HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Sahani, D. V. Articles by Hahn, P. F. Search for Related Content PubMed PubMed Citation Articles by Sahani, D. V. Articles by Hahn, P. F.

Intraductal Papillary Mucinous Neoplasm of Pancreas: Multi–Detector Row CT with 2D Curved Reformations—Correlation with MRCP¹

¹ From the Departments of Abdominal Imaging and Interventions (D.V.S., R.K., M.B., P.F.H.), Surgery (C.F.), and Pathology (G.Y.L.), Massachusetts General Hospital, 55 Fruit St, White 270, Boston, MA 02114. Received August 26, 2004; revision requested October 29; revision received February 1, 2005; accepted February 28; final version accepted June 13. Address correspondence to D.V.S. (e-mail: dsahani{at}partners.org).

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

 
Purpose: To retrospectively compare accuracy of multi–detector row computed tomography (CT), combined with two-dimensional (2D) curved reformations, and that of magnetic resonance (MR) cholangiopancreatography (MRCP) for characterization of intraductal papillary mucinous neoplasm (IPMN) as malignant, with pathologic examination as reference standard.

Materials and Methods: Institutional review board approval was obtained, informed consent was waived, and study was HIPAA compliant. Twenty-five patients (12 women, 13 men; age range, 44–88 years) with pathologically proved IPMN were examined with dual-phase CT with 1.25-mm-thick sections for pancreatic phase; 2D curved reformations along main pancreatic duct (MPD) were generated. T2-weighted MRCP included thick- and thin-slab single-shot fast spin-echo imaging and transverse fast spin-echo imaging. Two radiologists, blinded to surgical and pathologic findings, evaluated images for lesion location, septa, mural nodules, communication with MPD, extent and diameter of MPD dilatation, calcifications, and vascular encasement. Malignancy was suspected when one of the following was present: MPD diameter larger than 10 mm, mural nodules, vascular encasement, peripancreatic lymphadenopathy, or metastases. Sensitivity and specificity values for prediction of malignancy were calculated for CT and MRCP. Interobserver variability was determined ( analysis).

Results: Excellent correlation between modalities was observed. Cyst communication was seen in 20 and 21 of 24 branch pancreatic duct (BPD) IPMNs with CT and MRCP, respectively. Sensitivity, specificity, and accuracy for detection of malignancy were 70%, 87%, and 76% (CT) and 70%, 92%, and 80% (MRCP), respectively. Interobserver agreement was good to perfect for both readers in all comparisons (overall, = 0.70–1.00).

Conclusion: CT combined with 2D curved reformation can provide imaging details of IPMN, including communication of BPD IPMN with MPD, that are almost equivalent to those provided at MRCP. Presence of mural nodules, dilated MPD (>10-mm diameter), or thick septa at CT or MRCP may be used as independent predictors of malignancy.

© RSNA, 2006

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

 
Intraductal papillary mucinous neoplasm (IPMN) of the pancreas, once considered a rare tumor, now is increasingly recognized at cross-sectional imaging. This entity was first described by Ohashi et al (1) in 1982 and has since been the subject of much research and clinical interest. IPMN is one of the mucin-producing tumors of the pancreas, and it develops from the epithelial lining of the main pancreatic duct (MPD) or its side branches. Many pancreatic neoplasms, including those previously termed papillary carcinomas, ductectatic mucinous cystadenomas, villous adenomas, and mucin-producing tumors of the pancreas, are now classified as IPMN (2). The treatment decision with regard to IPMN often is based on the patient's age at presentation, the presence or absence of symptoms, the lesion location in the pancreas, the extent of ductal involvement, and the presence or absence of malignant features (3).

Although current thinking is that all IPMNs have the potential to be malignant, the frequency and rate of progression to carcinoma in situ and invasive carcinoma are not known and may be different for IPMNs of the MPD and the branch pancreatic duct (BPD) (4). Because these tumors frequently are discovered in elderly asymptomatic patients, prediction of the likelihood of malignancy is paramount. The surgical strategy may be altered because of the grade of malignancy of the tumor (5). Therefore, radiologic differentiation between benign and malignant lesions is important in the determination of the appropriate treatment. Even when frankly malignant, IPMNs often are resectable, and patients with these lesions have a better prognosis than do patients with ductal adenocarcinoma (5). Previous methods for determination of the likelihood of malignancy have been based mainly on clinical findings and endoscopic retrograde cholangiopancreatographic findings (6–8).

Magnetic resonance (MR) cholangiopancreatography (MRCP) is an excellent imaging technique for characterization of IPMN, and such characterization includes determination of whether a pancreatic cystic lesion communicates with the MPD and assessment of the extent of ductal involvement (9,10). In several reports, investigators compared MRCP and endoscopic retrograde cholangiopancreatography for evaluation of mucin-producing tumors of the pancreas (11–13). Most of these investigators reported the superiority of MRCP over endoscopic retrograde cholangiopancreatography for evaluation of IPMN of the pancreas and concluded that MRCP is the initial modality of choice among imaging techniques for assessment of these lesions. Improvements in computed tomographic (CT) technology, such as multi–detector row CT, as well as the evolution of postprocessing techniques, however, have enhanced the capability of CT for the evaluation of the pancreatic parenchyma and the pancreatic ducts in patients with IPMN (9,10). In addition, CT often is performed in many patients for initial detection and characterization of cystic lesions in the pancreas. The burgeoning use of multi–detector row CT also has led to the commonly incidental detection of these lesions on CT scans obtained for other indications.

