Converting blood TAC to plasma TAC
Most PET tracers equilibrate relatively slowly between plasma and red blood cells (RBCs). In these cases only the radiotracer in plasma is available to tissue extraction during the blood passage through the capillary. Thus, for most radiotracers, the blood TAC measured using “blood pump” with on-line detector, or extracted from the PET image, must be converted to plasma TAC before it can be used as input function in quantitative analysis. In addition to RBCs, blood also contains white blood cells (WBCs) and platelets, but they constitute normally <1% of the blood volume.
Plasma and blood TACs (PTACs and BTACs, respectively) are usually not similar; even if the equilibrium between blood cells and plasma is reached instantly, the water space is smaller inside the cells than in the plasma, or binding to plasma proteins or blood cell membrane lipids or haemoglobin (Paixao et al, 2009) affects the distribution of the tracer. Partition of a drug between plasma and RBC has even been utilized in measurement of the plasma protein binding.
Certain analysis methods require that the impact of vascular radioactivity must be considered. If only plasma TAC is measured, then the blood TAC must be calculated from it, e.g. using p2blood. Note that hematocrit in small vessels in tissue is usually lower than in large venous vessels, from where the sample for hematocrit determination is collected.
Radioactivity concentration in red blood cells cannot be measured directly from centrifuged blood samples, because RBC preparations always contain some plasma. If plasma is washed away, then part of the radioactivity inside the cells is also removed, and certainly the part that is absorbed on the blood cell membrane proteins and lipids.
However, if the TACs of whole blood and plasma, and hematocrit (HCT) are measured, then the TAC of RBC can be calculated from Eq (1) in Excel or using program b2rbc. Blood-to-plasma or plasma-to-blood ratio curve can be converted to RBC-to-plasma with bpr2cpr.
Rapid substrate transport between plasma and RBCs cannot be studied by separating plasma and blood cells from blood sample, but other methods, such as NMR, are needed (Harvey & Ellory, 1989).
When a new radiotracer is introduced, hematocrit and radioactivity concentrations in both plasma and blood have to be measured during the length of the PET imaging.
Based on the plasma-to-blood ratio curve a decision can be made whether a conversion can be applied, e.g. based on hematocrit or measured one-point ratio, or whether both plasma and blood samples need to be individually measured also in future.
The rate of partitioning of the parent tracer between plasma and RBC determines whether PTAC or BTAC should be used as reference fluid in analysis; this does not only apply to PET studies but to all drug studies (Hinderling, 1997).
Note also that partitioning may be different in different animal species, due to different properties of RBC and plasma proteins.
The kinetics of RBC transport can be studied in vitro by incubating blood samples at +37 °C with radiotracer for different times before separating plasma. The procedure should be repeated for the main metabolites (Nahmias et al., 2000).
Tracer distribution between plasma and blood cells
Blood cell to plasma partition coefficient (rRBC/P) can be measured In vitro in +37 °C by adding a known amount of radioactivity into a known volume of fresh blood (providing CB), centrifuging the sample, and sampling a known volume of plasma. Radioactivity of the plasma sample is measured to calculate CP. Blood cell to plasma ratio is then calculated as (Bower, 1982):
There are four main types of tracer distribution between plasma and RBC:
- Tracer persists in plasma
- Tracer persists in red blood cells
- Tracer penetrates RBC membrane instantly
- Tracer penetrates RBC membrane slowly
If PET tracer, or any radioactive metabolite in plasma, can not penetrate the red blood cell membrane, or the concentration inside RBCs is negligible because of high plasma protein binding, then CRBC=0. The concentrations in plasma and blood are related by the equation
Thus, plasma-to-blood ratio is:
, and hematocrit could even be calculated from the plasma-to-blood ratio:
Figure. Plasma-to-blood ratio is only dependent on hematocrit, if tracer, and possible label-carrying metabolites, both persist in plasma.
Radioactive metabolite penetrates RBC membrane
Although the authentic radioligand does not penetrate RBC membrane, it is possible that its radioactive metabolites do.
For example, Unchanged [11C]-R-PK11195 persists in plasma, but its radioactive metabolites equilibrate between RBC and plasma water spaces. In the case of [11C]palmitate, the unchanged tracer and most of its radioactive metabolites stay in plasma, but the first appearing metabolite, [11C]CO2 (or [11C]HCO3-) permeates the RBCs readily.
