Spotlight on Ibrutinib in PCNSL: Adding Another Feather to Its Cap

Ibrutinib, a first-in-class irreversible oral Bruton tyrosine kinase (BTK) inhibitor, is Pharmacyclic's blockbuster drug, with 2015 sales crossing $1.23 billion and projected 2022 sales of $8.29 billion (1). Phase I/II clinical trials with ibrutinib in mantle cell lymphoma (MCL), chronic lymphocytic leukemia (CLL), and small lymphocytic leukemia (SLL) began in 2009. Ibrutinib was first approved by the FDA for MCL in 2013. Approval of ibrutinib has since expanded to CLL, SLL, Waldenström's macroglobulinemia (WM), and most recently to marginal zone lymphoma (MZL) in 2017. There is growing in vitro and in vivo evidence that ibrutinib is also an IL2-inducible T-cell kinase (ITK) inhibitor, providing a broad alternative therapeutic utility as an immune modulating agent (2). Ibrutinib has been tested in clinical trials for multiple indications, including diffuse large B-cell lymphoma (DLBCL), multiple myeloma, T-cell lymphoma, a variety of solid tumors, and graft-versus-host diseases (GVHD). Herein, Grommes and colleagues report that ibrutinib as a single agent has activity in primary central nervous system lymphoma (PCNSL) and secondary central nervous system lymphoma (SCNSL).

Grommes and colleagues conducted an open-label, nonrandomized, single-center, dose-escalation study to establish the MTD of ibrutinib as a single agent in 20 patients with relapsed/refractory PCNSL/SCNSL (3). The study was expanded to further assess toxicity and clinical activity exploring doses of 560 and 840 mg administered once daily. Ibrutinib was acceptably tolerated with manageable adverse events, and no dose-limiting toxicity was observed. Ibrutinib reached therapeutic concentrations that inhibit BTK in cerebrospinal fluid, and the concentration increased 1 month into therapy in a dose-responsive manner among patients receiving 560 and 840 mg, respectively. Out of 13 patients with PCNSL, 5 showed complete response (CR), 5 showed partial response (PR), 1 discontinued by choice, and 1 experienced tumor regression but did not meet criteria for PR. Median progression-free survival (PFS) was 4.6 months and median overall survival (OS) was 15 months (7/13 patients still alive). Out of 7 patients with SCNSL, 4 showed CR and 1 showed PR. The median PFS was 7.43 months, and median OS has not been reached as 5 of 7 patients are still alive. Although patient numbers were small, response in each group was much higher than that expected in DLBCL arising from non-CNS compartment, suggesting differential biology may be at play. This study complements the recent report by Lionakis and colleagues demonstrating high single-agent responses and enhanced efficacy of chemotherapy with ibrutinib in PCNSL (4). In contrast to the observation by Lionakis and colleagues, fungal infections with Aspergillosis were not observed.

The authors elegantly complemented the clinical data with sequencing data characterizing their patients and mechanistic studies exploring how this therapy could be improved. First, they looked at the genomic landscape of 586 genes in 177 PCNSL biopsies and compared it with the genomic landscape of DLBCL reported in prior studies to identify any differences. They compared PCNSL with DLBCL outside of the CNS because of the differences in overall response rate to ibrutinib. Mutations in the Toll/IL1 receptor domain of MYD88 and immunoreceptor tyrosine-based activation motif (ITAM) of B-cell antigen receptor (BCR)–associated CD79B occurred at a higher frequency (58% and 41%) in PCNSL than in DLBCL. Second, they studied the association between disease subgroups and mutations in the BCR–NF-κB pathway genes MYD88, CD79B, CARD11, and TNFAIP3. Unlike DLBCL outside of the CNS, ∼89% of PCNSL fall under the non-germinal center B cell–like (non-GCB) subgroup, and both GCB and non-GCB subgroups carried mutations in the BCR–NF-κB pathway. These data suggested that BCR dependence and BCR pathway mutations are distinguishing features of PCNSL. Finally, by correlating BCR pathway mutations in pretreatment patient samples with clinical response, the patient who had complete ibrutinib resistance was identified to have a missense mutation within the coiled–coiled domain of CARD11 (R179Q). Such mutations have been shown to activate NF-κB independent of BTK in DLBCL outside of the CNS and in MCL. Three patients with incomplete ibrutinib responses showed mutations in CARD11 (R337Q) and inactivating lesions in TNFAIP3, a negative regulator of NF-κB.

