Abstract
INTRODUCTION: There is currently no evidence on the effectiveness of routine interval scanning compared to symptomatic scanning, of glioblastoma multiforme (GBM) patients, in detecting disease progression after completion of primary treatment. This issue has now been highlighted by the Neuro-Oncology Clinical Trials UK Research Network as a priority for clinical research. METHODS: We looked at all GBM patients treated at our unit from 2004 to 2012 (N=340), and analysed those who were scanned as part of the routine interval surveillance after completion of primary treatment (Stupp protocol) or those who were scanned due to symptoms, and correlated these with diagnosis of disease progression. RESULTS: Of the 340 GBM patients, 86 (25.3%) patients were commenced on adjuvant temozolomide (after surgery/biopsy and concurrent chemoradiotherapy) and only 38 (44.2%) patients completed all 6 cycles of adjuvant temozolomide. These patients then had further scans as part of the routine interval surveillance (145 scans), or if they became symptomatic (12 scans). More symptomatic scans compared to routine interval scans detected disease progression (58.3% vs 22.7%; P= 0.018). Median time to death after a progression scan was 5.7 ± 2.6 months. No patient had disease progression after a 5th stable routine interval surveillance scan. CONCLUSION: Routine interval surveillance imaging after completion of primary treatment may identify disease progression but whether this leads to better survival remains unproven. Frequent scans are costly, and “scanxiety” may affect quality of life; whether interval scanning is justified is up for debate, given that symptomatic scanning is more predictive of disease progression. The optimum frequency of interval scanning has also not been studied. More research using a larger dataset that could enable mathematical modelling is needed to clarify this important issue.