Clinical Efficacy and Safety of Chimeric Antigen Receptor T-Cell Therapy for Mantle Cell Lymphoma with Secondary Central Nervous System Involvement

Abstract

Data describing outcomes of chimeric antigen receptor (CAR) T-cell therapy in patients with secondary central nervous system (SCNS) involvement of mantle cell lymphoma (MCL) are limited. We identified 10 patients with MCL and SCNS involvement treated with anti-CD19 CAR T-cell therapy at 3 U.S. academic centers. Frequent objective responses were observed in the CNS (86%) and systemically (90%), and the 1-year progression-free survival was 47%. Seven patients developed immune-effector-cell-associated-neurotoxicity-syndrome (n=2 Grade 1, n=5 Grade 3). Our results suggest that anti-CD19 CAR T-cell therapy in this setting is feasible and additional data regarding neurotoxicity in this population may be warranted.

INTRODUCTION

Effective treatment of secondary central nervous system (SCNS) involvement in mantle cell lymphoma (MCL) represents an unmet clinical need.¹ The prognosis is generally poor, with the median overall survival (OS) from CNS involvement reported to be less than 5 months.² While the Bruton tyrosine kinase inhibitor (BTKi) ibrutinib has demonstrated improved survival outcomes compared to standard chemoimmunotherapy,³ robust data regarding alternate therapeutic strategies, and particularly post-BTKi, are currently lacking.

Chimeric antigen receptor (CAR) T-cell therapies have been demonstrated to have manageable safety profiles and favorable efficacy in diffuse large B-cell lymphoma with SCNS involvement.^4–6 However, there are limited data on efficacy and safety of anti-CD19 CAR T-cell therapy in patients with MCL and SCNS disease. In the ZUMA-2 trial, which led to regulatory approval of brexucabtagene autoleucel (brexu-cel), patients with CNS involvement were excluded.⁷ To date, a single-patient case report and one retrospective analysis have formally reported on a total of 17 patients with MCL and SNCS involvement treated with CAR T-cell therapy. In these patients, CAR T-cell therapy was feasible with a similar toxicity profile compared to that observed for patients without CNS involvement.^8,9 However, given the limited number of patients reported, additional data are needed to confirm these findings and guide clinical management for this uncommon patient subgroup. In this retrospective study, we report the efficacy and safety of anti-CD19 CAR T-cell therapy in 10 patients from 3 U.S. academic centers with MCL and SCNS involvement.

METHODS

Patients with MCL and SCNS disease at any point prior to CAR T-cell infusion were identified from institutional immune effector cell (IEC) therapy databases at three U.S. academic centers (Dana-Farber Cancer Institute, Memorial Sloan Kettering Cancer Center, and Massachusetts General Hospital Cancer Center). The following data were collected from the electronic medical record: patient demographics; MCL disease features; prior therapies; onset and duration of cytokine release syndrome (CRS) and IEC-associated neurotoxicity syndrome (ICANS) as documented by treating physicians based on consensus guidelines from the American Society of Transplantation and Cellular Therapy;¹⁰ results of available positron emission tomography (PET) /computed tomography (CT) scans, brain or spine magnetic resonance imaging (MRI) scans, and cerebral spinal fluid (CSF) testing; progression and survival status. Systemic disease response was based on Lugano criteria using available PET/CT results,¹¹ and CNS disease response was based on resolution of contrast enhancement within parenchymal or leptomeningeal lesions from available imaging results and/or morphologic or flow cytometric clearance of lymphoma cells from available CSF results (as per the International Primary CNS Lymphoma Collaborative Group guidelines).¹² Data cutoff was January 1, 2023. The study was approved by the Institutional Review Boards at participating centers. Progression-free survival (PFS) was calculated as the time from CAR T-cell infusion to disease progression, death, or last follow-up. OS was calculated as the time from CAR T-cell infusion to death or last follow-up.

