Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

DOI: 10.1148/radiol.2422041261
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hagspiel, K. D.
Right arrow Articles by Brayman, K. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hagspiel, K. D.
Right arrow Articles by Brayman, K. L.
(Radiology 2007;242:590-599.)
© RSNA, 2007

Vascular and Interventional Radiology

Evaluation of Vascular Complications of Pancreas Transplantation with High-Spatial-Resolution Contrast-enhanced MR Angiography1

Klaus D. Hagspiel, MD, Kiran Nandalur, MD, Timothy L. Pruett, MD, Daniel A. Leung, MD, J. Fritz Angle, MD, David J. Spinosa, MD, Alan H. Matsumoto, MD, Hossam Ahmed, MD, PhD, Hilary A. Sanfey, MD, Robert G. Sawyer, MD, Brian Burkholder, BA, RT(R)(MR) and Kenneth L. Brayman, MD, PhD

1 From the Department of Radiology (K.D.H., K.N., D.A.L., J.F.A., D.J.S., A.H.M., H.A., B.B.) and Division of Transplant Surgery, Department of Surgery (T.L.P., H.A.S., R.G.S., K.L.B.), University of Virginia Health System, PO Box 800170, 1215 Lee St, Charlottesville, VA 22908. Received July 26, 2004; revision requested October 1; revision received March 6, 2006; accepted April 3; final version accepted June 15. Address correspondence to K.D.H. (e-mail: kdh2n{at}virginia.edu).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
Purpose: To retrospectively evaluate high-spatial-resolution contrast material–enhanced three-dimensional (3D) magnetic resonance (MR) angiography for assessment of vascular complications of pancreas allografts.

Materials and Methods: The institutional review board approved the study and waived the requirement for informed patient consent owing to the retrospective nature of the study with use of an anonymous-subject database. The study was HIPAA compliant. The clinical and MR angiography findings in 11 patients (eight men, three women; mean age, 43 years; age range, 30–54 years) who had a history of pancreatic transplant dysfunction and underwent a total of 13 contrast-enhanced 3D MR angiography examinations were retrospectively reviewed. Comparison with conventional angiography findings was possible for four MR angiography examinations, comparison with surgical findings was possible for two examinations, and clinical follow-up was possible for all examinations. Two observers in consensus and blinded to the clinical results performed image analysis of the arterial and venous segments. Classification agreement was assessed with quadratic weighted {kappa} statistics.

Results: Ten MR angiography examinations revealed vascular complications or signs suggestive of rejection. Only three examinations were considered to have completely normal results. All major complications were detected and included complete or partial arterial graft occlusion, stenosis of the arterial Y-graft caused by a kink, complete venous thrombosis, and arteriovenous fistula with pseudoaneurysm formation. For 46 arterial segments and 15 venous segments with angiographic and/or surgical comparison, overall agreement with MR angiography findings was nearly perfect (mean {kappa}, 0.983; standard error of the mean, 0.128).

Conclusion: High-spatial-resolution MR angiography of pancreas allografts enables assessment of the arterial and venous vascular anatomy and can be used to reliably identify clinically relevant vascular complications.

© RSNA, 2007


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
Pancreatic transplantation is a definitive treatment for diabetes. It is a technically challenging procedure, and vascular complications are the most common cause of early transplant failure, with allograft rejection being the second most common cause (1). Despite improvements in surgical technique, vascular complications frequently result in transplant necrosis and necessitate transplant removal. Vascular thrombosis is the most common cause of vascular pancreatic transplant dysfunction (2,3). Arterial and venous thromboses can occur and typically cause pancreatic infarction and/or pancreatitis (4). Other vascular complications include pseudoaneurysms, mycotic aneurysms, arteriovenous fistulas, and complications related to surgical technique. Successful surgical and endovascular treatments have been described for a number of these complications (5–8), but timely diagnosis is essential for graft survival.

The clinical diagnosis of transplant dysfunction is based on the results of laboratory tests, including serum hyperglycemia, but these are relatively insensitive and nonspecific (9,10). A number of imaging modalities, including ultrasonography (US), scintigraphy, and computed tomography (CT), have been used to examine patients suspected of having graft dysfunction. To our knowledge, however, none of these modalities has proved to be consistently reliable and most of them require iodinated contrast material, which is undesirable in kidney transplant patients, to definitively exclude rejection or vascular complications (10–14). The role of magnetic resonance (MR) imaging in this clinical setting has been investigated previously, and it has been shown that dynamic contrast material–enhanced MR imaging enables the detection of acute transplant rejection and can be used to identify allografts with infarction (11).

Previous work involving lower-spatial-resolution MR angiography techniques has included studies of two-dimensional time-of flight and three-dimensional (3D) contrast-enhanced MR angiography sequences (1,4,14), and the results have been generally encouraging. However, technical advances in MR technology have made implementation of breath-hold high-spatial-resolution MR angiography techniques that enable detailed evaluation of the complex vascular allograft anatomy a clinical reality. Thus, the purpose of our study was to retrospectively evaluate breath-hold high-spatial-resolution contrast-enhanced MR angiography for assessment of vascular complications of pancreas allografts.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 References
 
Our institutional review board approved this study and waived the requirement for informed patient consent owing to the retrospective nature of the study with use of an anonymous-subject database. The study was Health Insurance Portability and Accountability Act compliant.

Patient Population
For all patients, the following data were collected as part of the study protocol: demographics, transplant type, clinical abnormalities, time of MR imaging and/or MR angiography in relation to transplantation, MR results, conventional angiography and surgical findings and outcomes, and final diagnosis. Graft function was evaluated if the patient had abnormal blood glucose levels, needed insulin, or had a persistently elevated serum amylase level. Other clinical signs, including rejection, infection, and hemorrhage, also were evaluated. All clinical findings were reviewed by two authors (K.D.H., K.N.).

