| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, Division of Abdominal Imaging and Intervention,
Brigham and Women's Hospital, Harvard Medical School, 75 Francis St., Boston,
MA 02115.
2 Department of Radiology, Dana Farber Cancer Institute, Boston, MA 02115.
3 Department of Radiologic Pathology, Armed Forces Institute of Pathology,
Washington, DC.
4 Department of Radiology, Uniformed Services University of Health Sciences,
Bethesda, MD.
5 Department of Pathology, Brigham and Women's Hospital, Harvard Medical School,
Boston, MA.
6 Division of Gastroenterology, Department of Medicine, Brigham and Women's
Hospital, Harvard Medical School, Boston, MA.
Received April 18, 2004;
accepted after revision July 19, 2004.
Address correspondence to S. Tatli
(statli{at}partners.org).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Eleven patients (six women and five men; mean age, 64 years) with acinar cell carcinoma, documented by pathologic examination of resected specimens, underwent CT (n = 9) or MRI (n = 2) examinations. Two radiologists evaluated imaging studies and determined, by consensus, the following data for each tumor: size, location, margination, internal density or signal intensity, and contrast enhancement pattern. In addition, they assessed the presence of calcification, pancreatic or bile duct dilation, and metastases. Imaging features were correlated with gross and microscopic pathologic features of the tumors.
RESULTS. Masses were distributed throughout the pancreas (head, n = 5; body, n = 2; and tail, n = 4). The mean largest dimensions were 6.0 x 5.3 cm (range, from 2 x 1.7 to 15 x 11 cm). Tumors were oval (n = 5), round (n = 4), or lobular (n = 2). Ten (91%) masses were well marginated; nine (82%) were exophytic. Five (45%) masses enhanced homogeneously; the remaining tumors contained cystic areas. All masses enhanced less than the surrounding pancreas. Three (27%) masses contained calcifications. Four (80%) masses invaded the duodenum. Common bile and pancreatic duct dilatation was present in two and three patients, respectively. One patient had metastatic liver disease at presentation.
CONCLUSION. Pure acinar cell carcinoma of the pancreas is usually an exophytic, oval or round, well-marginated, and hypovascular mass on CT and MRI. It typically is completely solid when small and contains cystic areas due to necrosis when large.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although the acinar cell carcinoma has been long recognized as a distinct clinicopathologic entity, to our knowledge, a comprehensive analysis of their CT and MRI features has not been reported. Case reports have described the CT appearances of acinar cell carcinoma as a poorly defined, dense mass [5], a well-defined mass with central necrosis [6], a cystic mass surrounded by a thick hypervascular wall [7], a well-defined hypodense mass with a thin, enhancing capsule [8], and a well-defined, hypervascular solid mass [9]. Acinar cell carcinoma showed avid uptake of mangafodipir trisodium on MRI in one patient [10].
The objective of this study was to describe the CT and MRI findings of acinar cell carcinoma of the pancreas in adults.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our study population included six women and five men with a mean age of 64 years (range, 44–79 years). Ten patients were white, and one was Hispanic. Nine patients (82%) presented with epigastric and right upper quadrant pain (n = 6), jaundice (n = 1), back pain (n = 1), or painful skin nodules and arthralgias (n = 1). The median time interval between onset of clinical symptoms and the pathologic diagnosis was 8 months (range, 1–36 months). In two patients, tumors were discovered incidentally on a chest CT scan obtained for shortness of breath (n = 1) and for chest tightness (n = 1). Tumors were palpable in two patients (18%). One of the patients with recurrent epigastric pain had been mistakenly diagnosed with chronic pancreatitis, and the tumor had been misdiagnosed as a pseudocyst. Of nine patients for whom laboratory data were available, three had elevated serum lipase or amylase levels, and one patient had an elevated serum bilirubin level.
All patients underwent surgical resection of the tumor either with distal pancreatectomy (n = 6), pancreaticoduodenectomy (Whipple procedure) (n = 4), or a palliative tumor resection with gastrojejunostomy (n = 1).
Imaging Technique and Analysis
Nine patients underwent CT examinations, eight with IV contrast material.
Two had both unenhanced and enhanced scans. One patient had only an unenhanced
CT scan. Two patients underwent MRI, one with IV gadolinium and one without.
Both MRI examinations included unenhanced T1-weighted spin-echo images and
T2-weighted fast spin-echo images. Spoiled gradient-recalled echo images with
fat suppression were obtained before and after the IV administration of
gadopentetate dimeglumine (2 mmol/kg of body weight). Contrast-enhanced CT or
MR images were obtained during the arterial phase (n = 3) and portal
venous phase (n = 6).
Two abdominal radiologists blinded to pathologic findings evaluated in consensus all images retrospectively. They evaluated the following morphologic features: location of the tumor in the pancreas (head, body, or tail); maximal transverse diameters of the tumor; shape (round, oval, or lobulated); and presence of calcifications. Internal density or signal intensity characteristics of the tumor were compared with those of surrounding pancreas and were described as hypo-, iso-, or hyperdense or intense and as homogeneous or heterogeneous. If areas of water density or signal intensity were seen, the tumor was classified as cystic. The fraction of tumor composed of cystic material versus solid material was estimated and expressed as a percentage. Finally, the enhancement pattern of the tumor (hypovascular or hypervascular compared with normal pancreas) was evaluated in nine patients. Images were also evaluated for common bile and pancreatic duct obstruction; vascular invasion; and tumor spread to regional lymph nodes, adjacent solid organs, and other abdominal sites.
Pathologic Examination and Analysis
Two pathologists, one at each institution, reviewed the gross tumor
specimen and H and E–stained microscopic slides. CT and MRI features
were correlated with gross pathologic and histologic findings in each case. In
addition to morphologic characteristics of the tumor, pathologists recorded
the presence of lymph node metastases, bile or pancreatic duct dilatation,
invasion of visceral vessels, and involvement of surrounding organs.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Both reviewers agreed in all cases for all MRI features studied. On unenhanced CT and MRI, the masses appeared homogeneous in five cases (45%) and heterogeneous in six cases (55%). Tumors appeared completely solid in five cases (Fig. 2A, 2B) and cystic in six. The cystic areas composed more than 75% of the mass in four cases, between 50% and 75% in one, and less than 25% in one. All cystic masses had at least one solid component, usually in the periphery (Fig. 3). The mean diameters of solid tumors and tumors with cystic areas were 3.5 and 10.1 cm, respectively. Calcifications were observed in three masses (27%) (small and punctate in two, thick and peripheral in one [Fig. 4A, 4B]). One mass showed hyperintense areas on T1-weighted images suggestive of hemorrhage. Intratumoral hemorrhage could not be evaluated in six patients who had only enhanced CT.
|
|
|
|
|
On enhanced CT, four (100%) of four solid masses enhanced homogeneously but less than the surrounding normal pancreas. Five (100%) of five cystic masses that underwent contrast-enhanced imaging showed homogeneous enhancement of the peripheral solid components. These solid components enhanced less than surrounding pancreas (Fig. 5A, 5B, 5C, 5D, 5E).
|
|
|
|
|
One tumor examined on MRI was solid and slightly hypointense on T1-weighted and hyperintense on T2-weighted images in comparison with pancreatic parenchyma. It enhanced homogeneously but less than surrounding pancreatic parenchyma (Fig. 6A, 6B, 6C, 6D). The second tumor was partially cystic and well marginated. It had a large central necrotic area with mixed T1 signal intensity and hyperintense T2 signal intensity.
|
|
|
|
Pancreatic and bile duct dilatation were seen in three and two patients, respectively. In two patients, both pancreatic and bile ducts were dilated. Four of five masses located in the pancreatic head showed duodenal invasion (Fig. 7A, 7B). In one patient, the duodenum was compressed by the mass but had not been invaded. On CT and MRI, the tumor was shown to have invaded the stomach (n = 1), transverse mesocolon (n = 1), portal venous confluence (n = 1), and splenic vein (n = 1). One patient had a subcentimeter indeterminate liver lesion at initial presentation that was proven to be a metastasis at surgery; one patient developed a liver metastasis during follow-up. No lymphadenopathy was detected.
|
|
Correlation with Pathologic Findings: Gross Morphologic and Immunohistochemical Features
The relative proportions of solid and cystic areas varied greatly among the
tumors (from totally solid to markedly cystic) but correlated well with CT and
MRI findings (Fig. 8A,
8B,
8C,
8D). Six tumors had necrotic
areas, and two had hemorrhagic areas, one of which was not visible on CT. The
duodenum was invaded by tumor in four patients and displaced markedly without
invasion in one patient, as could be predicted from the preoperative imaging.
Focal splenic artery invasion with tumor thrombus (n = 1), invasion
of the portal–splenic venous confluence (n = 1), invasion of
the adjacent stomach and transverse mesocolon including midcolic vessels
(n = 1), and focal lymph node metastasis (n = 1) were noted
during pathologic evaluation. All these findings, excluding splenic artery
invasion, were visible on CT or MRI.
|
|
|
|
On microscopic examination, all tumors consisted of cells arranged in nests
and acinar structures with lobulation and thin strands of fibrovascular
stroma. Tumor cells had focally abundant apical eosinophilic cytoplasm with
round to slightly irregular basally located nuclei. Results of a periodic
acid–Schiff (PAS) stain with diastase digestion were available in seven
cases and revealed fine zymogen granules in the cytoplasm of the tumor cells.
Results of immunostaining were positive for 
-1-antichymotrypsin
(n = 6), -1-antitrypsin (n = 5), keratin (CAM5.2 or
AE1/AE3) (n = 5), and lipase (n = 1) antibodies. Results of
synaptophysin and chromogranin staining were focally positive in some cases.
Tumor cells were negative for carcinoembryonic antigen (CEA) and mucicarmine
staining. These immunohistochemical results are consistent with the diagnosis
of acinar cell carcinoma. Electron microscopic evaluation revealed zymogen
granules in four patients, which also supports the diagnosis of acinar cell
carcinoma.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although no sex predilection was observed in our series, the tumor has been reported to be more frequent in men [1, 3]. Excluding children and adolescents, the mean age of patients in the largest series to date was 60 years [3], which is similar to that of our series. There seems to be no racial predilection. However, whites constitute most of the patients in clinicodemographic studies and in our patient population [1, 3].
Patients usually present with symptoms related to either local mass effect or metastases [1]. Symptoms usually are nonspecific and include abdominal pain, loss of appetite, weight loss, nausea, and vomiting [1–3]. The most common presenting symptom in our series was abdominal pain (55%), similar to the findings in prior reports [3]. The most common clinical sign at presentation is a palpable abdominal mass, whereas jaundice is considered rare. Although only one patient in our series presented with a liver metastasis, approximately half of the patients in prior series had metastases at presentation, with the liver being the most common site [3]. Extraabdominal metastases are rare [1].
Interestingly, acinar cell carcinoma can cause hyperlipasemia, which may
lead to diffuse subcutaneous nodules and polyarthropathy
[1,
3,
5,
6,
14]. Subcutaneous nodules are
generally widely distributed, erythematous, and painful and may be
misinterpreted as erythema nodosum or as metastases
[3,
6]. Arthropathy is caused by
periarticular fat necrosis and involves peripheral joints such as the ankles,
knees, wrists, and small joints of the hands and feet. Radiographs of the
osseous lesions typically show multiple lytic areas that might be mistaken for
metastases involving both cancellous and cortical bones
[6]. These lesions are often
located beneath subcutaneous lesions. Although frequently reported, these
lipase-induced stigmata are uncommon with acinar cell carcinoma
[2,
3]. In the series by Holen et
al. [3], they were seen in less
than 10% of patients. Only one of our patients (9%) presented with signs and
symptoms of hyperlipasemia. In addition to an elevated serum lipase level,
some patients with acinar cell carcinoma also may have an elevated serum
amylase level, peripheral eosinophilia, and a markedly elevated serum
-fetoprotein level [1,
15]. In fact, in our series,
elevated levels of serum lipase and amylase were found in three (27%) and two
patients (18%), respectively. An isolated case of acinar cell carcinoma
secreting insulinlike growth factors that caused hypoglycemia has also been
reported in the literature
[16].
Tumors that are amenable to surgical resection and are 10 cm or smaller are associated with a longer patient survival than are larger and unresectable tumors [2]. Older age (> 60 years), the presence of symptoms of lipase secretion, the location of tumor in the head of the pancreas, and the presence of metastasis at presentation are associated with a decreased chance of survival [2]. Radiation and chemotherapy are indicated only for palliative purposes [1].
In our series, a particular tumor location did not predominate, although the pancreatic head was the most common site. Approximately half of the tumors were also located in the head of the pancreas in the series by Holen et al. [3]. Most tumors in our series were well marginated (91%) and partially or completely exophytic (82%). These features can be used to differentiate acinar cell carcinoma from other pancreatic tumors because they are rare in more common pancreatic neoplasms.
Tumors in our series were slightly smaller (7.1 cm) than those described in prior reports (10.6 cm) in mean diameter [1, 2]. Tumors were homogeneous and solid when small. Five tumors in our series were solid, and all of them were less than 5 cm in largest diameter (mean diameter, 3.5 cm). Although cystic changes have been reported to be rare in a series of 28 patients [2], six of our patients (55%) had tumors with a varying degree of cystic change. In four, cystic areas constituted more than 75% of the tumor. Another interesting finding was that tumors with cystic areas (mean diameter, 10.1 cm) were substantially larger than purely solid ones (mean diameter, 3.5 cm). This finding is likely due to greater necrosis in the larger tumors resulting from impairment of the blood supply.
Limited published data are available regarding the enhancement pattern of acinar cell carcinoma [9, 10, 15]. In our series, the tumors typically enhanced homogeneously but less than surrounding pancreatic parenchyma. However, the patients were not routinely examined with arterial phase imaging. Mustert et al. [9] reported a case of acinar cell carcinoma that was hypervascular on arterial phase imaging.
On histopathologic examination, pure acinar cell carcinoma has two
predominant cellular patterns of growth: an acinar pattern consisting of cells
growing in well-formed acini and a solid pattern characterized by sheets and
cords of cells in a fibrovascular stroma
[3]. PAS staining after
diastase digestion characteristically reveals PAS-positive granules
corresponding to zymogen granules
[2]. Mucin stains are typically
negative. Acinar cell carcinoma displays a unique immunochemical staining
pattern: strongly positive for the digestive enzymes of exocrine pancreas such
as trypsin, chymotyripsin, lipase, and phospholipase A2 and
negative or only focally positive for neuroendocrine markers such
synaptophysin, chromogranin, glucagon, somatostatin, gastrin, and vasoactive
intestinal peptide [2]. Results
of keratin stains are always positive especially when the CAM5.2 antibody is
used. Another acinar cell secretory product, -1-antitrypsin, is
diffusely expressed in most tumors
[1]. In our series, all tumors
showed acinar differentiation by these criteria. In addition, staining for
mucicarmine and CEA produced negative results. Results for mucicarmine and CEA
are usually positive in ductal adenocarcinoma but not in acinar cell
carcinoma.
Immunohistochemical detection of a minor endocrine component in acinar cell carcinoma is not rare [1–3, 12, 17, 18], and we detected such a component in a minority of the cases in our series. Klimstra et al. [2] detected minor endocrine components in 42% of their cases. Tumors displaying both acinar and endocrine features that constitute more than 25% of the cells have been termed "mixed acinar–endocrine carcinomas." They are otherwise histologically similar to the pure acinar cell carcinomas and have been included in other published pathologic series studying acinar cell carcinoma [2, 3]. We also identified two cases of mixed acinar–endocrine carcinoma and one case of mixed acinar–ductal origin; however, we chose not to include these cases in our study group.
The radiologic differential diagnosis of acinar cell carcinoma includes ductal adenocarcinoma, neuroendocrine tumor, solid and pseudopapillary tumor, pancreaticoblastoma, mucinous cystic neoplasm, and pseudocyst [1]. It is important to differentiate these neoplasms because treatment and prognosis differs significantly for these various entities. Pancreatic ductal adenocarcinoma is the most common primary pancreatic malignancy. This tumor is usually smaller than acinar cell carcinoma and virtually never contains calcification or cystic degeneration [19]. Unlike acinar cell carcinoma, pancreatic ductal adenocarcinoma is not well marginated and is almost always locally invasive. Neuroendocrine tumor is typically more vascular than acinar cell carcinoma and, therefore, may enhance more than the pancreatic parenchyma; this tumor shows clinical evidence of abnormal hormone secretion. However, nonfunctioning endocrine tumors may present as large well-marginated masses with internal hemorrhagic-cystic areas and thus may not be distinguishable from acinar cell carcinoma [20]. Solid and pseudopapillary tumors may also mimic acinar cell carcinoma. These tumors are well-marginated, large, encapsulated tumors with solid and cystic areas; however, they are seen almost exclusively in young women, in which acinar cell carcinoma only rarely occurs, and have a better prognosis [21]. Pancreatoblastomas are extremely rare malignant epithelial tumors with acinar differentiation and are histologically similar to acinar cell carcinoma [1]. However, pancreatoblastomas usually occur in infants and children [1]. They are more aggressive than acinar cell carcinoma and often present with liver metastases. Finally, thick-walled pancreatic pseudocysts may resemble an acinar cell carcinoma, as in one of our patients. However, a history of pancreatitis is almost always present.
In conclusion, pure acinar cell carcinoma of the pancreas, although rare, has distinctive CT and MRI features that allow radiologists to render an accurate diagnosis. It typically presents as a well-marginated, exophytic mass that enhances homogeneously and less than surrounding pancreas when small and contains cystic areas due to necrosis when large.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  Y. Seki, T. Okusaka, M. Ikeda, C. Morizane, and H. Ueno Four Cases of Pancreatic Acinar Cell Carcinoma Treated with Gemcitabine or S-1 as a Single Agent Jpn. J. Clin. Oncol., November 1, 2009; 39(11): 751 - 755. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. A. Sahni and K. J. Mortele The Bloody Pancreas: MDCT and MRI Features of Hypervascular and Hemorrhagic Pancreatic Conditions Am. J. Roentgenol., April 1, 2009; 192(4): 923 - 935. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. M. Chung, M. D. Travis, and R. M. Conran From the Archives of the AFIP: Pancreatic Tumors in Children: Radiologic-Pathologic Correlation. RadioGraphics, July 1, 2006; 26(4): 1211 - 1238. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
