A system for accurate and automated injection of hyperpolarized substrate with minimal dead time and scalable volumes over a large range

Abstract

Over recent years hyperpolarization by dissolution dynamic nuclear polarization has become an established technique for studying metabolism in vivo in animal models. Temporal signal plots obtained from the injected metabolite and daughter products, e.g. pyruvate and lactate, can be fitted to compartmental models to estimate kinetic rate constants. Modeling and physiological parameter estimation can be made more robust by consistent and reproducible injections through automation. An injection system previously developed by us was limited in the injectable volume to between 0.6 and 2.4ml and injection was delayed due to a required syringe filling step. An improved MR-compatible injector system has been developed that measures the pH of injected substrate, uses flow control to reduce dead volume within the injection cannula and can be operated over a larger volume range. The delay time to injection has been minimized by removing the syringe filling step by use of a peristaltic pump. For 100μl to 10.000ml, the volume range typically used for mice to rabbits, the average delivered volume was 97.8% of the demand volume. The standard deviation of delivered volumes was 7μl for 100μl and 20μl for 10.000ml demand volumes (mean S.D. was 9 ul in this range). In three repeat injections through a fixed 0.96mm O.D. tube the coefficient of variation for the area under the curve was 2%. For in vivo injections of hyperpolarized pyruvate in tumor-bearing rats, signal was first detected in the input femoral vein cannula at 3-4s post-injection trigger signal and at 9-12s in tumor tissue. The pH of the injected pyruvate was 7.1±0.3 (mean±S.D., n=10). For small injection volumes, e.g. less than 100μl, the internal diameter of the tubing contained within the peristaltic pump could be reduced to improve accuracy. Larger injection volumes are limited only by the size of the receiving vessel connected to the pump.

Keywords: Dead space minimization; Hyperpolarization; In vivo automated injection system; Large volume range.

Graphical abstract

Fig. 1

(a) Overview of injector system design showing the control system with the stepper motor, flexible drive shaft and pump and the receive vessel, (b) pump chassis, drive shaft and 3-way stopcock, and (c) image of injector system.

Fig. 2

Schematic and image of hyperpolarized fluid receive vessel (not to scale). Shown are the receive vessel with pipe guide (into which the DNP polarizer dissolution pipe was inserted), air driven stirrer, pH electrode and vacuum pipe. Connection to an animal cannula is made via a Luer 3-way stopcock fitted to the outlet of the pump.

Fig. 3

Flow diverter configuration showing: (a) schematic of flow diverter cannula, pinch valve for controlling flow to the rat and normally closed (NC) diaphragm valve, (b) complete fluid path for flow diversion system, and (c) 2 pinch valves.

Fig. 4

(a) Sketch of rat showing location of surface coil and routing of i.v. cannula, (b) axial FLASH image of the tumor, reference phantom and femoral vein cannula. A cartoon representation (blue box) of the surface coil location and guide lines for 2 slices are also shown. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 5

Demand volume versus average delivered volume for at least 3 measurements by mass of water. Error bars represent the standard deviation in delivered volume. Inset shows zoomed region of main figure in the volume range 0.00–2.00 ml.

Fig. 6

In vitro delivery profile for injection of 1.50 ml of hyperpolarized pyruvate over 13 s as measured by the integral of the ¹³C PA signal. Temporal data acquisition commences simultaneously with injector start. 6.7 M ¹³C sodium acetate used as signal reference; (a) single injection into vial mounted on a 20 mm ¹³C/¹H surface coil, (b) 3 repeat injections through a fixed horizontal tube located 5 mm from ¹³C surface coil surface (color matched solid lines pyruvate signal, dotted colored lines acetate reference. Vertical dashed line at 13 s). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 7

Normalized integral for hyperpolarized ¹³C pyruvate versus time for 5 ml/kg injection over 13 s in three rats, (a) ¹³C signal from tail vein cannula and (b) slice localized ¹³C signal from tumor (dotted lines – lactate signal). In both figures the traces are color matched for individual animals. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)