“…Putative reasons which may contribute to affect the consistency and comparability of available findings in this setting (and more generally the assessment of PD-L1 expression as a biomarker in BC) are (1) intrinsic to the study population, including a small number of patients overall [50,58,62], an exceedingly low recurrence rate after BCG treatment, and/or unbalanced subgroup size (i.e., BCG responders vs. non responders) [53]; (2) intrinsic to the treatment, including the frequency and/or timing of BCG administration and type of BCG reagents; (3) intrinsic to the method, such as PD-L1 assessment through heterogeneous assays, antibody clones, cellular populations, scoring systems, and cut-off points (see next section); and (4) intrinsic to the tumor/tissue sample, due to the intra-tumoral heterogeneity of PD-L1 expression and the dynamic nature of the tumor microenvironment [66,67]. Furthermore, the constantly changing of the molecule over time (temporal heterogeneity) has been proven by the discrepancy over different tumor stages, as well as between metastatic sites and corresponding primary tumors, revealed by several authors [68,69].…”