The organization of RCNs in DLBCL NOS calculated using the average distance between the cells in different RCNs. A, A heatmap depicting the average distances from cells in the RCNs shown on the y-axis (neighborhood of interest) to the cells in the RCNs depicted on the x-axis. The distance from neighborhood x to neighborhood y represents the average distance of a cell in neighborhood x to the nearest cell in neighborhood y, whereas the distance from neighborhood y to neighborhood x signifies the average distance of a cell in neighborhood y to the nearest cell in neighborhood x. B, Graphical illustration of the organization of RCNs in DLBCL NOS. The RCN that is, on average, situated closest to another RCN is connected to that RCN by a line. RCNs not connected by a line are, on average, situated further apart from each other. The distances between cells in neighborhoods x and y are calculated as the average of the distance from cells in neighborhood x the neighborhood of interest in (A) to cells in neighborhood y, and the distance from cells in neighborhood y to cells in neighborhood x. The thickness of the line is inversely proportional to the distance between cells in the neighborhoods. C, Kaplan–Meier (log-rank test) survival plot depicts PFS in rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP)–treated patients with DLBCL NOS in whose lymphomas CD8+ T cell–rich RCN2 neighborhoods are situated close to immune-poor RCN5 neighborhoods, and patients in whose lymphomas these RCNs are situated far from each other or are not present. The optimal cutoff was sought using the R maxstat package. D, A forest plot visualizing the impact of the proximity of CD8+ T cell–rich RCN2 neighborhoods to immune-poor RCN5 neighborhoods on PFS in a Cox multivariable regression analysis with IPI and COO in R-CHOP–treated patients with DLBCL NOS. E, Kaplan–Meier (log-rank test) survival plot depicts PFS in R-CHOP–treated patients with DLBCL NOS in whose lymphomas PD-L1+ B cell–rich RCN7 neighborhoods are situated close to CD8+ T cell–rich RCN2 neighborhoods, and patients in whose lymphomas these RCNs are situated far from each other or are not present. The optimal cutoff was sought using the R maxstat package. F, A forest plot visualizing the impact of the proximity of PD-L1+ B cell–rich RCN7 neighborhoods to CD8+ T cell–rich RCN2 neighborhoods on PFS in a Cox multivariable regression analysis with IPI and COO in R-CHOP–treated patients with DLBCL NOS. G, Voronoi plots of representative samples in which CD8+ T cell–rich RCN2 neighborhoods are situated close to immune-poor RCN5 neighborhoods (left) and PD-L1+ B cell–rich RCN7 neighborhoods are situated close to CD8+ T cell–rich RCN2 neighborhoods (right). px, pixel. B, Created in BioRender. Autio, M. (2025) https://BioRender.com/dnqcyry
Autio et al. (Wed,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: