After successful reperfusion of acute ischaemic stroke, the salvaged penumbra is not necessarily intact: PET with 11C-flumazenil (FMZ, a marker of cortical GABA-receptor-bearing neurons) shows selective neuronal loss (SNL), and PET with 11C-PK11195 (a marker of the translocator protein TSPO overexpressed by activated microglia) shows neuroinflammation. Morris and colleagues (Brain, 2018) demonstrated in reperfused human stroke patients that SNL affects the surviving penumbra, that microglial activation is spatially dissociated from it, and that the two processes show no quantitative coupling — but their measurements were cross-sectional and left the therapeutic value of penumbral SNL unresolved. The problem left open is whether preventing penumbral SNL translates into improved functional outcome in humans, and whether any study beyond the cross-sectional work of Morris et al. has paired FMZ-PET and TSPO-PET in the same stroke patients.

Background

In acute ischaemic stroke, a region of hypoperfused but still viable tissue — the ischaemic penumbra — surrounds the infarct core and can be salvaged by intravenous thrombolysis and mechanical thrombectomy (endovascular removal of the occluding clot). Even with successful reperfusion, fewer than half of treated patients regain independence, motivating the search for complementary therapeutic targets. Two processes documented in rodent models affect the salvaged penumbra itself. First, selective neuronal loss (SNL): death of scattered neurons in tissue that never undergoes infarction and therefore appears normal on structural MRI. SNL is invisible to conventional imaging but can be mapped in vivo by positron emission tomography (PET) with 11C-flumazenil (FMZ), a radioligand that binds the central benzodiazepine site of GABAA_A receptors on cortical neurons, so reduced FMZ binding signals loss of neurons in MRI-normal tissue. Second, microglial activation: the brain's resident immune cells respond to injury and, depending on phenotype, can either exacerbate neuronal death or support repair. Activated microglia (and infiltrating macrophages) overexpress the 18 kDa translocator protein (TSPO), which can be imaged with PET radioligands such as 11C-PK11195; increased binding therefore marks neuroinflammation, though TSPO is not microglia-specific and does not report activation phenotype.

Human evidence accumulated in steps. Guadagno et al. (2008) used follow-up FMZ-PET after acute stroke with early recovery and showed FMZ binding loss in non-infarcted middle cerebral artery (MCA) territory that was proportional to the initial hypoperfusion, consistent with SNL in the rescued penumbra. Longitudinal TSPO-PET studies (Thiel et al. 2010, using serial 11C-PK11195 with diffusion tensor imaging; Gulyás et al. 2012, using 11C-vinpocetine) charted the time course of post-stroke microglial activation — local activation in and around the infarct that largely subsides, and remote activation along degenerating fibre tracts that persists — but none of these studies measured neuronal integrity. The pivotal human study pairing the two modalities is Morris et al. (2018, Brain), which recruited 16 acute MCA-stroke patients with admission CT-perfusion-documented penumbra and early neurological recovery, and acquired both FMZ-PET and 11C-PK11195 PET at follow-up, with final infarcts mapped by MRI and a dilated peri-infarct rim excluded to control partial-volume spill-over of tracer signal. They found: (i) significant FMZ binding reduction in the non-infarcted ipsilateral MCA territory and specifically within the CT-defined surviving penumbra, establishing penumbral SNL in man; (ii) increased PK11195 binding in the non-infarcted territory that was concentrated in non-penumbral tissue, with no significant increase within the penumbra itself; and (iii) no significant correlation between the FMZ and PK11195 measures in any compartment. The authors concluded that microglial activation may not contribute to penumbral neuronal loss in man — perhaps merely reflecting secondary remote degeneration — and that penumbral SNL could instead represent a novel, adjunctive therapeutic target, with FMZ-PET as a potential biomarker to monitor treatment. The measurements were, however, cross-sectional (11C-PK11195 PET acquired in the subacute window, day 13–28, and FMZ-PET in the early-chronic window, 6–21 weeks post-stroke, with no serial follow-up of either process in the same patient), confined to cortical GABAA_A-bearing neuronal populations, blind to inflammatory components other than TSPO-positive microglia/macrophages, and obtained in a selected reperfusion-candidate population. Approaches to preventing penumbral SNL that have been discussed in this literature include acutely improving penumbral perfusion and oxygenation as an adjunct to thrombectomy and preventing capillary no-reflow (Ter Schiphorst et al. 2021; Baron 2021).

