Page 68 - 2021_09-Haematologica-web

C.S. Tam et al.
duration of response (DOR), and safety (frequency and severity of all treatment-emergent adverse events [AE]). ORR was assessed per modified iwCLL criteria for CLL23,27 and per Lugano criteria for SLL.28 Measurable disease, defined as ≥1 lymph node >1.5 cm in the longest diameter and measurable in two perpendicular diame- ters, was assessed by computed tomography/magnetic resonance imaging. Response assessments were performed every 12 weeks after the first dose day for 96 weeks, then every 24 weeks until PD or initiation of new, nonprotocol therapy, whichever came first. Patients underwent bone marrow examination at baseline and for confirmation of complete response (CR), or CR with incomplete hematologic recovery (CRi), or if PD was suspected due to cytope- nia.
All treatment-emergent AE, including AE of interest (AEI) based on the known toxicity profile for BTK inhibitors occurring on or after day 1 until 30 days after treatment discontinuation were summarized. AEI were categorized in accordance with predefined MedDRA search criteria (Online Supplementary Table S1). AE sever- ity was assessed using the National Cancer Institute Common Terminology Criteria for Adverse Events v4.03 and the Grading Scale for Hematologic Toxicity in CLL Studies.23 Biomarkers were assessed at baseline and optionally at progression.
Statistical analyses
Primary efficacy and safety analyses included all patients with centrally confirmed del(17p) CLL/SLL receiving ≥1 dose of zanubrutinib. ORR was summarized as percentage of responders (CR, CRi, nodular partial response [nPR], partial response [PR], or PR with lymphocytosis [PR-L]) with corresponding 95% Confidence Interval (CI). An evaluation of ORR in subgroups defined by key demographic and baseline disease characteristics was conducted and summarized in a forest plot. DOR was defined as time from first response until PD or death due to any cause. PFS was measured from time of first dose to PD or death due to any cause. Median DOR, PFS, and event-free survival rates were esti- mated using Kaplan-Meier methodology with corresponding 95% CI.
Results
Patient characteristics and disposition
Between February 3, 2018 and February 20, 2019, 109 patients with centrally confirmed del(17p) CLL/SLL were enrolled from 59 sites in 13 countries (Online Supplementary Table S2) and received ≥1 dose of zanubru- tinib. Two additional patients without del(17p) were assigned in error to this study arm and are not included in the analysis. At the data cutoff date of April 15, 2020, the median duration of study follow-up was 18.2 months (range, 5.0–26.3). Median age at study entry was 70.0 years (range, 42–86); the majority of patients present- ed with CLL (90.8%). Reported reasons for treatment per iwCLL criteria in order of frequency included progressive marrow failure (41.3%); massive, progressive, or sympto- matic lymphadenopathy (41.3%); significant fatigue (33.0%); night sweats (32.1%); progressive lymphocytosis with increase of > 50% over 2 months or doubling time of <6 months (28.4%); massive, progressive, or symptomatic splenomegaly (23.9%); and unintentional weight loss (14.7%). More than one reason for treatment may have been given, with a median of two reasons given for each patient. Many patients had other high-risk disease charac- teristics, including 40 patients with CLL who presented as Binet stage C (40.4%), 42 patients with bulky disease ≥5
cm (38.5%), 78 patients with elevated β2-microglobulin (78.8% of 99 patients with available data), and 67 patients with an unmutated immunoglobu- lin heavy chain variable (IGHV) locus (65.0% of 103 patients with sufficient RNA for testing). Furthermore, 32 (37.2%) of 86 patients with sufficient metaphases avail- able for analysis had at least three distinct karyotypic abnormalities defined as complex karyotype (Table 1).
Efficacy
The ORR was 94.5%, which included three patients (2.8%) with CR, one patient (0.9%) with CRi, 95 (87.2%) with PR, and four (3.7%) with PR-L (Table 2). Five (4.6%) patients had a best response of stable disease (SD). One patient (0.9%) had PD at the first response assessment. Five (4.6%) patients met clinical CR criteria but did not undergo bone marrow biopsy. Ninety-seven patients (89.0%) remained on treatment at the time of analysis. Nine patients with an initial response (8.3%) progressed on study, four of whom had histologically confirmed Richter transformation. Median time to transformation was 13.7 months (time to transformation for each patient: 3.9, 13.6, 13.8, and 15.7 months). Two other patients had new lesions seen on CT with positron emission tomogra- phy (PET) avidity for which biopsy could not definitively confirm transformation, while one other patient had accelerated CLL. Seven patients who have progressed have discontinued treatment; two other patients with pro- gression remained on treatment at time of data cutoff. Four patients who progressed have died; two due to pro- gression, one due to an adverse event after progression (acute kidney injury), and one after progression due to septic shock. Four patients (3.7%) discontinued treatment due to AEs, of whom 2 have died (see Safety below), while one patient discontinued treatment after withdraw- al of consent and was lost to follow up.
Median PFS and OS were not reached. The estimated 18-month PFS rate was 88.6% (95% CI: 79.0–94.0) (Figure 1A), while the estimated 18-month OS rate was 95.1% (95% CI: 88.4–98.0) (Figure 1B). The median time to response was 2.8 months (range, 1.9–16.5) The median DOR was not reached; 92.8% of patients had a DOR ≥ 12 months (Figure 1C). ORR was consistent across all pre- specified demographic and baseline disease characteristics (Figure2).
Transient treatment-related lymphocytosis was observed (Online Supplementary Figure S1A), but there was a significant reduction in target lesion size by the first scheduled response assessment, consistent with the short median time to response (Online Supplementary Figure S1A). The peak median change in absolute lymphocyte count (ALC) and time to resolution of lymphocytosis both appeared to be decreased from previous experience with zanubrutinib.19 In an exploratory analysis, changes in ALC were compared between patients with mutated and unmutated IGHV locus. Patients with unmutated IGHV showed a slight trend towards less treatment-related lym- phocytosis (Online Supplementary Figure S1B), similar to previous reports.9,29
Sixty-one patients (56.0%) began the study with at least one cytopenia (Table 1), including 43 patients with anemia (39.4%), eight patients with neutropenia (7.3%), and 28 patients with thrombocytopenia (25.7%) (Online Supplementary Table S3). Eleven patients (10.1%) received at least one dose of a granulocyte colony-stimulating fac-
2356
haematologica | 2021; 106(9)