In multiple myeloma, two therapeutic formats now target the same validated surface antigen. Both redirect T cells. The clinical profiles — response rates, durability, toxicity, patient eligibility — are not the same. Explaining why requires looking beyond the format.
The comparison between CAR-T and T-cell engagers is one the field makes constantly, usually framed as a format competition. It is the wrong frame. What separates these therapies is not primarily whether the T-cell receptor was engineered ex vivo or bridged in vivo. It is what each format depends on, what it measures, and what it cannot address regardless of how well it is executed.
The comparison everyone makes
Ide-cel and cilta-cel are CAR-T therapies with FDA approval in relapsed and refractory multiple myeloma. Teclistamab is a bispecific T-cell engager approved for the same patient population. All three target the same validated myeloma surface antigen. All three redirect T cells against tumour cells expressing that antigen. Their pivotal trial response rates, durability curves, toxicity profiles, and eligibility criteria differ meaningfully.
The antigen is not the variable. The antigen is the same. What differs is what each format requires from the patient's biology, what it can measure before dosing, and what selective pressure it imposes on the tumour. Understanding the differences at that level — not the format level — is what determines whether a patient has an option after the first therapy stops working.
What is the same
At the mechanism level, all three therapies activate T cells through CD3 signalling against a tumour antigen. All three have produced complete responses in patients who had exhausted multiple prior lines of therapy, including proteasome inhibitors, immunomodulatory drugs, and anti-CD38 antibodies. All three carry cytokine release syndrome as a class effect, requiring managed step-up dosing or post-infusion monitoring protocols in specialist centres.
In a patient population that had no effective options five years ago, all three represent genuine clinical progress. The convergence on mechanism and on antigen has established that redirecting T cells against this target is a valid therapeutic approach in myeloma. The differences begin when the format constraints become clinically relevant.
What is different: manufacturing
CAR-T is a patient-specific product. After leukapheresis to collect the patient's own T cells, ex vivo genetic engineering installs the chimeric antigen receptor, and the engineered cell population is expanded, quality-tested, and returned — a process measured in weeks from collection to infusion. The patient receives a single administration. The manufacturing window creates eligibility constraints that are not present with off-the-shelf therapies.
Patients who progress rapidly during the wait may not be fit for infusion by the time the product is ready. Patients whose T cells are too exhausted by prior therapy to expand adequately in culture may not produce a viable product. Healthcare systems outside major academic centres may not have the logistical infrastructure to manage the cold-chain and coordination requirements.
T-cell engagers are manufactured at scale from stable cell lines, stored, and available immediately. Monthly subcutaneous or intravenous administration does not require leukapheresis. For patients who cannot wait, or for treatment settings where the CAR-T logistics chain is not available, the access profile is materially different.
What is different: the T-cell pool
CAR-T engineering creates a defined population of T cells with a specified receptor. After infusion, those cells expand in vivo. The durability of response correlates with the persistence of the engineered CAR-T population — and in heavily pre-treated patients, T-cell exhaustion acquired during prior lines of therapy frequently compromises both expansion capacity and long-term persistence.
T-cell engagers depend on the patient's existing T-cell repertoire. If that repertoire is exhausted — reduced in number, skewed toward dysfunctional phenotypes, or impaired in cytolytic capacity — the engager has fewer functional T cells available to form productive immune synapses with tumour cells. This is one reason why functional screening of T-cell engager candidates against primary human immune cells from donors with variable T-cell backgrounds matters: the molecule must perform across the range of T-cell states present in real patients, not just in well-rested cells cultured for assay convenience. A molecule selected for functional performance under these conditions has been evaluated against the biology it will encounter in the clinic.
What is the same problem: antigen escape
The convergence that matters most clinically is the shared architectural vulnerability. Both CAR-T therapies and single-target T-cell engagers direct immune pressure at one antigen. Under that pressure, tumour cells that reduce or eliminate target expression are not killed. They survive. They divide. They eventually repopulate the disease with a phenotype that resists the treatment. This pattern has been documented repeatedly across the approved myeloma therapies targeting this antigen.
This is not a CAR-T problem or a T-cell engager problem. It is an architecture problem. A single-target molecule — whatever format delivers it — imposes a selection pressure that tumour evolution can resolve through one mutational event: downregulate the target. The format does not change the biology of escape. The architecture does.
The right question
The clinical choice between CAR-T and a T-cell engager is real. It involves manufacturing logistics, patient fitness, treatment line, prior therapies received, and institutional capability. It is a decision that matters and that clinicians are navigating every day.
But it is not the question that determines whether a patient who progresses has a next option. A patient who has received and progressed on a single-target therapy — whether CAR-T or a T-cell engager — has already demonstrated that their disease can escape through antigen downregulation. The question for their next treatment is not which format delivers the same target at higher dose or with different scheduling. It is whether the next molecule was designed to address escape-resistant biology: independent killing mechanisms that each arm drives without requiring the other, selected for functional performance across the antigen-expression states present in advanced relapsed disease.
Format selection matters. Architecture matters more. The patients who will benefit most from the next generation of myeloma therapy are the ones for whom the current generation has already run its course — and whose disease requires a different design logic, not a different delivery mechanism for the same one.