Tumor-Associated Macrophages Predict Outcome in De Novo Diffuse Large B Cell Lymphoma Treated with R-CHOP
King L Tan, Graham W Slack, Laurie H Sehn, Joseph M Connors, David W Scott, Christian Steidl, Randy D Gascoyne. British Columbia Cancer Agency, Centre for Lymphoid Cancer, Vancouver, Canada
Background: The tumor microenvironment plays an important role in diffuse large B cell lymphoma (DLBCL). Gene expression profiling of DLBCL treated with R-CHOP defines two stromal signatures; the prognostically favorable stromal-1 signature reflecting extracellular-matrix deposition and macrophage infiltration, and the prognostically unfavorable stromal-2 signature reflecting in part, blood vessel density. While CD68 is used as a pan macrophage marker (both M1 and M2 subtypes), CD163 is reported to be a marker of alternatively-activated (M2) macrophages, which promote tumor growth and associated with worse prognosis in some solid tumors. We investigated the prognostic significance of macrophages in de novo DLBCL uniformly treated with R-CHOP.
Design: A tissue microarray was constructed using duplicate 0.6mm cores from diagnostic paraffin blocks including 186 patients with de novo DLBCL uniformly treated with R-CHOP. IHC for CD10, BCL6, MUM1, GCET1, FOXP1, LMO2, CD68 and CD163 were performed. Cases were classified as GCB and non-GCB subtypes using the Tally algorithm. CD68 and CD163 IHC were analyzed by computer image analysis (Aperio) and pathologist scoring. Optimum thresholds for CD68 and CD163 expression and survival were determined by X-tile. Survival analysis was determined by Kaplan-Meier method with differences evaluated by log-rank test. Multivariate analysis of survival predictors was performed by Cox regression.
Results: There was a significant correlation between Aperio and pathologist scores for CD68 (R=0.860, p<0.001) and CD163 (R=0.939, p<0.001). Increased CD68 expression was associated with superior survival, while increased CD163 expression was associated with inferior survival (Table1). CD163 expression was higher in non-GCB vs GCB subtypes (median 18.6% vs 5.8% respectively, p<0.001). No significant difference between non-GCB and GCB subtypes was seen with CD68 expression. In multivariate analysis with the International Prognostic Index (IPI), Tally algorithm, CD68, CD163; the IPI, CD68 and CD163 remained independent predictors of OS (p=0.002, p=0.007, and p=0.013 respectively) and PFS (p<0.001, p=0.006 and p=0.034, respectively).
|5 year PFS||40%||64%||0.017||78%||48%||0.005|
|5 years OS||53%||71%||0.013||85%||58%||0.006|