To our knowledge, there has been no publication in the literature in which the performance of state-of-the-art multi–detector row CT and that of MRCP for evaluation of IPMN were compared. The purpose of our study was to retrospectively compare the accuracy of multi–detector row CT, combined with two-dimensional (2D) curved reformations, with that of MRCP for characterization of IPMN as malignant, with pathologic examination as the reference standard.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

 
Patients
Twenty-five patients (12 women, 13 men; mean age, 69 years; range, 44–88 years) were examined between December 1999 and June 2004. These patients, who were suspected of having a diagnosis of IPMN that was based on findings from outside imaging studies that were not included as a part of our data, were referred to our institution. Multi–detector row CT and MRCP were performed as requested by the referring clinical services as standard-of-care examinations. Institutional review board approval was obtained for this retrospective analysis, and the need to obtain informed consent was waived. Our study was compliant with the Health Insurance Portability and Accountability Act. Patients included in the study had to have undergone both dedicated contrast material–enhanced multi–detector row CT and MRCP at our institution within 6 weeks prior to surgery. The interval between multi–detector row CT and MRCP ranged from 1 day to 4 weeks (average, 12 days).

CT Technique
Each CT examination was performed with a multisection scanner with four- or 16-section capability (Lightspeed QX/I; GE Medical Systems, Milwaukee, Wis). No oral contrast agent was administered for opacification of the gastrointestinal tract. Nonenhanced CT of the upper abdomen was performed with 10-mm section thickness and 10-mm spacing. A total of 130–150 mL of nonionic iodinated contrast material (Ultravist; Berlex, Montville, NJ), 300 mg of iodine per milliliter, was injected into the antecubital vein at a rate of 4 mL/sec with an 18–20-gauge cannula. Dual-phase CT of the pancreas was performed after the initiation of the contrast material injection as follows: Pancreatic phase imaging was performed at 35 seconds after initiation of contrast material injection; venous phase imaging of the pancreas and the liver was performed afterward at 65–70 seconds after initiation of contrast material injection.

The parameters for scanning were the following: For the pancreatic phase, a section thickness of 1.25 mm and a table speed of 11.25 mm/sec were selected. For the venous phase, a section thickness of 5.00 mm and a table speed of 18.75 mm/sec were selected. Other parameters that were maintained at a constant level for both phases included tube voltage of 140 kVp, tube current of 220–280 mA, and tube rotation at 0.5–0.8 second. Images were retrospectively reconstructed at a 50% overlap, 1.00–1.25-mm section thickness, and 0.625-mm spacing.

Postprocessing of Reconstructed Images
The reconstructed images were subsequently postprocessed at a commercially available workstation (ADW 4; GE Medical Systems). In all patients, coronal subvolume maximum intensity projections and 2D curved reformations were generated by a technologist trained in image postprocessing. Two-dimensional curved reformations were obtained by interactively placing a cursor on a stack of transverse, sagittal, coronal, or oblique sections along the course of the MPD.

MRCP Technique
MR imaging of the pancreas and MRCP were performed with a 1.5-T system (Signa; GE Medical Systems) by using a phased-array surface coil. T1-weighted gradient-recalled echo images were obtained through the pancreas before and after fat saturation (repetition time msec/echo time msec, 150–200/2.1; flip angle, 90°; number of signals acquired, one; matrix, 192 x 256; section thickness, 5.00 mm). T2-weighted MRCP was performed by using a single-shot fast spin-echo sequence. Both thin- and thick-slab MRCP were performed. The following parameters were selected for thick-slab MRCP: 2047/958.5; flip angle, 90°; slab thickness, 50 mm; number of signals acquired, 0.5; matrix, 256 x 256; echo train length, zero; and field of view, 260 mm. Likewise, for thin-slab single-shot fast spin-echo MRCP, the following parameters were selected: 2660/843.8; flip angle, 90°; 4-mm-thick consecutive sections; number of signals acquired, 0.5; matrix, 256 x 256; echo train length, zero; and field of view, 380 mm. Multiplanar imaging was performed in the coronal, coronal oblique, and transverse planes. No intravenous contrast material was administered for MRCP.

Image Analysis
Two readers (D.V.S. and M.B.) independently reviewed multi–detector row CT and MRCP images at a picture archiving and communication system workstation (Agfa, version 4.0; Agfa, Richmond, Va). Both readers were fellowship-trained abdominal radiologists with 5 and 7 years of experience in reading CT and MR images of the pancreas, respectively. The readers were aware of the diagnosis of an IPMN but were blinded to the findings at surgery and pathologic examination. In cases of interobserver disagreement, final decisions were determined with consensus.

Recording of Findings and Image Quality
At CT and MRCP, each reader recorded the following findings: extent of ductal involvement, location and number of lesions, communication of the lesions of the BPD with the MPD, and features suspicious for malignancy. Quality of the 2D curved reformation was assigned a grade subjectively, with a five-point scale as follows: grade 1, poor (image could not be evaluated); grade 2, suboptimal; grade 3, acceptable (minimal artifacts); grade 4, good; and grade 5, excellent (no artifacts).

Extent of ductal involvement.—IPMNs were classified as follows: MPD type with segmental or diffuse dilatation of MPD greater than 4 mm in diameter, BPD type with clustered lesions, and combined type that involved both MPD and BPD types. The maximum diameter of the pancreatic duct in the MPD type and the size of the lesion in the BPD type were recorded by using electronic calipers.

Location and number of lesions.—Description of location was documented according to the anatomic landmarks in the head, the uncinate process, the body, and the tail of the pancreas. Diffuse involvement was denoted when the lesion involved the entire gland.

Communication of the BPD lesions with the MPD.—The communication was defined as a neck or a channel that connected the pancreatic cyst to the normal or dilated MPD. The usefulness of 2D curved reformation in the demonstration of the cyst communication with the MPD also was evaluated.

