OBJECTIVE: To establish the first evidence-based diagnostic criteria for atopic myelitis (AM) enabling it to be discriminated from myelitis-onset multiple sclerosis (MS), which is a difficult differential diagnosis. METHODS: Sixty-nine consecutive AM patients examined from 1996 to 2010 at Kyushu University hospital, who fulfilled the empirical definition of AM (2003), and 51 myelitis-onset MS patients in whom allergen-specific IgE was measured, were enrolled. The first available brain MRI findings were compared between the two. Then, we compared the clinical and laboratory features between the 16 AM cases who did not meet the Barkhof brain MRI criteria for MS after more than 5 years follow-up and 51 myelitis-onset MS cases. Based on the discriminative findings, we established diagnostic criteria for AM and calculated the sensitivity and specificity. RESULTS: AM patients had a significantly lower frequency of Barkhof brain lesions on baseline MRI than myelitis-onset MS patients. AM patients had a significantly higher frequency of present and/or past history of atopic disease and hyperIgEemia, and higher cerebrospinal fluid levels of interleukin 9 and CCL11/eotaxin, but a lower frequency of oligoclonal IgG bands than myelitis-onset MS patients. Our proposed diagnostic criteria for AM demonstrated 93.3% sensitivity and 93.3% specificity for AM against myelitis-onset MS, with 82.4% positive predictive value and 97.7% negative predictive value. CONCLUSION: Our first evidence-based criteria for AM show high sensitivity and specificity, and would be useful clinically.
OBJECTIVE: To establish the first evidence-based diagnostic criteria for atopic myelitis (AM) enabling it to be discriminated from myelitis-onset multiple sclerosis (MS), which is a difficult differential diagnosis. METHODS: Sixty-nine consecutive AM patients examined from 1996 to 2010 at Kyushu University hospital, who fulfilled the empirical definition of AM (2003), and 51 myelitis-onsetMSpatients in whom allergen-specific IgE was measured, were enrolled. The first available brain MRI findings were compared between the two. Then, we compared the clinical and laboratory features between the 16 AM cases who did not meet the Barkhof brain MRI criteria for MS after more than 5 years follow-up and 51 myelitis-onsetMS cases. Based on the discriminative findings, we established diagnostic criteria for AM and calculated the sensitivity and specificity. RESULTS: AM patients had a significantly lower frequency of Barkhof brain lesions on baseline MRI than myelitis-onsetMSpatients. AM patients had a significantly higher frequency of present and/or past history of atopic disease and hyperIgEemia, and higher cerebrospinal fluid levels of interleukin 9 and CCL11/eotaxin, but a lower frequency of oligoclonal IgG bands than myelitis-onsetMSpatients. Our proposed diagnostic criteria for AM demonstrated 93.3% sensitivity and 93.3% specificity for AM against myelitis-onsetMS, with 82.4% positive predictive value and 97.7% negative predictive value. CONCLUSION: Our first evidence-based criteria for AM show high sensitivity and specificity, and would be useful clinically.