Mutations in MYD88 have been linked to ibrutinib sensitivity in WM, and mutations in CD79B were frequently associated with MYD88 mutations. It was surprising to note that none of the patients with PCNSL with concurrent MYD88 and CD79B mutations showed a complete response to ibrutinib. In WM, complete responses are rare with ibrutinib. Because Y196D mutation in the ITAM of CD79B impairs BCR downregulation, the authors hypothesized that these mutations attenuate BTK dependence by diversifying BCR signal output and provide a BTK-independent survival signal. To identify such signals, they performed RNA-sequencing analysis from CD79B wild-type and mutant PCNSL samples and, using gene set enrichment analysis, identified that mTOR signaling pathway genes were enriched in CD79B-mutant samples. They validated the activation of the mTOR pathway in patients with CD79B-mutant PCNSL samples by IHC of PI3K–mTOR pathway components as well as established patient-derived xenograft (PDX) models and showed regression of tumor upon selective inhibition of PIK3CA and PIK3CD. They established cell lines from the CD79B-mutant PDX models and demonstrated that PI3K inhibitors synergized with ibrutinib to induce cell death (Fig. 1A). They carefully matched the in vitro mechanistic work to clinically achievable dose ranges of ibrutinib. These data suggest a novel mechanism for PI3K–mTOR-mediated, BTK/NF-κB-independent signaling in CD79B-mutant PCNSL.

An aspect not considered in this work but which could be relevant to the higher response to PCNSL and SCNSL with ibrutinib is the secondary immune-modulating potential of this agent. Patients with both PCNSL and SCNSL have also been shown to be responsive to PDCD1 (PD-1) blockade with nivolumab (5). In addition to irreversibly inhibiting BTK, ibrutinib also targets ITK, and several studies have shown ibrutinib's potential to favorably modulate T-cell and myeloid-derived suppressor cell function (2, 6–9). Given that we now have two new active therapeutics for CNS lymphoma (ibrutinib and nivolumab) for which synergy exists in preclinical models (10), pursuit of this combination as well in clinical trials is warranted.

Not surprisingly, both primary and secondary resistance to ibrutinib is also seen in patients with PCNSL. In this study, primary resistance was noted in a patient with CARD11^R179Q mutation predicted to activate NF-κB downstream of BTK. Other mutations in the NF-κB pathway were also associated with incomplete responses. Another form of potential primary resistance not allowing complete response to ibrutinib is mutation of CD79B or MYD88, which both this study and the Lionakis study demonstrated were quite frequent in PCNSL. Precision medicine with pretreatment sequencing might allow assignment to either alternative therapies (in the case of CARD11^R179Q mutation) or combination-based treatment (in the case of CD79B or MYD88 mutations). Although the observation of activity of ibrutinib in PCNSL and SCNSL is relatively new, it will be of interest if this tumor mimics the pattern of secondary ibrutinib resistance observed in CLL (11, 12) and WM (13), where BTK^C481S and PLCG2 mutations are common, or follows a divergent pathway as seen in aggressive MCL (14).

In an era where developing a drug from bench to bedside takes more than a decade and ∼$2 billion (15), it is exciting that targeted therapy with agents such as ibrutinib has demonstrated an efficient trend of repurposing an FDA-approved drug for subsequent multiple indications. Especially in the context of cancer, where different cancers share similar mutations and activated pathways, repurposing is more often feasible. Ibrutinib use is currently expanding across most B-cell malignancies, and investigation in GVHD, solid tumors, and immune modulation could enhance its application even more (Fig. 1B). As ibrutinib use becomes more widespread, this offers the opportunity to examine drug costs required to gain return on corporate investment. Prices of targeted agents such as ibrutinib are often based upon a narrow niche and do not change over time to match more widespread use. A more economically driven regulatory assessment at each subsequent approval milestone for new therapeutics might offer an opportunity to better control the costs of new therapeutic agents in cancer. Such application of cost adjustment of therapeutic agents will be required to make it financially feasible to combine novel targeted agents. This point is quite important as it provides the next step in upcoming clinical trials where other targeted therapies such as approved PI3K inhibitors and PD-1–blocking antibodies are also expensive.

Although this paper adds another feather to ibrutinib's cap in terms of therapy, what is the next step forward? Combination-based approaches in relapsed patients are clearly indicated. More importantly, moving ibrutinib earlier into the treatment algorithm of PCNSL should be considered, provided it can be performed safely. Indeed, initial therapy for PCNSL is quite intense and often comes with great morbidity and some mortality. The Lionakis study showing Aspergillosis infections with ibrutinib calls attention to such safety concerns. Studies in other types of lymphoma have shown the ability to combine ibrutinib together with multiple chemotherapy regimens without increased aspergillosis. Future studies in a larger number of patients will precisely define the risk of Aspergillosis. A more provocative direction that might be pursued in newly diagnosed older patients and those who are not appropriate for intensive chemotherapy is one that combines ibrutinib together with checkpoint inhibitors or PI3K inhibitors that penetrate the CNS in appropriately genetic-screened patients. This would provide a window of acceptable clinical investigation to ascertain if this uncommon B-cell malignancy can be effectively treated without chemotherapy. From both this study and the recent Lionakis study, we certainly can see a bright future for the use of ibrutinib in CNS DLBCL.

J.C. Byrd is a consultant/advisory board member for Pharmacyclics. No potential conflicts of interest were disclosed by the other author.