Hashimoto's Thyroiditis Presenting as Acute Painful Thyroiditis and as a Manifestation of an Immune Reconstitution Inflammatory Syndrome in a Human Immunodeficiency Virus–Seropositive Patient

Introduction

A bnormal thyroid function tests are common among patients with advanced human immunodeficiency virus (HIV) infection (1). The association of Graves' disease with an immune reconstitution inflammatory syndrome (IRIS) after initiation of antiretroviral therapy (ART) is well recognized (2). We report a patient who presented with a painful thyroiditis and thyrotoxicosis 10 months after initiation of ART, which was accompanied by the appearance of a high anti-thyroid peroxidase (anti-TPO) antibody titer. The thyrotoxicosis evolved into a persistent hypothyroidism requiring thyroxine replacement therapy, consistent with Hashimoto's thyroiditis.

Patient

A 37-year-old Thai woman was diagnosed with an advanced HIV infection (CD4 cell count: 10 cells/μL and viral load: 800,000 copies/mL) complicated by a cryptococcal meningitis in January 2010. The cryptococcal meningitis was treated with 2 weeks of liposomal amphotericin B and flucytosine, followed by secondary prophylaxis with fluconazole 400 mg daily. Two weeks after presentation, ART (zidovudine/lamivudine and lopinavir/ritonavir) was started; trimethoprim-sulfamethoxazole and isoniazide were given as Pneumocystis jiroveci and tuberculosis prophylaxis.

While still on fluconazole, she presented with a new episode of headache and fever in July 2010. A 2-week course of amphotericin B and flucytosine was repeated for a presumed relapse of cryptococcal meningitis with a positive cerebrospinal fluid (CSF) India ink preparation for cryptococci. Because CSF cultures remained sterile and the headache and fever persisted despite anti-cryptocococcal treatment, an immune reconstitution syndrome against cryptococci was considered and the patient was started on prednisone (30 mg od) with a rapid clinical response. Zidovudine was replaced by raltegravir because of anemia and with tenofovir being contraindicated because of amphotericin B–induced renal dysfunction. The thyrotropin (TSH) level at that time was normal (1.81 mU/L).

In October 2010, the patient was readmitted with fever and anterior neck pain. There had not been any symptoms of an upper respiratory tract infection. Physical examination showed a temperature of 40.1°C, tachycardia, and a painful, not enlarged thyroid gland. C-reactive protein (CRP) levels were elevated (138 mg/L) and thyroid function tests revealed a suppressed TSH level (0.01 mU/L) with an elevated free thyroxine (FT₄) level (27.8 pmol/L), consistent with thyrotoxicosis (Table 1). The CD4 cell count had risen to 150 cells/μL and the viral load had been undetectable since July 2010. Anti-TSH receptor antibodies were negative. In contrast, anti-TPO antibodies were strongly positive (>1000 IU/mL). Retrospective testing for anti-TPO in stored plasma samples showed that the patient had been anti-TPO negative at the start of ART, 10 months earlier. Ultrasonography showed an inhomogeneous aspect of the thyroid gland with alternating areas of hyper- and hypoechogenicity, and multiple surrounding small lymph nodes. A radioactive ¹²³I scan revealed no uptake in the thyroid. It was considered that this episode was another complication of her immune reconstitution inflammatory syndrome and prednisone was increased from 5 mg to 20 mg daily. The neck pain and fever disappeared in the course of 2 weeks and CRP levels returned to normal. The thyrotoxic phase was followed by persistent hypothyroidism for which the patient remains in need of thyroxine supplementation.

Discussion

Our patient presented with an acute painful thyroiditis, which was associated with appearance of anti-TPO antibodies and evolved into a persistent hypothyroidism. The clinical course and thyroid autoantibodies fit the diagnosis of Hashimoto's thyroiditis. Remarkably, an IRIS against cryptococcal meningitis had been diagnosed shortly before the onset of the thyroiditis. That thyroid antibodies may appear after immune restoration on ART and development of Graves' disease, in the setting of immune restoration, is a well-known phenomenon (2,3). Therefore, we speculate that the Hashimoto's thyroiditis in this case was also a manifestation of an IRIS.

To our knowledge, one previous case of painful Hashimoto's thyroiditis following ART initiation has been reported by Chen et al. (3). This 55-year-old woman presented with an acute thyroiditis with anterior neck pain, accompanied by a high CRP level (153 mg/L) and decreased uptake on a technetium-99 scan (0.6%). Thyroiditis in this patient developed 8 months after initiation of ART and was accompanied by appearance of a high anti-TPO antibody titer (1249 IU/mL). As in our patient, overt hypothyroidism developed and the patient remained dependent on thyroxine replacement therapy.

Although Hashimoto's thyroiditis is a common illness, an initial presentation with a painful thyroiditis and thyrotoxicosis is rare (4). Strikingly, both our case and the case described by Chen et al. (3) presented with such a clinical picture and such a presentation might thus be more common in the setting of ART-associated immune restoration. We cannot rule out that the development of thyroiditis was merely coincidental. However, the fact that anti-TPO antibodies were absent before the start of ART makes this, as well as an exacerbation of pre-existing thyroid autoimmune disease, less likely.

The pathogenic mechanisms of autoimmune thyroiditis remain speculative; however, activated CD4 T-cells predominate in the thyroid infiltrate and appear to be a key factor. Initiation of ART restores cellular immunity by a rapid recruitment of memory CD4 T-cells and a slower increase of naive CD4 T-cells produced by the thymus. Especially this naive CD4 T-cell expansion is thought to be involved in ART-associated autoimmune thyroid disease (5). It has been suggested that immune restoration on ART, occurring on a markedly altered T cell repertoire and thymic microenvironment, may result in autoimmunity through disruption of peripheral tolerance and a failure to delete autoreactive T-cell clones in the regenerating thymus (6). Finally, restored immunity against antigens from (opportunistic) microorganisms and microorganism-induced thyroid tissue damage may be involved.

Upon presentation with the acute thyroiditis, our differential diagnosis included a subacute granulomatous thyroiditis (De Quervain's thyroiditis) and an infectious thyroiditis. An association between HLA-B35 and subacute granulomatous thyroiditis has been reported (7) and our patient tested negative for HLA-B35. Opportunistic infections, such as Pneumocystis jiroveci and Cryptococcus neoformans, may also induce thyroiditis (8); however, this diagnosis was considered unlikely given the patient's use of prophylaxis against these microorganisms and the excellent response to prednisone. In conclusion, our case suggests that autoimmune thyroiditis may develop as a complication of immune restoration in patients treated with ART for HIV infection.

Clinicians caring for HIV-infected patients should be aware of this possible association and consider the diagnosis of thyroiditis in patients with compatible symptoms. This is especially important because symptoms of thyroid disease are often nonspecific and may be attributed to ART side-effects or other HIV-related illnesses.