Initial Experience with Oral Contrast in PET/CT: Phantom and Clinical Studies

Data Acquisition

Scanning was performed using a PET/CT scanner (Discovery LS; GE Medical Systems, Waukesha, WI). Emission data were acquired for 5-7 positions at 5 min per position and were reconstructed using CT for attenuation correction (ordered-subset expectation maximum algorithm). The parameters of the multidetector helical CT were 140 kVp, 80 mA, 0.8 s per CT rotation, pitch of 6, and 22.5 mm/s table speed. The conversion scale used to transform Hounsfield units (HU) into attenuation coefficients was described previously (5).

Phantom Studies

The first 2 experiments were performed with a phantom filled with 74 MBq ¹⁸F activity (0.24 mCi/mL) and a saline bag filled with CT oral contrast (READY CAT barium sulfate suspension, 1.3% weight/volume [w/v]; E-Z-EM Inc., Westbury, NY) and high-density barium (E-Z-HD barium sulfate suspension, 98% w/v; E-Z-EM Inc.), respectively (to simulate a contrast-filled stomach). In the third experiment, 8 syringes filled with varying concentrations (100%, 80%, 60%, 40%, 20%, 10%, 1%, and 0%) of MD-Gastroview (Mallinckrodt Inc., St. Louis, MO) were placed in a cylindric phantom containing 74 MBq ¹⁸F activity diluted with water (0.35 mCi/mL).

In all experiments, CT was performed, followed by emission and transmission scanning with ⁶⁸Ge (3 min per position). Two datasets of PET images with ⁶⁸Ge and CT-based correction and 1 without attenuation correction were reviewed by an experienced nuclear medicine physician using eNTEGRA software (ELGEMS, Haifa, Israel). Images were reviewed for abnormal uptake in the region of the stomach present on the attenuation-corrected images and absent on the noncorrected images. For the third experiment, a 1.91-cm² region of interest (ROI) was used to measure the “activity” and contrast density for each syringe.

Patient Studies

A total of 108 clinical patients were evaluated retrospectively by whole-body ¹⁸F-FDG imaging after oral contrast utilization was initiated. Patients received 2 bottles (450 mL per bottle) of CT contrast (barium suspension; 1.3%) at least 10 min before receiving a 555- to 740-MBq ¹⁸F-FDG intravenous injection, followed by another bottle of contrast 40 min later.

Image Analysis

One experienced reader performed image analysis using eNTEGRA software. Patient compliance with contrast ingestion and quantity of contrast were reviewed. The extent of bowel opacification was categorized as absent, partial, or adequate on the CT images. For the first 30 patients receiving oral contrast, density of contrast in the digestive tract was measured in HU using a manually drawn ROI. The maximum diameter in millimeters of the gastric lumen, evaluated in a transverse section of the midportion of the stomach, was then measured for CT and PET images. PET window levels were adjusted for each patient to those used for clinical interpretation (liver activity set to moderate gray intensity). A soft-tissue window (minimum = − 250 HU; maximum = +250 HU) was used for CT. The possible effect of oral contrast ingestion on neck muscle uptake was evaluated in 2 groups of 20 patients (with and without contrast) by measuring metabolic activity in several neck muscles using a circular ROI (6-pixel area).

Phantom Studies

The first experiment with CT oral contrast (1.3% w/v) showed a cold area in the region of contrast in the nonradioactive stomach for all 3 sets of images. The maximum density in the stomach was 182 HU.

The second experiment with high-density contrast showed a cold area in the region of the nonradioactive stomach in the noncorrected and ⁶⁸Ge-corrected images. The CT-corrected images, however, showed an artifactual focus of intense “activity” with a maximal density measured at 3,071 HU in the nonradioactive stomach.

In the third experiment, ⁶⁸Ge-corrected and noncorrected images showed the cold syringes in a uniform background. The CT attenuation correction produced similar expected “cold”-appearing images of the syringes containing 0% and 1% contrast, but the higher concentrations showed increasing levels of activity, with intense artifacts at >60% concentration (not shown). Figure 2 shows that the measured activity is progressively overestimated with CT attenuation correction at high HU using the commercial software.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Phantom with syringes filled with variable concentrations of gastrografin, plotted as measured emission data activity (in Bq/mL) versus measured contrast density by CT (in Hounsfield units [HU]). Maximum (♦) and mean (▪) measured activity with CT attenuation correction, and maximum (▴) and mean (✖) measured activity with ⁶⁸Ge attenuation correction are displayed. Drop in measured activity is observed for high-density contrast. In this region, “real density” is >3,071 HU, but the PET/CT system plateaus at this level. Thus, overcorrection in presence of high-density contrast also has plateau corresponding to 3,071 HU density. However, because of increased true attenuation as result of increasing “real” density, measured counts decrease, resulting in drop in curve.

Patient Studies

Of the 108 patients (84%), 91 received oral contrast. Of these, 75 patients received the full regimen of contrast and 16 received only part of the contrast because of inability to drink the entire volume. Seventeen did not receive any contrast because of the study indication (head and neck cancer [n = 11]) or refusal (n = 6). Details on bowel opacification are reported in Table 1. Measured oral contrast density in the digestive tract is reported in Table 2.

The diameter of the stomach lumen measured on CT was significantly different (59.8 ± 20.6 mm) from that measured on PET emission images with CT attenuation correction (43.8 ± 23; P < 0.000001). Muscle activity tended to be higher in patients who received oral contrast than in those without contrast (Table 3). A systematic review of the diagnostic utility of oral contrast in enhancing the interpretation of PET/CT images (Fig. 3) will be reported subsequently. No obvious foci of artifactual “uptake” resulting from CT-based attenuation correction in the presence of CT oral contrast were found. There were thus no false-positive interpretations related to such potential artifacts.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Example of clinical value of oral contrast in patient with resected colorectal carcinoma evaluated with ¹⁸F-FDG PET for suspected recurrence. (A) Transverse CT image shows transverse colon with contrast (white arrow) and soft-tissue- density lesion adjacent to anterior abdominal wall (black arrow). (B) PET images with CT attenuation correction show focus of intense uptake in anterior abdomen (white arrow). (C) Fused PET/CT image shows that focus of uptake corresponds to soft-tissue mass (white arrow) and not to bowel. Diagnosis of peritoneal tumor implant was made. Patient underwent chemotherapy.