Introduction

Systemic lupus erythematosus (SLE) is a multi-systemic autoimmune disorder of connective tissue that is characterized by the production of autoantibodies against the nuclear and cytoplasmic antigens, diverse manifestations ranging from mild to severe, remissions and flares, and varied prognosis. It is more prevalent among African Americans, American Indians, Alaskans, and Arabians. It predominantly affects women with a reported female-to-male ratio (F : M) of 7-15 : 1. The peak incidence of SLE in women is between the third and fifth decade [1, 2]. A systematic review of the worldwide epidemiology of SLE by Rees et al. [3] suggested that North America had the highest incidence and prevalence of 23.2/100,000 person-years and 241/100,000 people respectively while the lowest incidence was recorded in Africa at 0.3/100,000 person-years and lowest prevalence in Northern Australia (0 in a sample of 847 people). People of African descent had the highest incidence and prevalence while Caucasians had the lowest incidence and prevalence of the disease. The global burden of SLE cannot be overemphasized as the mortality from the disease is 2–3 times higher than the general population and it runs a more aggressive course among Africans and Hispanics. It is also one of the leading causes of death in young women and the causes of mortality in SLE include cardiovascular disease (CVD), infection, and renal disease [4–6]. Systemic lupus erythematosus was previously reported to be rare in sub-Saharan Africa compared to the African-American counterparts and this was attributed to immune disruption by parasitic infections [7]. However, there has been increasing reporting of this disease from this region over time, suggesting otherwise [8, 9]. Although most of these studies are hospital-based, a recent meta-analysis of these studies from sub-Saharan Africa showed an SLE prevalence of 1.7% among 28,575 participants. Rheumatological, dermatological, and haematological manifestations were the most common clinical presentations while patients showed a high seroprevalence for extractable nuclear antigens (ENA) such as anti-Smith (anti-Sm), anti-ribonucleoprotein, anti-Ro, and anti-La [8]. A recent multicentred, descriptive, retrospective, hospital-based study by Osaze et al. was done where records of patients with SLE seen within 4 years (2017–2020) in 20 rheumatology centres in Nigeria were studied to determine the pattern of SLE manifestations in Nigeria. A total of 913 patients were seen during the period but 17 patients were excluded due to incomplete data. The mean age of presentation was 34.47 years with an F : M ratio of 8.1 : 1. Synovitis was the most common symptom, accounting for 61% of cases, followed by acute, subacute, and chronic cutaneous lupus accounting for 51%, 19.9%, and 11.4% of cases respectively. Antinuclear antibody (ANA) was positive in 98% of cases, with titres ranging from 1 : 80 to 1 : 64,000 [10]. It is important to note that this study heralded the Lupus Registry in Nigeria (LURIN) established in 2021 where all data on lupus diagnosed in Nigeria are continuously collated and is one of the very few SLE registries in Africa. This retrospective study gave an overview of the pattern of SLE among patients of multiple ethnicities from different geopolitical locations which may show some variations from the clinical and laboratory pattern of the disease in this study location. Also interesting in this study was the fact that the two major tribes in Nigeria, the Igbos and the Yorubas, share the majority of the patients with SLE. This opens a research opportunity as to whether the pattern of SLE manifestations in these tribes differs from the general picture. Two other similar studies have been done in Nigeria. Adelowo et al. [11] reported the pattern of SLE in a private clinic in Lagos, a metropolitan city with people of multiple ethnicities. Emorinken et al. [12] studied the pattern of SLE manifestations in a rural tertiary centre in Irrua, south-south Nigeria. Irrua is dominated by the Esan tribe. However, this study was carried out in an ancient suburban city of Ile-Ife, southwest Nigeria, predominantly occupied by the Yorubas. Therefore, this study aimed to provide more specific details about the pattern of SLE in this fairly homogeneous Yoruba population.

Material and methods

Results

Discussion

This study analysed 65 patients who were diagnosed with SLE since the inception of a rheumatology clinic in Ile-Ife, Nigeria 7 years ago. Ile-Ife is an ancient city predominantly populated by the Yoruba race, as reflected in this study, where they constituted about 89.2% of the patients. Most of the patients were females in their second to fourth decades of life. The median (IQR) MEX-SLEDAI score of the patients was 5.0 (0–12.5), while the highest proportion of the patients (47.9%) presented with a MEX-SLEDAI score of > 5, all suggesting active disease on presentation. A raised ESR and positive ANA were common, found in 91.7% and 100% of the study population. Glucocorticosteroids and HCQ were the anchor medications for the treatment of the disease.

The frequency of SLE among rheumatic diseases seen in the clinic during the study period was 15.8%. This was higher than the frequency of SLE in other similar studies by Adelowo et al. [11] and Emorinken et al. [12], where it was 5.28% and 4.7% respectively. This may be due to the increasing awareness of the disease among healthcare professionals in recent times and their prompt referral to rheumatologists. This further highlights the fact that SLE is much more common than was earlier reported and negates Deborah Simmons’s prevalent gradient theory, which tried to explain the rarity in Africa based on the endemicity of malaria infections [9, 13].

Our study also showed the predominant female pattern of involvement (9.8 : 1), similar to other studies done among Africans, African-Americans, Europeans, and Arabians, ranging between 7.9 : 1 and 32.5 : 1 [10–12, 14–20]. The mean duration of symptoms at presentation is similar to studies done in south-south Nigeria and Saudi Arabia but lower than in the multi-centred hospital-based study in Nigeria [10, 12, 17]. However, a Polish study showed a much longer duration of symptoms (5 years), which may be due to a milder, nonspecific initial course of the disease in the study population [20]. The mean age at presentation in this study was 33.8 years, similar to other studies [10, 11, 17], and most of them were engaged in productive activities. However, Emorinken et al. from south-south Nigeria reported a lower mean age at presentation of 28 years [12]. The sociodemographic pattern of our study also showed that 84.6% of the patients earned less than 83 dollars per month, suggesting that SLE disproportionately affects the disadvantaged population [21].

Arthritis was the most common clinical presentation, found in 82.6% of the patients. This is similar to other studies where arthritis was also the most common clinical feature [10–12, 16, 19]. This finding is also supported by a systematic review and meta-analysis of SLE in sub-Saharan Africa which showed that rheumatological manifestations were the most common features, ranging from 5.1% to 99.9%. Similarly, arthritis was the most common clinical manifestation across Polish, African-American, and Brazilian studies [19, 20, 22]. Cutaneous manifestations are common features in SLE. In our study, acute cutaneous rash was the most common cutaneous feature (53.8%), and the highest proportion of these patients presented with malar rash (32.3%). This is similar to other studies done in Nigeria but a considerably higher frequency of malar rash was found in studies involving largely homogeneous African American and South African populations. The same pattern was also observed for discoid rash in these studies as well [10, 12, 18, 19]. Interestingly, about 15.4% had photosensitivity rash, which is much lower than in other Nigerian studies [10, 12] but similar to what Adelowo et al. reported [11]. Bortolini et al. reported a much higher frequency of malar and photosensitivity rash among a Brazilian study population (53.5% and 72.5% respectively), which may be due to the interplay between the genetic and environmental factors (e.g., sunlight) [22]. A Polish study also reported a higher frequency of photosensitivity rash of 37% [20]. Nonetheless, photosensitivity rash and cutaneous lupus have been linked to photosensitivity [23].

Alopecia was a common presentation in our study (50.8%), similar to other Nigerian studies [10–12] but higher than in a Polish study (29.6%) [20]. Oral ulcers were recorded in 52.3% of the patients, which is similar to other African, African-American, Brazilian, and Arabian studies but it was lower in Polish and Moroccan studies [16, 17, 19, 20, 22, 24]. Serositis is another common presentation in this series, seen in 47.7%, similar to what was reported by Emorinken et al. and Kamen et al. (40.4% and 45.1% respectively) [12, 19], but much lower rates were found in studies involving Sudanese and South African populations (16.1% and 18.1% respectively) [15, 18] as well as in Polish and Brazilian studies (22.5% and 20.3% respectively) [20, 22].

Neuropsychiatric manifestation was an important feature, reported in 33.8% of the patients, with a similar finding reported by Osaze et al. and also an Ivorian study by Gbané-Koné et al. (33.8% and 36.75% respectively) [10, 25], but much lower rates were recorded in a comparative study of SLE between Sudanese and Swedish populations (15.6% and 11% respectively) [26]. Our finding was also higher than in African-American and Brazilian studies (15.7% and 12.1% respectively) but lower than in a Polish study (59.2%) [20, 22, 27]. Moreover, memory loss was the most common neurologic symptom reported in our cohorts (13.8%), but a similar multi-centre study in Nigeria and a Polish study reported headache as their most common neurologic symptom [10, 20]. In a meta-analysis of studies on neuropsychiatric symptoms, headache (28.3%) and cognitive dysfunction (19.7%) were the most frequently reported neurologic symptoms [28].

Renal involvement (proteinuria) was detected in about 37.5% in our study, similar to other studies done in Nigeria, US, Brazil, and Abidjan [10, 12, 19, 22, 25], but it was higher in a Malaysian study (58.6%) [29]. This may be due to the multi-ethnicity of the Malaysian population. Moreover, a Polish study reported a lower frequency of proteinuria (22.5%) [20]. Renal biopsy was done in 13 patients and the predominant histopathological pattern was Class IV (46.15%), similar to other studies done in Nigeria, Brazil, Poland, and Egypt [10, 22, 30, 31]. However, a South African study reported Class V while a Sudanese study reported Class III as their predominant renal pathologic features [15, 32].

Anaemia, leukopenia, and lymphopenia were the common haematological features in our study, similar to a study from south-south Nigeria and a meta-analysis of studies on SLE from sub-Saharan Africa [8, 12] as well as other studies outside Africa [20, 22]. Erythrocyte sedimentation rate was the commonest laboratory feature, found in 91% of the patients, with a median value of 103 mm/h. This highlights the sensitivity of ESR to make the diagnosis, especially in low-resource settings where the cost of serological investigations is unaffordable to many patients and the clinical probability of SLE is high. Moreover, ESR has been suggested as a more reliable biomarker of disease activity than anti-dsDNA and complement fractions, and this may help in the diagnosis of an active disease [33]. In this study, it is noteworthy that the median ESR is high. However, it is not unexpected as it is a marker of active disease which most of our patients presented with, and also corroborates the average MEX-SLEDAI score of the study population indicating an active disease. However, anaemia, infection, and hypoalbuminaemia could also have contributed to this raised ESR. Many of our patients presented with anaemia, another marker of disease activity, and unsurprisingly, there was a statistically significant negative correlation between ESR and anaemia in our patients. Some of our patients presented with infections, especially those with severe lupus. Infections may be a feature of SLE or a complication of its treatment. Hypoalbuminaemia occurred in our patients with lupus nephritis and nephrotic range proteinuria. Although many of our patients earned below $83 per month, it does not necessarily mean they are malnourished. Our legal tender in Nigeria is the naira, which has a much lower value than the US dollar, and therefore implies a lower cost of living in Nigeria. Moreover, the majority of our patients are well educated with a culturally inclined formidable family and social support.

Antinuclear antibody was the most common feature (100%) in this study, which corroborates other studies where their detection rate was > 90% [10, 17, 19, 25]. In a meta-analysis of studies of SLE in sub-Saharan Africa, the prevalence of ANA was statistically determined to be 89.7% [8]. This underlies the crucial importance of ANA in making the diagnosis of SLE. In our study, about 50.94% had ANA titres of 1 : 5,120 and above, higher than in a multi-centred study in Nigeria (15.7%) and a study from south-south Nigeria (17.3%) [10, 12]. This suggests that southwestern Nigerian patients, especially those of the Yoruba tribe, may have higher ANA titres than patients from other tribes and geopolitical zones in the country. This may have a diagnostic implication as a retrospective study of 1,297 patients with ANA results including 148 patients with SLE showed that higher ANA titres are highly suggestive and specific for SLE [34]. This may obviate the need for SLE-specific antinuclear autoantibody tests, which are becoming increasingly unaffordable to our patients, thereby relieving the costs of managing SLE in poorly insured low-income countries. Another important finding in this study is the predominant speckled immunofluorescence pattern (86.96%), which corroborates other studies done in Nigeria [10, 12, 35]. This may be attributable to the presence of antibodies against anti-ENA such as anti-Sm, which are common in black patients [35, 36], giving a speckled immunofluorescence pattern. In our study, anti-Sm was detected in 50% of the patients, while similar or higher rates were found in studies from Nigeria, Ivory Coast, Morocco, and the US [12, 19, 24, 25] and studies from Sudan and Brazil reported a lower rate of 19.3% and 26.4%, respectively [15, 22]. A meta-analysis of sub-Saharan African studies on SLE by Essouma et al. [8] estimated a similar pooled prevalence of anti-Sm at 53.5%. Furthermore, Essouma et al. [8] also estimated a pooled anti-dsDNA prevalence at 54.6%, similar to what was detected in this study (50%). Serological analysis in our study was restricted to ANA, anti-dsDNA, and anti-Sm because the extractable nuclear antibody (ENA) panel is very expensive as patients pay out of their pockets for these tests. The clinical and laboratory profile of SLE in cross-sectional studies across ethnicities and races is shown in Table V.

A number of co-morbidities have been associated with SLE, one of which is dyslipidaemia. Dyslipidaemia is one of the determinants of CVD, a predictor of mortality in SLE. Premature CVD is 50 times more common in premenopausal women with SLE within the 35–44-year age group than in the general population [37]. The prevalence of dyslipidaemia in SLE in studies ranged from 30% to 60% and the pathogenetic mechanisms implicated include antibodies against lipoprotein lipase and cytokines regulating the balance between pro-atherogenic and anti-atherogenic lipoproteins [38]. In our study, dyslipidaemia was found in 76% of cohorts. Therefore, traditional risk factors for CVD should be screened in every patient with SLE.

Glucocorticosteroids and HCQ are the anchor medications in SLE. Glucocorticosteroids are widely available and cheap. However, since the COVID-19 pandemic, there has been an astronomic rise in the price of HCQ due to its increased demand for the treatment of COVID-19 infections despite studies against its efficacy [39]. Even after the pandemic, its price has remained high with limited availability. Although HCQ is an antimalarial, the World Health Organization recommends artemisinin-based combination therapy, which we use to treat malaria in Nigeria. Therefore, all our patients in this study were commenced on oral HCQ in our clinic after the diagnosis of SLE was made by a rheumatologist. The use of biologics in Nigeria is largely limited to private rheumatology clinic settings where patients’ care is funded by large multinational companies and agencies [40]. Therefore, cheaper conventional immunosuppressants such as mycophenolate mofetil, cyclophosphamide, and azathioprine are preferred for treating more severe lupus. However, the costs of these medications have escalated in recent times, making them increasingly unaffordable as maintenance regimens and GC-sparing agents. This has resulted in poor drug compliance and the seeking of cheaper alternative medicine since patients pay out of their pockets. Consequently, some of these patients present with flares and complications, sometimes leading to mortality. Although this study is not aimed at assessing the accessibility to treatment and factors affecting drug compliance, further studies on these areas will help health policymakers in developing countries to understand the socioeconomic burden of this disease and allocate resources to relieve the cost of managing SLE. This will improve their prognosis, quality of life, and productivity.

Limitations

This study has some limitations. First, it had a small sample size and not all the patients were properly investigated. Second, it was a retrospective study, which precluded us from determining their disease course, treatment outcome and prognosis. Third, this is a single-centre study and thus the generalisability of these findings may be limited. Therefore, a prospective multi-centred study would be more appropriate to determine the epidemiology of lupus. This can be achieved with the establishment of LURIN, where data of patients with SLE on initial and follow-up visits in Nigeria are being collated electronically for research purposes. Fourth, most of our patients could not afford to do extensive serological investigations, which could have been helpful to carry out proper serological profiling.

Conclusions

Patients with SLE in this study population are mainly from the Yoruba tribe, predominantly females in their second to fourth decades of life. Arthritis, mucocutaneous manifestations, raised ESR, and positive ANA were the most frequent clinical, laboratory, and serological features, respectively. Dyslipidaemia was a common comorbidity among the cohorts.