• View in gallery
    Figure 1—

    Representative transverse CT images of the cranial abdominal region in a healthy dog depicting placement of ROIs for collection of general perfusion data from 3-mm CT slices and subsequent generation of perfusion maps. The same slice is shown in each image. The labels (eg, T1) for ROIs were enlarged on these images for easier viewing. A—The abdominal aorta is selected (cross) and 2 ROIs are placed in the pancreas (T1 and T2) and liver (T3 and T4). B, C, D and E—Representative images of color perfusion maps generated for calculation of perfusion, peak enhancement index, time to peak enhancement after the start of contrast medium injection, and blood volume, respectively, of the pancreas and liver. Colors indicate the scale from the lowest (blue) to highest (red) amount of perfusion. Data collected from the ROIs and the selected vessel were used to generate time-attenuation curves for calculation of general perfusion variables for the pancreas and liver.

  • View in gallery
    Figure 2—

    Representative transverse CT images of the cranial abdominal region in a healthy dog depicting placement of ROIs (A) and hepatic perfusion maps generated from 6-mm (reconstructed from 3-mm) slices (B through E). A—The abdominal aorta and portal vein are selected (red and green crosses, respectively), and ROIs are placed in the liver (T1 and T2) and spleen. B, C, D, and E—Representative color perfusion maps generated for calculation of hepatic arterial perfusion, hepatic portal perfusion, total hepatic perfusion, and hepatic perfusion index, respectively. Labels for ROIs were enlarged for easier viewing.

  • 1. Watson P. Pancreatitis in dogs and cats: definitions and pathophysiology. J Small Anim Pract 2015;56:312.

  • 2. Xenoulis PG. Diagnosis of pancreatitis in dogs and cats. J Small Anim Pract 2015;56:1326.

  • 3. Steiner JM, Newman S, Xenoulis P, et al. Sensitivity of serum markers for pancreatitis in dogs with macroscopic evidence of pancreatitis. Vet Ther 2008;9:263273.

    • Search Google Scholar
    • Export Citation
  • 4. Pieńkowska J, Gwoździewicz K, Skrobisz-Balandowska K, et al. Perfusion-CT—Can we predict acute pancreatitis outcome within the first 24 hours from the onset of symptoms? PLoS One 2016;11:e0146965.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 5. Rademacher N, Schur D, Gaschen F, et al. Contrast-enhanced ultrasonography of the pancreas in healthy dogs and in dogs with acute pancreatitis. Vet Radiol Ultrasound 2016;57:5864.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 6. Almeida RR, Lo GC, Patino M, et al. Advances in pancreatic CT imaging. AJR Am J Roentgenol 2018;211:5266.

  • 7. Balthazar EJ. Acute pancreatitis: assessment of severity with clinical and CT evaluation. Radiology 2002;223:603613.

  • 8. Adrian AM, Twedt DC, Kraft SL, et al. Computed tomographic angiography under sedation in the diagnosis of suspected canine pancreatitis: a pilot study. J Vet Intern Med 2015;29:97103.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 9. French JM, Twedt DC, Rao S, et al. Computed tomographic angiography and ultrasonography in the diagnosis and evaluation of acute pancreatitis in dogs. J Vet Intern Med 2019;33:7988.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 10. Yadav AK, Sharma R, Kandasamy D, et al. Perfusion CT: can it predict the development of pancreatic necrosis in early stage of severe acute pancreatitis? Abdom Imaging 2015;40:488499.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 11. Tsuji Y, Takahashi N, Tsutomu C. Pancreatic perfusion CT in early stage of severe acute pancreatitis. Int J Inflam 2012;497386.

  • 12. Grözinger G, Grözinger A, Horger M. The role of volume perfusion CT in the diagnosis of pathologies of the pancreas. Rofo 2014;186:10821093.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 13. Zwingenberger AL, Shofer FS. Dynamic computed tomographic quantitation of hepatic perfusion in dogs with and without portal vascular anomalies. Am J Vet Res 2007;68:970974.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 14. Kishimoto M, Tsuji Y, Katabami N, et al. Measurement of canine pancreatic perfusion using dynamic computed tomography: influence of input-output vessels on de-convolution and maximum slope methods. Eur J Radiol 2011;77:175181.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 15. Cáceres AV, Zwingenberger AL, Hardam E, et al. Helical computed tomographic angiography of the normal canine pancreas. Vet Radiol Ultrasound 2006;47:270278.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 16. Iseri T, Yamada K, Chijiwa K, et al. Dynamic computed tomography of the pancreas in normal dogs and in a dog with pancreatic insulinoma. Vet Radiol Ultrasound 2007;48:328331.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 17. Blomley MJ, Coulden R, Dawson P, et al. Liver perfusion studied with ultrafast CT. J Comput Assist Tomogr 1995;19:424433.

