Rheumatoid arthritis (RA) is a systemic inflammatory disease characterized by symmetric destructive arthritis that is often associated with extra-articular and systemic manifestations. We must also take into account an important morbidity and socio-economic impact [1, 2] of this disease as well as an increased cardio-vascular mortality [3–6]. Great strides have been made on RA in recent years, including a better understanding of the genetic aspects, environmental factors and the disease's pathogenesis, at least in the Western world. These developments also concern the clinical approach, including methods used to assess the impact of RA as well as treatment principles, such as the early intensive therapy and treat-to-target approach, and the use of biologicals. In contrast, there are still few data on RA in many other parts of the world. Little is known about the epidemiology of RA in black Africa, its severity, the cross-cultural validity of the different classification criteria and measurement scores, the socio-economic aspects as well as the genetic and/or environmental factors leading to the disease. Therefore, just copying Western principles might not be correct.

In the Democratic Republic of Congo very few studies have been published on RA. In 2008, JJ Malemba and JM Mbuyi-Muamba reported, in a retrospective study, that RA was rare and mild in its clinical presentation at the University Hospital of Kinshasa (UHK) [7]. These observations were in agreement with those reported by Bwanahali et al. in 1995 [8]. But these studies did not use "quantifiable" parameters to assess the severity of RA, and the first mentioned study had a hiatus due to its retrospective nature. In addition, no information was given on the frequency of shared epitope (SE), smoking and the possible influence of a tropical environment with its specific background of infections. In a recent epidemiological study [9] also, a rather mild phenotype was suggested. The current study was initiated to describe prospectively the phenotype and HLA-DRB1 genotype of newly diagnosed RA patients attending the rheumatology unit of the UHK in order to assess the severity of the disease in this part of the world and to test the potential involvement of amino acids 70 to 74 in the third hypervariable region of HLA-DRB1 (SE).

Table 1 shows that the patients whose serologic and/or radiographic data were available were similar to those for whom these data were not obtained. All had symmetric arthritis, arthritis of hand joints and involvement of at least three joint areas. No statistically significant difference was observed between the average age of patients whose X-rays were available compared to patients without X-rays (P = 0.21) or between patients with serological analyzes and patients without (P = 0.74).

HLA-DRB1 typing was performed in a subset of patients (n = 37) and controls (n = 24). The most frequent HLA-DRB1 alleles in the control group were HLA-DRB1*15:03, 13:01/02, 03:01/02 and 11:01. There was a very low carrier frequency of DRB1*04:01 and DRB1*13:03 in our Congolese population (patient group and controls). The SE alleles detected were DRB1*04:01, *13:03, *01:01/02 and *10. One copy of the SE was found in 13 patients (35.1%) and 3 controls (12.5%). Two copies of the SE were found in 1 patient (2.7%) and 1 control (4.2%).

Our results indicate that in a three-year time period 8.5% of outpatients suffering from rheumatic diseases and attending the rheumatology unit of the UHK were diagnosed with RA. The relative RA frequency of 8.5% observed in this study is higher than that reported in a previous paper showing RA to affect 3.6% of the patients attending the rheumatology unit of UHK from 1988 to 2002 [7]. This is probably because of the retrospective nature of the latter study.

Patients consulted following a rather long delay after experiencing typical symptoms and all presented with a high value of DAS-28. This may be attributed to several factors: the poverty of the population with difficulties in accessing and paying for healthcare, the use of traditional medicine, and lack of information and education. It is, therefore, understandable why many patients consult only when symptoms become intolerable as a consequence of high disease activity. The high DAS-28-ESR scores in our patient population are probably partly related to the low hemoglobin level found in this study. These low hemoglobin levels may not necessarily be related to the severity of the RA since they are similar in patients and controls. They are probably influenced by the nutritional deficit and tropical infections which affect a part of the Congolese population. Therefore, it is probably unwise to compare DAS-28-ESR scores from Congolese patients in absolute terms with scores obtained in a Western population. We cannot exclude that DAS-28-CRP would have allowed a more accurate interpretation of the disease activity, but unfortunately, for practical reasons, CRP could not be determined.

Mean HAQ index was rather high, probably related to the disease activity, but rheumatoid nodules and typical deformities were found in less than 20% of patients. Erosions and joint space narrowing were found in somewhat more than 50% of patients and were of low to moderate severity, taking into account the high disease activity and important delay before treatment.

The fact that more than 50% of Congolese RA patients are RF and anti-CCP negative may contribute to the mildness of the disease phenotype since the level of anti-CCP and RF titers is linked with the disease severity [15–17]. A mild phenotype was also reported in other African populations. Adewolo et al. [18] observed that in 200 patients followed for RA in Nigeria only 38.5% were RF positive and 29% had joint erosions on radiographs of hands, while the average duration of symptoms was 63 months before the first consultation. Ndongo S et al. [19] in Senegal reported that joint erosions were present in one-third of patients and that extra-articular manifestations were very rare, while the average time from the onset of symptoms to initiation of disease modifying antirheumatic drugs was 54 months. Anaya JM et al. [20] in Colombia have described that RA was less severe in blacks than Mestizos in terms of radiographic damage. Recently, Singwe-Ngandeu et al. in Cameroon have also reported a mild presentation of RA among their patients [21].

RF and anti-CCP were observed in only a few controls (2.98%) in our study, whereas in this population, higher numbers were expected due to the high frequency of parasitic and viral infections. Using European cut-off values for RF and anti-CCP antibodies, as done in this study, may be a problem in a sub-Saharan African population and for future studies it would be advisable to determine local cut-off values.

In our study there does not seem to be an association between the presence of the shared epitope and RF or anti-CCP antibodies (Table 4). Classic HLA-DRB1 S2 and S3P epitopes were very rare. Singwe-Ngandeu M et al. [21] observed also a discrepancy in a group of RA patients in Cameroon of whom more than 75% were RF positive and anti-CCP positive, but only 30% were carriers of the S2 and S3P HLA-DRB1 alleles. The four most frequent HLA-DRB1 alleles in our control group are those typically reported in sub-Saharan Africa [22].

Interestingly, there was a low prevalence of HLA-DRB1*04:01 in this study population (patients and controls) compared to other populations [23, 24]. This observation is in agreement with that of Mbayo et al. [25, 26], who reported that HLA-DR4 was less frequent in Congolese blood donors. This may be important since some authors have reported that HLA-DRB1*04:01 (S2) is associated with the most severe phenotype of RA [23, 27]. Other HLA-DRB1 alleles, on the contrary, were more prevalent. In agreement with the study of Mbayo [25, 26], we observed a high frequency of HLA-DRB1*15: 03 (X), DRB1*13:01/02 (S1) and DRB1*11:01 (S3D). The lower risk of these HLA-DRB1 alleles is confirmed in different studies [28–34].

The present study showed that 35.1% of the study RA population had S2 and S3P HLA-DRB1 alleles. However, compared to the control group, the association of SE with RA was statistically not significant. Singwe-Ngandeu et al. in Cameroon observed a lower frequency (30%) of SE-positive cases among RA patients [21]. A low frequency of HLA-DRB1 SE in blacks was also observed by LB Hughes et al. [35], who reported a frequency of 25.2% in African American patients. They found a high degree of European ancestry among African Americans with SE alleles, which suggests that a genetic risk factor for RA was introduced into the African American population through admixture. This is not in agreement with the study of PWA Meyer et al. [36] who reported a frequency of 59% of high-risk SE alleles in a predominant black South African female population with RA, and this frequency is similar to that reported by Barnetche et al. in a Caucasian population [37]. When considering our results, we cannot exclude that other genetic factors may be involved in Congolese RA patients. This needs further evaluation.

The effect of the environment on the incidence and phenotype of RA must also be considered. Smoking, the main environmental risk factor implicated in RA [38–40], is present only in a small proportion of our patients, and at a low level. Moreover, for cultural reasons women very rarely smoke in DR Congo. Beyond the effect of smoking, other environmental factors could be implicated. Previous studies conducted in sub-Saharan Africa have suggested a protective role of malaria against autoimmune diseases. This applies to the study of Greenwood [41] in Nigeria who observed that mice infected with Plasmodium falciparum showed fewer autoimmune disorders than mice free of P. falciparum. Similarly, Moolenburgh [42] in Lesotho observed that RA was more frequent and severe in areas of high altitude than on the warmer plains with more malaria. A similar survey in DR Congo could help to clarify this concept, and it remains to be seen whether the tropical environment with its many parasites could interfere with the phenotype of RA or be implicated in the pathogenesis of the disease.

Disease classification, disease activity and functionality of RA were evaluated in this paper using standardized methods that are classical for evaluation of patients seen in the Western world. Unfortunately, mainly because of the cost issue, detailed serology data and genetic typing were not feasible in all patients. Nevertheless, this study will provide a benchmark for measuring benefits of RA care in central Africa and will allow comparing disease patterns in rural compared to this urban population in the future. These preliminary results also make clear that perhaps specific validation is needed for classification criteria outside the Western world. For instance, the very recent new classification criteria for RA [43] are of no use in a specific situation where CRP and CCP measurement is not performed in daily practice and where intensive treatment and a treat-to-target approach are currently unrealistic goals. We also found aspects of timing (duration of morning stiffness, six weeks symptom duration) sometimes problematic to evaluate, as time perceptions are different compared to the Western world. Also proper cross-cultural validation of instruments like the HAQ needs to be carried out, not only because of translation issues, but also because of content problems; for instance, with concepts like 'shampooing hair' which is not often done by Congolese women.

In conclusion, the current preliminary data on the phenotype and genotype of RA are the first from this part of the world and according to our findings, we put forward the hypothesis that the particular genetic profile of RA in Western Europe is not present in sub-Saharan Africans and this could explain the decreased severity of the disease. A study in a larger sample population is needed to better assess the impact of HLA-DRB1alleles on the prevalence and phenotype of RA in Congolese, and the correlation between the shared epitope and the presence and titers of anti-CCP and RF.