Page 93 - Risk quantification and modification in older patients with colorectal cancer

Physical performance and poor surgical outcomes
Severe complications were defined as complications leading to an ICU admission (more than two days), the need for a (surgical) reintervention, a prolonged hospital stay (more than 14 days), or postoperative mortality. This definition is consistent with previous publications in which data from the DSCA were analysed.15 Date of death was retrieved through a linkage with the Municipal Personal Records Database. Follow-up of all patients was at least one year.
Skeletal muscle mass and density assessment
Computerised tomography (CT) was used to assess skeletal muscle mass and density (or muscle attenuation) as a proxy of muscle quality.16 Low muscle density reflects changes in skeletal muscle composition that occur with increased fat infiltration into skeletal muscle17 and has been associated with low physical performance in previous well-functioning older men and women.18 Picture Archiving and Communication System (PACS) radiological software were used to identify and extract preoperative contrast-enhanced CT-images. CT-image analysis software SliceOmatic version 4.3 (Tomovision, Montreal, Quebec, Canada) was used for assessment of skeletal muscle mass and muscle density. Skeletal muscle mass was evaluated on a single slice at the L3 level using Hounsfield Unit (HU) thresholds of−29 to 150 for skeletal muscle.19 Muscles measured at this level were the psoas, paraspinal, transverse abdominal, external oblique, internal oblique and rectus abdominis muscle. The sum of skeletal cross-sectional muscle areas was normalised for stature (m2)20 and is reported as cm2/m2. Skeletal muscle density is expressed as the mean of the HU of the skeletal muscle mass measured at L3.
All images were analysed by a trained investigator. To confirm the reliability of measurements, 25% (n=52) of all images were randomly selected and analysed by a trained second analyst (i.e. radiologist). Cohen’s kappa statistic was used to test interrater reliability.
Low skeletal muscle mass and low muscle density
Low skeletal muscle mass and muscle density were defined using two different methods. The first method was based on thresholds published by Martin et al.21 These thresholds were developed in a mixed cohort of cancer patients (40% CRC patients, mean age 65) accounting for gender and BMI differences. For skeletal muscle mass thresholds were: men BMI < 25 kg/m2 skeletal muscle mass< 43cm2/ m2, men BMI≥25 kg/m2 skeletal muscle mass < 53 cm2/m2, women independent of
91
5