Phase Ib study of combinations of avadomide (CC‐122), CC‐223, CC‐292, and rituximab in patients with relapsed/refractory diffuse large B‐cell lymphoma

Abstract There is a need for additional treatment options for patients with relapsed or refractory diffuse large B‐cell lymphoma (DLBCL) who do not benefit from available therapies. We examined combinations of the cereblon E3 ligase modulator (CELMoD) agent avadomide (CC‐122), the selective, ATP‐competitive mammalian target of rapamycin kinase inhibitor CC‐223, and the potent, selective, covalent Bruton tyrosine kinase inhibitor CC‐292 in patients with relapsed/refractory (R/R) DLBCL. In the multicenter, phase Ib CC‐122‐DLBCL‐001 study (NCT02031419), the dose‐escalation portion explored combinations of CC‐122, CC‐223, and CC‐292 administered as doublets or triplets with rituximab in patients with chemorefractory DLBCL. Primary endpoints were safety, tolerability, and dose‐limiting toxicities; additional endpoints included pharmacokinetics, pharmacodynamics, biomarkers, and preliminary efficacy. As of December 1, 2017, 106 patients were enrolled across four cohorts. The median age was 65 years (range 24–84 years), and patients had a median of 3 (range 1–10) prior to regimens. A total of 101 patients (95.3%) discontinued, most commonly due to disease progression (49.1%). The most common any‐grade adverse events (AEs) across treatment arms were gastrointestinal and hematologic; the most common grade 3/4 AEs were hematologic. CC‐122 was well tolerated, with no unexpected safety concerns. Preliminary efficacy was observed in three of four treatment arms. CC‐122 plus rituximab was considered suitable for dose expansion, whereas CC‐223 and CC‐292 combinations were associated with enhanced toxicity and/or insufficient improvement in responses. CC‐122 plus rituximab was well tolerated, with preliminary antitumor activity in patients with R/R DLBCL. This innovative study demonstrates the feasibility of assessing the tolerability and preliminary efficacy of novel combinations utilizing a multi‐arm dose‐finding design.


