Thalidomide dramatically improves the symptoms of early-onset Sarcoidosis/Blau syndrome: Its possible action and mechanism




Early-onset sarcoidosis (EOS), which occurs in children younger than 5 years of age, is associated with granulomatous lesions and a sporadic genetic mutation of the nucleotide-binding oligomerization domain 2 that causes constitutive NF-κB activation. The symptoms of EOS can be uncontrollable, progressive, and associated with profound complications. However, appropriate therapy is still under investigation. The aim of this study was to assess the efficacy of thalidomide in patients with severe EOS, based on etiology supporting an initial role of NF-κB in activation of this disease.


Thalidomide was given to 2 patients with EOS (a 16-year-old girl and an 8-year-old boy) at an initial dosage of 2 mg/kg/day, and the dosage was increased if necessary. To elucidate the mechanism of the drug, peripheral blood monocytes were isolated from the patients and stimulated with cytokines (macrophage colony-stimulating factor, tumor necrosis factor α, and interleukin-4), and their ability to form multinucleated giant cells (MGCs) and osteoclasts was measured.


Both patients showed dramatic improvement of their clinical symptoms (alleviation of fever and optic nerve papillitis, achievement of a response according to the American College of Rheumatology Pediatric 50 and Pediatric 70 criteria) and laboratory findings. Monocytes from patients with EOS had a greater ability to survive and induce MGCs and osteoclasts than those from healthy control subjects. The formation of MGCs and osteoclasts was inhibited by the presence of thalidomide.


The ability of thalidomide to improve clinical symptoms and laboratory findings in patients with EOS indicates a central role for NF-κB activity in this disorder. Inhibition of IKK might be a pharmacologic action by which thalidomide down-regulates NF-κB signaling. Thalidomide may be an effective medication in patients with severe complications of EOS, including ocular involvement.

Sarcoidosis is a multiorgan inflammatory disease with an unknown etiology, characterized by the histologic features of noncaseating epithelioid granulomas. Early-onset sarcoidosis (EOS), which occurs in children younger than 5 years of age, involves the triad of exanthema, arthritis, and uveitis, along with recurrent fever, without apparent pulmonary involvement (1). The symptoms of EOS are sometimes uncontrolled by corticosteroids, cyclosporine, antimalarials, and/or methotrexate (2). The disease is symptomatically progressive and can have profound complications, such as joint destruction, visceral involvement, and ocular involvement (uveitis and/or retinopathy) with a high probability of blindness (1–3). Severe growth retardation is observed in some patients.

EOS is associated with a sporadic genetic mutation of the nucleotide-binding oligomerization domain 2 gene (NOD2) that causes accelerated basal NF-κB activation (without muramyldipeptide stimulation), which is a genetic etiology it shares with Blau syndrome (4). It has been hypothesized that increased basal NF-κB activity is the key to the pathophysiology in EOS and Blau syndrome, and that such activity is likely related to the severity of the disease and sometimes to its complications. Lately, the term “pediatric granulomatous arthritis” refers to both EOS and Blau syndrome (5). Increased basal NF-κB activity could have a central role in the activation of monocytes and the release of proinflammatory cytokines in EOS.

It has been shown that treatment of adult-onset sarcoidosis with thalidomide decreases the size of the granulomatous lesion in vivo (6–8), then reduces epidermal thickness, modulates macrophage responses, recruits T cells into the granulomas, and finally changes the histologic appearance of granulomas (7). We therefore considered that thalidomide might be a possible alternative treatment step even for patients with EOS, and in the present study we examined the effects of thalidomide in 2 patients with high basal NF-κB activity. Both cases are genetically solitary, and the disease severity was the worst observed among the patients who had been recruited in a previous surveillance study for EOS (4, 9).

The transfected cells with their NOD2 mutants demonstrated accelerated basal NF-κB activities (4, 9). Our 2 patients showed a dramatic response to thalidomide, although the precise pharmacologic actions of the drug were unclear. Therefore, we also tried to elucidate the mechanism of the drug by means of monocyte-culturing analysis.


Patient procedures.

Written informed consent was obtained from the patients and their parents, and the treatment protocol and examination of peripheral blood samples were in accordance with the guidelines of the Institutional Review Board of Okayama University Hospital.

Experimental procedures.


