Immune restoration disease (IRD) and immune reconstitution inflammatory syndrome (IRIS) are essentially synonymous terms used to describe the paradoxical worsening of appropriately treated opportunistic infections or unmasking of a previously subclinical infection, malignancy, or autoimmune condition following the initiation of antiretroviral therapy (ART) in HIV-infected patients. Most cases of IRD represent an aberrant immune response to foreign antigen disseminated during periods of severe immunosuppression and are temporally associated with memory T-cell recovery occurring within the first 6 months of ART .
Graves’ disease is a common autoimmune thyroid disease that affects up to 0.5% of the population worldwide. Despite decades of investigation, the immunopathogenesis of Graves’ disease remains elusive. Although it is clear that inappropriate autoantibody production targeting the thyrotropin/thyroid stimulating hormone receptor (TSH-R) mediates disease, it is not understood how both B and T cells specific to these self-antigens escape elimination in the bone marrow and thymus, respectively.
At least 39 cases of Graves’ disease have been described as IRD events following initiation of active ART in the HIV literature [2–12]. In contrast to most IRD events, which occur early after ART initiation and are related to opportunistic infections, Graves’ disease-IRD occurs up to 3 years after ART suggesting that the immunopathogenesis of this form of IRD may be distinct. Some authors have speculated that this delay might be related to thymus-dependent generation of naive T cells [4,10]. In fact, in one study of Graves’ disease-IRD, immunophenotypic and T-cell receptor excision circle (TREC) analysis of a single patient with Graves’ disease-IRD revealed an increased numbers of naive CD4+ and CD8+ T cells and TREC levels in comparison with immune reconstituted HIV-infected individuals . Although the contributions of B cells to Graves’ disease-IRD in HIV patients have not been specifically addressed to date, HIV infection causes hyperactivation of B cells characterized by, among other things, increases in the proportions of immature B cells, differentiation of B cells to plasmablasts, and increases in the production of autoantibodies . We hypothesized, therefore, that Graves’ disease-IRD may be related to a break in tolerance related to both aberrant B-cell and T-cell recovery following ART initiation.
We designed this retrospective study to determine whether there were temporal relationships between naive T-cell recovery and the development of Graves’ disease in a case series of patients who developed Graves’ disease following ART initiation. We further sought to investigate whether there were differences in T-cell and B-cell recovery, TRECs, and serum cytokine levels between patients who developed Graves’ disease compared with those who did not and whether patients who developed Graves’ disease tended to develop other autoimmune antibodies following ART.
Study participants and study design
This study was performed at the National Institute of Allergy and Infectious Diseases (NIAID), National Institutes of Health (NIH) under an Institutional Review Board-approved protocol. All study participants signed informed consent. A retrospective review of all patients (N = 521) who received their primary HIV care at the National Institutes of Health from 1995 to 2009 was conducted. Participants who developed clinical and laboratory evidence of Graves’ disease following the initiation or change in ART were included in the study. Seven patients were identified, including one previously reported . For each Graves’ disease patient, two age, sex, and CD4+ T-cell-matched controls were identified. Each of the study participants had cryopreserved serum and peripheral blood mononuclear cells (PBMCs) available for retrospective laboratory evaluations.
Definitions and case identifications
Participants were considered to have Graves’ disease if they exhibited a sign or symptom of Graves’ disease including anxiety, depression, insomnia, tremor, proximal muscle weakness, weight loss, diarrhea, heat intolerance, sweats, palpitations, fatigue, proptosis, or lid lag; laboratory evidence of hyperthyroidism including decreased TSH and elevated free T4; and laboratory evidence of TSH-R antibodies (thyrotropin stimulating immunoglobulin). For the purposes of this study, month 0 was considered the day of initiation or change of ART.
