Influence of MRI Follow-Up on Treatment Decisions during Standard Concomitant and Adjuvant Chemotherapy in Patients with Glioblastoma: Is Less More?

SIMPLE SUMMARY: Glioblastomas are brain tumors with a poor prognosis, and early tumor progression occurs often. Therefore, patients are closely monitored with regular MRI scans, usually at 2–3 month intervals. However, there is no evidence for this strategy, and it is not known if patients benefit from this approach. Furthermore, effects from the treatment sometimes mimic tumor progression (pseudoprogression). Pseudoprogession can cause uncertainty and makes decision making about continuing or stopping treatment difficult. This study evaluated how often standard scheduled MRI scans influenced treatment decisions and how often MRI scans caused uncertainty. Standard scheduled follow-up MRI scans rarely led to treatment consequences (<10%). However, many MRI scans caused diagnostic uncertainty (>25%). When scans were made at unscheduled timepoints, e.g., in patients with new or worsening symptoms, they had more consequences. Our results do not support the current pragmatic follow-up strategy and suggest a more tailored follow-up approach for glioblastoma patients. ABSTRACT: MRI is the gold standard for treatment response assessments for glioblastoma. However, there is no consensus regarding the optimal interval for MRI follow-up during standard treatment. Moreover, a reliable assessment of treatment response is hindered by the occurrence of pseudoprogression. It is unknown if a radiological follow-up strategy at 2–3 month intervals actually benefits patients and how it influences clinical decision making about the continuation or discontinuation of treatment. This study assessed the consequences of scheduled follow-up scans post-chemoradiotherapy (post-CCRT), after three cycles of adjuvant chemotherapy [TMZ3/6], and after the completion of treatment [TMZ6/6]), and of unscheduled scans on treatment decisions during standard concomitant and adjuvant treatment in glioblastoma patients. Additionally, we evaluated how often follow-up scans resulted in diagnostic uncertainty (tumor progression versus pseudoprogression), and whether perfusion MRI improved clinical decision making. Scheduled follow-up scans during standard treatment in glioblastoma patients rarely resulted in an early termination of treatment (2.3% post-CCRT, 3.2% TMZ3/6, and 7.8% TMZ6/6), but introduced diagnostic uncertainty in 27.7% of cases. Unscheduled scans resulted in more major treatment consequences (30%; p < 0.001). Perfusion MRI caused less diagnostic uncertainty (p = 0.021) but did not influence treatment consequences (p = 0.871). This study does not support the current pragmatic follow-up strategy and suggests a more tailored follow-up approach.