RPMI 1640 and phosphate buffered saline (PBS) were obtained from Gibco (Grand Island, NY). Chemical compounds including thalidomide and muramyldipeptide were purchased from Sigma (St. Louis, MO), and human cytokines (macrophage colony-stimulating factor [M-CSF], interleukin-4 [IL-4], and tumor necrosis factor α [TNFα]) and the control rabbit antibody were obtained from R&D Systems (Abingdon, UK).

Monocyte separation and cell cultures.

Peripheral blood mononuclear cells were isolated by Histopaque (Sigma) density-gradient centrifugation. Human monocytes were isolated from healthy young volunteers (younger than age 20 years) by negative selection with monoclonal antibody (mAb)–coated immunologic magnetic beads and a cell sorter (Miltenyi Biotec, Bergisch Gladbach, Germany), after centrifugation. Purity was assessed by saturation with anti-CD14 mAb (IOM2; Immunotech, Marseilles, France) using a flow cytometer, and was >95%. Monocytes were cultured at 37°C in a 5% CO2 atmosphere in conditioned medium consisting of RPMI 1640 supplemented with 10% heat-inactivated fetal bovine serum, 2 mML-glutamine, 50 units/ml penicillin, 50 μg/ml streptomycin, and 10 mM HEPES at 2 × 106 cells/2 ml/well in 12-well tissue culture plates (Sumitomo Bakelite, Tokyo, Japan) with several cytokines. All culturing procedures were performed under sterile conditions, and each experiment was performed at least in triplicate. The cells were maintained for 4 days, at which time the medium was replaced with an equivalent volume of medium supplemented with various cytokines.

Cell staining and evaluation of multinucleated giant cells (MGCs) or osteoclast formation.

Culture plates were gently rinsed twice with PBS and dried well, and the cells were stained with May-Grünwald-Giemsa stain (Merck, Darmstadt, Germany). Digital pictures were obtained with a microscope (Keyence, Osaka, Japan) at low power (20× objective) under light vision, for the quantification of MGCs (Langerhans-type cells) or osteoclasts. Photomicrographic data are representative of a minimum of 3 separate experiments. The number of visible nuclei within MGCs or osteoclasts per 1 cm2 in area was expressed as the mean percentage of total nuclei counted in 4 representative microscopic fields by an experimenter who was blinded to the experimental protocol.

Generation of NOD2 mutants and NF-κB luciferase assay.

Expression plasmids of NOD2 and its mutants were subcloned into the p3xFLAG-CMV vector, as previously described (4, 9). Mutants were generated using the QuikChange Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA). HEK 293T human embryonic kidney cells (1 × 105) were transfected with 1,000 ng of plasmids containing 100 ng of NF-κB reporter plasmid, as previously described (4, 9). The cells were cultured with or without 5 μg/ml of muramyldipeptide for 12 hours after transfection and measured for NF-κB activity using the PicaGene Dual Luciferase Kit (Toyo Ink, Tokyo, Japan). Wild-type (WT) NOD2 was used as control.

Statistical analysis.

All statistical analyses were performed using Student's t-tests. All statistical tests were 2-tailed. Data are presented as the mean ± SD and were analyzed using statistical software for Excel (Atom, Tokyo, Japan). P values less than 0.05 were considered significant.


Case reports.

Patient 1.

The patient, a 16-year-old Japanese girl, consulted our clinic after experiencing remittent fever, skin lesions with tiny eruptions on the trunk and extremities, and arthritis with bone and joint deformity (knee, hip, hand, foot, fingers) since the age of 6 months. She showed acetabular dysplasia, malalignment of the legs, and lameness due to arthralgia. An initial diagnosis of systemic juvenile idiopathic arthritis had been made by a previous physician, and histopathologic analysis at the age of 9 months revealed noncaseating epithelioid cell granulomas in the skin.

Steroid treatment had been initiated at 14 months of age. Genetic analysis at the age of 13 years revealed a heterozygous missense mutation in the NOD-2 region of the CARD15/NOD2 gene (N670K) (4). At the age of 15 years, she was 135 cm tall and had a height SD score of less than −4.5. Her erythrocyte sedimentation rate (ESR; 45 mm/hour at age 15 years), white blood cell count (16–22 × 109/liter, with 75% segmented neutrophils), and C-reactive protein (CRP) level (1.5–3.0 mg/dl) had increased continuously over the years despite the various immunosuppressive therapies attempted (prednisolone, methotrexate, tacrolimus, infliximab); hypercalcemia was not detected. In addition, beginning at the age of 3 years, recurrent ocular symptoms (uveitis and optic nerve papillitis) had seriously affected her eyesight, with the result that she was nearly blind (visual acuity 0.04 [corrected eyesight 0.05] in the right eye and 0.02 [corrected eyesight 0.05] in the left eye) upon presentation to our clinic in August 2008.

Due to the severity of her symptoms, the risk of blindness, and the ineffectiveness of steroid therapy, we made a decision to administer thalidomide. The clinical use of thalidomide (100 mg/day) was started in September 2008, and it successfully improved the patient's laboratory results (Figure 1a). Her CRP level was decreased and normalized for the first time; surprisingly, her visual acuity was recovered to 0.3 in the right eye and 0.2 in the left eye, and an examination by an ophthalmologist indicated that her papillitis had been alleviated (i.e., suppressed activities of the inflammatory lesions). Arthralgia had been reduced during treatment and an American College of Rheumatology Pediatric 50 (ACR Pedi 50) response (10) was achieved and maintained.

Figure 1.

Clinical courses of patient 1 and patient 2. In patient 1, thalidomide treatment was started at a dosage of ∼2 mg/kg/day, beginning at the end of September 2008 (a). Significant clinical improvement, including resolution of lumbago, arthralgia, and fever, was observed within 3 weeks. Laboratory findings were also improved immediately. In patient 2, thalidomide was increased at a dosage of 3 mg/kg/day (b). PSL = prednisolone; WBC = white blood cell; CRP = C-reactive protein.

Patient 2.

The patient, a boy age 8 years and 8 months, was referred to our clinic after an 8-year history of EOS characterized by remittent fever (>38.5°C), skin lesions, and arthritis (knee, hip, hand, and fingers). His diagnosis had been made by a previous physician at ∼9 months of age, when a skin biopsy revealed a nonspecific granuloma. Prednisolone had been started at that time, and then methotrexate and cyclosporine were added, but an adequate response was not observed. A genetic analysis was performed at 5 years of age and revealed a missense mutation in the NOD-2 region of CARD15/NOD2 (C495Y) (4, 9). At 8 years of age, in the 6 months before his presentation at our clinic, his ESR was >60 mm/hour, and his white blood cell count (12–18 × 109/liter), CRP level (5.8–9.6 mg/dl), and matrix metalloproteinase 3 (MMP-3) level (450–600 ng/ml; reference range 40–121) were continuously high, with destruction of the knee and finger joints despite immunosuppressive treatment (Figure 2). In addition, a recent ocular examination had revealed ocular involvement (optic nerve papillitis). Upon presentation to our clinic in August 2008, he was 93 cm tall, weighed 18 kg, and had a height SD score of less than −6.2.

Figure 2.

Radiographs of the hand and knee joints of patient 2. Radiographs were obtained before thalidomide treatment was begun in August 2008. Osteopenia and bone destruction were observed in the small bones of the hand and knee joints.

Thalidomide was added beginning in October 2008 (final dosage 75 mg/day) and alleviated his arthralgia and high CRP levels. His CRP level was normalized (<0.2 mg/dl) for the first time since the disease began, and the level of MMP-3 was normalized (<120 ng/ml). Etanercept was withdrawn, but because the CRP level rose again, the drug was restarted. The patient's good condition has continued for the last 3 months (Figure 1b); we have been especially pleased that his height has increased by 8 cm over the previous 6 months. An ACR Pedi 70 response was achieved and maintained.

After beginning the thalidomide therapy, neither patient experienced arthralgia and/or lower back pain, and their symptoms, including ocular involvement, were relieved. Treatment was well tolerated by both patients. Major side effects and disturbance of nerve conduction were not seen.

Generation of NOD2 mutants and NF-κB activity.

Both cases are genetically solitary, and the disease severity of the 2 patients was the worst observed among patients who had been recruited in a previous surveillance study for EOS (4, 9). Increased basal NF-κB activity (without muramyldipeptide stimulation) was generated with mutated NOD2, and the spontaneous NF-κB activities were 0.61 and 0.79, respectively, and were significantly higher than the value of 0.05 for WT NOD2 (Figure 3a). R334W, the most frequent mutation in the patients with EOS/Blau syndrome, showed basal NF-κB activity ratio of 0.42.

Figure 3.