Problem Statement

Following Morris et al. (2018), the problem is to establish the human in vivo pathophysiology and therapeutic significance of selective neuronal loss (SNL) and microglial activation in the salvaged ischaemic penumbra, with one overarching objective: determining whether penumbral SNL is a therapeutically actionable lesion — actionable in the sense that an intervention that reduces penumbral SNL produces measurable functional benefit in humans. The following questions are open:

  1. Therapeutic translation. Assess, from the 2018–2026 human and preclinical evidence, whether penumbral SNL is a therapeutically actionable target — whether preventing penumbral SNL would translate into measurable functional benefit in humans.

  2. Microglia-targeted strategies. Establish, from the same evidence, whether microglia-targeted strategies for penumbral salvage are justified or refuted.

  3. Measurement specification. Identify and justify the biomarker and the functional outcome endpoint that would settle the question in interventional trials.

  4. State of the field. Establish, from the 2018–2026 literature, whether any study beyond the cross-sectional study of Morris et al. has paired FMZ-PET and TSPO-PET in the same human stroke patients; whether longitudinal TSPO studies have been accompanied by a neuronal-integrity tracer; whether serial imaging with benzodiazepine-receptor tracers has been performed alongside a microglia marker rather than only to monitor cell-transplantation plasticity (Kawabori et al. 2026, serial 123I-iomazenil SPECT); and whether any FMZ-PET-monitored interventional trial exists.

The source study's own measurements bound what current evidence can claim about the lesion: FMZ binds selectively to the neuron-specific GABAA_A receptor, so white-matter, striatal and non-GABAergic SNL could not be assessed and the extent/magnitude of SNL may have been underestimated; and PK11195 essentially does not reflect other components of post-stroke inflammation such as astrocytosis and infiltration by blood-borne immune cells, let alone microglial phenotypic states. These are source-stated caveats on the existing evidence, but they mean the true anatomical and cellular scope of penumbral SNL and of the accompanying inflammatory response remains to be characterized, and no validated in vivo marker currently exists for non-GABAergic or white-matter selective neuronal loss.

