Search

DOG1, a transmembrane calcium-regulated chloride channel protein, is a sensitive and specific marker for gastrointestinal stromal tumors compared with other spindle cell and epithelioid neoplasms. Overexpression has also been described in a variety of both benign and malignant epithelial neoplasms. Recently, DOG1 immunoreactivity has been reported in pancreatic solid pseudopapillary tumors (SPT), suggesting a role as a marker for SPT. Utilizing immunohistochemistry, we evaluated DOG1 expression in pancreatic neoplasms to determine the prevalence of staining and establish diagnostic utility. Multiple tissue microarrays (TMA) were created from cores of formalin-fixed paraffin-embedded blocks containing pancreatic adenocarcinomas (n=112), neuroendocrine tumors (n=99), serous cystadenomas (n=28), and SPT (n=14) as well as normal pancreas (n=12). Immunoreactivity for DOG1 (clone K9) was assessed for intensity (1 to 3+), percentage of tumor positivity and location. Of the 99 cases of neuroendocrine tumors, only 2 (2%) were focally positive. Patchy staining was identified in 8 cases (7%) of adenocarcinoma of 1 to 2+ intensity, involving 15% to 80% of the tumor cells and primarily seen in a membranous and luminal distribution. In contrast to a previous report, no DOG1 positivity was observed in SPT, evaluated by both TMA and full sections. The TMAs of serous cystadenomas and normal pancreas were negative for DOG1. Rarely, pancreatic islets displayed granular, cytoplasmic staining. DOG1 antibody clone K9 is not a useful marker for SPT or other primary pancreatic neoplasms. Additional studies may be helpful to evaluate differences between clones of DOG1.

The diagnosis of serous microcystic adenoma (SMA) is usually straightforward. For small biopsies and/or unusual variants, the differential diagnosis includes other pancreatic or metastatic neoplasms showing cystic or clear cell features. We evaluated immunostains for potential use in the diagnosis of SMA. Cases of SMA were identified from archival files. Tissue cores (2 per block) were arrayed to create a microarray of cores measuring 2mm each. Additionally, microarrays previously constructed from 56 pancreatic adenocarcinomas (PACs) and 64 pancreatic endocrine tumors (PENs) were studied. The microarrays were stained with calponin, chromogranin, CD10, alpha-inhibin, and monoclonal neuron-specific enolase (m-NSE). Subsequently, some were stained with MUC6, melan-A, D2-40, h-caldesmon, smooth muscle actin, and smooth muscle myosin. For SMAs, staining was seen with calponin (85.2%), alpha-inhibin (96.2%), and m-NSE (96.2%). Focal weak staining was seen with MUC6 (65%). All SMAs were negative with chromogranin, CD10, melan-A, D2-40, h-caldesmon, smooth muscle actin, and smooth muscle myosin. In contrast, calponin was negative in all PACs and PENs. Staining for alpha-inhibin was absent in PACs and present in 4.1% of PENs; whereas immunoreactivity for m-NSE was present in 26.8% of PACs and 73.7% of PENs. Chromogranin staining was present in 9.1% of PACs and 100% of PENs. An immunohistochemical profile of staining with calponin, alpha-inhibin, and m-NSE and absent staining with chromogranin supports the diagnosis of SMA, and distinguishes SMA from PAC and PEN. Calponin and alpha-inhibin are the most useful positive markers for SMA, and are negative in most entities in the differential diagnosis.

Mucin expression patterns have been studied in tumors from various sites. Previous studies have shown an association of MUC1 with poor prognosis and MUC2 and MUC5AC with a mucinous phenotype. The pattern of mucin expression in endometrial carcinomas has not been documented in a large series. We determined the mucin expression profile in endometrial carcinomas and evaluated the relationship between mucin expression and clinical-pathologic parameters. A tissue microarray of 310 cases of endometrial carcinoma with known clinical outcome was constructed from formalin-fixed, paraffin-embedded donor blocks using two 0.6 mm cores from each tumor. Sections were stained with monoclonal antibodies against MUC1, MUC2, MUC4, MUC5AC, and MUC6. Staining was considered positive if greater than or equal to 5% of cells stained positively in either core. Mucin expression was correlated with tumor type, histologic grade, International Federation Gynecology and Obstetrics stage, lymph node involvement, depth of myometrial invasion, patient age, ethnicity, and clinical outcome. MUC1 was expressed in 267/310 (86.1%) of endometrial carcinomas, MUC2 in 2/310 (0.6%), MUC4 in 73/310 (23.5%), MUC5AC in 1/310 (0.3%), and MUC6 in 4/310 (1.2%). Endometrioid endometrial carcinoma showed a higher rate of MUC1 expression than nonendometrioid endometrial carcinoma (227/258, 88.0% vs. 39/52, 75.0%, P=0.01). No significant differences in any of the mucins were noted among the other end points evaluated. Mucin expression did not correlate with tumor grade, stage, or patient outcome.