INTRODUCTION
Non-Hodgkin lymphoma (NHL) comprises a heterogeneous group of lymphoproliferative diseases with differing patterns of clinical presentation and responses to treatment [1,2]. Diffuse large B-cell lymphoma (DLBCL) is characterized by aggressive clinical behavior and is the most common type of NHL, accounting for approximately 40% of diagnoses [2,3]. The standard first-line treatment for DLBCL is the regimen R-CHOP (cyclophosphamide, doxorubicin, vincristine, and prednisone plus the CD20-directed monoclonal antibody [mAb] rituximab) [4].
Current guidelines recommend cytotoxic chemotherapy plus rituximab followed by autologous stem cell transplant (ASCT) in eligible patients as salvage treatment after first relapse [4]. However, those patients with DLBCL whose disease is refractory to salvage therapy and/or are ineligible for transplantation, or whose disease relapses after ASCT, have historically had few treatment options and typically have a poor prognosis [5][6][7][8]. A number of therapies have recently demonstrated efficacy in the third-line treatment setting and are now standard of care, including anti-CD19 chimeric antigen receptor (CAR) T-cell therapies, the anti-CD19 antibody-drug conjugate (ADC) loncastuximab tesirine, and the nuclear export inhibitor selinexor [4]. Initial data from two studies of anti-CD19 CAR T-cell therapies as second-line therapy in ASCT-eligible patients (TRANSFORM NHL-001 and ZUMA-9) have also shown improved event-free survival and response rates compared with high-dose chemotherapy and ASCT, suggesting that new therapies may offer improvements in outcome in earlier treatment lines as well [9,10].
Various targeted therapies have demonstrated single-agent activity in relapsed/refractory (R/R) DLBCL, including the immunomodulatory drug (iMiD) agent lenalidomide [11,12], the Bruton tyrosine kinase (BTK) inhibitor ibrutinib [13], and the mammalian target of rapamycin (mTOR) inhibitor temsirolimus [14]. These therapies target signaling pathways involved in B-cell growth and survival, and there is considerable interest in investigating combinations containing these drugs or compounds with similar mechanisms of action for the treatment of NHL [15]. To date, promising results have been observed with some combinations, including lenalidomide plus rituximab in follicular lymphoma (FL), lenalidomide plus tafasitamab in R/R DLBCL, everolimus plus R-CHOP in first-line DLBCL, and the triplet of ibrutinib, lenalidomide, and rituximab in first-line DLBCL [16][17][18][19].
A greater understanding of the pathology of DLBCL has led to the identification of molecular subgroups that are associated with distinct actionable targets and differential responses to some therapies, notably double-hit lymphoma and the germinal center B-cell (GCB) and activated B-cell (ABC) subtypes [20,21]. For example, the ECOG-ACRIN-1412 trial, which compared lenalidomide in combination with R-CHOP to R-CHOP alone, found that the addition of lenalidomide was associated with improved progression-free survival (PFS) in patients with ABC subtype DLBCL, but not in patients with the GCB subtype [22]. Knowledge of actionable targets associated with specific subtypes allows the selection of rational drug combinations with complementary mechanisms of action that are likely to have enhanced antitumor activity [23]. However, clinical investigation of such combinations must be done with care, due to the possibility that additive or synergistic effects may result in increased toxicity [20,23]. An example of such toxicities was reported from a study evaluating the combination of idelalisib, a phosphatidylinositol-3-kinase (PI3K) δ inhibitor, lenalidomide, and rituximab in mantle cell and FL [24]. Although doublets of idelalisib or lenalidomide with rituximab have previously demonstrated efficacy and safety in clinical trials, the combination of idelalisib and lenalidomide, with and without rituximab, was unexpectedly found to be poorly tolerated [24].
Avadomide (CC-122) is a novel cereblon E3 ligase modulator (CELMoD) agent that binds to cereblon in the cullin4 E3 ubiquitin ligase complex, leading to ubiquitination and subsequent degradation of the hematopoietic transcription factors Aiolos and Ikaros [25]. The degradation of Aiolos and Ikaros results in derepression of interferonresponse gene promoters and apoptosis in malignant B cells and derepression of interleukin-2 expression and secretion in T cells, resulting in T-cell activation [25][26][27][28]. CC-122 was well tolerated and demonstrated promising preliminary activity as monotherapy in R/R DLBCL [29][30][31], as well as preliminary activity in combination with the anti-CD20 mAb obinutuzumab in DLBCL [31]. A novel gene expression classifier has also been developed that identified a subgroup of patients with an improved clinical response to CC-122 monotherapy [29,32].
CC-223 is a selective, ATP-competitive inhibitor of mTOR, a kinase that plays a key role in the regulation of cell proliferation and survival [33][34][35]. CC-292 is a potent, selective, covalent inhibitor of BTK, a critical component of the B-cell receptor signaling network required for B-cell development [36].
As iMiD agents, mTOR/PI3K pathway inhibitors, and BTK inhibitors have demonstrated benefit as monotherapies and in combination with other agents, it is reasonable to suggest that combinations of CC-122, CC-223, and CC-292 may provide additional benefit over monotherapies. Rituximab has also been shown to enhance the antitumor activity of agents with these mechanisms of action in clinical studies [37][38][39], providing a rationale for the evaluation of rituximab in combination with all three agents.
The rationale for the addition of rituximab is further supported by preclinical studies, including enhancement of rituximab-mediated antibody-dependent cell-mediated cytotoxicity by CC-122 in two DLBCL cell lines ( Figure S1) and increased tumor growth inhibition (TGI) with CC-122 plus rituximab compared with either agent alone in a DLBCL xenograft model (55% TGI with combination therapy versus 28% and 19% TGI for CC-122 and rituximab monotherapies, respectively; Figure S1).
The CC-122-DLBCL-001 study was conducted to evaluate novel combinations of these agents administered as doublets and triplets in combination with rituximab in patients with R/R DLBCL in dose escalation and DLBCL or FL in dose expansion. Here, we report results from patients with DLBCL treated in the dose-escalation (part A) portion of the study.  Eligible patients were aged > 18 years, with histologically or cytologically confirmed R/R DLBCL (including transformed indolent lymphoma). All patients had chemorefractory DLBCL, defined as having a stable disease or progressive disease (PD) as the best response to their last chemotherapy and/or PD or recurrence within 12 months of prior ASCT. Patients must have received an anti-CD20 mAb therapy (unless the tumor was CD20-negative) and anthracycline-containing chemotherapy as well as ≥1 prior salvage treatment (unless ineligible for ASCT). See Supporting Information for additional eligibility criteria.

