Split-Bolus Spectral CT, Less Radiation Exposure, and Pancreatic Tumors

The use of split-bolus spectral multi-detector computed tomography (CT) for patients with suspected pancreatic tumor resulted in better tumor conspicuity and equivalent depiction of the arterial and portal venous system than standard CT. It also reduced the radiation dose by more than 40 percent.

Olga R. Brook, MD, of Beth Israel Deaconess Medical Center, and colleagues published the results of the comparison in Radiology.

“Split-bolus contrast material administration is used in a number of CT applications, most notably CT urography,” the researchers wrote. “Instead of a single continuous injection, various amounts of contrast material are injected in two or three parts with a variable pause between injections; scanning is performed only once.”

Brook and colleagues conducted the study to examine the efficacy of split-bolus contrast injection combined with spectral imaging in detecting tumor conspicuity of pancreatic tumors, and its effect on radiation dose.

Depending on scanner availability, 163 patients who required CT for a suspected pancreatic mass underwent imaging with either a standard (n=80) or split-bolus spectral CT protocol (n=83).

“Split-bolus spectral CT (CT unit with spectral imaging) combines pancreatic and portal venous phases in a single scan: 70 seconds before CT, 100 mL of contrast material is injected for the portal venous phase followed approximately 35 seconds later by injection of 40 mL of contrast material to boost the pancreatic phase,” the researchers explained. “Bolus tracking after the second bolus initiates scanning 15 seconds after aorta enhancement reaches 280 HU. Images were reconstructed at 60 and 77 keV.”

Eighty-five patients were found to have pancreatic lesions: 46 in the standard group and 39 in the experimental group.  The researchers found no significant difference between patients in the group in terms of age, weight, and body mass index.

There was a significant reduction in the mean per-examination estimated dose-length product with the split-bolus spectral protocol compared with the standard protocol (633 mGy/cm ± 105 vs. 1,112 mGy/cm ± 437; P<.001).

Greater mean tumor conspicuity was found with patients on the 60 keV split-bolus protocol than at the portal and venous phases of the multiphase standard protocol (89.1 HU ± 56.6 vs. 43.5 HU ± 28.4 and 51.5 HU ± 30.2, respectively; P<.001 for both). Although pancreatic cysts were more conspicuous than tumors for both protocols, the split-bolus protocol still resulted in greater conspicuity.

Mean tumor contrast-to-noise was also higher using the split-bolus spectral protocol at 60 keV (8.8 ± 6.2) compared with the pancreatic (4.5 ± 3.0; P=.002) or portal venous phases of the standard protocol (5.6 ± 4.0; P=.02).

The researchers pointed out that one of the weaknesses of the study was that a comparison of techniques was not performed in each patient leading to a possible bias due to patient selection.

“Further studies with larger groups of patients will be helpful to confirm our findings,” they wrote. “Further research in split-bolus spectral CT is also needed to characterize subgroups of pancreatic tumors (eg, hypoenhancing vs. hyperenhancing lesions).”