In the past few decades, rapid advances in genomics have revealed the genetic architecture of Mendelian and complex human diseases. However, the majority of these studies have been performed in populations of European ancestry and to a lesser degree, in African Americans, who share ancestry with West Africans. The translation of knowledge gained from genetic studies to improve health outcomes at the individual and public health levels is the overarching goal. The Human Heredity and Health in Africa Consortium, the African Wits–INDEPTH Partnership and other genetic initiatives are developing the infrastructure, supporting large-scale recruitment for genomic studies and building an ethical framework for genomics research on the African continent.(1,2) These initiatives herald the beginning of genetic-based personalized medicine and precision molecular diagnosis for clinical management and public health measures to reduce the burden of chronic and infectious diseases in Africa. The aim of this commentary is to discuss the advances and challenges of genetic screening and testing for monogenic and complex chronic kidney disease (CKD) in sub-Saharan Africa (SSA), drawing on the experiences and challenges in the United States for genetic testing in the clinical care setting.
BURDEN OF CHRONIC KIDNEY DISEASE IN SUB-SAHARAN AFRICA
CKD is the 12th leading cause of death globally, affecting 10 million people and causing more than 1.2 deaths per year.(3,4) CKD is defined as persistent abnormalities of kidney structure or function for greater than 3 months and is classified on the basis of aetiology, decreased estimated glomerular filtration rate (eGFR) <60 m/min/ 1.73 m2 and albuminuria (urine albumin/creatinine ratio) ≥30 mg/g.(5) Although HIV, malaria and tuberculosis remain the leading causes of death in SSA, many African countries and urban regions are experiencing an epidemiological transition, with the attendant increasing burden of diabetes, hypertension and obesity, which are risk factors for chronic cardiovascular and kidney disease.(6) The intersection of non-communicable and communicable diseases (i.e. HIV infection, diabetes, obesity and hypertension) is fuelling an alarming increase in CKD prevalence in SSA.(7,8) CKD is associated with increased cardiovascular risk and mortality and may progress to kidney failure with the need for kidney replacement therapy (i.e. dialysis or kidney transplant) to preserve life.(9–11) Because kidney replacement therapy is often unavailable and/or unaffordable in low or limited resource settings,(12,13) a diagnosis of CKD portends a decreased quality of life and survival for the majority patients and is often a death sentence for those requiring renal replacement therapy.(14) CKD also increases morbidity and mortality in a setting of malaria, HIV infection, diabetes, hypertension and cardiovascular disease, adding to the public health burden in SSA.(15,16) Unfortunately, treatment for most forms of CKD is limited, and management goals are to stem or slow progression to kidney failure through lifestyle modifications, reduction of proteinuria and lowering of blood pressure by the use of renoprotective renin–angiotensin II inhibitors (angiotensin-converting enzyme inhibitors (ACEis) or angiotensin II receptor blockers (ARBs)).(17)
Although there is a paucity of reliable population-based data on the prevalence of CKD in SSA, recent studies indicate that CKD is a substantial and growing public health burden. A 2018 systematic review of 152 articles published from 1995 to 2017 on the African continent reported an overall prevalence of CKD of 10% in the general population, driven primarily by hypertension and diabetes,(7) with higher rates in SSA (11%–16%) and lower rates in North Africa (~4%). A systematic review and meta-analysis reported an overall CKD prevalence of 15.8% on the African continent, with prevalence in SSA of 17.7% and nearly 5% having moderate or severe CKD.(18) A recent community-based survey of adults 40–60 years of age in five urban, rural and semi-rural sites across four countries reports an overall prevalence of 10.7% with risk factors being older age, female sex, having hypertension, diabetes and HIV infection.(19) However, considerable regional and site differences in CKD prevalence were observed, related in part to sociodemographic and epidemiological health transitions patterns, which differed between sites.(19) Although these studies have limitations, including selection bias and cross-sectional nature of the study, they point to the high prevalence rates of CKD in SSA, particularly in West and Southern Africa, and the need for additional tools for prevention, detection and treatment to prevent elevated risk of cardiovascular disease and progression to end-stage kidney disease (ESKD).(10)
