| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Gastrointestinal Imaging |
1 Departments of Diagnostic Radiology (T.L.K., B.D., J.J.W.Y.C., K.C.)
2 Surgery (S.T.B.), University of Maryland School of Medicine, 22 S Greene St, Baltimore, MD 21201.
 |
Abstract |
|---|
 TOP Abstract IntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: Thirty MR imaging studies were performed in 25 patients within 3 days of percutaneous biopsy or pancreatectomy. The mean percentage of parenchymal enhancement (MPPE) at dynamic contrast-enhanced MR imaging was calculated.
RESULTS: Biopsy findings were no evidence of rejection (n = 7 [23%]), mild rejection (n = 10 [33%]), moderate (n = 6 [20%]) and severe (n = 2 [7%]) acute rejection, and infarction (n = 5 [17%]). The corresponding MPPEs at 1 minute were 106%, 66%, 62%, 57%, and 3%, respectively. Overlap of cases in the normal and rejection groups occurred; however, using an MPPE cutoff of 100% resulted in a sensitivity of 96%. An MPPE over 120% was seen in the normal group only. The MPPE was significantly greater in the normal group than in the rejection or infarction group (P < .05).
CONCLUSION: Dynamic contrast-enhanced MR imaging is highly sensitive for the detection of acute pancreatic transplant rejection. Because of overlap of cases in the normal and rejection groups, percutaneous biopsy may be needed in some cases. Pancreatic allografts with infarction can be clearly identified.
Index terms: Magnetic resonance (MR), contrast enhancement, 77.121412, 77.12143 • Magnetic resonance (MR), tissue characterization, 77.121412, 77.12143 • Pancreas, MR, 77.121412, 77.12143, 77.12144, 77.12146 • Pancreas, transplantation, 77.458
|
Introduction |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Spin-echo MR imaging and gadopentetate dimeglumine–enhanced MR imaging have been inconsistent in the evaluation of pancreatic transplant dysfunction (7–9). In these series, only a few cases were correlated with histopathologic results; most of the acute rejection cases were diagnosed by using clinical or biochemical evaluation. Few pancreatic biopsy procedures were performed in prior studies because of the perceived risk of complications and the need for general anesthesia and open biopsy (10). Recent developments in percutaneous technique now enable safe and consistent biopsy of pancreatic transplants (5,11–14) and thereby enable standard-of-reference correlation with imaging findings. We analyzed the usefulness of dynamic contrast material–enhanced MR imaging in the diagnosis of acute rejection by correlating MR imaging enhancement patterns with histopathologic findings obtained at imaging-guided percutaneous core biopsy or surgical explantation.
|
MATERIALS AND METHODS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The surgical technique used for transplantation in all cases was intraperitoneal placement of a cadaveric whole-organ allograft and a portion of the duodenal C loop into the iliac fossa. Exocrine secretions were drained by either creating a pancreaticoduodenocystostomy whereby the donor duodenum was connected to the superior aspect of the recipient's bladder ("bladder drained") or draining the small intestine where the donor duodenum was anastomosed to a loop of recipient jejunum ("enteric drained"). All but four studies (in three patients) involved bladder-drained allografts.
Histopathologic Correlation
Correlation was made with the results of imaging-guided percutaneous pancreatic biopsy, which was performed in 24 cases (22 US guided, two CT guided) within 3 days after the MR imaging studies (mean, 1.4 days) by using a previously described technique (5,11–14). Three patients underwent biopsy within 2 weeks after transplantation, and most biopsy procedures were performed within 4 months after surgery (range, 10 days to 44.3 months; mean, 3.9 months). Indications for percutaneous pancreatic biopsy were a twofold increase in the serum amylase or lipase level, a sustained (ie, 40% or greater) decrease in the urinary amylase level, or clinical features suggestive of acute rejection such as fever, graft tenderness, or abdominal distention.
Patients with biochemical abnormalities that improved after urinary catheterization of bladder-drained allografts were considered to have reflux pancreatitis and thus did not undergo biopsy and were not included in this series. The MR imaging findings were correlated with the results of analyses of pancreatic specimens removed at surgery in six cases, with a mean time from imaging to surgery of 2.2 days. One patient underwent explantation within a week after transplantation. Most of the remainder of the explanted allografts were older than 1 year (range, 5 days to 50.6 months; mean, 18.3 months). Pancreatic explantation was performed because of unremitting abdominal pain, severe hyperglycemia that required insulin, imaging findings suggestive of infarction, or all three of these indications.
Biopsy results were evaluated by experienced histopathologists. The severity of acute rejection, if present, was graded as mild, moderate, or severe by using a qualitative analysis system that has been demonstrated to correlate with graft survival (15). This grading scale was based on the degree of vascular, septal, and acinar inflammatory changes (15). In addition, the presence of infarction or vascular thrombosis was noted. Biopsy specimens that did not have pancreatic parenchyma were not included in the study.
