CD22 Emerges as CAR T-Cell Therapy Target

OncologyLive, Vol. 19/No. 20, Volume 19, Issue 20

Although CD19 has proved to be an attractive and effective target for chimeric antigen receptor T-cell therapies in hematologic malignancies, a significant subset of patients treated with this groundbreaking form of immunotherapy eventually relapse.

Although CD19 has proved to be an attractive and effective target for chimeric antigen receptor (CAR) T-cell therapies in hematologic malignancies, a significant subset of patients treated with this groundbreaking form of immunotherapy eventually relapse, with some developing a more aggressive and harder-to-treat form of their disease.1-3

CAR therapies are composed of genetically engineered T cells designed to target tumor-associated antigens (TAAs). The 2 CAR T-cell therapies that the FDA has approved thus far, tisagenlecleucel (Kymriah) and axicabtagene ciloleucel (axi-cel; Yescarta), target the CD19 protein, which is highly expressed in a variety of B-cell malignancies.

As investigators seek to understand the mechanisms of resistance that develop with these and other CD19-directed anticancer therapies, some are turning their attention to alternative targets that could offer a way to stay 1 step ahead of evolving cancers. In this respect, CD22 shows particular promise.

CAR T-cell therapies that target CD22 are beginning to demonstrate potential value, particularly in patients who have relapsed following treatment with CD19 CARs. Additionally, CD22 is becoming an important target in its own right in hematologic malignancies. Two novel therapies directed against CD22, inotuzumab ozogamicin (Besponsa) and moxetumomab pasudotox-tdfk (Lumoxiti), have gained FDA approvals during the past year.

CD22 is a glycoprotein and a member of the immunoglobulin G gene superfamily. It is expressed across all stages of B-cell differentiation but absent from mature plasma cells. Like CD19, CD22 expression is restricted to B cells and helps regulate the immune response through B-cell receptor signaling (FIGURE).4 Additionally, CD22 is highly expressed in B-cell malignancies. Importantly, CD22 expression is usually retained in tumors that have lost CD19 expression. In the earlier stages of B-cell development, CD22 is found inside the cell, though later on it is expressed on the membrane. It plays a number of different roles. Significantly, it has an inhibitory function in B-cell signaling, establishing a base-line level of antigen binding that must be exceeded prior to B-cell activation, to help keep humoral immunity in check.5,6

CD19 Targeting Makes Impact

CAR T-cell therapies directed at CD22 are being studied in a growing number of clinicaltrials throughout the world and in preclinical programs. In the United States, academic centers are leading several phase I trials testing novel approaches, including CAR combinations (TABLE). Meanwhile, at least 2 dozen other trials are ongoing in China and Europe, according to clinicaltrials.gov.CAR T cells are a form of adoptive cell therapy in which the effector cells of the immune system are transplanted into a patient to boost the antitumor immune response. The T cells are genetically engineered outside the body to express a CAR that is designed to target a TAA. CARs are artificial T-cell receptors (TCRs) made of the single-chain variable fragment from an anti- body, which confers antigen specificity, fused to an intracellular portion composed of the CD3ζ chain of the TCR and 1 or more costimulatory domains, which trigger T-cell activation. In this way, an antibody-like sensitivity is essentially grafted onto a T cell.7,8

Figure. CAR T Cells in Signaling Network4

BCR indicates B-cell receptor; CAR, chimeric antigen receptor; CCR4, CC chemokine receptor 4. CAR-modified T cells targeting CD19 and CD22 transmembrane proteins on a malignant B cell. T-cell activation leads to apoptosis of the cancer cell.

Last year marked the first FDA approvals of this type of immunotherapy. These approvals and much of the research conducted to date have focused on targeting the TAA CD19—an ideal target for the treatment of leukemia and lymphoma thanks to its restriction to B cells and its frequent expression in B-cell malignancies.

