Fig 1: Validation of B-cell presence in FFPE tumor tissue of patients with PPBCPW, Pr-BC, and NP-BC and correlation with prognosis by means of multiplex IF (mIF) CD20+ labeling, TAPC scoring, and singleplex IHC CD38 labeling. A, Representative image of the two designed mIF panels showing staining for CD3, CD20, CD27, FoxP3, CD8, and CD138 markers. B, Cell density was estimated for all patient groups and visualized as a log10 transformation with a pseudocount of 1. Significance was calculated using Wilcoxon rank-sum test (pairwise) or a Kruskal–Wallis test (between all groups). CD20+CD3−CD8− cells—further described as CD20+ cells—(representing nonplasma B cells). No significant difference in CD20+ densities (number of cells/total tissue area) were observed between patients with PP-BCPW (n = 44) and NP-BC (n = 58; Wilcoxon, P = 0.10). mIF patient groups below the minimum power threshold of n = 15 (Pr-BC, n = 11 and PP-BCDL, n = 7) are visualized but not subjected to statistical tests (see decision tree visualized in Supplementary Fig. S1). C, Kaplan–Meier OS curve for all patient groups with adequate sample size and follow-up >20 months. Pr-BC and PP-BCDL were excluded from mIF survival curves due to low sample numbers delivering insufficient power (see decision tree visualized in Supplementary Fig. S1). Neither high nor low CD20+ mIF density was not significantly associated with OS between PP-BCPW and NP-BC, as determined via a log-rank test. D, CD20+/CD20+ clustering (normalized cellular colocalization within a fixed radius of 0.30 mm) indicate a significantly decreased clustering in patients with PP-BCPW compared with NP-BC (Wilcoxon, P = 0.03), whereas (E) no difference in survival can be observed based on CD20+/CD20+ clustering. F, Representative image of a patient with PP-BCPW with high and low TAPC counts in the tumor area as evaluated in an H&E-stained FFPE slide. G, Patients with PP-BCPW (n = 66) present with significantly higher TAPC counts compared with patients with Pr-BC (n = 51) and NP-BC (n = 76; Wilcoxon, P = 0.011). H, Although no significant difference in survival rates was seen between patient groups with low TAPC scores (score “0”–“1”), patients with PP-BCPW with high TAPC scores (score “2”–“3”) were associated with a significantly increased risk for death (log-rank P = 0.03) compared with patients with Pr-BC and NP-BC. Only PP-BCDL (n = 17) was excluded. I, Representative image of CD38+ IHC staining showing high and low densities of CD38+ plasma cells of a patient with PP-BCPW. J, Significantly increased intratumoral CD38+ plasma cell density in the tumor area of patients with PP-BCPW (n = 50) versus patients with NP-BC (n = 63; Wilcoxon, P < 0.01). K, Although no significant difference in survival rates was seen between patient groups with low intratumoral CD38+ densities, patients with PP-BCPW with high intratumoral CD38+ densities had a significantly increased risk for death (log-rank P = 0.02) compared with patients with NP-BC. Significant P values are indicated with an asterix (*).