Phase Ib Study of Tirabrutinib in Combination with Idelalisib or Entospletinib in Previously Treated Chronic Lymphocytic Leukemia

In the past two decades, signaling through the B-cell receptor (BCR) has been a major focus of pharmacologic research. Several BCR-targeted agents are now approved for patients with chronic lymphocytic leukemia (CLL), including the Bruton tyrosine kinase (BTK) inhibitors ibrutinib and acalabrutinib (1, 2), idelalisib and duvelisib (PI3K inhibitors), and venetoclax (a BCL2 inhibitor).

Ibrutinib, a first-generation BTK inhibitor, irreversibly inhibits at least 11 other kinases with an IC₅₀ of approximately 11 nmol/L or less: BTK, BLK, BMX, CSK, FGR, BRK, HCK, EGFR, YES, ErbB2, and ITK (3, 4). In addition, ibrutinib reversibly binds to other kinases that lack the active site cysteine residue, in some cases with comparable affinity as binding to BTK (3–5). Binding of ibrutinib to these additional kinases may potentially be responsible for side effects such as rash, diarrhea, bleeding, and atrial fibrillation (6–9).

Selective BTK inhibitors have been developed to minimize these off-target effects. Tirabrutinib (formerly ONO/GS-4059) is a selective, irreversible, second-generation, small-molecule BTK inhibitor (5, 10). Tirabrutinib irreversibly binds to C481 of BTK with greater target selectivity. In a recent study, the IC₅₀ values for inhibition of kinases by tirabrutinib were: BTK, 6.8 nmol/L; BMX, 6 nmol/L; BLK, 300 nmol/L; TEC, 48 nmol/L; EGFR, 3020 nmol/L; ErbB2, 7313 nmol/L; and ITK, >20,000 nmol/L. Tirabrutinib showed a 440-fold and >2,940-fold selectivity for BTK over EGFR and ITK, respectively (11). In a phase I dose-escalation trial, tirabrutinib demonstrated an overall response rate (ORR) of 96% in 28 patients with relapsed/refractory (R/R) CLL, with no MTD identified up to 600 mg every day (5). Long-term follow-up of that study (for a median of 32.5 months) demonstrated an estimated median progression-free survival (PFS) of 38.5 months and median overall survival of 44.9 months (12). Despite the high ORR achieved with BTK inhibitors, most patients with CLL will progress while on therapy. Thus, new therapeutic strategies are needed to achieve an increased depth of response to improve remission duration and reduce the emergence of resistant subclones, without additional toxicities.

The BCR signaling cascade consists of divergent signaling pathways. BCR cross-linking leads to activation of proximal tyrosine kinases, including BTK, PI3K, and spleen tyrosine kinase (SYK). The signal is further transmitted through multiple downstream mediators, leading to upregulation of antiapoptotic proteins such as MCL1 and BCL2 (13, 14). The potential benefit of combining agents targeting BTK, PI3K, and SYK has been well documented in CLL and non-Hodgkin lymphoma (NHL) preclinical models. Concurrent targeting of BTK and PI3K in a murine CLL model led to improved survival and reduction in tumor burden compared with either agent alone (15). Additive or synergistic effects of the combination of ibrutinib and idelalisib have been reported in CLL, mantle cell lymphoma (MCL), and diffuse large B-cell lymphoma (DLBCL) cell lines (16, 17). Combined inhibition of PI3K and BTK induced apoptosis of the DLBCL cell line TMD8, which was resistant to inhibition of either kinase alone (18). Concurrent inhibition of alternative kinases may rely on unique mechanisms: unlike other BCR-signaling inhibitors, the selective reversible inhibitor of SYK entospletinib led to downregulation of MCL1 in CLL cells in microenvironment-mimicking conditions in vitro, thereby interrupting prosurvival signaling (13). Given this data, several clinical trials combining BTK and PI3K inhibitors in B-cell malignancies are ongoing. These include a phase I study of umbralisib + ibrutinib in patients with R/R CLL (19), a phase I study of umbralisib, ublituximab, and ibrutinib in patients with CLL and NHL (20), and a phase I/II study of pan-PI3K inhibitor copanlisib and ibrutinib in patients with MCL (NCT03877055).

