[11C]PK11195 is a peripheral benzodiazepine receptor (PBR) ligand. PBR expression is enhanced in the presence of increased glial activity, thus PK11195 binding is a marker of glial activation. PK11195 is a very difficult tracer to quantify because there is no reference region that is devoid of PBR binding. In addition, PK11195 suffers from poor brain penetration and very high non-specific binding. The most common method used for quantification of PK11195 brain uptake is cluster analysis which was developed originally by the Hammersmith group. Blood input compartmental modeling has been used, as well as white matter normalization. The latest method developed by Federico Turkheimer has been termed a "Supervised" clustering method. Here, we discuss the rationale, advantages, drawbacks, techniques, and specific implementation of each method.
Rationale: Compartmental modeling with a blood input function is always considered the gold standard method of quantitation.
Advantages: The most thorough method, and generally, the most irrefutable technique.
Drawbacks: Arterial blood sampling is required; both the experiment and analysis are relatively difficult.
Techniques: PK11195 modeling was developed and validated by Marc Kropholler. The key to applying the method is in assuming that all subjects have the same non-specific binding. That is the value of K1/k2 is held fixed to the same value for all subjects, then the parameters R1, k2, BP2, are determined.
Normalization to white matter
Rationale: PK11195 uptake in white matter achieves nearly a plateau, which provides a stable level activity that can be used as a reference for evaluating the binding level in other regions.
Advantages: Extremely simple method.
Drawbacks: The least accurate method. It is prone to bias and is less sensitive than the other methods.
Techniques: Generate TACs, compute activity levels, take ratio relative to white matter
Normalization to gray matter
The uptake per unit gray matter was used for normalization in a study of MS subjects.
Drawing gray matter ROI's to define reference TAC
Rationale: By sampling different gray matter regions, the region with lowest DV can be determined, and that region can be taken as the reference region.
Advantages: More simple technically than cluster analysis.
Drawbacks: Even the lowest DV region will still have some binding, and it the result will depend in part on how the regions are selected.
Techniques: Sample ROI's, estimate relative DV, use TAC from low DV region for reference-tissue modeling.
Cluster analysis (to get reference region)
Rationale: There is no anatomic reference region, thus one cannot simply draw an ROI in a specific brain region to get a reference TAC. To derive a pseudo-reference curve, we note that PK11195 binding in control subjects represents a minimum or baselevel of glial activity/PBR binding. Thus if we can characterize the PK11195 uptake in control subjects we can assume that a reference TAC in patients would have a similar shape. By applying cluster analysis to patient studies, several TACs can be extracted then their shapes can be compared against the control curve. The cluster TAC that appears most consistent with the control TAC may be used as a reference curve for the patient study.
Advantages: Blood sampling not needed.
Drawbacks: Method is not as deterministic as we would like.
Techniques: Generate cluster reference TAC, then apply desired reference tissue method.
SuperPK (Supervised clustering)
Rationale: Kinetic classes of PK11195 uptake can be characterized, and based on those kinetic curves, the PK11195 uptake can be determined.