MR Imaging and Analysis
All MR imaging studies were acquired on a 1.5-T system (Signa, GE Medical Systems, Milwaukee, Wis) with a phased-array coil. Initially, a fast multiplanar spoiled gradient-recalled-echo (GRE) MR image in the sagittal plane (20/1.2 [TR msec/TE msec], 60° flip angle, 256 x 128 matrix, two signals averaged, 10-mm section thickness without gap) was obtained to determine the area of coverage. Before and immediately after injection of the contrast material, a coronal 23-second, breath-hold T1-weighted fast multiplanar spoiled GRE image (150/4.2, 60° flip angle, 256 x 128 matrix, 5-mm section thickness without gap, 24-cm field of view) was acquired, and imaging was repeated every minute for 5 minutes. Thirteen sections per breath hold were obtained; this permitted coverage of the entire pancreatic allograft. After written informed consent was obtained, an intravenous bolus injection (0.1 mmol per kilogram of body weight injected for 5–10 seconds) of gadopentetate dimeglumine (Magnevist; Berlex Laboratories, Wayne, NJ) followed by a saline flush was administered through a 21-gauge needle. These two sequences required approximately 15 minutes of imaging time. In addition, axial T2-weighted chemical fat-suppressed fast spin-echo MR imaging and flow-sensitive MR angiography were performed, but these images were not evaluated for the purposes of this study.
During the unenhanced breath-hold study, a mean glandular signal intensity for each pancreatic transplant was derived by averaging two 1-cm-diameter regions of interest from a representative image section that encompassed the middle portion of the tail and head of the transplanted pancreas. The representative areas were selected by one investigator (K.C.) without knowledge of the results of histopathologic analysis. Pancreatic allograft enhancement may be heterogeneous, and focal areas of signal intensity abnormality (eg, intraparenchymal fluid collection) were excluded from sampling. The same regions of interest were measured during the subsequent contrast enhancement portion of the study. In most cases, a single section was adequate for analyzing both the head and tail of the transplanted pancreas. These measurements were used to calculate a mean percentage of parenchymal enhancement (MPPE) for each breath-hold gadolinium-enhanced GRE study by using the following formula: MPPE = (contrast-enhanced MGSI - unenhanced MGSI) x 100%/unenhanced MGSI, where MGSI is the mean glandular signal intensity. A time-enhancement curve was created by using these values plotted against the histopathologic findings.
Statistical Analyses
An unpaired Student t test was performed with the MPPE measurements during the 1-minute breath-hold (90 seconds after the bolus injection) portion of the gadolinium-enhanced MR imaging study, which was stratified by using histopathologic subgroups and Excel version 7 (Microsoft, Redmond, Wash). A statistically significant difference in means was considered to be present when P was less than .05. Sensitivity, specificity, and accuracy were calculated with standardized formulas.
|
RESULTS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Comparison of the time-MR enhancement curves with the histopathologic findings revealed three basic patterns (Fig 1). The average MPPE in the normal group was greater for each time measurement than in all other histopathologic groups. A peak mean enhancement of 106% occurred at 1 minute, with a subsequent minimal decrease over time. For each period, the average MPPE in the rejection subgroups was similar between these groups; it remained well below the average MPPE in the normal group and never exceeded 67%. Enhancement in mild and moderate grades of rejection peaked at 1 minute and subsequently diminished. The MPPE in the severe rejection group rose slightly over time. Transplanted organs that demonstrated infarction had an average MPPE that was well below that in the rejection and normal subgroups and remained less than 6% for all periods.
|
|
|
DISCUSSION |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Experience with contrast-enhanced MR imaging for assessment of pancreatic allograft rejection has been limited. Fernandez et al (9) obtained dynamic, single-section contrast-enhanced MR images through the pancreatic transplant during breath holding. They demonstrated that the percentage of enhancement in normally functioning pancreatic transplants was greater than that in dysfunctional allografts (with either rejection or infarction). In six cases of normally functioning allografts, they recorded a mean percentage of enhancement at 1 minute of 98% compared with 42% in six cases with acute dysfunction. In their series, the findings of five MR imaging studies were correlated with pancreatic transplant biopsy results.
Although many transplantation centers now use percutaneous pancreatic transplant biopsy to reliably diagnose acute rejection (11,13,18), this previously was not the case. Acute pancreatic rejection was thought to occur synchronously with acute renal rejection in simultaneously transplanted pancreas and kidney allografts from the same donor, and pancreatic rejection was not thought to occur in isolation (19). Because renal rejection could easily be diagnosed by using serum creatinine levels and percutaneous renal biopsy, the accepted practice was to use renal rejection to determine pancreatic rejection (19–22). However, it has been demonstrated that episodes of acute rejection after simultaneous renal and pancreatic transplantation may be discordant—that is, may occur in one organ but not in the other—in 22%–47% of cases (23,24). A sensitive noninvasive imaging study would be preferable because it would enable the small but substantial (ie, up to 3%) risk of major complications after percutaneous pancreatic biopsy to be avoided (5,14).