Immunophenotyping of cryopreserved PBMCs from cases and controls was performed at baseline prior to ART (month 0) and yearly following ART initiation for up to 4 years when specimens were available. T lymphocyte activation and differentiation (naive/memory) were evaluated by multiparameter flow cytometry. The fluorochrome-conjugated antibodies used were as follows: anti-CD8 QD655, and Live/Dead Fixable Blue Dead Cell Stain Kit with UV excitation from Invitrogen (Carlsbad, California, USA); anti-CD3 APC-Cy7, anti-CD4 Pacific Blue, anti-CD25 PE-Cy7, anti-CD38 Phycoerythrin, anti-CD40L Phycoerythrin, anti-CD45RO APC, anti-CD45RO PE-Cy7, anti-CTLA4 APC, anti-HLA-DR PE-Cy5, and anti-Ki67 FITC from Becton Dickinson (Franklin Lakes, New Jersey, USA); anti-CD27 Alexa Fluor 700 and anti-programmed cell death protein 1 (PD-1) biotinylated from BioLegend (San Diego, California, USA); and anti-FoxP3 PE-Cy5 from Ebioscience (San Diego, California, USA). Samples were acquired in a BD LSR-II flow cytometer and analysis was performed using FlowJo software version 8. A forward scatter area (FSC-A) vs. forward scatter height (FSC-H) gate was used initially to exclude doublets and nonviable cells. Remaining cells were then gated based on FSC-A and side scatter area (SSC-A) to capture lymphocytes. Events were sequentially gated on CD3+, CD4+, or CD8+ lymphocytes, and analysis of naive and memory subpopulations was done based on CD27 and CD45RO expression: naive (CD45RO−CD27+), central memory (CD45RO+CD27+), effector memory (CD45RO+ CD27−), and effector (CD45RO−CD27−) cells. CD4+CD25high FoxP3+ cells were identified as T-regulatory cells (Tregs). Activation markers [PD-1, CD38, human leukocyte antigen (HLA)-DR, CTLA4, and CD40L] and the cycling marker Ki67 were analyzed within CD4+ and CD8+ T cells and within each memory subpopulation.
PBMCs were obtained and immunophenotyped as previously described  with the following modifications. Fluorochrome-conjugated monoclonal antibodies used for staining included allophycocyanin (APC) anti-CD10; APC-H7 anti-CD20; peridinin chlorophyll protein-Cy5.5 anti-CD19; APC-eFluor 780 anti-CD19; PE-Cy7 anti-CD27; and FITC anti-CD21 (Beckman Coulter, Fullerton, California, USA). Immunophenotyping was performed on a BD FACSCanto II (BD Biosciences San Jose, California, USA) flow cytometer. Analyses were performed with FlowJo Version 9.4.11 software (TreeStar, Ashland, Oregon, USA). Analysis of naive and memory subpopulations was based on CD10, CD27, and CD21 expression: immature/transitional B cells (CD10+CD27−), naive B cells (CD10−CD27−CD21hi), resting memory B cells (CD27+CD21lo), activated memory B cells (CD10−CD27+CD21lo), tissue-like memory B cells (CD10−CD27−CD21lo), or plasmablasts (CD27+CD20−CD21lo).
T-cell receptor excision circle analysis
TRECs in PBMCs were quantified by real-time PCR using a cell-lysis method as described . The consistency of the DNA content of the cell lysate was evaluated by real-time PCR using a ribosomal protein gene and the Taqman assay kit from Applied Biosystems (Life Technologies, Carlsbad, California, USA). The number of TRECs is reported per 106 PBMCs.
Luciferase immunoprecipitation systems
Plasma autoantibody testing was performed at baseline, year 2, and year 4 when specimens were available for cases and controls using luciferase immunoprecipitation system (LIPS). The Renilla luciferase-autoantigen targets included glutamic decarboxylase-65 (GAD65), the β-subunit of the gastric ATPase, and Ro52, which have been previously described [17,18]. LIPS testing with these different Renilla luciferase antigen fusions was performed in a 96-well plate format at room temperature as described . Briefly, a ‘master plate’ was constructed by diluting patient plasma 1 : 10 in assay buffer A (20 mmol/l Tris, pH 7.5, 150 mmol/l NaCl, 5 mmol/l MgCl2, 1% Triton X-100) in a deep-well microtiter plate. For evaluating antibody titers by LIPS, 40 μl of buffer A, 10 μl of diluted human sera (1 μl equivalent), and 50 μl of 1 × 107 light units (LU) of the Renilla luciferase-antigen Cos1 cell extract diluted in buffer A were added to each well of a polypropylene plate and incubated for 1 h at room temperature. Next, 7 μl of a 30% suspension of Ultralink protein A/G beads (Pierce Biotechnology, Rockford, Illinois, USA) in phosphate-buffered saline were added to the bottom of each well of a 96-well filter HTS plate (Millipore, Bedford, Massachusetts, USA). To this filter plate, the 100 μl antigen-antibody reaction mixture was transferred and incubated for 1 h at room temperature on a rotary shaker. The washing steps of the retained protein A/G beads were performed on a BioMek FX work station (Beckman Coulter) using an integrated vacuum manifold. After the final wash, light units were measured in a Berthold LB 960 Centro microplate luminometer (Berthold Technologies, Bad Wildbad, Germany) using coelenterazine substrate mix (Promega, Madison, Wisconsin, USA). All light unit data were obtained from the average of at least two independent experiments and the light unit values were used without subtracting the buffer blank. Positivity for each of the antigens was based on cut-offs from previous studies [18–20].