Biologic effects of CARD15 variants and osteoclast formation in patients with early-onset sarcoidosis (EOS). a, Muramyldipeptide (MDP)–independent and MDP-dependent NF-κB transactivation by CARD15/NOD2 variants discovered in patients with EOS. HEK 293T cells were cotransfected with the 30-ng expression construct of a CARD15 variant together with the NF-κB reporter plasmid and measured for NF-κB activity after 12 hours of incubation with or without 5 μg/ml of MDP. b, Osteoclast formation by CARD15 variants discovered in patients with EOS. Peripheral blood monocytes were incubated with macrophage colony-stimulating factor (M-CSF; 20 ng/ml) and tumor necrosis factor α (TNFα; 10 ng/ml) or TNFα alone for 14 days. The number of formed osteoclasts (>1,500 μm2) per squared centimeter was counted. Bars show the mean and SD results from triplicate cultures.

Monocytes cultured with cytokines and the effect of thalidomide.

Exposure of peripheral blood monocytes to IL-4 (20 ng/ml) plus M-CSF (20 ng/ml) for 14 days led to their morphologic change into fibrous cells (dendritic cells) and their assembly into large MGCs (Langerhans-type cells), and the combination of TNFα (10 ng/ml) and M-CSF (20 ng/ml) induced the formation of RANKL-positive giant foam cells (osteoclasts) (for review, see refs.11–13). May-Grünwald-Giemsa staining was used to assess the morphology of monocytes incubated with different combinations of cytokines on the culture plate. M-CSF by itself induced survival of monocytes and their differentiation into petty “foamy” macrophages, but neither TNFα alone nor IL-4 alone induced macrophage differentiation in WT monocytes (Fig- ures 4a–h). These experiments were highly reproducible. The patients' monocytes survived longer than WT cells, and the cells differentiated into MGCs and giant “foamy” osteoclasts (Figure 3b) in the presence of an individual cytokine (TNFα or IL-4) without survival factor (M-CSF).

As shown in Figure 5, thalidomide at a pharmacologically relevant concentration of 25 μg/ml significantly inhibited the formation of MGCs and foamy macrophages for 14 days in WT monocytes, and it reduced the percentage of visible nuclei within osteoclasts from 50% to 5% (90% inhibition). Thalidomide is poorly soluble and requires an intermediate stock dilution in DMSO (285 mg/ml:1 mM), followed by storage at −30°C. DMSO is known as an agent to induce differentiation and morphologic changes in monocytic leukemia cell lines (14). However, the final concentration of DMSO was <0.02% in our experiments; it had little effect on the pH of the culture medium and little effect on monocyte–macrophage differentiation at a concentration of 0.02%, as previously reported (15). 4

Figure 5.

Effects of thalidomide on monocytes from patient 1 cultured in the presence of 20 ng/ml of macrophage colony-stimulating factor (M-CSF), with 20 ng/ml of interleukin-4 (IL-4) or 10 ng/ml of tumor necrosis factor α (TNFα) for 14 days. Representative images of May-Grünwald-Giemsa staining of monocyte-derived cells are shown. Inhibition of multinucleated giant cell (Langerhans-type cell) formation and a reduction in the number of osteoclasts (large cells) by thalidomide were observed. These results were highly reproducible.

Figure 4.

Characterization of peripheral blood monocytes cultured with a combination of cytokines for 14 days. Representative images of May-Grünwald-Giemsa staining of monocyte-derived cells cultured with TNFα (10 ng/ml), M-CSF (20 ng/ml), and interleukin-4 (IL-4; 20 ng/ml) are shown. Peripheral blood monocytes were cultured in RPMI 1640 medium supplemented with 10% heat-inactivated fetal bovine serum, 100 units/ml penicillin, and 100 μg/ml streptomycin, at 1 × 106 cells/ml/well in 24-well tissue culture plates for 14 days. Monocytes from healthy volunteers (a–d and h) and monocytes from patient 1 (e–g) were cultured with TNFα alone (a and e), M-CSF alone (b and f), IL-4 alone (c and g), IL-4 plus M-CSF (d), or TNFα plus M-CSF (h). See Figure 3 for other definitions.


EOS/Blau syndrome is a rare systemic granulomatosis that has been associated with mutated NOD2 (4). EOS shares a common genetic etiology with Blau syndrome. In our patients, it was demonstrated that NOD2 genotyping may help predict disease progression, and that clinical disease severity is associated with basal NF-κB activity in transfected cells. In these patients with EOS, thalidomide seemed to lead to complete resolution of granulomatous inflammation and clinical symptoms, an effect that, to our knowledge, has not been previously reported. After starting thalidomide therapy, the patients showed dramatic improvements in clinical symptoms and laboratory parameters. We believe that an improved understanding of the mechanisms by which NOD2 acts in the disease pathogenesis should help in discovering therapeutic targets for the treatment of EOS.

The classic EOS symptom triad is skin rash, arthritis, and uveitis. Histopathologic analysis typically reveals proliferating giant cells and epithelioid cells in noncaseating granulomas in the lesion, which are common to both adult-onset sarcoidosis and EOS. Multinucleated giant cells (Langerhans-type cells) and epithelioid cells are derived from monocyte–macrophage–lineage cells that are key cells in the development and maintenance of sarcoidal granulomatous lesions (16). Diffuse osteopenia and bone destruction are recognized as radiographic manifestations of skeletal sarcoidosis (17), and these findings were also observed in our patients. Osteolytic lesions are associated with the histologic features of accelerated bone remodeling, i.e., increased numbers of osteoclasts and osteoblasts (17). We speculated that increased basal NF-κB activity influences the development of sarcoidal granulomatous lesions and skeletal destruction in patients with EOS. Indeed, monocytes from our 2 patients survived longer than WT cells, and the cells could differentiate into MGCs and giant foam osteoclasts in the absence of survival factor (M-CSF). The monocytes of patients with EOS have been shown to have the potential to easily produce MGCs and osteoclasts in response to inflammatory stimuli. A heightened ability to form MGCs has also been observed in monocytes from patients with adult-onset sarcoidosis (16).

Thalidomide was developed in the 1950s as a sedative drug but was withdrawn in 1961 because of its teratogenic effects (18–20). Recently, this agent has been recognized as a potent immune response–modifying drug (21–24). The suppressive effect of thalidomide on the activation of NF-κB may explain its immune effects. NF-κB is retained in the cytoplasm with IκB and is activated by a wide variety of inflammatory stimuli, including muramyldipeptide and some infectious agents, via the Toll-like receptor, and then is translocated to the nucleus. Thalidomide has been shown to selectively suppress the NF-κB activation induced by mediators of inflammation (25). In our experimental studies, thalidomide at a pharmacologically relevant concentration of 25 μg/ml significantly inhibited the formation of Langerhans-type cells and “foamy” osteoclasts developed from monocytes. Taken together, these results indicate that thalidomide is a promising agent for the reduction of granulomatous inflammation through its suppression of NF-κB activation and interference with the proliferation and differentiation of monocytes.

In conclusion, we performed a pilot study assessing the efficacy of thalidomide in patients with severe EOS, based on etiology supporting an initial role of NF-κB activation in the disease, and conclude that thalidomide is a useful therapeutic option. From the perspective of quality of life, the ocular symptoms of EOS/Blau syndrome require the closest attention. In a previous study, one-third of patients with EOS/Blau syndrome and NOD2 mutations had a poor or extremely poor visual outcome (3, 5, 9); however, it is possible that the extent of visual impairment could be modified by treatment with thalidomide. In the current study, patient 1 was about to go blind, but examination by an ophthalmologist indicated that her retinopathy was improved by thalidomide therapy. Thalidomide can reduce TNFα production by stimulated macrophages (26). In patient 2, the combination of anti–TNFα therapy and thalidomide resulted in improvement of clinical symptoms and laboratory parameters.

Finally, we believe that an improved understanding of basic mechanisms of inflammation (Figure 6) could result in targeted therapies not only for patients with EOS/Blau syndrome but also for those with other autoinflammatory disorders. Thalidomide itself has potential for patients with conditions other than EOS/Blau syndrome and should be considered as a treatment for adult-onset sarcoidosis with severe ocular involvement, to prevent blindness.

Figure 6.

Proposed mechanism of thalidomide in monocyte activation. Thalidomide inhibits activation of IKKα/β and IKKα/α, and then both enzymes induce phosphorylation of IκB and further activate NF-κB and its translocation into nuclei. It is known that the knockout of IKKα/α can induce limb defects in mice. NOD2 is upstream of NF-κB activation and has a leucine-rich repeat portion. TNFR = tumor necrosis factor receptor; MDP = muramyldipeptide; TLR-4 = Toll-like receptor 4; MyD88 = myeloid differentiation factor 88; IL-6 = interleukin-6.


All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Yasui had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Yasui, Manki, Morishima.

Acquisition of data. Yasui, Yashiro, Tsuge, Takemoto, Yamamoto, Morishima.

Analysis and interpretation of data. Yasui, Morishima.


We appreciate the kind assistance of Drs. Ikuo Okafuji and Ryuta Nishikomori (Kyoto University, Japan) in the evaluation of NF-κB activity.