  • 18. Miles KA, Hayball MP, Dixon AK. Functional images of hepatic perfusion obtained with dynamic CT. Radiology 1993;188:405411.

  • 19. Tsushima Y, Blomley MJ, Kusano S, et al. Measuring portal venous perfusion with contrast-enhanced CT: comparison of direct and indirect methods. Acad Radiol 2002;9:276282.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 20. Motosugi U, Ichikawa T, Sou H, et al. Multi-organ perfusion CT in the abdomen using a 320-detector row CT scanner: preliminary results of perfusion changes in the liver, spleen, and pancreas of cirrhotic patients. Eur J Radiol 2012;81:25332537.

    • Crossref
    • Search Google Scholar
    • Export Citation

Advertisement

Computed tomographic evaluation of pancreatic perfusion in healthy dogs

Timothy B. Kloer1Department of Environmental and Radiological Health Sciences, College of Veterinary Medicine and Biomedical Sciences, Colorado State University, Fort Collins, CO 80523.

Search for other papers by Timothy B. Kloer in
Current site
Google Scholar
PubMed
Close
 DVM
,
Sangeeta Rao2Department of Clinical Sciences, College of Veterinary Medicine and Biomedical Sciences, Colorado State University, Fort Collins, CO 80523.

Search for other papers by Sangeeta Rao in
Current site
Google Scholar
PubMed
Close
 PhD
,
David C. Twedt2Department of Clinical Sciences, College of Veterinary Medicine and Biomedical Sciences, Colorado State University, Fort Collins, CO 80523.

Search for other papers by David C. Twedt in
Current site
Google Scholar
PubMed
Close
 DVM
, and
Angela J. Marolf1Department of Environmental and Radiological Health Sciences, College of Veterinary Medicine and Biomedical Sciences, Colorado State University, Fort Collins, CO 80523.

Search for other papers by Angela J. Marolf in
Current site
Google Scholar
PubMed
Close
 DVM

Abstract

OBJECTIVE

To evaluate the feasibility of contrast-enhanced CT for assessment of pancreatic perfusion in healthy dogs.

ANIMALS

6 healthy purpose-bred female Treeing Walker Coonhounds.

PROCEDURES

Contrast-enhanced CT of the cranial part of the abdomen was performed with 3-mm slice thickness. Postprocessing computer software designed for evaluation of human patients was used to calculate perfusion data for the pancreas and liver by use of 3-mm and reformatted 6-mm slices. Differences in perfusion variables between the pancreas and liver and differences in liver-specific data of interest were evaluated with the Friedman test.

RESULTS

Multiple pancreatic perfusion variables were determined, including perfusion, peak enhancement index, time to peak enhancement, and blood volume. The same variables as well as arterial, portal, and total perfusion and hepatic perfusion index were determined for the liver. Values for 6-mm slices appeared similar to those for 3-mm slices. The liver had significantly greater median perfusion and peak enhancement index, compared with the pancreas.

CONCLUSIONS AND CLINICAL RELEVANCE

Measurement of pancreatic perfusion with contrast-enhanced CT was feasible in this group of dogs. Hepatic arterial and pancreatic perfusion values were similar to previously published findings for dogs, but hepatic portal and hepatic total perfusion measurements were not. These discrepancies might have been attributable to physiologic differences between dogs and people and related limitations of the CT software intended for evaluation of human patients. Further research is warranted to assess reliability of perfusion variables and applicability of the method for assessment of canine patients with pancreatic abnormalities.

Abstract

OBJECTIVE

To evaluate the feasibility of contrast-enhanced CT for assessment of pancreatic perfusion in healthy dogs.

ANIMALS

6 healthy purpose-bred female Treeing Walker Coonhounds.

PROCEDURES

Contrast-enhanced CT of the cranial part of the abdomen was performed with 3-mm slice thickness. Postprocessing computer software designed for evaluation of human patients was used to calculate perfusion data for the pancreas and liver by use of 3-mm and reformatted 6-mm slices. Differences in perfusion variables between the pancreas and liver and differences in liver-specific data of interest were evaluated with the Friedman test.

RESULTS

Multiple pancreatic perfusion variables were determined, including perfusion, peak enhancement index, time to peak enhancement, and blood volume. The same variables as well as arterial, portal, and total perfusion and hepatic perfusion index were determined for the liver. Values for 6-mm slices appeared similar to those for 3-mm slices. The liver had significantly greater median perfusion and peak enhancement index, compared with the pancreas.