MOST common FORMS OF CHRONIC KIDNEY DISEASE SHOW POLYGENIC INHERITANCE
CKD is heterogenous with many underlying causes, and multifactorial with physiological, environmental and genetic risk factors contributing to disease onset and rate of kidney function loss. CKD clusters in families, and creatinine-based eGFR, a measure of kidney function, has a heritability of 21%–31%.(20,21) Genome-wide association studies (GWAS) and meta-analyses of hundreds of thousands of individuals have identified hundreds of genes associated with eGFR, proteinuria, CKD and other renal phenotypes.(22–25) Translating these GWAS findings to patient care has proven challenging due in large part to the underlying heterogeneity of CKD.(26,27)
Although Mendelian (monogenic) mutations are an important cause of CKD, for the most common aetiologies of CKD (e.g. diabetic nephropathy and hypertensive nephropathy), polygenic inheritance is likely to play as large a role as monogenic variants.(27,28) Polygenic risk scores (PRSs) are based on the cumulative effect of thousands of variants across the genome, identified by GWAS to be associated with polygenic diseases.(29) Most common variants, with a few notable exceptions, have very small individual effect sizes, but the aggregate effects, captured by PRSs, are highly predictive for several common complex diseases. Khera et al. developed and validated a genome-wide polygenic score that could identify 8.0%, 6.1%, 3.5%, 3.2% and 1.5% of individual's European ancestry at 3-fold or more increased risk for type-2 diabetes mellitus, breast cancer, inflammatory bowel disease, coronary heart disease and atrial fibrillation, respectively.(30) Although PRSs may prove less predictive for CKD due to the heterogenous nature of CKD, a PRS using a weighted average of 1.2 million variants for eGFR was recently reported to predict incident CKD, ESKD (dialysis), kidney failure and acute kidney injury.(26,29) As shown in Figure 1, propensity scores incorporating traditional risk factors, biomarkers and predictive genetic variants might identify individuals at-risk for progressive CKD and guide clinical management through modifiable risk reduction and could guide clinical management even for those reaching kidney failure and requiring renal replacement therapy.(31) It is worth noting that the PRS was based on the summary statistics of over 850,000 individuals and validated in a cohort of nearly 9000 European–American participants, a relatively homogenous population.(29)
Disease-associated variants and PRSs derived from these variants (identified by GWAS of European populations) frequently fail to replicate or perform well in non-European populations due to differences in allelic frequencies, linkage disequilibrium and haplotype structure, and confounding by environmental risk factors.(32) Genetic diversity between African ethnolinguistic groups is greater than the diversity between Europeans and Asians, which represent only a subset of African diversity.(33) A recent whole genome sequencing analysis of 426 individuals from 50 ethnolinguistic groups highlights the breadth of genetic diversity and complex patterns of admixture and genetic stratification in SSA, indicating that the development of pan-African PRSs for multifactorial complex diseases like CKD will be challenging.(34) However, using whole genome genotyping arrays or whole genome sequencing to capture African-specific variants and to develop PRSs based on ancestry-related clusters should overcome this hurdle while increasing sensitivity and specificity for risk prediction using PRSs.(34)
APOL1 RENAL RISK VARIANTS AND KIDNEY DISEASE
African Americans have 3- to 4-fold higher risk of ESKD compared to their White counterparts.(35) Much of this excess risk in African Americans prevails due to the carriage of two variant alleles (hereafter referred to as high-risk (HR) genotype) in APOL1. APOL1 encodes apolipoprotein L1, a protein which conveys protection against Trypanosoma brucei brucei, but not the two strains causing African human trypanosomiasis.(36) Carriage of two coding variant alleles, termed G1 (comprising two missense variants (p.S342G and p.I384M)) and G2, a six base per deletion (N388del:y389del), are strongly associated with focal segmental glomerulosclerosis (FSGS), HIV-associated nephropathy (HIVAN) and non-diabetic CKD and ESKD (Tables 1 and 2).(37–41) The association of APOL1 variants with CKD is the strongest in a setting of untreated or undertreated HIV infection with odds ratios (ORs) of 29 and 89, respectively, in African Americans and South Africans.(38,42)