Figure. Functions representing population average of plasma-to-blood ratio in [11C]-R-PK11195 and [11C]palmitate studies (unpublished data from TPC). For [11C]-R-PK11195 see also similar plot by Kropholler et al. (2009), Fig 2a.
[18F]Fluoride is used as a PET tracer, but it is also a common metabolite of many [18F]-labelled radioligands, including [18F]FMPEP-d2. Transport of [18F]F- between plasma and RBCs is very fast, but concentration inside RBCs is lower than in plasma.
Figures. Plasma-to-blood ratio in [18F]NaF (Hawkins et al., 1992), and in [18F]FMPEP-d2 (unpublished data from TPC) studies.
These tracers include [15O]O2 and [15O]CO. The plasma-to-blood ratio of authentic tracers is then zero.
Note that the concentration of metabolite of [15O]O2, [15O]H2O, is in equilibrium between blood and plasma water spaces.
The transport rate of a molecule across the red blood cell membrane is correlated with its lipophilicity (Schanker et al., 1961).
Figure. Plasma-to-blood ratio is only dependent on hematocrit, if tracer, and possible label-carrying metabolites, are distributed instantly and equally in the water spaces of plasma and red blood cells.
Even if the tracer can penetrate RBC membrane quickly, the concentration inside RBCs may still be negligible because of high plasma protein binding. In that case follow the case for tracers that persist in plasma.
Note that some radioactive metabolites may persist in the plasma, or are trapped in blood cells, and that the ratio at equilibrium may be very different from 1. In these cases, the blood-to-plasma ratio may have to be measured at several time points (for example [11C]L-deprenyl-D2), fitted to an appropriate function (for example a sigmoidal function), and thereafter blood-to-plasma conversion is done using the individual or population average function. [11C]MDL 100,907 is rapidly transported into RBC, but it additionally is bound to 5-HT2ARs on platelets, leading to plasma-to-blood ratio less than 1, but the ratio later increases due to metabolites that do not enter RBCs (Ito et al., 1998; Hinz et al., 2007). Platelets express also SERT, which may decrease the plasma-to-blood ratio of many radiotracers targeting serotonin and dopamine systems.
[11C]diprenorphine plasma-to-blood ratio is initially close to 1, and increases then slowly in linear fashion (Jones et al., 1994). [11C]Nomifensine is rapidly equilibrated between plasma and erythrocytes, but the metabolites (glucuronides) do not enter RBCs, and therefore blood-to-plasma ratio can be used to estimate the fraction of non-metabolized and not protein bound radiotracer in plasma (Salmon et al., 1990).
[11C]Metomidate is an example of tracer which, along with its radioactive metabolites, stays in the blood water spaces; the blood-to-plasma ratio is only dependent on blood hematocrit and water contents of blood cells (63%) and plasma (94%).
[11C]Carfentanil and other fentanyl analogs penetrate the RBC membrane easily. Its plasma-to-blood ratio is dependent on the relative binding to plasma proteins and proteins in RBCs. The radioactive metabolites of [11C]carfentanil seem to be also found inside the red blood cells, with less affinity to the plasma proteins:
Figure. “Hill type” functions representing population average of plasma-to-blood and RBC-to-plasma ratio in [11C]carfentanil studies (unpublished data from TPC).
[11C]PBR28 and its radioactive metabolites equilibrate between RBC and plasma water spaces, but in addition the parent tracer binds specifically to its target receptor TSPO in blood cells and plasma proteins. Therefore, the blood-to-plasma ratio is not only time-dependent, but also different in the high-, mixed-, and low affinity binding (HAB, MAB, and LAB, respectively) subpopulations.
Figure. “Hill type” functions representing subpopulation averages of blood-to-plasma and plasma-to-blood ratio in [11C]PBR28 studies (unpublished data from TPC).
These PET tracers include [18F]FDOPA, [11C]DOPA, [18F]FBPA, [11C]MeAIB, and [11C]methionine, and probably also most other amino acid tracers. A rapidly exchanging component is can be present because of substances adsorbed on blood cell membranes; for instance amino acids are known to be adsorbed on RBC membranes (Picó et al., 1991). Equilibration of 6-[18F]fluoro-L-m-tyrosine is slow, but its main metabolite is equilibrated within 5 min (Nahmias et al., 2000); plasma and blood curves and parent tracer fractions can be described with a compartmental model (Asselin et al., 2002). [18F]fluoromethylcholine and its major metabolite [18F]fluorobetaine both seem to penetrate RBC membrane, and at least the unchanged tracer slowly in mice (Slaets et al., 2013).
