Behçet’s syndrome (BS) shows a peculiar distribution, with a much higher prevalence in countries along the ancient Silk Road compared with rest of the world. BS also seems to follow a more severe course in ethnic groups with higher prevalence. Diagnosis depends on clinical findings. Criteria sets may not help in patients with less frequent types of involvement. Management strategies should be modified according to the age and sex of the patient and the organs involved. Being a serious health problem in endemic areas, BS also attracts global attention as a model to study inflammatory diseases of unknown cause.
Key points
- •
Behçet’s syndrome (BS) is a condition of unknown etiology, most prevalent in countries along the Old Silk Road.
- •
Epidemiologic data points to the role of both genetic and environmental factors in the etiology of BS.
- •
Pulmonary artery involvement in BS may consist of pulmonary artery aneurysms and/or pulmonary artery thrombosis, which may be accompanied by pulmonary hypertension during the acute stages.
- •
Pulmonary parenchymal lesions, which may accompany pulmonary vascular lesions, may be mistaken for pulmonary embolism or infections.
- •
There are distinct clusters in disease expression, such as the acne/arthritis/enthesopathy and the dural sinus thrombosis/pulmonary artery aneurysm/deep venous thrombosis/superficial venous thrombosis clusters, which possibly have distinct pathogenetic mechanisms.
- •
BS has a complex genetic background, and the most significant association is with HLA B51.
- •
Diagnosis of BS depends on clinical findings. There are no specific laboratory, radiologic, or histologic findings that help in diagnosing BS.
- •
Subsequent mucocutaneous and joint involvement may be bothersome and impair the quality of life, but do not cause organ-threatening damage. Their management depends on the type, frequency, and severity of the symptoms.
- •
Management of serious organ involvement such as ocular, vascular, gastrointestinal, and neurologic disease is with corticosteroids and immunosuppressives, which should be started early to prevent irreversible damage.
- •
More aggressive treatment of serious organ involvement with immunosuppressives and biologics has improved the outcome in BS patients.
Epidemiology
The prevalence of Behçet’s syndrome (BS) is highest in Turkey (1 in 250), where it was first described. The comparative prevalences in other regions are shown in Table 1 . Although the prevalence among Turkish immigrant workers and their offspring is lower than what has been reported from their native country, it is still higher than has been reported among ethnic Germans, pointing to combined genetic and environmental factors in the pathogenesis. A comparative study of the frequency of BS among the ethnic Armenian population living in Istanbul, Turkey has shown that the frequency of familial Mediterranean fever (FMF), a condition with a much better defined genetic component, was more common among Armenians than in the ethnic Turks, whereas the reverse was true for BS.
| Country | Authors, Year | Prevalence (per 100,000 Population) | HLA B51 Positivity (%) |
|---|---|---|---|
| Japan | Kurosawa et al, 2004 | 11.9 | |
| China | Zhang et al, 2012 | 14 | 17 |
| Iran | Davatchi et al, 2008 | 80 | NA |
| Iraq | Al-Rawi et al, 2003 | 17 | NA |
| Israel—Overall | Krause et al, 2007 | 15.2 | 81 |
| Druzes | 146.4 | NA | |
| Arabs | 26.2 | NA | |
| Jews | 8.6 | NA | |
| Saudi Arabia | Al-dalaan et al, 1997 | 20 | NA |
| Egypt | Assaad Khalil et al, 1997 | 7.6 | 58 |
| Portugal | Crespo et al, 1993 | 1.5 | 75 |
| Spain | Gonzalez-Gay et al, 2000 | 6.4 | NA |
| Italy | Salvarani et al, 2007 | 3.8 | 75 |
| France—Overall | Mahr et al, 2008 | 7.1 | 33 |
| Europeans | 2.4 | NA | |
| North African | 34.6 | NA | |
| Asian | 17.5 | NA | |
| Sub-Saharan African | 5.1 | NA | |
| Noncontinental French | 6.2 | NA | |
| Germany | Papoutsis et al, 2006 | ||
| German | 1.5 | NA | |
| Non-German | 26.6 | NA | |
| Turks | 77.4 | NA | |
| UK | Chamberlain, 1977 | 0.64 | 18 |
| Scotland | Jankowski et al, 1992 | 0.3 | 13 |
| Sweden | Ek et al, 1993 | 3.5 | 80 |
| USA | Calamia et al, 2009 | 5.2 | 0 |
In the United States, whereas an estimated rate of 0.33 in 10 5 was reported in Olmsted County, Rochester in 1978, a higher rate (5.2 in 10 5 ) was calculated in the same region between 1996 and 2005. The Behçet’s Syndrome Evaluation, Treatment and Research Center at the NYU Hospital for Joint Diseases in New York has been collecting data on BS patients since 2005. Several interesting observations about disease manifestations have been noted. In this dedicated center, 197 consecutive patients were divided into 2 groups, Group 1 comprising patients with a northern European background and Group 2 consisting of patients ethnically from areas where Behçet’s prevalence is high (Turkey, Greece, Israel, Middle East, and Far East). These groups were compared regarding their demography and disease manifestations. There were significantly more females (78% vs 54%) in Group 1, made up of predominantly patients with skin and mucosal disease. About one-third of patients had eye disease in both groups; interestingly there were no patients who were blind in the whole cohort. Vascular involvement was seen in 3 patients in Group 2 and in none in Group 1. These data suggest that even though most manifestations of BS were similar in frequency between the 2 groups, some manifestations might be more severe in patients with backgrounds from endemic areas, such as the Middle East. Similarly, milder disease has also been reported among Italian patients from northern Italy.
Epidemiology
The prevalence of Behçet’s syndrome (BS) is highest in Turkey (1 in 250), where it was first described. The comparative prevalences in other regions are shown in Table 1 . Although the prevalence among Turkish immigrant workers and their offspring is lower than what has been reported from their native country, it is still higher than has been reported among ethnic Germans, pointing to combined genetic and environmental factors in the pathogenesis. A comparative study of the frequency of BS among the ethnic Armenian population living in Istanbul, Turkey has shown that the frequency of familial Mediterranean fever (FMF), a condition with a much better defined genetic component, was more common among Armenians than in the ethnic Turks, whereas the reverse was true for BS.
| Country | Authors, Year | Prevalence (per 100,000 Population) | HLA B51 Positivity (%) |
|---|---|---|---|
| Japan | Kurosawa et al, 2004 | 11.9 | |
| China | Zhang et al, 2012 | 14 | 17 |
| Iran | Davatchi et al, 2008 | 80 | NA |
| Iraq | Al-Rawi et al, 2003 | 17 | NA |
| Israel—Overall | Krause et al, 2007 | 15.2 | 81 |
| Druzes | 146.4 | NA | |
| Arabs | 26.2 | NA | |
| Jews | 8.6 | NA | |
| Saudi Arabia | Al-dalaan et al, 1997 | 20 | NA |
| Egypt | Assaad Khalil et al, 1997 | 7.6 | 58 |
| Portugal | Crespo et al, 1993 | 1.5 | 75 |
| Spain | Gonzalez-Gay et al, 2000 | 6.4 | NA |
| Italy | Salvarani et al, 2007 | 3.8 | 75 |
| France—Overall | Mahr et al, 2008 | 7.1 | 33 |
| Europeans | 2.4 | NA | |
| North African | 34.6 | NA | |
| Asian | 17.5 | NA | |
| Sub-Saharan African | 5.1 | NA | |
| Noncontinental French | 6.2 | NA | |
| Germany | Papoutsis et al, 2006 | ||
| German | 1.5 | NA | |
| Non-German | 26.6 | NA | |
| Turks | 77.4 | NA | |
| UK | Chamberlain, 1977 | 0.64 | 18 |
| Scotland | Jankowski et al, 1992 | 0.3 | 13 |
| Sweden | Ek et al, 1993 | 3.5 | 80 |
| USA | Calamia et al, 2009 | 5.2 | 0 |
In the United States, whereas an estimated rate of 0.33 in 10 5 was reported in Olmsted County, Rochester in 1978, a higher rate (5.2 in 10 5 ) was calculated in the same region between 1996 and 2005. The Behçet’s Syndrome Evaluation, Treatment and Research Center at the NYU Hospital for Joint Diseases in New York has been collecting data on BS patients since 2005. Several interesting observations about disease manifestations have been noted. In this dedicated center, 197 consecutive patients were divided into 2 groups, Group 1 comprising patients with a northern European background and Group 2 consisting of patients ethnically from areas where Behçet’s prevalence is high (Turkey, Greece, Israel, Middle East, and Far East). These groups were compared regarding their demography and disease manifestations. There were significantly more females (78% vs 54%) in Group 1, made up of predominantly patients with skin and mucosal disease. About one-third of patients had eye disease in both groups; interestingly there were no patients who were blind in the whole cohort. Vascular involvement was seen in 3 patients in Group 2 and in none in Group 1. These data suggest that even though most manifestations of BS were similar in frequency between the 2 groups, some manifestations might be more severe in patients with backgrounds from endemic areas, such as the Middle East. Similarly, milder disease has also been reported among Italian patients from northern Italy.
Clinical presentation
BS is seen with roughly equal frequency in males and females. Males have a distinctly more severe course, which particularly manifests itself in the morbidity associated with the disease as well as pulmonary vascular disease. BS is relatively rare in childhood.
Comprehensive accounts of the clinical symptoms and signs of BS are available. This article emphasizes the more recent developments in the understanding of the phenotype of BS, and discusses why BS might be a syndrome that is true to type, with perhaps more than one pathogenetic pathway responsible for its clinical expression.
Vascular Involvement
The most mortal manifestation of BS is the pulmonary artery aneurysm, which carries a mortality of around 25% to 30%. It is now better appreciated that the pulmonary arterial aneurysm (PAA) is a manifestation of a more general pulmonary vascular disease. Some BS patients present mainly with the well-recognized PAA; in some the pulmonary artery thrombi (PAT) can accompany the PAA, and there is a third group who present with PAT only. The main clinical feature of the third group is that they present with a less copious hemoptysis in comparison with the patients with PAA.
About 80% of patients with pulmonary vascular disease also have peripheral vascular disease, mainly in the form of thrombophlebitis. It has long been debated whether some of the lung lesions seen in BS patients are indeed pulmonary emboli. The authors’ group has opposed this view mainly because: (1) in a condition with at least 30% to 40% venous thrombosis, pulmonary embolism at post mortem has not been observed ; (2) the venous thrombosis in BS, unlike common-variety venous thrombosis, involves long segments in the veins causing sticky thrombi; (3) although formal prospective studies testing the utility of anticoagulation are lacking, in 2 retrospective surveys it was found not to be useful. The authors were able to observe the long-term (2–6 years) outcome of ventilation/perfusion mismatch in 6 patients whose initial lung scans had been interpreted as having pulmonary emboli. This outcome is distinct from what one sees in pulmonary thromboembolism, where the defects seen on lung scans in true pulmonary embolism resolve mostly in a few months.
Many patients with pulmonary vascular disease in BS have parenchymal lesions in the lungs, which can easily arise as secondary infections in a patient using glucocorticoids and/or immunosuppressives, and some patients with pulmonary vascular disease also have mild to moderate pulmonary hypertension, as the authors have shown in a controlled study. It may also be possible that among patients in the acute stage with PAA and/or PAT, the degree of this pulmonary hypertension will increase.
Symptom Clusters: Acne, Arthritis, and Enthesopathy
Some of the clinical features of BS tend to go together. The most studied of these clusters has been the acne-arthritis cluster, which is interesting in the context of the well-known association of arthritis with acne ( Fig. 1 ). However, it was noted that acne lesions were not sterile, as had been thought, and, rather unexpectedly, Staphylococcus aureus grew in the associated pustules, an uncommon occurrence in acne vulgaris. This recurring theme of acne and arthritis led the authors to reconsider their initial stand on excluding BS from the seronegative spondarthritides, which some had thought was within the spectrum of BS. At this point the hypothesis that enthesitis, a common denominator in seronegative spondarthritides, was more frequent in BS was tested. Only those patients with arthritis and associated acne were found to have enthesitis more often. However, on formally checking whether this same group also had an increased frequency of sacroiliitis or were more frequent HLA B27, it was discovered that this was not the case.
More recently the authors have also shown that there was significant familial clustering of the acne/arthritis cluster, suggesting that a common genetic pathway is involved in its clinical expression.
Symptom Clusters: Deep Venous Thrombosis, Superficial Venous Thrombosis, and Dural Sinus Thrombosis
In BS, superficial venous thrombosis (SVT) and deep venous thrombosis (DVT) also go together, as well as dural sinus thrombosis (DST) and peripheral DVT. The clustering of DST, DVT, and SVT is conceptually easy to understand. What is perhaps more intriguing is the association between PAA and DVT. It has repeatedly been shown that about 80% of patients with PAA also have peripheral DVT, an association that has for many years led many to incorrectly consider pulmonary emboli as an integral part of pulmonary vascular disease in BS. On the other hand, this association is perhaps not so difficult to surmise if one remembers that the pulmonary arteries, in their structure and content, are very similar to veins. As an extension of this PAA and DVT association, it is also worth noting that one-third of patients with DST have PAA, whereas the frequency of PAA in the general population of BS patients is no more than 2% to 3%.
The authors consider that within the phenotype of BS, acne and arthritis is surely one clinical cluster and, most probably, DST/PAA/DVT/SVT is another. The presence of such clusters obviously suggests that there might be more than one pathogenetic pathway in what today is known as BS.
What is identified as BS is probably the result of several different pathogenetic mechanisms, also finding support from the fact that, the same drug curiously can have different effects on different manifestations of this syndrome. As also discussed in the section on management, etanercept is effective for skin-mucosal manifestations but curiously has no effect on the pathergy reaction or the dermal response to sodium monourate crystals. Gevokizumab, an interleukin (IL)-1β blocker, was reported as effective for eye disease in a small series of patients, whereas it seemed not very effective for oral ulcers. In controlled studies colchicine, widely considered to be beneficial in BS, turned out to be mainly effective for arthritis, erythema nodosum, and genital ulceration, only among females. On the other hand, if one looks at the effect of a single agent, such as glucocorticoids (GCS), in other inflammatory diseases such as rheumatoid arthritis (RA) or systemic lupus erythematosus (SLE), one finds that they are effective in managing active disease in any organ in either condition. In this context it is also important that in the only controlled study of GCS for BS, the authors’ group showed that low-dose methylprednisolone was ineffective in controlling skin and mucosal manifestations. In brief, there is good reason to suspect that pathways of inflammation are most probably not uniform and somewhat different to what is seen in other, more common inflammatory diseases such as like RA and SLE.
Genetics and disease mechanisms
BS is genetically a complex disorder. A phenotype is associated with more than one gene locus. As the authors previously pointed out for BS, it is, however, possible that at least a portion of this complexity might stem from attempts to ascribe multiple genetic associations to a single phenotype, whereas the phenotype at hand might not be homogeneous to begin with. Returning to the discussion about BS, it is well accepted that a simple Mandelian inheritance does not exist among adult patients with BS. On the other hand, an autosomal recessive mode of inheritance has been suggested in childhood BS. Although similar clinical features have been described in the pediatric and the adult forms, the frequency of genital ulceration and eye disease are less frequent among the pediatric cases.
The most significant association of BS is with HLA B51. However, this association explains around 20% to 30% of the heritability in a condition whereby only 1 in 10 patients identify a diseased relative. In a small (but the only) twin study, the concordance frequency among 4 monozygotic twins was 2 patients, and this did not change over the course of 8 years. Recently IL-10 gene mutations have also been described in BS. The contention that the mutation that causes a defect in the production of IL-10, an anti-inflammatory cytokine, is a risk factor in an inflammatory disease surely makes good sense. There is evidence that IL-10 levels do decrease with age, whereas in BS disease activity distinctly lessens with age.
Diagnosis
There are no specific laboratory, radiologic, or histologic findings that help in diagnosing BS. The diagnosis, for all purposes, is entirely clinical. There have been various and heatedly debated sets of diagnostic and classification criteria, the most widely used of which are still the International Society for Behçet’s Disease (ISBD) criteria of 1990. While not aiming to describe the individual clinical manifestations, this article does discuss several points concerning the ISBD criteria that are worth reiterating for the purposes of diagnosis.
- 1.
The ISBD criteria do not include gut disease, central nervous system disease, and pulmonary vascular disease, merely because these organ involvements had not been either specific or sensitive enough to have discriminatory value in separating BS from other diseases in the control groups.
- 2.
In the ISBD criteria oral ulceration is assumed to be always present, so in a patient with no oral ulceration (<1%–2% of BS patients) these criteria are, strictly speaking, inapplicable.
- 3.
Oral ulceration is the most sensitive lesion and genital ulceration is the most specific ( Fig. 2 ). The latter therefore, like in all relatively rare conditions, is the most clinically useful lesion in diagnosing BS, according to the ISBD scheme.