These data provide a rationale for exploring combined inhibition of multiple kinases in the BCR pathway for treatment of CLL. In this study we evaluated the safety, tolerability, and preliminary efficacy of the combinations tirabrutinib/idelalisib and tirabrutinib/entospletinib, as well as tirabrutinib monotherapy, in patients with R/R CLL.

This was a phase Ib, open-label, multicenter, sequential, dose-escalation and dose-expansion study (NCT02457598) conducted in the United States, the United Kingdom, and France. Patients with CLL reported here represent one histologic cohort of a larger study (Supplementary Fig. S1) evaluating tirabrutinib combinations in subjects with relapsed or refractory NHL. This article reports safety and efficacy data only in patients with CLL. Eligible patients were age ≥18 years with Eastern Cooperative Oncology Group (ECOG) performance status ≤2 and either documented disease progression or stable disease on the most recent of ≥1 chemotherapy- or immunotherapy-based CLL treatment regimen, and no prior exposure to AKT, BTK, PI3K, JAK, mTOR, or SYK inhibitors (see Supplementary Table S1 for more details). Institutional review boards at each of the study sites approved the protocols. All patients provided written informed consent. This study was conducted in accordance with the Declaration of Helsinki.

A standard 3+3 dose-escalation schema was followed (Supplementary Table S2). Patients receiving the tirabrutinib/idelalisib combination were treated on 28-day cycles with either idelalisib 50 mg twice a day or 100 mg every day and tirabrutinib ranging from 20 to 160 mg every day. Patients receiving the tirabrutinib/entospletinib combination were treated with either entospletinib 200 mg or 400 mg every day and tirabrutinib ranging from 40 to 150 mg every day. Tirabrutinib monotherapy was given at 80 mg every day. Patients received a single dose of tirabrutinib on cycle 1, day 1, before initiating idelalisib or entospletinib in combination with tirabrutinib on cycle 1, day 2 (or continuation with tirabrutinib monotherapy). After completion of study treatment, patients attended a 30-day safety follow-up visit.

Determination of CLL response and progression was based on the 2008 standardized International Workshop on CLL (iwCLL) Criteria, which were current at the time the study protocol was finalized (21). Patients with CLL had CT or MRI scans performed at baseline, at 24 weeks, and at the time of progression. Bone marrow aspirates were collected at the time of suspected complete response (CR) for minimal residual disease (MRD) testing using the CLL ERIC MRD flow cytometry panel at Covance/Labcorp, Indianapolis, IN, based on Rawstrom and colleagues, 2007, and Rawstrom and colleagues, 2013 (22, 23). For prognostic biomarkers, peripheral blood was collected prior to therapy at cycle 1, day 1, at disease progression, and at the time of suspected CR for MRD testing.

The primary endpoint of the dose-escalation phase was safety, evaluated by the occurrence of adverse events (AE) and laboratory abnormalities defined as dose-limiting toxicities (DLT; see Supplementary Table S3). The disease and dose chosen for expansion cohorts were based on emerging safety, pharmacokinetic, and pharmacodynamic results of the dose-escalation phase. In the dose-expansion phase (Supplementary Table S4), the primary endpoint was ORR, defined as the proportion of patients who achieve a CR (including those with CR with undetectable MRD), partial response (PR), or PR with lymphocytosis (24). Secondary endpoints included PFS, duration of response, time to response, proportion of subjects who achieve undetectable MRD (defined as <1 leukemia cell/10,000 leukocytes), and pharmacokinetic parameters. Sum of the products of greatest perpendicular lesion diameters (SPD) change from baseline was evaluated as an exploratory endpoint.