Despite the strong associations with specific aetiologies of CKD and hypertension-attributed ESKD, the lifetime risk of CKD in carriers of HR genotypes (defined as the carriage of two risk haplotypes) is only about 20% among the 13% of African Americans carrying HR genotypes, suggesting that other environmental or genetic modifiers are yet to be discovered. APOL1 is up-regulated by cytokines such as interferon, and high inflammatory states in a setting of untreated HIV infection, COVID-19, systemic lupus erythematosus and interferon therapy and are strongly associated with collapsing nephropathy.(43) In the African American population, both in the general population and in those with CKD, APOL1 HR genotypes are associated with a 2-fold risk of incident proteinuria, a marker of renal injury, and subsequently experience a steeper decline in eGFR than carriers of low-risk genotypes.(44,45)
The association of APOL1 with renal phenotypes in SSA has not been established, but it is likely that differences in environmental exposures, genetic background and traditional risk factors for CKD will modify APOL1 penetrance in different settings, as has been observed for FSGS and HIVAN in South Africans and African Americans (Table 1).(38,42) Interestingly, studies conducted in SSA and among African ancestry individuals living in the Americas, who have differing proportions of West African ancestry, suggest that APOL1 effect sizes may differ for different forms of kidney disease (Table 1).(46,47) For example, a study of stage-5 hypertension-attributed CKD in South Africa found no evidence of association with APOL1 HR genotypes, whereas APOL1 HR genotype is associated with 7.3-fold and 10.9-fold increased odds of ESKD in the USA and Brazil.(37,47) The South African study was only powered to detect associations with ORs of 3.2 or greater, suggesting that APOL1 HR are not strongly predictive of kidney failure or advanced CKD associated with hypertension in South Africa.(47) Among the Igbo in Nigeria, who share ancestry with African Americans, APOL1 HR genotypes were observed in 66% of individuals with CKD compared to 23% in controls.(48) On the other hand, APOL1 HR genotypes are more strongly associated with HIVAN (OR 89) in South Africans, compared to OR of 29 in African Americans, although the prevalence of APOL1 HR genotypes was similar between the two groups (79% and 72%, respectively) (Table 2).(38,42) Regional and population-specific differences in effect sizes in SSA point to the need for well-powered longitudinal and case–control studies to identify populations at-risk for APOL1-associated CKD and to identify population-specific genetic and regional environmental modifiers of APOL1 penetrance.
|Condition and setting||Effect size||Country||References|
|FSGS (mostly adults)||OR 10–17||USA||(37,38)|
|ESKD (hypertensive CKD)||OR 7||USA||(37)|
|Prevalent ESKD (diabetic)||No effect||USA||(85)|
|Incident ESKD, community-based, no CKD at baseline||HR 1.9||USA||(40)|
|Incident ESKD, community-based, with CKD||HR 2.2||USA||(40)|
|Prevalent ESKD, dialysis centres||OR 10.9||Brazil||(86)|
|Prevalent stage 5 CKD, hospital||No effect||South Africa||(47)|
|Prevalent CKD, hospital||OR 6.5||Nigeria||(48)|
|Incident CDK, community-based, no CKD at baseline||HR 1.5||USA||(40)|
|Incident ESKD (diabetic CKD), HR||OR 3.5||USA||(39)|
|Incident ESKD (hypertensive CKD), HR||HR 2.2||USA||(39)|
|Allograft survival, donor genotype||HR 3.8||USA||(87)|
|Allograft survival, recipient genotype||No effect||USA||(88)|
|Sickle cell nephropathy||OR 3.4||USA||(89)|
|Lupus nephritis, ESKD||OR 2.6||USA||(90)|
|Lupus nephritis, collapsing glomerulopathy||OR 5.4||USA||(91)|
|COVID-nephropathy, case series||7/7||USA||(92)|
|Incident albuminuria (general young adult population)||OR 2.9||USA||(44)|
|Microalbuminuria (general paediatric population)||OR 2.1||DRC||(93)|
APOL1 associations with different forms of CKD. Odds ratio (OR) are shown for case–control studies and hazard ratios (HR) are shown for time-to-event analysis; all results are for carriage of two APOL1 renal risk haplotypes (high-risk genotypes).
|Effect size||APOL1 high-risk genotype (two risk alleles)||Reference|
|Effect size||Pop. freq. (%)||Case freq. (%)|
|HIVAN||South Africa||OR 89||3||79||(42)|
|HIV+ FSGS||USA||OR 5.3||13||63||(94)|
|HIV+FSGS||South Africa||OR 2.1*||3||8||(42)|
|Progression to ESKD, biopsied CKD||USA||HR 2.9||13||NA||(61)|
|Microalbuminuria (Children with HIV)||DRC||OR 22||7||78.3||(93)|
APOL1 associations with different forms of HIV-associated CKD. Odds ratios (ORs) are shown for case–control studies and hazard ratios (HRs) are shown for time-to-event analysis; all results are for carriage of two APOL1 renal risk haplotypes (high-risk genotypes).
