Usefulness of Topically Applied Sensors to Assess the Quality of 18F-FDG Injections and Validation Against Dynamic Positron Emission Tomography (PET) Images

Abstract

Background: Infiltrations of 18F-fluorodeoxyglucose (FDG) injections affect positron emission tomography/computed tomography (PET/CT) image quality and quantification. A device using scintillation sensors (Lucerno Dynamics, Cary, NC) provides dynamic measurements acquired during FDG uptake to identify and characterize radioactivity near the injection site prior to patient imaging. Our aim was to compare sensor measurements against dynamic PET image acquisition, our proposed reference in assessing injection quality during the uptake period. Methods: Subjects undergoing routine FDG PET/CT imaging were eligible for this Institutional Review Board approved prospective study. After providing informed consent, subjects had sensors topically placed on their arms. FDG was injected into subjects' veins directly on the PET imaging table. Dynamic images of the injection site were acquired during 45 min of the uptake period. These dynamic image acquisitions and subjects' routine standard static images were evaluated by nuclear medicine physicians for abnormal FDG accumulation near the injection site. Sensor measurements were interpreted independently by Lucerno staff. Dynamic image acquisition interpretation results were compared to the sensor measurement interpretations and to static image interpretations. Results: Twenty-four subjects were consented and enrolled. Data from 21 subjects were gathered. During dynamic image acquisition review, physicians interpreted 4 subjects with no FDG accumulation at the injection site, whereas 17 showed evidence of accumulation. In 10 of the 17 cases that showed FDG accumulation, the FDG presence at the injection site resolved completely during uptake corresponding to venous stasis, the temporary sequestration of blood from circulation. Static image interpretation agreed with dynamic images interpretation in 11/21 (52%) subjects. Sensor measurement interpretations agreed with the dynamic images interpretations in 18/21 (86%) subjects. Conclusions: Sensor measurements can be an effective way to identify and characterize infiltrations and venous stasis. Comparable to an infiltration, venous stasis may produce spurious and clinically meaningful measurement bias and possibly even scan misinterpretation. Since the quality and quantification of PET/CT studies are of clinical importance, sensor measurements acquired during the FDG uptake may prove to be a useful quality control measure to reduce infiltration rates and potentially improve patient care. Registration: Clinicaltrials.gov, Identifier: NCT03041090.

Keywords: FDG; PET/CT; extravasation; infiltrations; quality control.

Figures

Figure 1

TAC from right antecubital fossa injection, butterfly access, 24 gauge needle. Black TAC, injection arm; Red TAC, reference arm; Green TAC, reference wrist; Blue TAC, Liver.

Figure 2

Subject flow chart.

Figure 3

Dynamically acquired images (bottom row) taken at various times (blue arrows) during uptake, static image (upper right), and black TAC (top curve) from injection arm sensor, all reflect the resolving nature of an infiltration. Red TAC (bottom curve) reflects increasing uptake as captured by the reference sensor on the non-injection arm during the uptake period. Blue TAC reflects liver uptake and plays a negligible role in injection arm sensor TAC interpretation due to injection site location and nature of the liver uptake. Patient was positioned on PET imaging table to ensure the injection site would be located near the caudal edge of the PET imaging bed. Red arrows on dynamic image acquisition frames and static image indicate approximate injection site (right hand).

Figure 4

Dynamically acquired images (bottom row) taken at various times (blue arrows) during uptake, static image (upper right), and black TAC (top curve) from injection arm sensor all reflect the complete resolution of a prolonged venous stasis. Red TAC reflects uptake captured by reference arm sensor on the non-injection arm during the uptake period. Blue TAC reflects liver uptake. Proximity to injection arm sensor and nature of the liver uptake played some role in injection arm sensor TAC interpretation. Patient was positioned on PET imaging table to ensure the injection site would be located near the caudal edge of the PET imaging bed. Red arrows on dynamic image acquisition frames and static image indicate approximate injection site (right antecubital fossa).

Figure 5

Dynamically acquired images (bottom row) taken at various time points (blue arrows) during uptake, static image (upper right), and black TAC from injection arm sensor all reflect an ideal radiotracer injection. Injection counts drop to very low levels immediately after bolus injection and saline flush, until the time the injection arm is placed in plexiglass table extenders and uncollimated sensors capture background torso radioactivity. At this time the injection curve climbs slightly and then levels off. Reference arm was placed in table extender before the injection. Red TAC reflects uptake captured by reference arm sensor on the non-injection arm during the uptake period. Blue TAC reflects liver uptake. Proximity to the reference arm sensor and nature of the liver uptake results in higher reference arm counts. Patient was positioned on PET imaging table to ensure the injection site would be located near the caudal edge of the PET imaging bed. Red arrows on dynamic image acquisition frames and static image indicate approximate injection site (left antecubital fossa).