ALBERT

Description: Background: Diffuse large B-cell lymphoma (DLBCL) is a genetically heterogeneous disease of malignant B cells most often classified by tumor gene expression and/or mutations. DLBCL is also characterized by a tumor microenvironmental influence that has not been well described. Immuno-oncology-targeting agents such as immune checkpoint inhibitors have limited clinical activity in DLBCL, highlighting the need for a better understanding of the DLBCL tumor microenvironment for rational drug development, combinations, and disease stratification. Here, we systematically characterize the immune composition of more than 100 DLBCL tumors using 2 imaging technologies and provide insight into the complexity of DLBCL disease biology. Methods: A total of 110 cases of newly diagnosed DLBCL were analyzed by multiplex immunohistochemistry (IHC; n=70) and multiplexed ion beam imaging (MIBI) (n=40), each with 10 common markers (CD20, CD3, CD8, Foxp3, CD56, CD163, CD11c, CD56, PD-1, PD-L1) and a few platform-specific markers. Both IHC and MIBI images were digitalized to generate marker-positive cell counts (including single, double, or triple positivity), cell density (cell count/mm2), and population fraction (% of total nucleated cells). Marker density was analyzed for the major components of tumor-infiltrating cell types and correlation between any pair of markers. In addition, RNAseq data and fluorescence in situ hybridization (FISH) data on MYC, BCL-2, and BCL-6 translocation were generated. Results: In the IHC cohort, T cells (CD3+), dendritic cells (DCs; by CD11c+), and macrophages (CD163+)were the major immune components, with median population fractions of 22%, 16%, and 2.7%, respectively. Natural killer cells (CD56+CD20−) were a minor component at a median of 0.1%. A significant negative correlation was observed between tumor cells (CD20+) and CD4+ (Spearman ρ = −0.47; P = 1.3 × 10−04), and CD8+ T cells (Spearman ρ = −0.42; P = 1.4 × 10−03) cells, and an unexpectedly negative correlation between DCs (CD11c+) and macrophages (CD163+, r = −0.63; P = 9.8 × 10−06) was found. Similar to follicular lymphoma, 2 PD-1+ T-cell populations were identified: PD-1bright and PD-1dim. The PD-1bright cells co-stained with CXCR5, indicating T follicular helper cells (Tfh). The PD-1dim were expressed on exhausted effector cells that co-stained with Tim3 or Lag3. The median population fraction of Tim3+ or Lag3+ among T cells was 25%, whereas that of PD-1dim among T cells was 0.2%, indicating that PD-1 was not a useful marker for exhausted T cells. PD-L1 was predominantly found on DCs and macrophages, and the median population fraction of PD-L1+ among tumor cells was only 6.2%. By unsupervised hierarchical clustering on marker density, 3 major immune-infiltration patterns (P1, P2, and P3) were identified. The first 2 segments (P1 and P2) were 20% and 25% of the total cases, respectively. Both were characterized by high T-cell, macrophage, and DC infiltration. P1 was enriched for PD-1+ T cells, whereas P2 was void of any PD-1+ cells. The third segment (P3) that comprised 55% of the cases was predominantly tumor cells with low T-cell, macrophage, and DC infiltration. PD-L1+ cells were primarily found in segments P1 and P2 but rare in segment P3. Additional analysis on the associations between the 3 immune-infiltration patterns and prognostic DLBCL molecular features such as cell-of-origin, double-hit gene signature, and double-hit FISH will also be presented. Conclusions: These data show the complexity of DLBCL disease biology and show classification of DLBCL at the immune-infiltration level as 3 distinct patterns. The overall low expression of PD-1+ T cells and the restricted pattern of PD-L1+ tumor cells provide a possible explanation for the lack of clinical activity of PD-1/PD-L1 blockade in DLBCL. We also observed other exhausted T cells expressing Lag3 and Tim3, suggesting alternative therapeutic opportunities in stratified populations. These data also highlight the opportunity to develop rational immuno-oncology-targeted agents based on the immune infiltration pattern of DLBCL and selection of patients who may respond more favorably to particular agents. Disclosures Huang: Celgene Corporation: Employment, Equity Ownership. Nakayama:Celgene Corporation: Employment, Equity Ownership. Stokes:Celgene Corporation: Employment, Equity Ownership. Towfic:Celgene Corporation: Employment, Equity Ownership. Lee:Celgene Corporation: Employment, Equity Ownership. Ren:Celgene Corporation: Employment, Equity Ownership. Marella:Celgene Corporation: Employment, Equity Ownership. Wang:Celgene Corporation: Employment, Equity Ownership. Hagner:Celgene Corporation: Employment, Equity Ownership, Patents & Royalties. Couto:Celgene Corporation: Employment, Equity Ownership, Patents & Royalties. Newhall:Celgene Corporation: Employment, Equity Ownership. Gandhi:Celgene Corporation: Employment, Equity Ownership, Patents & Royalties.