RESULTS

Patient and Disease Characteristics

Ten patients with MCL were identified who had a history of SCNS and received treatment with CAR T-cell therapy. The median age at time of CAR T-cell therapy was 60 years (range, 44–79), 50% of patients were male, and all patients received brexu-cel except for one patient who received tisagenlecleucel (Table 1). The median number of lines of prior therapy was 4 (range, 2–6) and 8 patients received bridging therapy. All patients had previously received BTKi therapy for systemic disease, 3 had undergone prior hematopoietic cell transplantation (2 autologous, 1 allogeneic), and 9 had disease progression within 24 months of diagnosis (POD24). Seven patients had active CNS disease at time of CAR T-cell infusion based on imaging and/or CSF results; the remaining 3 patients last had confirmed active CNS disease 2, 5, and 18 months prior to CAR T-cell infusion (patients F, I, and B, Table 1, respectively). Due to either high-risk features (blastoid/pleomorphic MCL), or history of highly refractory disease, it was felt per treating physicians that the likelihood of durable CNS disease remission was low at time of proceeding to CAR T-cell therapy. Six patients had leptomeningeal disease only, two patients had parenchymal involvement only, and two patients had dual involvement. Eight patients had received prior CNS-directed therapy (median lines of prior therapy 1, range 1–3), including 3 patients who received CNS-directed radiation therapy (Supplemental Table 1). Of note, 6 additional patients with MCL and secondary CNS involvement who were intended to proceed with treatment with brexu-cel were identified. Five patients did not undergo successful leukapheresis (3 due to death, 1 due to interim worsening disease progression and transition to hospice, 1 with development of acute subdural hematoma and hygroma with subsequent decline in performance status precluding CAR T-cell candidacy). One patient died shortly after leukapheresis due to septic and hemorrhagic shock.

Table 1:

Patient and CAR T-cell Treatment Characteristics

Patient ID	Age at CAR T-cell Therapy, Sex	MCL Histology & Ki67 (%) at Diagnosis	TP53 Status at Diagnosis	Prior Lines of Therapy: Total (CNS-directed)	Previous HCT	POD24	CNS Disease Location & Symptoms	Disease Presence at Time of CAR T-cell Therapy	CAR T-cell Product	CRS: Max Grade, Duration (days); Treatment	ICANs: Max Grade, Duration (days); Treatment
A	48, F	Blastoid; 95%	IHC: NA FISH: NA NGS: NA	6 (3)	Yes, auto	Yes	LMD & Parenchyma; lower extremity weakness	Systemic: No CNS: Yes	Brexu-cel	-	-
B	57, F	Blastoid; 50%	IHC: NA FISH: NA NGS: mutation absent	3* (0)	Yes, allo	Yes	LMD; none	Systemic: Yes CNS: No	Brexu-cel	Grade 2, 5d; steroids, toci., anakinra	Grade 3, 16d; steroids, toci., anakinra
C	51, F	NA 26%	IHC: NA FISH: NA NGS: mutation present	4 (2)	No	Yes	Parenchyma; pan-hypopituitarism	Systemic: No CNS: Yes	Brexu-cel	Grade 1, 1d; no treatment	Grade 1, 8d; no treatment
D	57, M	NA	IHC: NA FISH: del(17p) present NGS: NA	4* (1)	Yes, auto	Yes	LMD; seizures	Systemic: No CNS: Yes	Brexu-cel	Grade 1, 8d; toci.	Grade 3, 3d; dex., anakinra
E	79, M	Blastoid; 90%	IHC: NA FISH: no del(17p) NGS: mutation absent	4* (0)	No	Yes	LMD & parenchyma; AMS	Systemic: Yes CNS: Yes	Brexu-cel	Grade 3, 8d; steroids, toci.	Grade 3, 23d; steroids, toci., IVIG
F	64, M	NA	IHC: NA FISH: no del(17p) NGS: mutation absent	2* (1)	No	No	Parenchyma; headache, CN palsy	Systemic: No CNS: No	Brexu-cel	Grade 2, 7d; steroids, toci., anakinra	-
G	78, M	Classic; 10–20%	IHC: Positive (80%) FISH: NA NGS: NA	7* (2)	No	Yes	LMD; vision loss	Systemic: No CNS: Yes	Brexu-cel	Grade 2, 6d; dex., toci.	Grade 3, 13d; dex., toci., anakinra
H	64, M	Pleomorphic; 20%	IHC: NA FISH: NA NGS: mutation absent	4* (1)	No	Yes	LMD; headache	Systemic: Yes CNS: Yes	Brexu-cel	Grade 1, 3d; toci.	-
I	75, F	Pleomorphic; >95%	IHC: Positive (>95%) FISH: NA NGS: NA	4* (1)	No	Yes	LMD; double vision (CN III palsy)	Systemic: Yes CNS: No	Brexu-cel	Grade 1, 4d; dex., toci.	Grade 3, 10d; dex., toci.
J	44, F	Blastoid; 50%	IHC: NA FISH: NA NGS: NA	4* (1)	No	Yes	LMD; bilateral CN VII palsies	Systemic: Yes CNS: Yes	Tisa-cel	Grade 1, 2d; dex.	Grade 1, 5d; dex.

