Split-Bolus Spectral Multidetector CT of the Pancreas: Problem Solving in the Detection of "Isoattenuating" Pancreatic Cancer?

We read with great interest the article by Dr Brook and colleagues in the October 2013 issue of Radiology (1). We greatly appreciate the protocol for a 43% reduction in radiation dose, but there are some concerns to be discussed. Pancreatic ductal adenocarcinoma (PDA), which is “isoattenuating” at computed tomography (CT), is reported in 5%–45% of cases (2–5). In the study by Dr Brook and colleagues, an equal percentage of tumors were detected with both standard and split-bolus CT techniques, and results of short-term follow-up did not reveal any missed lesions. For quantitative analysis, Dr Brook and colleagues erroneously considered surrounding pancreatic parenchyma to the tumor as “normal pancreas.” Histologic demonstration of normal pancreas is not reported. In our experience with multiphasic CT in 38 surgically resected PDAs (6,7), coexisting pancreatitis in the surrounding pancreas (upstream to the tumor) was revealed histologically in all cases, making diagnosis difficult because of the overlapping attenuation values between the tumor and pancreas upstream; four of the 38 PDAs (10%) were unrecognizable at quantitative analysis. In our study, the mean attenuation values at the arterial (pancreatic) phase of multiphase CT were lower than that reported in the same phase of the multiphase protocol by Dr Brook and colleagues (83 HU ± 27 vs 105.1 HU ± 29.3, respectively). In addition, we believe that the mean attenuation values reported by Dr Brook and colleagues with split-bolus spectral multi detector CT at 60 keV are excessively high with respect to those reported in the arterial (pancreatic) phase (212.1 HU ± 64.7 vs 83 HU ± 27, respectively). To increase the sensitivity of the split-bolus technique in the detection of isoattenuating PDA, in addition to combined-phase images in a single scan we suggest a delayed phase in the upper abdomen at 5 minutes (8).