Biomarkers and In Vivo Responses to the BH3 Mimetic, ABT-263, in Patients (pts) with Chronic Lymphocytic Leukemia (CLL).

Abstract 2374

Poster Board II-351

ABT-263 is a first-in-class BH3 mimetic inhibitor of 3 pro-survival members of the BCL2 protein family (BCL2, BCLX_L, BCLW). Consistent with potent activity against BCL2-overexpressing cell lines and primary CLL cells in vitro, and BCL2-overexpressing lymphoid murine tumors in vivo, ABT-263 demonstrated significant antitumor activity in pts with relapsed, refractory CLL and small lymphocytic lymphoma (SLL) in 2 phase 1/2a studies. To date, 42 pts (35 evaluable) with CLL/SLL have been treated with ABT-263 (40-440 mg/d). 51% (18/35) with a baseline lymphocytosis >5,000 have achieved '50% reduction in lymphocytosis, while 34% (12/35; 5 bulky disease) with measurable nodal disease at initiation have achieved a partial response. Overall, some manifestation of antitumor activity was observed in 60% (21/35). Responses tend to be durable, with the median PFS not reached as yet while the median time on study for M06-873 pts is 9 mos for pts on the M06-873 study. Responses were observed in pts who received >4 lines of prior therapy and in those who were fludarabine refractory and/or had bulky disease. BCL2 is highly expressed in all CLL/SLL, yet only a subset of pts responded. We tested whether additional intrinsic biological characteristics of CLL were associated with response to ABT-263 in vivo. Potential biomarkers, both standard (FISH for 17p13, 11q22.3) and investigational (in vitro sensitivity; quantitative expression of BCL2, MCL1, BIM, BAX) were measured at study entry, during therapy and where possible at progression in a subset of pts treated in the phase 1 study in CLL (M06-873). Of 21 pts receiving >40mg/d ABT-263 for >7d, FISH data for 17p13 and 11q22.3 were available for 16, of which only 5 were normal; 4 and 6 had deleted 17p13 or 11q22.3 respectively, and 2 had deletions for both in >5% of cells examined. Similar majorities of pts with del17p13 (4/6) or del11q22.3 (5/8), or with neither abnormality (4/5), achieved either a >50% fall in peripheral blood lymphocytes, reduction in nodal size, or both. None of the pts without del17p13 or del11q22.3 have progressed after median 285 d (range 167-484 d) on drug, while only 1/6 with del17p13 and 2/8 with del11q22.3 have progressed (all 3 within 3 mos). CLL cells from 12 pts were tested for in vitro sensitivity to ABT-737, a BH3 mimetic with the same specificity and activity in vitro as ABT-263. LC₅₀ were all <50 nM at baseline (mean 5.5±5.1 nM). LC₅₀ at baseline correlated inversely with % reduction of lymphocytosis at both Cycle1/Day 14 (C1D14) and C3D1 (r²=-0.51 & 0.44, p=0.01 & 0.03, respectively), although in vitro LC₅₀ was not predictive of whether a partial response was ultimately achieved. Reassessment of in vitro LC₅₀ after 14 days on ABT-263 and at C3D1 revealed a modest reduction in sensitivity of residual CLL cells: C1D14 12±9.4 nM, C3D1 11±9.6 nM (p<0.05 for difference between baseline and later timepoints; repeated measures ANOVA). Expression levels of BCL2, BIM and BAX did not correlate with % reduction in lymphocytosis at either C1D14 or C3D1 in vivo responses. Higher basal expression of MCL1 was negatively correlated with % reduction in lymphocytes after 2 cycles (n=8; r² = -0.55, p=0.03). When MCL1 expression was measured in residual CLL cells collected at C1D14 or C3D1, no change from baseline was observed. These preliminary data indicate that ABT-263 is active in pts with relapsed refractory CLL carrying adverse genetic markers, and that expression patterns of BCL2 family member proteins do not strongly predict response to this drug. While in vitro LC₅₀ did correlate with the degree of initial fall in lymphocytosis, residual cells after 2 treatment cycles were only modestly less intrinsically sensitive in vitro, expressing higher levels of MCL1. Additional factors, presumably extrinsic to the CLL cells appear to significantly influence in vivo responses to ABT-263.