CONCLUSIONS AND CLINICAL RELEVANCE

Measurement of pancreatic perfusion with contrast-enhanced CT was feasible in this group of dogs. Hepatic arterial and pancreatic perfusion values were similar to previously published findings for dogs, but hepatic portal and hepatic total perfusion measurements were not. These discrepancies might have been attributable to physiologic differences between dogs and people and related limitations of the CT software intended for evaluation of human patients. Further research is warranted to assess reliability of perfusion variables and applicability of the method for assessment of canine patients with pancreatic abnormalities.

Pancreatitis is the most common exocrine pancreatic disorder found clinically in dogs.1 It is considered idiopathic in most cases and is typically divided into acute or chronic disease with or without necrosis.1,2 Accurate diagnosis of pancreatitis is difficult because of the wide variety of nonspecific clinical signs, potential for subclinical disease, and concurrent or complicating diseases such as diabetes mellitus.1 Dogs with pancreatitis can have normal findings or nonspecific changes on a CBC, serum biochemical analysis, and urinalysis.1 Currently, pancreatic lipase immunoreactivity assays are considered the most sensitive and specific serum tests for diagnosis of pancreatitis in dogs; however, reported sensitivities range from 21% to 78%.1,3,4 Abdominal ultrasonography is considered the diagnostic imaging modality of choice for diagnosis of pancreatitis in dogs; however, the highest reported sensitivity for diagnosis of acute pancreatitis is approximately 68%.1,5 The gold standard for diagnosis of pancreatitis and determination of acute versus chronic disease is histologic examination, which is not routinely performed because of associated clinical problems.1,2

Contrast-enhanced CT is considered the gold standard for diagnosis of acute pancreatitis in human patients.4,6 This imaging modality has the ability to differentiate necrotizing from edematous forms of pancreatitis.6,7 Studies8,9 have been performed to evaluate CT angiography in the diagnosis of pancreatitis in dogs and found it feasible and useful for this purpose. Perfusion CT is used to assess the probability of early necrosis in human patients with acute pancreatitis through evaluation of pancreatic microcirculation.6,10–12 Early detection of pancreatic necrosis is beneficial because necrosis is associated with systemic complications and higher mortality rates and may require different treatment than nonnecrotic pancreatitis.10–12

Zwingenberger and Shofer13 determined a single pancreatic perfusion variable for dogs by retrospective analysis of dynamic CT images in a study focused on hepatic perfusion in clinical patients with and without portal vascular anomalies. The authors of another study14 compared 2 different methods of quantitatively analyzing CT pancreatic perfusion data from dogs and determined that results for some variables were significantly correlated between the 2 methods and that vessel selection may alter these results. Lastly, dynamic CT has been used to evaluate vascular enhancement patterns of the vessels adjacent to the pancreas and to determine time-attenuation curves for those vessels and the pancreas in dogs.15,16 An easier method to determine perfusion parameters involves the use of proprietary perfusion analysis software, which can automatically calculate hemodynamic values for multiple organs. Most multi-slice CT scanners have this software capability, but to the authors’ knowledge, no study has been performed to report pancreatic hemodynamic data in healthy dogs with perfusion CT software.

Pancreatic perfusion data for healthy dogs are needed prior to use of CT for assessment of these values in dogs with pancreatitis and other pancreatic diseases. The purpose of the study reported here was to perform a prospective preliminary (pilot) investigation of the feasibility of CT for assessment of pancreatic perfusion in healthy dogs and to collect data for perfusion, peak enhancement index, time to peak enhancement after the start of contrast injection, and blood volume of the pancreas and liver in these animals by use of a CT software package designed for assessment of these variables in human patients. We additionally used the software algorithms to calculate hepatic arterial perfusion, hepatic portal perfusion, total hepatic perfusion, and hepatic perfusion index. Liver-specific variables were also investigated because the software automatically performs these calculations with the obtained data, and this could allow comparison with previously reported data in dogs.13

Materials and Methods

Animals

Six purpose-bred sexually intact female Treeing Walker Coonhounds with a mean age of 7.7 months (range, 7 to 8 months) and mean weight of 23.8 kg (range, 21.4 to 28.2 kg) were used in the study. All dogs were deemed healthy and were obtained from approved sources by Colorado State University Laboratory Animal Resources for inclusion in a larger graduate surgery and endoscopy laboratory. A CBC and serum biochemical analysis were performed with results within the respective reference ranges for all dogs. Prior to perfusion CT assessment, focal cranial abdominal ultrasonographya was performed for each dog by a veterinary radiology resident (TBK), with no abnormalities identified. The study was conducted in accordance with a protocol approved by the Colorado State University Animal Care and Use Committee (Protocol ID 17-7087A).

