Comparison of Plasma Cell Type of Castlemans Disease and IgG4-Related Sclerosing Disease: A Histopathological and Immunohistochemical Study

Objectives: Castleman’s disease (CD) is a group of rare atypical lymphoproliferative disorders classified as hyaline vascular (HV) and plasma cell (PC) types. CD may be closely mimicked by IgG4-related sclerosing disease (IgG4-SD) involving the lymph nodes. We retrospectively analyzed findings in patients with CD to elucidate the relationship between CD and IgG4-SD. Methods: Clinicopathological and immunophenotypical characteristics, including IgG+ and IgG4 expression by plasma cells, were analyzed in 87 consecutive patients diagnosed with CD from 1999 to 2010 at two major Korean hospitals. Results: The numbers of IgG+ (p ! 0.001) and IgG4+ (p ! 0.001) cells and the IgG4:IgG ratio (p = 0.003) were significantly higher in the PC than in the HV group. The mean IgG4+:IgG+ plasma cell ratio in the PC group was 25.1%, with 10 patients having a ratio 1 40%, the threshold IgG4:IgG ratio in patients with IgG4-SD. Conclusions: Patients with the PC form of CD and IgG4-related lymphadeReceived: December 6, 2010 Accepted after revision: March 14, 2011 Jooryung Huh, Department of Pathology Asan Medical Center, University of Ulsan College of Medicine 388-1 Poongnap-Dong, Songpa-Ku Seoul 138-736 (South Korea) Tel. +82 2 3010 4553, E-Mail jrhuh @ amc.seoul.kr © 2011 S. Karger AG, Basel 1015–2008/11/0784–0227$38.00/0 Accessible online at: www.karger.com/pat D ow nl oa de d by : 54 .7 0. 40 .1 1 11 /6 /2 01 7 7: 09 :3 3 P M Jo /Park /Jeon /Nam /Huh Pathobiology 2011;78:227–232 228 In this study, we evaluated the numbers and the distribution pattern of IgG-positive (IgG+) and IgG4-positive (IgG4+) plasma cells in CD to find relationship between CD and IgG4-SD. Patients and Methods Data from 87 patients with CD, including 55 consecutive patients diagnosed at the Asan Medical Center, Seoul, Korea, from 1999 to 2010, and 32 consecutive patients diagnosed at Seoul National University from 2001 to 2010, were analyzed in the study. The medical records of each patient were reviewed and clinical histories as well as follow-up data were recorded, including patient age, gender, original tumor site, tumor size and evidence of human immunodeficiency virus (HIV) infection. The multicentricity was assessed using the results of computed tomography; it was defined as a multicentric case when enlarged lymph nodes were identified in at least three different sites. Laboratory tests evaluated the presence in serum of human herpes virus 8 (HHV8) and serum concentrations of IgG and IgG4 if available. The histological features of each patient were used to reclassify patients into two groups, a hyaline vascular (HV) and a plasma cell (PC) group. Patients in the PC group showed at least one region with sheets of plasma cells and mixed cell types. To preserve nodal structures, we made multispecimen blocks [12] from formalin-fixed paraffin-embedded tissue sections and incubated these blocks with antibodies to IgG (1: 200; Dako, Glostrup, Denmark) and IgG4 (1: 2,000; Binding Site, Birmingham, UK). Immunohistochemical staining results were evaluated by determining the distribution patterns of IgG+ plasma cells, the numbers of IgG+ and IgG4+ plasma cells, and the ratio of IgG4+ to IgG+ plasma cells. The distribution patterns of IgG+ plasma cells were classified into a diffuse type, if at least one high-power field (HPF) was filled with IgG+ plasma cells, and a focal type, if IgG+ plasma cells were present, but not to the extent seen in the diffuse type. If no IgG+ cells were evident, the sample was classified as negative. The numbers of IgG+ and IgG4+ cells were counted using the UTHSCSA Image Tool (University of Texas Health Science Center, San Antonio, Tex., USA). The average numbers of IgG+ and IgG4+ cells per HPF were evaluated by enumeration of three areas showing the highest densities of positive cells. Samples were also immunostained with antibodies to (1: 100; Dako) and  (1: 4,000; Dako) light chains to exclude the possibility of malignant lymphoma. Epstein-Barr virus (EBV) status was assessed by in situ hybridization for EBV-encoded RNA, and the samples were also immunostained with antibody to HHV8 (1: 20; Cell Marque, Rocklin, Calif., USA). The  2 test was used to assess the relationships among multicentricity, distribution patterns of IgG+ cells and the new subgroups of CD. Student’s t test was employed to compare the numbers of IgG+ and IgG4+ cells per HPF and the ratio of IgG4+ to IgG+ cells in subgroups. All statistical analyses were performed using SPSS version 12.0 (SPSS Inc., Chicago, Ill., USA).