Blood samples were collected at protocol prespecified sampling times. Patients in the dose-escalation cohorts underwent intensive pharmacokinetic sampling on day 1, day 2, and day 8 of the first treatment cycle to assess potential drug–drug interactions between tirabrutinib and idelalisib or entospletinib. Dose-escalation cohorts enrolled subjects with various B-cell malignancies, with the exception of the last dose-escalation cohort/highest tirabrutinib dose (160 mg) that enrolled DLBCL subjects only. Patients with CLL in the dose-expansion cohorts had sparse pharmacokinetic samples collected at predose and 1.5–4.0 hours postdose of tirabrutinib and entospletinib or idelalisib throughout the study.

Plasma concentrations of tirabrutinib, entospletinib, and idelalisib were determined using validated bioanalytic assays. Pharmacokinetic parameters were estimated by standard noncompartmental methods using Phoenix WinNonlin 7.0 Software (Certara). Pharmacokinetic parameters and concentrations were summarized using descriptive statistics.

A panel of genetic aberrations common in CLL was assessed using next-generation sequencing [NGS, CGI Focus CLL panel (NGS mutation panel consisting of 7 genes: TP53, ATM, BIRC3, NOTCH1, SF3B1, CARD11, and MYD88)], IGVH mutational status analysis, and FISH (both performed at Cancer Genetics, Rutherford, NJ). The following FISH probes were used: 11q22.3 (ATM); 17p13 (TP53); CEP12; 13q14(D13S319)/13q34; CEP6/6q23 (c-MYB); and t(11;14)(CCND1/IGH). Cytogenetic risk was categorized as high for patients with TP53 aberrations (deletions and/or mutations in the TP53 gene determined by NGS or FISH panels), and standard risk for those with no detectable TP53 aberrations.

BTK occupancy was evaluated in a duplexed, homogeneous time-resolved fluorescence resonance energy transfer assay that measures total and free BTK in peripheral blood mononuclear cells (25).

Analysis results are presented using descriptive statistics. For categorical variables, the number and percentage of subjects in each category are presented; for continuous variables, this may include the number of subjects (n), mean, SD or SE, median, first quartile (Q1), third quartile (Q3), minimum, and maximum. Best overall response was defined as the best response recorded from the start of treatment until progressive disease (PD)/recurrence. ORR is defined as the proportion of subjects who achieve CR or PR during the study based on iwCLL 2018 criteria (24). PFS was analyzed by Kaplan–Meier methods and defined as the interval from the start of the study therapy to the earlier of the first documentation of definite disease progression (radiographic or clinical progression) or death from any cause.

The follow-up time for PFS was summarized using descriptive statistics and defined as the interval from the study therapy start date to the last follow-up date. For patients who were lost to follow-up without a PFS event, the last efficacy assessment date was used as the last follow-up date. All others were assigned the data extraction date (February 21, 2019) as the follow-up date.

Anonymized individual patient data will be available upon request to qualified external researchers 6 months after FDA and European Medicines Agency approval per Gilead's Clinical Trial Disclosure & Data Transparency Policy as posted at https://www.gilead.com/research/disclosure-and-transparency.

A total of 53 patients with CLL were enrolled: 29 patients in the tirabrutinib cohort, 14 in the tirabrutinib/idelalisib cohort, and 10 in the tirabrutinib/entospletinib cohort. Baseline demographics are summarized in Table 1. The median number of prior therapies was one in each treatment cohort. All patients in the combination therapy groups and 27 of 29 patients assigned to tirabrutinib had an ECOG performance status of ≤1. There was a higher proportion of patients with Rai stage III–IV in the tirabrutinib/entospletinib cohort (50%), compared with the tirabrutinib/idelalisib (21%) and tirabrutinib monotherapy (31%) cohorts.

