Radiotracer kinetic modeling in small animals with PET allows absolute quantification of physiologic and biochemical processes in vivo. It requires blood and tissue tracer concentrations as a function of time. Manual sampling, the reference method for blood tracer concentration measurements, requires fairly large amounts of blood besides being technically difficult and timeconsuming. An automated microvolumetric b blood counter (mBC) was designed to circumvent these limitations by measuring the blood activity in real time with PET scanning. Methods: The mBC uses direct b-particle detection to reduce its footprint and is entirely remote controlled for sampling protocol selection and real-time monitoring of measured parameters. Sensitivity has been determined for the most popular PET radioisotopes ( 18 F, 13 N, 11 C, 64 Cu). Dispersion within the sampling catheter has been modeled to enable automatic correction. Blood curves obtained with the mBC were compared with manual samples and PET-derived data. The mBC was used to estimate the myocardial blood flow (MBF) of mice injected with 13 N-ammonia and to compare the myocardial metabolic rate of glucose (MMRG) of rats injected with 18 F-FDG for arterial and venous cannulation sites. Results: The sensitivity limit ranges from 3 to 104 Bq/mL, depending on the isotope and the catheter used, and was found to be adequate for most small-animal studies. Automatic dispersion correction appears to be a good approximation of dispersionfree reference curves. Blood curves sampled with the mBC are well correlated with curves obtained from manual samples and PET images. With correction for dispersion, the MBF of anesthetized mice at rest was found to be 4.84 6 0.5 mL/g/min, which is comparable to values found in the literature for rats. MMRG values derived from the venous blood tracer concentration are underestimated by 60% as compared with those derived from arterial blood. Conclusion: The mBC is a compact automated counter allowing real-time measurement of blood radioactivity for pharmacokinetic studies in animals as small as mice. Reliable and reproducible, the device makes it possible to increase the throughput of pharmacokinetic studies with reduced blood sample handling and staff exposure, contributing to speed up new drug development and evaluation.

Small-animal nuclear imaging modalities have become essential tools in the development process of new drugs, diagnostic procedures, and therapies. Quantification of metabolic or physiologic parameters is based on pharmacokinetic modeling of radiotracer biodistribution, which requires the blood input function in addition to tissue images. Such measurements are challenging in small animals because of their small blood volume. In this work, we propose a microfluidic counting system to monitor rodent blood radioactivity in real time, with high efficiency and small detection volume (∼1 μL). Methods: A microfluidic channel is built directly above unpackaged p-i-n photodiodes to detect β-particles with maximum efficiency. The device is embedded in a compact system comprising dedicated electronics, shielding, and pumping unit controlled by custom firmware to enable measurements next to small-animal scanners. Data corrections required to use the input function in pharmacokinetic models were established using calibrated solutions of the most common PET and SPECT radiotracers. Sensitivity, dead time, propagation delay, dispersion, background sensitivity, and the effect of sample temperature were characterized. The system was tested for pharmacokinetic studies in mice by quantifying myocardial perfusion and oxygen consumption with 11 C-acetate (PET) and by measuring the arterial input function using 99m TcO 4 − (SPECT). Results: Sensitivity for PET isotopes reached 20%-47%, a 2-to 10-fold improvement relative to conventional catheter-based geometries. Furthermore, the system detected 99m Tc-based SPECT tracers with an efficiency of 4%, an outcome not possible through a catheter. Correction for dead time was found to be unnecessary for small-animal experiments, whereas propagation delay and dispersion within the microfluidic channel were accurately corrected. Background activity and sample temperature were shown to have no influence on measurements. Finally, the system was successfully used in animal studies. Conclusion: A fully operational microfluidic blood-counting system for preclinical pharmacokinetic studies was developed. Microfluidics enabled reliable and high-efficiency measurement of the blood concentration of most common PET and SPECT radiotracers with high temporal resolution in small blood volume. Radi onuclide-based molecular imaging using PET and SPECT is a leading diagnostic tool in oncology, cardiology, and neurology (1). In research applications, small-animal models are needed to facilitate the development of new drugs, radiotracers, and therapies that can eventually be translated to humans (2). Quantification of metabolic or physiologic disorders using radiotracer pharmacokinetic models requires dynamic blood analysis in addition to tissue imaging data (3). Whole-blood radioactivity as a function of time, the so-called arterial input function (AIF), can be extracted from PET images (4,5), using an intravascular b-microprobe (6,7) or an external blood counter connected to an artery via a catheter (8-10). Fo...

No abstract

New radiotracer developments for nuclear medicine imaging require the analysis of blood as a function of time in small animal models. A microfluidic device was developed to monitor the radioactivity concentration in the blood of rats and mice in real time. The microfluidic technology enables a large capture solid angle and a reduction in the separation distance between the sample and detector, thus increasing the detection efficiency. This in turn allows a reduction of the required detection volume without compromising sensitivity, an important advantage with rodent models having a small total blood volume (a few ml). A robust fabrication process was developed to manufacture the microchannels on top of unpackaged p-i-n photodiodes without altering detector performance. The microchannels were fabricated with KMPR, an epoxy-based photoresist similar to SU-8 but with improved resistance to stress-induced fissuring. Surface passivation of the KMPR enables non-diluted whole blood to flow through the channel for up to 20 min at low speed without clotting. The microfluidic device was embedded in a portable blood counter with dedicated electronics, pumping unit and computer control software for utilisation next to a small animal nuclear imaging scanner. Experimental measurements confirmed model predictions and showed a 4- to 19-fold improvement in detection efficiency over existing catheter-based devices, enabling a commensurate reduction in sampled blood volume. A linear dose-response relationship was demonstrated for radioactivity concentrations typical of experiments with rodents. The system was successfully used to measure the blood input function of rats in real time after radiotracer injection.

A rapid prototyping method was developed to integrate a microfluidic plasma separation device over a Si based radiation detector. KMPR photoresist was used to produce high pressure resistant channels sealed by a glass cover. Two methods were investigated to make access holes in the cover: laser micromachining and a simplified glass wet etching process. The lab-on-chip concept where a plasma separation device is coupled to an embedded on-chip p-i-n photodiode used as a high sensitivity beta particle detector was investigated. A first detector prototype was fabricated and characterized. KMPR autofluorescence was investigated to optimize microfluidic function characterization with fluorescent beads.

No abstract

No abstract

No abstract