*Not statistically significant.
CLINICAL UTILITY OF APOL1 GENETIC SCREENING AND TESTING
Franceschini et al. calculated the sensitivity and specificity of APOL1 genetic testing for the diagnosis of FSGS and screening for the prediction of incident CKD or CKD progression.(49) They found that a positive APOL1 genetic test for FSGS would have a low diagnostic value, whereas a negative test would rule out APOL1-related FSGS, suggesting other aetiologies for FSGS. On the other hand, a positive APOL1 test has low sensitivity for screening or prognosis of CKD or progression to ESKD, respectively, and a negative test would also not be helpful for screening because it lacks sensitivity for the prognosis of CKD and progressive CKD.(49)
In African Americans with sporadic FSGS, APOL1 HR status is associated with earlier age of onset, with 70% developing FSGS between 15 and 49 years of age, and more rapid progression to kidney failure than carriers of low-risk genotypes of APOL1. Whether APOL1 HR status predicts steroid-sensitive FSGS is not resolved. Kopp et al. reported that the frequency of steroid sensitivity in FSGS does not differ by APOL1 genotype but the sample size was small.(38) Another study showed that in African American children with nephrotic syndrome, APOL1 was associated with 4-fold higher odds of steroid resistance, whereas two HLA-DQA1 variants were highly predictive of steroid sensitivity. Since the prevalence of these variants differs by race, it may explain in part racial differences in steroid sensitivity.(50,51) Currently, treatments for APOL1-associated CKD and FSGS do not differ from standard-of-care management, so there is little benefit for genetic testing for APOL1 risk alleles, although this might change when treatments targeting APOL1 gene or protein become available.
MONOGENIC KIDNEY DISEASES
Monogenic or Mendelian kidney disease is classified by mode of inheritance as autosomal dominant (AD) or autosomal recessive (AR). To date, more than 500 monogenic causes of CKD have been identified.(52,53) Endogamy (the custom of marriage within a specific group), consanguinity or demographic events (i.e. founder events) may inflate the prevalence of recessive mutations in particular populations or in a geographical region.(54) This is particularly relevant in Africa which has numerous ethnolinguistic, endogamous populations.(34)
Whole exome sequencing of children and adults has revealed that a precise genetic diagnosis can be achieved in 20%–40% of individuals with CKD or ESKD.(55–57) Whole exome sequencing of 114 Irish families, with at least one affected adult with CKD of unknown aetiology, identified a pathogenic mutation in a known CKD gene in 69% of families with extra-renal features, 36% with a positive family history of CKD and 15% in those with no extra-renal features or family history, for an overall detection rate of 37%.(58) In another study comprising 3315 mostly adult patients, a genetic diagnosis was identified in ~9% (52). However, unlike the previous study, individuals were included with diabetic or hypertensive CKD and a smaller number reported family history. Features associated with the identification of a causal pathogenic mutation (diagnostic yield) included a family history of kidney disease, congenital or cystic renal disease, nephropathy of unknown aetiology and glomerulopathy.(52) Whole exome sequencing to determine monogenic causation of childhood and adult onset CKD, particularly CKD with unknown aetiology, may provide a definitive genetic diagnosis and inform clinical management.(58)
GENETIC BASIS FOR PRIMARY NEPHROTIC SYNDROME IN CHILDREN
Primary nephrotic syndrome is one of the most frequent glomerular diseases in childhood and is characterized by nephrotic range proteinuria causing peripheral oedema, hypoalbuminuria and hyperlipidaemia. Nephrotic syndrome varies in histological pattern, steroid response and long-term outcomes by race, ethnicity and geography, reflecting in part differences in the prevalence of underlying common and rare genetic variants predisposing or causing nephrotic syndrome, respectively.(38,42,59–66) Primary nephrotic syndrome in children is classified by response to corticosteroid therapy, pattern of relapse, histopathology and the presence or absence of causative genetic mutations.(67) The majority of patients are steroid sensitive, but approximately 10%–20% are steroid-resistant (SRNS) and are at a significantly greater risk of developing ESKD, with 50%–70% of children developing ESKD within 5–10 years of diagnosis.(68)
The incidence of nephrotic syndrome is higher in children of Asian descent compared to White or Black children, but African American and Black African children are more likely to be steroid resistant.(69–75) Black African children are 2-fold more likely to be steroid resistant and have a histology of FSGS, and among children who are steroid resistant, they are less likely to go into remission following intensive immunosuppressive therapy compared to Asian children.(73)
