High Early Mortality and Contributing Factors in Seropositive Neuromyelitis Optica Spectrum Disorder at a Tertiary Hospital in Zambia

Abstract

Background/Objectives:

Mortality of aquaporin-4 positive neuromyelitis optica spectrum disorder (NMOSD) in Africa is understudied. This study sought to report preliminary findings of one-year mortality among a cohort of seropositive NMOSD patients in Zambia and characterize key contributing factors.

Methods:

This cross-sectional study included seropositive patients with NMOSD (via ELISA) at a tertiary care hospital in Zambia between 2019 to 2025. Demographic, clinical, treatment characteristics and mortality at one year are reported using descriptive statistics.

Results:

Among 13 seropositive NMOSD patients, the mean age at symptom onset was 33.7 years, and most were female (n=11, 85%) and black African (n=12, 92%). Of ten patients with at least one year of follow-up, four died within one year of symptom onset. All deceased patients were female, had at least two relapses within the first year of disease onset, and all developed longitudinally extensive transverse myelitis (LETM). Half of the deceased patients had been initiated on immunosuppression, both with rituximab. One-year mortality was higher in seropositive NMOSD as compared with multiple sclerosis and stroke in Zambia.

Conclusions:

Higher mortality among those with LETM, more relapses, immobility-related complications, and limited access to immunosuppression all likely contributed to the observed high early mortality among seropositive NMOSD in Zambia. Urgent action to address contributing factors is needed.

Introduction

Neuromyelitis optica spectrum disorder (NMOSD), a rare autoimmune astrocytopathy, carries a high mortality, risk factors for which include aquaporin-4 antibody positivity,¹ older age at onset,² Afro-Caribbean ethnicity,^{2, 3} transverse myelitis,¹ and delayed initiation of immunosuppression,⁴ with cause of death typically being neurogenic respiratory failure.¹ In sub-Saharan Africa, where patients are younger at disease onset^{5, 6} and availability of aquaporin-4 antibody testing is low,^{7, 8} mortality remains understudied. This study sought to report preliminary findings of one-year mortality among a cohort of seropositive NMOSD patients in Zambia and characterize key contributing factors.

Methods

Study Setting

Zambia is a low-middle income country in sub-Saharan Africa where primary healthcare facilities abolished user fees in 2006 and a National Health Insurance Scheme was launched in 2019. Despite these policy changes, transportation, medication shortages and coverage for specialty medications remain significant challenges.^{9, 10} The national referral hospital, University Teaching Hospital, located in the capital city of Lusaka, provides the highest level of care in the country; while patients may self-refer, most are typically referred from district and provincial hospitals. At least four MRI units are available within Lusaka, though high out-of-pocket costs and consistent operation are barriers.⁸ Antibody testing for aquaporin-4 and MOG is only available via commercial laboratories in Lusaka, costs approximately 5 USD and are performed via ELISA; gold standard live cell-based assays are not available in the country. Acute treatments available include intravenous and oral methylprednisolone, and plasma exchange, though the latter is only available at the national referral hospital and often delayed due to equipment shortages. Availability of maintenance immunosuppression for NMOSD is limited mostly to azathioprine or mycophenolate mofetil (unpublished data, Sahu). Rituximab is available at the national referral hospital, and may be covered under the national insurance plan, though there are often delays to insurance enrollment and coverage approval.

Study Design and Analysis

This cross-sectional study was conducted among adults at the University Teaching Hospital, the national referral hospital in Lusaka, Zambia from October 2019 to December 2025. Only individuals who met the 2015 consensus diagnostic criteria¹¹ for NMOSD and had a positive aquaporin-4 antibody test (ELISA). Antibody titers were not consistently reported by the laboratory so were excluded. Demographic, clinical, and treatment characteristics and mortality at one year after symptom onset were reported using descriptive statistics. Listwise deletion was used to address missing data. This study was approved by the ERES Converge Institutional Review Board, the Zambia National Health Research Authority, and the Johns Hopkins University Institutional Review Board.

Results

Thirteen seropositive NMOSD patients were included (Table 1). Approximately half were initiated on either rituximab (n=6, 46%) or azathioprine (n=2, 15%) for immunosuppression, mainly limited by cost or inconsistent follow-up. Out of ten patients with at least one year of follow up, four died within one year of symptom onset (n=4/10, 40%). All patients who demised had at least two relapses within one year of disease onset, all developed LETM and all were female. Two had not been initiated on immunosuppression before death. Of the two that had initiated treatment, both had started rituximab (one had well-controlled HIV and received the two-dose induction; the other only received one dose prior to death). Of the eight total patients with ≥5 years of follow-up, two died, both at ≥10 years after disease onset and one after defaulting on rituximab. Identifying cause of death was limited in the absence of electronic health records, and, upon follow-up, most family members only noted generalized weakness. When compared to other non-communicable neurologic conditions in Zambia, work from our group found that seropositive NMOSD carries the highest all-cause one-year mortality compared to antibody-unknown suspected NMOSD (15%), multiple sclerosis (MS) (0%), and stroke (26%) (Figure 1A, unpublished data, Saylor).

Table 1:

Demographic and clinic characteristics of seropositive NMOSD patients: overall and by one-year survival status

	Overall (n=13)*	Died at 1 year (n=4)	Alive at 1 year (n=6)
Age at first symptom onset (years)
Mean ± SD	33.7 ± 6.4	33.5 ± 10.6	32 ± 6.1
Disease duration (years)
Median (interquartile range)	1 (1, 6)	1 (1, 1)	7 (4, 10)
Sex (%)
Female	11 (85%)	4 (100%)	4 (67%)
Race/Ethnicity (%)
Black African	12 (92%)	4 (100%)	5 (83%)
Asian	1 (8%)	-	1 (17%)
Disease course included (%)
Optic neuritis	9 (69%)	2 (50%)	5 (83%)
Longitudinally extensive transverse myelitis	11 (85%)	4 (100%)	4 (67%)
Antibody status (%)
Aquaporin-4 antibody positive	13 (100%)	4 (100%)	6 (100%)
Number of new relapses within first year of disease onset **
Median (interquartile range)	1.5 (1, 3)	3 (2.5, 3.5)	1 (1, 1.75)
Treatment initiated (%)
Rituximab	6 (46%)	2 (50%)	3 (50%)
Azathioprine	2 (15%)	-	2 (33%)
No treatment initiated	5 (38%)	2 (50%)	1 (17%)

^*Three patients have not had follow up for at least one year and are excluded from the subgroups.

^**Number of relapses was not available for two of the patients who died within the first year of symptoms onset.

Figure 1.

Discussion:

Reports of five-year mortality due to seropositive NMOSD vary globally, ranging from nearly 10% in the United Kingdom, Japan² and Denmark¹ to 16% in the Caribbean population.¹² While one-year mortality is not explicitly documented, cohort studies have demonstrated that mortality increases with time.¹ To our knowledge, this is the first report of estimated mortality among a predominantly black African cohort of seropositive NMOSD patients. In this study conducted in Zambia, preliminary estimate of one-year mortality is 40%, both far exceeding estimates from other world regions and surpass the one-year mortality of two other non-communicable neurologic conditions in Zambia. However, notable similarities between the Zambian cohort and others include a similar number of relapses within first year of disease onset and presence of LETM in all deceased patients.^{1, 2, 12}

Numerous factors from the pre- to post-diagnosis period may contribute to the high early mortality observed in this cohort (Figure 1B). Pre-diagnosis factors include differences arising from intrinsic disease severity, referral bias, delays in diagnosis or misdiagnosis,¹³ and limitations in antibody testing. While international cohort studies have reported that those of African ancestry may present with more severe relapses at onset, it is not clear whether this reflects intrinsic disease biology, genetic susceptibility or unmeasured social and structural determinants.^2–4 The role of ancestry-related biological factors cannot be disentangled from healthcare access disparities based upon this study, but represents an area for future investigation. However, mortality among people with stroke in Zambia, who are on average nearly twice as old, is lower than mortality in people with seropositive NMOSD. Additionally, MS co-occurs with more comorbidities than NMOSD in Zambia¹⁴ but carries a far lower mortality. This suggests that age and chronic comorbidities are unlikely to be driving mortality in NMOSD. Secondly, referral to the national hospital, where this study took place and where individuals can access specialized neurological care, necessitates recognition of disease complexity from primary-level healthcare facilities, and resource constraints result in prioritization of more severe cases. Therefore, it is possible that a combination of low awareness amongst healthcare providers and resource constraints led to an over-representation of severe cases in this cohort. Third, delays in diagnosis or misdiagnosis may have resulted in a higher pre-diagnosis morbidity, even if number of relapses was overall similar to other regions. Fourth, due to inclusion of only seropositive patients and all testing completed via ELISA, which carries a low sensitivity (60%) compared to cell-based assays, (80–94%),¹⁵ it is very likely that additional cases were excluded, including among our center’s 20 suspected NMOSD cases with unknown or negative serostatus. Lack of gold standard antibody testing further highlights that the true burden of seropositive NMOSD in Zambia is underestimated, and this cohort represents only a fraction of diagnosable cases.

Post-diagnosis factors contributing to high early mortality include delayed initiation of acute treatment with plasma exchange, delays or lapses in immunosuppressive therapy, and higher rates of complications arising from immobility and infections. Limitations in rapid, consistent access to acute treatment and preventative immunosuppression also affect people with MS in Zambia. However, the marked differences in mortality between MS and seropositive NMOSD in Zambia supports that the latter is a uniquely aggressive disease that is far less forgiving of delays and lapses in treatment. High rate of infectious complications among people with neurological disorders in Zambia¹⁶ further predisposes patients with LETM-associated immobility to worse outcomes. Infection likely plays an important role in mortality within the first year of symptom onset, given that mortality due to NMOSD itself typically results from repeated relapses over years, eventually affecting the upper cervical cord or brainstem and resulting in neurogenic respiratory failure.¹

Despite the high early estimates of mortality demonstrated in this cohort, there are currently no consensus guidelines for the diagnosis and management of NMOSD in sub-Saharan Africa. To address this gap, we propose potential solutions by synthesizing evidence from several complementary sources in the context of factors contributing to mortality in this study. These include published data on diagnostic barriers for demyelinating diseases in Zambia,^{13, 17} consensus guidelines for NMOSD management developed in other resource-limited settings (Latin America and Middle East North Africa),^{18, 19} emerging literature on novel diagnostic approaches (such as dried blood spot testing and rapid immunodot assay),^{20, 21} and adaptations from evidence-based inpatient care guidelines designed for stroke in resource-limited settings, particularly regarding immobility-related complications.²² These sources were supplemented by the clinical experience of the authors and their colleagues who routinely manage patients with NMOSD and other neurological conditions in Zambia. Proposed solutions corresponding with stage of diagnostic journey are summarized in Figure 1B. Notable recommendations include implementation of low-cost antibody testing in community hospitals, inclusion of gold standard antibody testing at referral hospital, educational interventions to prompt early referral and initiation of acute treatment, adaptation of local protocols to reduce acute complications, and collaboration with national and regional stakeholders to improve availability and cost of immunosuppressive medications.

Limitations to this study include referral bias, lack of access to cell-based antibody testing, recall bias, and lack of confirmation of cause of death. Additionally, individual data on health insurance coverage, out-of-pocket healthcare expenditures, caregiver availability, employment status, and distance from referral center were not collected. These structural factors are known to influence healthcare access and outcomes and may contribute to the high mortality observed in this cohort. Due to small sample size, the precision and generalizability of mortality estimates are limited, and these findings should be interpreted as preliminary, hypothesis-generating observations rather than definitive conclusions. Regardless, this study highlights a novel population in a resource-limited setting. While additional studies from this region may help corroborate or clarify findings, urgent action is needed to address the underlying systematic factors that contribute to a high proportion of early, preventable deaths in seropositive NMOSD.

Highlights.

• Four out of ten patients with seropositive NMOSD in Zambia died within one year.

• All patients who died had two or more early relapses and developed LETM.

• NMOSD one-year mortality exceeded that of both MS and stroke in Zambia.

• Immobility complications and limited immunosuppression likely contribute.

• Contributing factors and proposed solutions for resource-limited settings are outlined.

DATA AVAILABILTITY STATEMENT

Anonymized data not published within this article will be made available by request from any qualified investigator that follows Zambian regulations regarding data sharing.