Treatment
CC-122 hydrochloride, referred to here as CC-122, as an active pharmaceutical ingredient in capsule (AIC) formulation, was administered orally once daily (QD). An alternate intermittent schedule of CC-122 QD for five of every seven consecutive days (5/7), which improved tolerability and reduced the frequency and severity of neutropenia in a study of CC-122 monotherapy in DLBCL, was also explored [29].

Study assessments
AEs

Safety
Treatment-emergent AEs (TEAEs) occurring in ≥10% of patients by treatment arm are reported in Table 2 (Table S4).
At least 50% of patients in each arm reported a serious AE (SAE); however, the occurrence of SAEs in >1 patient was uncommon (  in a formulated capsule, given QD for 5/7 days plus 375 mg/m 2 rituximab was considered suitable for expansion.

Efficacy
The objective response rate (ORR) among treated patients was 29 arms, of the responders, three proceeded to ASCT. The median PFS and 6-and 12-month estimated PFS rates were substantially higher in arms A and B relative to arms C and D (Table 3). Prolonged responses of >500 days were observed in 10 patients (Figure 2).

Biomarker analysis
In a prior clinical study, Aiolos was shown to be a dose-dependent    was reduced by approximately 30% and the pAKT levels remained at an approximate 50% reduction.

PK and pharmacodynamics
Mean plasma concentration-time profiles and steady-state PK param-

CONCLUSIONS
Despite continuing advances in the treatment of B-cell malignancies, acquired and de novo resistance to treatment remain significant issues. treatment for patients whose disease has progressed despite treatment with available standard therapies [42][43][44]. Furthermore, initial data show that CAR T-cell therapies have promising efficacy compared with high-dose chemotherapy and ASCT, the current standard-of-care second-line treatment for DLBCL [9,10] the comparability of the patient populations enrolled in these trials is unclear [31]. The dose expansion provides an opportunity to assess CC-122 as a next-generation cereblon-modulating drug in the context of historical data with lenalidomide-containing doublet regimens.
In recently reported clinical studies, a novel gene expression classifier identified tumors with a high T cell and macrophage infiltrate that showed an enhanced response to CC-122 in patients with R/R DLBCL [29,32]. rates of 40%-58% as third-line treatment in patients with R/R large B-cell lymphomas, including DLBCL [42][43][44]. Importantly, responses to CAR T-cell therapies have proven to be durable. The estimated 12-month PFS rate was 65% in patients with a complete response to lisocabtagene maraleucel, 24-month PFS was 72% in patients with a complete response to axicabtagene ciloleucel, and 36-month PFS was 70% in patients with a complete response to tisagenlecleucel at 3 months [42,44,45]. In comparison, CR rates with loncastuximab tesirine and selinexor monotherapies in the third-line setting were 24% and 12%, respectively [46,47].
Several doublet and triplet combinations have been evaluated in patients who have received two or more prior therapies. Treatment with the ADC polatuzumab vedotin in combination with bendamustine and rituximab (pola-BR) resulted in a significantly higher CR rate and reduced the risk of death by 58% compared with BR alone in patients with transplantation-ineligible R/R DLBCL [48]. The combination of the CD19 mAb tafasitamab plus lenalidomide produced an ORR of 60% with CRs observed in 43% of patients with R/R DLBCL ineligible for ASCT [17]. These regimens are recommended as second-line or later therapy in the NCCN guidelines [4]. Clinical activity was also observed in a study of ibrutinib plus lenalidomide in DLBCL, with an ORR of 38% and median DOR of 15.9 months; of particular interest, increased ORR was observed in non-GCB-like DLBCL [19]. These studies indicate not only that novel triplet therapies are feasible, but that toxicity management and patient selection are key to identifying combination doses that may lead to transformational efficacy in specific patient populations.
In conclusion, this dose-escalation study evaluated multiple combi-

FUNDING INFORMATION
This work was funded was provided by Celgene, a Bristol Myers Squibb company.