As of February 21, 2019, 46 of 53 patients continue to receive treatment on study: 26 patients on tirabrutinib, 10 patients on tirabrutinib/idelalisib, and all 10 patients assigned to tirabrutinib/entospletinib. Median (range) exposures were 67.4 (0.3–104.6), 135.0 (36.0–185.3), and 132.1 (107.6–144.3) weeks in the three treatment cohorts, respectively (Fig. 1; Supplementary Table S5). Three patients discontinued the study in the tirabrutinib group (Supplementary Fig. S2): 1 due to a treatment-emergent adverse event (TEAE, aphasia, deemed not related to study drug), 1 per investigator discretion, and 1 due to progressive disease. Four patients discontinued the study in the tirabrutinib/idelalisib group: 1 due to progressive disease, 2 per investigator discretion, and 1 death. There were no study discontinuations in the tirabrutinib/entospletinib cohort.

No DLTs were observed in patients with CLL receiving either combination, and hence, no MTD was identified in any cohort at the doses evaluated. In each treatment cohort, all patients had at least one TEAE and/or at least one laboratory abnormality. Across all treatment cohorts, the most common TEAEs were diarrhea, constipation, nausea, neutropenia, and contusion (Table 2; Supplementary Table S6). Overall, neutropenia was the most common TEAE and the most common grade ≥3 laboratory abnormality (Table 3).

No patients receiving tirabrutinib or tirabrutinib/entospletinib experienced pneumonitis; however, 2 patients on tirabrutinib/idelalisib developed grade 1–2 pneumonitis. One patient receiving tirabrutinib/idelalisib died; the patient was reported to have stopped breathing while on a long car ride. No autopsy was performed and the cause of death remains unknown. No patients receiving tirabrutinib or tirabrutinib/entospletinib died during the study. There were no cases of Richter transformation in this study.

Diarrhea and nausea were the most common TEAEs with tirabrutinib monotherapy (Table 2), occurring in 9 (31%) patients each. Six (21%) patients had neutropenia. Serious TEAEs occurred in 5 (17%) patients (Supplementary Table S6): these included pneumonia, pneumonia pseudomonal, and pneumonia staphylococcal (3 patients); and aphasia and lower respiratory tract infection (1 patient each). Grade ≥3 TEAEs were reported in 10 (35%) patients, most commonly neutropenia, occurring in 5 (17%) patients.

The most common TEAEs on tirabrutinib/idelalisib combination therapy were diarrhea in 8 (57%) patients, and neutropenia, cough, rash, and bronchitis, occurring in 5 (36%) patients each. Seven (50%) patients had serious TEAEs, the most common being pyrexia, pneumonia, and febrile neutropenia (2 patients each). Grade ≥3 TEAEs were reported in 10 (71%) patients, most commonly neutropenia, which occurred in 5 (36%) patients.

In the tirabrutinib/entospletinib cohort, the most common TEAEs were diarrhea, fatigue, and contusion, occurring in 6 (60%) patients each. Serious TEAEs were reported in 5 (50%) patients, the most common being upper respiratory tract infection (2 patients). Grade ≥3 TEAEs were reported in 7 (70%) patients, most commonly neutropenia in 3 (30%) and upper respiratory tract infection in 2 (20%).

Diarrhea and liver function test abnormalities are recognized complications of therapy with PI3K inhibitors, while hemorrhagic events, atrial fibrillation, and hypertension have been associated with BTK inhibitor therapy. In this study, only 1 patient receiving tirabrutinib/idelalisib experienced a grade 3 AE of diarrhea, which resolved upon dose interruption. There was also 1 patient with a TEAE of diarrhea leading to treatment discontinuation in this study. Increases in alanine aminotransferase and aspartate aminotransferase and decreases in creatinine clearance occurred at higher rates in the combination cohorts. All such TEAEs resolved with treatment interruption.

Across treatment groups, the overall frequency of bleeding/hemorrhage was 53% (based on a broad medical search for terms such as contusion, petechiae, ecchymosis, conjunctival and hemorrhoidal hemorrhage, epistaxis, hematoma, and purpura). One patient receiving tirabrutinib/idelalisib had a grade 3 subdural hemorrhage. Atrial fibrillation was detected in 6% of patients on this study (none with grade ≥3). Meanwhile, the frequency of hypertension was 4%.