Distribution of glucose between plasma and blood cells differs between animal species (Andreen-Svedberg, 1933), and is very fast in humans (Lowe & Walmsley, 1986). In humans and pigs glucose tracer [18F]FDG equilibrates between RBC and plasma water spaces rapidly (less than one minute), but continues to be slowly accumulated in the RBCs; in rats and some other animals the transport may be slower (Gjedde, 1983; Buxton, 1999; Weber et al., 2002; Wu et al., 2007; Huang et al., 2017). Accurate analysis in rodents may require that RBC uptake of [18F]FDG is included in the model (Alf et al., 2013).
Figure. Functions representing population average of plasma-to-blood ratio in [18F]FDOPA and [18F]FBPA studies (unpublished data from TPC).
Also with these tracers, plasma-to-blood ratio often starts to increase during the PET scan. This is indicative of the appearance of labelled metabolites which do not penetrate the RBC membrane.
Software for converting blood to plasma and plasma to blood
- [11C]Metomidate ([11C]MTO)
Notice that (arterial) blood TAC calculated this way does not represent the average TAC of blood in tissue, which consists of arterial and venous blood in unknown proportion, and also the local venous blood TAC is unknown.
- Plotting curves
- Fick’s principle
- Erythrocytes in PET studies
- Plasma protein binding
- Blood sampling
- Processing plasma data
- Processing ABSS data
Berezhkovskiy LM, Zhang X, Cheong J. A convenient method to measure blood-plasma concentration ratio using routine plasma collection in in vivo pharmacokinetic studies. J Pharm Sci. 2011; 100(12): 5293-5298. doi: 10.1002/jps.22709.
Buxton DB. Glucose permeability in nonprimate erythrocytes. J Nucl Med. 1999;40(12):2125-2126.
Gjedde A. Modulation of substrate transport to the brain. Acta Neurol Scand. 1983; 67: 3-25. doi: 10.1111/j.1600-0404.1983.tb04541.x.
Hinderling PH. Red blood cells: a neglected compartment in pharmacokinetics and pharmacokinetics. Pharmacol Rev. 1997; 49(3): 279-295.
Hinz R, Bhagwagar Z, Cowen PJ, Cunningham VJ, Grasby PM. Validation of a tracer kinetic model for the quantification of 5-HT2A receptors in human brain with [11C]MDL 100,907. J Cereb Blood Flow Metab. 2007; 27: 161-172. doi: 10.1038/sj.jcbfm.9600323.
Koeppe RA. Quantitative functional imaging using positron computed tomography and rapid parameter estimation techniques. Thesis (Ph.D.), The University of Wisconsin, Madison, 1984.
Lee J-S, Su K-H, Lin J-C, Chuang Y-T, Chueh H-S, Liu R-S, Wang S-J, Chen J-C. A novel blood-cell-two-compartment model for transferring a whole blood time activity curve to plasma in rodents. Comput Methods Programs Biomed. 2008; 92(3): 299-304. doi: 10.1016/j.cmpb.2008.02.006.
Lubberink M, Boellaard R, Greuter HNJM, Lammertsma AA. Effect of uncertainty in plasma metabolite levels on kinetic analysis of [11C]flumazenil and [11C](R)-PK11195 PET studies. Neuroimage 2004; 22: T119-T120.
Nahmias C, Wahl LM, Amano S, Asselin M-C, Chirakal R. Equilibration of 6-[18F]fluoro-L-m-tyrosine between plasma and erythrocytes. J Nucl Med. 2000; 41: 1636-1641. PMID: 11037992.
Paixão P, Gouveia LF, Morais JAG. Prediction of drug distribution within blood. Eur J Pharm Sci. 2009; 36: 544-554. doi: 10.1016/j.ejps.2008.12.011.
Weber B, Burger C, Biro P, Buck A. A femoral arteriovenous shunt facilitates arterial whole blood sampling in animals. Eur J Nucl Med. 2002; 29: 319-323. doi: 10.1007/s00259-001-0712-2.
Updated at: 2019-01-21
Written by: Vesa Oikonen, Tuula Tolvanen, Kaisa Liukko, Pauliina Luoto, Anne Roivainen