Consistent and selective expression of the discoidin domain receptor-1 tyrosine kinase in human brain tumors.

OBJECTIVE: Few molecular targets are both consistently and selectively expressed in a majority of central nervous system (CNS) neoplasms. Receptor tyrosine kinases have been implicated in brain tumor oncogenesis. We previously isolated one such receptor, discoidin domain receptor-1 (DDR1), from high-grade pediatric brain tumors. Here, we analyze the cellular origin and distribution of DDR1 expression in human brain tumors and its expression in tumor cells relative to surrounding brain. METHODS: By use of a digoxigenin-labeled DDR1 riboprobe, we investigated the expression of DDR1 messenger ribonucleic acid in a prospective series of 30 resected human primary and metastatic brain neoplasms, nonneoplastic human brain, and mouse embryonic brain, as well as a mouse glioblastoma model, by in situ hybridization. RESULTS: All the high-grade primary brain and metastatic brain tumors showed unequivocal, intense DDR1 expression within the majority of tumor cells, whereas expression was not observed in hyperplastic tumor blood vessels, normal brain blood vessels, inflammatory cells, or in the normal brain tissue that surrounded the tumor. Receptor expression was limited to tumor cells located within solid tumor tissue. Overall, 27 of 29 resected CNS tumors exhibited tumor cell-specific DDR1 expression, whereas one specimen composed of isolated glioblastoma cells within invaded brain parenchyma showed no detectable staining for this receptor. DDR1 was also expressed preferentially in the ventricular zone (a region of highly proliferating precursor cells) of mice at embryonic Day 15.5. CONCLUSION: We found that DDR1 is consistently expressed in all high-grade brain neoplasms studied and is selective for tumor cells in the specimens analyzed. The expression of DDR1 by tumor cells of CNS neoplasms and by primitive cells of the embryonic ventricular zone suggests that DDR1 is a potentially useful marker of tumor cells within the CNS.

Preoperative abdominal ultrasound may be misleading in risk stratification for presence of common bile duct abnormalities.

BACKGROUND: Following the advent of laparoscopic cholecystectomy (LC), the preoperative predictors of common bile duct (CBD) abnormalities became more important in perioperative decision making. Preoperative transabdominal ultrasound (US) is used to assess the preoperative risks associated with CBD abnormalities. This study attempts to determine the sensitivity and specificity of US in determining CBD abnormalities in patients prior to LC. METHODS: US measurements of the CBD diameter and presence of stones were ascertained from radiology reports in 100 patients who had LC with a routine intraoperative cholangiogram (IOC). The same information was obtained from the patients' IOC. A supraduodenal CBD diameter of >8 mm was considered dilated. RESULTS: US demonstrated a sensitivity of 25% and a specificity of 70% for the detection of CBD dilatation compared to IOC. The sensitivity of US for predicting CBD dilatation was 55% when the IOC-derived diameter was >10 mm and 100% when it was >15 mm. The overall sensitivity of US for detection of stones was 10%; it improved to 17% in patients with a dilated CBD on US. CONCLUSIONS: Preoperative ultrasound is neither sensitive nor specific for detecting CBD dilatation or presence of stones. A negative preoperative US report may be misleading in risk stratification for the presence of these CBD abnormalities. In order to avoid missing any CBD pathology, we recommend the routine use of intraoperative cholangiography.