The Impact of Tirabrutinib Monotherapy for the Treatment of Bing-Neel Syndrome: A Multicenter Retrospective Study

Abstract

Waldenström macroglobulinemia (WM) is a rare lymphoproliferative disorder, and Bing-Neel syndrome (BNS) is characterized by the infiltration of WM cells into the central nervous system (CNS) without large cell transformation. Two retrospective studies demonstrated that BNS occurs in approximately 1% of WM patients and may develop at any stage of the disease [1, 2]. Although a standard treatment strategy has yet to be established for BNS, drugs that cross the blood–brain barrier are commonly used. With conventional chemotherapy regimens, the 3-year overall survival (OS) rate for patients with BNS was reported to be 60%, and a significant number of deaths occurred within 2 years of the diagnosis of BNS, 75% of which were attributed to its progression [2]. Castillo et al. [3] examined 28 BNS patients treated with the Bruton's tyrosine kinase inhibitor (BTKi), ibrutinib, and showed symptomatic improvement in 86% of patients and a 2-year event-free survival (EFS) rate of 80%. A recent multicenter study was performed on BNS treated by the second-generation BTKi, zanubrutinib, and the findings obtained indicated its high efficacy and manageable toxicities [4]. Tirabrutinib, a second-generation BTKi developed in Japan, has been authorized for the treatment of WM [5] as well as primary CNS lymphoma [6]; however, evidence for the efficacy and safety of tirabrutinib for BNS in real-world clinical practice is limited. Therefore, we conducted a multicenter retrospective study to examine the outcomes of BNS treated with tirabrutinib.

Patients with BNS treated with at least one dose of tirabrutinib were retrospectively enrolled in this study, which included previously published case reports with extended follow-ups and re-evaluations (the Supporting Information). A definitive diagnosis of BNS was defined by any of the following criteria: atypical lymphocytes, lymphoplasmacytic cells, or plasma cells detected by cytology in the CSF, the confirmation of clonal B-cells in the CSF by flow cytometry, the infiltration of atypical lymphocytes, lymphoplasmacytic cells, or plasma cells in the histopathological diagnosis of CNS tissue, or the detection of the MYD88 ^L265P mutation in the CSF. Additionally, a probable diagnosis of BNS was applied to patients who did not fulfill the criteria for a definitive diagnosis of BNS but were deemed appropriate for a diagnosis of BNS based on the exclusion of other diseases, such as only evidence of brain masses, leptomeningeal enhancement, or abnormal spinal cord signals using magnetic resonance imaging (MRI). A definitive diagnosis and probable diagnosis were equally evaluated in the present study. This study was approved by the centralized Ethics Review Committee of the National Hospital Organization Disaster Medical Center, the lead research institution, and by the collaborating research institutions. Clinical data at the time of the diagnosis of WM, BNS, and tirabrutinib initiation for BNS were collected from medical records. Response criteria, a mutation analysis, toxicity assessments, statistical analyses, and informed consent are available in the Supporting Information.

Twenty-one patients were included in the present study (patient inclusion details are provided in the Supporting Information), and their clinical characteristics are summarized in Table S1. Fourteen patients (66.7%) were male. Median age at the diagnosis of WM was 61 years (range: 45–77 years), while median age at the diagnosis of BNS was 63 years (range: 45–77 years). The MYD88 ^L265P mutation was detected in 4 of 7 bone marrow (BM) samples at the diagnosis of WM. The median interval from the diagnosis of WM to that of BNS was 41.3 months (range: 0–145.9 months), and 6 patients (28.6%) were simultaneously diagnosed with BNS at the diagnosis of WM. The other 15 patients received the therapeutic intervention for WM before the BNS diagnosis, and the median number of chemotherapies for WM was 2 (range: 1–5). Neurological symptoms of BNS included sensory deficits in 15 patients (71.4%), motor deficits in 14 (66.7%), ataxia in 8 (38.1%), cranial nerve deficits in 5 (23.8%), cognitive deficits in 3 (14.3%), impaired consciousness in 3 (14.3%), neurogenic bladder in 3 (14.3%), and seizures in 2 (9.5%). Sixteen and 5 patients received a definitive diagnosis and a probable diagnosis, respectively. Among patients with a definitive diagnosis of BNS, CSF cytology was positive in 15 patients, CSF flow cytometry revealed clonal B-cells in 9, and the MYD88 ^L265P mutation (the analysis was not performed on all patients) was detected in the CSF of 7 of the 8 patients analyzed. Three patients were diagnosed with BNS based on pathological findings in the brain or cauda equina. No patients were tested for CXCR4 mutations in any samples, including BM, CSF, or CNS tissue. Four patients with a probable diagnosis of BNS were diagnosed based on imaging findings, and another was diagnosed by an increase in CSF protein concentrations, clinical symptoms, and the exclusion of other diseases. MRI findings of the brain and spinal cord at the time of the diagnosis of BNS, prior therapies for BNS before starting tirabrutinib, and the starting dosage of tirabrutinib are available in the Supporting Information.

