Background.- In this study, we investigated clinical features, the etiology, the predictive value of the clinically established diagnosis, prospective follow-up results and also predictive factors of secondary erythema nodosum (EN) in patients with EN.
Design.- A total of 44 patients (mean age :38) with a diagnosis of EN between 1993 to 2001 in our clinic were included in the study prospectively. A biopsy was performed in 27 of the patients. Patients were considered to have secondary when an underlying condition was found; and to have primary when no such condition was found. For the diagnosis of the underlying diseases, the diagnostic criteria and/or diagnostic methods were used. Categoric and continous variables were compared by using chi-square and Mann-Whitney U tests respectively.
Results.- The majority of the patients were female (Female/Male: 8/1) and half (50%) had a determined etiology. The leading etiology was primary tuberculosis (20%) followed in decreasing order by poststreptococcal (14%), sarcoidosis (9%), inflammatory bowel diseases (IBD) (5%), and Behçet.s syndrome (2%). Eleven (25%) patients had a complaint of cough; the diagnosis was primary tuberculosis in 8 and sarcoidosis in 3. Two patients with arthritis were diagnosed as having Behçet.s disease and Crohn.s disease. Tuberculin skin test was positive in 72% (13/18) and 53% (10/19) of the patients with idiopathic and secondary EN respectively (p:0.22). All the patients were followed for a mean duration of 6.5 years. The nodosities relapsed every year in 73% of idiopathic EN patients but in only one (Behçet.s syndrome) of the secondary ones. A histology consistent with EN was seen in 25 out of 27 patients. The predictive value of clinical diagnosis was determined as 93%. Predictive variables for secondary EN were found as follows: fever (p=0.000), cough (p=0.000), sore throat (p=0.008), not relapsing EN (p=0.000), an abnormal chest X-ray (p=0.000), leukocytosis (p=0.019), an erythrocyte sedimentation rate of higher than 50 mm/hour (p=0.000) and a CRP value of higher than 6-fold (p=0.02). All of the patients had bed rest and was given naproxen. The outcomes were usually favorable within 7 days. The patients with an underlying disease were given the specific treatment.
Conclusion.- EN has been associated with numerous diseases. In order to reduce cost and duration of diagnosis, every population should determine its own most frequent etiologic factors. Predictive variables for secondary EN should also be determined. Our study revealed primary tuberculosis, streptococcal pharyngitis, sarcoidosis, IBD, and Behçet.s syndrome as the main etiologies of EN.