Features suspicious for malignancy.—Malignancy was suspected when there was presence of mural nodules, papillary projections, or a solid mass in the dilated duct or within the cystic lesion. In cases of MPD- or combined-type IPMN, a diameter of the MPD greater than 10 mm was considered another predictor of malignancy. Likewise, we also looked for the presence of vascular encasement, peripancreatic lymphadenopathy, and metastases as supporting evidence of malignancy. Lesions were considered malignant when one criterion or a combination of the criteria mentioned previously was present. Septa and peripheral or septal calcifications also were noted, and their association with malignancy was assessed. Finally, we evaluated the presence of malignancy in side-branch IPMNs that were greater than 30 mm in maximum diameter. On the basis of the imaging findings, readers predicted whether the lesion was benign or malignant.

Surgery
All patients underwent resection of the IPMN performed by one surgeon (C.F.) with more than 10 years of experience in pancreatic surgery. The mean interval between imaging and surgery was 35 days. A Whipple procedure was performed in 16 patients; distal pancreatectomy, in six patients; middle segment pancreatectomy, in two patients; and total pancreatectomy, in one patient. The type of resection performed was based on the pancreatic lesion location, extent of MPD involvement at CT and at MRCP, and findings at frozen-section analysis in regard to the presence or absence of malignancy in the specimen. Pathologic examination was performed in gross specimens. Marking stitches were made on the resected specimens to help the pathologist delineate the head, the body, and the tail of the pancreas.

Pathologic Examination
One gastrointestinal pathologist (G.Y.L.) with more than 12 years of experience in the evaluation of the pancreas reviewed each pathologic specimen. The extent of ductal involvement and the extent of the tumor were assessed. We looked for epithelial cell proliferation, mucin production, location of the lesion, presence of mural nodules, presence of septa, and whether there was communication of the cyst with the MPD. The tumors were classified into histologic subtypes as follows: benign lesion, low-grade malignancy, and invasive carcinoma. Low-grade malignancy was diagnosed when evidence of carcinoma in situ or borderline malignant changes were noted. Invasive carcinoma was diagnosed when there was evidence of invasion into the pancreatic parenchyma and/or surrounding structures. The degree of extrapancreatic extent and of peripancreatic lymphadenopathy (diameter of >1 cm in the short axis) also were assessed. The pancreatic parenchyma also was evaluated for evidence of pancreatitis or pancreatic atrophy.

Statistical Analysis
Sensitivity and specificity values for the prediction of malignancy were calculated for the findings at both multi–detector row CT and MRCP. We analyzed the value of individual findings in the prediction of malignancy and also the overall prediction of malignancy based on the combination of findings at multi–detector row CT and at MRCP. Analysis with the value was performed by using statistical software (Medcalc, version 7.4.2.0, 1993–2004; MedCalc, Marikerke, Belgium) to determine interobserver variability between the two readers. Forward conditional logistic regression analysis was performed by using software (SPSS, version 11.5.1; SPSS, Chicago, Ill) to determine the better predictor for malignancy.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

 
At pathologic examination, eight patients had benign IPMNs, 11 had low-grade malignancies, and six had invasive cancers.

Results for evaluation of the extent of ductal involvement among 25 patients with IPMN were as follows: Five patients had MPD-type lesions (two were diffuse and three were segmental), seven had combined-type lesions, and the remaining 13 had BPD-type lesions. In 13 patients with IPMNs of the BPD, a total of 17 lesions were present: Ten patients had one lesion, two patients had two lesions, and one patient had three lesions.

Findings at multi–detector row CT with 2D curved reformations were used to correctly predict the extent of ductal involvement in all 25 patients (Figs 1, 2). Likewise, results of MRCP were concordant with the pathologic findings of involvement of the MPD and BPD in 24 of 25 patients. In one patient with combined-type IPMN, the readers failed to recognize the involvement of the MPD at MRCP.

View larger version (120K): [in this window] [in a new window] [Download PPT slide]   Figure 1a: Combined-type IPMN in 59-year-old man. (a) Transverse multi–detector row CT image shows the plane in which 2D curved reformations were obtained along the course of the MPD. (b) Two-dimensional curved reformation demonstrates the communication between the cystic lesion (arrow) and the dilated MPD (arrowhead). (c) Gross pathologic specimen obtained at surgery shows translucent cyst (arrow) that projects from the surface of the pancreas.

View larger version (112K): [in this window] [in a new window] [Download PPT slide]   Figure 1b: Combined-type IPMN in 59-year-old man. (a) Transverse multi–detector row CT image shows the plane in which 2D curved reformations were obtained along the course of the MPD. (b) Two-dimensional curved reformation demonstrates the communication between the cystic lesion (arrow) and the dilated MPD (arrowhead). (c) Gross pathologic specimen obtained at surgery shows translucent cyst (arrow) that projects from the surface of the pancreas.

View larger version (118K): [in this window] [in a new window] [Download PPT slide]   Figure 1c: Combined-type IPMN in 59-year-old man. (a) Transverse multi–detector row CT image shows the plane in which 2D curved reformations were obtained along the course of the MPD. (b) Two-dimensional curved reformation demonstrates the communication between the cystic lesion (arrow) and the dilated MPD (arrowhead). (c) Gross pathologic specimen obtained at surgery shows translucent cyst (arrow) that projects from the surface of the pancreas.

View larger version (101K): [in this window] [in a new window] [Download PPT slide]   Figure 2a: MPD-type IPMN in 64-year-old man who presented with epigastric pain. (a) Transverse multi–detector row CT image demonstrates a dilated MPD in the head of the pancreas with an enhancing mural nodule (arrowhead). (b) Curved reformation demonstrates extent of segmental dilatation (arrow) of the MPD in the head and the proximal body of the pancreas. (c) T2-weighted transverse MR image shows a dilated MPD with a hypointense mural nodule (arrowhead) along the anterior wall. (d) MRCP image shows extent of ductal dilatation (arrow). (e) Gross specimen of the pancreas dissected along the MPD shows the mucin (arrows) in the dilated MPD.

