Positron emission tomography to image cerebral neuroinflammation in ischaemic stroke: a pilot study

Research output: Contribution to journalArticlepeer-review

  • External authors:
  • Eszter Visi
  • Martin Punter
  • Arshad Majid
  • Alexander Gerhard


Activated microglia play a complex role in neuroinflammation associated with acute ischaemic stroke. As a potential target for anti-inflammatory therapy, it is crucial to understand the association between intensity, extent and the clinical outcome of a stroke. The 18-kDa translocator protein is a marker of cerebral microglial activation and of macrophage infiltration after damage to the brain. It can be imaged by positron emission tomography. Therefore, the recently developed radiopharmaceutical [18F]-GE180 was used in patients after a mild to moderate stroke and compared with [11C]-(R)-PK11195, which has already been established in research but cannot be used in routine clinical settings because of its very short half-life.

Objectives for phase 1 were to evaluate the tolerability of positron emission tomography scanning, to assess the technical feasibility of imaging the 18-kDa translocator protein using [18F]-GE180 as radiopharmaceutical, to compare [18F]-GE180 with [11C]-(R)-PK11195 as reference. Objectives for phase 2 were examining the relation of positron emission tomography imaging with clinical outcome, magnetic resonance imaging and systemic inflammation. However, the study was ended after phase 1 because of the results obtained in that phase and did not enter phase 2.

Ten participants (aged 24–89 years, median 68 years) (eight male and two female) with a history of recent ischaemic stroke of mild to moderate severity (modified Rankin scale score of 2–3) in the middle cerebral artery territory were scanned 18 to 63 days (median 34.5 days) after the stroke by magnetic resonance imaging (Philips 1.5 T; Philips, Amsterdam, the Netherlands), [18F]-GE180 (200 MBq, 30-minute dynamic scan) and [11C]-(R)-PK11195 (740 MBq, 60-minute dynamic scan) positron emission tomography (Siemens HRRT; Siemens, Munich, Germany). The two positron emission tomography scans were performed on 2 separate days (mean 3.4 days apart). Five patients were randomised to receive the [18F]-GE180 scan at the first session and five patients were randomised to receive it at the second session. Participants were genotyped for the rs6971 18-kDa translocator protein polymorphism, which is known to affect binding of [18F]-GE180 but not of [11C]-(R)-PK11195. All positron emission tomography and magnetic resonance data sets were co-registered with T1-weighted magnetic resonance image scans. Binding of [18F]-GE180 was compared with [11C]-(R)-PK11195 for the infarct and contralateral reference regions. Spearman’s rank-order correlation was used to compare tracers, t-tests to compare patient subgroups.

Tolerability of scans was rated as 4.36 (range 4–5) out of a maximum of 5 by participants, and there were no serious adverse events. There was a close correlation between [18F]-GE180 and [11C]-(R)-PK11195 (r = 0.79 to 0.84). The 18-kDa translocator protein polymorphism had a significant impact on the uptake of [18F]-GE180, which was very low in normal cortex. Ischaemic lesions with contrast enhancement on magnetic resonance as an indicator of blood–brain barrier damage showed a significantly higher uptake of [18F]-GE180 than the lesions without enhancement, even in low-affinity binders.

[18F]-GE180 was safe and well tolerated. However, strong dependency of uptake on blood–brain barrier damage and a genetic 18-kDa translocator protein polymorphism, as well as a high contribution of vascular signal to the uptake and evidence of non-specific binding in ischaemic lesions with blood–brain barrier damage, limits the clinical applicability of [18F]-GE180 as a diagnostic marker of neuroinflammation.

As the study was ended after phase 1, this was only a small pilot trial. Further studies are warranted to fully understand the influence of blood–brain barrier damage on positron emission tomography microglia imaging.

Bibliographical metadata

Original languageEnglish
Pages (from-to)1-26
JournalEfficacy and Mechanism Evaluation
Issue number1
Publication statusPublished - 1 Feb 2020