Rapid Early Progression in Glioblastoma Patients: Missing the Target

Purpose: A post-operative MRI (MRIpost-op) is routinely used for radiation treatment planning in glioblastoma (GBM) patients, with radiotherapy starting within 6 weeks of surgery. Patients may undergo an additional pre-radiotherapy MRI (MRIpre-RT). We sought to analyze the incidence and extent of rapid early progression (REP) between surgery and initiation of radiotherapy and its impact on radiation treatment volumes.

Methodology: Patients with GBM diagnosed between 2018-2020 who had an MRIpost-op and MRIpre-RT were retrospectively identified. Criteria for REP was based on Modified RANO criteria for GBM patients [1]. Radiation treatment volumes were created per RTOG guidelines [2] using the MRIpost-op and MRIpre-RT and compared to assess changes in treatment volumes. Overall survival (OS) and progression free survival (PFS) was calculated using Kaplan-Meier method with log rank analysis.

Results: We identified 136 patients who were diagnosed with GBM between 2018-2020 at our institution. Of these patients, 55 patients met our inclusion criteria. The median time between MRIpost-op and MRIpre-RT was 26 days. Indications for MRIpre-RT included clinical trial enrollment in 45/55 (82%), new symptoms in 5/55 (9%), and unspecified in 5/55 (9%). REP was identified in 37/55 (67%) of patients with 10/37 (27%) of these patients having disease progression outside of the post-operative clinical target volume 2 (CTV2post-op) and 2/37 (6%) had REP that extended beyond the CTV1post-op radiotherapy field. Using an MRIpost-op for treatment planning led to a median 21.1cc, or 11.2% of the pre-radiotherapy CTV2 (CTV2pre-RT), being undertreated. Patients without REP had a 7% median volume reduction in CTV2pre-RT versus CTV2post-op. Additionally, 34/55 (62%) of patients had a smaller CTV1pre-RT than CTV1post-op. Among patients without REP, there was a 38% median volume reduction in CTV1pre-RT versus CTV1post-op which correlated with a median of 117cc of normal brain tissue being spared from inclusion into the 46Gy CTV1pre-RT volume. Patients with REP had worse median OS (13.1 vs 21.6 months, p=0.0035) and median PFS (4.2 vs 7.6 months, p=0.037) compared to patients without REP.

Conclusions: This is the largest study assessing the impact of REP on radiation treatment volumes. Two-thirds of patients with GBM in our study had REP, and nearly all were asymptomatic, suggesting that REP is being underdetected prior to adjuvant therapy. Utilization of an MRIpre-RT leads to increased identification of REP, improved coverage of gross and subclinical disease, and smaller treatment volumes leading to increased brain sparing for patients without REP. For these reasons, we recommend that all GBM patients undergo a MRIpre-RT for radiotherapy treatment planning.

References: 1. Ellingson, B.M., P.Y. Wen, and T.F. Cloughesy, Modified Criteria for Radiographic Response Assessment in Glioblastoma Clinical Trials. Neurotherapeutics, 2017. 14(2): p. 307-320.
2. Kruser, T.J., et al., NRG brain tumor specialists consensus guidelines for glioblastoma contouring. Journal of Neuro-Oncology, 2019. 143(1): p. 157-166.