View larger version (111K): [in this window] [in a new window] [Download PPT slide]   Figure 2b: MPD-type IPMN in 64-year-old man who presented with epigastric pain. (a) Transverse multi–detector row CT image demonstrates a dilated MPD in the head of the pancreas with an enhancing mural nodule (arrowhead). (b) Curved reformation demonstrates extent of segmental dilatation (arrow) of the MPD in the head and the proximal body of the pancreas. (c) T2-weighted transverse MR image shows a dilated MPD with a hypointense mural nodule (arrowhead) along the anterior wall. (d) MRCP image shows extent of ductal dilatation (arrow). (e) Gross specimen of the pancreas dissected along the MPD shows the mucin (arrows) in the dilated MPD.

View larger version (69K): [in this window] [in a new window] [Download PPT slide]   Figure 2c: MPD-type IPMN in 64-year-old man who presented with epigastric pain. (a) Transverse multi–detector row CT image demonstrates a dilated MPD in the head of the pancreas with an enhancing mural nodule (arrowhead). (b) Curved reformation demonstrates extent of segmental dilatation (arrow) of the MPD in the head and the proximal body of the pancreas. (c) T2-weighted transverse MR image shows a dilated MPD with a hypointense mural nodule (arrowhead) along the anterior wall. (d) MRCP image shows extent of ductal dilatation (arrow). (e) Gross specimen of the pancreas dissected along the MPD shows the mucin (arrows) in the dilated MPD.

View larger version (89K): [in this window] [in a new window] [Download PPT slide]   Figure 2d: MPD-type IPMN in 64-year-old man who presented with epigastric pain. (a) Transverse multi–detector row CT image demonstrates a dilated MPD in the head of the pancreas with an enhancing mural nodule (arrowhead). (b) Curved reformation demonstrates extent of segmental dilatation (arrow) of the MPD in the head and the proximal body of the pancreas. (c) T2-weighted transverse MR image shows a dilated MPD with a hypointense mural nodule (arrowhead) along the anterior wall. (d) MRCP image shows extent of ductal dilatation (arrow). (e) Gross specimen of the pancreas dissected along the MPD shows the mucin (arrows) in the dilated MPD.

View larger version (114K): [in this window] [in a new window] [Download PPT slide]   Figure 2e: MPD-type IPMN in 64-year-old man who presented with epigastric pain. (a) Transverse multi–detector row CT image demonstrates a dilated MPD in the head of the pancreas with an enhancing mural nodule (arrowhead). (b) Curved reformation demonstrates extent of segmental dilatation (arrow) of the MPD in the head and the proximal body of the pancreas. (c) T2-weighted transverse MR image shows a dilated MPD with a hypointense mural nodule (arrowhead) along the anterior wall. (d) MRCP image shows extent of ductal dilatation (arrow). (e) Gross specimen of the pancreas dissected along the MPD shows the mucin (arrows) in the dilated MPD.

View larger version (151K): [in this window] [in a new window] [Download PPT slide]   Figure 3a: BPD-type IPMN in 44-year-old woman. (a) Transverse multi–detector row CT image shows a multiseptate cystic lesion, in the middle of the body of the pancreas, with a narrow neck (white arrow) that communicates with MPD (black arrow). (b) Two-dimensional curved reformation shows layout of the MPD (black arrow), as well as the communication (white arrow) between IPMN and MPD. (c) Coronal oblique single-shot fast spin-echo MRCP image clearly demonstrates communication (arrow) between cyst and MPD.

View larger version (148K): [in this window] [in a new window] [Download PPT slide]   Figure 3b: BPD-type IPMN in 44-year-old woman. (a) Transverse multi–detector row CT image shows a multiseptate cystic lesion, in the middle of the body of the pancreas, with a narrow neck (white arrow) that communicates with MPD (black arrow). (b) Two-dimensional curved reformation shows layout of the MPD (black arrow), as well as the communication (white arrow) between IPMN and MPD. (c) Coronal oblique single-shot fast spin-echo MRCP image clearly demonstrates communication (arrow) between cyst and MPD.

View larger version (97K): [in this window] [in a new window] [Download PPT slide]   Figure 3c: BPD-type IPMN in 44-year-old woman. (a) Transverse multi–detector row CT image shows a multiseptate cystic lesion, in the middle of the body of the pancreas, with a narrow neck (white arrow) that communicates with MPD (black arrow). (b) Two-dimensional curved reformation shows layout of the MPD (black arrow), as well as the communication (white arrow) between IPMN and MPD. (c) Coronal oblique single-shot fast spin-echo MRCP image clearly demonstrates communication (arrow) between cyst and MPD.

View larger version (147K): [in this window] [in a new window] [Download PPT slide]   Figure 4a: Images in 56-year-old man who presented with a vague pain in the upper abdomen. (a) Transverse CT image shows a small cyst (arrow) in the head of the pancreas without a definite communication with MPD. (b) Corresponding 2D curved reformation demonstrates a clear communication (arrow) between cyst and MPD. (c) Coronal 2D MRCP image clearly shows morphologic features of cyst and an obvious communication (arrow) with the normal-sized MPD.

View larger version (145K): [in this window] [in a new window] [Download PPT slide]   Figure 4b: Images in 56-year-old man who presented with a vague pain in the upper abdomen. (a) Transverse CT image shows a small cyst (arrow) in the head of the pancreas without a definite communication with MPD. (b) Corresponding 2D curved reformation demonstrates a clear communication (arrow) between cyst and MPD. (c) Coronal 2D MRCP image clearly shows morphologic features of cyst and an obvious communication (arrow) with the normal-sized MPD.

