151751-Najiba-Chargi

243 Systemic therapy: skeletal muscle mass and anti-cancer drug toxicity SKELETAL MUSCLE MASS ASSESSMENT There were several differences between the studies in the method used to measure SMM. All included studies used CT to evaluate SMMwith one study also using MRI. 50 However, there was a difference in the selected vertebrae used for the SMM assessment as shown in Table 1. Most studies used lumbar level 3 (L3); other vertebrae that were used were cervical level 3 (C3) and thoracic level 4 (T4). Supplementary Table S1 shows other differences between studies such as the time between CT and treatment start, as well as the software used to measure SMM. The included studies also used different cut-off values for low SMM; this can be seen in Table 1. Most studies used cut-off values cited from previous articles. The most commonly used cut-off values were established by Prado et al., 2008 46 (<52.4 cm 2 /m 2 for men and <38.5 cm 2 / m 2 for women), followed by Martin et al., 2013 47 (<43 cm 2 /m 2 for men if body mass index (BMI) ≤ 24.9 kg/m 2 or <53 cm 2 /m 2 for men if BMI > 25 kg/m 2 and <41 cm 2 /m 2 for women), Fujiwara et al., 2015 48 (≤36.2 cm 2 /m 2 for men and ≤29.6 cm 2 /m 2 for women), and Caan et al., 2018 51 (<40 cm 2 /m 2 ). It is noteworthy that five studies cited the cut-off values of Prado et al. 46 but used other cut-off values in their analysis than those published by Prado et al. 5,7,37,39 Four studies did not use cut-off values for low SMM and instead used continuous SMI during analysis. 28,29,34,36 STUDY QUALITY ASSESSMENT The results of the QUIPS assessment of all included studies are summarized in Figure 2. Out of the 31 included studies, seven studies had a low risk of bias, 6,16,20,21,30,34,41 16 studies had a moderate risk of bias, 5,7,8,22,24,26,28,31,33,39,40,43–45,50,52 and eight had a high risk of. 23,29,31,36–38,42 The domains study participation, study confounding, and statistical analysis and reporting were most frequently assessed as having a high risk of bias. Whereas the domains study attrition, prognostic factor measurement, and outcome measurement were most frequently assessed as having a low risk of bias. ASSOCIATION BETWEEN LOW SMM AND TOXICITY Figure 3A shows the forest plot for the OR of 13 studies that used DLT as the measure of toxicity. Kurk et al., 2019 27 performed two separate analyses in the same patient population receiving sequential treatments, 232 patients treated with Capox-B and 182 patients treated with Cap-B. These results were entered into the forest plot separately. Patients with low SMM had a significantly higher risk for DLT compared to patients without low SMM (OR 2.24; 95%CI 1.28–3.92, p < 0.001). Heterogeneity across studies was high ( χ 2 = 60.97 and I2 = 79%). Figures 3B,C show a selection of the 13 studies that used DLT as an endpoint. To create an analysis with less heterogeneity, studies were matched together based on identical cut-off values, measurement techniques, and vertebrae level analyzed. The studies included in Figure 3B all used the cut-off values established by Martin et al., 2013 47 and measured SMM at L3 using CT. There was no association between low SMM and DLT (OR 1.98; 95% CI 0.76–5.22, p = 0.16). Het - erogeneity across studies was high ( χ 2 = 24.48 and I2 = 84%). The studies included in Figure 3C all used cut-off values established by Prado et al., 2008 46 as well as the same measurement technique at L3 using CT. There was no associated between low SMM and DLT (OR 1.87; 95% CI 0.32–10.93, p = 0.49). Heterogeneity across studies was high ( χ 2 = 60.97 and I2 = 79%). 13