Assessment of radiopharmaceutical retention for vascular access ports using positron emission tomography imaging

Abstract Purpose The purpose of this study was to resolve the issue of whether various generations of CR Bard peripheral vascular access ports and catheters are prone to retain PET radiopharmaceuticals. The study evaluates the residual radioactivity remaining following injection for two PET radiopharmaceuticals currently used extensively in the clinic, FDG and Na18F. Methods FDG was purchased from a local cyclotron facility and Na18F was prepared in‐house. Three generations of currently marketed vascular access ports were tested. A total of five (n = 5) of each model was tested. Radiopharmaceutical of 2–3 mCi of each was injected into each port and flushed with 10, 30, 60, and 120 ml of saline. MicroPET scans were performed after each flush to detect the residual radioactivity on each port. A dose calibrator was used to detect the retention of radioactivity after each flush. Results Radioactivity retention for all vascular port models measured by microPET imaging was similar for both FDG and Na18F, with less than 1% residual activity following a 10 ml saline flush. Based on the microPET images, all the subsequent flushes of 30, 60, and 120 ml were also considered. Dose calibrator activity measurements validated microPET measurements as negligible for all the ports, even with the first 10 ml flush. Conclusions MicroPET imaging was more sensitive than the dose calibrator in determining the radioactivity retention of the vascular access ports from CR Bard. These ports may be used for the injection of FDG and Na18F to track glucose metabolism and bone uptake with PET imaging. It is recommended to apply at least a 10 ml flush after radiopharmaceutical administration, to reduce residual activity to baseline levels.

chest region. With a minor surgery technique, surgeons can quickly insert the catheter of the port into a vein, while subcutaneously positioning the port body such that the septum is immediately under the skin surface.
Computerized tomography examinations are commonly performed with the intravenous use of iodinated contrast media, either as a fundamental part of the examination or to improve soft-tissue distinction. 4 Port systems allow for liquids to be injected into the bloodstream at a higher rate than by peripherally inserted central catheter injection. The use of vascular access ports is considered by many physicians to be a significant alternative to prevent potentially damaging attempts at percutaneous indwelling access. Physicians must weigh the relative risks and benefits of using a vascular access port for power injection of contrast material, as they will have a clearer knowledge of the relative importance of contrast enhancement for accurate physiological and anatomical interpretation. 4 Positron emission tomography (PET) is a noninvasive diagnostic tool that similarly provides tomographic images, while enabling observation of perfusion, cell viability, proliferation, and metabolic activity in tissues. The use of miniature PET imaging systems for molecular imaging in preclinical research has been shown to be a valuable tool and is gaining increased use in many fields including drug development. 5 There have been conflicting anecdotal comments on administration of PET radiotracers through ports with regard to activity retention within the port body and the attached catheter following injection. The majority of the risk in using a port is whether the radiopharmaceutical is taken up by the materials of the device that may result in residual image artifacts. Undesirable contrast, resolution, and artifacts can lead to a potential error for physiological PET tracer uptake errors in diagnosis. 6,7 An international team has already investigated similar issues, leading up to the need for this study. They looked at hydrodynamics and temperature as key parameters for the efficiency of a vascular access device to rid contrast media directly. The group evaluated six devices (n = 6) and indicated using a Mann-Whitney U-test that port cavities are incompletely rinsed after three consecutive flushes of 10 ml saline solution for a net 30 ml flush. 8 Here, we explore the ramifications of introducing various port and catheter designs as well as two commonly used radiotracers using PET imaging: FDG and Na 18 F. Similarly, testing of a total of five devices (n = 5) of each port model were explored. Results from consecutive flushes of injected saline are described for each port and tracer type.