Tisagenlecleucel gained the FDA’s approval in August 2017 for the treatment of patients up to 25 years of age with B-cell precursor acute lymphoblastic leukemia (ALL) that is refractory or in second or later relapse. The decision was based on findings from the single-arm, phase II ELIANA trial, in which the overall response rate (ORR) was 81%, including complete remission (CR) in 60% of patients.9 The drug subsequently was approved for treating adults with relapsed or refractory (r/r) B-cell lymphoma after 2 or more lines of systemic therapy.

In October 2017, axi-cel was approved for treating patients with r/r large B-cell lymphoma after 2 or more lines of systemic therapy. The phase II ZUMA-1 trial demonstrated an ORR of 82%, with CR in 54% of patients 23 years and older (median age, 58 years; range, 23-76) with refractory large B-cell lymphomas.10

Mechanisms of Resistance Identified

CD19 has also proved an effective target in other forms of immunotherapy. In 2014, the FDA approved blinatumomab (Blincyto) for treating adult patients with r/r Philadelphia chromosome—negative precursor B-cell ALL. Blinatumomab is a new type of antibody format: a bispecific T-cell engager that binds to CD19 on B-lineage cells and CD3 expressed on T cells. In a single-arm, multicenter, open-label study of 189 patients, 43% of patients achieved CR or CR with partial hematologic recovery of peripheral blood counts within 2 cycles of blinatumomab.11Although CD19-targeted CAR therapies have undoubtedly produced dramatic and unprecedented responses in patients with B-cell malignancies, some patients do not respond, and among those who do respond, a significant proportion relapse. Because these therapies have been introduced only in the past several years, understanding of their durability is still evolving. Relapse rates of 10% to 20% have been observed in pediatric and adult patients with B-cell ALL who were treated with CARs, although rates in other studies have been higher.1,2 Some experts have observed differences in relapse rates and patterns, depending on the cancer type.12 As a result, research efforts have begun to focus on understanding the driving forces behind this resistance.

Although the field is still in its infancy, 2 major types of resistance have been uncovered: CD19-positive and CD19-negative relapse. In the first scenario, the tumor retains CD19 expression and resistance has been linked to poor T-cell activity or persistence, allowing tumors to thrive in the presence of CD19-targeted therapies.

In the second scenario, the recurring tumor is a different phenotype from the original pretreatment tumor, having apparently lost expression of the CD19 antigen. A number of possible mechanisms explaining how this occurs have been proposed. These involve genetic or epigenetic alterations (deletions, mutations, or alternative splicing of the mRNA) that result in the outright loss of the CD19 gene or in the expression of a CD19 gene that lacks the epitope recognized by CAR T cells.

CD19 Loss is a Major Factor

Alternatively, CD19 loss has been reported to occur via a process known as lineage switching. B cells are highly plastic, and their development is greatly influenced by other developing cells and by cues from the surrounding microenvironment. Acute leukemias that are initially classified as lymphoid at diagnosis can show the opposite myeloid lineage at relapse. This mecha- nism of resistance is frequently associated with the co-occurrence of MLL gene rearrangements.1,13-16Studies conducted to date suggest that CD19 antigen loss is a major form of resistance associated with CD19-targeted CAR T-cell therapy. This is also observed, albeit likely at a lesser frequency, with other types of CD19-targeted therapy. In a study of CD19-targeted CAR T cells in 59 pediatric patients with r/r B-cell ALL, 93% of patients achieved CR.

Table. CD22-Directed CAR T-Cell Trials in the United States10

ALL indicates acute lymphoblastic leukemia; B-ALL, B-cell acute lymphoblastic leukemia; CAR, chimeric antigen receptor; DLBCL, diffuse large B-cell lymphoma; EGFRt, truncated form of the human epidermal growth factor; FL, follicular lymphoma; NCI, National Cancer Institute; NHL, non-Hodgkin lymphoma; r/r, relapsed or refractory; TCR, T-cell receptor.

