Skip to main content
. 2014 Dec 31;6(7):5022–5040. doi: 10.18632/oncotarget.3217

Figure 4. Forest plots show association between CXCR4 over-expression and PFS in breast cancer.

Figure 4

(A) Summary for all thirteen trials, the estimates is 1.44(1.23-1.69) using fixed effects model. (B) Although three studies appeared to be outliers (Andre, 2009; Mirisola, 2009; Holm, 2009), we did not find clinical heterogeneity justifying their exclusion. Excluding three studies yield similar results but without significant heterogeneity. The estimate is 1.75(1.41-2.17) using fixed effects model. (C) Funnel plots showing association of CXCR4 and PFS in breast cancer. Visual inspection of the Begg funnel plot did not identify substantial asymmetry.