A swimmer plot shows the duration of tirabrutinib therapy, the time of responses to tirabrutinib, the time of tirabrutinib discontinuation, and the BNS status during the observation period for each patient (Figure 1A). Treatment responses to tirabrutinib were assessed in 18 patients, excluding three (Nos. 1, 2, and 5), with reasons detailed in the Supporting Information. All other analyses, including those for survival and safety, were performed on the entire cohort of 21 patients. Among the 18 patients who were evaluable for responses, all responded to tirabrutinib (ORR 100%), and there were 10 CR (55.5%). The median time from tirabrutinib initiation to the best response was 5 months (range, 0.2–44.9 months). Tirabrutinib was continued without BNS progression in 18 of 21 patients during the observation period; the remaining three discontinued treatment due to the progression of BNS (n = 2) or WM (n = 1). Patient No. 11 discontinued tirabrutinib due to WM progression, although CR was maintained for BNS. The patient died of pneumonia associated with WM progression. Two patients discontinued tirabrutinib due to BNS progression (patient Nos. 7 and 17), and subsequent treatment for BNS in both patients consisted of ibrutinib and rituximab. Patient No. 7 continued ibrutinib and rituximab until the last follow-up, while patient No. 17 did not respond and died due to BNS. The median follow-up duration from tirabrutinib initiation to the last observation was 30.9 months (range: 4.5–49.5 months). Estimated 30-month EFS and OS rates from tirabrutinib initiation were 90.5% (95% CI, 67.0%–97.5%; Figure 1B) and 90.2% (95% CI, 66.2%–97.5%; Figure 1C), respectively. Hematological responses to tirabrutinib and survival from the BNS diagnosis to the last observation are available in the Supporting Information and Figure S1. AEs of any grade developed in 16 patients (76.2%), while grade 3 or higher AEs occurred in 7 (33.3%). A summary of AEs is shown in Table S2. Skin eruptions were observed in 3 patients (all grade 2) and were managed through dose reductions and the temporary interruption of tirabrutinib. Details of AEs grade ≥ 3, dose reductions, and the interruption of tirabrutinib are available in the Supporting Information. No patients discontinued tirabrutinib due to AEs.

The present results showed the efficacy and safety of tirabrutinib. Two factors may explain the favorable treatment outcomes of tirabrutinib for BNS. The first factor is the high CNS penetration rate of tirabrutinib, reported to be 13%–18% based on a phase I/II study on patients with primary CNS lymphoma [6], which may account for the high response rate observed in the present study. The second factor is the difference in the AE profiles of BTKis. In the present study, no cases required the discontinuation of tirabrutinib due to AEs. In contrast, 2 of 28 patients in a previous study discontinued ibrutinib due to AEs [3]. This difference may be attributed to the high selectivity of second-generation BTKis, which reduces off-target effects and minimizes AEs, thereby contributing to the favorable EFS observed in this study.

There are several limitations that need to be addressed. This was a retrospective study with a small sample size. Another limitation is that testing for the MYD88 ^L265P mutation was performed on only a limited number of cases using CSF or BM samples. Among the three patients who underwent testing on both compartments, one showed concordant positivity, while two showed discordant results with CSF-positive and BM-negative findings. In both discordant cases, the BM tumor burden was less than 20%, and Sanger sequencing was used, raising the possibility of false-negative results due to limited assay sensitivity. The lack of testing for the CXCR4 mutation and molecular analyses using peripheral blood cell-free DNA or the droplet digital polymerase chain reaction represents another significant limitation. Furthermore, the CSF examination at the response assessment was not considered to be an absolute requirement, and efficacy was evaluated by defining CR as the resolution of reversible symptoms, improvements in imaging findings, or CSF clearance. In addition, a definitive diagnosis and probable diagnosis of BNS were examined equally in this study.

In conclusion, tirabrutinib exhibited acceptable toxicity and durable efficacy for the treatment of BNS and, thus, has potential as a treatment option for BNS.

Masuho Saburi and Naohiro Sekiguchi were responsible for the design of the study, the interpretation of data, and writing the manuscript. Taro Masunari, Noriko Fukuhara, Yuichiro Inagaki, Arika Shimura, Naoto Imoto, Yuta Hasegawa, Masao Hagihara, Nobuhiko Kobayashi, Hanae Kumekawa, Hideyuki Nakazawa, Kana Miyazaki, Toshiro Kawakita, Takahiro Isshiki, Atsushi Katsube, Shin Fujisawa, Yuichi Horigome, Yusuke Koba, and Fumiaki Jinnouchi provided patient data. All authors contributed to the revision of the manuscript and approved the final version.

M.S. has received honoraria from Janssen and Ono. N.F. has received research grants from Abbvie, Chugai, Chordia Therapeutics, Genmab, Haihe, Incyte, Kyowa Kirin, Loxo Oncology, Ono, and Takeda, and honoraria from Abbvie, AstraZeneca, BMS, Chugai, CSL Behring, Eisai, Eli Lilly, Genmab, Janssen, Kyowa Kirin, Nippon Shinyaku, Novartis, Ono, Sanofi, Symbio, and Takeda. Y.I. has received honoraria from Kyowa Kirin, Sanofi, BMS, Jansen, Pharmaessensia, and Novartis. K.M. has received research grants or contracts from Takeda, Otsuka, Chugai, Kyowa Kirin, Sumitomo, and Zenyaku Kogyo, and honoraria from Chugai, Janssen, AstraZeneca, Novartis, Incyte, Asahi Kasei, Abbvie, SymBio, Ono, Genmab, Meiji Seika, BMS, Kyowa Kirin, Daiichi Sankyo, Gilead Sciences, and Nippon Shinyaku. Y.H. has been involved in advisory board participation with Janssen and received honoraria from Janssen, Sanofi, Ono, BMS, Takeda, CSL Behring, Novartis, and Chugai. N.S. has received honoraria from Janssen, Ono, and BeiGene, and research funding from Incyte Biosciences Japan, Janssen, Mitsubishi Tanabe Pharma Corporation, MSD, and Ono.