Molecular Biomarkers for the Diagnosis of Primary Vitreoretinal Lymphoma

Abstract

Primary vitreoretinal lymphoma (PVRL) or primary intraocular lymphoma, a subtype of primary central nervous system lymphoma, often masquerades as uveitis. The diagnosis of PVRL requires identification of lymphoma cells inside the eye, which is often challenging due to the frequent necrosis and admixing of PVRL cells with reactive lymphocytes. Therefore, detection of immunoglobulin heavy chain (IgH) and T-cell receptor (TCR) gene rearrangements provide molecular diagnosis of B- and T-cell lymphoma, respectively. We retrospectively evaluated 208 cases with a clinical diagnosis of masquerade syndrome from 1998 to 2010. In 200 cases with molecular analyses using microdissection and polymerase chain reaction, we found that 110 cases had IgH gene rearrangement, 5 cases had TCR gene rearrangement, and 85 cases were negative for these two gene arrangements. The molecular data corroborated the cytopathological diagnoses of PVRL and uveitis in the majority of cases. Cytokine above the detected levels in the specimens were also measured in 80 of the 208 cases. A ratio of vitreous IL-10 to IL-6 greater than 1, suggesting PVRL, was found in 56/80 cases; 53/56 had the correct diagnosis. A ratio less than 1, suggesting uveitis, was found in 24/80 cases; 17/24 correctly confirmed the diagnosis. Moreover, the molecular data corresponded well with the clinical course of the diseases. The sensitivity and specificity of these molecular biomarkers for the diagnosis of PVRL are higher than 95%.

Primary Intraocular Lymphoma

“…Cellular microdissection and PCR amplification can be applied together if malignant cells are too scanty to yield a greater cell population from vitreous specimens. 36,182,195,210 Chan et al, in 60 cases of PIOL using this technique, showed rearrangement of the CDR3 of IgH gene in all. 32 In an updated report of 114 cases of B-cell PIOL, all cases had IgH gene rearrangement, but this change was found in only 1 of 86 cases of uveitis.…”

“…26 Moreover, PIOL is reported in some studies to exhibit a high IL-10:IL-6 ratio (>1). 24,38,210,214,219 Wolf et al found the ratio was 74.7% accurate (sensitivity 74%, specificity 75.0%) for 35 patients with PIOL, with the reverse in 64 patients with uveitis. 219 Wang et al reported sensitivity of 88% and specificity of 85% in 80 patients with uveitis or PIOL.…”

“…219 Wang et al reported sensitivity of 88% and specificity of 85% in 80 patients with uveitis or PIOL. 210 The use of the IL-10:IL-6 or IL-10:IL-12 ratios, however, has been controversial as elevation may occur in eyes with non-neoplastic uveitis. 3,4,153 In one report the IL-10:IL-6 or IL-10:IL-12 ratio was elevated in 8 out of 14 vitreous samples from non-neoplastic vitritis eyes.…”

“…51 For lymphoma, the positive predictive value of cytologic evaluation was 99e100% and the negative predictive value was 61e81%. 63,210 Sparse number of cells is the main reason for an inconclusive result. 78,123 Vitreous specimens contain many reactive T-lymphocytes, necrotic cells, debris, and fibrin that can also confound the identification of malignant cells.…”

“…32 In an updated report of 114 cases of B-cell PIOL, all cases had IgH gene rearrangement, but this change was found in only 1 of 86 cases of uveitis. 210 The size of the CDR3 region can also be studied. 97 PCR is not always reliable.…”