For comparisons of variables repeated over time following ART initiation in cases vs. controls, values were grouped in 10-month intervals and were compared with the Wilcoxon signed rank test using SAS version 9.3 (SAS Institute Inc., Cary, North Carolina, USA). To account for the correlation due to repeated observations over time, we confirmed the results when appropriate using repeated measures mixed model analysis. Because of the exploratory nature of the study, there was no correction for multiple comparisons, and only unadjusted P values are reported. Median values with interquartile ranges (IQRs) are reported throughout.
Timing and presentation of clinical cases
Demographic and clinical characteristics for each Graves’ disease patient are described in Table 1. All seven patients were men between the ages of 30 and 45 years. Four patients identified themselves as white, two as black, and one as Latino. All but one met the 1993 revised Centers for Disease Control criteria for AIDS. None had a personal or family history of thyroid disease and one had a personal history of Addison's disease. At the time of ART initiation, four patients had never experienced an AIDS-defining illness, one had been diagnosed with pneumocystis pneumonia (PCP), and two patients had a history of several opportunistic diseases (Table 1). The median number of months between ART initiation and Graves’ disease diagnosis was 34 (range 17–46). One patient interrupted ART from months 26 to 29 and from month 31 to the time of Graves’ disease diagnosis at 38 months. The clinical presentations of Graves’ disease were as follows: two patients presented with diarrhea (one of whom had significant weight loss); one with proptosis alone; one patient with proptosis, tremor, lid lag, and insomnia; one with weight loss, tremulousness, clonus, and new-onset hypertension; one with new-onset dementia; and one with thyrotoxicosis.
All seven Graves’ disease-IRD patients had excellent immunologic response to ART including the patient with intermittent nonadherence; CD4+ T-cell counts increased from a median of 12 cells/μl (IQR 3–53) pre-ART to 513 cells/μl (IQR 383–567) at Graves’ disease diagnosis. Plasma viremia decreased from a median of 5.7 log copies/ml (IQR 5.5–6.7) pre-ART to 1.69 log copies/ml (IQR 1.69–1.95) at Graves’ disease-IRD diagnosis. Naive T-cell recovery following ART was temporally associated with emergence of TSH-R antibodies (Fig. 1). The proportion of Treg (CD4+CD25highFoxP3+) was comparable or higher than historical healthy controls at the time of ART initiation and at the time of Graves’ disease diagnosis (Fig. 1).
Naive T-cell recovery in cases and controls
To further investigate relationship between naive T-cell recovery and Graves’ disease-IRD, we identified controls for each of the seven Graves’ disease-IRD patients. Because the time to Graves’ disease-IRD following ART varied from patient to patient, we compared cases matched with HIV-positive controls by variables grouped in 10-month intervals up to 30 months after ART.
Cases and controls were well matched on pre-ART CD4+ T-cell counts [cases 28 (IQR 12–231) vs. controls 49.5 (IQR 24–243) cells/μl], and there were no significant differences in absolute CD4+ T-cell recovery following ART (data not shown). At the time of ART initiation, patients who would go on to develop Graves’ disease had a higher proportion of naive CD4+ T cells compared with controls [26.9% (8.43–39.3%) vs. 7.64% (5.51–11.1%), P = 0.0572]. Furthermore, the proportion of naive CD4+ T cells rose significantly in cases compared with controls after the first 10 months of ART (months 10–20, P = 0.0010; months 20–30, P = 0.0027; Fig. 2a). Repeated measures mixed model analysis confirmed this finding (overall P = 0.0162, data not shown). There were no significant differences in the change in proportion of effector or memory T cells between the two groups. Although TREC values between the two groups were not different at baseline, levels in cases compared with controls increased significantly 20–30 months (7634 vs. 504/106 cells, P = 0.0071) and 30–40 months following ART initiation (16 766 vs. 1010/106 cells, P = 0.0253; Fig. 2b). Repeated measures mixed model analysis confirmed this finding (overall P = 0.0408, data not shown).
