Risk of Rash in PD-1 or PD-L1-Related Cancer Clinical Trials: A Systematic Review and Meta-Analysis

Figure 6

Forest plots of comparison groups for grades 3–5. (a) The OR of rash for grades 3–5 checked using the random effect (RE) model in Group A (PD-1 or PD-L1 versus Chemotherapy): subgroup analyses were carried out based on tumor types. (b) The OR of rash for grades 3–5 checked using the random effect (RE) model in Group B (PD-1 or PD-L1 plus Chemotherapy versus Chemotherapy): subgroup analyses were carried out based on tumor types. (c) The OR of rash for grades 3–5 checked using the random effect (RE) model in Group E (PD-1 or PD-L1 versus Placebo): subgroup analyses were carried out based on tumor types. (d) The OR of rash for grades 3–5 checked using the random effect (RE) model in Group G (PD-1 or PD-L1 plus CTLA-4 versus PD-1 or PD-L1): subgroup analyses were carried out based on tumor types. (e) The OR of rash for grades 3–5 checked using the random effect (RE) model in Group F (PD-1 or PD-L1 plus Chemotherapy versus PD-1 or PD-L1): subgroup analyses were carried out based on the types of immune checkpoint inhibitors (PD-1 or PD-L1). (f) The OR of rash for grades 3–5 checked using the random effect (RE) model in Group H (PD-1 or PD-L1 versus CTLA-4). (g) The odds ratio of rash for grades 3–5 checked using the random effect (RE) model in Group M (PD-1 or PD-L1 versus Methotrexate/docetaxel/cetuximab). (h) The OR of rash for grades 3–5 checked using the random effect (RE) model in Group K (PD-1 or PD-L1 plus Bevacizumab versus Sorafenib): subgroup analyses were carried out based on the name of immune checkpoint inhibitors. (i) The OR of rash for grades 3–5 checked using the random effect (RE) model in Group J (PD-1 or PD-L1 plus CTLA-4 plus Chemotherapy versus Chemotherapy): subgroup analyses were carried out based on treatment regimens.

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)