Peak lymphocyte count after cilta-cel flags delayed neurotoxicity risk

Design

Single-centre kinetics study plus multicentre retrospective cohort

Population

532 adults with relapsed or refractory myeloma given cilta-cel (256) or ide-cel (276)

Primary outcome

Delayed neurotoxicity by peak absolute lymphocyte count

Effect

Peak ALC ≥3 × 10³/µL (or ≥2.5 after twofold rise): sensitivity 81%, specificity 59%

Delayed neurotoxicity, including parkinsonism, is a recognised complication of BCMA-directed CAR-T therapy in myeloma, but there has been no routine laboratory warning sign. This study measured CAR-T expansion by flow cytometry in 90 patients and then tested the absolute lymphocyte count as a substitute in a multicentre cohort of 532 (256 given ciltacabtagene autoleucel, 276 idecabtagene vicleucel).

Cilta-cel expanded more than ide-cel (median peak 106 vs 49 cells/µL). With ide-cel, higher expansion tracked response; with cilta-cel it did not, and very rapid expansion was linked to delayed neurotoxicity (median peak about 1,009 vs 96 cells/µL in those with and without it).

In the larger cohort, a higher peak lymphocyte count was associated with delayed neurotoxicity, particularly parkinsonism after cilta-cel. A peak of at least 3 × 10³/µL, or at least 2.5 × 10³/µL after a twofold or greater rise, identified those at raised risk with 81% sensitivity and 59% specificity.

For laboratories supporting CAR-T programmes, the full blood count becomes a monitoring tool. A high peak should be reported promptly to the treating team; whether acting on it prevents neurotoxicity has not been tested.

• Track the absolute lymphocyte count serially after cilta-cel, not just at single time points.
• Flag a peak of 3 × 10³/µL or more, or 2.5 × 10³/µL after a twofold rise, to the treating team.
• Expect many false alarms: specificity was 59%.
• Note that the threshold was derived mainly for cilta-cel; ide-cel showed a different pattern.
• Report flow-based CAR-T counts where available, as the lymphocyte count is only a surrogate.

Why it matters

It turns a test every laboratory already runs into an early warning for a complication that can be permanent.

Don't overread it

The threshold identifies risk; no study yet shows that acting on it prevents neurotoxicity.

The statistics, in plain English

A sensitivity of 81% means about four in five patients who developed delayed neurotoxicity crossed the threshold; a specificity of 59% means about two in five who did not also crossed it. That makes it a screening flag, not a diagnosis. The association is observational, so it does not show that high expansion causes the neurotoxicity.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.