Leukemia cell proliferation and death in chronic lymphocytic leukemia patients on therapy with the BTK inhibitor ibrutinib

Abstract

BACKGROUND. Ibrutinib is an effective targeted therapy for patients with chronic lymphocytic leukemia (CLL) that inhibits Bruton's tyrosine kinase (BTK), a kinase involved in B cell receptor signaling. METHODS. We used stable isotopic labeling with deuterated water (2H2O) to measure directly the effects of ibrutinib on leukemia cell proliferation and death in 30 patients with CLL. RESULTS. The measured average CLL cell proliferation ("birth") rate before ibrutinib therapy was 0.39% of the clone per day (range 0.17%-1.04%); this decreased to 0.05% per day (range 0%-0.36%) with treatment. Death rates of blood CLL cells increased from 0.18% per day (average, range 0%-0.7%) prior to treatment to 1.5% per day (range 0%-3.0%) during ibrutinib therapy, and they were even higher in tissue compartments. CONCLUSIONS. This study provides the first direct in vivo measurements to our knowledge of ibrutinib's antileukemia actions, demonstrating profound and immediate inhibition of CLL cell proliferation and promotion of high rates of CLL cell death. TRIAL REGISTRATION. This trial was registered at clinicaltrials.gov (NCT01752426). FUNDING. This study was supported by a Cancer Center Support Grant (National Cancer Institute grant P30 CA016672), an NIH grant (CA081554) from the National Cancer Institute, MD Anderson's Moon Shots Program in CLL, and Pharmacyclics, an AbbVie company.

Conflict of interest statement

K.W. Li, M.K. Hellerstein, S.M. Turner, and C.L. Emson are employees of or consultants to KineMed Inc. J.A. Burger, N. Chiorazzi, and S. O’Brien received research funding from Pharmacyclics, and J.A. Burger received speaking fees from Janssen Pharmaceuticals.

Figures

Figure 1. Heavy water labeling in CLL patients before ibrutinib therapy: workflow and trial outline.

(A) Workflow diagram to illustrate the principles of the study. Patients ingest heavy water (2H2O), and deuterium (2H) becomes incorporated into the DNA of dividing cells, including dividing CLL cells. Repeated blood draws and purification of CLL cells, followed by DNA isolation, allows quantification of the fraction of labeled DNA in the CLL cells at each time point via gas chromatography/pyrolysis/isotope ratio–mass spectometry. (B) The three phases of the trial — labeling, resting, and treatment — and the duration of each are indicated on the horizontal axis (months, days); 30 patients participated in all three phases of the trial. Blood samples were drawn every 2 weeks at the days indicated by the arrows. Ibrutinib treatment started after the resting period with 420 mg orally daily. Patients benefiting from ibrutinib were allowed to continue on treatment.

Figure 2. Ibrutinib inhibits CLL cell birth and permits increased CLL cell death.

(A) 2H2O enrichment data from the plasma of CLL patients (n = 30), showing enrichment of 2H2O during the first 4 weeks of ingestion, followed by dilution of 2H2O during the washout phase, which fits an exponential decay model. (B) Percentage of 2H enrichment in the DNA of peripheral blood CLL cells from the 30 CLL patients studied was measured and converted into a fraction (f) of newly divided cells, as described in Methods. Data are plotted for the first 12 weeks of the study, comprising the first 4 weeks of labeling, followed by washout. (C) After starting ibrutinib therapy, there is a plateau in the proportion of labeled CLL cells, indicating an arrest of CLL cell birth based on the absence of dilution in f (the fraction of previously divided, labeled CLL cells) by newly divided, unlabeled CLL cells. (D) The increase in ALC before the start of ibrutinib therapy is due to birth rates that are higher than death rates (see F). (E) After the start of ibrutinib, the ALC initially increases briefly, due to redistribution lymphocytosis and then continuously declines. (F) Calculated birth and death rates before start of ibrutinib therapy (pre-dose) and after the start of ibrutinib therapy (post-dose) demonstrate profound inhibition of CLL proliferation in addition to accelerated CLL cell death and ALC clearance after the start of ibrutinib.

