Analysis of diagnostic criteria in adamantiades-behçet disease: a retrospective study.

Adamantiades-Behçet's disease (ABD) is a chronic-relapsing, inflammatory and multi-systemic disease. Any organ or system may be involved: ABD presents a great variety of cutaneous and mucosal lesions, ocular manifestations, central and peripheral nervous system abnormalities, joint as well as gastrointestinal involvement. Since clear pathognomonic clinical features and laboratory tests are lacking, the diagnosis of ABD mainly relies on the characteristic clinical features. Several sets of diagnostic criteria have been used. The International Study Group for Behçet Disease (ISGBD) in 1990 formulated a set of criteria to warrant uniformity of both diagnosis and classification. Therefore, in 2006, a new set was proposed by the International Team for the Revision of the International Criteria for Behçet's Disease (ITR-ICBD) not only to uniform the previous criteria but also to establish best accuracy, along with an optimum sensivity and specificity. The aims of this study are both to analyze the clinical features of ABD patients and to validate the ISGBD and ITR-ICDB criteria for the diagnosis of ABD in our cohort.

What was known?

International Study Group for Behçet Disease (ISGBD) in 1990 formulated a set of criteria to warrant uniformity of both diagnosis and classification of ABD. A new set was proposed in 2006 by the International Team for the Revision of the International Criteria for Behçet's Disease (ITR-ICBD) not only to uniform the previous criteria but also to establish best accuracy, along with an optimum sensivity and specificity.

Introduction

Adamantiades-Behçet's disease (ABD) is a chronic-relapsing, inflammatory and multi-systemic disease whose etiopathogenesis remains unknown.[1] Any organ or system may be involved: ABD presents a great variety of cutaneous and mucosal lesions, ocular manifestations, central and peripheral nervous system abnormalities, joint as well as gastrointestinal involvement.[2] The mucocutaneous lesions of ABD exhibit histological changes of vascular reaction or vasculitis.[3]

The disease is a universal disorder, although it has a higher prevalence and incidence between the Mediterranean area and East of Asia, in a historic trading region called the “silk route”.[2]

Since clear pathognomonic clinical features and laboratory tests are lacking, the diagnosis of ABD mainly relies on the characteristic clinical features and the judgment of the experienced physician.[4] Several sets of diagnostic criteria have been used since Marson and Barnes first relied their criteria in 1969.[5] The International Study Group for Behçet Disease (ISGBD) in 1990 formulated a set of criteria to warrant uniformity of both diagnosis and classification. According to ISGBD, the presence of oral aphtosis was mandatory. Two of the following symptoms were required for the diagnosis of ABD: Genital aphthosis, skin manifestations, ocular lesions and positive pathergy reaction.[6] For more than 10 years, the ISGBD criteria were the most used criteria for diagnosing ABD, however the majority of validation studies failed. Therefore, in 2006, a new set was proposed by the International Team for the Revision of the International Criteria for Behçet's Disease (ITR-ICBD) not only to uniform the previous criteria but also to establish best accuracy, along with an optimum sensitivity and specificity.[2] According to ITR-ICBD, oral aphtosis, skin manifestations, vascular manifestations and pathergy-positive reaction score one point each. Genital ulcers and ocular manifestations get two points each. The diagnosis of ABD is made when three or more points are collected.[7]

The aims of this study are both to analyze the clinical features of ABD patients been addressed to the Clinica Dermatologica of Trieste and to validate the ISGBD and ITR-ICDB criteria for the diagnosis of ABD in our cohort.

Materials and Methods

In the period from 2003 through 2010, 42 (13 male, 29 female) consecutive Italian-caucasians patients were recruited at the Clinica Dermo - Venereologica of Trieste in this validation study. Every patient must have shown at least: Genital ulcers consistent with ABD or ocular lesions consistent with ABD or recurrent oral aphthosis plus at least one of other characteristic clinical features of ABD (skin manifestations, positive pathergy reaction, vascular manifestations). The sensitivity, specificity and accuracy for both ISGBD criteria and ITR-ICBD were determined.[8] Sensitivity, specificity, accuracy and receiver operator characteristic (ROC) curve were calculated for ISGBD and ITR-ICBD criteria. ITR-ICBD criteria exhibited higher sensitivity, specificity and accuracy than ISGBD criteria. ITR-ICBD also had a better level of optimization. Regarding to the ABD diagnosis according to the new criteria of the ITR-ICBD 29 patients (11 male, 18 female) satisfied them. So detailed data on the presence or absence of the clinical features of ABD in each patient were prospectively recorded and then analyzed. The statistical analysis required Pearson Chi-square test, Fisher exact test when more appropriate and Kruskal-Wallis test. Data were assessed using STATA SE 9 software (Stata Corporation, Tx, USA).

Results

Validation study

Senisivity, specificy and accuracy were calculated both for ISGBD and ITR-ICBD criteria. Regarding the ISGBD criteria, the sensivity was 75%, the specificy 100% and the diagnostic accuracy 83%. For optimum criteria set the gap between sensitivity and specificity: Should be as small as possible.[9] The optimization of ISGBD criteria was 25%.

The sensitivity for the ITR-ICBD was 100%, the specificy 92.8%, the accuracy 97.6% and the optimization was 7.2%. For both criteria ROC curves were calculated [Figure 1]: The area under the curve of the ISGBD criteria was 0.875, the area under the curve of ITR-ICBD criteria was 0.964.

Figure 1

Receiver operator characteristic

Analysis of the clinical manifestations

The male to female ratio of our 29 patients was 0.61. The mean age is 41.34 years [standard error (SE) 2.54, standard deviation (SD) 13.70 and 95% confidence interval (CI) 36.16-46.56]. The female patients present a statistically significant higher age (P = 0.037) than the male patients: The females’ mean age is 44.89 years (SE: 3.25, SD: 13.81, 95% CI: 38.02-51.76); males’ mean age is 35.55 (SE: 3.59, SD: 11.90, 95% CI: 27.55-43.54).

Oral aphthosis was seen in 28 patients (97%), with no significant difference between males and females (P = 1.000). Genital ulcers were reported in 24 patients (83%), with no significant difference in prevalence between males and females (P = 0.339). Eight patients (28%) presented ocular manifestations such as anterior or posterior uveitis, retinal vasculitis and conjunctivitis; regarding the prevalence in eye involvement there was no difference between sexes (P = 0.694). The presence of skin manifestations was described in 17 subjects (66%, in details: Erythema nodosum 24%, pseudofolliculitis 31%, papulopustolosis 10%) with no significant difference between male and female patients (P = 0.717). The positive pathergy reaction patients were 11, with no significant sex differentiation (P = 1.000). Sixteen patients (55%) in this survey presented joint involvement such as arthritis and arthralgia, in this case there was a significant higher prevalence of articular manifestation in female patients (P = 0.003). With the purpose of better investigating the characteristics of skin and joint manifestations, a relationship analysis between the two kinds of symptoms was made. In this work, a strong association between cutaneous and articular features was reported (P = 0.047), this kind of association was even stronger in female patients (P = 0.023).

Concerning gastrointestinal features, 12 patients (45%) reported chronic gastritis, cholitis, Crohn-like symptoms and oesophageal ulcers, with no significant difference in prevalence between males and females (P = 0.717). Headache was the only symptom of central nervous system involvement reported, and it was found in 10 patients (34%), there was no difference between male and female subjects. Peripheral nervous system involvement such as neuralgia and paresthesia was found in 8 patients (28%), with no significant different prevalence between sexes (P = 1.000) [Table 1].

Table 1

Analysis of the clinical manifestations

Finally, an analysis of the summation of the whole symptoms per patient was made. The mean number of symptoms per patient was 4.31 (SE: 0.31, SD: 1.67, 95% CI: 3.67-4.95), the mean number of symptoms in the male patients was 4.36 (SE: 0.56, SD: 1.86, 95% CI: 3.11-5.61) and in female patients 4.28 (SE: 0.38, SD: 1.60, 95% CI: 3.48-5.07). Using Kruskal-Wallis test, there was no significant difference between males and females (P = 0.87).

Discussion

The mystery of ABD etiology raised a heated debate about its diagnosis and classification over years. Several sets of diagnostic criteria have been proposed. This study was planned to compare the efficiency of the ISGBD criteria, which were released in 1990, and widely used in the past with the new criteria of the ITR-ICBD, released in 2006 in an Italian cohort of ABD patients.

ITR-ICBD criteria exhibited higher sensitivity, specificity and accuracy than ISGBD criteria. ITR-ICBD also had a better level of optimization. Similar results have been reported by Davatchi et al. in a recent validation study in Iran.[9]

The ROC can compare the precision of accuracy of the criteria.[9] In this work, the area under the curve of ITR-ICBD criteria is larger than the area of ISGBD, this means that ITR-ICBD set of criteria has a better efficiency.

Oral aphthosis represented the main clinical feature in our ABD patients, like in most previous studies.[10] According to the ITR-ICBD criteria oral aphthosis is no more a mandatory diagnostic clinical manifestation of ABD, so that in the future, the diagnosis of ABD without oral aphthous ulcers can be established.

What is new?

ITR-ICBD criteria exhibited higher sensitivity, specificity and accuracy than ISGBD criteria. ITR-ICBD also had a better level of optimization. According to the ITR-ICBD criteria oral aphthosis is no more a mandatory diagnostic clinical manifestation of ABD, so that in the future, the diagnosis of ABD without oral aphthous ulcers can be established.