Key Ideas

CD19-directed CAR T-cell therapy has transformed care for B-cell lymphomas, but most patients still relapse, often due to loss of the CD19 target. Several multispecific CAR T-cell products targeting CD19 and CD20 together are now showing high response rates and low toxicity in early trials, though it remains too soon to say whether they will become the new standard of care.

A Real and Measurable Problem

CAR T-cell therapy directed against CD19 has revolutionized care for patients with B-cell malignancies, with randomized trials demonstrating clear benefits in progression-free survival, overall survival, and patient-reported outcomes. A single infusion can be curative for some patients with aggressive B-cell lymphoma and several other diseases. Despite this, the majority of patients who receive CD19 CAR T-cell therapy experience either a relapse or progression event, or non-relapse mortality, after treatment.

Over roughly the past decade, since the earliest approvals began in 2017, understanding of resistance mechanisms has deepened considerably. One of the clearest mechanisms is antigen loss: down-regulation or complete loss of CD19 on the surface of lymphoma cells after CAR T-cell exposure. This was first recognized through immunohistochemistry, which showed loss of surface expression or a shift to intracytoplasmic CD19 in a meaningful fraction of relapses. Consortium case series confirmed this in roughly 30% of relapsed cases.

More recently, quantitative flow cytometry using the QuantiBRITE PE assay (BD Biosciences) on more than 100 patients has added precision to this picture. Comparing pre-CAR T-cell antigen expression to relapse samples shows a clear drop in CD19 site density, while CD20, CD22, and CD79b are largely preserved on the same relapsing cells. That pattern points to genuine immunologic selection pressure from a therapy that targets only one antigen. It compounds an existing problem: patients with lower baseline CD19 expression, below roughly 3,000 molecules per cell, tend to have inferior outcomes even before therapy begins, a relationship that holds up in multivariable modeling after adjusting for other prognostic factors.

Why Two Targets Instead of One

Several multispecific CAR T-cell products for lymphoma are now in registration-track development, each combining CD19 and CD20 targeting through a different construct strategy.

Zamtocabtagene autoleucel (zamto-cel), developed by Miltenyi, uses a monocistronic tandem construct with a single 4-1BB signaling domain spanning both CD19 and CD20 binding domains. In third-line large B-cell lymphoma, six-month progression-free survival has been around 55%, with a favorable safety profile: no grade 3 or higher cytokine release syndrome, and grade 3 or higher neurotoxicity in fewer than 5% of patients.

Rondecabtagene autoleucel (ronde-cel), developed by Lyell Immunopharma, also uses a tandem construct, but adds a manufacturing step that selects for CD62L-expressing T cells to enrich the product for memory phenotype cells before infusion. Across second- and third-line settings, complete remission rates have ranged from roughly 60% to 76%, with overall response rates of 80% to 90%, and a similarly favorable safety profile.

The product with the most personal involvement in development is KITE-753, a dual CD19/CD20 CAR T-cell therapy using two independent single-chain variable fragments and two separate signaling domains: CD28 for the CD19 arm and 4-1BB for the CD20 arm. Unlike a monocistronic tandem design, this bicistronic approach gives each target its own complete signaling pathway. KITE-753 is also manufactured in three days, compared with the longer process used for its predecessor, KITE-363.

That distinction between one shared signaling domain and two independent ones is not incidental. In an earlier trial combining CD19 and CD22 targeting in a single tandem construct, with CD22 as the distal domain and CD19 proximal, outcomes looked no different from single-target CD19 or CD20 therapy. When the CD22 arm was tested alone in a monocistronic construct, signaling was markedly impaired, apparently because the proximal CD19 domain interfered with signal transduction through the distal domain. That experience is part of the rationale for a bispecific-signaling design like KITE-753's: reliable, independent signaling through each target, rather than one domain potentially blocking the other.

Early Clinical Results

KITE-753 was evaluated in a phase 1a/1b dose-escalation and dose-expansion trial, structured as an umbrella and basket study alongside its predecessor KITE-363, primarily in large B-cell lymphoma. KITE-363, manufactured through a conventional process similar to existing approved CD19 products, produced a complete response rate of close to 78%, with 71% of complete responses remaining durable.

Building on that, KITE-753 moved to a memory-enriched product with a substantially lower dose, roughly two orders of magnitude below KITE-363, using a highest dose level of 200,000 CAR-positive cells per kilogram. The patient population reflected a typical relapsed/refractory phase 1 lymphoma cohort: median age approaching 60 (oldest patient treated was 84), a substantial proportion with stage III or IV disease, and a median of two prior lines of therapy (range 1 to 7). No bridging chemotherapy or radiotherapy was permitted, aside from steroids.

Safety was notable. Only one patient, who had heavy bulky disease, experienced grade 3 cytokine release syndrome; no patients experienced grade 3 or 4 ICANS, with most neurotoxicity graded 1 or 2, and roughly 60% of patients developing no neurotoxicity at all. On efficacy, the best complete response rate at the highest, recommended phase 2/3 dose was 78%, with an overall response rate of about 89%. At six months, the complete response rate remained around 71%, with data still maturing; longer follow-up on durability was presented separately.

Pharmacodynamically, despite the substantially lower cell dose, CAR T-cell expansion was comparable to axicabtagene ciloleucel (axi-cel) from the ZUMA-1 experience, though peak expansion was somewhat delayed (day 14 versus day 7 for axi-cel), consistent with the bicistronic, memory-enriched product taking a different expansion trajectory. Cytokine profiling showed interleukin-2, interferon gamma, and granzyme B levels coinciding with CAR T-cell expansion, while interleukin-15 levels were comparatively low, in line with the low rate of higher-grade neurotoxicity observed.