Twenty patients subsequently relapsed after CR, and 13 of these patients (65%) had CD19-negative disease at relapse.17 Two separate studies of adult patients with B-cell ALL were conducted at Memorial Sloan Kettering Cancer Center (MSK) in New York, New York, and Fred Hutchinson Cancer Research Center in Seattle, Washington. In the MSK study, among 87% of responders, the rate of CD19-negative relapse was 14%.18 The Seattle research group reported that 2 of the 29 evaluable patients experienced CD19-negative relapse after first achieving CR, including 1 case of lineage switch.19

Thus, CD19-negative relapses have been reported across different patient populations treated with different CAR constructs. In addition, evidence suggests that CD19 loss could occur in other types of B-cell malignancy beyond ALL following treatment with CD19-targeted CAR T-cell therapy, including patients with chronic lymphocytic leukemia and diffuse large B-cell lymphoma.

To date, the frequency of CD19 antigen loss appears to be highest in patients with ALL; however, it is not yet clear whether this reflects the greater clinical experience with this type of malignancy or genuine differences in disease biology.

In patients who relapsed after blinatumomab treatment, several studies have examined the frequency of CD19 negativity. In 16 of 20 patients with adult B-cell ALL who achieved CR with blinatumomab, there were 2 CD19-negative relapses. All 4 relapses occurred in patients who did not undergo stem cell transplantation, of which 2 were CD19 negative and 2 were extramedullary relapses.20,21 In a separate study, in adult patients with B-cell ALL, there were 10 relapses, and 3 of them were CD19 negative.22

Most recently, the outcomes of 68 adult patients with r/r B-cell ALL after treatment with blinatumomab were evaluated. Overall, only 8% of patients lost CD19 expression, although the authors pointed out the short duration of exposure to blinatumomab of patients in this trial. The lower selective pressure could have influenced the frequency of CD19 negativity.23

However, if the results of these studies are borne out in larger trials following longer exposure to blinatumomab, then a substantial proportion of patients who relapse after blinatumomab treatment may still be eligible for CD19-targeted CAR T-cell therapy.

The current understanding of CD19 antigen loss is limited, but investigators are homing in on this area to learn how and when it occurs and the potential risk factors for this type of relapse. Emerging evidence suggests that preexisting CD19-negative clones may be present prior to treatment and that they are selected for during CD19-targeted treatment.2,15

CD22 Joins Next Wave of New Targets

CD19 antigen loss also has implications for monitoring disease progression in patients with B-cell ALL. Minimal residual disease and disease progression are often evaluated by quantifying the presence of CD19-positive cells using flow cytometry. However, if the loss of CD19 drives progression, these cells are not captured by conventional flow cytometry. In cases of CD19- negative relapse, it will be necessary to move away from exclusively relying on CD19 to monitor disease progression. Results from several studies have suggested CD22 as among potential alterna- tives or additional markers in this respect.24,25Two Novel Drug Approvals

Regardless of the underlying mechanism, the prognosis for patients with r/r disease after CD19-targeted therapy is poor. This has prompted investigators to seek out alternative targets for CAR T cells and other immunotherapies for the treatment of B-cell malignancies.

CD22 has already demonstrated its value as a therapeutic target in patients with B-cell malignan- cies. Inotuzumab ozogamicin is an antibody—drug conjugate (ADC) in which a CD22 monoclonal antibody is linked to a cytotoxin from the calicheamicin class. The drug received regulatory approval in 2017 for the treatment of patients with r/r ALL irrespective of their Philadelphia chromosome status.

The approval was based on the results of the phase III INOVATE ALL trial in which inotuzumab ozogamicin was compared with chemotherapy in 326 patients. CR was observed in 35.8% of patients treated with the ADC, and the median duration of remission (DOR) was 4.6 months compared with a 17.4% CR rate and a 3.1-month median DOR in those treated with chemotherapy. The most common adverse events experienced by patients treated with inotuzumab ozogamicin included thrombocyto- penia, neutropenia, infection, anemia, and fatigue.26

In September 2018, the FDA approved moxetumomab pasudotox-tdfk for treating adults with r/r hairy cell leukemia (HCL) who have received at least 2 prior systemic therapies, including treatment with a purine nucleoside analogue. The drug is a first-in-class CD22-directed cytotoxin composed of a murine immunoglobulin variable domain genetically fused to a Pseudomonas exotoxin that inhibits protein synthesis.

