Assessment of the hypervascularized fraction of glioblastomas using a volume analysis of dynamic susceptibility contrast-enhanced MRI may help to identify pseudoprogression

PLoS One. 2022 Oct 13;17(10):e0270216. doi: 10.1371/journal.pone.0270216. eCollection 2022.

Abstract

Purpose: Although perfusion magnetic resonance imaging (MRI) is widely used to identify pseudoprogression, this advanced technique lacks clinical reliability. Our aim was to develop a parameter assessing the hypervascularized fraction of glioblastomas based on volume analysis of dynamic susceptibility contrast-enhanced MRI and evaluate its performance in the diagnosis of pseudoprogression.

Methods: Patients with primary glioblastoma showing lesion progression on the first follow-up MRI after chemoradiotherapy were enrolled retrospectively. On both initial and first follow-up MRIs, the leakage-corrected cerebral blood volume (CBV) maps were post-processed using the conventional hot-spot method and a volume method, after manual segmentation of the contrast-enhanced delineated lesion. The maximum CBV (rCBVmax) was calculated with both methods. Secondly, the threshold of 2 was applied to the CBV values contained in the entire segmented volume, defining our new parameter: %rCBV>2. The probability of pseudoprogression based on rCBVmax and %rCBV>2 was calculated in logistic regression models and diagnostic performance assessed by receiving operator characteristic curves.

Results: Out of 25 patients, 11 (44%) were classified with pseudoprogression and 14 (56%) with true progression based on the Response Assessement in Neuro-Oncology criteria. rCBVmax was lower for pseudoprogression (3.4 vs. 7.6; p = 0.033) on early follow-up MRI. %rCBV>2, was lower for pseudoprogression on both initial (57.5% vs. 71.3%; p = 0.033) and early follow-up MRIs (22.1% vs. 51.8%; p = 0.0006). On early follow-up MRI, %rCBV>2 had the largest area under the curve for the diagnosis of pseudoprogression: 0.909 [0.725-0.986].

Conclusion: The fraction of hypervascularization of glioblastomas as assessed by %rCBV>2 was lower in tumours that subsequently developed pseudoprogression both on the initial and early follow-up MRIs. This fractional parameter may help identify pseudoprogression with greater accuracy than rCBVmax.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Brain Neoplasms* / pathology
  • Contrast Media
  • Disease Progression
  • Glioblastoma* / pathology
  • Humans
  • Magnetic Resonance Imaging / methods
  • Reproducibility of Results
  • Retrospective Studies

Substances

  • Contrast Media

Grants and funding

“Olea innovators” research grant (Olea medical, La Ciotat, France) received by M.R. in September 2018. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.