Dutz et al recently stated that X-linked lymphoproliferative disease (XLP) is a disease that presents as an Epstein-Barr virus (EBV)–specific immune defect.1 In contrast, we and others clearly have shown that prior EBV infection is not necessary for the development of XLP phenotypes dysgammaglobulinemia (mostly low immunoglobulin G [IgG]; increased IgM) and/or malignant lymphoma.2-4 It is likely that EBV may only trigger the most serious complication, fulminant mononucleosis. Moreover, in 2 boys with XLP, normal EBV–specific HLA-restricted cytotoxicity was clearly demonstrated.5 Furthermore, the authors stated that EBV-associated hemophagocytosis (VAHS) is a less common expression of XLP.1 In contrast, VAHS has been reported in 80% to 90% of XLP patients with fatal mononucleosis.6 7VAHS is therefore the second most common phenotype in patients with XLP.
The diagnosis of hypogammaglobulinemia of the reported patient was already made at age 13 months.1 The authors do not discuss why this boy received regular intravenous immunoglobulin (IVIG) for the first time at age 10 years when severe bronchiectasis had already developed and chorioiditis resulted in gradual blindness. Monthly IVIG substitution initiated at an earlier age would probably have prevented these serious complications.
The authors further report very low copy numbers of EBV genomes in vascular tissue of the reported patient with XLP.1 It is arguable whether these findings really play a significant role in the pathogenesis of XLP-associated lymphocytic vasculitis. EBV has recently been shown to infect endothelial cells.8 It remains to be elucidated whether low amounts of latent EBV may also be identified in vascular tissue of deceased immunocompetent EBV-seropositive individuals.
The expanding clinical phenotype of SAP/SH2D1A mutations
The specific aims of our paper were to report a new mutation in the gene for X-linked lymphoproliferative disease (XLP), to draw attention to a rare clinical manifestation of XLP (systemic vasculitis, including chorioretinitis), and to provide evidence that Epstein-Barr viral (EBV) infection or reactivation within endothelial cells may precipitate cytotoxic T lymphocyte (CTL)–mediated vasculitis. Schuster and colleagues raise a number of interesting points with regard to our paper, none of which detract from the above points.
The majority of patients with XLP present clinically upon exposure to EBV and the vast majority have had a history of EBV infection.1-1 The study of defined kindreds and the advent of molecular diagnostics with the identification of the gene responsible for the disease have allowed an expansion of the XLP clinical phenotype and have confirmed that in some cases, prior EBV infection is not required for all manifestations of the disease.1-2 This has been supported, in abstract form1-3 and by papers published after the submission of our manuscript.1-4 Similarly, while a number of patients present with a clinical picture compatible with virus-associated hemophagocytic syndrome (VAHS), until recently, VAHS in a single male patient with no family history would certainly be considered a less recognized expression of XLP. Arico et al have now shown that VAHS is linked with mutations of the XLP gene and that clinical presentations compatible with VAHS should prompt molecular investigations for XLP.1-5
Intravenous immunoglobulin (IVIG) can be used to treat significant hypogammaglobulinemia in some patients with XLP. The patient in question did not receive IVIG because the attending staff did not consider the moderate dysgammaglobulinemia significant enough to warrant IVIG treatment initially. Moreover, when IVIG was considered, compliance with therapy was problematic due to the remoteness of the patient's home and attendant difficulties in accessing health care.
EBV has been known for some time to infect endothelial cells in vitro,1-6 and murine γ-herpesvirus 68 has also been shown to infect endothelium in mice lacking the IFN-γ receptor.1-7 This is the reason we set forth to identify viral genome in the endothelium of the vasculitic lesions. Cell-mediated immune (CMI) dysfunction, including defective natural killer cell function has been documented in patients with XLP, and defective CMI may result in altered EBV tissue tropism as we suggest. We agree that many more patients, as well as controls, would need to be studied to definitively answer this issue. The functionality of EBV-specific CTL remains a subject for debate.1-8 1-9 Normal EBV-specific CTL function, as has been detected in some patients with XLP, might indeed explain the targeted assault and damage of the infected vascular structures that we have demonstrated.
