Dynamic BH3 Profiling Predicts Patient Response and MRD Status in Chronic Lymphocytic Leukemia (CLL) Patients Undergoing Frontline Treatment with Kinase Inhibitor Plus FCR (KI+FCR)

Background

We previously found that the combination of ibrutinib (IBR) or duvelisib (DUV) with FCR leads to a high rate of MRD undetectability when given as frontline therapy for younger, fit CLL patients (Davids et al., ASH 2017 and EHA 2018), a strategy we call KI+FCR. However, not all patients achieve deep response, and identifying predictive biomarkers may further improve efficacy rates while not exposing patients unlikely to benefit from FCR to the risks of this therapy. Dynamic BH3 profiling (DBP) is a functional assay that measures the net pro-apoptotic signaling induced in a cell in response to a drug treatment (delta priming). Here, we evaluated whether DBP could predict the likelihood of achieving MRD undetectability on two KI+FCR trials.

Methods

Mononuclear cells were isolated by Ficoll separation from peripheral blood (PB) and bone marrow (BM) of CLL patients in two different investigator-initiated trials of frontline therapy in patients age ≤ 65: IBR+FCR (iFCR, n=30) or DUV+FCR (dFCR, n=25). Samples were isolated pre-treatment and after one week of KI monotherapy (just prior to the first cycle of KI+FCR). BH3 profiling was performed on untreated cells by measuring %cytochrome C release in cells exposed to BH3-only peptides or small molecules targeting pro-survival BCL-2 family proteins. Untreated CLL cells were also co-cultured with NK.TERT stromal cells (1:10 ratio) with 1 µM drug treatments, and DBP was performed as previously described (Montero et al., Cell, 2015) to measure the net pro-apoptotic signaling induced by each drug. Venetoclax (VEN) was used to measure BCL-2 dependence, and the BH3-only peptides HRK, MS1, and FS1 were used to measure BCL-xL, MCL-1, and BFL-1 dependence, respectively. BIM BH3 peptide was used to measure overall mitochondrial priming for apoptosis. Clinical response was assessed by 2008 IW-CLL criteria.

Results

In ex vivo experiments, pre-treatment PB-derived CLL cells showed similar increased BCL-2 dependence after 24 hrs of treatment with IBR or DUV (IBR, n=22, delta priming= +11.6%, CI=5.5-17.7, p= 0.0007 and DUV, n=13, delta priming= +8.4%, CI=1.1-15.7, p=0.0274). Similarly, pre-treatment BM-derived CLL cells treated ex vivo with IBR or DUV for 24 hrs. had increased BCL-2 dependence (IBR, n=13, delta priming= +15.4%, CI= 9.1-21.7, p=0.0002 and DUV, n=13, delta priming= +9.3%, CI =1.5-17.2, p=0.0235). No significant changes were observed in overall priming for apoptosis, nor in BCL-XL, MCL-1, or BFL-1 dependence.

In in vivo experiments performed on CLL cells from patients treated with one week of IBR or DUV monotherapy prior to combination with FCR, no significant change in BCL-2 dependence was observed in aggregate comparing the pre-treatment and one week post-treatment PB-derived CLL cells. However, PB-derived CLL cells from patients in both trials who later went on to achieve complete response (CR) had significantly higher delta priming for BCL-2 than those achieving a partial response (PR) as best response. (Fig. 1: iFCR, CR, n=20, mean delta priming= +9.9%, PR, n=10, mean delta priming= -6.2%, CI= -27.6 to -4.55 and p = 0.0084, dFCR, CR, n=10, mean delta priming=+12.1%, PR, n=9, mean delta priming= -1.7%, CI= -26.6 to -1.02, p = 0.0359). As in the ex vivo experiments, no significant alteration in overall priming or in dependence on other anti-apoptotic proteins was observed.

Pre-treatment PB-derived CLL cells from patients treated with dFCR who reached BM-MRD undetectability (MRD-) also had significantly higher baseline BCL-2 dependence than those who were BM-MRD detectable (MRD+) as best response (MRD-, n=11, priming=+36.9%, MRD+, n=5, priming=21.7%, p=0.0377). Similar trends were observed with iFCR, though these results did not reach statistical significance, possibly due to the small number of MRD+ patients (n=5).

Conclusions

Ex vivo treatment of CLL cells with IBR or DUV leads to increased BCL-2 dependence, and the level of in vivo delta-priming for BCL-2 dependence predicts depth of response to KI+FCR based frontline therapy in CLL. Additional experiments are ongoing to better define the mechanism underlying the altered anti-apoptotic dependencies seen with KI therapy in CLL cells. DBP should be further evaluated for its potential as a clinical decision-making tool to identify those young, fit CLL patients who are the best candidates for frontline KI+FCR therapy and who may have a more favorable risk/benefit ratio from a novel agent only regimen.

View largeDownload slide

View largeDownload slide

Close modal