Tocilizumab for the treatment of large-vessel vasculitis (giant cell arteritis, Takayasu arteritis) and polymyalgia rheumatica




The interleukin-6 pathway is up-regulated in giant cell arteritis (GCA), Takayasu arteritis (TA), and polymyalgia rheumatica (PMR). We retrospectively assessed the outcomes of 10 patients with relapsing/refractory GCA, TA, or PMR treated with tocilizumab (TCZ).


Patients with GCA (n = 7), TA (n = 2), and PMR (n = 1) received TCZ. Seven subjects had failed at least 1 second-line agent. The outcomes evaluated were symptoms of disease activity, inflammatory markers, ability to taper glucocorticoids, and cross-sectional imaging when indicated clinically.


The mean followup time of this cohort since diagnosis was 27 months (range 16–60 months). The patients were treated with TCZ for a mean period of 7.8 months (range 4–12 months). Before TCZ therapy, the patients experienced an average of 2.4 flares/year. All patients entered and maintained clinical remission during TCZ therapy. The mean daily prednisone dosages before and after TCZ initiation were 20.8 mg/day (range 7–34.3 mg/day) and 4.1 mg/day (range 0–10.7 mg/day), respectively (P = 0.0001). The mean erythrocyte sedimentation rate declined from 41.5 mm/hour (range 11–68 mm/hour) to 7 mm/hour (range 2.2–11.3 mm/hour; P = 0.0001). The adverse effects of TCZ included mild neutropenia (n = 4) and transaminitis (n = 4). One patient flared 2 months after TCZ discontinuation. An autopsy on 1 patient who died from a postoperative myocardial infarction following elective surgery revealed persistent vasculitis of large and medium-sized arteries.


TCZ therapy led to clinical and serologic improvement in patients with refractory/relapsing GCA, TA, or PMR. The demonstration of persistent large-vessel vasculitis at autopsy of 1 patient who had shown a substantial response requires close scrutiny in larger studies.


Glucocorticoids (GCs) remain the principal therapy for large-vessel vasculitis (LVV; giant cell arteritis [GCA] and Takayasu arteritis [TA]) and polymyalgia rheumatica (PMR) (1, 2). Although these agents generally control clinical evidence of inflammation when administered in moderate to high doses, a sizeable subset of patients (27–85%) with these conditions experience recrudescence of disease activity upon GC tapering or discontinuation (3–7). Such patients require prolonged or repetitive GC courses that result in a host of undesirable side effects. In GCA, long-term GC therapy is associated with adverse effects in the majority of patients, including fractures, diabetes mellitus, infection, cataracts, hypertension, avascular necrosis, and gastrointestinal bleeding (4). Therefore, a major unmet need in this spectrum of disease exists for more specific drugs to induce remission, maintain remission, and reduce the cumulative adverse effects of long-term GC exposure.

Experience using conventional immunosuppressive drugs has been mixed in LVV and PMR, but a clearly effective GC-sparing alternative has not been identified (6–14). In the setting of conflicting results (6–8, 15), methotrexate (MTX) has shown a weak positive signal in terms of remission maintenance in GCA, but prevention of GC-related adverse events could not be demonstrated. In a study by Jover et al (6), despite the lower cumulative dose of prednisone used, 90% of patients in the MTX group had at least 1 adverse event that was definitely or probably related to GCs, and 45% of these patients flared at least once. Tumor necrosis factor α (TNFα) blockade in GCA and PMR failed in controlled studies (16, 17), and the reported beneficial effects of MTX and TNFα inhibitors in TA (18, 19) have not been subject to scrutiny in randomized trials.

Interleukin-6 (IL-6) is a pleiotropic cytokine with a myriad of metabolic, regenerative, and proinflammatory effects that vary according to the target cell (20). IL-6 participates in the regulation of immune responses, acute-phase reactions, hematopoiesis, and bone metabolism, and there is increasing evidence that this molecule is important in the pathogenesis of diverse inflammatory conditions, including LVV and PMR. Patients with GCA, TA, and PMR have elevated concentrations of IL-6 in both their peripheral circulation and their inflamed tissues, and serum levels of IL-6 correlate with disease activity (21–24). Therefore, the IL-6 pathway is an attractive target for therapy.

IL-6 exerts its biologic activity through 2 molecules, an IL-6–specific receptor (IL-6R) and a signal transducer, gp130. The IL-6R exists in transmembrane (mIL-6R) and soluble (sIL-6R) isoforms generated by alternative splicing or enzymatic cleavage. Most cells express gp130 and are able to respond to the complex IL-6/sIL-6R by a process called trans-signaling, but only certain cell types (particularly neutrophils, macrophages, hepatocytes, and some T cell subsets) carry the transmembrane form (mIL-6R) and are directly stimulated by this mediator (classic signaling) (20).

Tocilizumab (TCZ) is a humanized anti-human IL-6R antibody that inhibits both isoforms of IL-6R and has antiinflammatory effects in conditions such as rheumatoid arthritis (25), juvenile idiopathic arthritis (26), Crohn's disease (27), adult-onset Still's disease (28), and Castleman's disease (29). Although TCZ has been studied in large randomized clinical trials of rheumatoid arthritis and juvenile idiopathic arthritis and in smaller trials of inflammatory bowel disease, there are only few studies in the literature regarding its use in GCA, TA, or PMR. In this study, we assessed the clinical, serologic, and radiographic outcomes of 10 patients with LVV or PMR treated with TCZ.

Significance & Innovations

  • Interleukin-6 (IL-6) receptor blockade may maintain disease remission and decrease the cumulative glucocorticoid exposure in patients with the giant cell arteritis spectrum of disease.

  • Our experience is the largest reported to date and the only one to provide information on treatment for at least 6 months in the majority of cases.

  • The autopsied case that we report also provides an important cautionary note, indicating the possibility that large-vessel vasculitis may persist in some cases despite apparent clinical control with anti-IL-6 receptor therapy.


The institutional review board at our hospital approved the retrospective collection of the clinical data reported in this study. This retrospective investigation was intended to evaluate the potential utility of anti–IL-6R therapy in facilitating the induction of remission, the tapering of GCs, and the prevention of disease exacerbation (flare) in a group of patients with LVV or PMR. Ten patients (7 women and 3 men) met the American College of Rheumatology classification criteria for GCA (30) or TA (31), or had a clinical diagnosis of PMR by the Jones and Hazleman criteria (32). All cases were refractory to conventional therapies or included intolerable adverse effects from treatment with these agents. TCZ was offered as a compassionate off-label option after discussing the risk/benefit ratio of this strategy with each particular individual.

Eight patients were treated with TCZ 8 mg/kg/month and 2 (cases 9 and 10, both GCA patients) (Tables 1 and 2) received TCZ 4 mg/kg/month. The patients studied all had a history of refractory/relapsing disease; at the time of TCZ initiation, 8 patients had clinical signs or symptoms of active disease despite a median prednisone dosage of 15 mg/day (range 5–60 mg/day). A subject with TA (case 3) received TCZ, given the radiographic evidence of arterial inflammation by positron emission tomography–computed tomography angiography (PET-CTA), and pulmonary artery stent stenosis despite a previous course of prednisone (Figure 1). An individual with PMR who was previously treated with prednisone for 19 months (case 7) was offered TCZ after she declined further GC therapy for persistently active disease.

Table 1. Baseline characteristics of 10 patients with large-vessel vasculitis or PMR treated with IL-6 receptor blockade*
Case no., diagnosisHistorical disease featuresBiopsyCross-sectional imaging (CTA, MRA)Before TCZ treatment
Followup, monthsNo. of flaresMean prednisone dosage, mg/dayMean ESR, mm/hourMean CRP level, mg/literMean IL-6, pg/mlOther specific treatments
  • *

    Disease flare was defined as the unequivocal presence of signs or symptoms of active giant cell arteritis (GCA), polymyalgia rheumatica (PMR), or Takayasu arteritis (TA) that required therapy augmentation to be controlled (i.e., increment in the glucocorticoid dose and/or the addition of another immunomodulatory therapy). See the Methods for a detailed description of parameters of disease activity. IL-6 = interleukin-6; TCZ = tocilizumab; CTA = computed tomography angiography; MRA = magnetic resonance angiography; ESR = erythrocyte sedimentation rate; CRP = C-reactive protein; TAB = temporal artery biopsy; N/A = not applicable; CYC = cyclophosphamide; MTX = methotrexate; AZA = azathioprine.

1, GCAHeadache, jaw claudication, amaurosis fugax, PMRTAB (+)N/A11534.330.320.99.3CYC
2, GCAHeadache, fever, PMRTAB (+)N/A22724.135.524.42.8MTX, AZA
3, TAPulse deficit, upper extremity claudication, pulmonary hypertensionN/APulmonary, carotid, and subclavian artery stenosis15Pulmonary stent stenosis8.7119.315.1None
4, GCAHeadache, jaw claudication, upper extremity claudication, PMRTAB (−)Carotid, subclavian, distal aorta, and iliac artery stenosis25417.536.924.312Infliximab, CYC
5, GCAHeadache, jaw claudication, diplopia, PMR, peripheral neuropathyTAB (+)N/A1832332.932.834CYC
6, TAPulse deficit, lower extremity claudication, hypertensionAorta (+)Aortic arch aneurysm and external iliac artery stenosis12Active inflammation on biopsy13.34518.66.3Infliximab
7, PMRPMRN/AN/A193766.530.115.25None
8, GCAHeadache, jaw claudicationAorta (+)Aortic arch aneurysm10232.249.1338.7MTX
9, GCAHeadache, jaw claudication, PMRTAB (+)N/A53818.240.324.4N/AMTX, adalimumab, etanercept
10, GCAJaw claudication, fever of unknown etiologyTAB (+)N/A10329.86872N/ANone
Table 2. Parameters of disease activity at the time of IL-6 receptor blockade initiation and response to treatment in 10 patients with large-vessel vasculitis or PMR*
Case no., diagnosisAt the time of TCZ initiationDuring TCZ treatment
Active featuresPrednisone dosage, mg/dayESR, mm/hourCRP level, mg/literFollowup, monthsNo. of flaresMean prednisone dosage, mg/dayPrednisone dosage at end of followup, mg/dayMean ESR, mm/hourMean CRP level, mg/literMean IL-6, pg/mlOther specific treatment
  • *

    Disease flare was defined as the unequivocal presence of signs or symptoms of active giant cell arteritis (GCA), polymyalgia rheumatica (PMR), or Takayasu arteritis (TA) that required therapy augmentation to be controlled (i.e., increment in the glucocorticoid dose and/or the addition of another immunomodulatory therapy). See the Methods for a detailed description of parameters of disease activity. IL-6 = interleukin-6; TCZ = tocilizumab; ESR = erythrocyte sedimentation rate; CRP = C-reactive protein; FDG = fluorodeoxyglucose; PET = positron emission tomography; N/A = not applicable.

  • The patient had a fatal postoperative myocardial infarction.

  • The patient declined glucocorticoid therapy.

  • §

    The patient received TCZ 4 mg/kg.

1, GCAHeadache, PMR30120.6110407.60.747.9None
2, GCAHeadache153113602.704.60.4114.2None
3, TAPulmonary stent stenosis; bilateral subclavian, right carotid, and left lower lobe pulmonary artery FDG uptake0157.6100 (PET abnormalities resolved)003.60.330None
4, GCAHeadache, jaw claudication156451.6120 (adrenal insufficiency)3.8170.7155.6None
5, GCAHeadache, jaw claudication8506260 (postoperative myocardial infarction)5.445.22.6239None
6, TALower extremity claudication53417901.808.75.668None
7, PMRPMR 08125.8400011.318.6None
8, GCAAortitis609033704.809.20.6124None
9, GCA§PMR154034708.262.25N/ANone
10, GCA§Jaw claudication, myalgias20113726010.7510.821.2N/ANone
Figure 1.

Positron emission tomography with computed tomography (PET-CT) before and after interleukin-6 receptor blockade. A,Sketch representing the location of the cross-sectional image cut. B, Merged PET-CT image and C, PET image of a patient with Takayasu arteritis showing intense fluorodeoxyglucose uptake associated with intimal thickening of the left subclavian artery just distal to its origin. D, Followup PET scan showing interval resolution of the previously seen uptake after 4 infusions of tocilizumab.

Before treatment with TCZ, 7 of the patients had failed treatment with up to 3 cytotoxic, immunomodulatory, or biologic agents, including MTX (n = 3), azathioprine (n = 1), cyclophosphamide (n = 3), adalimumab (n = 1), etanercept (n = 1), and infliximab (n = 2). The GCA patients were all taking aspirin (81 mg/day) at baseline. None of the patients increased their GC dose or received steroid pulses at the time of TCZ administration.

Clinical improvement was assessed by evaluating signs and symptoms of disease activity, ability to taper prednisone, and serial cross-sectional imaging when clinically indicated before and during anti–IL-6R therapy. Remission was defined as the absence of any clinical sign or symptom of active disease. A flare or active disease was defined as the unequivocal presence of signs or symptoms of active GCA, PMR, or TA that required therapy augmentation in order to be controlled (i.e., the increment of the GC dose and/or addition of another immunomodulatory therapy). The signs and symptoms of active GCA included fever (temperature ≥38°C or ≥100.4°F) without alternative explanation (i.e., infection), localized headache, temporal artery or scalp tenderness, visual signs or symptoms (such as acute vision loss due to arteritic anterior ischemic optic neuropathy, transient blurry vision, amaurosis fugax, or diplopia), jaw claudication, new or worsened extremity claudication, new or worsened vascular lesions on cross-sectional imaging or conventional angiography, or evidence of active vascular inflammation on arterial biopsy (arterial wall mononuclear cell infiltrates with or without the presence of giant cells). Active PMR was defined as shoulder and hip girdle pain associated with significant morning stiffness lasting for >45 minutes. Active TA was considered to be present in the setting of new or worsened vascular lesions on cross-sectional imaging or conventional angiography, new or worsened transient and otherwise unexplained ischemic manifestations (i.e., extremity claudication), appearance of new bruits or new peripheral pulse asymmetries, fever without alternative explanation, or active inflammation seen on blood vessel biopsy.

The erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) level were measured serially, even though these acute-phase reactants are expected to normalize with adequate IL-6R blockade and therefore their utility in monitoring response to treatment is theoretically limited. Before TCZ therapy, ESR and CRP level were measured every 1–3 months in all cases. In order to account for missing data and to obtain monthly ESR and CRP level, gaps were completed by averaging the closest available values. Once TCZ was initiated, acute-phase reactants were assessed monthly. With these results, mean ESR and CRP level before and after TCZ therapy were calculated for each particular patient. In 8 patients, serum IL-6 concentrations were checked at baseline and before each monthly TCZ infusion thereafter. As an estimate of the cumulative GC exposure, the mean prednisone dose used during the course of the disease before and during TCZ therapy was calculated for each patient.


Baseline patient characteristics.

In total, 10 patients (7 with GCA, 2 with TA, and 1 with PMR) received TCZ. Their mean ages were 69 years (range 60–83 years) for the GCA/PMR patients and 41 years (range 35–48 years) for the TA patients. The mean duration of disease at the time of anti–IL-6R therapy initiation was 20 months (range 10–53 months), and the mean followup time during TCZ treatment was 7.8 months (range 4–12 months). The patients experienced an average of 2.4 disease flares in the year that preceded their TCZ treatment (range 1–8 flares). Other baseline clinical characteristics of the patients are shown in Table 1.

Clinical response.

Within 8–12 weeks, all patients had clinical improvement and were able to taper their daily dose of prednisone to ≤6 mg. The disease activity features in each individual patient by the time of the first TCZ infusion and their response to treatment are shown in Table 2. Clinical remission was achieved in all patients and was maintained throughout the duration of TCZ treatment. The mean ESR declined from 41.5 mm/hour (range 11–68 mm/hour, normal value <17) in the period before TCZ therapy to 7 mm/hour (range 2.2–11.3 mm/hour) during TCZ treatment (P = 0.0001) (Figure 2). The mean CRP concentration declined from 29 mg/liter (range 9.3–72 mg/liter, normal value <8.0) before anti–IL-6R therapy to 3.8 mg/liter (range 0.3–21.2 mg/liter) during IL-6R blockade (P = 0.0001) (Figure 2). PET scans before and at 4 months after the start of TCZ therapy in 1 patient with TA showed complete resolution of high 18-fluorodeoxyglucose uptake in the carotid, subclavian, and pulmonary arteries (Figure 1).

Figure 2.

Acute-phase reactants and prednisone dose before and after interleukin-6 receptor (IL-6R) blockade. Average of the mean erythrocyte sedimentation rate (ESR), C-reactive protein (CRP) level, and prednisone dose before and after the initiation of tocilizumab (TCZ) therapy in a group of patients with relapsing/refractory giant cell arteritis, Takayasu arteritis, and polymyalgia rheumatica. (Note that CRP level and ESR do not reliably reflect active inflammation during anti–IL-6R therapy).

GC sparing.

All patients achieved substantial reductions in their daily prednisone doses once TCZ was initiated. To estimate the cumulative exposure to GCs, we calculated the average of the mean GC doses for each patient over their entire disease course prior to the initiation of TCZ, and calculated the same measurement during anti–IL-6R therapy. This calculation showed that the cohort had been exposed to a mean prednisone dosage of 20.8 mg/day (range 7–34.3 mg/day) before starting IL-6 blockade (Table 1 and Figure 2) and 4.1 mg/day (range 0–10.7 mg/day) during the time they received TCZ (P = 0.0001) (Table 2 and Figure 2). The 6 patientswho were taking the highest prednisone dosages at the time of the first TCZ administration (≥15 mg/day of prednisone) were able to taper their GC dose by 60–90% within 8–12 weeks. No patients needed an increase in the amount of prednisone to control their disease following the initiation of TCZ. Other immunosuppressants were discontinued in all cases. At the end of the followup, only 3 subjects remained on a low-dose prednisone taper at a mean followup time of 8 months after starting TCZ (Table 2). One of these patients received GCs for adrenal insufficiency rather than GCA.

Serum IL-6 concentrations.

Despite the patients' baseline treatments, serum IL-6 concentrations were elevated before TCZ therapy in all subjects except 1 GCA patient. The mean pretreatment IL-6 concentration was 12.9 pg/ml (range 2.8–34 pg/ml, normal range 0.31–5). The mean IL-6 level during anti–IL-6R treatment was 98.4 pg/ml (range 8.6–239 pg/ml). In 7 of 8 cases (5 GCA and 2 TA), the mean IL-6 level significantly increased after TCZ infusion to 116.4 pg/ml (range 30.9–308 pg/ml). In 6 cases, this increment was seen within 1 month; in 1 subject with TA, this increment was seen after 3 infusions. IL-6 levels remained elevated and demonstrated wide fluctuations that ranged from 4.8–530 pg/ml (data not shown). Except for a transient normal value in 1 TA patient, the IL-6 concentrations never normalized during the followup period of our cohort.

Adverse outcomes.

All patients tolerated the study medication without infusion reactions. No infectious complications or gastrointestinal perforations occurred. TCZ was associated with leukopenia and neutropenia in 4 patients. The total white blood cell count nadir in these cases was 2,100–3,400/mm3 (normal value >4,000), with a corresponding absolute neutrophil count of 940–1,300/mm3 (normal value >1,500). Four subjects developed mild transaminitis. The aspartate aminotrasferase peaked at values ranging from 43–92 units/liter (normal range 9–32), and the alanine aminotrasnferase peaked at values between 50 and 74 units/liter (normal value 7–30). None of these abnormalities led to a cessation of the medication or modification of the dose. One patient (case 4) developed adrenal insufficiency after the completion of her prednisone taper and had to resume low-dose GC replacement therapy. One GCA patient (case 2) had a disease flare after discontinuing TCZ therapy for 2 months. Case 7 (who had PMR) experienced a liver contusion during a motor vehicle accident, and TCZ therapy was stopped because of her hepatic injury.

Finally, we report the single death in this series (case 5) in detail. An 82-year-old woman with GCA and a history of diverticulitis underwent elective colostomy reversal and died following a postoperative myocardial infarction. A postmortem examination showed evidence of active vasculitis in the large and medium-sized vessels despite normal inflammatory markers. This patient had a history of temporal artery biopsy–proven GCA that had responded initially to prednisone therapy, but flared with an extensive vasculitic neuropathy when her prednisone taper reached 10 mg/day. Subsequent treatment with high doses of GCs and daily cyclophosphamide led to multiple complications, including osteoporosis, severe leukopenia, and multilobar pneumonia. After 2 years of relapsing disease despite immunosuppressive therapy, we elected to treat her with TCZ.

During TCZ therapy (6 months), her inflammatory markers normalized and she was able to taper the prednisone dose by 1 mg every 2 weeks without having an exacerbation of her symptoms. When the daily dose of 4 mg was reached, a reversal of her colostomy was performed. Two days after the elective surgery, she developed shortness of breath unaccompanied by chest pain and died of a myocardial infarction.

The autopsy revealed active GCA involving the brachiocephalic, subclavian, carotid, vertebral, and femoral arteries. The most severe involvement was in the right subclavian artery (Figure 3). The right superficial temporal artery also showed active GCA, as did both the thoracic and abdominal aorta. There was focal mild involvement of the coronary arteries by GCA. The coronary arteries also showed severe atherosclerosis with 80% stenosis. The cause of death, an acute myocardial infarction of the left ventricle, was attributed to the rupture of an atherosclerotic plaque.

Figure 3.

Persistent vasculitis in one patient at autopsy who died of a postoperative myocardial infarction.A, Right subclavian artery at autopsy. A low-power image of the right subclavian artery shows active vasculitis with extensive inflammation in the media and with intimal hyperplasia and narrowing of the lumen.B, Right subclavian artery at autopsy. A high-power histologic image of the right subclavian artery shows extensive inflammatory infiltration of the media with necrosis and giant cell formation (arrow head) indicative of active giant cell arteritis.


We observed good clinical and serologic responses in this group of patients whose LVV and PMR had been severe and highly refractory to GCs and other conventional and biologic immunosuppressive agents. The response to TCZ in our patient cohort is consistent with the concept that IL-6 is an important mediator in the pathogenesis of disorders in this disease spectrum. Our experience is the largest reported to date and the only one to provide information on treatment for at least 6 months in the majority of cases (33–38). The autopsied case that we described also provides an important cautionary note, indicating the possibility that LVV may persist in some cases despite apparent clinical control.

Our results are in concordance with other small series and case reports using TCZ for the treatment of LVV and PMR, but histologic evidence of inflammation despite TCZ treatment has not been described before. Seitz et al (33) reported rapid and complete clinical improvement in a group of 7 patients with relapsing and newly diagnosed LVV (5 with GCA, 2 with TA) treated with TCZ 8 mg/kg for a mean period of 4.3 months. Two subjects received TCZ monotherapy and 5 patients were able to taper their GC dosage significantly from a mean of 26.5 mg/day at the first TCZ administration to 4.5 mg/day at 12 weeks. Evidence of active LVV on baseline magnetic resonance angiography entirely disappeared in 3 GCA patients and improved in 2 TA subjects after 3 months of treatment. Salvarani et al (38) published a pilot study of 4 patients with LVV (2 with GCA, 2 with TA) treated with TCZ 8 mg/kg for 6 months followed by MTX maintenance therapy. Three patients entered complete clinical remission and 1 case achieved a partial response. All cases had improvement of vascular fluorodeoxyglucose uptake seen on baseline PET-CT. In 2 relapsing patients who were taking GCs at the time of TCZ initiation, prednisone was tapered off or gradually decreased to 2.5 mg/day by the end of the study period. Beyer et al (34) successfully treated 3 refractory LVV patients with TCZ 8 mg/kg for 6 months. Radiographic signs of activity assessed in 2 cases by PET-CT resolved, and all individuals were able to taper their daily prednisone dose below 7.5 mg.

IL-6R blockade has a sound pathophysiologic basis for the treatment of LVV. Vascular inflammation in GCA is hypothesized to be driven primarily by 2 lineages of CD4-positive T cells: an interferon-γ–producing Th1 cell line, and IL-17–secreting Th17 lymphocytes (39, 40). These cells infiltrate large and medium-sized blood vessels in response to chemotactic stimuli produced by dendritic cells after recognition of unknown danger signals within the vascular wall (41). IL-6 appears to act in concert with IL-1, TNFα, IL-23, IL-21, and transforming growth factor β (TGFβ) to promote the differentiation of naive T helper cells toward a Th17 phenotype with autoimmune characteristics (39). IL-17, in turn, stimulates macrophages and other resident cells within the vascular wall (endothelium, fibroblasts, and smooth muscle cells) to produce proinflammatory cytokines, including IL-6, thereby creating a positive feedback loop. The blockade of the IL-17/IL-6 cycle is one putative mechanism whereby TCZ could exert beneficial effects in patients with GCA (42, 43).

A second (and not mutually exclusive) potential mechanism of IL-6R antagonism in this disorder is the generation of Treg cells that counterbalance the effects of autoreactive lymphocytes. Th17 and Treg cells share a reciprocal developmental pathway. FoxP3, the master transcription factor of Treg cells, is induced by TGFβ and inhibited by IL-6. Therefore, TCZ could shift the Th17/Treg balance in favor of regulatory cell commitment, leading to the amelioration of disease (44–47).

In 2001, Nishimoto et al successfully treated a patient with refractory TA in the context of ulcerative colitis with TCZ (36). In pulseless disease, early active vascular infiltrates mainly consist of γδ T lymphocytes, natural killer cells, macrophages, and cytotoxic (CD8+) and T helper cells. In contrast, late-stage lesions demonstrate extensive fibrosis, intimal hyperplasia, and luminal narrowing. Patients with TA have increased serum concentrations of IL-6 (48) and strong expression of IL-6 within their aortic tissue (49). Several effects of IL-6 might be relevant in the pathogenesis of TA. IL-6 is known for its ability to trigger T cell and fibroblast proliferation, as well as cytotoxic lymphocyte differentiation. Furthermore, IL-6 stimulates the endothelial expression of intercellular adhesion molecule 1 and monocyte chemotactic protein 1, which in turn regulates the extravasation and chemotaxis of monocytes into inflamed tissues. Suzuki et al demonstrated that the rate of monocyte (U937 cell) adhesion to human umbilical vessel endothelial cells (HUVEC) was increased by pretreating HUVEC with IL-6 + sIL-6R, and this process was inhibited by TCZ (50). Infliximab, an inhibitor of TNFα, has been reported in open-label studies to be effective in TA (19). Because TNFα is upstream of IL-6 in the inflammatory cascade, one potential mechanism of action of infliximab in this setting could be the indirect reduction of IL-6 production.

TCZ binds to either sIL-6R or mIL-6R with an affinity similar to that of IL-6 and competes with the cytokine for its receptor. The administration of TCZ to our cohort of patients increased serum IL-6 levels significantly, an effect previously reported in other conditions (51). This phenomenon is explained by the fact that IL-6 normally undergoes receptor-induced catabolism, and this process is inhibited by IL-6R blockade (52). Circulating TCZ saturates IL-6R active sites, delaying IL-6 clearance and leaving its synthesis unopposed. In this regard, during pathologic states, serum IL-6 concentrations immediately after TCZ administration might reflect the actual state of endogenous cytokine production more accurately than do the concentrations before the initiation of treatment. Conversely, as data from a recently reported study of rheumatoid arthritis patients in Japan suggest (53), it is possible that asymptomatic patients receiving TCZ whose IL-6 levels decline over time are good candidates for lengthening the TCZ dosing interval or even discontinuing this therapy altogether. At this time, the meaning of the IL-6 levels during TCZ treatment for LVV remains unclear. Long-term followup of a larger cohort is required to have a better insight into this problem.

The finding of active LVV in 1 patient at autopsy confirms that TCZ requires rigorous testing in larger trials before it can be regarded as efficacious in this disease spectrum. It is known that LVV can often have an asymptomatic course (54). Moreover, persistent vascular inflammation despite apparent clinical control of the disease with GCs is well described in GCA patients, and some experts believe that GCs do not fully suppress vascular inflammation, even at high doses (55). Deng et al obtained a second temporal artery biopsy in 8 GCA patients 3–9 months after the initiation of treatment with prednisone, and found that all second-side biopsies had vasculitis by histomorphology (mainly interferon-γ–producing Th1 cells) (39). In this regard, the impact of anti–IL-6R therapy on the different inflammatory cell populations in GCA and its pathophysiologic consequences will require further study. It is presumed that blocking IL-6 will impact the Th17 arm, but the effects of this strategy on the Th1 and Treg cell compartments are less predictable. On the other hand, the unreliability of traditional acute-phase reactants for disease monitoring while receiving anti–IL-6 therapy may provide evidence in favor of exploring new biomarkers (e.g., Th1/Th17 signatures, vascular imaging) in order to assess the degree of disease activity in patients receiving TCZ.

Our study has certain limitations. Our conclusions must be circumscribed because of the retrospective nature of this study, the small number of subjects treated, the absence of controls, and the short followup period on the study medication. All of these factors could have distorted the magnitude of the effects and introduced bias. Nevertheless, the swiftness with which our patients' previously refractory clinical manifestations were controlled and their ability to taper GCs clearly give a positive preliminary signal that needs further study.

In summary, our findings are consistent with the concept that IL-6 may be a key mediator in GCA, TA, and PMR. Treatment with TCZ led to prompt clinical, serologic, and radiologic improvement in a group of patients with highly refractory/relapsing disease despite treatment with GCs and second-line immunomodulatory agents. At this juncture, adequately powered randomized controlled trials are needed to properly evaluate the safety and efficacy of TCZ for this spectrum of disease.


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. J. H. Stone 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. Unizony, Arias-Urdaneta, Khosroshahi, Keroack, J. R. Stone, J. H. Stone.

Acquisition of data. Unizony, Arias-Urdaneta, Miloslavsky, Arvikar, Khosroshahi, Keroack, J. R. Stone, J. H. Stone.

Analysis and interpretation of data. Unizony, Arias-Urdaneta, Khosroshahi, Keroack, J. R. Stone, J. H. Stone.