| MATERIALS AND METHODS
Radioactive fluoride ( 18 F) and 18 F-fluorodeoxyglucose (FDG) was first purchased from PETNET Solutions, Inc. (Louisville, KY, USA). Radiolabeled 18 F-Sodium Fluoride (Na 18 F) was then prepared from the acquired 18 F in-house as described in the literature. 9,10 Briefly, the process began with 18 F-fluoride and 0.1 ml H 18 O diluted with 5 ml sterile water and passed through a cation exchange (in the form of H + ) cartridge connected to a Waters Corporation Sep-Pak Accell Plus QMA cartridge (Milford, MA, USA). The cation exchange cartridge was then removed. The QMA cartridge was rinsed with 10 ml sterile water and air dried. Finally, Na 18 F was eluted with 5-10 ml saline and passed through a 0.2 lm filter to provide the end product for imaging. The yield for Na 18 F was better than 90%.
The imaging system of choice for this study was a Siemens R4 MicroPET scanner (Knoxville, TN, USA). Images were reconstructed using an OSEM2D algorithm and analyzed with ASIPro software.
The dedicated third-generation PET scanner was capable of obtaining millimeter-level high-spatial resolution at the center of the field of view. Calibration and quality assurance testing were performed prior to any acquisition. To accompany PET activity detection, radioactivity was also measured using a dose calibrator (Biodex Atomlab 500).  Table 1.
Each port and catheter pairing was primed with 2 ml saline and heated at 37°C for 24 hr prior to use. Five sets (n = 5) for each of the three port designs were tested, using both FDG and Na 18 F, resulting in a total of 30 ports used in this study. For each type, one was evaluated as a negative control. Ports were then subject to injection of 2-3 mCi (74-111 MBq) of FDG or Na 18 F, respectively, where they served as positive controls. The ports were individually and identically taped to a minified cardboard couch and positioned to best fit within the field of view for the scanner (Fig. 1).
There was no time lag in between flushes with imaging. Volumes were introduced cumulatively. For example, the first flush was 10 ml with imaging, followed by an additional 20 ml to additively arrive at 30 ml for the second level with imaging, and so on to 60 and 120 ml. A PET scan was then performed following each repetitive flush. Regions-of-interest (ROIs) were manually drawn over the images using ASIPro software that is used specifically with the microPET-R4 scanner. From these ROIs, the mean voxel values of the radioactive decay counts were determined. The net activity of each port and catheter type was obtained by summation of activity in 20 consecutive coronal slices. Retention rates were calculated as 100% x (Counts in each flushed sample-Counts of negative control)/ (Total counts of positive control). The residual radioactivity for each sample was also measured using a well-counter detector system to further confirm the PET results.
In order to understand whether dead-time and cross-talk from adjacent slices played a significant role in our measurements, tubing was prepared and filled with 1.65 mCi of FDG in 1.2 ml volume. It was then put on top of another piece of tubing without being filled with FDG inside the field of view of the microPET scanner. Micro-PET images for this entire assembly were taken serially every hour for 7 hr.
An analysis of variance statistical model was used to compute one-way test P-values. 11 Consistent with standard usage, only P-values less than a = 0.05 were considered significant. Results validate the hypothesis that specific volumes of saline are necessary to remove residual activity of either FDG or Na 18 F in each of the port models considered.

| RESULTS
The retention rates for different port models based on PET quantification are shown in Table 2 for FDG and Table 3 for Na 18 F. Background scan PET scan saline-filled port and catheter for 10 min as background control  was reduced from 0.66% to 0.61%, while for PowerPort â ClearVUE â (model 1618000) it was reduced from 0.66% to 0.54% when the flushing volume went from 10 to 120 ml. The average percentage retention for all three models after a 10 ml saline flush was 0.71 AE 0.22%. For Na 18 F, a similar trend was observed in all three port models. The percentage retention did not show any significant difference among all three models, with an average of 0.69 AE 0.26%. Moreover, there was no remarkable benefit to flushing more than 10 ml, although some reduction was noted. The average change for ports evaluated with either contrast agent between 10 and 120 ml was a mere 0.10 AE 0.10% on average. For both tracers, radioactivity for all port models measured by the well-counter was close to background after flushing with 10 ml of saline.
The spill-over of counts from the chosen PET image slice to adjacent slices were expectedly of the same order of magnitude as our retention measurements.
Statistical P-values for all models of ports tested with both FDG and Na 18 F are reported in Tables 2 and 3. Measured activities were deemed not statistically different from one another when the Pvalue >0.05. All P-values were estimated to be larger than 0.05 among all the groups with different flushing amount for all three port models. From this data, it was concluded that only a 10 ml saline flush was required following the administration of either FDG or Na 18 F.

| DISCUSSION
In this study, three types of CR Bard vascular access port types were evaluated and compared for their inherent ability to retain or release radioactive contrast agents used for PET imaging. We note In order to evaluate if there might be a chance of having activity contamination, which could affect the ROI values, ports were re-taped onto the cardboard after each flush prior to PET imaging. We also considered the possibility of slice numbers within the field of view being different or angled. Care was applied so as to duplicate alignment of the cardboard couch and to determine retention in the same couch position for all measurements, using the same technique.
Dead-time was also an unlikely contributing factor in this process. The measured decay over time for 18 F is shown in Fig. 4 We caution that similar testing could potentially reveal the presence of above background remaining activity with only a single 10 ml saline flush. If remarkable activities remain even after saline flushing, such contrast could give rise to image artifacts that affect diagnostic reading, in the form of false positives or negatives, and even further affect treatment options for patients. 14,15 F I G . 4. Time activity curve generated from a series of microPET images taken every hour for 7 hr. 1.65 mCi of FDG in 1.2 ml filled up the tubing.

| CONCLUSION
We examined three of the most common vascular access port systems from CR Bard ex vivo in order to characterize the ability of each port in retaining FDG and Na 18 F. Both PET radiotracers are routinely used in the clinic. The PET scan analysis and dose calibrator assays revealed that no significant residual radioactivity remained after only a 10 ml flush. No remarkable benefit was identified when flushing more than 10 ml. The average change for ports evaluated with either contrast agent between 10 and 120 ml was mere 0.1%. These evaluated vascular access ports from CR Bard may be used for the injection of FDG and Na 18 F used for PET imaging.

CONF LICTS OF INTEREST
The authors have no relevant financial interests in the manuscript and no other potential conflicts of interest to disclose.