General anesthesia

All dogs were anesthetized and monitored (including measurement of heart rate and blood pressure) by the anesthesia service personnel. An 18-gauge IV catheter was placed in a cephalic vein according to standard protocol. Each dog was premedicated with atropine (0.02 mg/kg, SC) and morphine (1 mg/kg, SC), and some dogs also received acepromazine (0.01 mg/kg, SC). Anesthesia was induced with propofol (IV, to effect). The dogs were intubated with a cuffed endotracheal tube, and general anesthesia was maintained with isoflurane in oxygen. Each dog was mechanically ventilated and received lactated Ringer solution (5 mL/kg/h, IV) throughout the anesthetic episode.

Perfusion CT protocol

All dogs were scanned in dorsal recumbency in a trough. A 16-slice positron emission tomography–CT scannerb was used to acquire all CT scans. A breath-hold technique was used during the perfusion assessment portion of the CT scan, with controlled mechanical ventilation discontinued to briefly allow expiratory apnea to develop. Immediately after the perfusion data were obtained, mechanical ventilation was restarted. After the procedures, the dogs were monitored by the anesthesia service for any potential adverse effects until fully recovered prior to return to their housing.

An initial CT scan was performed without the use of contrast medium to identify the pancreas and liver for the subsequent collection of perfusion data. A 24-mm section of the cranial aspect of the abdomen that included portions of the pancreas, liver, and spleen was selected. Inclusion of the spleen was necessary for specific hepatic value calculations with the perfusion software. A power injectorc was used to deliver nonionic iodinated contrast mediumd (2.2 mL/kg, IV) at a rate of 5 mL/s, followed by sterile saline (0.9% NaCl) solution (1.1 mL/kg, IV) at the same rate. Dynamic evaluation included multiple CT scans performed over the same 24-mm region of the abdomen every 2 seconds for 1 minute, initiated at the exact same time as contrast medium delivery, with the following settings: 120 kV, 150 mAs, and 512 × 512 matrix. The scans were performed with 3-mm slice thickness, and the images were subsequently reconstructed into 6-mm slices. The scanner used could perform perfusion analysis in 1.5-, 3-, and 6-mm slices, and most clinical abdominal CT scans of dogs at our institution were performed with 2-mm slice thickness. In human medicine, 5-mm slices are typically used during perfusion CT scanning.4,10–12 Therefore, 3-mm and 6-mm slices were chosen when the study protocol was developed. The raw CT data obtained were used to determine the perfusion variables through postprocessing analysis with software. A full abdominal postcontrast CT scan was performed immediately following the perfusion scan to ensure no abnormalities were identified that would impact the determination of all dogs being healthy.

Perfusion CT data analysis

Raw data were evaluated in the proprietary software programe on a dedicated workstation. Structures on the CT images used for perfusion analysis included an artery, a vein, the spleen, and ROIs in the liver and pancreas, depending on the structure evaluated. One to 3 ROIs were selected in the pancreas (typically 2 regions), and 2 ROIs were selected in the liver of each dog to obtain perfusion data. The total number of ROIs was 12, with a mean of 2/dog, for both the pancreas and liver. The pancreatic and hepatic ROIs varied in size from 5.7 mm2 to 13 mm2, depending on the amount of parenchyma present on the CT slices. If a smaller portion of the pancreas was present on the slice analyzed for perfusion, a larger ROI was drawn. The specific portion of the pancreas evaluated was determined on the basis of pancreatic tissue availability within CT slices that contained the other relevant features needed for perfusion calculations; for most dogs, the ROIs were placed in the body of the pancreas. Similar ROIs were placed in the right aspect of the liver (specific lobe not identified) for each dog. When a splenic ROI was required for calculations, the region was similar in size to those placed in the liver for that specific dog. Care was taken to avoid any large blood vessels and organ boundaries when selecting ROIs for all organs. Manually calculated time-attenuation curves were made for each vessel and ROI of the dogs to ensure the values the computer used for calculations were accurate.

Functional general perfusion analysis

Functional general perfusion analysis was performed with the maximum slope method14 to determine perfusion, peak (contrast) enhancement index, time to peak enhancement after contrast medium injection, and blood volumes for the liver and pancreas (Figure 1). The abdominal aorta was selected as the artery along with ROIs in the liver and pancreas for these calculations.

Figure 1—
Figure 1—

Representative transverse CT images of the cranial abdominal region in a healthy dog depicting placement of ROIs for collection of general perfusion data from 3-mm CT slices and subsequent generation of perfusion maps. The same slice is shown in each image. The labels (eg, T1) for ROIs were enlarged on these images for easier viewing. A—The abdominal aorta is selected (cross) and 2 ROIs are placed in the pancreas (T1 and T2) and liver (T3 and T4). B, C, D and E—Representative images of color perfusion maps generated for calculation of perfusion, peak enhancement index, time to peak enhancement after the start of contrast medium injection, and blood volume, respectively, of the pancreas and liver. Colors indicate the scale from the lowest (blue) to highest (red) amount of perfusion. Data collected from the ROIs and the selected vessel were used to generate time-attenuation curves for calculation of general perfusion variables for the pancreas and liver.

