A method of studying pharmacokinetics in man at picomolar drug concentrations.

1. We describe a new method that enables the tissue kinetics of picomolar concentrations of drugs to be measured in man. The method is based on the administration of a drug, labelled with a short-lived positron-emitting radioisotope, such as carbon-11 (t1/2 = 20.4 min, beta + = 99.8%) or fluorine-18 (t1/2 = 109.8 min, beta + = 96.9%), which is then detected in vivo by an array of 10 large uncollimated sodium iodide scintillation detectors, arranged as five opposing pairs, with each pair collecting data over one major organ or region of the body. 2. To illustrate the scope of the new method we report the results of administering [O-methyl-11C]-diprenorphine, an established radioligand for central opiate (mu, kappa, and delta) receptors and L-6-[18F]-fluoro-DOPA, a marker for dopaminergic neurons. 3. Only 2-10 muCi (74-370 kBq) of radioactivity are used and, as a consequence of the high specific activities with which carbon-11 and fluorine-18 labelled compounds can be prepared, the method requires less than a nanomole of drug to be administered. In many cases, this amount of drug might be considered low enough to avoid any adverse biological effect. Furthermore repeat studies are possible in many without delivering unacceptable radiation burdens. 4. The high sensitivity realised for both radioactivity and mass suggests a mean for determining the human biodistribution of a new drug at a very early stage in its development. This has potential benefit to drug discovery programmes and to ensuing drug therapies.

Asymmetric synthesis of a precursor for the automated radiosynthesis of S-(3'-t-butylamino-2'-hydroxypropoxy)-benzimidazol-2-[11C]one (S-[11C]CGP 12177) as a preferred radioligand for beta-adrenergic receptors.

S-[1-(2,3-Diaminophenoxy)]-3'-(N-t-butylamino)propan-2'-ol has been synthesized in three steps from 2,3-dinitro-phenol and the chiral auxiliary, S-glycidyl-3-nitrobenzenesulphonate, to provide a precursor for labelling S-(3'-t-butylamino-2'-hydroxypropoxy)-benzimidazol-2-one (S-CGP 12177) with the short-lived positron-emitting radionuclide, carbon-11 (t 1/2 = 20.4 min; beta+ = 99.8%). Reaction of the diamine with [11C]phosgene, itself derived from no-carrier-added cyclotron-produced [11C]methane, provides radiochemically and chemically pure S-[carbonyl-11C]CGP 12177 in greater than 95% enantiomeric excess after HPLC. Automated apparatus is described for safely producing up to 5.9 GBq (160 mCi) of S-[11C]CGP 12177 with high sp. act. (20-40 GBq/mu mol or 0.54-1.08 Ci/mu mol) in a form suitable for human intravenous injection at only 30 min from the end of radionuclide production. S-[11C]CGP 12177 is preferred to the formerly described racemate as a radioligand for the study of beta-adrenergic receptors in vivo by positron emission tomography.

Labelled agents for PET studies of the dopaminergic system--some quality assurance methods, experience and issues.

Practical methods are described for the quality assurance of three labelled agents (L-6-[18F]fluoro-DOPA, S-[N-methyl-11C]nomifensine and [O-methyl-11C]raclopride) now produced regularly for PET studies of the dopaminergic system in man. These include indirect methods for the initial determination of label position (e.g. 13C-NMR spectroscopy) and also direct methods for the assessment of chiral purity (TLC and HPLC) and the routine determination of radiochemical purity, chemical purity and specific activity (HPLC). Mass spectrometry has been used to identify some impurities. L-6-hydroxy-DOPA (a precursor in vivo of the neurotoxin, L-6-hydroxydopamine) has been detected by HPLC in some preparations of L-6-[18F]fluoro-DOPA. Formulated S-[N-methyl-11C]nomifensine has been found to decrease in radiochemical purity with storage, whereas formulated [O-methyl-11C]raclopride has been found to be stable. Some quality assurance issues are discussed in relation to experience in the application of the described methods and the obtained results.

Use of short- and long-lived rubidium tracers for the study of transient ischemia.