A COMMON AFRICAN-SPECIFIC VARIANT OF SRNS-FSGS IN BLACK CHILDREN IN SOUTHERN AFRICA
Using targeted exome sequencing, an AR mutation in the NPHS2 gene encoding podocin was identified in 30% of Black, but not Asian, children with SRNS living in KwaZulu-Natal.(66) Subsequently, Govender et al. extended the finding to Black children of multiple ethnic groups (i.e. Tsonga, Tswana, Swati, Sotho, Pedi and Zulu) of the Bantu ancestry, suggesting that the variant may be broadly distributed in Southern Africa.(65,66) NPHS2 encodes podocin, a critical protein required for the integrity of the podocyte slit diaphragm, as an essential component of the tripartite glomerular filtration barrier. The p.V260E mutation, previously reported in consanguineous families in the former Omani empire, disrupts the transport of the protein from the endoplasmic reticulum to the plasma membrane.(76)
CLINICAL CHARACTERISTICS OF NPHS2 p.V260E SRNS
In two studies enrolling children <18 years of age with SRNS, NPHS2 homozygosity for NPHS2 p.V260E provided a precision molecular diagnosis in 33% and 55% of children with SR-FSGS.(65,66) Age-of-onset for p.V260E homozygotes ranged from less than 2 to 12 years of age.(65,66) Although neither African study assessed children with congenital nephrotic syndrome, previous studies have reported five congenital cases (age-of-onset in the first 3 months of age) homozygous for p.V260E from the Comoros Island and Saudi Arabia.(77,78) This variable age-of-onset is also seen within families with affected siblings developing SR-FSGS at 3 and 10 years in one family and 6 and 9 years in a second family.(65) This phenotypic variability suggests a role for possible genetic or environmental modifiers. In the Govender et al. study, three children carrying two APOL1 risk alleles presented with SR-FSGS before 4 years of age, one of the affected children was homozygous for the common V allele and two were homozygous for E mutation.(65) In the Durban study, only one affected child carried two APOL1 risk alleles.(65) Further study is required to confirm APOL1 as a potential effect modifier or a cause of childhood nephrotic syndrome in African populations.
These and previous studies have shown that affected children carrying AR podocin mutations do not respond to conventional therapy or intensive immunosuppressive therapy (i.e. glucocorticoid treatment, cyclophosphamide, tacrolimus and other calcineurin inhibitors).(66,77–79) As a consequence, children with NPHS2 mutations are more likely to progress to ESKD; however, there is considerable variability in time to kidney failure. In the African studies, mean kidney survival was 26.9 months in patients homozygous for the mutation and 69.2 months in patients without p.V260E.(65)
CLINICAL UTILITY OF SCREENING BLACK CHILDREN WITH NEPHROTIC SYNDROME FOR NPHS2 p.V260E
Genetic testing for NPHS2 p.V260E has the potential to obviate the need for a renal biopsy for differential diagnosis and to optimize treatment, but further clinical trials are required to evaluate its clinical utility. Since non-Mendelian forms of SR-FSGS recur in approximately 70% of patients following renal transplantation, assessment of monogenic causes of SR-FSGS identifies those who are unlikely to have recurrent SR-FSGS following kidney transplantation.(79,80) Testing with a point-of-care, validated assay would provide a cost-effective alternative to invasive renal biopsy and provide a precision, molecular diagnosis of SR-FSGS (Figure 2). It would also spare children who are unresponsive to conventional or intensive immunosuppressive therapy (i.e. oral glucocorticoid treatment, cyclophosphamide and calcineurin inhibitors) and the toxic effects associated with these therapies. The benefit of alternative treatments, including renoprotection using angiotensin-converting enzymes or ARBs, should be considered in this group of children with SR-FSGS. Precision genetic diagnosis will enable clinicians to provide genetic counselling and to plan for renal replacement therapy or renal transplantation. Validation of genetic assays for genetic testing and clinical trials to determine the clinical utility and benefit compared to standard of care are needed prior to implementation of genetic testing for NPHS2 p.V260E or other genetic testing strategy.(49) As reviewed by Franceschini et al. and shown in Figure 2, there are three key aspects to the implementation of genetic testing to assure genetic test validity and clinical utility before a test is implemented. These criteria include the development and performance of a test to accurately and reproducibly identify genetic variants; proof of clinical validity and clinical utility – does genetic testing improve clinical outcomes for the patient above the usual standard of care?
GENETIC TESTING FOR PERSONALIZED MEDICINE AND PUBLIC HEALTH IN AFRICA
As genomics capacity is developed in Africa, genes and gene variants will increasingly be identified that have the potential for genetic screening, diagnosis, prognosis and therapeutic management of both complex and monogenic diseases.(1,2,81,82) Genetic-based approaches will not only benefit individuals through personalized medicine but will also identify populations at-risk for Mendelian and polygenic diseases through their having a high prevalence of disease-associated variants, which could affect public health decisions at the local or national levels. Examples for personalized public health include common drug metabolizing genetic variants that vary in prevalence among different regions or ethnolinguistic groups. Several genetic variants in drug metabolizing genes and transporters have been investigated that might affect treatment choices in persons with kidney disease. CYP3A4 and ABCB1 have been associated with response to cyclosporine therapy for renal transplant recipients; CYP3A4 and CYP3A5 haplotypes are associated with tacrolimus clearance, an immunosuppressive therapy sometimes used to treat SR-FSGS; and variants in CYP3A4 and ADRB2 have been associated with the blood pressure–lowering drugs amlodipine and ramipril, respectively.(83) The implementation of personalized pharmacogenetics in SSA requires additional genetic and clinical research, including clinical trials to identify population-specific variants that affect drug metabolism, clearance and clinical efficacy. An example is renal risk reduction (R3)-randomized clinical trial in Nigeria that is designed to study the efficacy of blocking the renin–angiotensin–aldosterone system with inexpensive ACEis to retard progression of kidney disease in patients with treated HIV infection and to determine if APOL1 HR status is associated with worse kidney outcomes among study participants.(84)
Using kidney disease as an example, I have shown the potential and limitations of personalized and precision medicine to improve patient outcomes for polygenic and monogenic Mendelian disorders. Figure 3 is a schematic to demonstrate how personalized genetic approach might be applied to improve health outcomes for individuals with monogenic or polygenic kidney disease by genetic screening and targeted genetic sequencing of gene panels (e.g. genes causative for polycystic kidney disease or SR-FSGS). The translation of genetic findings to the clinic has proven difficult in part because of the underlying complexity of polygenic inheritance and the influence of environmental modifiers on phenotypic expression. On the other hand, genetic screening for monogenic disorders provides a precision diagnosis for the family and may inform clinical management of the disorder. Testing for founder pathogenic variants, such as NPHS2 p.V260E, may provide a cost-effective alternative to more invasive or toxic diagnostic procedures (i.e. steroid treatment or renal biopsy). Africa is on the cusp of a genomics revolution, but hand-in-hand with advances in gene discovery is the need to build clinical research capacity to exploit genetic findings for clinical benefit, to provide health-care providers with the training and tools to use and interpret genetic results, to train clinical geneticists for family counselling and to engage communities, patients and other stake-holders.
ACKNOWLEDGEMENTS AND FUNDING
The project has been supported in part by the National Institutes of Health and the National Cancer Institute Intramural Research Program (CAW) and under contract HHSN26120080001E. The content of this publication does not necessarily reflect the view or policy of the Department of Health and Human Services nor does mention of trade names, commercial products or organizations imply endorsement by the government. I thank Dr. George Nelson for critically reading this manuscript and my collaborators, Drs. Raj Bhimma, June Fabian and Saraladevi Naicker for welcoming me to the South African nephrology community.