Citation: American Journal of Veterinary Research 81, 2; 10.2460/ajvr.81.2.131

Functional liver perfusion analysis

Functional liver perfusion analysis included assessment of arterial perfusion (ie, perfusion during the arterial phase), portal perfusion (perfusion during the portal-venous phase), total perfusion (arterial perfusion plus portal perfusion), and hepatic perfusion index (the ratio of arterial perfusion to total perfusion). For the computer to perform these calculations, identification of an artery, a vein, and the spleen was required. The abdominal aorta was chosen as the artery; the caudal vena cava or the portal vein was chosen as the vein, and ROIs were placed on the spleen and liver (Figure 2). After discussion with technical support stafff for the equipment used, results for calculations when the caudal vena cava was selected were compared with results for calculations when the portal vein was selected. For these measurements, calculations were performed with the method described by Blomley et al.17 The perfusion software included options for use of the Blomley method or that described by Miles et al18 for functional liver perfusion data, with the Blomley method set as the default. These methods calculate arterial perfusion similarly; however, the Blomley method removes the arterial component of the portal time-attenuation curve and thus from the portal perfusion value. In clinical tests, neither method was proven superior, but the Blomley method is preferred if portal venous data are available.19 Given this information and after discussion with the technical support staff, only the Blomley method was used in this study.

Figure 2—
Figure 2—

Representative transverse CT images of the cranial abdominal region in a healthy dog depicting placement of ROIs (A) and hepatic perfusion maps generated from 6-mm (reconstructed from 3-mm) slices (B through E). A—The abdominal aorta and portal vein are selected (red and green crosses, respectively), and ROIs are placed in the liver (T1 and T2) and spleen. B, C, D, and E—Representative color perfusion maps generated for calculation of hepatic arterial perfusion, hepatic portal perfusion, total hepatic perfusion, and hepatic perfusion index, respectively. Labels for ROIs were enlarged for easier viewing.

Citation: American Journal of Veterinary Research 81, 2; 10.2460/ajvr.81.2.131

Histologic examination

The dogs underwent general anesthesia on another day after the CT analyses were complete and were included in a graduate surgical laparoscopy teaching laboratory. At the end of the laboratory session, the dogs were euthanized by IV administration of pentobarbital sodium while still anesthetized. Necropsy was performed, and pancreatic biopsy samples were obtained and submitted for histologic examination by 1 board-certified veterinary pathologist.

Statistical analysis

The Friedman test was used to compare mean ranks for perfusion variables between the liver and pancreas. The Friedman test was also used to compare mean ranks between hepatic perfusion variables calculated with the caudal vena cava and portal vein selected during user-defined input. A nonparametric approach was used because of the small sample size and because the data did not meet normality assumptions (as assessed by the Shapiro-Wilk test). Statistical analysis was performed with commercially available software.g Values of P ≤ 0.05 were considered significant.

Results

The pancreas had a normal appearance on abdominal ultrasonographic images for all dogs on the basis of echogenicity and echostructure. Heart rates varied among the 6 dogs during the perfusion CT scans with ranges of 92 to 115, 95 to 115, 100 to 110, 120 to 135, 140 to 160, and 150 to 180 beats/min. Mean arterial blood pressure ranges during this time were 60 to 70, 55 to 62, 60 to 72, 62 to 77, 65 to 78, and 60 to 75 mm Hg, respectively, for the same dogs. No dogs had evidence of adverse reactions to anesthesia or contrast medium. Examination of biopsy samples indicated that the pancreas was histologically normal in all dogs.

Perfusion CT data

No dogs or data were excluded from the analyses. The perfusion values were similar between 3-mm and 6-mm CT slices (Table 1). The manually calculated time-attenuation curves were similar to the time-attenuation curves generated by the software program for perfusion parameters of the pancreas and the liver (data not shown).

Table 1—

Pancreatic and hepatic perfusion values obtained with 3-mm and (reconstructed) 6-mm slice contrast-enhanced CT in 6 healthy young female dogs.