View larger version (78K): [in this window] [in a new window] [Download PPT slide]   Figure 4c: Images in 56-year-old man who presented with a vague pain in the upper abdomen. (a) Transverse CT image shows a small cyst (arrow) in the head of the pancreas without a definite communication with MPD. (b) Corresponding 2D curved reformation demonstrates a clear communication (arrow) between cyst and MPD. (c) Coronal 2D MRCP image clearly shows morphologic features of cyst and an obvious communication (arrow) with the normal-sized MPD.

View larger version (114K): [in this window] [in a new window] [Download PPT slide]   Figure 5a: Images show incidentally detected pancreatic cyst in 76-year-old man. Diagnosis was BPD-type IPMN. (a) Transverse CT image through the pancreas shows septate cystic lesion (arrow) in the head of the pancreas. It was difficult to observe an obvious communication of this lesion with the pancreatic duct on the transverse images alone. (b) A definite communication (arrow) between the lesion and mildly dilated MPD (arrowhead) was noted on the 2D curved reformation. Note a small pocket of air in the pancreatic duct that was likely caused by endoscopic ultrasonographically guided aspiration performed at an outside institution. (c) Similar findings of a communication (arrow) between the lesion and the dilated MPD (arrowhead) were noted on the corresponding coronal MRCP image. At pathologic analysis, this lesion was diagnosed as a benign side-branch IPMN.

View larger version (114K): [in this window] [in a new window] [Download PPT slide]   Figure 5b: Images show incidentally detected pancreatic cyst in 76-year-old man. Diagnosis was BPD-type IPMN. (a) Transverse CT image through the pancreas shows septate cystic lesion (arrow) in the head of the pancreas. It was difficult to observe an obvious communication of this lesion with the pancreatic duct on the transverse images alone. (b) A definite communication (arrow) between the lesion and mildly dilated MPD (arrowhead) was noted on the 2D curved reformation. Note a small pocket of air in the pancreatic duct that was likely caused by endoscopic ultrasonographically guided aspiration performed at an outside institution. (c) Similar findings of a communication (arrow) between the lesion and the dilated MPD (arrowhead) were noted on the corresponding coronal MRCP image. At pathologic analysis, this lesion was diagnosed as a benign side-branch IPMN.

View larger version (111K): [in this window] [in a new window] [Download PPT slide]   Figure 5c: Images show incidentally detected pancreatic cyst in 76-year-old man. Diagnosis was BPD-type IPMN. (a) Transverse CT image through the pancreas shows septate cystic lesion (arrow) in the head of the pancreas. It was difficult to observe an obvious communication of this lesion with the pancreatic duct on the transverse images alone. (b) A definite communication (arrow) between the lesion and mildly dilated MPD (arrowhead) was noted on the 2D curved reformation. Note a small pocket of air in the pancreatic duct that was likely caused by endoscopic ultrasonographically guided aspiration performed at an outside institution. (c) Similar findings of a communication (arrow) between the lesion and the dilated MPD (arrowhead) were noted on the corresponding coronal MRCP image. At pathologic analysis, this lesion was diagnosed as a benign side-branch IPMN.

View larger version (117K): [in this window] [in a new window] [Download PPT slide]   Figure 6a: Diffuse IPMN in 83-year-old man who had abdominal discomfort and weight loss. (a) Two-dimensional curved reformation demonstrates the extent of the dilatation of MPD. Note presence of mural nodules (arrow) along the dilated MPD and bulging of the duodenal papilla (arrowhead). (b) Comparative coronal oblique single-shot fast spin-echo MRCP image shows diffuse pancreatic duct dilatation, and the dilated duct contained multiple T2 dark mural nodules (arrow) along the periphery.

View larger version (98K): [in this window] [in a new window] [Download PPT slide]   Figure 6b: Diffuse IPMN in 83-year-old man who had abdominal discomfort and weight loss. (a) Two-dimensional curved reformation demonstrates the extent of the dilatation of MPD. Note presence of mural nodules (arrow) along the dilated MPD and bulging of the duodenal papilla (arrowhead). (b) Comparative coronal oblique single-shot fast spin-echo MRCP image shows diffuse pancreatic duct dilatation, and the dilated duct contained multiple T2 dark mural nodules (arrow) along the periphery.

View larger version (137K): [in this window] [in a new window] [Download PPT slide]   Figure 7a: Malignant MPD-type IPMN in 63-year-old man who had epigastric pain. (a) Transverse CT image depicts a diffusely dilated MPD along with a cystic lesion (arrow) in the tail of the pancreas that communicates with the dilated MPD. In addition, a soft-tissue mass (arrowhead) is present in the proximal MPD. (b) Transverse fast spin-echo T2-weighted MR image shows features of IPMN (arrow) and a dark mural nodule (arrowhead) in the proximal MPD.

View larger version (116K): [in this window] [in a new window] [Download PPT slide]   Figure 7b: Malignant MPD-type IPMN in 63-year-old man who had epigastric pain. (a) Transverse CT image depicts a diffusely dilated MPD along with a cystic lesion (arrow) in the tail of the pancreas that communicates with the dilated MPD. In addition, a soft-tissue mass (arrowhead) is present in the proximal MPD. (b) Transverse fast spin-echo T2-weighted MR image shows features of IPMN (arrow) and a dark mural nodule (arrowhead) in the proximal MPD.

Both readers considered the 2D curved reformations to be of good quality, and the average score from two readers was 4.1. Excellent interobserver agreement was achieved for findings at multi–detector row CT and MRCP (Tables 4, 5). The presence of mural nodules, of thick septa, and of calcifications and a diameter of the MPD greater than 10 mm were all good predictors of malignancy in the present study (Table 6).

Multi–detector row CT had an overall sensitivity, specificity, and accuracy of 70%, 87%, and 76%, respectively, for detection of malignancy; in comparison, MRCP had values of 70%, 92%, and 80%, respectively.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

 
IPMN accounts for 1% of all exocrine pancreatic neoplasms, and this neoplasm has been reported with increasing frequency (14). In its typical form, IPMN is characterized by dilatation of the MPD or BPD, with excretion of mucin through the patulous orifice of an enlarged papilla of Vater. The surgical treatment of IPMN differs from that of serous cystadenomas and mucinous cystic neoplasms. Although the surgeon can usually locate the tumor preoperatively in patients with serous cystadenomas and mucinous cystic neoplasms and, accordingly, can plan a segmental pancreatic resection, this is not always the case in patients with IPMN. In patients with IPMN of the MPD, the presence of dilatation in the MPD with or without an intraductal mass is a classic feature. Because of the overproduction of mucus, MPD dilatation can occur both proximal and distal to the tumor, and this occurrence causes problems in determining the location of the tumor.

Helical CT often is used in the evaluation of pancreatic neoplasms (15,16). This technique can be used to image the entire pancreas with thin sections during a single breath hold, and it is less subject to respiratory motion and partial-volume effects than are previously used technologies. There are few reports available in the literature about the usefulness of thin-section CT in the evaluation of IPMN (17,18). Our aim in this study was to evaluate the usefulness of state-of-the-art thin-section multi–detector row CT compared with MRCP for the comprehensive evaluation of IPMN.

Our results indicated that multi–detector row CT is a very effective technique for detailed evaluation of the IPMN. Good correlation was established with both multi–detector row CT and MRCP for the determination of features of the IPMN seen at gross pathologic examination.

With respect to extent of ductal involvement, that in IPMN may be classified into three groups: MPD type, BPD type, and combined type. In our study, seven (28%) patients had the combined type, five (20%) patients had the MPD type, and 13 (52%) patients had only BPD involvement. The involvement of the MPD in one patient with the combined-type IPMN was overlooked at MRCP; however, the involvement was correctly identified in all patients at multi–detector row CT with curved reformations.

In the past, several authors (19,20) have noted that a cyst diameter of 30 mm or greater suggests malignancy. In our experience, however, a diameter of the BPD lesion of greater than 30 mm did not correlate well with the presence of malignancy in this study; sensitivity was 42% (five of 12) and specificity was 75% (nine of 12). A similar observation was made by Chari et al (21). In their study, 73 noninvasive IPMNs had an average diameter of 5.2 cm, whereas 40 invasive IPMNs had an average diameter of 6.6 cm. Thus, the size of a side-branch IPMN alone cannot be used to predict malignancy.

On the other hand, a dilated MPD (>10-mm diameter) correlated well with the presence of malignancy in the lesion. A total of eight of nine patients with MPD with a diameter of greater than 10 mm had evidence of malignancy (specificity of 87% for prediction of malignancy). This was also previously shown by Taouli et al (18). An MPD greater than 10 mm in diameter had a specificity of 96% for the prediction of malignancy. Thus, it is prudent to suspect malignancy if the MPD diameter is more than 10 mm.

With respect to communication of the IPMN of the BPD with the MPD, we determined that communication of the pancreatic cystic lesion with the MPD is one of the most reliable findings for the diagnosis of IPMN (6,22,23). In our study, multi–detector row CT images along with 2D curved reformations showed communication in 20 (83%) of 24 IPMNs of the BPD type and the combined type. The 2D curved reformations were useful for establishment of a clear communication, which had not been obvious on the transverse images, in eight of 19 lesions.

The thickness of the curved plane is the voxel dimension perpendicular to the curved plane and depends on the orientation of the section on which it is drawn. It should be noted that curved reformations are highly dependent on the accuracy of the operator who draws the curve, and artifactual lesions may be created or eccentric lesions may not be displayed (24).

In previous studies in which thin-section CT or helical technology was not used, it was difficult to demonstrate the cyst communication with the MPD (6,25–27). In another study, however, Fukukura et al (17) demonstrated communication between the MPD and the side-branch cyst at thin-section CT in 80% of the cases. The improved sensitivity for recognition of cyst communication in our study may have been caused by the improved resolution of multi–detector row CT technology and the use of an even thinner section thickness with 1.25-mm-thick sections with overlapping reconstructions. This technology and the use of thinner section thickness with overlapping reconstructions also explain the esthetically pleasing quality of the 2D curved reformations generated in this study. The performance of MRCP also was good, and the communication was depicted in 21 of 24 patients with BPD- and combined-type IPMNs (sensitivity, 87%), and the depiction of the communication was similar to that for the performance of multi–detector row CT.

In regard to features suggestive of malignancy, various predictive signs of malignancy have been quoted in the literature. They include a history of diabetes, male sex, a history of alcohol abuse, the presence of marked MPD dilatation (>10-mm diameter), a large size of tumors with different threshold diameter of 3–5 cm, the presence of thickened septumlike structures, and the presence of protruding masses in the dilated duct (5,28,29). In our study, the presence of mural nodules, thick septa, and calcifications and a diameter of greater than 10 mm of the MPD were all good predictors of malignancy. Our results are similar to those in the study of Taouli et al (18). These researchers suggested a combination of four CT findings for the detection of invasive IPMNs: a solid mass, an MPD dilatation with a greater than 10-mm diameter, either diffuse or multifocal involvement, and attenuating or calcified intraluminal content. Calcifications, however, were not mentioned as a sign of malignant IPMN in two studies with large series of patients—1379 and 62 patients, respectively (30,31). The discrepancy between their findings and those in the present study may have been caused by the small sample size in our study. In the study by Suzuki et al (30 with the largest series of patients in the literature about IPMN, the indicators of malignant IPMN included advanced age, positive symptoms, abundant mucous secretion, presence of large nodules and/or large cysts, marked dilatation of the MPD, and MPD- or combined-type IPMN.

Our study had some limitations. First, the patients included in this study were from a select group with IPMN because they had undergone multi–detector row CT and MRCP before surgery, and, therefore, our data contain an inadvertent patient selection bias. The lack of a substantial difference in the sensitivity and specificity of multi–detector row CT and MRCP may be partly caused by the small sample size. In addition, we did not address the issue of distinguishing an IPMN from other causes of cystic lesions or pancreatic ductal dilatation. The addition of a three-dimensional fast spin-echo T2-weighted MRCP sequence and sequences performed with gadopentetate dimeglumine (Magnevist; Berlex) could have potentially enhanced the performance of MRCP. Finally, our institution is a tertiary referral center for pancreatic surgery, and, therefore, the data from our study do not reflect the disease prevalence in a general hospital.

In conclusion, multi–detector row CT can enable excellent evaluation of the pancreatic parenchyma—with delineation of ductal dilatation, calcifications, mural nodules, and mass lesions—with a good predictive value for the diagnosis and the extent of ductal involvement of malignant IPMN. Just as with MRCP, 2D curved reformations can provide additional imaging details of IPMN, which include display of the ductal anatomy and of the communication between the IPMN of the BPD and the MPD. The presence of a mural nodule and of septa greater than 2 mm thick and an MPD greater than 10 mm in diameter may be used as predictors of malignancy. Diameter of IPMNs of the BPD greater than 30 mm did not correlate well with the presence of malignancy. In the present study, multi–detector row CT with 2D curved reformations, compared with MRCP, provided similar diagnostic accuracy for detection of IPMN. Multi–detector row CT with 2D curved reformations can provide imaging details of the IPMN that are almost equivalent to those provided with MRCP, and this technique may be used advantageously as a comprehensive evaluation tool.

	FOOTNOTES

 

Abbreviations: BPD = branch pancreatic duct • IPMN = intraductal papillary mucinous neoplasm • MPD = main pancreatic duct • MRCP = MR cholangiopancreatography • 2D = two-dimensional

Authors stated no financial relationship to disclose.

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

	References

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

 