In our study, we compared the results of dynamic breath-hold contrast-enhanced pancreatic allograft MR imaging with those of standard-of-reference histopathologic analysis. Normally functioning transplanted organs tended to enhance avidly with contrast material administration, whereas the transplanted organs with rejection enhanced less actively. At 1 minute, the percentage of enhancement in the normal group was substantially greater, 106%, compared with 50% in the dysfunctional group (ie, transplants with rejection and infarction); these results are similar to those of the study by Fernandez et al (9). However, in their study, dysfunctional transplanted organs were not separated into rejection versus infarction groups, presumably because of the limited number of cases. In our study, the MPPE in the transplanted organs with rejection at 1 minute (63%) was markedly different from that in the transplanted organs with infarction (3%).
One of the mechanisms of dysfunction that occurs with acute pancreatic rejection is an alloimmune vasculitis (25). Decreased contrast enhancement in the transplanted organ, as occurred in the allografts with acute rejection in our study, may be due to narrowed or occluded small vessels, which result in a decreased rate and degree of accumulation of extracellular contrast material. For this purpose, gadolinium-based extracellular contrast materials for MR imaging have an advantage over the iodinated contrast materials that are used for CT, because most recipients have a coexistent renal transplant, and there is a reduced risk of nephrotoxicity.
The difference in enhancement between normal transplanted organs and dysfunctional allografts on gadolinium-enhanced GRE MR images may be useful in the clinical assessment and treatment of allograft recipients and in helping to guide the selective use of biopsy. In our study, despite the overlap of MPPE data points in the normal and rejection groups, the MPPE in the normal group was significantly different from that in the rejection group (P < .05). Gadolinium-enhanced GRE MR imaging appears to be highly sensitive for the detection of rejection; an MPPE cutoff of 100% resulted in a sensitivity of 96% (Figs 3, 4). In our series, the one false-negative case was due to infarction in a transplanted organ without acute rejection. Failure to diagnose this case by using MR imaging was not clinically relevant, because the allograft needed to be removed regardless of whether acute rejection was present. The two histopathologically normal false-positive cases, which occurred in transplanted organs that had diminished enhancement, were more problematic. If MR imaging had been used as the sole guide for antirejection treatment instead of biopsy, unnecessary therapy may have been instituted. These allografts may have had areas of acute rejection that were not detected because of biopsy sampling error. An MPPE of greater than 120% was demonstrated in normally functioning allografts only; this indicates that this group may not require biopsy. There was a very small number of such cases in our series. The converse may also be useful; allografts with an MPPE of less than 60% were dysfunctional owing to either acute rejection or infarction.
|
|
|
Graft thrombosis that leads to allograft infarction is another major cause of graft loss. This event can occur early (ie, in less than 1 month) or late after surgery. Early graft thrombosis, as occurred in one patient in this series, is usually the result of vascular graft anastomotic error or microvascular damage from preservation injury. Late thrombosis (ie, that which occurs more than a month after surgery), as occurred in four cases in this series, usually results from severe acute rejection with alloimmune arteritis and causes occlusion of small vessels, which leads to complete major vessel occlusion. Regardless of the causes, all cases with an MPPE of less than 10% in our study had infarction and were identified by using gadolinium-enhanced GRE MR imaging (Fig 5). Identification of total graft thrombosis and infarction is important, because the transplanted organ should be removed immediately to avoid the severe systemic effects of graft autolysis (21). Acute rejection, regardless of its severity, is usually treated medically, and the allograft usually does not need to be removed; however, it may be difficult to differentiate severe rejection from graft infarction in the clinical setting. In our study, there was no overlap of cases between the rejection subgroups and the infarction group. Therefore, the described technique appears to be useful in separating these two diagnoses. This supports the findings in previous MR imaging literature (27), which have demonstrated a lack of enhancement in pancreatic allografts with infarction in a small number of cases.
|
In summary, we correlated dynamic contrast-enhanced GRE MR imaging findings in pancreas allografts with histopathologic results, as determined by using pancreatic biopsy specimens, and found significant differences in the enhancement patterns. The mean percentage of enhancement in normal pancreatic transplants on dynamic MR imaging studies at 1 minute was 104%, which was significantly greater than the 67% and 3% in the pancreatic transplants that had acute rejection and infarction, respectively. Transplanted organs with infarction can be clearly differentiated from viable allografts for prompt surgical intervention. Despite the overlap of acute rejection and normal cases, decreased enhancement on MR images appears to be highly sensitive for rejection. Such imaging findings in the appropriate clinical setting and/or with associated biochemical abnormalities may allow greater confidence in clinical decision making. By using an MPPE cutoff of 100%, gadolinium-enhanced GRE MR imaging was 96% sensitive for the detection of acute rejection. Biopsy may not be required in transplanted organs with an MPPE of greater than 120%. For clinically indeterminate cases, in particular those with an MPPE of 100%–120%, biopsy may be required. Sensitive detection of acute rejection is likely to be important in portal venous–drained pancreatic transplants, in which percutaneous biopsy cannot be readily performed.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
From the 1997 RSNA scientific assembly.
