KEY POINTS
- This multi-institutional retrospective study included 221 patients with IDH-wildtype glioblastoma and leptomeningeal dissemination, including 66 with dissemination at initial diagnosis and 155 who developed it at recurrence.
- Prognosis differed markedly by timing. From diagnosis of leptomeningeal dissemination, median overall survival was 12.1 months for primary versus 4.4 months for recurrent LMD, while median progression-free survival was 6.7 versus 2.9 months, respectively.
- Treatment associations were very different in the recurrent setting. Only 28/155 patients received RT for LMD, including 15 treated with comprehensive fields; systemic therapy was given to 74/155, most commonly bevacizumab.
- After multivariable adjustment, comprehensive RT field coverage was associated with substantially better survival in recurrent LMD: HR 0.40 for overall survival and 0.46 for progression-free survival. Systemic therapy was independently associated with HR 0.35 for overall survival and 0.53 for progression-free survival.
- Bevacizumab did not appear uniquely responsible for the systemic-therapy signal. Bevacizumab and non-bevacizumab regimens showed similar associations with survival, suggesting that receipt of active systemic therapy itself may be more important than a specific regimen in this dataset.
- Broader neuraxial irradiation was not benign. Among 18 patients receiving craniospinal or whole-spine RT, grade 3-4 lymphopenia occurred in 88.9%, leukopenia in 83.3%, neutropenia in 72.2% and thrombocytopenia in 44.4%, frequently requiring growth-factor or transfusion support.
- The apparent treatment benefit is highly vulnerable to selection bias: fitter patients were more likely to receive aggressive therapy. LMD diagnosis was also predominantly MRI-based, spinal imaging was incomplete, treatment schedules were heterogeneous, and nonhematologic toxicity may have been underreported.
CLINICAL TAKEAWAY
Recurrent leptomeningeal dissemination in IDH-wildtype glioblastoma remains a very poor-prognosis situation, but automatic transition to supportive care may not be appropriate for every patient. In carefully selected patients with good performance status, broader RT coverage plus systemic treatment may be reasonable to consider, although the retrospective associations cannot prove treatment benefit and must be balanced against substantial hematologic toxicity.
SOURCE
International Journal of Radiation Oncology, Biology, Physics