Eleven consecutive patients (eight men, three women; mean age, 43 years; age range, 30–54 years) with a history of pancreatic allograft dysfunction underwent a total of 13 3D contrast-enhanced MR angiography examinations at our institution between January 2000 and March 2004. The examinations were performed 2 days to 13 years after transplantation. Indications for MR angiography were hyperglycemia in nine cases and one case each of the following abnormalities: abnormal mercaptoacetyltriglycine renal scintigraphy findings, lower gastrointestinal bleeding with a CT finding of pancreatic pseudoaneurysm, fever and malaise with enlarged allograft head at unenhanced CT, and fever with elevated amylase and lipase levels. All 11 patients underwent both kidney and pancreas transplantation, which was performed simultaneously in 10 patients and sequentially in one patient. One patient underwent a second pancreatic transplantation after the first one failed owing to rejection.

Imaging Protocols
All MR examinations were performed with 1.5-T systems and a four-element phased-array body coil. Eleven examinations were performed by using a high-performance (40 mT/m, minimum rise time of 200 µsec) gradient system (Sonata; Siemens Medical Systems, Malvern, Pa) and the following sequences: transverse T1-weighted fast low-angle shot, transverse T2-weighted fast spin echo with fat suppression, and transverse T1-weighted 3D gadolinium-enhanced volumetric interpolated brain examination with fat suppression. Coronal 3D contrast-enhanced MR angiography was performed with a fast low-angle shot sequence by using 3.3/1.2 (repetition time msec/echo time msec), a 25° flip angle, a 390 Hz/pixel bandwidth, a 211 x 512 matrix, and a 6:8 rectangular field of view with a maximum dimension of 400 mm. The resulting voxel size was 0.8 x 1.3 x (1.3–1.8) mm. The imaging duration was 24 seconds.

Two examinations were performed with an older (25 mT/m, minimum rise time of 600 µsec) gradient system (Magnetom Vision; Siemens Medical Systems) and the following sequences: transverse T1-weighted fast low-angle shot, transverse T2-weighted fast spin echo with fat suppression, and T1-weighted two-dimensional fast low-angle shot gadolinium enhanced with fat suppression. Coronal 3D contrast-enhanced MR angiography was performed with a fast imaging with steady-state precession sequence by using 4.6/1.8, a 30° flip angle, a 390 Hz/pixel bandwidth, a 195 x 512 matrix, and a 6:8 rectangular field of view with a maximum dimension of 390 mm in sequential fashion. The resulting voxel size was 0.8 x 1.5 x (1.5–1.9) mm. The imaging duration was 31 seconds.

For MR angiography, 40 mL of gadodiamide (Omniscan; Amersham Health, Princeton, NJ) was injected at 1.7 mL/sec with a power injector (Medrad Spectris; Medrad, Indianola, Pa). An injection delay was calculated to allow peak arterial enhancement during the acquisition of the central lines of k-space for the first acquisition. A timing bolus technique was used for all patients. A minimum of two acquisitions were performed with an interimaging delay of 10 seconds to depict both the arterial and the venous enhancement phases. All examinations were performed by using a breath-hold technique. For the 11 examinations performed with the high-performance gradient system, a subtraction technique was used: A precontrast mask MR angiography acquisition was subtracted from the arterial or venous phase acquisition. Three-dimensional reconstructions were performed on a workstation (Siemens Medical Systems) by using maximum intensity projection and multiplanar projection rendered algorithms.

Image Analysis
Two observers experienced in cardiovascular MR imaging (K.D.H., K.N., 16 and 3 years of experience, respectively) in consensus and blinded to the clinical, surgical, and conventional angiographic results randomly assessed the examination results, including those of MR imaging and MR angiography. These observers then reviewed the medical charts and other relevant clinical and surgical documents to obtain data on patient outcomes. The duration of follow-up was 2–494 days (mean, 67 days).

The visibility and patency of the lumina of the arteries and veins supplying and draining the transplants and their respective anastomoses were assessed on the workstation by using source, multiplanar reformatted, and maximum intensity projection images in a semiquantitative fashion with use of a grading scale: Grade 1 meant completely visualized and patent; 2, completely visualized with 50% or less stenosis or filling defect; 3, completely visualized with greater than 50% stenosis or filling defect; 4, completely visualized with complete occlusion; and 5, not visualized. The following vessels were assessed:

Arteries.—The distal aorta and common, external, and internal iliac arteries were assessed in the transplant recipient. The iliac Y-graft (where applicable), splenic artery, and superior mesenteric artery (SMA), as well as the respective anastomoses, were assessed in the pancreas transplant.

Veins.—The inferior vena cava; common, external, and internal iliac veins; superior mesenteric vein; and portal vein were assessed in the transplant recipient. The mesenteric, splenic, and portal veins, as well as the respective anastomoses, were assessed in the pancreas transplant.

Pancreatic parenchymal enhancement was assessed according to the qualitative criteria proposed by Krebs et al (10): Grade 1 indicated homogeneous normal enhancement; 2, inhomogeneous enhancement; 3, decreased enhancement; and 4, absent enhancement. The first-order intraparenchymal branches of the main allograft arteries and veins were assessed and recorded as visualized or not visualized.

Overall image quality, contrast material bolus timing, and presence of artifacts were assessed qualitatively by using a four-point scale. For image quality, grade 1 meant excellent, with interpretability sufficient for treatment planning; 2, good, with minor image quality impairments and interpretability sufficient for treatment planning; 3, poor, with some image quality impairment but interpretability sufficient for treatment planning and images diagnostic; and 4, nondiagnostic and interpretability not sufficient for treatment planning. For bolus timing, grade 1 meant excellent—that is, pure arterial phase enhancement without venous contamination, enabling depiction of all relevant arterial structures; 2, good—that is, predominant arterial enhancement and minimal venous contamination, enabling identification of all arterial structures; 3, fair—that is, predominant arterial enhancement and substantial venous contamination, enabling identification of all major arterial structures but not the side branches; and 4, poor—that is, poor arterial enhancement and/or severe venous contamination, resulting in nondiagnostic images. For presence of respiratory or motion artifacts, grade 1 meant absent; 2, present but not affecting image interpretation; 3, present and affecting image interpretation; and 4, severe.

