Arterial spin labeling MR imaging for differentiation between high- and low-grade glioma - A meta-analysis

Forskningsoutput: TidskriftsbidragArtikel i vetenskaplig tidskrift

Abstract

Background. Arterial spin labeling is an MR imaging technique that measures cerebral blood flow (CBF) noninvasively. The aim of the study is to assess the diagnostic performance of arterial spin labeling (ASL) MR imaging for differentiation between high-grade glioma and low-grade glioma. Methods. Cochrane Library, Embase, Medline, and Web of Science Core Collection were searched. Study selection ended November 2017. This study was prospectively registered in PROSPERO (CRD42017080885). Two authors screened all titles and abstracts for possible inclusion. Data were extracted independently by 2 authors. Bivariate random effects meta-analysis was used to describe summary receiver operating characteristics. Trial sequential analysis (TSA) was performed. Results. In total, 15 studies with 505 patients were included. The diagnostic performance of ASL CBF for glioma grading was 0.90 with summary sensitivity 0.89 (0.79-0.90) and specificity 0.80 (0.72-0.89). The diagnostic performance was similar between pulsed ASL (AUC 0.90) with a sensitivity 0.85 (0.71-0.91) and specificity 0.83 (0.69- 0.92) and pseudocontinuous ASL (AUC 0.88) with a sensitivity 0.86 (0.79-0.91) and specificity 0.80 (0.65-0.87). In astrocytomas, the diagnostic performance was 0.89 with sensitivity 0.86 (0.79 to 0.91) and specificity 0.79 (0.63 to 0.89). Sensitivity analysis confirmed the robustness of the findings. TSA revealed that the meta-analysis was adequately powered. Conclusion. Arterial spin labeling MR imaging had an excellent diagnostic accuracy for differentiation between high-grade and low-grade glioma. Given its low cost, non-invasiveness, and efficacy, ASL MR imaging should be considered for implementation in the routine workup of patients with glioma.

Detaljer

Författare
Enheter & grupper
Externa organisationer
  • University of Chieti-Pescara
  • Skåne University Hospital
  • Uppsala universitet, Historiska institutionen
  • Karolinska Institute
Forskningsområden

Ämnesklassifikation (UKÄ) – OBLIGATORISK

  • Radiologi och bildbehandling
  • Cancer och onkologi

Nyckelord

Originalspråkengelska
Sidor (från-till)1450-1461
Antal sidor12
TidskriftNeuro-Oncology
Volym20
Utgåva nummer11
StatusPublished - 2018
PublikationskategoriForskning
Peer review utfördJa