There were no significant differences between cases and controls in the proportion of B cells, immature/transitional B cells, naive B cells, resting memory B cells, activated memory B cells, tissue-like memory B cells, or plasmablasts at day 0 (the day of ART initiation). Similarly, no significant differences were found between cases and controls in these variables over time following ART initiation (Fig. 3).
We investigated whether the development of Graves’ disease-IRD might be associated with cytokines associated with B-cell activation [interferon-γ (IFNγ), interleukin-10 (IL-10), IL-6, IL-8, tumor necrosis factor-α (TNFα), BAFF, sCD40L], B-cell precursor proliferation (IL-7), or B-cell proliferation and retention in germinal centers (IL-21). Serum levels of IFNγ, IL-10, IL-12p70, IL-1β, IL-6, IL-8, TNFα, IL-7, BAFF, sCD40L, and IL-21 were not different between cases and controls at day 0 and no significant differences between cases and controls were found in cytokine levels following ART initiation (data not shown).
Autoantibody testing using luciferase immunoprecipitation system
Patients who developed Graves’ disease-IRD in association with increased proportions of naive CD4+ T cells could also be prone to the development of other autoantibodies after ART initiation. Thus, we assessed for the presence of antibodies to other self-antigens in the Graves’ disease-IRD cases and controls at baseline and over time following ART using LIPS. GAD65 is a neuroendocrine enzyme and anti-GAD65 antibodies are detected in several autoimmune diseases including new-onset type I diabetes mellitus  and neurological syndromes such as stiff man's syndrome [20,21]. Gastric ATPase is an enzyme produced by gastric parietal cells  and antibodies against it are also associated with type I diabetes mellitus. Ro52 is an autoantibody associated with autoimmune hypothyroidism and type I diabetes mellitus . At baseline, LIPS analysis demonstrated the presence of antibodies to GAD65 or gastric ATPase in two (28.5%) of Graves’ disease-IRD patients and five (35.7%) of control patients. After ART initiation, three Graves’ disease-IRD patients developed antibodies to these two antigens and none the controls. In the Graves’ disease-IRD group as a whole, antibody titers increased significantly compared with controls both to gastric ATPase (change from baseline; 1369 vs. −583 LU, P = 0.0051) and to GAD65 (change from baseline; 2684 vs. −631 LU, P = 0.0174). No participants developed diabetes or any autoimmune neurologic syndrome. Ro52 antibodies were not detected in any study participant at the tested time points.
Our data show that the development of Graves’ disease-IRD is temporally associated with the recovery of naive CD4+ T lymphocytes following ART initiation and that patients who go on to develop Graves’ disease-IRD experience greater naive CD4+ T-cell recovery in comparison with controls. Consistent with this finding, TREC levels, which were similar between cases and controls at the time of ART initiation, increased significantly following ART initiation in Graves’ disease-IRD patients compared with controls. These data suggest that immune restoration of lymphopenic patients that follows effective ART may predispose HIV-infected individuals to the development of autoreactive T lymphocytes and inappropriate autoantibody production. Thus, robust naive CD4+ T-cell recovery, possibly with an abundance of primary thymic emigrants, may be a key risk factor for a break in immune tolerance.
T lymphocytes play an essential role in the development of Graves’ disease and may also play an integral role in preventing autoimmune disease in unaffected individuals . Graves’ disease is mediated by autoantibodies whose production is dependent on the presence of autoreactive B cells and their activation and differentiation to plasmablasts. We, thus, hypothesized that patients who developed Graves’ disease-IRD might have higher proportions of B-cell phenotypes associated with hyperactivation, immature/transitional B cells, and plasmablasts. Although we did not find differences between cases and controls in proportions of B-cell phenotypes at baseline or following ART initiation, we did observe a relatively high proportion of immature/transitional B cells in both cases and controls at baseline during the period of continued viremia and normalization in both groups over time after ART. Immature and transitional B cells are prone toward polyreactivity and B-cell maturity serves as an important checkpoint in selecting against self-reactive antibodies . It is feasible, therefore, that the break in B-cell immune tolerance required for the development of Graves’ disease-IRD occurs during this period prior to ART and that this increase in immature/transitional B cells is necessary, but not sufficient, for the development of Graves’ disease-IRD. Supporting this hypothesis is our finding that antibody titers to unrelated autoantigens increased significantly in Graves’ disease-IRD patients compared with controls after ART.