 3–mm slices6–mm slices
VariableMedianSERangeMedianSERange
Pancreas      
 Perfusion (mL/min/100 mL)37.83*3.3523.88–60.3543.74*3.5427.42–67.07
 Peak enhancement index (HU)47.67*2.5437.62–69.9249.20*2.3134.26–56.85
 Time to peak enhancement (s)39.293.9711.47–50.0631.813.518.34–50.47
 Blood volume (mL/100 g)18.151.5710.35–26.6620.842.4113.37–40.93
Liver      
 Perfusion (mL/min/100 mL)52.123.7946.54–85.2457.183.8647.83–90.32
 Peak enhancement index (HU)68.351.8852.33–75.0065.861.5854.21–69.88
 Time to peak enhancement (s)31.283.2118.78–50.0535.464.1517.73–56.30
 Blood volume (mL/100 g)21.112.3515.00–40.0622.363.2619.15–51.68
 Arterial perfusion (mL/min/100 mL)21.493.6312.41–51.5817.583.8014.13–52.72
 Portal perfusion (mL/min/100 mL)10.145.970.00–53.053.875.830.00–58.68
 Total perfusion (mL/min/100 mL)40.896.7616.81–91.5622.219.4514.13–111.40
 Perfusion index (%)76.378.3018.98–100.0087.456.2348.03–100.00

Twelve sets of measurements were obtained for each value (mean of 2/dog). Total hepatic perfusion represents arterial perfusion plus portal venous perfusion. Hepatic perfusion index indicates the percentage of total perfusion attributed to arterial perfusion.

Within a column for a given variable, the result is significantly (P < 0.05) different from that for the liver.

The caudal vena cava was selected as the source of venous input for these calculations.

General perfusion analysis revealed significant differences between the liver and pancreas for perfusion and peak enhancement index values (Table 1). The liver had significantly greater perfusion, compared with the pancreas, on both 3-mm (P < 0.001) and 6-mm (P < 0.001) CT slices. The liver also had significantly greater peak enhancement index on 3-mm (P = 0.002) and 6-mm slices (P < 0.001). No other significant differences in perfusion variables were identified between the liver and pancreas.

Comparison of perfusion values calculated when the caudal vena cava and portal vein were selected as the source of venous input revealed significant differences in hepatic portal perfusion and hepatic perfusion index (Table 2). When the portal vein was selected, the portal perfusion value was significantly greater than that determined when the caudal vena cava was used on both 3-mm (P = 0.05) and 6-mm (P = 0.048) CT slices. When caudal vena cava was selected, the hepatic perfusion index was significantly greater than that calculated when the portal vein was used on 3-mm (P = 0.010) and 6-mm (P = 0.021) slices. No significant difference in hepatic arterial perfusion or total perfusion was present when comparing perfusion values calculated with the caudal vena cava versus the portal vein selected as the source of venous input.

Table 2—

Comparisons of hepatic portal perfusion and hepatic perfusion index measurements with the caudal vena cava or portal vein selected as the source for venous input in Blomley method calculations.

  Portal veinCaudal vena cava
VariableSlice thickness (mm)MedianSERangeMedianSERange
Portal perfusion (mL/min/100 mL)341.216.530.00–171.9010.14*5.970.00–53.05
 645.7716.170.00–184.853.87*5.830.00–58.68
Perfusion index (%)343.948.7910.51–100.0076.37*8.3018.98–100.00
 637.027.9913.37–100.0087.45*6.2348.03–100.00

Within a row, value is significantly (P < 0.05) different from that for the same variable when the portal vein was selected as the input source.

See Table 1 for remainder of key.

Discussion

Results of the present study indicated that CT assessment of pancreatic perfusion is feasible in dogs. Each CT scan took approximately 15 minutes to complete from the time of induction to completion of the scan, and the duration for data evaluation with the software was approximately 5 minutes. Multiple variables including perfusion, peak enhancement index, time to peak enhancement measured from initiation of contrast medium injection, and blood volume were calculated for the pancreas and liver as well as liver-specific variables of arterial perfusion, portal perfusion, total perfusion, and hepatic perfusion index. No dogs had adverse reactions attributed to the procedure.

One of the challenges in developing this protocol was determining the 24-mm section of the abdomen that included the pancreas, liver, and spleen. This was difficult in some dogs owing to the small size of the pancreas on non–contrast-enhanced images. However, each section chosen did include all anatomic features needed for the computer program to calculate the perfusion data. Median perfusion and peak enhancement index each differed significantly between the liver and pancreas, which was consistent with findings in previous studies13,20 of dogs and people. The software-derived hepatic arterial perfusion and pancreatic perfusion values in our study were similar to those reported for dogs without portal vascular anomalies by Zwingenberger and Shofer,13 especially when taking into account the 95% confidence intervals in that study. Differences in the calculation methods used might explain the slight differences between values in the 2 studies.

