Neuromyelitis Optica Spectrum Disorders in Africa

Discussion

We have reviewed the epidemiology, clinical presentation, additional investigations, treatment options, and outcomes of patients with NMOSD on the African continent. We found only limited data available in the literature, with only 10 of the 54 countries reporting on patient cohorts, and most data were hospital based. NMOSD occurs in all regions of the continent, suggesting that it is a pan-African occurring disorder. Indeed, a number of case reports, including reports from countries not represented in our review, have been published (Burkina Faso, Ghana, South Africa, Uganda, Nigeria, and Togo).^{46,54,-,60,e1-e6} Prospective data are scarce, and most studies were retrospective in nature, with inclusion criteria that varied over time in accordance with evolving diagnostic criteria of NMO and NMOSD.¹¹

Two recent reviews have emphasized the high prevalence of NMOSD in populations of African ancestry that may be as high as 10/100,000 in black people.^15,e7,e8 All studies cited in these reviews have been performed off the African Continent in French Martinique, the United States, the United Kingdom, Cuba, and New Zealand. The exact incidence and prevalence of NMOSD on the African continent remain unknown to date. The prevalence estimates that resulted from the analysis of the collected data from the published literature likely represent a massive underestimation of the real prevalence. Indeed, the estimated prevalence data from South Africa are lower than those reported in the query data from Multiple Sclerosis International Federation.¹⁸ This underscores the need for prospective nationwide or even continent-wide data collection initiatives. Limited human and financial resources, lack of systematic data collection on neurologic disorders, and numerous barriers to health care services may contribute to the lack of knowledge on epidemiologic data on NMOSD in Africa.^e9

NMOSD predominantly affects women and young adults in Africa, similar to other parts of the world. However, the female-to-male ratio of 3.5:1 is lower than reported in other review articles, which may be as high as 9:1 in AQP4-IgG–positive NMOSD.^e7 The age at diagnosis in our review was younger than the average of 40 years stated in other reviews.^15,e7,e8 In comparison to a study that evaluated racial differences, Afro-Americans and Afro-Europeans had a younger age of onset than patients of European descent,^e10 which may mean that the African ancestry may predispose to a younger age at onset. Sex, however, did not differ between races in this study, and between 84% and 90% were females.^e10 Especially in African countries, the early age of onset and more severe disease course may have a high impact, as comprehensive rehabilitation services are inequitable and inaccessible in Africa.^e11 Older patients may not come to the attention of health care services, and the average life expectancy of 62 years for men (range 54–72 years) and 65 years for women (range 59–75 years) may be another explanation for the young age at onset in our review.^e12 Women in Africa may be less likely to have access and attend to the hospitals^e13 in comparison to men, which may explain the higher proportion of men in our study. However, this hypothesis needs to be validated by community-based studies.

There seems a delay in the diagnosis as most of the studies identified patients after they had relapsed, which may translate into worse outcomes. Others reported that the median time to first relapse was about 5 months,^e14 which would mean that most of African patients are diagnosed only after a first relapse, as demonstrated in 8/19 of the studies.^{36,-,39,47,49,-,51} This calls for early and accurate diagnosis with adequate relapse treatment and institution of a maintenance therapy to prevent subsequent relapses and disability. Indeed, it is well known that more relapses before and after treatment are independent predictors of severe motor disability.^e10 Outcomes were only rarely reported with variable measures, including general terms, modified Rankin Scale score, or EDSS score. The data collected in our review were insufficient to describe outcomes in patients reliably.

The NMO:MS ratio increases when moving from the northern parts of Africa toward the south. This is not unexpected as it is known that the prevalence of MS increases with increasing latitude and lower levels of vitamin D.^e15,e16

The co-occurrence of HIV infection with NMOSD has been rarely described. However, the disease course does not differ from HIV-negative patients with NMOSD.^e17 Causality has not been proven yet, and potential pathophysiologic mechanisms need further investigation. Conflicting evidence exists also for TB and NMOSD.^e18-e20 A strong association of NMOSD with other autoimmune diseases, especially SLE and Sjogren syndrome, is well known and has also been found in some of the studies reviewed here.^e21-e23

NMOSD can present with several clinical syndromes, which include TM, ON, and area postrema syndrome.^6,e24-e26 In our review, TM and ON were the most frequent presenting symptoms, whereas a combined opticospinal presentation and brainstem symptoms were less frequent. Whether other presentations occur less or are merely underreported in the African population is unknown and an area open to further research.

Most studies in our review have been published recently and used established diagnostic criteria with additional investigations including at least an MRI and serology.⁶ There was a high frequency of LETM on spinal cord MRI and atypical MS-like brain lesions in most of these studies.^1,e27-e29 AQP4-IgG seropositivity was around 55%, which is slightly lower than has been reported in other studies,^e30-e32 probably due to the various inclusion criteria and assays used.^e33,e34 VEPs in NMO are useful in longitudinal follow-up of patients with NMO.^e35,e36 The frequency of an abnormal VEP was high in our review, which may be due to late presentation with residual deficits from previous relapses.

Relapse treatment in NMOSD consists of high-dose IVMP and/or plasmapheresis.^e37,e38 The latter is not readily available in most African countries and therefore hardly ever used according to our review in contrast to high-dose IVMP. Maintenance or preventive treatment recommendations for NMOSD include rituximab and mycophenolate as first-line agents, which are both not readily available in most African countries. Second-line agents include azathioprine, methotrexate, ciclosporin, mitoxantrone, and cyclophosphamide.^e37,e38 The most common preventive treatments used were steroids and azathioprine.

Both severe and mild cases may be missed, as many patients do not receive medical care in many parts of Africa. Both access to health care and universal health coverage are strikingly low compared to Europe and the United States and worst in sub-Saharan Africa.^e39,e40 Although the WHO recommends a ratio of 1 neurologist per 100,000 inhabitants, there is only a median number of 0.043 neurologist per 100,000 inhabitants in the WHO African region.^e41 Most of them (97%) practice in the capital city, and only 3% are based in rural areas.^e41 Some African countries have no neurologists at all.^e42 This problem starts at the level of (post)graduate education, with limited options for neurologic training and scarce neurologic faculty available for teaching.^e43 Moreover, access to MRI, which is crucial in making a correct diagnosis, is not available in many regions of Africa. Although there has been an increase in availability of MRI in the past few years on the African continent, in West sub-Saharan Africa, there is still a very low number of facilities (0.30–0.48/million inhabitants,^e44) compared with Europe or the United States (4.91–40 per million in 2019,^e45). No access to AQP4 or MOG antibody tests has been reported by approximately half (48%, 13/27) of African and Eastern Mediterranean countries.^e46 This study also highlighted that patients with NMO in LMICs cannot afford immunomodulatory treatment, despite local availability of the treatments.^e46 The same problem likely applies to the diagnostic tests and facilities: despite improving availability, they may remain unaffordable for many patients. International and political efforts are ongoing to improve health care coverage in LMICs, like on the African continent (e.g., UHC2030—accelerating progress toward Universal Health Coverage).

Because of limited resources, applicability of the 2015 IPND NMOSD criteria in the African setting remains challenging. Furthermore, possible relations between HIV, TB, and NMOSD need further research, and as these diseases are highly prevalent and incident on the African continent, this could lead to novel insights. Immunogenetic studies could aid in confirming or finding new genetic risk factors, such as HLA associations in the African population. In Afro-Caribbeans, an association with HLA class II allele DRB1 03 has been reported before.^e47 Efforts toward development of a global neuroscientific community such as TreND in Africa^e48 may help to build educational and scientific capacity.

We have to address several limitations of our systematic review: most included studies were retrospective, provided incomplete data, described different populations, and used different NMO criteria over time. We also acknowledge the differences in race, genetic background,^e49 cultures, and access to medical care on the vast geographical African continent. Research on risk factors and outcomes in LMICs may benefit from these variability and lead to optimization of the diagnostic and treatment pathway taking into account the limited resources. All these aspects may affect the quality of the data in a significant way. However, a major strength of our review is that it captures all available data on NMOSD, excluding single case reports, from the literature from the African continent.

Clearly, prospective data registries from the African continent would help to solve some issues raised in this article. Initiatives such as NMOBase and MOGBase, both as part of the MSBase international registry, may serve as easy-to-use data collection platforms. This registry is an international collaboration dedicated to sharing, documenting, and evaluating outcomes data in MS and other neuroimmunologic diseases, including NMO and MOG. The only 2 African countries currently actively participating in MSBase are Egypt and Tunisia. Efforts to increase awareness and knowledge on NMOSD in Africa have included dedicated conferences and support to perform AQP4 antibody assays, in which The Guthy-Jackson Charitable Foundation has played an important role. The CIRCLES initiative is collecting biospecimens and clinical data, currently only in North American sites, including the United States and Canada.^e50 An NMOSD African network could be envisaged in the future. The recently inaugurated African Stroke Organization may serve as an example to establish within Africa collaborations and to strengthen the African continent on its mission to reduce the burden of neurologic diseases.

There is a clear need for prospective data collection on NMOSD in the African continent to assess the real prevalence, incidence, and burden of this disease, which may turn out not to be ultra-rare in this continent at all (ultra-rare disease affects <1 per 50 ,000 persons). The lower proportion of women and younger age at onset need further investigation. Without increasing the number of neurologists per capita, educating the population, improving accessibility and affordability of diagnostics (AQP4 and MOG antibody tests and MRI) and acute and preventive treatments, cases will continue to be missed, patients will remain undertreated and their functional outcomes will not improve. We hope that our article serves as a call to action to direct research efforts and resources toward NMOSD in Africa, as the knowledge we gain in this way may benefit all people affected by NMOSD throughout the world.