In contrast to two recent studies that found lower proportions of Treg cells in HIV-uninfected individuals with recent diagnoses of Graves’ disease [24,25], we observed proportions of Treg cells that were comparable to historical healthy controls and no clear changes in these proportions in the time leading up to Graves’ disease-IRD events. Due to sample availability, we were not able to investigate whether Treg cells from Graves’ disease-IRD patients had impaired function as has been described in two recent studies of Graves’ disease in HIV-uninfected adults [25,26].
The primary limitation of our study is that it was performed retrospectively. Although we know that none of these participants had a family or personal history of thyroid disease prior to enrollment in this cohort, we cannot prove that they would not have developed Graves’ disease in the absence of HIV infection. It is possible that some or all of these participants had the genetic and/or environmental predisposition for Graves’ disease at the time of enrollment but were protected from its development by HIV-induced CD4+ lymphopenia.
In summary, we describe the clinical presentation, T-cell and B-cell subset evolution, and serum cytokine levels of HIV-positive patients who developed Graves’ disease during immune reconstitution. In contrast to most IRD syndromes that are associated with recovery of memory CD4+ T cells, Graves’ disease-IRD is associated with naive CD4+ T-cell and primary thymic emigrant recovery in the absence of elevations in inflammatory cytokines typically associated with IRD.
I.S., B.O.P., S.M., and V.S. developed the hypotheses. J.M.M. provided clinical care for the patients and helped B.O.P. identify patients for the study. I.S. and V.S. developed the study design. A.G. performed analysis of serum cytokines. P.B.D. performed autoantibody testing using LIPS. L.K., S.M., and B.H.S. performed B-cell immunophenotyping. A.G.R. and V.S. performed T-cell immunophenotyping. V.N. performed analysis of TREC content. R.D. performed statistical analysis. V.S. wrote the article. All authors contributed to article preparation. The authors would like to acknowledge Cathy Rehm for her help with specimen processing.
This work was supported in part by the Intramural Research Programs of National Institute of Allergy and Infectious Diseases (NIAID) and National Institute of Dental and Cranfacial Research (NIDCR) and with federal funds from the NCI, NIH under Contract No. HHSN261200800001E. The views expressed in this article are those of the authors and views or policies do not necessarily reflect those of the Department of Health and Human Services, nor does mention of trade names, commercial products.
Conflicts of interest
The authors report no conflicts of interest.
1. Antonelli LR, Mahnke Y, Hodge JN, Porter BO, Barber DL, DerSimonian R, et al. Elevated frequencies of highly activated CD4+ T cells in HIV+ patients developing immune reconstitution inflammatory syndrome
2. Jimenez C, Moran SA, Sereti I, Wynne S, Yen PM, Falloon J, et al. Graves’ disease after interleukin-2 therapy in a patient with human immunodeficiency virus infection
3. Gilquin J, Viard JP, Jubault V, Sert C, Kazatchkine MD. Delayed occurrence of Graves’ disease after immune restoration with HAART. Highly active antiretroviral therapy
4. Jubault V, Penfornis A, Schillo F, Hoen B, Izembart M, Timsit J, et al. Sequential occurrence of thyroid autoantibodies and Graves’ disease after immune restoration in severely immunocompromised human immunodeficiency virus-1-infected patients
. J Clin Endocrinol Metab
5. Sereti I, Sarlis NJ, Arioglu E, Turner ML, Mican JM. Alopecia universalis and Graves’ disease in the setting of immune restoration after highly active antiretroviral therapy
