Kaposi's sarcoma-associated herpesvirus/human herpesvirus type 8-positive solid lymphomas: a tissue-based variant of primary effusion lymphoma

Abstract

Kaposi's sarcoma-associated herpesvirus (KSHV), also termed human herpesvirus type 8, is consistently identified in Kaposi's sarcoma, primary effusion lymphoma (PEL), and multicentric Castleman's disease. Here we report four cases of KSHV-bearing solid lymphomas that occurred in AIDS patients (cases 1 to 3) and in a human immunodeficiency virus (HIV)-seronegative person (case 4). The patients presented extranodal masses in the abdomen (cases 1, 3, and 4) or skin (case 2), and nodal involvement, together with Kaposi's sarcoma (case 3). The gastrointestinal tract was involved in two patients (cases 1 and 3). The patients did not develop a lymphomatous effusion. KSHV was detected in the tumor cells of all cases by immunohistochemistry and by polymerase chain reaction. Epstein-Barr virus was detected in two of the HIV-related cases. All KSHV-positive solid lymphomas exhibited PEL-like cell morphology. To investigate the relationship of these disorders to PEL and to other AIDS-associated diffuse large cell lymphomas, KSHV-positive solid lymphomas were tested for the expression of a set of genes that were previously shown by gene profiling analysis to define PEL tumor cells. The results showed that expression of this set of genes in KSHV-positive lymphomas is similar to that of PEL but distinct from KSHV-negative AIDS-associated diffuse large cell lymphomas. Because pathobiological features of KSHV-positive solid lymphomas closely mimic those of PEL, our results suggest that KSHV-positive solid lymphomas should be considered as a tissue-based variant of classical PEL, irrespective of HIV status.

Figure 1

KSHV/HHV8-positive solid lymphomas. A: H&E stain. B: Immunohistochemistry for KSHV ORF73 protein. C: PCR for KSHV sequences (K330₂₃₃). In cases 3 and 4 DNA was extracted from both formalin and frozen tissue samples. C+, positive control (CRO-AP/6 PEL cell line); M, DNA molecular weight marker. Original magnifications, ×40.

Figure 2

KSHV-positive solid lymphomas positively immunostain for molecules that are specifically expressed by PEL tumor cells. Original magnifications, ×40.

Figure 3

KSHV-positive solid lymphomas display expression of a subset of genes selected among the genes specifically up-regulated in PEL tumor cells. The results of a relative quantification assay is reported as a fold-difference relative to a calibrator sample. Relative quantification of target gene expression was calculated from ABI Prism 7700 Relative Quantification software that calculates relative levels of gene expression using the comparative C_T method of data analysis. A sample of centroblastic AIDS-DLCL, already studied by gene expression profiling analysis is used as calibrator. The calibrator expressed all of the tested genes, but at lower levels than those reported for PELs. A PEL cell line termed CRO-AP/5, whose parental sample was previously tested by gene expression profiling analysis, is included in the real-time RT-PCR relative quantification experiments to determine the expression of the same subset of genes. The x axis of the graph lists all of the target involved in the analysis. Within each target group, the software displays the relative quantity of the target in each sample. In the y axis are the relative quantities graphed on a logarithmic scale. The quantities are shown relative to the calibrator sample. Each increment corresponds to a 10-fold difference in gene expression. Because the calibrator sample is compared to itself, the level of cDNA expression in the calibrator always appears as 1 (1E + 00).