The approval was based on findings from a single-arm study that included 80 patients with HCL in which moxetumomab pasudotox induced an ORR of 75%, including a CR lasting for over 180 days (durable CR) for 30% of patients. The agent was approved with a boxed warning regarding the potential for grade 3/4 capillary leak syndrome and hemolytic uremic syndrome, which occurred at the rate of 3.8% each in the safety population.27

Early CD22 CAR Findings

Because the development of CARs directed at CD22 is in the preclinical and early stages, few reports of efficacy have been published.

Earlier this year, investigators at the National Cancer Institute reported findings from a first- in-human trial of CD22-targeting CAR T cells. The results of the small, dose-escalation study, published in Nature Medicine, suggested that this type of CAR T-cell therapy may have comparable efficacy to that of CARs targeting CD19 in patients with B-cell ALL.28

Safety and efficacy were evaluated in 21 pediatric and adult patients, with a median age of 19 years, with r/r B-cell ALL, all of whom had undergone at least 1 prior hematopoietic cell transplantation. Most patients had also previously received CD19-targeted therapy.

Patients were treated with either 3 x 105 cells/ kg, 1 x 106 cells/kg, or 3 x 106 cells/kg. Sixteen patients experienced cytokine release syndrome (CRS); however, this was most commonly grade 1 or 2. Dose-limiting toxicities occurred in 1 patient at the first dose level. The recommended phase II dose was established at 1 x 106 cells/kg. The CR rate was 73% among patients receiving this dose and 57% across all dose levels, with a median DOR of 6 months. This study was the first to demonstrate the potential of CD22-targeted CAR T cells in their own right in patients with CD19 CAR—naïve tumors and—most excitingly—in overcoming resistance to CD19-targeted therapies. CD19 negativity was observed in 10 patients in the study, 9 of whom had undergone CAR T-cell therapy and 1 of whom had been treated with blinatumomab.

In another study, CD22-targeted T cells were specifically tested as salvage therapy in 15 pediatric patients with B-cell ALL who relapsed after or were refractory to CD19-targeted CAR T-cell therapy. In the open-label, single-center, single-arm study, CAR T cells were administered at a dose of 8.2 x 10.05 cells/kg in patients who had not undergone stem cell transplantation and 0.9 x 10.05 cells/kg in those who had.

Avoiding Antigen Loss

The results were presented at the 23rd Congress of the European Hematology Association in Stockholm, Sweden, in June. Thirty days after infusion, the CR rate was 80% and there was also 1 partial response. CD22-targeted CARs were well tolerated, with only mild cases of CRS.29Studies of CD22-targeted CARs have demonstrated that some patients also relapsed after treatment and that loss of CD22 expression is a mechanism of relapse; for example, 7 of 8 patients who relapsed in the first-in-human trial demonstrated dimin- ished CD22 expression.28

This has prompted questions about how CD22 targeted CAR T-cell therapies should best be used. Should they be used as salvage therapy in patients who have relapsed on CD19-targeted therapy, or should they be used simultaneously with CD19-targeted therapy to prevent the development of resistant variants and reduce the risk of treatment failure? Ongoing clinical trials are testing a variety of scenarios, including sequential treatment with CD19-targeted therapy followed by CD22-targeted therapy, mixed infusions of CAR T cells targeting both antigens at the same time, and bispecific and even trispecific CAR constructs.

In a presentation at the 2018 ASCO Annual Meeting, investigators from Baylor College of Medicine detailed the development of 2 trivalent CAR T-cell products that target CD19 and CD22 in addition to CD20. In the first, dubbed a TriCAR, CAR T cells express 3 different CAR constructs, while in the second, called a sideCAR, T cells express a single CAR targeting CD19 and a second bispe- cific CAR targeting CD20 and CD22. Preclinical data demonstrated that both types of CAR T cells were more effective at killing cancer cells than CD19-targeted CAR T cells alone, at lower effector cell-to-target cell ratios.30