Interleukin-2-unresponsive immune defects in good syndrome: Letter to the Editor

To the Editor:

Good Syndrome is a rare and poorly understood syndrome that has been reviewed recently in this journal [1], consisting of profound immunodeficiency associated with thymoma, for which there is no known treatment [2]. We report a case of Good Syndrome, and demonstrate for the first time specific defects in T cell differentiation and trial of interleukin-2 therapy.

A 48 year-old woman presented with Cytomegalovirus (CMV) encephalitis. She had had excision of a large benign thymoma 7 years prior, Pneumocystis pneumonia 5 years prior, unilateral CMV retinitis 4 years prior, with recurrence 3 years prior, and had been maintained chronically on trimethoprim-sulfamethoxazole and valganciclovir prophylaxis. Multiple HIV tests had been negative. She had completed a course of decitabine for possible aplastic anemia about 6 months prior. Ten months before admission, she had blood total CD4⁺/CD8⁺ T cell counts of 57/174 cells/ mm³ (ratio 0.33), B cell (CD19⁺) count of 0 cells/mm³, and NK cell (CD56⁺/CD3⁻) count of 80 cells/mm³. The history of thymoma, absent B cells, reduced NK cells, and CD4⁺ T lymphopenia with inverted CD4 to CD8⁺ T cell ratio, and opportunistic infections with CMV and Pneumocystis were all consistent with past reports of Good Syndrome [2].

She presented with CMV encephalitis diagnosed by cerebrospinal fluid CMV PCR and brain MRI, as well as bilateral active CMV retinitis. Foscarnet was begun; genotyping of CMV later confirmed ganciclovir resistance. Anti-CMV-enriched immunoglobulin (Cytogam) was also administered.

Initially, her blood total CD3⁺/CD4⁺/CD8⁺ T cell counts were 283/94/183 cells/mm³ (CD4:CD8 ratio 0.51), NK (CD16⁺/CD56⁺) cell count was 26 cells/mm,³ and B cell count was 0 cells/mm.³ Hypogammaglobulinemia was present with a total IgG/IgM/IgA of 311/14/17 mg/dl. A bone marrow biopsy showed hypocellularity with normal trilineage maturation without malignant or dysplastic cells.

Cytofluorometric delineation of T cell memory subsets [3] was performed serially starting day 44 (Fig. 1A and B). For CD4⁺ T cells, percentages of naïve (CD45 RA⁺/CD62L⁺) and central memory (CD45 RA-/CD62L⁺) cells were relatively normal, but more differentiated effector-memory (CD45 RA⁻/CD62L⁻) and terminal effector T cells (CD45 RA⁺/CD62L⁻) were depressed compared to healthy individuals [3]. For CD8⁺ T cells, naïve cells were relatively elevated, and central memory/effector-memory/terminal effector cells were reduced. Because an ongoing systemic viral infection should bias toward effector memory and terminal effector cells, the reduced maturation of T cells in this subject suggested a marked differentiation block in both CD4⁺ and CD8⁺ T cell types.

Figure 1.

T cell memory subsets. (A) Peripheral blood CD4⁺ and CD8⁺ T cells, assessed for expression of CD62L and CD45RA by cytofluorimetry. (B) Percentages of each memory T cell subset on day 44 versus normal values previously determined in healthy control subjects. (C) Percentages of memory subsets of CD4⁺ and CD8⁺ T cells over time. Gray shading indicates intervals where IL-2 was administered. B and NK cell subsets also did not normalize with IL-2 therapy (not shown).

Despite initial control of viremia, her neurological status steadily deteriorated. Cidofovir was added at about 60 days but her clinical status continued to decline. Thus, low dose s.c. interleukin (IL)-2 was administered in an attempt to improve the underlying immunodeficiency by boosting T/ NK/B cell production and/or differentiation. Doses were given on days 63–66 (3.5 × 10⁵ IU q12h), 76–99 (2.0 × 10⁵ IU q12h for two days then 3.5 × 10⁵ IU q12h), and 123–126 (3.5 × 10⁵ IUq12h). Dosing was not escalated further due to ongoing superimposed hemodynamic instability, fever, and anasarca. During treatment, there was no clear rise in T cell counts (not shown) or normalization of T cell memory distribution (Fig. 1C). A transient increase in NK and T cells between days 99–106 (not shown) appeared to be temporally correlated to the initiation of combination foscarnet and cidofovir therapy on day 88. Given her continuing decline and the development of renal failure, care was withdrawn on day 163, and the patient expired.

There are two novel observations from this Good Syndrome case. To our knowledge, we provide the first detailed examination of memory T cell differentiation. Despite active disseminated CMV infection, the percentages of effector memory and terminal effector T cells were low compared to healthy controls, suggesting marked impairment of both T cell production and differentiation. While both the CD4⁺ and CD8⁺ T cell populations were low in memory subsets, a primary defect in the CD4⁺ T cells could be sufficient to result in defective memory in CD8⁺ T cells, as well as B and NK cells, through decreased helper activity.

This is the first report of IL-2 administration to a patient with Good Syndrome. IL-2 is FDA-approved for treatment of melanoma and hypernephroma, and is believed to exert antitumor effects through promoting the growth and differentiation of T, B, and NK cells through signaling via the IL-2 receptor [4]. Interestingly, decreased IL-2 production from T cells in vitro has been reported for Good syndrome [2] suggesting dysfunction of memory T cells. We found that low dose IL-2 administration had no clear impact on the immunologic defects in our patient, indicating that this amount of IL-2 does not compensate for the underlying mechanism of Good Syndrome. With the caveat that the dose was low compared to that demonstrated to boost CD4⁺ T cell counts in HIV-1-infected patients [5], this suggests that reduced IL-2 production due to lymphopenia is not solely responsible for B and NK cell defects. If abnormality of the IL-2 axis is involved in the immunopathogenesis of Good Syndrome, this would suggest a defect in responsiveness to IL-2 such as a receptor or signaling defect. The typical adult onset and lack of familial pattern of this disease, however, argue against a simple genetic defect such as IL-2 receptor mutations that have been associated with cases of severe combined immunodeficiency [6].

Finally, this is only the fourth case of CMV encephalitis, and second case of CMV encephalitis concurrent with retinitis, to be reported for Good Syndrome [7]. This illustrates the profound level of immunosuppression that can result from the immunologic defects of T cells in this syndrome. A full discussion of the clinical manifestations and pathogenesis of Good Syndrome can be found in this issue of Clinical Immunology [1].

In conclusion, we find that the reduced CD4⁺ T cells typically noted in Good Syndrome are also associated with qualitative defects in memory subsets, and that these abnormalities are not responsive to low dose IL-2 administration.