Many patients with pancreatic carcinoma have stent placement for biliary obstruction before resection. Stent-associated atypia, found in common bile duct (CBD) margins at the time of resection, may be confused with malignancy. We evaluated histologic and immunohistochemical changes in CBD margins from resection specimens for pancreatic carcinoma. Histologic findings in CBDs, including ulcer and inflammation; epithelial metaplasia, atypia, and gland complexity; and increased wall thickness, nerve entrapment, and smooth muscle content, were compared in 30 stented and 31 nonstented CBD margins from pancreaticoduodenectomies for carcinoma and 13 normal CBDs from autopsy material. The proliferation index was calculated for stented and nonstented CBDs after Ki-67 immunohistochemical staining. Immunostaining for Ki-67, p53, and c-erbB-2 was performed in stented CBDs and corresponding carcinomas. All the histologic changes occurred more frequently in stented and nonstented CBD margins from carcinoma patients than in normal CBDs. Stented CBDs had significantly increased epithelial changes and Ki-67 proliferation rate as compared with nonstented CBDs. The stented CBDs had significantly less p53 and c-erbB-2 expression as compared with corresponding pancreatic carcinomas. Caution should be applied when interpreting atypia in CBD margins from patients with a history of CBD stenting. Changes found in stented CBDs are characteristic, and in most cases can be distinguished from malignancy. In difficult cases, immunohistochemistry may be useful.

Three separate groups, using MRI, have reported basal ganglia abnormalities in Tourette's syndrome (TS). We found similar abnormalities in boys with attention-deficit/hyperactivity disorder (ADHD). Because TS and ADHD are frequently comorbid, we contrasted ADHD boys with and without TS along with control subjects. As expected, we found a significant loss of the normal globus pallidus asymmetry in the patients, but presence or absence of TS did not differentiate the ADHD groups. We conclude that accounting for ADHD comorbidity will be important in future TS morphometric studies.

Analysis of cerebral magnetic resonance images of 24 subjects with Sydenham's chorea and 48 age-, height-, weight-, gender-, and handedness-matched controls demonstrated increased sizes of the caudate, putamen, and globus pallidus in the Sydenham's chorea group. In contrast, neither total cerebral, prefrontal, or midfrontal volumes or thalamic area were increased. These results indicate the selective involvement of the basal ganglia in Sydenham's chorea.

Staging laparotomy was performed at the University of Virginia Medical Center on 111 patients with Hodgkin's disease. The operation included multiple liver and lymph node biopsies and, excepting three patients, splenectomy. The histopathology was reviewed and the 111 patients were classified as follows: nodular sclerosis, 74; mixed cellularity, 28; lymphocyte predominance, 7; and undetermined, 2. There were no deaths. Wound, pulmonary or urinary tract complications occurred in 11 patients. One case of postoperative thrombophlebitis occurred and in another case small bowel obstruction developed, and resolved without reoperation. The pathologic stage (PS) following laparotomy was unchanged from the clinical stage (CS) in 64%, reduced in 20%, and advanced in 16%. The therapy, however, was altered in 38% of the patients. Lymphangiography in 103 patients was interpreted as showing lymph node involvement in 38, equivocal involvement in 11, and no involvement in 54. Among the 92 examinations reported as either positive or negative, 77% were confirmed histopathologically, 21% were falsely positive, and 2% were falsely negative. The spleen was positive for Hodgkin's disease in 39% of cases, and in these patients with positive spleens there was no reason to suspect intra-abdominal involvement preoperatively in 21%.