^*Indicates bridging therapy was received (defined as any therapy from time of apheresis to CAR T-cell infusion). allo, allogeneic; auto, autologous; brexu-cel, brexucabtagene autoleucel; CAR, chimeric antigen receptor; CNS, central nervous system; CRS, cytokine release syndrome; d, days; dex., dexamethasone; F, female; HCT, hematopoietic cell transplantation; ICANS, immune effector cell-associated neurotoxicity syndrome; IHC, immunohistochemistry; IVIG, intravenous immunoglobulin; LMD, leptomeningeal disease; M, male; MCL, mantle cell lymphoma; max, maximum; NA, not available; NGS; next-generation sequencing; POD24, progression of disease within 24 months of diagnosis; tisa-cel, tisagenlecleucel; toci., tocilizumab

Efficacy

At day 30 assessment by PET/CT, 9 patients were in a complete response (CR) with respect to systemic disease, while 1 patient had progressive disease (PD) (Figure 1A). Regarding CNS disease, 6 of 7 patients (86%) with active CNS disease at time of CAR T-cell infusion achieved an objective response within the first ~30 days: 2 patients had a CNS CR, 4 patients had a CNS PR, and 1 patient had stable CNS disease. At time of data cutoff, with a median follow-up of 15.4 months for survivors, 7 patients remain alive (5 of whom are in remission), with 3 patients deceased due to PD (2 patients with systemic PD only, 1 patient with both systemic and CNS PD). In the entire cohort, the median PFS was 11.7 months (95% CI: 0.6-Not reached [NR]); median OS NR. The 12-month PFS was 47% (95% CI: 15%−74%) and the 12-month OS was 79% (95% CI: 38%−94%). In the 7 patients with active CNS disease at time of CAR T-cell infusion, the median PFS was 11.7 months (95% CI: 0.6-NR); median OS NR. The 12-month PFS was 36% (95% CI: 5%−70%) and the 12-month OS was 71% (95% CI: 26%−92%). Subsequent therapies for those patients who developed disease progression are included in the legend for Figure 1A.

Figure 1. Clinical and Toxicity Courses.

A) Clinical courses for all 10 patients, with systemic and CNS-specific disease status shown separately. For the 5 patients who developed disease progression after CAR T-cell therapy, subsequent therapies are as follows: Patient D: radiation therapy to nasopharyngeal mass followed by pirtobrutinib, Patient E: radiation therapy followed by venetoclax, Patient H: whole brain radiation therapy, Patient I: obinutuzumab in combination with lenalidomide, Patient J: palliative radiation therapy to spine. B) Toxicity courses with specific interventions for 5 representative patients who developed neurotoxicity. Patient C and Patient J developed Grade 1 ICANS which resolved without any intervention. Patient D, who had a history of recent seizures in the setting of new CNS disease, developed neurotoxicity symptoms on Day +5. He had increased somnolence and difficulty following commands, which then progressed to intermittent garbled speech and worsening somnolence (electroencephalogram [EEG] was negative). Following treatment with dexamethasone and anakinra, this patient returned to his neurological baseline by Day +8. Patient G developed a new decrement in immune effector cell encephalopathy (ICE) score to 8/10 on Day +6. He was treated with tocilizumab, dexamethasone, and later anakinra, however his ICE scored dropped to 0 on Day +12, prompting transfer to the Neurological ICU. EEG monitoring showed no evidence of seizures. Dexamethasone and anakinra were continued and the patient’s neurological status improved over the course of a week, with transfer out of the Neurological ICU on Day +18. Patient I developed a decrement in mental status with ICE score of 0 on Day +2. EEG showed slowing but no epileptiform activity. Following initiation of dexamethasone, symptoms gradually improved with ICE score 8–9 for several days, until resolution by Day +12.

