151751-Najiba-Chargi

248 CHAPTER 13 Figure 5. Forest plots for association between low skeletal muscle mass (SMM) and toxicity specifically for monotherapies used in multiple studies. DISCUSSION In this review, 31 studies were evaluated, of which 19 were used in the meta-analysis. The meta-analysis showed that low SMM has predictive value for toxicity (DLT OR = 2.24 and ≥grade 3 toxicity OR = 4.08). Heterogeneity across studies using DLT as the outcome was very high, which can be explained by the differences in the definition of DLT. The general definition of DLT is any toxicity leading to dose reduction, treatment delay, or discontinuation. However, studies differed in the level of detail of this definition, for example, some studies included any dose reduction 20,41 , others applied aminimumof 50% reduction 8,42 , some studies also included toxicity leading to hospitalization 35 , some studies included any treatment delay 20,41 , and others included delays over 4 days 8 or 7 days 24 . Even when creating subgroups by matching cut-off values, measurement techniques, and vertebrae level analyzed, the heterogeneity remained high because of the difference in the definition of DLT and could not provide accurate evi - dence of the association between low SMM and DLT. The results presented in Figure 4 for the association of low SMM with toxicity ≥ grade 3 were much more reliable because of the low level of heterogeneity with an I2 of 0%. This can be explained by the clear definition of grade 3 toxicity according to CTCAE. Therefore, the meta-analysis for the association between low