Surgery and Conventional Angiography
Five patients had surgical and/or conventional angiography findings for comparison. Three patients underwent conventional angiography only, one underwent surgery only, and one underwent both angiography and surgery.

Surgical evaluation.—Open laparotomy with visual inspection of graft viability and assessment of the patency of the inflow arteries, Y-graft, proximal splenic and superior mesentery transplant arteries, transplant portal vein, and draining systemic or mesentery vein was performed in both the patients. All surgeries were performed by experienced transplant surgeons (K.L.B., T.L.P., each with 11 or more years pancreas transplantation experience).

Angiographic evaluation.—Conventional angiographic evaluation consisted of nonselective pelvic arteriography (A.H.M.) performed in a patient with complete arterial thrombosis and graft necrosis and nonselective pelvic arteriography and superselective transplant arteriography (K.D.H.) performed in three patients. The venous drainage of the transplants in all four patients was evaluated indirectly—that is, by imaging the venous phase after the arterial injection. Selective transplant venography was not performed in any of the four patients. The angiograms were reviewed (K.D.H., K.N.), and the same analysis criteria that were used to assess the contrast-enhanced MR angiograms were applied.

Statistical Analyses
For the arterial and venous segments for which angiographic and/or surgical comparison was possible, classification agreement was assessed with quadratic weighted {kappa} statistics with standard errors by using MedCalc, version 4.16e, for Windows 3.1 (MedCalc Software 2004, Mariakerke, Belgium). The weighted {kappa} statistic is used to estimate the proportion of agreement beyond that expected by chance. Strength of agreement was judged to be slight ({kappa} = 0.00–0.20), fair ({kappa} = 0.21–0.40), moderate ({kappa} = 0.41–0.60), substantial ({kappa} = 0.61–0.80), or almost perfect to perfect ({kappa} = 0.81–1.00), according to an assessment method adapted from Landis and Koch (15). We also used a generalized estimating equation model (SAS Proc GENMOD, SAS 9.1; SAS Institute, Cary, NC) to determine whether agreement existed, with the within-subject correlation taken into account. P ≤ .05 was considered to indicate significance.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 References
 
All clinical and imaging findings documented according to the study protocol are summarized in the Table. MR angiography was technically successful in all 13 examinations. There were no complications.


View this table:
[in this window]
[in a new window]

  		Clinical Characteristics of Patients

 
Image Quality and Visualization of Vascular Segments and Anastomoses
A grade of 1 (excellent) for image quality and bolus timing was assigned in all patients. There were no artifacts at any examinations; all 13 were assigned a grade of 1 (absent). A total of 134 arterial segments and anastomoses and 90 venous segments and anastomoses could be analyzed in the 13 examinations.

Vascular Segments with Surgical and/or Angiographic Comparison
Surgical and/or conventional angiographic comparison was possible for 46 arterial segments and corresponding anastomoses and for 15 venous segments and anastomoses in five examinations. For the arterial segments in these examinations, contrast-enhanced MR angiography correctly depicted 15 segments with complete occlusion (Fig 1), two with grade 2 (<50%) partial stenosis or thrombosis, one with grade 3 (>50%) partial stenosis, and 25 without disease. There was one misclassification, in which stenosis or thrombus was overestimated with MR angiography, and two false results involving underestimated stenosis or thrombosis. These results yielded a mean {kappa} of 0.983 ± 0.147 (standard error of mean) (P < .001). For the 15 venous segments, contrast-enhanced MR angiography correctly depicted four segments with complete thrombosis, one with grade 3 partial thrombosis, and nine without disease. There was only one misclassification—that of overestimated occlusion. These results yielded a mean {kappa} of 0.982 ± 0.258 (P = .001).


Figure 1A
View larger version (104K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1a: Images in 42-year-old man with hyperglycemia in whom US (not shown) on 2nd postoperative day depicted normal pancreatic allograft, with limited detection of intraparenchymal blood vessels but normal resistive indexes. Contrast-enhanced MR angiography (3.3/1.2, 25° flip angle) on 3rd postoperative day depicted acute occlusion of Y-graft off right common iliac artery, with complete thrombosis and infarction of allograft. (a) Maximum intensity projection of left anterior oblique contrast-enhanced MR angiogram shows stump of Y-graft (arrow) and normal kidney transplant on contralateral side. (b) Coronal oblique multiplanar projection of contrast-enhanced MR angiogram shows filling defect in Y-graft (arrow). (c) Findings on left anterior oblique digital subtraction angiography (DSA) image of right iliac artery confirm MR angiography findings: patent iliac arteries and occluded Y-graft (arrow). (d) Coronal source contrast-enhanced MR angiogram shows nonenhancing infarcted pancreas (arrows) with adjacent normally enhancing kidney.

 

Figure 1B
View larger version (124K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1b: Images in 42-year-old man with hyperglycemia in whom US (not shown) on 2nd postoperative day depicted normal pancreatic allograft, with limited detection of intraparenchymal blood vessels but normal resistive indexes. Contrast-enhanced MR angiography (3.3/1.2, 25° flip angle) on 3rd postoperative day depicted acute occlusion of Y-graft off right common iliac artery, with complete thrombosis and infarction of allograft. (a) Maximum intensity projection of left anterior oblique contrast-enhanced MR angiogram shows stump of Y-graft (arrow) and normal kidney transplant on contralateral side. (b) Coronal oblique multiplanar projection of contrast-enhanced MR angiogram shows filling defect in Y-graft (arrow). (c) Findings on left anterior oblique digital subtraction angiography (DSA) image of right iliac artery confirm MR angiography findings: patent iliac arteries and occluded Y-graft (arrow). (d) Coronal source contrast-enhanced MR angiogram shows nonenhancing infarcted pancreas (arrows) with adjacent normally enhancing kidney.