CRS and ICANS

Nine of 10 patients experienced CRS (max grade 1, 2, and 3 in 5 patients, 3 patients, and 1 patient, respectively). The median time to CRS onset was 4 days (range, 1–14) and the median duration was 5 days (range, 2–9). Seven patients developed ICANS (grade 1 in 2 patients and grade 3 in 5 patients), with a median time to onset of 6 days (range, 2–11) and a median duration of 11 days (range, 4–23). Toxicity courses for 5 representative patients with granular information available regarding timing of interventions are shown in Figure 1B. One patient required neurologic intensive care unit admission (stay of 6 days). ICANS occurred at a similar rate among patients with active CNS disease at time of CAR T-cell infusion (5/7) and among patients without active CNS disease (2/3). All 6 patients with leptomeningeal disease, 1 of 2 patients with both leptomeningeal and parenchymal disease, and 1 of 2 patients with parenchymal disease only developed any grade ICANS. All 3 patients ≥ 65 years old developed grade 3 ICANS. All cases of neurotoxicity were reversible. There was no treatment-related mortality.

DISCUSSION

The results from our cohort of 10 patients with MCL and SCNS disease demonstrate that anti-CD19 CAR T-cell therapy in this population is feasible and efficacious. As in the recent U.S. Lymphoma CAR T Consortium retrospective study of brexu-cel,⁹ we found that CAR T-cell therapy was effective in heavily pre-treated, refractory patients (4 median lines of prior therapy in our cohort; 90% with POD24). The systemic disease CR rate in our cohort of 90% (9/10) is comparable to the 82% reported in the consortium study⁹ and the 68% reported in ZUMA-2.⁷ Likewise, durable disease control was similar with a 12-month PFS in our cohort of 47% compared to 59% in the recently published real-world cohort⁹ and 61% in ZUMA-2.⁷ CNS-specific disease responses were not included in the consortium study cohort and thus our study provides more detailed information on this aspect; 86% of patients with active CNS disease at the time of CAR T-cell infusion achieved a response in the CNS.

While our analyses are limited by the sample size of our cohort and retrospective nature, this cohort provides additional results on a rare patient population that bear further investigation. In our cohort, 50% of patients developed grade 3 ICANS, which is higher than the rates reported in patients without CNS disease in the U.S. Consortium study (33%)⁹ and in the registrational ZUMA-2 study (31%).⁷ While we observed frequent high-grade ICANS, there were no fatal complications. Our results also differ from those of the recent U.S. Consortium study, in which only 25% of patients with CNS involvement (n=16) were reported to have high-grade ICANS. Additional studies are thus needed to determine if rates of severe ICANS are higher in patients with either active CNS disease or a prior history of treated CNS disease, and if there are risk factors within this population that can predict more severe neurotoxicity. If higher rates of ICANS are confirmed in patients with SCNS involvement, then strategies that may potentially warrant consideration in this population could include a pre-treatment screening assessment for CNS involvement and interventions such as prophylactic corticosteroids to potentially decrease the rate of severe neurotoxicity, as was investigated in cohort 6 of the ZUMA-1 study, with promising results.¹³ 4–1BB-costimulatory based CAR-T constructs are additionally an important area of ongoing investigation in MCL as rates of ICANS may be lower with lisocabtagene maraleucel.¹⁴ The ongoing TRANSCEND NHL 001 trial (NCT02631044) notably includes patients with CNS involvement and the awaited results may add an important dimension to the choice of CAR product in this patient population. In conclusion, results from this study support ongoing investigation of anti-CD19 CAR T-cell therapy for patients with MCL and SCNS involvement, with careful attention to neurotoxicity.