6. Perez N, Del Bianco G, Murphy JR, Heresi GP. Graves’ disease following successful HAART of a perinatally HIV-infected 11-year-old
7. Vos F, Pieters G, Keuter M, van der Ven A. Graves’ disease during immune reconstitution in HIV-infected patients treated with HAART
. Scand J Infect Dis
8. Knysz B, Bolanowski M, Klimczak M, Gladysz A, Zwolinska K. Graves’ disease as an immune reconstitution syndrome in an HIV-1-positive patient commencing effective antiretroviral therapy: case report and literature review
. Viral Immunol
9. Clark BM, Krueger RG, Price P, French MA. Compartmentalization of the immune response in varicella zoster virus immune restoration disease causing transverse myelitis
10. Chen F, Day SL, Metcalfe RA, Sethi G, Kapembwa MS, Brook MG, et al. Characteristics of autoimmune thyroid disease occurring as a late complication of immune reconstitution in patients with advanced human immunodeficiency virus (HIV) disease
. Medicine (Baltimore)
11. Crum NF, Ganesan A, Johns ST, Wallace MR. Graves disease: an increasingly recognized immune reconstitution syndrome
12. Rasul S, Delapenha R, Farhat F, Gajjala J, Zahra SM. Graves’ disease as a manifestation of immune reconstitution in HIV-infected individuals after initiation of highly active antiretroviral therapy
. AIDS Res Treat
13. French MA, Lewin SR, Dykstra C, Krueger R, Price P, Leedman PJ. Graves’ disease during immune reconstitution after highly active antiretroviral therapy for HIV infection: evidence of thymic dysfunction
. AIDS Res Hum Retroviruses
14. Malaspina A, Moir S, Ho J, Wang W, Howell ML, O'Shea MA, et al. Appearance of immature/transitional B cells in HIV-infected individuals with advanced disease: correlation with increased IL-7
. Proc Natl Acad Sci U S A
15. Moir S, De Ravin SS, Santich BH, Kim JY, Posada JG, Ho J, et al. Humans with chronic granulomatous disease maintain humoral immunologic memory despite low frequencies of circulating memory B cells
16. Lempicki RA, Kovacs JA, Baseler MW, Adelsberger JW, Dewar RL, Natarajan V, et al. Impact of HIV-1 infection and highly active antiretroviral therapy on the kinetics of CD4+ and CD8+ T cell turnover in HIV-infected patients
. Proc Natl Acad Sci U S A
17. Burbelo PD, Leahy HP, Issa AT, Groot S, Baraniuk JN, Nikolov NP, et al. Sensitive and robust luminescent profiling of anti-La and other autoantibodies in Sjogren's syndrome
18. Burbelo PD, Lebovitz EE, Bren KE, Bayat A, Paviol S, Wenzlau JM, et al. Extrapancreatic autoantibody profiles in type I diabetes
. PLoS One
19. Burbelo PD, Hirai H, Issa AT, Kingman A, Lernmark A, Ivarsson SA, et al. Comparison of radioimmunoprecipitation with luciferase immunoprecipitation for autoantibodies to GAD65 and IA-2beta
. Diabetes Care
20. Ching KH, Burbelo PD, Carlson PJ, Drevets WC, Iadarola MJ. High levels of anti-GAD65 and anti-Ro52 autoantibodies in a patient with major depressive disorder showing psychomotor disturbance
. J Neuroimmunol
21. Peltola J, Kulmala P, Isojarvi J, Saiz A, Latvala K, Palmio J, et al. Autoantibodies to glutamic acid decarboxylase in patients with therapy-resistant epilepsy
22. Volpe R. The pathogenesis of Graves’ disease
. Endocr Pract
23. Wardemann H, Yurasov S, Schaefer A, Young JW, Meffre E, Nussenzweig MC. Predominant autoantibody production by early human B cell precursors
24. Bossowski A, Moniuszko M, Dabrowska M, Sawicka B, Rusak M, Jeznach M, et al. Lower proportions of CD4+CD25(high) and CD4+FoxP3, but not CD4+CD25+CD127(low) FoxP3+ T cell levels in children with autoimmune thyroid diseases
25. Mao C, Wang S, Xiao Y, Xu J, Jiang Q, Jin M, et al. Impairment of regulatory capacity of CD4+CD25+ regulatory T cells mediated by dendritic cell polarization and hyperthyroidism in Graves’ disease
. J Immunol
26. Glick AB, Wodzinski A, Fu P, Levine AD, Wald DN. Impairment of regulatory T cell function in autoimmune thyroid disease