References

  1. Morris RS, Jones PS, Alawneh JA, Hong YT, Fryer TD, Aigbirhio FI, Warburton EA, Baron JC (2018). Relationships between selective neuronal loss and microglial activation after ischaemic stroke in man. Brain 141(7):2098–2111. doi:10.1093/brain/awy121. https://doi.org/10.1093/brain/awy121
  2. Guadagno JV, Jones PS, Aigbirhio FI, Wang D, Fryer TD, Day DJ, Antoun N, Nimmo-Smith I, Warburton EA, Baron JC (2008). Selective neuronal loss in rescued penumbra relates to initial hypoperfusion. Brain 131(Pt 10):2666–2678. doi:10.1093/brain/awn175. https://doi.org/10.1093/brain/awn175
  3. Thiel A, Radlinska BA, Paquette C, Sidel M, Soucy JP, Schirrmacher R, Minuk J (2010). The temporal dynamics of poststroke neuroinflammation: a longitudinal diffusion tensor imaging–guided PET study with 11C-PK11195 in acute subcortical stroke. Journal of Nuclear Medicine 51(9):1404–1412. doi:10.2967/jnumed.110.076612. https://doi.org/10.2967/jnumed.110.076612
  4. Gulyás B, Tóth M, Schain M, Airaksinen A, Vas A, Kostulas K, Lindström P, Hillert J, Halldin C (2012). Evolution of microglial activation in ischaemic core and peri-infarct regions after stroke: a PET study with the TSPO molecular imaging biomarker [11C]vinpocetine. Journal of the Neurological Sciences 320(1–2):110–117. doi:10.1016/j.jns.2012.06.026. https://doi.org/10.1016/j.jns.2012.06.026
  5. Kim YK, Yang EJ, Cho K, Lim JY, Paik NJ (2014). Functional Recovery After Ischemic Stroke Is Associated With Reduced GABAergic Inhibition in the Cerebral Cortex: A GABA PET Study. Neurorehabilitation and Neural Repair 28(6):576–583. doi:10.1177/1545968313520411. https://doi.org/10.1177/1545968313520411
  6. Schaechter JD, Hightower BG, Kim M, Loggia ML (2021). A pilot [11C]PBR28 PET/MRI study of neuroinflammation and neurodegeneration in chronic stroke patients. Brain, Behavior, & Immunity – Health 17:100336. doi:10.1016/j.bbih.2021.100336. https://doi.org/10.1016/j.bbih.2021.100336
  7. Hughes JL, Beech JS, Jones PS, Wang D, Menon DK, Aigbirhio FI, Fryer TD, Baron JC (2020). Early-stage 11C-Flumazenil PET predicts day-14 selective neuronal loss in a rodent model of transient focal cerebral ischemia. Journal of Cerebral Blood Flow and Metabolism 40(10):1997–2009. doi:10.1177/0271678X19883040. https://doi.org/10.1177/0271678X19883040
  8. D'Anna L, Searle G, Harvey K, Matthews PM, Veltkamp R (2023). Time course of neuroinflammation after human stroke – a pilot study using co-registered PET and MRI. BMC Neurology 23(1):193. doi:10.1186/s12883-023-03178-7. https://doi.org/10.1186/s12883-023-03178-7
  9. Jensen P, Ozenne B, Meden P, Feng L, Thomsen G, Knudsen L, Steglich-Arnholm H, Møller K, Thomsen C, Svarer C, Beliveau V, Mikkelsen J, Knudsen G, Pinborg LH (2025). Poststroke Translocator Protein Expression Dynamics and Correlations to Chronic Infarction: A [123I]-CLINDE-SPECT Study. Journal of Neuroimaging 35(1):e70002. doi:10.1111/jon.70002. https://doi.org/10.1111/jon.70002
  10. Zatcepin A, Kopczak A, Holzgreve A, Hein S, Schindler A, Duering M, Kaiser L, Lindner S, Schidlowski M, Bartenstein P, Albert N, Brendel M, Ziegler SI (2024). Machine learning-based approach reveals essential features for simplified TSPO PET quantification in ischemic stroke patients. Zeitschrift für Medizinische Physik 34(2):218–230. doi:10.1016/j.zemedi.2022.11.008. https://doi.org/10.1016/j.zemedi.2022.11.008
  11. Becker G, Debatisse J, Rivière M, Crola Da Silva C, Beaudoin-Gobert M, Eker O, Wateau O, Cho TH, Wiart M, Tremblay L, Costes N, Mérida I, Redouté J, Léon C, Langlois JB, Le Bars D, Lancelot S, Nighoghossian N, Mechtouff L, Canet-Soulas E (2023). Spatio-Temporal Characterization of Brain Inflammation in a Non-human Primate Stroke Model Mimicking Endovascular Thrombectomy. Neurotherapeutics 20(3):789–802. doi:10.1007/s13311-023-01368-2. https://doi.org/10.1007/s13311-023-01368-2
  12. Wang Z, Song Y, Bai S, Xiang W, Zhou X, Han L, Zhu D, Guan Y (2023). Imaging of microglia in post-stroke inflammation. Nuclear Medicine and Biology 118–119:108336. doi:10.1016/j.nucmedbio.2023.108336. https://doi.org/10.1016/j.nucmedbio.2023.108336
  13. Orhii PD, Haque ME, Fujita M, Selvaraj S (2023). Advances in PET imaging of ischemic stroke. Frontiers in Stroke 1:1093386. doi:10.3389/fstro.2022.1093386. https://doi.org/10.3389/fstro.2022.1093386
  14. Shi K, Tian DC, Li ZG, Ducruet AF, Lawton MT, Shi FD (2019). Global brain inflammation in stroke. The Lancet Neurology 18(11):1058–1066. doi:10.1016/S1474-4422(19)30078-X. https://doi.org/10.1016/S1474-4422(19)30078-X
  15. Baron JC (2021). The core/penumbra model: implications for acute stroke treatment and patient selection in 2021. European Journal of Neurology 28(8):2794–2803. doi:10.1111/ene.14916. https://doi.org/10.1111/ene.14916
  16. Ter Schiphorst A, Charron S, Hassen WB, Provost C, Naggara O, Benzakoun J, Seners P, Turc G, Baron JC, Oppenheim C (2021). Tissue no-reflow despite full recanalization following thrombectomy for anterior circulation stroke with proximal occlusion: a clinical study. Journal of Cerebral Blood Flow and Metabolism 41(2):253–266. doi:10.1177/0271678X20954929. https://doi.org/10.1177/0271678X20954929
  17. Baron JC (2025). Fifty Years of Deciphering Stroke Pathophysiology. Stroke 56(7):1947–1957. doi:10.1161/STROKEAHA.125.051163. https://doi.org/10.1161/STROKEAHA.125.051163
  18. Kawabori M, Hirata K, Shichinohe H, Watanabe S, Miura A, Ito YM, Kudo K, Kuroda S, Houkin K, Fujimura M (2026). Serial brain FDG-PET and IMZ-SPECT following intracerebral MSC transplantation in patients with subacute ischemic stroke. Stem Cell Research & Therapy 17(1):243. doi:10.1186/s13287-026-05048-8. https://doi.org/10.1186/s13287-026-05048-8