Calculation of portal perfusion with the Blomley method revealed a significant difference when the caudal vena cava was selected as the venous input for the software, compared with the same measurement when the portal vein was selected. Use of the portal vein data resulted in a significantly higher portal perfusion value. During method evaluation, it became apparent that depending on which slice was used and where the ROI was placed within the vessel, portal perfusion values changed substantially from a range of 0 to > 1,300 mL/min/100 mL. None of the portal perfusion or total hepatic perfusion values obtained with either vessel selected as the venous input source in the study were similar to previously published values.13 Because of this and the wide variability in values obtained when either the portal vein or caudal vena cava was selected, the software-derived values for total hepatic perfusion and portal perfusion may not be reliable or consistently repeatable in dogs.

The significant and substantial differences in data for some liver-specific variables derived by use of measurements that included the different venous inputs in the present study may have reflected anatomic and blood flow differences between dogs and people. The algorithms used by the software were formulated for evaluation of perfusion in human patients and may not necessarily be applicable to other species. Additionally, our study was a preliminary investigation that included a small number of dogs of 1 breed and similar sizes, and variations in weight and size may also need to be considered when using computer software calculations for these variables in dogs. Further studies are needed to evaluate specific perfusion variables in dogs to determine reliability and accuracy, particularly of portal and total hepatic perfusion.

Most multislice CT scanners come with software to calculate hemodynamic data, which increases the feasibility of performing perfusion analyses for veterinary patients. Perfusion assessment by CT is a measurement of parenchymal attenuation of contrast medium and thus, organ perfusion.6 Manufacturers of CT equipment offer distinct software packages that may rely on different mathematical kinetic models to calculate the data, and dissimilarities might make comparison of perfusion parameters across different CT scanners more challenging.12

Maintaining appropriate heart rate and blood pressure is important when evaluating perfusion. The anesthesia service at our facility developed an anesthesia protocol with a drug combination that is widely used in clinical settings and, in their experience, results in less cardiovascular depression than some other drug combinations while allowing the procedure to be performed. Heart rate and arterial blood pressures were maintained within fairly narrow ranges for all dogs as well. Intravenous fluids were administered to all dogs to help maintain blood pressure within acceptable ranges as is common in clinical practice.

Contrast-enhanced CT perfusion analysis of the pancreas can be used in future studies in dogs with suspected pancreatitis to evaluate for microcirculation changes that may develop during pancreatic inflammation. In people, CT perfusion analysis can be used concurrently with standard CT examination to assess for acute pancreatitis.10 With the protocol used in the present study, an entire abdominal precontrast and postcontrast CT scan is performed as well as perfusion analysis. Full abdominal CT scanning with IV contrast medium administration allows clinicians to assess the pancreas and evaluate peripancreatic tissues and vasculature for any sequelae of pancreatitis that may be present.9

There were several limitations to the study reported here. As a preliminary investigation, only a small number of dogs was used, and the group was homogeneous in regard to breed, body weight, age, and sex. Healthy dogs of different breeds, sizes, and ages may have different perfusion values, and further studies will need to be performed to assess applicability of these findings or differences in perfusion values for other groups of dogs. The use of anesthetic drugs and general anesthesia decreases heart rate and blood pressure, which can alter perfusion variables. Currently, most abdominal CT scans in dogs are performed with general anesthesia at our institution, and CT perfusion analysis of sedated dogs could be performed to investigate whether perfusion variables differ significantly under those conditions. However, given that a dog must be motionless for 1 minute to obtain accurate values, heavy sedation would likely be required, and this would likely have some impact on perfusion variables. Another limitation of the present study was the lack of a gold standard for evaluating pancreatic perfusion in dogs with the computer software. However, pancreatic perfusion and hepatic arterial perfusion values generated in our study were similar to the findings in a previous study,13 and manually calculated time-attenuation curves did not appear substantially different from those generated with software-derived calculations. Additionally, the perfusion values were obtained with a 16-slice CT scanner, which limited the data obtained to a 24-mm section of the cranial portion of the abdomen and did not allow for assessment of perfusion throughout the entirety of the pancreas. Use of a multislice CT scanner with a greater number of slices can include more tissue (creating a thicker ROI) for more thorough evaluation and allow these data to be obtained more rapidly.

Assessment of pancreatic perfusion CT was simple to perform and appeared safe in the sample of healthy dogs that we evaluated. However, our findings indicated that human medical software–derived hepatic portal and total perfusion data may not be appropriate for evaluation of liver perfusion in dogs. Additional studies involving dogs of various breeds with and without pancreatic diseases are needed to further elucidate the potential relevance and clinical application of pancreatic perfusion analysis by CT for dogs.

Acknowledgments

No third-party funding or support was received in connection with this study or the writing or publication of this manuscript. The authors declare that there were no conflicts of interest.

The authors thank Anthony Nguyen for technical assistance in acquiring CT images.

ABBREVIATIONS

ROI

Region of interest

Footnotes

a.

Antares Siemen, Siemens Medical Solutions, Mountain View, Calif.

b.