 

Figure 1C
View larger version (150K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1c: Images in 42-year-old man with hyperglycemia in whom US (not shown) on 2nd postoperative day depicted normal pancreatic allograft, with limited detection of intraparenchymal blood vessels but normal resistive indexes. Contrast-enhanced MR angiography (3.3/1.2, 25° flip angle) on 3rd postoperative day depicted acute occlusion of Y-graft off right common iliac artery, with complete thrombosis and infarction of allograft. (a) Maximum intensity projection of left anterior oblique contrast-enhanced MR angiogram shows stump of Y-graft (arrow) and normal kidney transplant on contralateral side. (b) Coronal oblique multiplanar projection of contrast-enhanced MR angiogram shows filling defect in Y-graft (arrow). (c) Findings on left anterior oblique digital subtraction angiography (DSA) image of right iliac artery confirm MR angiography findings: patent iliac arteries and occluded Y-graft (arrow). (d) Coronal source contrast-enhanced MR angiogram shows nonenhancing infarcted pancreas (arrows) with adjacent normally enhancing kidney.

 

Figure 1D
View larger version (118K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1d: Images in 42-year-old man with hyperglycemia in whom US (not shown) on 2nd postoperative day depicted normal pancreatic allograft, with limited detection of intraparenchymal blood vessels but normal resistive indexes. Contrast-enhanced MR angiography (3.3/1.2, 25° flip angle) on 3rd postoperative day depicted acute occlusion of Y-graft off right common iliac artery, with complete thrombosis and infarction of allograft. (a) Maximum intensity projection of left anterior oblique contrast-enhanced MR angiogram shows stump of Y-graft (arrow) and normal kidney transplant on contralateral side. (b) Coronal oblique multiplanar projection of contrast-enhanced MR angiogram shows filling defect in Y-graft (arrow). (c) Findings on left anterior oblique digital subtraction angiography (DSA) image of right iliac artery confirm MR angiography findings: patent iliac arteries and occluded Y-graft (arrow). (d) Coronal source contrast-enhanced MR angiogram shows nonenhancing infarcted pancreas (arrows) with adjacent normally enhancing kidney.

 
For all 61 vascular segments with angiographic and/or surgical comparison, the overall mean {kappa} was 0.983 ± 0.128 (P < .001). After taking into account the within-subject correlation, we found that close agreement still existed (P < .001). In fact, with the assumption of compound symmetry covariance structure in the generalized estimating equation model, the within-subject correlation was close to 0, suggesting that the weighted {kappa} obtained under the independence assumption was a reasonable estimate. Separate analysis of the arterial segments that were misdiagnosed at MR angiography revealed one false case, in which MR angiography depicted grade 2 (<50%) stenosis at a splenic artery anastomosis that was angiographically normal (grade 1). Two other misclassifications involved a splenic artery and a splenic artery anastomosis that were judged to be patent at MR angiography (grade 1) but were angiographically shown to have grade 2 (<50%) stenosis. The incorrect classification of the venous segment represented a left common iliac vein that was judged to be occluded (grade 4) at contrast-enhanced MR angiography but was angiographically found to have grade 3 (>50%) stenosis. Of note, all misclassifications involved hemodynamically insignificant differences in the degree of stenosis, and correct classification would not have affected the clinical management.

Vascular Segments without Surgical and/or Angiographic Comparison
In the eight examinations without comparison data, which revealed a total of 88 arterial and 75 venous segments, all 88 (100%) arterial and all 75 (100%) venous segments were identified. All except 10 arteries were judged to be fully patent (grade 1).

One patient had an occlusive thrombus in the distal end of an allograft SMA and stenosis in the proximal allograft superior mesenteric vein (Fig 2). This patient was treated with anticoagulation on the basis of the MR angiography results and subsequently improved clinically. Another patient who previously had undergone angioplasty owing to a kink had a Y-graft with grade 2 (<50%) stenosis. In another patient with a Y-graft, the SMA limb was completely occluded (grade 4), with occlusion of the distal anastomosis and the allograft SMA. The splenic limb of the Y-graft had 50% stenosis (Fig 3). This patient also received anticoagulation therapy on the basis of the MR angiography results and improved clinically. In three other cases, one patient had complete occlusion (grade 4) of the SMA anastomosis and two had complete SMA occlusion (grade 4). Finally, two patients had findings of grade 2 (<50%) stenosis: In one patient the stenosis involved a Y-graft, and in the other it involved a proximal Y-graft anastomosis.


Figure 2A
View larger version (146K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2a: Three-dimensional contrast-enhanced MR angiograms (3.3/1.2, 25° flip angle) in 54-year-old man with hyperglycemia 12 days after portal enteric drainage allograft placement show complete distal SMA thrombosis and proximal allograft superior mesenteric vein stenosis. (a) Maximum intensity projection of entire MR angiography data set (left anterior oblique projection) shows portal enteric drainage allograft in right iliac fossa (arrow) and kidney transplant (arrowhead) in left iliac fossa. (b) Subvolume maximum intensity projection of same data set (left anterior oblique projection) shows occluded distal allograft SMA (arrow) and splenic artery (arrowhead). (c) Curved multiplanar projection rendered from venous phase MR angiography data set shows thrombus as low-signal-intensity filling defect (arrows) within SMA lumen. (d) Subvolume maximum intensity projection of same venous phase data set shows the stenosed proximal allograft superior mesenteric vein (arrow) in anteroposterior projection. The allograft splenic vein (arrowhead) is patent.

