CD22-directed CAR T-cell therapy for large B-cell lymphomas progressing after CD19-directed CAR T-cell therapy: a dose-finding phase 1 study

Abstract

Background: Outcomes are poor for patients with large B-cell lymphoma who relapse after CD19-directed chimeric antigen receptor (CAR) T-cell therapy (CAR19). CD22 is a nearly universally expressed B-cell surface antigen and the efficacy of a CD22-directed CAR T-cell therapy (CAR22) in large B-cell lymphoma is unknown, which was what we aimed to examine in this study.

Methods: In this single centre, open-label, dose-escalation phase 1 trial, we intravenously administered CAR22 at two dose levels (1 million and 3 million CAR22-positive T cells per kg of bodyweight) to adult patients (aged ≥18 years) who relapsed after CAR19 or had CD19-negative large B-cell lymphoma. The primary endpoints were manufacturing feasibility, safety measured by the incidence and severity of adverse events and dose-limiting toxicities, and identification of the maximum tolerated dose (ie, the recommended phase 2 dose). This study is registered with ClinicalTrials.gov (NCT04088890) and is active, but closed for enrolment.

Findings: From Oct 17, 2019, to Oct 19, 2022, a total of 41 patients were assessed for eligibility; however, one patient withdrew. 40 patients underwent leukapheresis and 38 (95%) had CAR T-cell products manufactured successfully. The median age was 65 years (range 25-84), 17 (45%) were women, 32 (84%) had elevated pretreatment lactate dehydrogenase, 11 (29%) had refractory disease to all previous therapies, and patients had received a median of four lines of previous therapy (range 3-8). Of the 38 patients treated, 37 (97%) had relapsed after previous CAR19. The identified maximum tolerated dose was 1 million CAR T cells per kg. Of 29 patients who received the maximum tolerated dose, no patients developed a dose-limiting toxicity or grade 3 or higher cytokine release syndrome, immune effector cell-associated neurotoxicity syndrome, or immune effector cell-associated haemophagocytic lymphohistiocytosis-like syndrome.

Interpretation: This trial identifies CD22 as an immunotherapeutic target in large B-cell lymphoma and demonstrates the durable clinical activity of CAR22 in patients with disease progression after CAR19 therapy. Although these findings are promising, it is essential to recognise that this is a phase 1 dose-finding study. Further investigations are warranted to establish the long-term efficacy and to delineate the patient subgroups that will derive the most benefit from this therapeutic approach.

Funding: National Cancer Institute, National Institutes of Health, Stanford Cancer Institute, Leukemia & Lymphoma Society, Parker Institute for Cancer Immunotherapy, Lymph & Co, and the European Hematology Association.

Figure 1. ∣. Objective Response, Duration of Response, Progression-free Survival, and Overall Survival

(A) Best response achieved among 38 patients treated. (B) Kaplan-Meier estimate of duration of response in patients who achieved a complete response or partial response as a best response. (C) Kaplan-Meier estimate of progression free survival in all patients, subdivided by those who achieved a CR, PR, or no response as their best response. (D) Kaplan-Meier estimate of overall survival in all patients, subdivided by those who achieved a CR, PR, or no response as their best response. (E) Swimmer plot showing the duration of response and survival post-infusion for all treated patients (N= 38). CR=complete response; PR=partial response; SD=stable disease; PD=progressive disease; DL1=dose level 1; DL2=dose level 2; CI=confidence interval; NE=not evaluable.

Figure 2 ∣. Subgroup Analysis of Complete Response

Shown is the analysis of complete response according to key baseline and clinical covariates. The Clopper–Pearson method was used to calculate the 95% confidence interval and are not adjusted for multiplicity. Tumor burden was assessed as the sum of the product diameters (SPD). CD19-negative disease was defined as an H-score <150 (corresponding to ≤50% staining by IHC), and/or undetectable surface CD19 by flow cytometry. SCT=Stem Cell Transplantation; CAR19=CD19-directed chimeric antigen receptor T-cell therapy; PD=progressive disease; DLBCL=Diffuse large B-cell lymphoma; GCB= Germinal Center B-cell-like; LDH=Lactate dehydrogenase; ULN=Upper limit of normal range; AE mgmt.=Adverse event management

Figure 3 ∣. CAR T-Cell Expansion and Correlations with Response and Adverse Events.

(A) CAR+ cellular kinetic profile, illustrating in vivo expansion and persistence of CAR22 cells in patients' peripheral blood, stratified by dose levels. Median values are shown by the interconnected dots, and interquartile ranges (Q1 and Q3) are represented by the shaded regions. Peak expansion occurred at a median of 14 days post-infusion for both dose levels. (B) Association between peak CAR22 expansion and the best objective response achieved (CR/PR vs SD/PD, median 129 vs 8.8, p-value 0.0022). (C) Association between peak CAR22 expansion and the maximum grade of CRS achieved (Grade 2-3 vs Grade 0-1, median 203 vs 46, p-value 0.026). (D) Association between peak CAR22 expansion and the maximum grade of ICANS achieved (Grade 1-2 vs None, median 68 vs 98, p-value NS). (E) Association between peak CAR22 expansion and the maximum grade of IEC-HS achieved (Grade 2-4 vs Grade 0-1, median 2378 vs 189, p-value 0.0016). (F) Baseline tumor CD22 expression by semiquantitative H-scoring of immunohistochemistry for all available patients (n=27) including those who did (n=12) or did not (n=15) achieve a CR. H-score was calculated as the percentage of cells with positive staining multiplied by the intensity of staining on a scale from 0 to 3+. No patients (n=6) whose tumor sample had a H-score below 200 achieved a CR (p-value 0.02) (G) Proportion of baseline tumor surface CD22 expression at low or absent levels by quantitative flow cytometry for all available patients (n=27) including those who did (n=14) or did not (n=13) achieve a CR. Patients with higher proportions (>15%) of cells carrying fewer than 1000 molecules of CD22 on their cell surface (n=4) had a higher likelihood of CAR22 failure (p-value 0.04). DL1=dose level 1; DL2=dose level 2; ORR=objective response rate; CRS=cytokine release syndrome; ICANS=immune effector cell associated neurotoxicity syndrome; IEC-HS=immune effector cell-associated hemophagocytic lymphohistiocytosis-like syndrome; CR=complete response; PR=partial response; SD=stable disease; PD=progressive disease.