Gemini TF Big Bore System, Philips Medical Systems, Cleveland, Ohio.

c.

Medrad Inc, Whippany, NJ.

d.

Omnipaque, 350 mg I/mL, GE Healthcare Inc, Princeton, NJ.

e.

Philips IntelliSpace Portal, version 8, Lot 8.0.1.20640, Philips Healthcare, Andover, Mass.

f.

Sharon Moen, Clinical Support Engineer, Philips Healthcare, Andover Mass: Personal communication, 2017 and 2018.

g.

SAS, version 9.4, SAS Institute Inc, Cary, NC

References

  • 1. Watson P. Pancreatitis in dogs and cats: definitions and pathophysiology. J Small Anim Pract 2015;56:312.

  • 2. Xenoulis PG. Diagnosis of pancreatitis in dogs and cats. J Small Anim Pract 2015;56:1326.

  • 3. Steiner JM, Newman S, Xenoulis P, et al. Sensitivity of serum markers for pancreatitis in dogs with macroscopic evidence of pancreatitis. Vet Ther 2008;9:263273.

    • Search Google Scholar
    • Export Citation
  • 4. Pieńkowska J, Gwoździewicz K, Skrobisz-Balandowska K, et al. Perfusion-CT—Can we predict acute pancreatitis outcome within the first 24 hours from the onset of symptoms? PLoS One 2016;11:e0146965.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 5. Rademacher N, Schur D, Gaschen F, et al. Contrast-enhanced ultrasonography of the pancreas in healthy dogs and in dogs with acute pancreatitis. Vet Radiol Ultrasound 2016;57:5864.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 6. Almeida RR, Lo GC, Patino M, et al. Advances in pancreatic CT imaging. AJR Am J Roentgenol 2018;211:5266.

  • 7. Balthazar EJ. Acute pancreatitis: assessment of severity with clinical and CT evaluation. Radiology 2002;223:603613.

  • 8. Adrian AM, Twedt DC, Kraft SL, et al. Computed tomographic angiography under sedation in the diagnosis of suspected canine pancreatitis: a pilot study. J Vet Intern Med 2015;29:97103.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 9. French JM, Twedt DC, Rao S, et al. Computed tomographic angiography and ultrasonography in the diagnosis and evaluation of acute pancreatitis in dogs. J Vet Intern Med 2019;33:7988.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 10. Yadav AK, Sharma R, Kandasamy D, et al. Perfusion CT: can it predict the development of pancreatic necrosis in early stage of severe acute pancreatitis? Abdom Imaging 2015;40:488499.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 11. Tsuji Y, Takahashi N, Tsutomu C. Pancreatic perfusion CT in early stage of severe acute pancreatitis. Int J Inflam 2012;497386.

  • 12. Grözinger G, Grözinger A, Horger M. The role of volume perfusion CT in the diagnosis of pathologies of the pancreas. Rofo 2014;186:10821093.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 13. Zwingenberger AL, Shofer FS. Dynamic computed tomographic quantitation of hepatic perfusion in dogs with and without portal vascular anomalies. Am J Vet Res 2007;68:970974.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 14. Kishimoto M, Tsuji Y, Katabami N, et al. Measurement of canine pancreatic perfusion using dynamic computed tomography: influence of input-output vessels on de-convolution and maximum slope methods. Eur J Radiol 2011;77:175181.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 15. Cáceres AV, Zwingenberger AL, Hardam E, et al. Helical computed tomographic angiography of the normal canine pancreas. Vet Radiol Ultrasound 2006;47:270278.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 16. Iseri T, Yamada K, Chijiwa K, et al. Dynamic computed tomography of the pancreas in normal dogs and in a dog with pancreatic insulinoma. Vet Radiol Ultrasound 2007;48:328331.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 17. Blomley MJ, Coulden R, Dawson P, et al. Liver perfusion studied with ultrafast CT. J Comput Assist Tomogr 1995;19:424433.

  • 18. Miles KA, Hayball MP, Dixon AK. Functional images of hepatic perfusion obtained with dynamic CT. Radiology 1993;188:405411.

  • 19. Tsushima Y, Blomley MJ, Kusano S, et al. Measuring portal venous perfusion with contrast-enhanced CT: comparison of direct and indirect methods. Acad Radiol 2002;9:276282.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 20. Motosugi U, Ichikawa T, Sou H, et al. Multi-organ perfusion CT in the abdomen using a 320-detector row CT scanner: preliminary results of perfusion changes in the liver, spleen, and pancreas of cirrhotic patients. Eur J Radiol 2012;81:25332537.

    • Crossref
    • Search Google Scholar
    • Export Citation

Contributor Notes

Address correspondence to Dr. Marolf (Angela.Marolf@colostate.edu).