 

Figure 2B
View larger version (108K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2b: Three-dimensional contrast-enhanced MR angiograms (3.3/1.2, 25° flip angle) in 54-year-old man with hyperglycemia 12 days after portal enteric drainage allograft placement show complete distal SMA thrombosis and proximal allograft superior mesenteric vein stenosis. (a) Maximum intensity projection of entire MR angiography data set (left anterior oblique projection) shows portal enteric drainage allograft in right iliac fossa (arrow) and kidney transplant (arrowhead) in left iliac fossa. (b) Subvolume maximum intensity projection of same data set (left anterior oblique projection) shows occluded distal allograft SMA (arrow) and splenic artery (arrowhead). (c) Curved multiplanar projection rendered from venous phase MR angiography data set shows thrombus as low-signal-intensity filling defect (arrows) within SMA lumen. (d) Subvolume maximum intensity projection of same venous phase data set shows the stenosed proximal allograft superior mesenteric vein (arrow) in anteroposterior projection. The allograft splenic vein (arrowhead) is patent.

 

Figure 2C
View larger version (139K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2c: Three-dimensional contrast-enhanced MR angiograms (3.3/1.2, 25° flip angle) in 54-year-old man with hyperglycemia 12 days after portal enteric drainage allograft placement show complete distal SMA thrombosis and proximal allograft superior mesenteric vein stenosis. (a) Maximum intensity projection of entire MR angiography data set (left anterior oblique projection) shows portal enteric drainage allograft in right iliac fossa (arrow) and kidney transplant (arrowhead) in left iliac fossa. (b) Subvolume maximum intensity projection of same data set (left anterior oblique projection) shows occluded distal allograft SMA (arrow) and splenic artery (arrowhead). (c) Curved multiplanar projection rendered from venous phase MR angiography data set shows thrombus as low-signal-intensity filling defect (arrows) within SMA lumen. (d) Subvolume maximum intensity projection of same venous phase data set shows the stenosed proximal allograft superior mesenteric vein (arrow) in anteroposterior projection. The allograft splenic vein (arrowhead) is patent.

 

Figure 2D
View larger version (150K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2d: Three-dimensional contrast-enhanced MR angiograms (3.3/1.2, 25° flip angle) in 54-year-old man with hyperglycemia 12 days after portal enteric drainage allograft placement show complete distal SMA thrombosis and proximal allograft superior mesenteric vein stenosis. (a) Maximum intensity projection of entire MR angiography data set (left anterior oblique projection) shows portal enteric drainage allograft in right iliac fossa (arrow) and kidney transplant (arrowhead) in left iliac fossa. (b) Subvolume maximum intensity projection of same data set (left anterior oblique projection) shows occluded distal allograft SMA (arrow) and splenic artery (arrowhead). (c) Curved multiplanar projection rendered from venous phase MR angiography data set shows thrombus as low-signal-intensity filling defect (arrows) within SMA lumen. (d) Subvolume maximum intensity projection of same venous phase data set shows the stenosed proximal allograft superior mesenteric vein (arrow) in anteroposterior projection. The allograft splenic vein (arrowhead) is patent.

 

Figure 3A
View larger version (137K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3a: Contrast-enhanced MR angiograms (3.3/1.2, 25° flip angle) in 39-year-old woman with hyperglycemia 51 days after pancreas allograft placement show Y-graft with complete occlusion of SMA limb; distal anastomosis and allograft SMA are occluded. Splenic limb of Y-graft has 50% stenosis. (a) Anteroposterior maximum intensity projection of entire MR angiography data set shows systemic enteric drainage allograft (arrow) in right iliac fossa and kidney transplant (arrowhead) in left iliac fossa. (b) Subvolume maximum intensity projection of same arterial phase data set shows occluded SMA limb of Y-graft, occluded SMA anastomosis, and occluded proximal allograft SMA (large arrows). There is also 50% stenosis (arrowhead) in splenic limb of Y-graft (left anterior oblique projection). Note reconstituted peripheral transplant SMA (small arrow) via collateral vessels.

 

Figure 3B
View larger version (103K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3b: Contrast-enhanced MR angiograms (3.3/1.2, 25° flip angle) in 39-year-old woman with hyperglycemia 51 days after pancreas allograft placement show Y-graft with complete occlusion of SMA limb; distal anastomosis and allograft SMA are occluded. Splenic limb of Y-graft has 50% stenosis. (a) Anteroposterior maximum intensity projection of entire MR angiography data set shows systemic enteric drainage allograft (arrow) in right iliac fossa and kidney transplant (arrowhead) in left iliac fossa. (b) Subvolume maximum intensity projection of same arterial phase data set shows occluded SMA limb of Y-graft, occluded SMA anastomosis, and occluded proximal allograft SMA (large arrows). There is also 50% stenosis (arrowhead) in splenic limb of Y-graft (left anterior oblique projection). Note reconstituted peripheral transplant SMA (small arrow) via collateral vessels.

 
Sixty-nine venous segments were judged to be fully patent (grade 1). Patient 7 (Table), in whom previous conventional angiographic comparison was possible, had four venous segments with complete occlusion (grade 4) at follow-up examination. Patient 10 had two stenotic venous segments (one grade 2, one grade 3) and consequently received anticoagulation therapy; he had full recovery of pancreatic function.

Although no surgical or angiographic comparison was possible for these patients, they had either a clinical course of complete normalization of pancreatic function or a benign clinical course that did not warrant intervention, thus appearing to corroborate the 3D contrast-enhanced MR angiography findings. In two patients—one with arterial thrombosis and one with venous stenosis—the decision to treat with anticoagulation was based on the MR angiography findings.