All patients were evaluable for response. Among all patients on study, ORR was 88.7% (47/53 patients). Four patients achieved CR as defined by iwCLL criteria: 2 on tirabrutinib and 1 each in the combination therapy groups. Of these, none had undetectable MRD. SPD reduction (best change from baseline) is shown in Fig. 2. Out of the patients with valid baseline and post-baseline SPD measurements, only 1 patient in the tirabrutinib arm and 2 patients treated with the combination of tirabrutinib/idelalisib did not reach a 50% decrease in SPD from baseline as the best response.

The ORR in the tirabrutinib group was 83% (Table 4). Median duration of response was not reached; mean (SD) time to response was 4.6 (1.3) months. Median PFS has not been reached; median (range) follow-up time of PFS was 15.5 (0.0–24.0) months.

In patients treated with the tirabrutinib/idelalisib combination, the ORR was 93%. Median duration of response was 27 months (95% CI, 15–27 months). Mean (SD) time to response was 5.5 (1.1) months. Median PFS was 32 months (95% CI, 8–32 months), and the median (range) follow-up time of PFS was 34 (26, 43) months. Two patients had disease progression on tirabrutinib/idelalisib during the time course of this study.

Patients in the tirabrutinib/entospletinib treatment group had an ORR of 100%. Median duration of response was not reached, and the mean (SD) time to response was 5.8 (0.7) months. Median PFS also was not reached, and the median (range) follow-up time of PFS was 30.4 (24.7–33.2) months.

Patients were classified as high risk if a TP53 mutation and/or a del(17p) was detected, in contrast to patients with no aberration in TP53 who were classified as standard risk. Data were available for all but 4 patients across all three arms. On the basis of the presence of TP53 gene aberrations, 12 patients were categorized as high risk (Fig. 2). Ten of these patients achieved either complete or PR (responders), while 2 were nonresponders (Supplementary Table S7). Furthermore, 37 patients were categorized as standard risk; 34 of these were responders, 3 were nonresponders. Reduction in tumor burden in response to tirabrutinib alone or in combination with idelalisib or entospletinib was observed among patients with wild-type as well as aberrant TP53 (Fig. 2). Across all treatment groups, only 7 patients had IGHV mutations; none of these patients achieved CR. Additional cytogenetic data correlated with treatment response can be found in Supplementary Table S8.

The pharmacokinetics of tirabrutinib was consistent with previous studies (5). No accumulation of tirabrutinib was observed after every day dosing, which is expected given the short tirabrutinib half-life of 4–7 hours (5). Systemic tirabrutinib exposure in patients with CLL is shown in Supplementary Table S9. Tirabrutinib plasma concentrations in CLL subjects were above levels required to inhibit BTK in peripheral blood (∼20 ng/mL protein-adjusted IC₅₀, see Supplementary Fig. S3; ref. 5).

Assessment of drug–drug interactions between tirabrutinib and entospletinib or idelalisib was carried out in the dose-escalation cohorts, where intensive pharmacokinetic sampling was performed (see Patients and Methods). On the basis of these data we conclude that idelalisib and entospletinib did not affect tirabrutinib pharmacokinetics (Supplementary Fig. S3; Supplementary Table S10). Tirabrutinib did not affect entospletinib pharmacokinetics, but increased idelalisib exposures at higher doses (tirabrutinib 160 mg every day; see Supplementary Table S10). Note that patients treated at the tirabrutinib 160 mg every-day dose level were those with DLBCL. This observation is consistent with a potential inhibitory effect of tirabrutinib on the CYP3A metabolizing enzyme and/or the P-glycoprotein transporter for which idelalisib is a substrate (26, 27).