Figure 3. Peripheral blood cell counts, bone marrow infiltration, and survival during ibrutinib treatment.

(A) Median ALC (±SEM) before and after starting ibrutinib therapy (black arrow: start of labeling phase; gray arrow: start of ibrutinib therapy). The trend of the ALC shows a characteristic transient increase after the start of ibrutinib therapy due to redistribution of tissue CLL cells into the peripheral blood. (B) Redistribution lymphocytosis was seen only in patients with mutated IGHV (M-CLL, n = 11); in contrast, redistribution lymphocytosis was absent patients with unmutated IGHV (U-CLL, n = 17). (C) Median hemoglobin levels and (D) platelet counts (±SEM) normalize during ibrutinib therapy. The gray shaded areas indicate normal values. (E) There was a continuous improvement in bone marrow infiltration by CLL cells after starting ibrutinib therapy, indicated as the percentage of bone marrow lymphocytes that are displayed for each patient together with the median (±SEM, n = 29). (F) When comparing patients with M-CLL with patients with U-CLL, there was faster bone marrow clearance of leukemia cells in patients with U-CLL. (G) No progression events were seen in the entire patient population after a median follow-up of 26 months. (H) Only one CLL- and treatment-unrelated death occurred; the remaining patients are alive and almost all (27 of 30) continue on ibrutinib therapy. PFS, progression-free survival; OS, overall survival.

Figure 4. Volumetric analyses of CLL lymph node and spleen manifestation before and after 2 months of therapy with ibrutinib.

CT images from a patient with CLL from our series with superimposed reconstruction of main areas of CLL involvement highlighted in color. The volumes (in cubic centimeters) of the axillary (green), mesenteric (blue), other intraabdominal (red), and spleen (orange) disease sites are displayed next to each involved area. (A) Axillary lymph nodes; (B) spleen; and (C) abdominal lymph nodes.

Figure 5. Modeling and estimating CLL kinetics.

(A) A 2-compartment mathematical model (given by ordinary differential equations) was used to estimate CLL cell death rates in tissue and blood, according to previously established methods. CLL cells in the tissue are assumed to die with a rate d1 and redistribute to blood with a rate m. CLL cells in the blood are assumed to die with a rate d2. The parameter c is included to account for the observation that the rate of CLL cell decline during ibrutinib treatment slows down over time and converges to a plateau. In accordance with data, it is assumed that no meaningful amount of cell proliferation and homing occurs during therapy. (B) Two examples of the treatment responses to ibrutinib. Circles show measured ALC in the blood, and the line is the best model fit to the data. The top graph is a typical patient with “slow” disease clearance, with a pronounced lymphocytosis phase that lasts for about 3 weeks. The bottom graph is a typical “fast” responder, with limited lymphocytosis that only lasts for a few days and a subsequent rapid decline. (C) Distribution of tissue death rates of tumor cells among patients, separated according to M-CLL (blue) and U-CLL (red). Each symbol represents a patient. The death rate is expressed both as the average life span of CLL cells in a patient (left), and as the percentage of CLL cells that die per day (right). (D) Average tissue decline dynamics among patients with M-CLL (blue) and U-CLL (red), as predicted by the parameterized mathematical model. The computer simulation of the model was run for individual patients, and the predicted tissue sizes were averaged and plotted every 30 days. The tissue size before start of treatment is given by 100, and the other sizes are scaled accordingly. For the purpose of comparison to previous studies, we plotted the predicted 2D tissue size, which scales with the number of cells to the power of 2/3. Note that this assumes complete correspondence between tissue size shrinkage and the decline of cell numbers in tissue, which should be reasonably accurate for large tissue sizes. Deviations are expected at smaller tissue sizes.