Influence of corticosteroids on contrast enhancement in patients with untreated glioblastoma

The study by Sanvito et al investigated the effects of corticosteroids on MRI features of glioblastoma, emphasizing how corticosteroid-induced changes can mimic tumor response.¹ The researchers analyzed serial MRI scans from patients with untreated glioblastoma, focusing on volumetric and diffusion metrics before and after corticosteroid dose alterations. They found that an increase in corticosteroid dosage led to a significant reduction in contrast-enhancing tumor volume, averaging around 44%, and caused a notable decrease in the apparent diffusion coefficient (ADC) by 14% on average within the tumor. These changes reflect the capacity of corticosteroids to restore the integrity of the blood-brain barrier, reducing vascular permeability and vasogenic edema, which results in tumor shrinkage and decreased contrast enhancement on MR imaging. This phenomenon is consistent with a pseudoresponse, as the apparent tumor regression is driven by corticosteroid effects rather than actual antineoplastic effects.

Additionally, study results suggest that corticosteroids induce a reduction in blood-brain barrier permeability by inhibiting the release of vascular mediators such as VEGF. Diffusion MRI findings support this mechanism, showing decreased ADC values, most probably linked to reduced vasogenic edema and vascular leakage.

The findings underscore the importance of accounting for corticosteroid-induced imaging alterations to improve the accuracy of glioblastoma response assessment and refine imaging criteria such as RANO. This understanding can help clinicians to look more closely at pseudoresponse, thereby improving patient management and the evaluation of treatment effects.

Moreover, the observed extent of contrast-enhancing tumor shrinkage following corticosteroids aligns with historical studies in patients with recurrent malignant glioma^2,3 and provides novel and contemporary data in patients with newly diagnosed, untreated glioblastoma.

Nevertheless, the present study has several limitations that warrant consideration. One key shortcoming is its observational and retrospective nature, which limits the ability to establish causality and may introduce selection bias. The sample size, particularly for certain subgroup analyses such as patients undergoing corticosteroid dose reduction, was relatively small, restricting the power to detect subtle effects and to generalize findings broadly. Additionally, the study focused on untreated glioblastomas at diagnosis, which means the results may not fully apply to the recurrent setting or those previously exposed to corticosteroids, limiting clinical applicability across different disease stages. The accuracy of corticosteroid dosing data depended on patient compliance and retrospective chart reviews, which might not precisely reflect actual medication intake or the exact timing relative to MRI scans, potentially affecting corticosteroid effects on MRI metrics. Moreover, technical variability in MRI protocols and measurements across patients could influence the quantification of tumor volume and ADC changes, leading to heterogeneity in results. Finally, the study did not evaluate how these imaging changes correlate with clinical outcomes or neurological function, leaving uncertainty about the direct clinical impact of corticosteroid-induced radiographic alterations.

It should be pointed out that the authors acknowledged the above-mentioned limitations, such as MRI protocol heterogeneity, the observational design, small sample size, and challenges in precise corticosteroid dosing timing, explicitly in their discussion. They mitigate some issues by adjusting for baseline tumor volume, which yielded similar results, and note that focusing on untreated glioblastomas provides a clearer assessment of corticosteroid effects. They stress that their findings are preliminary and emphasize the need for future prospective studies with larger cohorts to validate these results and explore clinical correlations. Despite limitations, their study offers important benchmarks for understanding corticosteroid-induced imaging changes.

To overcome the issue of dexamethasone-induced alterations on the contrast-enhancing MRI volume, another imaging method would be desirable. For example, PET with radiolabeled amino acids, e.g., O-(2-[¹⁸F]-fluoroethyl)-L-tyrosine (FET), represents one neuroimaging approach for the assessment of treatment response that is largely independent of the effects of dexamethasone.^4,5 An experimental study evaluated the effect of dexamethasone on blood-brain barrier permeability on FET uptake in rat glioma and human xenograft models.⁶ Although dexamethasone markedly reduced blood-brain barrier permeability, FET uptake was largely unaffected. These results indicate that FET uptake in gliomas is essentially independent of dexamethasone-induced changes in blood-brain barrier integrity.

The present dataset, comprising serial MRI scans from patients with untreated glioblastoma, is unique and has enabled quantitative assessments such as corticosteroid-induced volumetric shrinkage of contrast enhancement and reductions in the ADC on diffusion-weighted MRI following corticosteroid administration. These findings highlight the important clinical observation that corticosteroids can induce a significant pseudoresponse in glioblastoma. Prospective studies are warranted to further investigate and elucidate this phenomenon.

Conflict of interest statement. N.G.: Honoraria for lectures from Blue Earth Diagnostics, for advisory board participation from Telix Pharmaceuticals and Servier, and for consultancy services from Telix Pharmaceuticals.

Funding

None declared.