Carvykti: half of early-line myeloma patients in five-year treatment-free remission

Half of the patients in a 20-person cohort are alive, progression-free, and taking no myeloma therapy at all, five years after one infusion of Carvykti. Johnson & Johnson and Legend Biotech reported on 25 September 2026 that in CARTITUDE-2 Cohort A, at a median follow-up of 60.7 months, 10 of 20 patients (50%) remained alive and progression-free without further anti-myeloma treatment at least five years after a single ciltacabtagene autoleucel infusion ^[1][2]. Median progression-free survival was 60.5 months, median overall survival was not reached, and the five-year overall survival rate was 69.2% ^[1]. The data were presented at the 2026 International Myeloma Society Annual Meeting in Glasgow (Abstract #PA-288) ^[1].

Why it matters

A BCMA-directed autologous CAR-T given once, with half the cohort off all treatment at five years, reframes the durability question for the whole class. Carvykti's US label covers relapsed or refractory multiple myeloma after at least one prior line including a proteasome inhibitor and an immunomodulatory agent, in patients refractory to lenalidomide ^[3]. Endpoints News framed the update as adding to the argument that Carvykti could be a cure ^[4]. That word is press and analyst framing, not a regulatory or peer-reviewed conclusion — but the underlying benchmark a competitor now has to clear is concrete: treatment-free remission measured in years, not response rate measured at first assessment.

Safety with longer follow-up was consistent with the known Carvykti profile, with no new CAR-T-related neurotoxicity reported ^[1]. The cohort is small, and the signal to watch sits in the tail: since the roughly 30-month analysis, one patient developed a new hematologic malignancy (acute myeloid leukemia) and two deaths occurred, from progressive disease and a new cancer ^[1]. In a 20-patient cohort, a single second primary malignancy is a differentiation lever any challenger will press.

Who is exposed

Allogeneic BCMA CAR-T developers are building the challenger case on response depth at much earlier follow-up. Caribou also reported a case in which a patient previously treated with ciltacabtagene autoleucel achieved a complete response with CB-011 after eight prior lines ^[5]. The read for that program: the response numbers are strong, but the bar Carvykti just set is durability off treatment at five years, and CB-011's public follow-up does not yet reach that horizon.

Commercially, the incumbent is not standing still.

What to watch

The five-year data are a conference abstract (PA-288); Prognyx could not confirm from public sources whether they will appear in a peer-reviewed journal. The prior CARTITUDE-4 subgroup analysis at ASCO 2026 showed progression-free and overall survival benefit across high-risk and standard-risk cytogenetics among bridging-therapy responders, with 30-month overall survival above 85% and no cases of IEC-associated Parkinsonism ^[9]. Watch whether the second-malignancy tail in longer follow-up shapes payer and physician framing as the cure narrative advances.

The now-what

Three moves are on the table for a competing myeloma developer. First, reset the internal durability benchmark from response rate to treatment-free remission at five years — that is the number your data will be measured against. Second, if you run an allogeneic or bispecific program, decide now whether to power a study long enough to speak to durability, because response depth alone no longer answers the incumbent. Third, mine the second primary malignancy signal: in a small cohort it is a genuine open question, and a cleaner long-term safety profile is a defensible wedge.

Verdict — threat level: high for BCMA challengers; window to act: this development cycle. Carvykti has moved the myeloma CAR-T conversation from depth of response to years of treatment-free survival, and any program that cannot speak to durability is now arguing on the wrong axis.

What Prognyx could not verify

"Cure" and "functional cure" are framing from J&J, Legend and Endpoints News, not a regulatory or peer-reviewed finding ^[1][2][4]. No ClinicalTrials.gov record for this cohort was available to confirm trial status directly. The causes of relapse or progression among the 10 of 20 patients not in treatment-free remission are not detailed in the public materials, and peer-reviewed publication of the five-year data is not confirmed.