Faculty Bibliography

Bodies have continuous reticular networks, comprising collagens, elastin, glycosaminoglycans, and other extracellular matrix components, through all tissues and organs. Fibrous coverings of nerves and blood vessels create structural continuity beyond organ boundaries. We recently validated fluid flow through human fibrous tissues, though whether these interstitial spaces are continuous through the body or discontinuous, confined within individual organs, remains unclear. Here we show evidence for continuity of interstitial spaces using two approaches. Non-biological particles (tattoo pigment, colloidal silver) were tracked within colon and skin interstitial spaces and into adjacent fascia. Hyaluronic acid, a macromolecular component of interstitial spaces, was also visualized. Both techniques demonstrate interstitial continuity within and between organs including within perineurium and vascular adventitia traversing organs and the spaces between them. We suggest that there is a body-wide network of fluid-filled interstitial spaces that has significant implications for molecular signaling, cell trafficking, and the spread of malignant and infectious disease.

BACKGROUND:Recent studies have suggested the important roles of CD47 and tumor-associated macrophages in the prognosis and immunotherapy of various human malignancies. However, the clinical significance of CD47 expression and CD163+ TAMs in pancreatic neuroendocrine tumor (PanNET) remains unclear. METHODS:In this study, 47 well-differentiated PanNET resection specimens were collected. CD47 expression and CD163+ macrophages were evaluated using immunohistochemistry and correlated with clinicopathologic properties. RESULTS:Positive CD47 staining was seen in all PanNETs as well as adjacent normal islets. Compared to normal islets, CD47 overexpressed in PanNETs (p = 0.0015). In the cohort, lymph node metastasis (LNM), lymphovascular invasion (LVI), and perineural invasion (PNI) were found in 36.2, 59.6, and 48.9% of the cases, respectively. Interestingly, PanNETs with LNM, LVI, or PNI had significantly lower H-score of CD47 than those without LNM (p = 0.035), LVI (p = 0.0005), or PNI (p = 0.0035). PanNETs in patients with disease progression (recurrence/death) also showed a significantly lower expression of CD47 than those without progression (p = 0.022). In contrast, CD163+ macrophage counts were significantly higher in cases with LNM, LVI, and PNI. CONCLUSIONS:Our data suggest relative low CD47 expression and high CD163+ TAMs may act as indicators for poor prognosis of PanNETs.

Simple mucinous cysts of the pancreas have an epithelial lining resembling pancreatic intraepithelial neoplasia but may have a clinical presentation similar to premalignant mucinous neoplasms such as intraductal papillary mucinous neoplasms. Whether the epithelial lining shares genomic alterations with other pancreatic preinvasive neoplasms such as PanIN and intraductal papillary mucinous neoplasm has not been determined. We performed targeted sequencing analysis using a custom-designed MiSeq panel including the full coding regions of 18 pancreatic cancer genes on 13 clinically and pathologically well-characterized simple mucinous cysts. We detected 59 mutations in 15 genes in the cohort, with a median of 4 mutations per cyst (range = 0-16 mutations per cyst). The mutated genes and rate of detected mutations were as follows: KMT2C (MLL3) (62%), KRAS (15%), BRAF (8%), RNF43 (8%), CDKN2a (8%), TP53 (15%), and SMAD4 (8%). No GNAS mutations were detected. Four cases (31%) had no mutations detected. These findings place the majority of simple mucinous cysts of the pancreas in the spectrum of early, low-grade mucinous neoplasia, albeit with a different spectrum of genomic alterations compared with PanIN and intraductal papillary mucinous neoplasm.

OBJECTIVES/OBJECTIVE:To investigate the diagnostic potential of AEG-1 and GPC-3 in hepatocellular carcinoma (HCC). METHODS:AEG-1 and GPC-3 immunohistochemistry were performed on HCC, adjacent nontumor tissue (ANT), and dysplastic nodules (DN). RESULTS:H score of AEG-1 or GPC-3 in HCC was significantly higher than in ANT or DN. In HCC, 92% and 54% showed AEG-1 and GPC-3 positivity, respectively. In ANT, 16.2% were AEG-1 and 7.6% GPC-3 positive. AEG-1 staining was mostly diffuse, whereas GPC-3 frequently showed focal staining. AEG-1 alone showed high sensitivity but low specificity and accuracy. GPC-3, on the other hand, showed high specificity but low sensitivity and accuracy. Combination of both stains boosted the sensitivity, specificity, and accuracy to 94.6%, 89.5%, and 90.5%, respectively, when only diffuse staining was considered as positive. CONCLUSIONS:AEG-1 or GPC-3 alone seemed not an ideal marker for HCC. The combination of AEG-1 and GPC-3 might improve early diagnosis of HCC.