Pharmacodynamic changes were assessed by measuring free and total BTK levels. Among the 25 patients with CLL with evaluable data, the range of measured free BTK at baseline was 31–139 ng/mL. Free BTK levels decreased rapidly on treatment with 19 of 22 patients for whom assay data were available, having no detectable free BTK (lower limit of quantification, 12 ng/mL) 2 hours after the first dose. Free BTK levels were also not detectable in patients at trough. These observations were consistent across tirabrutinib dose levels (20 mg twice a day, 40 mg every day, and 80 mg every day); however, few samples were available at lower doses (Supplementary Table S11). Combining tirabrutinib with idelalisib or entospletinib appeared to have the same effect on the free BTK levels as tirabrutinib alone, although only limited data are available for combination treatment (Supplementary Fig. S4).

BTK inhibition in CLL is associated with an improved rate of PFS compared with standard chemoimmunotherapy regimens (bendamustine plus rituximab; fludarabine, cyclophosphamide, and rituximab; and chlorambucil), particularly in patients with unmutated IGHV (6, 28) and del(17p) (7, 29). Recently the ELEVATE-TN study also showed that for patients with CLL with mutated IGHV, acalabrutinib + obinutuzumab led to significantly improved PFS when compared with obinutuzumab/chlorambucil (2). However, there is a need for therapies for patients with CLL that lead to deeper and longer remission. CRs with single-agent BCR pathway inhibitors are infrequent (<10%; refs. 6, 30–33). Persistent low-level residual disease allows the development of resistance, which can be difficult to treat. Combination therapy has the potential to meet these needs through broader elimination of B-cell clones, increased depth of response, and hence, shortened treatment duration. BTK, PI3K, and SYK are tractable targets within the BCR signaling cascade, and combining inhibitors of these pathways is a logical option to achieve those goals. The results presented here are the first to evaluate the combination of selective BTK inhibition with SYK or PI3K inhibition in patients with R/R CLL.

In this study, the objective was to evaluate the safety and preliminary efficacy of tirabrutinib as monotherapy as well as in combination with entospletinib or idelalisib for patients with R/R CLL. The results demonstrate that tirabrutinib in combination with idelalisib or entospletinib was well tolerated, with no significant potentiation of the previously characterized side effects associated with the individual agents. No DLTs were observed in patients receiving combination treatment, and no MTD was identified in any cohort at the doses evaluated. In previous studies, diarrhea and hepatic toxicity have been prevalent TEAEs observed with idelalisib monotherapy or entospletinib monotherapy, respectively (5, 34). In this study, rates of grade 1–2 diarrhea were 31%, 57%, and 60% with tirabrutinib, tirabrutinib/idelalisib, and tirabrutinib/entospletinib, respectively, but only 1 patient discontinued therapy due to this AE. A study evaluating dual therapy with idelalisib/entospletinib in patients with CLL and NHL found severe treatment-emergent pneumonitis to be a prohibitive toxicity. Pneumonitis is a well-described complication of therapy with PI3Kδ inhibitors and has been reported in clinical trials of idelalisib, duvelisib, and umbralisib (20, 35, 36). Cases of pneumonitis have been reported in patients with CLL treated with ibrutinib (37). Pneumonitis was also reported with the SYK inhibitor fostamatinib (38). No patients receiving tirabrutinib or tirabrutinib/entospletinib experienced pneumonitis; however, 2 patients on tirabrutinib/idelalisib developed grade 1–2 pneumonitis. Therefore, while uncommon, physicians treating patients with CLL should be aware of pulmonary complications that can arise when using novel agents.

