Anti-CD22, a promising new CAR-T target for R/R ALL: Results from a phase I trial

Chimeric antigen receptor (CAR) T-cell therapy has proven very efficacious for the treatment of patients with relapsed/refractory (R/R) B-cell acute lymphoblastic leukemia (ALL). CAR T-cell therapies for ALL have mainly targeted the B-cell CD19 surface antigen.¹ Unfortunately, a proportion of patients still relapse following anti-CD19 CAR T-cell therapy, having very limited treatment options and poor prognosis.^1,2

In an attempt to investigate novel treatment options for ALL patients who relapse following CD19-directed immunotherapies, Shah et al. published the results of a phase I trial in the Journal of Clinical Oncology.² In this study (NCT02315612), the safety and preliminary efficacy of anti-CD22 targeted CAR T-cell therapy was assessed in patients with B-cell ALL.

Study design

• Single center, 3 + 3 dose-escalation, phase 1 trial in children and young adults from 3–30 years old with R/R CD22⁺ malignancies
• All but two evaluable patients had B-ALL diagnosis (n = 55)
• Patients who received prior CAR-T therapy needed to have < 5% circulating CAR T cells
• Anti-CD22 CAR-T dose levels (DL):
  • DL1: 3 × 10⁵/kg (n = 6)
  • DL2: 1 × 10⁶/kg (n = 18)
  • DL3: 3 × 10⁶/kg (n = 2)
• CD4/CD8 T-cell selection (TCS) from all apheresis material was performed to maximize the efficacy and safety of the CAR-T cells. Due to enhanced inflammatory responses observed after DL2-TCS (n = 7), all subsequent patients were de-escalated to DL1-TCS of 3 × 10⁵/kg (n = 25)
• All patients received fludarabine 25 mg/m²/day on Days -4, -3, and -2 and cyclophosphamide 900 mg/m² on Day -2 prior to CAR T-cell infusion (Day 0)
• The primary endpoint was safety and toxicity, while secondary endpoints included efficacy, CAR T-cell persistence, cytokine profiling, and reinfusion evaluation
• The key baseline characteristics of the patients are shown below in Table 1

Table 1. Patient baseline characteristics (adapted from Shah et al.²)

CAR, chimeric antigen receptor; CNS, central nervous system; HSCT, hematopoietic stem cell transplantation * Bone marrow leukemic blasts 5–25%
Baseline characteristic, % unless otherwise specified	All treated patients (n = 58)
Median age, years (range)	17.5 (4.4-30.6)
Prior therapy
HSCT CD19-targeted therapy CD19 CAR-T Blinatumomab Inotuzumab CD22 CAR-T exposure	67.2 87.9 62.0 39.7 24.1 8.6
Any CD19– population	56.9
≥ M2 marrow*	75.9
Isolated CNS disease at infusion time	1.7

Key findings

• Cytokine release syndrome (CRS) was detected in 86.2% of patients, with the majority being of Grade 1–2 (90%). The average time to CRS onset was 7 days postinfusion (range, 3–16), and the median duration of CRS was 5 days.
• Neurotoxicity was generally mild without seizures, encephalopathy, or severe toxicity in the first 22 participants. One or more neurological events were detected in 32.8% of all patients with the majority being of Grade 1–2, with the exception of one patient who developed Grade 4 intracranial hemorrhage.
• Two deaths occurred at DL2, leading to protocol modification for earlier tocilizumab and/or corticosteroids in patients with pulmonary toxicity and no other Grade 5 events. Reported toxicity rates and CRS management are detailed in Table 2.

Table 2. Toxicity profile and CRS management (adapted from Shah et al.²)

AE, adverse events; aHUS, atypical hemolytic uremic syndrome; ASTCT, American Society for Transplantation and Cellular Therapy; CLS, capillary leak syndrome; CRS, cytokine release syndrome; DIC disseminated intravascular coagulation; HLH, hemophagocytic lymphohistiocytosis
Variable, %	All treated patients (n = 58)
Patients with CRS	86.2
CRS Grade 1–2 CRS Grade ≥ 3 CRS Grade ≥ 3 ASTCT scale	90.0 10.0 24.0
Any neurotoxicity	32.8
Severe neurotoxicity	1.7
CRS management
Received tocilizumab Received corticosteroids	39.7 31.0
Other AEs
DIC Symptomatic coagulopathy HLH CLS aHUS	24.1 15.5 32.7 5.2 5.2
Grade 5 events	3.4

 

• Following CD4/CD8 TCS at DL2, there was a higher incidence of hemophagocytic lymphohistiocytosis/macrophage activation syndrome-like toxicities and coagulopathies (p = 0.017) but without any change in CRS rates. Therefore, the team de-escalated to DL1-TCS.
• Hemophagocytic lymphohistiocytosis/macrophage activation syndrome toxicities were treated with anakinra.
• Peak CAR expansion was detected between Days 14–21 after infusion with a median percentage of CAR-positive T cells at peak expansion of 77% (median absolute number of CAR T cells, 480.5 cells/μL; range, 39.7–11,346).

Preliminary Efficacy

• Efficacy data are shown in Table 3. In brief, amongst the 57 evaluable patients, 70.2% of them achieved complete remission (CR). Of those, 87.5% were measurable residual disease (MRD) negative as detected by flow cytometry.
• Responses were not significantly affected by prior CD19-targeted therapy (p = 0.24) or prior hematopoietic stem cell transplantation (p = 0.76).
• CR rates of > 70% were observed at DL2, DL2-TCS, and DL1-TCS levels.
• Patients with prior CD22-targeted therapy or CD22 CAR had significantly lower MRD-negative CR rates (p = 0.039) and shorter remission duration.
• CD22 expression was lower in those with MRD positivity or at the time of relapse (p ≤ 0.001), indicating that CD22 is an important player in immune escape mechanisms.

Table 3. Preliminary efficacy results²

ALL, acute lymphoblastic leukemia; CI, confidence interval; CR, complete response; HSCT, hematopoietic stem cell transplantation; MRD, measurable residual disease; RFS, relapse-free survival; OS, overall survival.
Responses	Evaluable patients
CR rate (total cohort, n = 57) CR rate (ALL patients only, n = 55)	70.2% 72.7%
MRD-negative CR (total cohort, n = 57) MRD-negative CR (ALL patients, n = 55)	87.5% 63.6%
Median OS of those in CR, months (95% CI)	13.4 (7.7-20.3)
Median RFS of those in CR, months (95% CI)	6.0 (4.1-6.5)
Patients proceeded to HSCT, n	13

Conclusion

The authors highlight the reduced neurotoxicity seen with CD22 CAR T cells compared with CD19 CAR T cells, though they also mention distinct toxicities such as atypical hemolytic uremic syndrome, severe capillary leak syndrome, and ocular manifestations. In addition, the group comment on how manufacturing modifications can alter outcomes. They conclude that preliminary data from this phase I trial are promising and indicate that CD22 is a potentially effective target for CAR-T therapy in patients with ALL, especially those who relapse following CD19-targeted therapies.