Positron emission tomography (PET) with rubidium-82 (82Rb) has been developed to measure regional myocardial perfusion and to detect transient ischemia both in the experimental laboratory and in humans. There are known and separate contaminating effects of the 82Rb signal by disturbances in wall motion, wall thinning, and the partial volume effect that occur during transient ischemia. In nine anesthetized greyhounds, PET with 82Rb (T1/2 = 78 sec) was used to determine the regional myocardial uptake of this cation during a control period that consisted of a mild stenosis of the left anterior descending coronary artery in the absence of ischemia (to limit reactive hyperemia), during 10 min of total occlusion and, finally, at 30 and 60 min of recovery with release of the occlusion but not of the stenosis. Separately, rubidium-81 (81Rb); T1/2 = 4.58 hr) was given as a peripheral intravenous injection 2 hr before the study to allow this long-lived tracer to distribute in the potassium space of the myocardium. Observations during control and ischemia revealed marked decreases in 82Rb uptake (0.84 +/- 0.12 to 0.28 +/- 0.12, p = 0.001) in affected regions and were paralleled by similar decreases in microsphere blood flow (0.88 +/- 0.08 to 0.12 +/- 0.10 ml/min/g, p = 0.003), which gradually recovered by 60 min postischemia. Lesser decreases in 81Rb activity (0.84 +/- 0.11 to 0.76 +/- 0.17, p = 0.83) were observed in the same regions during ischemia, but these were immediately reversible. Separate in vitro postmortem experiments in eight rabbits confirmed a linear relationship between plasma and myocardial activities of stable potassium and 81Rb although there was a greater concentration of 81Rb in the myocardium that in the plasma relative to potassium (y = -3.29 +/- 0.79 x, s.e.e. 1.91, r = 0.95). These studies demonstrate that if 81Rb is given intravenously to distribute into the potassium pool, tomograms of the heart may be recorded to measure the potassium-rich mass of myocardium providing information about the acute effects of wall thinning during ischemia. Rubidium-81 used in this way may be helpful in assessing the effects of wall thinning and/or scar when other tracers are being used to assess perfusion or metabolism.

Experience with a 82Sr/82Rb generator for clinical use.

A 82Sr/82Rb generator system is described which is shown to be suitable for continuous intravenous infusion in man. The breakthrough of the 82Sr parent has been closely monitored and remained less than 18.5 Bq mL-1 of infusate. A method using a 0.05% solution of sodium hypochlorite to disinfect the generator resulted in a sterile and pyrogen free eluate. Recommendations are made for the setting up of the generator to ensure the maintenance of its pharmaceutical integrity.

Quantitative measurement of blood-brain barrier permeability using rubidium-82 and positron emission tomography.

In normal brain, the blood-brain barrier (BBB) is highly impermeable to K+ cations, their transport being controlled by ATPases situated in the endothelial cell membranes. 82Rb+ is a positron-emitting analogue of K+ with a half-life of 75 s. Using a steady-state model and positron emission tomography, quantitative extraction data for 82Rb+ transport across the BBB have been obtained both in normal human subjects and in a variety of conditions of cerebral pathology. A mean cerebral Rb extraction of 2.1% was found for normal subjects, corresponding to a mean value of 1.1 x 10(-6) cm s-1 for 82Rb+ cation permeability across the BBB. No increase in cerebral Rb extraction was observed for patients with diffusely raised intracranial pressure secondary to obstructive hydrocephalus and benign intracranial hypertension, or for patients with multiple sclerosis or cerebral systemic lupus erythematosus. Cerebral tumours that were enhanced on computed tomography scanning showed a significant increase in local Rb uptake. No correlation between tumour size, or grade of glioma, and tumour Rb extraction was found. Nonenhancing tumours showed no increase in local Rb extraction, and regions of perifocal tumour oedema also had Rb extraction values in the normal range. It is concluded that increased Rb extraction occurs only where tight junction integrity in the BBB breaks down locally, that is, in the microcirculation of enhancing tumours but not in that of perifocal regions of tumour oedema or nonenhancing tumours.

