Demyelinating Disease Associated with Use of Etanercept in Patients with Seronegative Spondyloarthropathies
To the Editor:
Etanercept is a recombinant tumor necrosis factor-α (TNF-α) receptor immunoglobulin fusion protein. Its anti-TNF properties have gained approval for treatment of rheumatoid arthritis (RA), juvenile RA, psoriatic arthritis (PsA), Crohn's disease, and ankylosing spondylitis1. Etanercept binds specifically to TNF and blocks its interaction with surface TNF receptors. It inhibits binding of both TNF-α and TNF-ß to cell-surface TNF receptors, making TNF biologically inactive. Etanercept also modulates biological responses that are induced or regulated by TNF, including expression of adhesion molecules responsible for leukocyte migration, serum levels of cytokines, and serum levels of matrix metalloproteinase-3.
Data demonstrate a sustained benefit of etanercept treatment in patients with PsA, including inhibition of radiographic progression2. However, through clinical experience, etanercept has been linked to a potential increased risk of demyelinating disease. We describe a case series of 3 patients followed at the Carilion Clinic Rheumatology Clinic, Roanoke, Virginia, who were all diagnosed with demyelinating disease during the course of treatment with etanercept for PsA, iridocyclitis, and ankylosing spondylitis (AS). Our clinic follows 179 patients with spondyloarthritis; 81 patients (45%) are receiving TNF blockers, and 54 of them are receiving etanercept. We also follow 295 patients with RA reciving TNF blockers (36% of the RA population of the clinic). One hundred twelve patients of the 295 receive etanercept.
Case 1 was a 53-year-old man with a history of PsA, degenerative joint disease, dyslipidemia, and allergic rhinitis, who presented in 2001 for followup of severe PsA. He had been diagnosed 15 years before presentation. He had previously been treated with prednisone, methotrexate (MTX), and multiple nonsteroidal antiinflammatory agents, but continued to show progression of his disease. In 2002, he started taking etanercept and continued MTX, with some improvement of his symptoms. However, 6 months later he complained of numbness to his left anterior thigh, which spread to his entire left lower extremity and lower abdomen. Later he developed numbness to his right foot that was associated with mild back pain. He denied any family history of demyelinating disease. He was initially evaluated with lumbosacral spine radiographs, which showed prominent lumbar lordosis, and a diagnosis of acute sciatica, degenerative joint disease, and questionable S1 root versus pyriformis syndrome was made. He continued to have worsening back pain, with new symptoms of numbness to the left side of his face, lower extremity parasthesias, and urinary incontinence. Magnetic resonance imaging (MRI) of the thoracic spine revealed an intramedullary lesion at the C7-T1 level consistent with demyelination. Spinal fluid analysis was abnormal, with mildly elevated proteins, presence of oligoclonal bands, and increased IgG index. Other laboratory studies were not supportive of any other etiology of his symptoms. A diagnosis of demyelinating disease was made and etanercept was discontinued. He was treated with interferon-ß (IFN-ß) 3 times weekly and given baclofen for leg cramps, with improvement of his symptoms. He had 2 subsequent flares of his multiple sclerosis (MS)-like symptoms that were treated with steroids, and he was started on IFN-ß-1a for prophylaxis of his demyelinating process in early 2004. His last flare requiring corticosteroids was in mid-2004 and his disease process has been stable since then.
Case 2 was a 42-year-old man who had a history of hyperlipidemia and PsA diagnosed 13 years before presentation in January 2004. He too had been treated previously with MTX and steroids. In addition, he had been given sulfasalazine, but continued to show progression of his disease. Nine months later, routine followup revealed worsening synovitis and he was started on etanercept. He was clinically improved taking etanercept, but 21 months later complained of headache, fatigue, and visual disturbance. This was later followed by back pain and numbness. He had no known family history of demyelinating disease. He was evaluated by a neurologist, who proceeded with investigation appropriate for MS. Laboratory studies revealed erythrocyte sedimentation rate 47 ml/h; and Lyme titer, dsDNA, antinuclear antibody, methylmalonic acid, homocysteine, and angiotensin-converting enzyme levels were all within normal limits. Cerebrospinal fluid (CSF) showed mildly increased protein and an increased IgG index. MRI of the spinal cord showed plaques at C5-C6, with subtle enhancement suggestive of MS. Etanercept was discontinued and the patient continued MTX therapy. No IFN was administered, but neurological symptoms improved with cessation of etanercept. He restarted daily steroids for PsA, but it remains poorly controlled without etanercept. In September 2007 he was given IFN by his neurologist after having a relapsing remitting course of his neurological symptoms.
Case 3 was a 51-year-old woman with a history of iron deficiency anemia and rhinosinusitis. She was HLA-B27-positive and based on the modified New York criteria fit the diagnosis of AS. Her symptoms included lower back pain and hip pain for > 3 months, morning stiffness, and iridocyclitis. A plain lumbosacral radiograph revealed she had loss of her sacroiliac joints, consistent with grade 4 sacroiliitis. She was being treated with MTX, prednisone, and nonsteroidal antiinflammatory drugs. She continued to show disease progression over a 3-year period and was started on etanercept; 18 months after initiation of etanercept she began to have paresthesias and progressive numbness of the fingertips. This was followed by numbness to the lower chest. Neurologic consultation was sought and there was a suspicion that her symptoms represented the MS-like syndrome associated with TNF-α agents. Investigations revealed an increased serum protein. CSF showed an increased immunoglobulin/albumin ratio. VDRL, Cryptococcal antigen, and India ink and Lyme studies were all negative. Thyroid-stimulating hormone and folate were within normal limits. MRI of the spine was positive for increased signal from C1–T4 with definite enhancement. MRI of the brain showed changes consistent with advanced demyelination in keeping with MS. Etanercept was discontinued in December 2006, with mild improvement in her neurological symptoms. At followup in May 2007 she continued to have abdominal numbness and pain and was scheduled to be seen again by her neurologist. To date she has not been started on a course of interferon.
The longterm safety of etanercept was compared between 597 elderly subjects (≥ 65 yrs) and 3296 younger subjects (< 65 yrs) in a metaanalysis of 22 etanercept clinical studies3. The analysis included 22 rheumatoid arthritis (RA), 2 psoriatic arthritis (PsA), and 2 AS studies. Eight cases of demyelinating disease were observed in subjects aged < 65 years, 6 in RA and 2 in PsA; no cases were found in patients age > 65 years. Another publication reporting data from September 1998 and June 2003 evaluated the rate of adverse events associated with the use of etanercept, infliximab, leflunomide, and MTX, in which the rate of demyelinating disorders was 29.96 per 100,000 patient-years in the etanercept patients3. A publication from the US Food and Drug Administration's FDA Medwatch system in December 2001 reported 20 cases of neurological disease, 18 after etanercept treatment and 2 after infliximab4.
Cases of transverse myelitis, optic neuritis, MS, and new onset or exacerbation of seizure disorders have been observed in association with etanercept therapy. The causal relationship between these events and etanercept therapy remains unclear. The reported rate of demyelination from postmarketing experience likely underestimates the incidence of these events due to underreporting, and the total number of patients exposed to etanercept being an estimate3. Therefore it is difficult to state whether the rate of demyelinating disease noted in our patient population is higher than expected, with 2.5% of our patients receiving etanercept demonstrating a clinical picture in keeping with demyelinating disease.
No clinical trials have been performed evaluating etanercept therapy in patients with MS. However, other TNF-α antagonists administered to patients with MS have been associated with increases in disease activity. The exacerbation of MS appeared to occur as a result of neutralization of TNF-α rather than lymphotoxin. In addition, TNF-α antagonists cannot penetrate the blood-brain barrier to neutralize CNS TNF demyelination. This association has been seen with infliximab, in which clinical studies have demonstrated an increase in disease activity with MRI evidence of worsening disease5. One report on demyelination occurring during anti-TNF therapy showed that the time interval from receiving the drug to the presentation of symptoms was on average 2–6 months for those with PsA, but had an even wider range in patients with RA, from 1 dose of the medication to 10 months6. Interestingly, some of our cases showed a much longer time lag. Data from Amgen reported cases in which the interval between introduction of etanercept and neurologic complaints was 9–17 months. No case of demyelinating disease was observed in patients older than age 65 years3.
Our case series emphasizes that the neurologic deficits from etanercept occurred in an unpredictable manner when viewed in relation to the initiation of the drug. Also, a wide spectrum of deficits appeared to be associated with etanercept. Our 3 patients had no personal or family history of MS, which favors the postulate of de novo occurrence of these neurologic deficits with use of TNF-α antagonist. We concur with recommendations that patients be diligently screened using a comprehensive history and neurological examination before starting TNF-α antagonists such as etanercept, and that the physician remain vigilant for symptoms of development of demyelinating disease when seeing patients for routine followup during treatment with these agents. All cases of new-onset demyelination should be identified to determine what if any risk factors they might share7. It would also seem reasonable to avoid the use of anti-TNF-α agents in patients with established demyelinating disease and to immediately discontinue therapy and pursue diagnostic tests in any patient with suspected demyelination8. Current guidelines are to avoid the use of anti-TNF drugs in individuals with a history of MS or demyelinating disease. However, although the overall number of individuals with new-onset demyelinating events undergoing anti-TNF therapy appears to be small, these episodes can be clinically silent, making it difficult to assess the actual numbers of affected individuals7. This emphasizes the need to establish registries of patients receiving these agents; these registries should include a comparison cohort of patients not receiving biological treatments. It will likely be many years before definitive answers are available9.
STACY A. DAVIS, MD, Department of Internal Medicine, Carilion Clinic; ROBERT R. JOHNSON, MD, FACP, FACR, Department of Rheumatology, Carilion Clinic, Associate Professor of Clinical Medicine, University of Virginia, Associate Professor of Clinical Medicine, Edward Via Virginia College of Osteopathic Medicine; JOHN W. PENDLETON, MD, FACR, Department of Rheumatology, Carilion Clinic, Assistant Professor of Clinical Medicine, University of Virginia School of Medicine, Roanoke, Virginia, USA.
2. Mease PJ, Kivitz AJ, Burch FX, et al. Continued inhibition of radiographic progression in patients with psoriatic arthritis following 2 years of treatment with etanercept. J Rheumatol 2006;33:712-21. [MEDLINE]
3. Amgen Medical Information Department; data on file, Feb 2007,V5.0.
5. Weisman MH. What are the risks of biologic therapy in rheumatoid arthritis? An update on safety. J Rheumatol 2002;29 Suppl 65:33-8.