Abbreviations: GRE = gradient-recalled echo MPPE = mean percentage of parenchymal enhancement
Author contributions: Guarantor of integrity of entire study, T.L.K.; study concepts and design, T.L.K., B.D., J.J.W.Y.C.; definition of intellectual content, T.L.K., B.D., J.J.W.Y.C.; literature research, T.L.K.; clinical studies, T.L.K., B.D., J.J.W.Y.C., K.C., S.T.B.; data acquisition, K.C.; data analysis, T.L.K., B.D., J.J.W.Y.C., K.C.; statistical analysis, T.L.K., B.D., J.J.W.Y.C.; manuscript preparation, T.L.K.; manuscript editing, B.D., J.J.W.Y.C., S.T.B.; manuscript review, T.L.K., B.D., J.J.W.Y.C., S.T.B.
Received March 30, 1998;
revision requested June 17, 1998; revision received July 24, 1998;
accepted September 28, 1998.
|
References |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. D. Hagspiel, K. Nandalur, T. L. Pruett, D. A. Leung, J. F. Angle, D. J. Spinosa, A. H. Matsumoto, H. Ahmed, H. A. Sanfey, R. G. Sawyer, et al. Evaluation of Vascular Complications of Pancreas Transplantation with High-Spatial-Resolution Contrast-enhanced MR Angiography Radiology, February 1, 2007; 242(2): 590 - 599. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Dobos, D. A. Roberts, E. K. Insko, E. S. Siegelman, A. Naji, and J. F. Markmann Contrast-enhanced MR Angiography for Evaluation of Vascular Complications of the Pancreatic Transplant RadioGraphics, May 1, 2005; 25(3): 687 - 695. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. D. Hagspiel, K. Nandalur, B. Burkholder, J. F. Angle, K. L. Brayman, D. J. Spinosa, A. H. Matsumoto, O. L. Veldhuis, H. Sanfey, R. G. Sawyer, et al. Contrast-Enhanced MR Angiography After Pancreas Transplantation: Normal Appearance and Vascular Complications Am. J. Roentgenol., February 1, 2005; 184(2): 465 - 473. [Full Text] [PDF]
|
|
|
|
|
|
J. T. Heverhagen, H.-J. Wagner, H. Ebel, A. L. Levine, K. J. Klose, and A. Hellinger Pancreatic Transplants: Noninvasive Evaluation with Secretin-augmented MR Pancreatography and MR Perfusion Measurements--Preliminary Results Radiology, October 1, 2004; 233(1): 273 - 280. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. C. Freund, W. Steurer, E. M. Gassner, K. M. Unsinn, M. Rieger, A. Koenigsrainer, R. Margreiter, and W. R. Jaschke Spectrum of Imaging Findings After Pancreas Transplantation with Enteric Exocrine Drainage: Part 2, Posttransplantation Complications Am. J. Roentgenol., April 1, 2004; 182(4): 919 - 925. [Full Text] [PDF]
|
|
|
|
|
|
L. Johansson, C. Johnsson, E. Penno, A. Bjornerud, and H. Ahlstrom Acute Cardiac Transplant Rejection: Detection and Grading with MR Imaging with a Blood Pool Contrast Agent—Experimental Study in the Rat Radiology, October 1, 2002; 225(1): 97 - 103. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S J Green, P S Sidhu, and C R Deane Imaging of simultaneous kidney pancreatic transplants Imaging, August 1, 2002; 14(4): 299 - 307. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Hariharan, J. D. Pirsch, C. Y. Lu, L. Chan, T. E. Pesavento, S. Alexander, G. L. Bumgardner, G. Baasadona, D. E. Hricik, M. D. Pescovitz, et al. Pancreas after Kidney Transplantation J. Am. Soc. Nephrol., April 1, 2002; 13(4): 1109 - 1118. [Full Text] [PDF]
|
|
|
|
|
|
W. B. Eubank, U. P. Schmiedl, A. E. Levy, and C. L. Marsh Venous Thrombosis and Occlusion After Pancreas Transplantation: Evaluation with Breath-Hold Gadolinium-Enhanced Three-Dimensional MR Imaging Am. J. Roentgenol., August 1, 2000; 175(2): 381 - 385. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