In vivo measurement of regional cerebral haematocrit using positron emission tomography.

A method is described for measuring the regional cerebral-to-large vessel haematocrit ratio using inhalation of carbon-11-labelled carbon monoxide and the intravenous injection of carbon-11-labelled methyl-albumin in combination with positron emission tomography. The mean value in a series of nine subjects was 0.69. This is approximately 20% lower than the value of 0.85 previously reported. It is concluded that previous measurements of regional cerebral blood volume using a haematocrit ratio of 0.85 will have underestimated the value of regional cerebral blood volume by 20%.

PAC studies of 111In binding to transferrin, tropolone and acetylacetone in aqueous solutions.

Time integral and time-differential PAC measurements have been made over a wide temperature range in aqueous solutions of [111In]tropolonate and [111In]acetylacetonate. The quadrupole frequency in the latter is approximately 30% higher than that in the former and the molecular volumes derived from rotational correlation times show the expected differences. Apo-transferrin was separately added to the two 111In-chelates and the transfer of activity from chelate to transferrin followed as a function of relative molar concentrations. Very much larger molar ratios of transferrin to tropolone than of transferrin to acetylacetone were required before substantial transfer of 111In from chelate to transferrin took place. This difference in affinity for transferrin could be one significant factor in explaining the enhanced ability of [111In]tropolonate to label blood cells in the presence of plasma. The determination of PAC parameters in [111In]transferrin over a range of temperatures showed that the values of quadrupole frequency obtained depended on the number of binding sites assumed. For only one 111In site per molecule, the quadrupole frequency increases by over 50% as the temperature is reduced below the freezing point of the solution. If two 111In sites are assumed there appears to be a change in the percentage occupancy of the two sites on either side of the transition.

Radioassay problems associated with the clinical use of a 82Rb radionuclide generator.

The short-lived positron emitting radionuclide 82Rb (t1/2 1.27 min) is conveniently available from a 82Sr/82Rb generator system. The parent nuclide (t1/2 25.5d) produced from the spallation of molybdenum, has associated with it varying quantities of other long-lived strontium radionuclides, namely 85Sr, 89Sr and 90Sr. It is important therefore in the clinical use of such material that the levels of strontium radionuclides being administered to patients is carefully assayed and controlled. The problems associated with these measurements are discussed with special reference to the radiation dose received by the patient and the problems in resolving overlapping peaks with different FWHMs.

Spleen function and platelet kinetics.

In patients suffering from various platelet abnormalities, quantitative scanning after injection of indium-111 (111In) labelled platelets showed three different patterns of platelet destruction and distribution. In patients with a normal platelet life span but with evidence of increased splenic pooling, the spleen tended to be the main site of destruction. In patients with a moderately reduced platelet life span, the distribution of destruction in the system and variable destruction in the marrow. However, because of its rapidity this destruction was difficult to quantify, and it was difficult in these cases to distinguish reliably between spleen pool, sequestration, and destruction. Destruction of platelets on the liver appeared to be unimportant in all three groups. 111In, because of its physical characteristics, is preferable to chromium-51 as a platelet label in the assessment of abnormal platelet kinetics.

Radionuclide distribution following injection of 111Indium-labelled platelets.

Platelets labelled with 111In displayed similar survival curves after incubation with 111In-oxine in plasma, plasma-saline and dextrose saline media. The use of autologous red cells to cushion platelets during high-speed centrifugation facilitated platelet resuspension without greatly affecting the duration of the labelling procedure. Quantitative scanning after reinjection of labelled platelets in haematologically normal subjects showed that, initially, splenic indium amounted to about 35% of the injected dose and hepatic indium about 12%; these levels rose only slightly over the subsequent duration of the platelet life span. Subtraction of the signal from indium in platelets thought to be normally pooled within the spleen from the total indium signal gave splenic indium uptake curves which reflected splenic platelet destruction. Initially, the sum of indium levels in spleen, liver and blood equalled 100% of the dose. Thereafter, the sum fell progressively at a rate thought to be approximately equal to the rate of bone marrow uptake.