Triple-Class-Refractory Myeloma: CAR-T, a Bispecific, or a Second Transplant
A durable first remission after autologous transplant is itself evidence about this patient's biology — but it competes against two newer options with a much closer match to her current, triple-class-refractory disease.
Rosa M., a 63-year-old retired seamstress, still hems and mends clothes for half the families on her street, work she says keeps her hands busy while her body recovers from four years of treatment. Her multiple myeloma was diagnosed nine years ago and achieved a deep, four-year remission after her first autologous transplant — an unusually durable response that tells the team something real about how her disease behaves when it responds. Since then she has relapsed twice, most recently after a daratumumab-based regimen that controlled her disease for only eight months — eight against forty-eight, the same patient, the same disease, which is the shape of the problem in a single ratio. This relapse was caught on routine surveillance rather than by symptoms, a rising free light chain ratio confirmed on two consecutive monthly draws with no new skeletal lesions and stable renal function, and her repeat cytogenetics show no new high-risk markers beyond what was there at diagnosis. Her disease has not changed character. It has only stopped listening. She is now refractory to an immunomodulatory drug, a proteasome inhibitor, and an anti-CD38 antibody — triple-class-refractory, and squarely inside the population studied in the pivotal trials for the newest myeloma therapies.
Those trials describe her more precisely than her own history does at this point. CARTITUDE-1 (Berdeja), the pivotal trial for the BCMA-directed CAR-T product ciltacabtagene autoleucel, enrolled triple-class-exposed patients after at least three prior lines — Rosa is on her fourth — and reported a 97% response rate with a median progression-free survival near 35 months, unusually durable for disease at this stage. Teclistamab, a bispecific antibody redirecting T cells to BCMA-expressing plasma cells without the weeks-long manufacturing wait CAR-T requires, produced a 63% response rate in MajesTEC-1 (Moreau) in a comparable population — real, but well short of CARTITUDE-1's figure. Its long-term cost is a different one from CAR-T's, and worth naming precisely: because BCMA sits on plasma cells rather than on B cells generally, continuous teclistamab produces profound hypogammaglobulinemia rather than the B-cell aplasia a CD19-directed product causes, with a serious infection rate that typically commits the patient to indefinite monthly immunoglobulin replacement. Her own four-year first remission argues for something else entirely — that whatever gave her that response once might do so again, and a second autologous transplant is a real, lower-technology option in patients whose first transplant bought genuine years rather than months.
Choosing fourth-line therapy for triple-class-refractory disease
CARTITUDE-1 enrolled triple-class-exposed patients with a prior-line count close to Rosa's, and reported an overall response rate near 98% with remissions that, at longer follow-up, are running past three years in a meaningful share of patients. That's not just a better number than the alternatives — it's a population match, not an extrapolation, and it's the deepest, most durable response available for disease at her stage.
I agree CAR-T has the strongest efficacy data, and I'm not arguing against it on those grounds. My concern is timing: her M-protein has been rising over six weeks, and CAR-T manufacturing realistically takes four to six weeks from apheresis to infusion. Teclistamab doesn't carry that wait — it's available essentially immediately, with real, if numerically smaller, response rates in the same triple-class-refractory population.
The cost isn't B-cell aplasia, to be precise about it — BCMA is a plasma-cell antigen, so what continuous teclistamab produces is profound hypogammaglobulinemia, and the infection risk that comes with it usually means monthly immunoglobulin replacement indefinitely. That's a real, open-ended burden, not a one-time toxicity like CAR-T's own acute cytokine release risk.
Both of those arguments are about which trial population fits her best, and I think that's the wrong level to settle this at entirely. Rosa's own first transplant gave her four years — that's not a population statistic, it's a direct, individual data point about how her disease behaves when a treatment actually works for her. Second autologous transplants in general produce shorter remissions than a first one, and shorter than what CAR-T is now showing in this disease stage — I'm not arguing it should be the default here.
I'm arguing it shouldn't be dismissed without weighing what her own history already told us, even while the disease-kinetics argument for not waiting on CAR-T's manufacturing timeline is a real and separate point worth taking seriously.
Agreed: proceed to apheresis and ciltacabtagene autoleucel manufacturing, with low-intensity corticosteroid bridging rather than cytotoxic bridging chemotherapy, given her rising but not rapidly accelerating M-protein.
Not fully agreed: whether her own four-year first remission should have carried more weight in the initial discussion than it ultimately did. The transplant physician noted it for the record as a real, individual data point that a population-level trial comparison can't incorporate, without pressing for it to change today's plan — a distinction the team agreed was worth stating rather than letting the final choice read as though her own treatment history had no bearing on the decision at all.