Will This Change the Standard of Care?

The problem multispecific CAR T-cell therapy is trying to solve is real and measurable: more than half of patients with lymphoma progress via antigen loss or down-regulation. Early efficacy with multiantigen targeting is encouraging, and toxicity across these products has been strikingly low, with complete response rates clustering in the 70% to 80% range.

It is not yet possible to call this a change in standard of care. These remain early-phase trials with relatively short follow-up, and there is no randomized data yet comparing multispecific CAR T-cell therapy against single-target CD19 products. It is also difficult to fully separate the effect of dual-antigen targeting from other manufacturing changes happening in parallel, including shorter manufacturing times and T-cell memory enrichment, both of which have independently been associated with better outcomes.

The benchmark to beat matters here too. In the ZUMA-7 trial, one-year progression-free survival was 55% and three-year progression-free survival was 45%, without routine use of bridging or debulking therapy. Real-world data from one institution, in a cohort of nearly 500 patients treated with the same FDA-approved axi-cel and liso-cel products used in that trial, showed meaningfully better outcomes: roughly 65% one-year progression-free survival and 60% at three years. Multispecific CAR T-cell products will need to exceed that real-world benchmark, not just the original trial numbers, to represent a genuine advance. Durability beyond one or two years is the open question, and randomized second-line trials are now underway, including registrational head-to-head comparisons against investigator's choice of an approved CD19 product, to help answer it.

For Patients

CAR T-cell therapy has transformed treatment for many people with aggressive B-cell lymphoma, sometimes curing the disease with a single infusion. But most patients who receive it still eventually relapse, and one common reason is that the lymphoma stops making the specific target protein, CD19, that the therapy was designed to find.

Researchers are now testing a new generation of CAR T-cell therapies designed to recognize two targets on the lymphoma cell, CD19 and CD20, instead of just one. The idea is that even if the cancer loses one target, the therapy can still find and attack it through the other. Early studies of several of these dual-target therapies are showing high remission rates with low rates of serious side effects like cytokine release syndrome.

These are still early-stage trials, and it is not yet known whether dual-target CAR T-cell therapy will become the new standard treatment or how durable these remissions will be over several years. Larger, randomized trials comparing these newer therapies directly against current standard treatments are now underway. Patients interested in these approaches should ask their oncology team whether a clinical trial testing dual-target CAR T-cell therapy is available and appropriate for their situation.

Key Takeaways

  • Antigen loss, most often loss of CD19, drives a substantial share of relapses after CD19 CAR T-cell therapy, and low baseline CD19 expression before treatment is independently associated with worse outcomes.

  • Several CD19/CD20 dual-targeting CAR T-cell products (zamtocabtagene autoleucel, rondecabtagene autoleucel, and KITE-753) are showing complete response rates in the range of 60% to 89% in early-phase trials, with low rates of severe cytokine release syndrome and neurotoxicity.

  • Construct design matters: a bicistronic design with two independent signaling domains appears to avoid the signaling interference seen in earlier tandem CD19/CD22 constructs, where a proximal domain impaired distal-domain function.

  • Manufacturing changes, including shorter production times and memory T-cell enrichment, are occurring alongside dual-antigen targeting and make it difficult to isolate how much of the benefit comes from targeting two antigens specifically.

  • Real-world outcomes with currently approved single-target CD19 products already exceed the original pivotal trial benchmarks, raising the bar that multispecific products will need to clear; randomized second-line trials, including head-to-head comparisons, are underway to determine whether they do.

References

  1. Neelapu SS, Locke FL, Bartlett NL, et al. Axicabtagene ciloleucel CAR T-cell therapy in refractory large B-cell lymphoma. New England Journal of Medicine. 2017;377(26):2531-2544.

  2. Westin JR, Oluwole OO, Kersten MJ, et al. Survival with axicabtagene ciloleucel in large B-cell lymphoma. New England Journal of Medicine. 2023;389(2):148-157.

  3. Murakami JL, Guevara CI, Cai Q, et al. KITE-753: an autologous rapid manufactured anti-CD19/CD20 CAR-T product for the treatment of B-cell malignancies. Presented at the American Society of Hematology Annual Meeting, 2024.

  4. Dahiya S, et al. A phase 1 study of KITE-753 or KITE-363 in patients with relapsed/refractory B-cell lymphoma: initial safety and preliminary efficacy of KITE-753 and updated results of KITE-363. Presented at the American Society of Hematology Annual Meeting, 2025.

  5. Borchmann P, Michael Overstijns T, et al. Zamtocabtagene autoleucel, a tandem CD20-CD19 directed CAR-T cell therapy as second-line treatment for relapsed/refractory large B-cell lymphoma: primary analysis of the randomized, pivotal DALY 2-EU study. Presented at the American Society of Hematology Annual Meeting, 2025.

  6. Lyell Immunopharma. Positive new clinical data demonstrating high rates of durable complete responses from the phase 1/2 trial of LYL314 (now rondecabtagene autoleucel) for the treatment of aggressive large B-cell lymphoma. Company announcement, 2025.

  7. ClinicalTrials.gov. PiNACLE-H2H: a phase 3 randomized controlled trial of rondecabtagene autoleucel versus investigator's choice of CD19 CAR T-cell therapy. NCT07188558.