Visualization of Intraparenchymal Branches
The first-order intraparenchymal arterial pancreatic branches could be visualized in 11 (85%) of 13 examinations. The first-order intraparenchymal venous pancreatic branches could be identified in 10 (77%) of 13 examinations. Angiographically, first-order branch comparison was possible for two patients, and the results confirmed the 3D contrast-enhanced MR angiography finding of patency.

Parenchymal Enhancement
Parenchymal enhancement was present in all viable grafts. The two examinations at which enhancement was absent (grade 4) were performed in the two patients with complete graft thrombosis, and in both patients, surgical or angiographic comparison confirmed complete graft thrombosis. Of the 11 examinations that revealed graft enhancement, seven revealed homogeneous complete enhancement (grade 1) and four revealed inhomogeneous enhancement (grade 2). The four examinations that revealed inhomogeneous enhancement were performed in the one patient with an arteriovenous fistula, mycotic pseudoaneurysm, and graft infection; in two patients with a clinical diagnosis of mild graft rejection that responded to immunosuppressive therapy; and in one patient with partial arterial and complete venous allograft thrombosis.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
References
 
Pancreatic transplantation is increasingly being used to treat type 1 diabetes mellitus (16). From a vascular standpoint, two types of procedures exist: The original procedure involves intraperitoneal placement of the allograft in the pelvis and an anastomosis between the transplant splenic artery and SMA and the recipient's iliac arteries by way of a donor Y-graft, with pancreatic venous outflow typically into the common or external iliac vein and thus the systemic circulation. A modified form of pancreatic transplantation, reintroduced in 1995, is commonly referred to as portoenterically drained pancreas transplantation. As in the standard technique, in the portoenteric drainage technique the pancreatic allograft is placed intraperitoneally but higher in the recipient's abdomen. The pancreatic arteries, splenic arteries, and SMAs are anastomosed to the iliac arteries or aorta as well but owing to the higher position in the abdomen by way of a longer Y-graft formed from the donor's common, internal, and external iliac arteries. Pancreatic venous outflow is achieved by anastomosing the transplant portal vein with the recipient's portal venous system, typically the superior mesenteric vein and less commonly the splenic vein (17–19). This fairly complex vascular anatomy and the small size of the vessels make imaging technically very challenging.

Although conventional angiography is generally considered the reference standard for evaluation of the vascular anatomy of pancreatic allografts (20), it is invasive and involves the use of potentially nephrotoxic iodinated contrast agents. Postoperative pancreatic allograft dysfunction is not infrequent. Vascular thrombosis is the most common cause of pancreatic allograft dysfunction, with a reported incidence of 2%–19%, followed by rejection (21). Arterial and venous thromboses occur, as do inflow vessel occlusions, owing to surgical clamp injuries or preexisting atherosclerotic disease, pseudoaneurysms, and arteriovenous fistulas. Surgical and/or endovascular treatment of these conditions is sometimes possible if the diagnosis is made early (5,7,8).

Since clinical and laboratory tests are unreliable and nonspecific in the setting of vascular complications, imaging is needed to detect and delineate them (10,17). US is the key imaging modality for initial evaluation of pancreas allografts. Color Doppler US and power Doppler US have been shown to enable detection of the allograft vasculature and vascular complications such as thrombosis, stenosis, and pseudoaneurysm (22,23). Resistive indexes have been recorded from the intraparenchymal pancreatic vessels, but there is a large overlap of resistive indexes between normal and diseased allografts. In addition, it is often difficult to differentiate the pancreatic tissue from the surrounding tissues, and because of the intraperitoneal position of the pancreas, overlying gas-filled loops of bowel can make visualization impossible (24).

In some centers, US-guided allograft biopsy is used with good success to diagnose rejection, but the procedure is not without risk and yields no information about vascular complications unless the organ is infarcted (25). If US fails, CT is often used. CT is particularly well suited to depict perigraft fluid collections (11). When performed with modern multidetector scanners and iodinated contrast agents, CT has been shown to enable evaluation of the transplanted pancreas and assessment of vascular complications such as thrombosis or stenosis of the arterial graft or pancreatic infarction (26,27). The major disadvantage of CT angiography, as compared with MR angiography, is the need to use iodinated contrast agents, which frequently are contraindicated in patients with kidney and pancreas transplants owing to renal insufficiency and the radiation dose, especially with multiphasic CT protocols.

Scintigraphic techniques also have been used to evaluate pancreas allografts. A number of different radiopharmaceutical agents have been assessed, and scintigraphy generally is considered a helpful but nonspecific tool for diagnosing pancreatic dysfunction. However, once this diagnosis has been made, other imaging modalities generally are required to determine the specific cause (28,29).

For the above reasons, there has been marked interest in MR imaging and MR angiography as primary modalities for evaluating malfunctioning pancreas transplants (1,4,9,10,14,20,30,31). In addition, the gadolinium-based contrast agents used for MR imaging are only minimally nephrotoxic (32). Dynamic contrast-enhanced MR imaging performed with a T1-weighted fast spoiled gradient-echo sequence after the intravenous bolus injection of gadopentetate dimeglumine has been shown to be highly sensitive for the detection of acute pancreatic transplant rejection and enables clear identification of allograft infarction (10). The role of MR angiography in the detection of vascular complications has been investigated in a small number of studies. In a previous study involving two-dimensional time-of flight MR sequences, the absence of flow-related enhancement in the main vascular pedicle in four patients correlated with venous or arterial thrombosis detected at surgery (4). The use of two-dimensional time-of-flight MR angiography has been abandoned clinically owing to the advent of newer contrast-enhanced MR angiography techniques that enable evaluation of the vasculature in less time and with greater detail.

To our knowledge, investigators in only four studies (1,5,14,20) have reported on the value of contrast-enhanced 3D MR angiography in the assessment of vascular complications in patients with pancreas transplants. One group prospectively compared contrast-enhanced MR angiography with DSA in seven patients with pancreas and kidney transplants and found MR angiography to be suitable for evaluation of the vascular anatomy (20). The aorta, inferior vena cava, and main arterial and venous branches of the transplant were reliably depicted. In the 21 pancreas transplant vessels in these seven patients (seven iliac grafts, seven splenic arteries, and seven mesenteric arteries), seven occlusions were depicted by both DSA and contrast-enhanced MR angiography. Only those grafts with both arteries occluded showed no enhancement, whereas the five patients with at least one patent artery to the allograft had homogeneous parenchymal enhancement. In addition, these investigators found depiction of the relevant venous anatomy with contrast-enhanced MR angiography to be superior to depiction with DSA. They concluded that contrast-enhanced MR angiography has advantages over DSA and could replace it in patients who have impaired graft function or are suspected of having vascular complications (20). The spatial resolution achieved by these authors was fairly low, with a voxel size of 8.1 mm3, and the number of vascular complications in their series was small; however, the diagnostic potential of MR angiography in this setting was established.

Eubank et al (1) reported the successful detection of venous allograft thrombosis in five patients with contrast-enhanced MR angiography, but they used an older, very-low-spatial-resolution technique (voxel size > 10 mm3). Venous thrombus appeared as serpentine voids within the graft parenchyma or at the venous anastomosis during the venous phase. A 3-cm pseudoaneurysm associated with arteriovenous fistula was reported in a case report (5)—again, the investigators used a lower-spatial-resolution technique. Huber et al (14) published their experience in using a 3D contrast-enhanced MR angiography technique in eight patients. They achieved a voxel size of 6 mm3, which is more than four times larger than the voxel size achieved in our current work, 1.3 mm3. They had angiographic comparison in two patients, which confirmed the MR angiography findings. These authors used venous phase imaging that was not optimized and consequently visualized the larger allograft veins in only 50% of their patients (14).

All of these data were suggestive that the development of an optimized pancreas allograft protocol and implementation on a high-performance cardiovascular MR imager for routine clinical cases would likely be of diagnostic and clinical benefit. To our knowledge, our contrast-enhanced MR angiography protocol yields higher spatial resolution than all previously investigated techniques. Used in conjunction with MR imaging, this technique enables comprehensive assessment of allografts. The relevant vascular anatomy, the majority of vascular complications, and—as confirmed by our results—all clinically relevant complications can be detected. As observed by others, dynamic gadolinium-enhanced MR imaging can be used to diagnose rejection with reasonable accuracy, and this examination has been implemented into our MR angiography protocol (10).

Our study had several limitations: The sample size was relatively small, and there was potential referral bias due to the retrospective nature of the study. Furthermore, a limited number of patients had angiographic and/or surgical findings for comparison, and long-term clinical follow-up data were missing. Nevertheless, our pancreas transplant surgeons believe that the described MR angiography protocol facilitates the treatment of pancreas transplant recipients, and presently we rely heavily on MR angiography for assessment of potential vascular complications in these patients.

In conclusion, high-spatial-resolution breath-hold 3D contrast-enhanced MR angiography evaluation of the pancreas is very useful for assessing vascular complications of pancreas transplantation. Where available, this examination should be the secondary modality when the findings of US, which is more readily available and less expensive, are inconclusive.


   FOOTNOTES
 

Abbreviations: DSA = digital subtraction angiography • SMA = superior mesenteric artery • 3D = three-dimensional

Authors stated no financial relationship to disclose.

Author contributions: Guarantor of integrity of entire study, K.D.H.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, K.D.H., K.N.; clinical studies, K.D.H., K.N., T.L.P., D.A.L., H.A.S., R.G.S., B.B., K.L.B.; statistical analysis, K.D.H., K.N.; and manuscript editing, K.D.H., K.N., H.A., B.B.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 

  1. Eubank WB, Schmiedl UP, Levy AE, Marsh CL. Venous thrombosis and occlusion after pancreas transplantation: evaluation with breath-hold gadolinium-enhanced three-dimensional MR imaging. AJR Am J Roentgenol 2000;175:381–385.[Abstract/Free Full Text]
  2. Grewal HP, Garland L, Novalk K, et al. Risk factors for postimplantation pancreatitis and pancreatic thrombosis in pancreas transplant recipients. Transplantation 1993;56:609–612.[Medline]
  3. Douzdjian V, Abecassis MM, Cooper JL, et al. Pancreas transplant salvage after acute venous thrombosis. Transplantation 1993;56:222–223.[CrossRef][Medline]
  4. Krebs TL, Daly B, Wong JJ, Chow CC, Bartlett ST. Vascular complications of pancreatic transplantation: MR evaluation. Radiology 1995;196:793–798.[Abstract/Free Full Text]
  5. Orsenigo E, De Corbelli F, Salvioni M, et al. Successful endovascular treatment for gastroduodenal artery pseudoaneurysm with an arteriovenous fistula after pancreas transplantation. Transpl Int 2003;16:694–696.[CrossRef][Medline]
  6. Woo EY, Milner R, Brayman KL, Fairman RM. Successful PTA and stenting for acute iliac arterial injury following pancreas transplantation. Am J Transplant 2003;3:85–87.[CrossRef][Medline]
  7. Tan M, Di Carlo A, Stein A, Cantarovick M, Tchervenkov JI, Metrakos P. Pseudoaneurysm of the superior mesenteric artery after pancreas transplantation treated by endovascular stenting. Transplantation 2001;72:336–338.[CrossRef][Medline]
  8. Tzakis AG, Carroll PB, Gordon RD, Yokoyama I, Makowka L, Starzl TE. Arterial mycotic aneurysm and rupture: a potentially fatal complication of pancreas transplantation in diabetes mellitus. Arch Surg 1989;124:660–661.[Abstract]
  9. Yuh WT, Hunsicker LG, Nghiem DD, et al. Pancreatic transplants: evaluation with MR imaging. Radiology 1989;170:171–177.[Abstract/Free Full Text]
  10. Krebs TL, Daly B, Wong-You-Cheong JJ, Carroll K, Bartlett ST. Acute pancreatic transplant rejection: evaluation with dynamic contrast-enhanced MR imaging compared with histopathologic analysis. Radiology 1999;210:437–442.[Abstract/Free Full Text]
  11. Moulton JS, Munda R, Weiss MA, Lubbers DJ. Pancreatic transplants: CT with clinical and pathologic correlation. Radiology 1989;172:21–26.[Abstract/Free Full Text]
  12. Wong JJ, Krebs TL, Klassen DK, et al. Sonographic evaluation of acute pancreatic transplant rejection: morphology-Doppler analysis versus guided percutaneous biopsy. AJR Am J Roentgenol 1996;166:803–807.[Abstract/Free Full Text]
  13. Nelson NL, Largen PS, Stratta RJ, et al. Pancreas allograft rejection: correlation of transduodenal core biopsy with Doppler resistive index. Radiology 1996;200:91–94.[Abstract/Free Full Text]
  14. Huber A, Holzknecht N, Heuck A, Stangl M, Reiser M. Erste erfahrungen mit der kontrastverstärkten MR-angiographie nach nieren- und pankreastransplantation. Radiologe 1997;37:239–242.[CrossRef][Medline]
  15. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159–174.[CrossRef][Medline]
  16. Neri E, Cappelli C, Boggi U, Salvatori L, et al. Multirow CT in the follow-up of pancreas transplantation. Transplant Proc 2004;36:597–600.[CrossRef][Medline]
  17. Heyneman LE, Keogan MT, Tuttle-Newhall JE, et al. Pancreatic transplantation using portal venous and enteric drainage: the postoperative appearance of a new surgical procedure. J Comput Assist Tomogr 1999;23:283–290.[CrossRef][Medline]
  18. Gaber AO, Shokouh-Amiri MH, Hathaway DK, et al. Results of pancreas transplantation with portal venous and enteric drainage. Ann Surg 1995;221:613–614.[Medline]
  19. Bloom RD, Olivares M, Rehman L, Raja RM, Yang S, Badosa F. Long-term pancreas allograft outcome in simultaneous pancreas-kidney transplantation: a comparison of enteric and bladder drainage. Transplantation 1997;64:1689–1695.[CrossRef][Medline]
  20. Boeve WJ, Kok T, Tegzess AM, et al. Comparison of contrast enhanced MR-angiography-MRI and digital subtraction angiography in the evaluation of pancreas and/or kidney transplantation patients: initial experience. Magn Reson Imaging 2001;19:595–607.[CrossRef][Medline]
  21. Fernandez MP, Bernardino ME, Neylan JF, Olson RA. Diagnosis of pancreatic transplant dysfunction: value of gadopentetate dimeglumine-enhanced MR imaging. AJR Am J Roentgenol 1991;156:1171–1176.[Abstract/Free Full Text]
  22. Foshager MC, Hedlund LJ, Troppmann C, Beneditti E, Gruessner RW. Venous thrombosis of pancreatic transplants: diagnosis of duplex sonography. AJR Am J Roentgenol 1997;169:1269–1273.[Abstract/Free Full Text]
  23. Tobben PJ, Zajko AB, Sumkin JH, et al. Pseudoaneurysms complicating organ transplantation: roles of CT, duplex sonography, and angiography. Radiology 1988;169:65–70.[Abstract/Free Full Text]
  24. Nikolaidis P, Amin RS, Hwang CM, et al. Role of sonography in pancreatic transplantation. RadioGraphics 2003;23:939–949.[Abstract/Free Full Text]
  25. Klassen DK, Weir MR, Cangro CB, Bartlett ST, Papadimitriou JC, Drachenberg CB. Pancreas allograft biopsy: safety of percutaneous biopsy—results of a large experience. Transplantation 2002;73:553–555.[CrossRef][Medline]
  26. Cappelli C, Neri E, Boggi U, et al. Pancreas transplantation: multislice computed tomography follow-up. Radiol Med (Torino) 2003;106:191–200.[Medline]
  27. Eschwege P, Rocher L, Rondeau Y, et al. Complications of pancreas transplantation: radiological evaluation. Transplant Proc 1998;30:2795.
  28. van der Hem LG, van der Linden CJ, Ticheler CH, Hoitsma AJ, Corstens FH. Early detection of post-transplant pancreatic graft dysfunction with technetium-99m-HMPAO scintigraphy. J Nucl Med 1994;35:1488–1490.[Abstract/Free Full Text]
  29. Lee HK, Chung DH, Jung J, et al. Three cases of pancreas allograft dysfunction. J Korean Med Sci 2000;15:105–110.[Medline]
  30. Contis JC, O'Connor TP, Holland GA, et al. Noninvasive evaluation of bladder-drained whole pancreaticoduodenal transplants with magnetic resonance angiography. Transplant Proc 1994;26:464–465.[Medline]
  31. Vahey TN, Glazer GM, Francis IR, et al. MR diagnosis of pancreatic transplant rejection. AJR Am J Roentgenol 1988;150:557–560.[Abstract/Free Full Text]
  32. Niendorf HP, Haustein J, Louton T, Beck W, Laniado M. Safety and tolerance after intravenous administration of 0.3 mmol/kg Gd-DTPA: results of a randomized, controlled clinical trial. Invest Radiol 1991;26(suppl 1):S221–S223.[Medline]




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hagspiel, K. D.
Right arrow Articles by Brayman, K. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hagspiel, K. D.
Right arrow Articles by Brayman, K. L.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE