Alireza Mansouri, Associate Professor at Penn State Health, shared on X:
“With SNO2026 coming up, a question to think about leading into some of the oral sessions I am excited about: high-risk patient, suspicious symptoms or equivocal MRI, LP cytology negative, but CSF ctDNA is positive. Do you call it leptomeningeal disease and treat, or repeat the LP and wait? Cytology is still the standard but the molecular signal is hard to ignore!
The data: in a prospective breast cancer cohort (n=30), CSF ctDNA was detected in all 24 patients with LMD, including the half whose first cytology was negative or suspicious. None of the 6 without LMD had detectable ctDNA. (Fitzpatrick et al, Clin Cancer Res 2022)
In a retrospective MGH study (White et al, JAMA Netw Open 2021), across 48 CSF samples, cfDNA sensitivity was 93% vs 72% for cytology, with 100% specificity for both. cfDNA also identified LMD in patients with persistently negative cytology.
The limitation: these are small, mostly single-center cohorts. In the MGH study, 3 controls with a parenchymal tumor abutting the CSF were cfDNA-positive. A positive result shows tumor DNA in the CSF. It does not prove the leptomeninges are seeded necessarily.
The EANO-ESMO framework right now calls LMD confirmed only with positive cytology, and advises one repeat LP when cytology is negative or equivocal. Until we get more data, perhaps we should treat positive CSF ctDNA as a reason to repeat the LP and re-image, not as a diagnosis on its own yet.
Question for colleagues
CSF ctDNA positive, cytology negative, MRI equivocal. What would make you start LMD-directed therapy? A second positive sample, a rising tumor fraction, a mutation matching the primary? Curious how others are handling this.”
You may also be interested in: