The utility of a novel antibody in the pathological diagnosis of pancreatic acinar cell carcinoma
- Makiko Yasumoto1,
- Masato Hamabashiri2,
- Jun Akiba1,
- Sachiko Ogasawara1,
- Yoshiki Naito1,
- Tomoki Taira3,
- Masamichi Nakayama1,
- Aya Daicho2,
- Fumio Yamasaki4,
- Kazuhide Shimamatsu5,
- Yusuke Ishida6,
- Ryohei Kaji6,
- Yoshinobu Okabe6,
- Osamu Nakashima1,
- Koichi Ohshima1,
- Manabu Nakashima2,
- Michio Sata6,
- Hirohisa Yano1
- 1Department of Pathology, Kurume University School of Medicine, Kurume, Japan
- 2Department of Immunological and Molecular Pharmacology, Faculty of Pharmaceutical Science, Fukuoka University, Fukuoka, Japan
- 3Department of Diagnostic Pathology, Kurume University Hospital, Kurume, Japan
- 4Department of Pathology, Saga Medical School, Faculty of Medicine, Saga University, Saga, Japan
- 5Division of Histopathology, Omuta City General Hospital, Omuta, Japan
- 6Division of Gastroenterology, Department of Medicine, Kurume University School of Medicine, Kurume, Japan
- Correspondence to Dr Jun Akiba, Department of Pathology, Kurume University of Medicine, 67 Asahi-machi, Kurume 830-0011, Japan;
Contributors Makiko Yasumoto have a central role in immunohistochemical staining and these assessment. Jun Akiba and Sachiko Ogasawara led immunohistochemical staining and these assessment. Masato Hamabashiri, Aya Daicho and Manabu Nakashima participated in the process of establishing this novel antibody. Fumio Yamasaki and Kazuhide Shimamatsu were provided specimens of the pancreatic tumors. Yusuke Ishida, Ryohei Kaji and Yoshinobu Okabe were provided samples, such as pancreatic juice etc. Yoshiki Naito, Tomoki Taira, Masamichi Nakayama, Osamu Nakashima and Koichi Ohshima participated in immunohistochemical staining. Michio Sata and Hirohisa Yano organized this mamanuscript.
- Accepted 16 November 2011
- Published Online First 7 January 2012
Aims Acinar cell carcinomas (ACCs) are rare tumours of the exocrine pancreas accounting for about 1–2% of all pancreatic neoplasms in adults. It is therefore difficult to come across a large number of ACC cases in a single medical institution, and only a few serial studies have been published. Since ACCs present a wide variety of morphological patterns, immunohistochemical analysis is useful. In this study, the authors established a novel monoclonal antibody 2P-1-2-1 by means of a subtractive immunisation method.
Methods Immunohistochemical staining was performed using 50 primary pancreatic tumors, including 7 ACCs, 7 neuroendocrine tumours (NETs), 5 solid-pseudopapillary neoplasms (SPNs), and 31 ductal carcinomas and organs other than the pancreas.
Results Non-neoplastic acinar cells were stained diffusely, but epithelial cells of the pancreatic duct and the islets of Langerhans were not stained. In pancreatic tumours, all the seven ACCs were diffusely positive for the 2P-1-2-1 antibody. However, no positive staining was found in other pancreatic tumours including NETs, SPNs and ductal adenocarcinomas. The sensitivity and specificity of the 2P-1-2-1 antibody for ACCs were both 100%. In other organs studied, positive staining was observed only in the ectopic pancreas.
Conclusions It was shown that the 2P-1-2-1 antibody specifically stained the pancreatic acinar cells and tumours of acinar cell origin, such as ACCs. Although it remains unclear at this time to which proteins the monoclonal antibody 2P-1-2-1 is directed, it is suggested to be useful for the pathological diagnosis of ACCs and for the exclusion of other pancreatic tumours.
- gastrointestinal disease
- monoclonal antibody
- image analysis
- pancreatic cancer
- cancer stem cells
Tumours of pancreatic origin are classified as tumours of the exocrine pancreas or the endocrine pancreas in the WHO classification.1 Exocrine tumours are composed of ductal adenocarcinomas (DCs), acinar cell carcinomas (ACCs) and solid-pseudopapillary neoplasms (SPNs). Endocrine tumours include neuroendocrine neoplasms.
ACCs are rare pancreatic tumours accounting for approximately 1–2% of pancreatic exocrine tumours in adults.1–3 It is therefore difficult to come across a large number of ACC cases in a single medical institution, and series reports are limited in number. Therefore, the nature of ACCs is poorly understood.
ACCs can exhibit a wide spectrum of histological features. The most characteristic is the acinar pattern. The other common patterns are the solid pattern or trabecular pattern.1 These different patterns frequently coexist in the same tumour, making it difficult to reach a definite pathological diagnosis with H&E stain alone. ACCs that show acinar cell differentiation have zymogen granules in the cytoplasm in general, and the presence of these zymogen granules has proven useful for pathological diagnosis. Traditionally, the diagnosis of ACCs has required periodic acid-Schiff (PAS) stain and diastase digestion PAS (dPAS) stain to confirm the presence of dPAS-positive granules in the cytoplasm,1 4 5 or electron microscopy to confirm the presence of zymogen granules in tumour cells.1 4 6 Recently, it has become possible to immunohistochemically demonstrate the presence of pancreatic enzymes in the cytoplasm.1 3 4 7
In this study, we established a novel monoclonal antibody 2P-1-2-1 using pancreatic juice collected during endoscopic retrograde pancreatography performed for diagnostic purposes. Using the 2P-1-2-1 antibody, we conducted the immunohistochemical (IHC) analysis in various pancreatic tumours.
Materials and methods
Establishment of a novel antibody
We followed an immunological approach to establish novel monoclonal antibodies specific for the pancreatic juice from patients, which would be potentially useful in the diagnosis of pancreatic disorders. Mice were immunised by subtractive immunisation using mixed pancreatic juices from chronic pancreatitis and DC patients as the tolerogen and the immunogen, respectively.8 This antibody, 2P-1-2-1, detected a protein with a molecular weight of 120 kDa in western blot analysis.9 A novel monoclonal antibody 2P-1-2-1 was used in the following examination.
Patients and tumour samples
The pancreatic tumours used in this study included 7 ACCs, 7 neuroendocrine tumours (NETs), 5 SPNs and 31 DCs. Tissue samples obtained from preoperative endoscopic ultrasound-guided fine needle aspiration (EUS-FNA) were also included in one out of seven ACCs. In addition, a series of other organs including five salivary glands, five breast carcinomas (all invasive ductal carcinomas containing normal areas), five prostatic carcinomas (all adenocarcinomas of different histological grades containing normal areas), seven gastric carcinomas (four adenocarcinomas and three poorly cohesive carcinomas containing normal areas), seven colonic carcinomas (adenocarcinomas of different histological grades containing normal areas), five extranodal marginal zone lymphoma of mucosa-associated lymphoid tissue (MALT lymphoma; biopsy specimens of stomach) and seven ectopic pancreas tissues were also examined.
The tissues examined in this study were fixed in 10% buffered formalin. Sections were embedded in paraffin and cut into 4 μm thick slices. IHC staining was performed with the BenchMark XT (Ventana Automated Systems, Inc, Tucson, Arizona, USA). The antibodies used in the present study are listed in table 1.
Expression of antibodies was graded on four levels according to the distribution of the immunoreactive tumour area, that is, 0 when positive cells were present in <10% of the entire area, 1+ when the area was 10–39%, 2+ when 40–69% and 3+ when 70–100%.
Staining intensity was graded on three levels, that is, 0 when the intensity in the tumour area was negative, +1 when the intensity in the tumour area was less than that of normal pancreatic acinar cells and +2 when the intensity in the tumour area was almost equal to that of normal pancreatic acinar cells. The total score was obtained as the distribution grade added to the staining intensity score. This total score was then grouped into two levels, that is, 0–1, negative, and 2–5, positive. The IHC analysis was evaluated independently by two pathologists (MY and JA).
Data are expressed as the mean±SD. Comparisons between groups were performed using the Mann–Whitney U test. Differences were considered significant at p<0.05.
Clinicopathological characteristics of ACCs
The clinicopathological characteristics of the seven patients diagnosed with ACCs are described in table 2.
The preoperative diagnosis included DC, ACC, NET, SPN and/or cystic neoplasm. Electron microscopic examination was performed for two cases and revealed zymogen granules in the cytoplasm.6 The remaining cases were diagnosed based on morphological and IHC analyses.
Findings of IHC analysis
The acinar cells in non-neoplastic pancreatic tissue stained diffusely with the 2P-1-2-1 antibody. Both cytoplasmic dot-like and membranous immunoreactivity were observed in acinar cells. The epithelial cells of the pancreatic duct and the islets of Langerhans showed completely negative staining (figure 1).
The results of the expression score in ACCs, NETs, SPNs and DCs are summarised in table 3. In all the ACC cases, the antibody 2P-1-2-1 was diffusely and strongly positive in tumour cells identically with background normal tissue, with a sensitivity of 100% (figure 2). The location of the positive signal of antibody 2P-1-2-1 was observed in the cytoplasm and cell membrane of tumour cells. The pancreatic enzyme trypsin was positive in 5/7 (71.4%) cases, followed by amylase in 3/7 (42.9%) and lipase in 1/7 (14.3%). Meanwhile, protease inhibitors alpha-1-antitrypsin (α1-AT) and alpha-1-antichymotrypsin (α1-ACT), which were also used in our investigation, had a high sensitivity for ACCs in 7/7 (100%) cases. Cytokeratin AE1/AE3, a marker of epithelial cells, was observed in 5/7 (71.4%) cases. CD56, a marker of neuroendocrine differentiation, was negative in all cases, while synaptophysin and chromogranin A were both positive in 1/7 (14.3%) cases. The antibody 2P-1-2-1 had a mean staining score of 5 and stained diffusely and strongly, with statistical significance compared with other antibodies except α1-ACT (figure 3).
In seven NETs, the antibody 2P-1-2-1 was completely negative in all cases and had a specificity of 100% for NETs. Trypsin was negative in all cases, whereas lipase, amylase, α1-AT and α1-ACT were positive in 1/7, 2/7, 4/7 and 3/6 cases, respectively. Trypsin had a specificity of 100%. By contrast, the specificity of lipase, amylase, α1-AT and α1-ACT were 85.7% (6/7), 57.1% (5/7), 42.9% (3/7) and 50.0% (3/6), respectively. These antibodies had a lower specificity than the antibody 2P-1-2-1.
In five SPNs, the antibody 2P-1-2-1 was negative for tumour cells in all cases and had a specificity of 100%. Trypsin, lipase and amylase were all negative (specificity 100%). Specificity of both α1-AT and α1-ACT was 20% (1/5) and that of AE1/AE3, chromogranin A, synaptophysin and CD56 was 60% (3/5), 80% (4/5), 20% (1/5) and 0% (0/5), respectively.
Although DCs are less difficult to distinguish from other pancreatic tumours with H&E stain, the antibody 2P-1-2-1 did not exhibit significant staining of neoplastic pancreatic ductal epithelium in 31 DCs (specificity 100%) (figure 4).
Preoperative EUS-FNA was performed in one case (case 4 in table 3) out of seven ACCs. The biopsy specimen was sparse and relatively altered with haemorrhage and fibrin deposition in the background. The tumour was composed of cuboidal cells with acinar arrangement (figure 5A). dPAS-positive granules were observed in the cytoplasm of tumour cell. IHC findings of the biopsy specimen were almost the same as that of the resected tissue described in table 3. These findings were compatible with ACC. In addition, the biopsy specimen was diffusely positive for antibody 2P-1-2-1 (figure 5B).
The results of IHC findings for various organs except pancreas are summarised in table 4.
Using the novel antibody 2P-1-2-1 produced by patients' pancreatic juice, we investigated its expression in pancreatic tumours using an immunohistological approach. The study demonstrated that the antibody 2P-1-2-1 was useful as a marker of pancreatic acinar cells because it specifically stained pancreatic acinar cells and cells showing pancreatic acinar differentiation. Moreover, this antibody was also very useful for pathological diagnosis of pancreatic tumours because of its high sensitivity and specificity for ACCs.
With the recent advancement of imaging technology, clinical diagnostic accuracy of pancreatic tumours has been further improved. On imaging, DCs exhibit typical imaging features of infiltrative growth. Since pancreatic tumours other than DCs occur infrequently and may often not show typical imaging features, pathological examination is recommended for definitive diagnosis. The histological differential diagnosis of ACCs includes NETs, SPNs and pancreatoblastomas, and immunohistological investigation has proven useful for the differentiation of these tumours.10–12 Currently, immunostaining using antibodies related to pancreatic exocrine enzymes such as trypsin, lipase, chymotrypsin and amylase has proven useful in the diagnosis of ACCs.
Among the large variety of pancreatic enzymes that have been investigated, trypsin has been found to be one of the most commonly expressed in ACCs. Trypsin has been reported to be expressed at a frequency of 71%13 to 100%,3 7 and another pancreatic enzyme chymotrypsin has been also shown to have a high sensitivity (85%13 to 100%7). However, the sensitivity varies widely. In the review by Ordonez NG,14 the sensitivity of trypsin, lipase, chymotrypsin and amylase was reported to be 97%, 84.5%, 66.1% and 14.3%, respectively. In our study, in contrast to previous reports, the sensitivity of trypsin was 71.4%, but that of lipase and amylase was 14.3% and 42.9%, respectively. The reasons for these discrepancies have been attributed to differences in the antibodies used in the studies.3
α1-AT and α1-ACT are two antigens that have often been reported to be expressed in ACCs and it has been suggested that they may be of some value in the diagnosis of this tumour. α1-AT and/or α1-ACT have been reported to be expressed in a wide variety of pancreatic tumours, including DCs, NETs, SPNs and pancreatoblastomas.3 5 15 16 These markers are also expressed in adenocarcinomas of the lung and stomach, renal cell carcinoma17 and hepatocellular carcinoma.18 Although α1-AT and α1-ACT showed a high sensitivity also in our study (α1-AT 100%, α1-ACT 100%), these two antibodies also expressed other pancreatic tumours. The specificity of these antibodies for ACCs was low (α1-AT 33.3%, α1-ACT 36.4%). Thus, it might be difficult to make a definitive diagnosis of ACCs based on only these markers.
More recently, BCL10 antibody has been identified as a useful antibody for a pathologic diagnosis of ACCs.1 8 19 20 BCL10 was closely associated with MALT lymphomas and a protein with a molecular weight of 105 kDa.19 21 We therefore performed IHC examination of MALT lymphomas using this 2P-1-2-1 antibody, but no staining was observed. Moreover, molecular weights of trypsin, lipase, amylase, α1-AT and α1-ACT were 23, 48, 54, 52 and 68 kDa, respectively. The protein recognised by the antibody 2P-1-2-1 has a molecular weight of 120 kDa and thus it may be inferred that the antibody 2P-1-2-1 recognises proteins other than BCL10 and other pancreatic enzymes. In our literature research using PubMed, the protein with a molecular weight of 120 kDa was not found in the component of pancreatic acinar cell. At this time, however, it is not known which proteins the antibody 2P-1-2-1 recognises.
Recently, EUS-FNA has been widely used to assist the histological diagnosis of pancreatic tumours. In EUS-FNA, ACCs are usually highly cellular,22–24 and tissue samples are frequently sparse and sometimes small in volume, which may make it very difficulty to diagnose pancreatic tumours. Our series included a case of ACC that was diagnosed using EUS-FNA samples. This sample obtained by EUS-FNA showed diffuse immunoreactivity for the antibody 2P-1-2-1. Thus, the antibody 2P-1-2-1 is expected to be useful for pathological diagnosis of a small amount of sample obtained using EUS-FNA.
In conclusion, it is not known which proteins the antibody 2P-1-2-1 recognises, so further investigations should be carried out to determine the proteins to which this antibody is directed. However, the antibody 2P-1-2-1—established by means of a subtractive immunisation method using pancreatic juice obtained for diagnostic purposes, which recognises normal pancreatic acinar cells or cells showing acinar cell differentiation and shows high sensitivity and specificity for pancreatic ACCs—is useful for the pathological diagnosis of ACCs and for the exclusion of other pancreatic tumours.
The antibody 2P-1-2-1 is useful for the pathological diagnosis of acinar cell carcinomas and for the exclusion of other pancreatic tumours.
The antibody 2P-1-2-1 is expected to useful for pathological diagnosis of a small amount of sample obtained using endoscopic ultrasound-guided fine needle aspiration.
The authors gratefully thank Ms Sachiyo Maeda, Akiko Tanaka and Akemi Fujiyoshi for their excellent technical assistance and Ms Mika Hanada for manuscript preparation.
Funding This work was supported by the Japan Society for the Promotion of Science (JPSP) KAKENHI (Grant-in-Aid Challenging Exploratory Research: 21659151).
Competing interests None.
Patient consent Obtained.
Ethics approval This study was approved by the ethical committee of Kurume University (09034).
Provenance and peer review Not commissioned; externally peer reviewed.