Description: Background: The availability of chimeric antigen receptor (CAR)-modified T cells (CAR T) has profoundly increased therapeutic options for patients (pts) with B cell malignancies, including DLBCL. Liso-cel is an investigational, anti-CD19, defined composition, 4-1BB, CAR T cell product administered at a target dose of CD4+ and CD8+ CAR T cells. To understand tumor microenvironmental (TME) factors affecting short-term and durable responses in pts with R/R DLBCL who received liso-cel in the TRANSCEND NHL 001 study, we conducted multiplexed IF analyses of 111 DLBCL biopsies for 83 pts obtained at baseline (n=58) and approximately 11 days (D11) (n=53; 28 paired) after liso-cel infusion (NCT02631044). Methods: We employed three 5-plex IF panels, consisting of antibodies detecting (1) B cell (CD19, CD20) and T cell lineage (CD4, CD8, EGFR) markers, (2) immunosuppressive markers (CD163, FoxP3, CD73, IDO1, PD-L1), and (3) functional markers (CD3, Ki67, GZMB, PD-1, EGFR). Liso-cel expresses a truncated EGFR (EGFRt), and fluorescent anti-EGFR was used to identify CAR T cells within the tumor biopsies. We also performed bulk tumor RNA profiling for an overlapping subset of 50 baseline biopsies and 37 D11 biopsies (11 paired). We investigated the association of differences in marker densities for pts with best overall response (BOR) of complete response (CR), and progressive disease (PD). Baseline and D11 biopsy findings were correlated with early responses at ~1 month (M1) posttreatment (PD n=16; CR n=42) and durable responses at ~9 months (M9) posttreatment (PD n=76; CR n=32; 55 pts evaluated at both M1 and M9). We investigated how baseline and D11 densities, with spatial distinction between tumoral and peritumoral regions, correlated with early and durable responses. All comparisons describe differences in median densities, and have statistical significance reported with uncorrected P values assessed via the (unpaired) Wilcoxon-Mann-Whitney nonparametric test. Results: Signals in baseline biopsies that correlated with early (M1) response differed from those that correlated with durable (M9) CR. A 21% higher baseline presence of PD-1+ T cells was associated with pts who achieved early CR at M1 vs pts who had PD at M1 (P=0.007). Pts with durable CR at M9 had 39% lower baseline levels of CD163+ macrophages (P=0.019) and 270% higher levels of CD73+ cells (P=0.028) than those with PD at M9. On-treatment (D11) tumors of pts with both early and durable CR had 28% higher levels of EGFRt+ (CAR T) CD8+ T cells (P=0.022), and 810% higher EGFRt- (non-CAR T) CD4+ (but notably, not CD8+; P=0.28) T cells (P=0.009). We also investigated changes in marker densities between baseline and on-treatment (D11) biopsies, and found that pts with durable CR at M9 had decreased on-treatment B cell densities (P=0.029), and increased densities of CD8+ GZMB+, Ki67+, and/or PD-1+ CAR (P=0.001) as well as non-CAR T (P=0.017) cells. Pts with durable CR also had a 29% increase in tumor-associated CD163+ macrophages at D11 relative to baseline (P=0.033). While the accessibility of spatial arrangements and multilabeled cells from IF enables a more nuanced picture of the TME, many of the general trends described above are concordant with those observed in bulk tumor RNA sequencing. Lower baseline expression of CD163 (P=0.021) and higher expression of CD73 (P=0.054) were seen in pts with durable CR. Additionally, elevated on-treatment (D11) expression of CD3E, CD4, and liso-cel (P

Description: Diffuse large B-cell lymphoma is a heterogeneous disease, commonly described by cell-of-origin (COO) molecular subtypes. We sought to identify novel patient subgroups through an unsupervised analysis of a large public dataset of gene expression profiles from newly diagnosed de novo DLBCL patients, yielding two biologically distinct subgroups characterized by differences in the tumor microenvironment. Pathway analysis and immune deconvolution algorithms identified higher B-cell content and a strong proliferative signal in subgroup A and enriched T-cell, macrophage, and immune/inflammatory signals in subgroup B, reflecting similar biology to published DLBCL stratification research. A gene expression classifier, featuring 26 gene expression scores, was derived from the public dataset to discriminate subgroup A (classifier-negative, immune-low) and subgroup B (classifier-positive, immune-high) patients. Subsequent application to an independent series of diagnostic biopsies and replicated the subgroups with immune cell composition confirmed via immunohistochemistry. Avadomide, a CRL4CRBN E3 ubiquitin ligase modulator, demonstrated clinical activity in relapsed/refractory DLBCL patients, independent of COO subtypes. Given the immunomodulatory activity of avadomide and need for a patient selection strategy, we applied the gene expression classifier to pretreatment biopsies from relapsed/refractory DLBCL patients receiving avadomide (NCT01421524). Classifier-positive patients demonstrated an enrichment in response rate and progression-free survival of 44% and 6.2 months versus 19% and 1.6 months for classifier-negative patients (HR=0.49 [0.280, 0.86]; P=.0096). The classifier was not prognostic for R-CHOP or salvage immunochemotherapy. The classifier described here discriminates DLBCL tumors based on tumor and non-tumor composition and has potential utility to enrich for clinical response to immunomodulatory agents including avadomide.

Description: The generalized fractal dimensions are measured for the time series based on two complete earthquake catalogues: one with M≥6 earthquakes occurring in the north-south seismic belt of mainland China during the 1900-1990 period published by Ma et al. (1992) and the other with M≥5.5 earthquakes occurring in southern California, USA during the 1915-1994 period compiled by Press and Allen (1995). The log-log plot of Cq versus t, where Cq(t) is the generalized correlation integral and t is the interoccurrence time in years between two events, at positive q shows a linear distribution when t〈tc. Dq is the slope of this linear portion. The value of tc decreases from 50.1 to 39.8 years for Chinese earthquakes and from 50.1 to 31.6 years for southern California events as q is increased from 0 to 15. For M≥6 Chinese earthquakes, the well-distributed, monotonically decreasing function of Dq with increasing q would imply that such earthquakes have formed a multifractal time series. In contrast, the M≥5.5 southern California earthquakes might have not yet formed a complete multifractal time series or the number of these events is too small to accurately estimate the multifractal dimensions, especially for large qs. Different degrees of complexity of fault distributions in the two seismic regions might also be a factor in causing the difference in the Dq-q relations. In addition, the results also suggest that a Dq-q relation is better than the first three commonly-used values of Dq to completely represent a multifractal time series.