These phase I safety findings are significant when put in context with use of currently approved BTK inhibitors. Between 12% and 21% of patients with R/R CLL have stopped therapy with ibrutinib due to AEs when treated in clinical trials (with median follow-up between 9 and 62 months; refs. 29, 33, 39). Furthermore, the overall ibrutinib discontinuation rate was as high as 41% in a real-world study (median follow-up 17 months), with drug-associated toxicity accountable for 50% of the discontinuations among the relapsed patients (8). This is particularly important in CLL because the majority of these patients have multiple comorbidities (40), which are known to negatively impact outcomes following chemoimmunotherapy (41) or treatment with ibrutinib (42). For example, age is a significant independent risk factor for therapy discontinuation (43), and higher comorbidity burden increases the likelihood of drug discontinuation and/or death (42). Consistent with the favorable tolerability of tirabrutinib monotherapy and combinations in this study, we observed low overall discontinuation rates (13%) due to AEs among all patients. Tirabrutinib monotherapy discontinuation was particularly infrequent (6%).

Tirabrutinib pharmacokinetics were consistent with previous reports. Pharmacodynamic data demonstrate that tirabrutinib doses of 40 mg and above lead to full target occupancy, with no detectable free BTK, implying complete inhibition of BTK-mediated signaling. The reduction in free BTK was rapid (within hours of the first dose). Combining tirabrutinib with either entospletinib or idelalisib did not influence the levels of free BTK. An objective response was achieved by 83%, 93%, and 100% of patients in the tirabrutinib, tirabrutinib/idelalisib, and tirabrutinib/entospletinib treatment cohorts, respectively. The benefit of treatment is also evident in that responses to treatment are ongoing—46 patients still continue on study at the time of this report. Median PFS was 32 months in the tirabrutinib/idelalisib cohort and was not reached in patients with CLL treated with tirabrutinib or tirabrutinib/entospletinib. On the basis of the PD data showing that full BTK occupancy was achieved at 40 mg tirabrutinib, and safety and efficacy data from this study in conjunction with that of a previous study (Study ONO-4059POE001, NCT01659255; ref. 5), the recommended phase II dose is 80 mg tirabrutinib every day.

The preliminary efficacy findings of this study are consistent with the results of other early-phase clinical trials of BCR signaling pathway inhibitors. A phase I study of the PI3K inhibitor umbralisib in combination with ibrutinib reported an ORR of 90% in patients (N = 21) with R/R CLL (19). In a combination phase I study of umbralisib, ublituximab (a second-generation anti-CD20 antibody), and ibrutinib in patients with CLL and NHL, ORR was 84% (20). Meanwhile, an ORR of 95% was reported among patients with R/R CLL treated with acalabrutinib, a second-generation BCR signaling inhibitor with relative selectivity toward BTK (31).

Durable responses in patients with TP53 aberrations continue to represent an unmet medical need in the era of targeted therapies. Presence of del(17p) was a statistically significant independent negative predictor of outcomes among patients treated with ibrutinib in a large cooperative group trial (39, 44). In these data, responses were independent of TP53 or del(17p) status. Among the 12 high-risk patients, 10 achieved a response. Thus, consistent with previous evaluation of tirabrutinib as monotherapy, in this study tirabrutinib given as monotherapy as well as in combination with idelalisib or entospletinib appears to benefit high-risk patients with TP53 mutations or del(17p). This is consistent with the previously observed efficacy of BTK inhibition in patients with 17p/TP53 aberrations (7, 29, 31).

The limitations of this study include the short follow-up and the small number of patients, which make comparisons between treatment groups difficult. While a panel of mutations and other aberrations was assessed, because of the limited number of patients and the multiple combinations of these aberrations per patient, additional conclusions on the efficacy of tirabrutinib in patients with specific aberrations besides TP53mt/17pdel could not be drawn.

BTK and in particular PI3K inhibition may increase genomic instability in preclinical models through enhanced expression of activation-induced cytidine deaminase, an enzyme involved in class switch recombination of the immunoglobulin genes (45). It is possible that long-term therapy with BTK inhibitors alone or in combination with PI3K inhibitors may result in increased incidence of secondary cancers through this mechanism. This is particularly relevant in patients with CLL who demonstrate an increased risk of secondary malignancies due to underlying immune dysregulation (46). On the other hand, treatment with BTK inhibitors may lead to the reversal of the immunosuppressive state in CLL (47), potentially enhancing antitumor immunity. Long-term follow-up will be needed to fully evaluate the risks of secondary malignancies among patients treated with BCR signaling inhibitors.