1. Ohashi K, Murakami Y, Maryuama M, et al. Four cases of mucous secreting pancreatic cancer. Prog Dig Endosc 1982;20:348–351.
2. Adsay NV, Longnecker DS, Klimstra DS. Pancreatic tumors with cystic dilatation of the ducts: intraductal papillary mucinous neoplasms and intraductal oncocytic papillary neoplasms. Semin Diagn Pathol 2000;17:16–30.[Medline]
3. Gigot JF, Deprez P, Sempoux C, Metairie S, Glineur D, Gianello P. Surgical management of intraductal papillary mucinous tumors of the pancreas: the role of routine frozen section of the surgical margin, intraoperative endoscopic staged biopsies of the Wirsung duct, and pancreaticogastric anastomosis. Arch Surg 2001;136:1256–1262.[Abstract/Free Full Text]
4. Salvia R, Fernandez-del Castillo C, Bassi C, et al. Main-duct intraductal papillary mucinous neoplasms of the pancreas: clinical predictors of malignancy and long-term survival following resection. Ann Surg 2004;239(5):678–685.[CrossRef][Medline]
5. Yamaguchi K, Ogawa Y, Chijiiwa K, Tanaka M. Mucin-hypersecreting tumors of the pancreas: assessing the grade of malignancy preoperatively. Am J Surg 1996;171:427–431.[CrossRef][Medline]
6. Procacci C, Graziani R, Bicego E, et al. Intraductal mucin-producing tumors of the pancreas: imaging findings. Radiology 1996;198:249–257.[Abstract/Free Full Text]
7. Nickl NJ, Lawson JM, Cotton PB. Mucinous pancreatic tumors: ERCP findings. Gastrointest Endosc 1991;37:133–138.[Medline]
8. Maeshiro K, Nakayama Y, Yasunami Y, Furuta K, Ikeda S. Diagnosis of mucin-producing tumor of the pancreas by balloon-catheter endoscopic retrograde pancreatography: compression study. Hepatogastroenterology 1998;45:1986–1995.[Medline]
9. Irie H, Honda H, Aibe H, et al. MR cholangiopancreatographic differentiation of benign and malignant intraductal mucin-producing tumors of the pancreas. AJR Am J Roentgenol 2000;174:1403–1408.[Abstract/Free Full Text]
10. Sugiyama M, Atomi Y, Hachiya J. Intraductal papillary tumors of the pancreas: evaluation with magnetic resonance cholangiopancreatography. Am J Gastroenterol 1998;93:156–159.[CrossRef][Medline]
11. Irie H, Honda H, Tajima T, et al. Optimal MR cholangiopancreatographic sequence and its clinical application. Radiology 1998;206:379–387.[Abstract/Free Full Text]
12. Barish MA, Soto JA. MR cholangiopancreatography: techniques and clinical application. AJR Am J Roentgenol 1997;169:1295–1303.[Free Full Text]
13. Irie H, Honda H, Jimi M, et al. Value of MR cholangiopancreatography in evaluating choledochal cysts. AJR Am J Roentgenol 1998;171:1381–1385.[Abstract/Free Full Text]
14. Kloppel G, Schluter E, Pathology of pancreas. In: Baert A, ed. Radiology of the pancreas. 2nd ed. Berlin, Germany: Springer-Verlag, 1999; 69–100.
15. Bluemke DA, Cameron JL, Hruban RH, et al. Potentially resectable pancreatic adenocarcinoma: spiral CT assessment with surgical and pathologic correlation. Radiology 1995;197:381–385.[Abstract/Free Full Text]
16. Stafford-Johnson DB, Francis IR, Eckhauser FE, Knol JA, Chang AE. Dual-phase helical CT of nonfunctioning islet cell tumors. J Comput Assist Tomogr 1998;22:59–63.[CrossRef][Medline]
17. Fukukura Y, Fujiyoshi F, Sasaki M, Inoue H, Yonezawa S, Nakajo M. Intraductal papillary mucinous tumors of the pancreas: thin-section helical CT findings. AJR Am J Roentgenol 2000;174:441–447.[Abstract/Free Full Text]
18. Taouli B, Vilgrain V, Vullierme M, et al. Intraductal papillary mucinous tumors of the pancreas: helical CT with histopathologic correlation. Radiology 2000;217:757–764.[Abstract/Free Full Text]
19. Sugiyama M, Atomi Y, Kuroda A. Two types of mucin-producing cystic tumors of the pancreas: diagnosis and treatment. Surgery 1997;122:617–625.[CrossRef][Medline]
20. Obara T, Maguchi H, Saitoh Y, et al. Mucin-producing tumor of the pancreas: surgery or follow-up? [in Japanese]. Nippon Shokakibyo Gakkai Zasshi 1994;91:66–74.[Medline]
21. Chari ST, Yadav D, Smyrk TC, et al. Study of recurrence after surgical resection of intraductal papillary mucinous neoplasm of the pancreas. Gastroenterology 2002;123:1500–1507.[CrossRef][Medline]
22. Ohta T, Nagakawa T, Akiyama T, et al. The "ductectatic" variant of mucinous cystic neoplasm of the pancreas: clinical and radiologic studies of seven cases. Am J Gastroenterol 1992;87:300–304.[Medline]
23. Onaya H, Itai Y, Niitsu M, Chiba T, Michishita N, Saida Y. Ductectatic mucinous cystic neoplasms of the pancreas: evaluation with MR cholangiopancreatography. AJR Am J Roentgenol 1998;171:171–177.[Abstract/Free Full Text]
24. Vargas R, Murcia MN, Trueblood W, Jeffrey RB. MDCT in pancreatic adenocarcinoma: prediction of vascular invasion and respectability using a multiphasic technique with curved planar reformations. AJR Am J Roentgenol 2004;182:419–425.[Abstract/Free Full Text]
25. Itoh S, Ishiguchi T, Ishigaki T, Sakuma S, Maruyama K, Senda K. Mucin-producing pancreatic tumor: CT findings and histopathologic correlation. Radiology 1992;183:81–86.[Abstract/Free Full Text]
26. Itai Y, Kokubo T, Atomi Y, Kuroda A, Haraguchi Y, Terano A. Mucin-hypersecreting carcinoma of the pancreas. Radiology 1987;165:51–55.[Abstract/Free Full Text]
27. McDowell RK, Gazelle GS, Murphy BL, et al. Mucinous ductal ectasia of the pancreas. J Comput Assist Tomogr 1997;21:383–388.[CrossRef][Medline]
28. Yanagisawa A, Ohashi K, Hori M, et al. Ductectatic-type mucinous cystadenoma and cystadenocarcinoma of the human pancreas: a novel clinicopathological entity. Jpn J Cancer Res 1993;84:474–479.[CrossRef]
29. Kobayashi G, Fujita N, Lee S, Kimura K, Watanabe H, Mochizuki F. Correlation between ultrasonographic findings and pathological diagnosis of the mucin-producing tumor of the pancreas [in Japanese]. Nippon Shokakibyo Gakkai Zasshi 1990;87:235–242.[Medline]
30. Suzuki Y, Atomi Y, Sugiyama M, et al. Cystic neoplasm of the pancreas: a Japanese multiinstitutional study of intraductal papillary mucinous tumor and mucinous cystic tumor. Pancreas 2004;28(3):241–246.[CrossRef][Medline]
31. Sugiyama M, Izumisato Y, Abe N, Masaki T, Mori T, Atomi Y. Predictive factors for malignancy in intraductal papillary-mucinous tumours of the pancreas. Br J Surg 2003;90(10):1244–1249.[CrossRef][Medline]

This article has been cited by other articles:

				H. Ogawa, S. Itoh, M. Ikeda, K. Suzuki, and S. Naganawa Intraductal Papillary Mucinous Neoplasm of the Pancreas: Assessment of the Likelihood of Invasiveness with Multisection CT Radiology, September 1, 2008; 248(3): 876 - 886. [Abstract] [Full Text] [PDF]

				C. Bassi, S. Crippa, and R. Salvia Intraductal papillary mucinous neoplasms (IPMNs): is it time to (sometimes) spare the knife? Gut, March 1, 2008; 57(3): 287 - 289. [Full Text] [PDF]

				M. D'Onofrio, A. J. Megibow, N. Faccioli, R. Malago, P. Capelli, M. Falconi, and R. P. Mucelli Comparison of Contrast-Enhanced Sonography and MRI in Displaying Anatomic Features of Cystic Pancreatic Masses Am. J. Roentgenol., December 1, 2007; 189(6): 1435 - 1442. [Abstract] [Full Text] [PDF]

				D. S. Katz, D. M. Friedel, D. Kho, N. Georgiou, and J. J. Hines Relative Accuracy of CT and MRI for Characterization of Cystic Pancreatic Masses Am. J. Roentgenol., September 1, 2007; 189(3): 657 - 661. [Full Text] [PDF]

				R. Salvia, S. Crippa, M. Falconi, C. Bassi, A. Guarise, A. Scarpa, and P. Pederzoli Branch-duct intraductal papillary mucinous neoplasms of the pancreas: to operate or not to operate? Gut, August 1, 2007; 56(8): 1086 - 1090. [Abstract] [Full Text] [PDF]

This Article Abstract