Background IMGN529 is an antibody drug conjugate (ADC) consisting of an anti-CD37 antibody with direct anti-tumor activity conjugated via a thioether linker to the cytotoxic maytansinoid antimicrotubule agent DM1. IMGN529 has shown pre-clinical (Deckert et al, Blood 2013) and clinical activity in lymphoma (Stathis et al, ASH 2014; NCT01534715). Here, we assessed the anti-tumor activity of IMGN529 on a large panel of B cell and T cell human lymphomas to identify potential biomarkers of response.

Methods Fifty-four lymphoma cell lines [diffuse large B cell lymphoma (DLBCL), n.=27; mantle cell lymphoma (MCL), n.=10; anaplastic large T-cell lymphoma, n.=5; marginal zone lymphomas, n=6, others, n=6] were exposed to increasing doses of IMGN529 or to the unconjugated DM1 for 72h. Cell proliferation was measured using the MTT. Apoptosis induction was defined by at least 1.5-fold increase in caspase 3/7 signal activation with respect to controls using the Promega ApoTox-Glo Triplex Assay. CD37 surface expression was assessed by cytofluorimetry. Gene expression profiling (GEP) was done with the Illumina HumanHT-12 Expression BeadChips on untreated cell lines followed by GSEA (NES > |2|, P<0.05, FDR<0.25) and limma t-test (FC> |1.2|; P< 0.05; top 200 up and top 200 down).

Results. The IMGN529 median IC50 in the 54 cell lines was 780pM (95%C.I., 263pm-11.45nM). Activity was stronger (P<0.001) in B cell lymphoma cell lines (n= 46; median IC50=450pM; 95%C.I., 150-800pM) than in T cell lymphoma cell lines (n=8; median IC50=22.5nM; 95%C.I., 14-40nM). The median IC50 for DM1 was 30pM (C.I.95%, 20-40pM) with no differences between B and T cell lymphoma origin. IMGN529 induced apoptosis in 33/54 (61%) lymphoma cell lines. Surface CD37 expression was higher in cell lines derived from B than from T cells (P< 0.0001): IMGN529 IC50 values, but not of DM1, were negatively correlated with surface CD37 expression across all cell lines (R=-0.39; P= 0.018), but not within the individual B or T cell subgroups. Among B cell lines, DLBCL cell of origin, TP53 status or the presence of BCL2 translocation did not affect the sensitivity to IMGN529, while IC50s were higher in the presence of MYC translocation (P= 0.043). No association was seen between IMGN529-induced apoptosis or the sensitivity to DM1 with DLBCL cell of origin, TP53 status or the presence of BCL2 or MYC translocations.

We then compared the baseline gene expression profiling of DLBCL cell lines that were highly sensitive to IMGN529 (IC50< 800pM; "S") versus less sensitive/resistant DLBCL cell lines (IC50>10nM, "R"), separately for germinal center B cell type (GCB) (S, n=11; R, n=8) and for activated B cell like (ABC) (S, n=4; R, n=3). In both DLBCL groups, MYC targets, genes involved in unfolded protein response, glycolysis and DNA repair were enriched in transcripts more expressed in R than S cell lines. Transcripts associated with low sensitivity included CD44, VIM, ANXA2, BCL2, ANXA2P1, HSP90B1, NFKBIZ, CDK6, BIRC5 in GCB and HSPA1B, HSP90AA1, CADM1, CD86, TUBB2A, TUBG1, NOTCH1 in ABC cell lines. HEBP1, PHB, PSME3, RNU6-15, RPL13 were more expressed in both GCB and ABC R. Genes involved in PI3K/AKT/mTOR, hypoxia, INF-gamma, TNFA signaling via NFKB and in complement were more expressed in S than in R cell lines. Genes associated with sensitivity to IMGN529 comprised: CD37 (IMGN529 target), CD79A, CHI3L2, FAM117B, LPAR5, NFATC1, PTPN22, RBM38, SGPP1, SLC6A16 in both GCB and ABC cell lines; BASP1, CXCR5, BIK, LY86, TLR10, CD86, LCK, CD22, PTPN22, BCL6, PIK3IP1, CDKN2A in GCB; AFF3, PIM1, MGMT, PDE4B, NFKBIE, SYK, FOXO1in ABC.

Conclusions. IMGN529 showed a very strong anti-tumoral activity in pre-clinical lymphoma models. High expression of CD37 and mostly genes involved in BCR signalling were associated with sensitivity to IMGN529. Conversely, the presence of MYC translocation, a high expression of MYC targets and of genes known to be involved in drug resistance (BCL2, BIRC5, CDK6, heat-shock proteins, annexins, proteasome and tubulin components) appeared to negatively affect the response to the ADC but also represent therapeutic targets for novel combinations to be explored.

Disclosures

Rossi:Gilead: Honoraria, Research Funding; Abbvie: Honoraria; Janseen: Honoraria. Sloss:Immunogen Inc: Employment.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution