAT for late toxicity on multivariable analysis p=0.02, the negative beta coefficient indicating increasing SMI reduces the risk of late toxicity. Prescribed dose was also significantly associated with increased risk of late toxicities (beta coefficient 0.015, p=0.002). A significant interaction with BMI was observed on multivariable analysis, p=0.01, table 1. Figure 1c includes a contrast plot for SMI for individuals with a healthy, overweight or obese BMI. Patients in the healthy category showed a significantly higher risk of toxicities for low SMI, with a significantly lower risk at higher SMI (colour bands representing 95% confidence intervals). For individuals in the overweight and obese categories, there was no significant change in risk with SMI. This may suggest that increased weight at the time of radiotherapy was protective against toxicity. Interestingly, BMI was not significant in univariable or multivariable analysis (with and without SMI included) indicating that SMI is not acting as a surrogate for BMI and provides independent prognostic information. Multivariable analysis of individual late toxicity endpoints identified lower SMI as significantly associated with haematuria (p=0.03), urinary obstruction (p=0.04), and proctitis (p<0.001). Results: