Session Item

Treatment of diffuse gliomas typically consists of surgical resection and subsequent radiotherapy and/or chemotherapy. High-dose radiation can however lead to new enhancing lesions on follow-up scans, which are challenging to distinguish from tumor progression. These post-radiation lesions, including pseudoprogression and radiation necrosis, occur in 20-30% of glioma patients and are thought to stem from cerebral inflammatory responses and structural changes in the cerebral vasculature. As development of these lesions might relate to the effectivity of the treatment, we have investigated the association between post-radiation lesions and overall survival (OS) in patients with diffuse glioma.

We accessed a retrospective database of 144 adult cases with WHO grade II-IV supratentorial gliomas, who received surgery and postoperative MRI within 3 days, and identified 65 patients who received radiotherapy after first surgery and had subsequent MRI follow-up of at least 6 months. Patient follow-up was up to 60 months after start of radiotherapy. Post-radiation injury was defined as a new enhancing lesion on CE-T1 MRI that stabilized or decreased after a minimum of 3 months follow-up (Figure 1). OS was analyzed in a Cox-regression model, consisting of development of post-radiation lesions, age, WHO grade, prescribed dose and concurrent or additional chemotherapeutic treatment.

Figure 1. Patient with post-operative CE-T1 scan (left), onset of post-radiation lesion (middle) and stabilization after 6 months (right).

Post-radiation injury occurred in 26 patients (36%) and had a median volume of 9.9cc. Development of post-radiation lesions was significantly associated with a hazard ratio (HR) of 0.538 [95% confidence interval (CI) 0.296 – 0.977] (Table 1) and these patients had a longer median OS (25 months [95% confidence interval (CI) 20.0-30.0] versus 16 months [95% CI 10.8-21.2]). Factors significantly associated with OS were: 1) WHO grade 4 (compared to WHO grade 2), 2) prescribed dose of 60Gy and 3) chemotherapeutic treatment (Table 1). Among patients with post-radiation injury, lesion volume was not associated with OS

Table 1. Multivariable analysis for OS in diffuse glioma.

We have investigated the relation between development of post-radiation lesions and OS. Despite the therapeutic impasse these lesions often impose, we found them to be a favorable prognostic factor, even after correcting for predisposing factors. This finding could elude that post-radiation lesions represent an effective treatment of diffuse glioma. Future prospective studies with larger sample sizes should validate this finding.