MRD-Negative After CAR-T for B-ALL: Consolidate With Transplant, or Trust the Cells Still Working
Consolidative transplant would address her leukemia's known tendency to relapse early — and would also eliminate the CAR-T cells that are the reason she's in remission today.
Ava R., a 17-year-old high school senior finishing her last semester before graduation, was diagnosed with B-cell acute lymphoblastic leukemia at 13 and achieved a first remission that held for just under eighteen months before relapsing — a marrow relapse detected on a routine surveillance visit, not because she felt unwell. That timing matters clinically: relapse within eighteen months of initial diagnosis is a recognized higher-risk feature, one some programs treat as reason enough to pursue consolidative transplant even after a strong second response. She received tisagenlecleucel, a CD19-directed CAR-T product, and achieved a bone marrow remission with no measurable residual disease by next-generation flow cytometry six weeks later — about as deep a response as this therapy produces.
What makes the next decision genuinely hard, not just cautious, is that pursuing transplant now would destroy the very thing keeping her in remission. Her ongoing B-cell aplasia — the near-total absence of normal B cells in her blood — is not itself a problem to fix; it's a marker that her CAR-T cells are still alive, active, and patrolling for CD19-positive cells, malignant or otherwise. In ELIANA (Maude), the pivotal trial for this exact product, a meaningful subset of patients who maintained B-cell aplasia this way stayed in durable remission without ever proceeding to transplant — and ELIANA enrolled patients aged 3 to 21, which is to say it enrolled Ava. The data usually raised against that come from ZUMA-3, brexucabtagene autoleucel in adult B-ALL, where a majority of responders went on to transplant and the relapse picture without consolidation looks worse. ZUMA-3's median age was 40 — a real finding about a different product in a different population, pointing the opposite way, which is why the room is divided rather than merely cautious. Proceeding to transplant would ablate her existing CAR-T cells along with her own marrow, trading an active, ongoing immune surveillance mechanism that is currently working for the toxicity, GVHD risk, and long-term effects — including on fertility and growth — of allogeneic transplant in a 17-year-old.
One thing that would have ended this discussion before it started is absent from her workup. Pre-CAR-T imaging showed disease confined to the marrow, no extramedullary involvement, and no new high-risk cytogenetic lesions beyond diagnosis — extramedullary relapse after CAR-T carries a distinct and generally worse prognosis, and had it been present it would have pushed toward transplant regardless of what her aplasia was doing. Fertility preservation counseling and oocyte cryopreservation, discussed and completed before her CAR-T infusion at her family's request, mean that at least one long-term cost of eventual transplant has already been partly addressed, whenever that decision is finally made.
Deciding on consolidative transplant at week 6 of remission
Ava relapsed within eighteen months of her original diagnosis — a recognized higher-risk feature that predicts a real chance of relapsing again even after a strong response to salvage therapy. Her current remission is as deep as this therapy produces, MRD-negative by flow cytometry, and she has a fully matched unrelated donor already identified. Consolidating that response with transplant now addresses the underlying risk her relapse timing already told us about, rather than hoping the current remission holds on its own.
I take the early-relapse risk seriously, but I don't think it settles this on its own. Her ongoing B-cell aplasia isn't just a side effect — it's direct, measurable evidence that her CAR-T cells remain active and are actively patrolling for CD19-positive cells right now. In ELIANA, the trial for this exact product in patients aged 3 to 21, a meaningful share of those who maintained aplasia this way stayed in remission without ever needing transplant.
Proceeding to transplant doesn't just add a risk — it eliminates the mechanism that's currently working, in exchange for graft-versus-host disease risk and real long-term harm to her fertility and growth at seventeen. The adult data suggesting higher relapse without consolidation is ZUMA-3, brexucabtagene autoleucel, median age 40 — a different product in a different population, not patients like her.
Both of you are reading real, different pieces of evidence correctly, and I don't think either fully resolves the other's concern. What I'd propose is using the thing you both agree is genuinely informative — her B-cell aplasia — as an explicit decision trigger rather than forcing a yes-or-no today. If her aplasia resolves, meaning her CAR-T cells are no longer active, that's real, new evidence her surveillance mechanism has stopped working, and transplant should proceed promptly. If it persists through a defined monitoring period, that's evidence favoring continued observation.
This doesn't split the difference for its own sake — it waits for the one piece of information that would actually change which position is right for her.
Agreed: no transplant today. B-cell aplasia will be monitored monthly by flow cytometry alongside routine marrow surveillance; her donor and a conditioning plan are held in reserve so transplant can proceed promptly if aplasia resolves, rather than restarting a donor search from scratch at that point.
Not agreed, and stated as an explicit open disagreement rather than resolved by the monitoring plan: whether this was the right call at all, or whether it should have been consolidative transplant now regardless. The pediatric hematologist-oncologist views the monitoring plan as a reasonable compromise but would have recommended transplant outright given her relapse-timing risk if the decision had been hers alone; the oncologist views waiting for the aplasia signal as the more genuinely evidence-based path. Both agreed to the trigger-based plan without agreeing on what it should show.