Ming Zhao1, Lu Sun2, Jenny Z Lai3, Huaiyin Shi2, Kaiyong Mei4, Xianglei He1, Xiaolong Jin5, Jinping Lai6, Dengfeng Cao7. 1. Department of Pathology, Zhejiang Provincial People's Hospital & People's Hospital of Hangzhou Medical College, Hangzhou, China. 2. Department of Pathology, Chinese PLA General Hospital, Beijing, China. 3. University College, Washington University in Saint Louis, MO, USA. 4. Department of Pathology, The Second Affiliated Hospital, Guangzhou Medical University, Guangzhou, China. 5. Department of Pathology, Shanghai Ruijin Hospital, Shanghai Jiaotong University School of Medicine, Shanghai, China. 6. Department of Pathology and Laboratory Medicine, University of Florida, Gainesville, FL, USA. 7. Department of Pathology and Immunology, Washington University School of Medicine, Saint Louis, Missouri, USA. Electronic address: dcao@path.wustl.edu.
Abstract
INTRODUCTION: Gastric hepatoid carcinomas (GHCs) include type I (classic) and type II (fetal type gastrointestinal adenocarcinoma). The classic type shows overlapping morphologic features with those of hepatocellular carcinoma (HCC). The aim of this study is to investigate expression of LIN28 in GHCs and explore its utility to distinguish classic GHC from HCC. METHODS: We investigated immunohistochemical expression of LIN28 in 93 primary GHCs (47 type I, 46 type II) and 60 HCCs with comparison to SALL4, AFP, glypican-3, Hep Par1, p-CEA and CK7. We also stained LIN28 and SALL4 in 52 conventional gastric adenocarcinomas to assess their specificity in gastric carcinomas. RESULTS: Classic GHCs and fetal type gastrointestinal adenocarcinomas showed positive LIN28 in 21/47 (45%) and 10/46 (22%), SALL4 in 41/47 (87%) and 36/46 (78%), AFP in 30/46 (65%) and 33/46 (72%), glypican-3 in 31/41 (76%) and 24/38 (63%), Hep Par1 in 27/41 (66%) and 28/37 (76%), and CK7 in 15/40 (38%) and 25/38 (66%), respectively. p-CEA staining was seen in 19/44 (43%) classic GHCs. Among HCCs, LIN28, SALL4, AFP, glypican-3, Hep Par1, p-CEA and CK7 was seen in 1/60 (2%), 0/60 (0%), 6/30 (20%), 23/30 (77%), 29/30 (97%), 28/30 (93%) and 21/30 (70%) cases, respectively. LIN28 and SALL4 staining was seen in 2/52 (4%) and 14/52 (27%) gastric conventional adenocarcinomas, respectively. The sensitivity and specificity of distinguishing classic GHCs from HCCs was 45% and 98% for LIN28, 87% and 100% for SALL4, 65% and 80% for AFP, 76% and 30% for glypican-3, 66% and 3% for Hep Par1, 43% and 7% for p-CEA, and 38% and 30% for CK7, respectively. Combining LIN28 and SALL4 increased the sensitivity to 96% with 98% specificity to distinguish classic GHCs from HCCs. CONCLUSIONS: LIN28 is a very specific marker (98% specificity) for distinguishing classic GHCs from HCCs though it is not as sensitive as SALL4. AFP, glypican-3, Hep Par1 and p-CEA are not useful in distinguishing classic GHCs from HCCs. Combining LIN28 and SALL4 increased the sensitivity to distinguish classic PHCs from HCCs.
INTRODUCTION:Gastric hepatoid carcinomas (GHCs) include type I (classic) and type II (fetal type gastrointestinal adenocarcinoma). The classic type shows overlapping morphologic features with those of hepatocellular carcinoma (HCC). The aim of this study is to investigate expression of LIN28 in GHCs and explore its utility to distinguish classic GHC from HCC. METHODS: We investigated immunohistochemical expression of LIN28 in 93 primary GHCs (47 type I, 46 type II) and 60 HCCs with comparison to SALL4, AFP, glypican-3, Hep Par1, p-CEA and CK7. We also stained LIN28 and SALL4 in 52 conventional gastric adenocarcinomas to assess their specificity in gastric carcinomas. RESULTS: Classic GHCs and fetal type gastrointestinal adenocarcinomas showed positive LIN28 in 21/47 (45%) and 10/46 (22%), SALL4 in 41/47 (87%) and 36/46 (78%), AFP in 30/46 (65%) and 33/46 (72%), glypican-3 in 31/41 (76%) and 24/38 (63%), Hep Par1 in 27/41 (66%) and 28/37 (76%), and CK7 in 15/40 (38%) and 25/38 (66%), respectively. p-CEA staining was seen in 19/44 (43%) classic GHCs. Among HCCs, LIN28, SALL4, AFP, glypican-3, Hep Par1, p-CEA and CK7 was seen in 1/60 (2%), 0/60 (0%), 6/30 (20%), 23/30 (77%), 29/30 (97%), 28/30 (93%) and 21/30 (70%) cases, respectively. LIN28 and SALL4 staining was seen in 2/52 (4%) and 14/52 (27%) gastric conventional adenocarcinomas, respectively. The sensitivity and specificity of distinguishing classic GHCs from HCCs was 45% and 98% for LIN28, 87% and 100% for SALL4, 65% and 80% for AFP, 76% and 30% for glypican-3, 66% and 3% for Hep Par1, 43% and 7% for p-CEA, and 38% and 30% for CK7, respectively. Combining LIN28 and SALL4 increased the sensitivity to 96% with 98% specificity to distinguish classic GHCs from HCCs. CONCLUSIONS:LIN28 is a very specific marker (98% specificity) for distinguishing classic GHCs from HCCs though it is not as sensitive as SALL4. AFP, glypican-3, Hep Par1 and p-CEA are not useful in distinguishing classic GHCs from HCCs. Combining LIN28 and SALL4 increased the sensitivity to distinguish classic PHCs from HCCs.