ALBERT

Description: Ibrutinib is a promising targeted therapy for chronic lymphocytic leukemia (CLL). However, a small subset of patients progress on ibrutinib either through progressive CLL or Richter's transformation. Patients responding to ibrutinib and then progressing with Richter's transformation do so most commonly within the first 2 years of treatment and have an extremely poor prognosis. Identifying biomarkers associated with this transformation is of utmost importance. Near-tetraploidy (4 copies of most chromosomes within a cell) has been reported in various lymphomas; however, its incidence in CLL has not been described. We investigated the prevalence of near-tetraploidy in CLL patients prior to starting ibrutinib and identified it as a pre-treatment biomarker for Richter's transformation. We examined near-tetraploidy in a large series of CLL patients enrolled across four ibrutinib clinical trials at the Ohio State University, for which extensive correlative studies and follow up data are available (previously described by Maddocks et al., JAMA Oncol, 2015). We identified this abnormality in 9 of 300 patients (3.0%, 95% CI: 1.4-5.6) in blood or bone marrow samples taken prior to starting therapy. Near-tetraploidy was detected by the presence of four signals with four or more fluorescence in situ hybridization (FISH) probes and confirmed in the metaphase karyotype of each patient in at least one cell. Near-tetraploidy was associated with aggressive disease characteristics including: Rai stage 3/4 (p=0.03), deletion 17p (p=0.03), and complex karyotype (p=0.01), as well as trisomy 12 (p=0.05). With a median follow-up time of 40.5 months, in patients positive with near-tetraploidy, one patient (11%) progressed with CLL on ibrutinib, six patients (67%) progressed with Richter's transformation, and two patients (22%) were still on treatment. Cumulative incidence of Richter's transformation was significantly higher in patients with near-tetraploidy (Figure; p

Description: Chronic Lymphocytic Leukemia (CLL) has a varied clinical course; some patients experience a long survival and others succumb to disease in a short time. Clinical factors correlated with either time to first treatment (TFT) and/or overall survival include Rai stage, IGHV somatic hypermutation status, fluorescence in situ hybridization (FISH) abnormalities, especially del(17p), karyotypic complexity and the presence of a cytogenetic translocation. Previous studies have included patients both at diagnosis and at various times throughout their diseases, and many included limited numbers of patients, precluding extensive analyses of relationships between the prognostic factors and their relative impact on clinical outcome. We sought to identify which factors determined within a short time of diagnosis (i.e., 1 year) were prognostic for TFT in untreated CLL patients. We identified 329 untreated CLL patients who had stimulated karyotypic and FISH analyses within 1 year of diagnosis seen at The Ohio State University (OSU). Patient characteristics and outcome were obtained from patient records. The studies were approved by the OSU IRB and were conducted according to the Declaration of Helsinki. A complex karyotype was defined as ≥ 3 unrelated aberrations by karyotype. Patient characteristics are given in Table 1. Translocations occurred in 87 (26.4%) patients: 38 balanced and 49 unbalanced translocations. Initial statistical analyses showed no large difference in TFT between balanced and unbalanced translations, so they were combined for final analyses. 144 patients (49 with and 95 without a translocation) had unmutated IGHV, and 144 patients (22 with and 122 without a translocation) had mutated IGHV. IGHV data were not available for 41 patients. TFT was calculated from date of diagnosis to date of first treatment. Untreated patients were censored at last known untreated date. Kaplan-Meier curves estimated TFT probability, and proportional hazard models were used to examine the association between potential risk factors and TFT. Using backward selection, variables with statistical significance when adjusting for all other covariates were included in the final model. To evaluate potential effect modifications, pairwise interactions among all the variables in the final model were examined and retained if statistically significant. Stata 14 (College Station TX) was used, and all tests were two-sided with statistical significance set at p