While the preliminary efficacy data are promising, the reported combinations have not resulted in attaining rates of deeper responses that were hoped for in CLL. Whether a longer-term follow-up of these patients will result in higher rates of CRs or undetectable MRD remains to be seen. A potential approach to further deepen responses is to add an anti-CD20 regimen with the current combinations. Phase II studies of triple-combination therapy to evaluate tirabrutinib/entospletinib ± obinutuzumab (NCT02983617) and tirabrutinib/idelalisib ± obinutuzumab (NCT02968563) in patients with R/R CLL are currently underway. Overall, we report the favorable safety profile, low rates of discontinuation, and promising preliminary efficacy data of tirabrutinib both as monotherapy and in combination with other BCR signaling pathway inhibitors.

A.V. Danilov is an employee/paid consultant for Gilead Sciences, Pharmacyclics, AstraZeneca, Verastem Oncology, Beigene, Bayer Oncology, Genentech, TG Therapeutics, Celgene, Rigel Pharm, and Seattle Genetics, and reports receiving commercial research grants from Gilead Sciences, Aptose Bioscences, AstraZeneca, Takeda Oncology, Verastem Oncology, and Bayer Oncology. H.S. Walter is an employee/paid consultant for and reports receiving speakers bureau honoraria from AbbVie, and reports receiving other commercial research support from Janssen, Gilead Sciences, and AbbVie. P. Hillmen reports receiving commercial research grants from Gilead Sciences, Janssen, AbbVie, Roche, and Pharmacyclics, and speakers bureau honoraria from Janssen and AbbVie. M.J.S. Dyer reports receiving speakers bureau honoraria from Kite/Gilead Sciences. S.S. Mitra, R. Humeniuk, Z. Zhou, and J.M. Jürgensmeier are employees/paid consultants for Gilead Sciences. P. Bhargava is an employee/paid consultant for and holds ownership interest (including patents) in Gilead Sciences. No potential conflicts of interest were disclosed by the other authors.

Conception and design: P. Hillmen, S.A. Rule, R. Humeniuk, J.M. Jürgensmeier

Development of methodology: A.V. Danilov, S.S. Mitra, P.C. Yi, J.M. Jürgensmeier

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): A.V. Danilov, C. Herbaux, H.S. Walter, P. Hillmen, S.A. Rule, E.A. Kio, L. Karlin, M.J.S. Dyer, P.C. Yi, X. Huang, P. Bhargava, C.D. Fegan

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): A.V. Danilov, P. Hillmen, S.A. Rule, M.J.S. Dyer, S.S. Mitra, P.C. Yi, R. Humeniuk, X. Huang, Z. Zhou, P. Bhargava, J.M. Jürgensmeier, C.D. Fegan

Writing, review, and/or revision of the manuscript: A.V. Danilov, C. Herbaux, H.S. Walter, P. Hillmen, S.A. Rule, M.J.S. Dyer, S.S. Mitra, P.C. Yi, R. Humeniuk, X. Huang, Z. Zhou, P. Bhargava, J.M. Jürgensmeier, C.D. Fegan

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): X. Huang, P. Bhargava

Study supervision: A.V. Danilov, H.S. Walter, E.A. Kio, S.S. Mitra, X. Huang, P. Bhargava, J.M. Jürgensmeier

Other (chief investigator of this study in the UK): M.J.S. Dyer

This study was funded by Gilead Sciences, Inc. We extend our thanks to the patients and their families. We would also like to thank Jeff Silverman, Helen Yu, Hoa Truong, Julie Lin, Donovan Verrill, Wanying Li, and Nishan Raj for advice, and for experimental, programming, and statistical contributions to the article. Writing and editorial support was provided by Impact Communication Partners, Inc. A.V. Danilov is a Leukemia & Lymphoma Society Scholar in Clinical Research.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.