The epidemiology of IgA nephropathy: East versus West

IgA nephropathy (IgAN) is reported to be the most common type of glomerular disease worldwide.1 Kidney biopsy registry studies have shown that the proportion of biopsies with glomerular disease that is due to IgAN is highest in Asia (39.5%), intermediate in Europe and North America (22.1% and 11.8%) and lowest in Latin America and Africa (6.1% and 2.8%). Although these results are impacted by factors unrelated to IgAN including biopsy practices and the regional frequency of other glomerular diseases, they nonetheless suggest that the incidence of IgAN may vary worldwide.2, 3 A recent systematic review evaluated the incidence of IgAN in different geographic regions using studies that accurately reported both the number of incident cases confirmed on kidney biopsy and the size of the population at risk.4 In the United States, the incidence ranged from 0.39 to 1.4 per 1000,000 people per year, and amongst the Black population the incidence was as high as 1.02 per 100 000 people per year. By comparison, the incidence in South Africa was extremely low at 0.06 per 100 000 people per year. In Western Europe, the incidence was similar to that in the United States, being the lowest in the United Kingdom and Northern Ireland (both 0.99 per 1000,000 people per year), and the highest in France (2.7 per 100 000 people per year). The incidence in Asia was highly variable, ranging from 1.8 per 100 000 people per year in Singapore to 4.2 per 100 000 people per year in Japan, and the incidence was substantially higher in Australia at 10.5 per 100 000 people per year. As such, there was a 175-fold difference in the incidence rates worldwide, from the lowest rate in South Africa to the highest in Australia. The large magnitude of observed differences in incidence rates that did not follow correlations with ethnic groups traditionally considered at higher risk of disease susceptibility suggest that other factors may be contributing to the variability in disease incidence. Genome wide association studies have identified genetic risk loci for the development of IgAN that are more common in East Asians or South Asians, and less common in Blacks.5 However, this did not match the observed trends in disease incidence. Australia is a multicultural country and had the highest observed incidence of IgAN. Even within Asia there was substantial variability, with an incidence rate over twofold higher in Japan compared with Singapore. The disease incidence amongst the Black population in the United States was similar to that seen in Europe or Singapore. Childhood screening programmes for hematuria and proteinuria in Japan may trigger kidney biopsy investigations and therefore contribute to a higher observed incidence of disease. Whether this reflects a truly increased risk of developing IgAN or early diagnosis of more mild disease is not clear. Reduced access to health care in socioeconomically disadvantaged populations in Brazil or South Africa may prevent kidney biopsy diagnosis and thus reduce the observed incidence of IgAN. However, another contributing factor may the lack of adequate population-level data infrastructure required to generate accurate estimates of disease incidence. To calculate an incidence rate, it is necessary to accurately capture both the number of new cases in a fixed time period and the population at risk. Error in the measurement of the population at risk can have a disproportionately large effect because it is in the denominator of the incidence rate. To overcome this limitation, the incidence of disease needs to be ascertained at the population-level in a well-defined region. This is often challenging when using biopsy data from renal pathology centres that may have a poorly defined source population. As such, the true incidence of IgAN in different geographic regions may not be accurately known. A recent study from Canada sought to address some of these limitations.6 Health care in Canada is organized at the provincial level, such that each province represents a relatively closed population with similar access to a single health-care system. One such province, British Columbia, has a single repository of all kidney biopsies performed from 2000 to 2012. This offers the opportunity to study the incidence of IgAN within a well-defined population at risk. The overall incidence of IgAN was 1.79 per 100 000 people per year (95% CI 1.66, 1.91) in a population that was 34% White, 48% Asian and 12% South Asian. The incidence of disease was not uniform across the province, and instead clustered in four distinct regions which was not explained by region-level demographics including age, sex or ethnicity. This suggests that other factors, such as environmental exposures, may explain the clustering of incident IgAN and demonstrates that regions with a disproportionately high incidence of IgAN cannot be identified based on demographic characteristics alone. Instead, each country should develop the necessary data infrastructure to accurately capture the population-level incidence of IgAN so that health-care resources targeting this high-risk kidney population can be effectively deployed to regions with an increased burden of disease. Over sufficiently long follow up, the majority of patients with IgAN are at risk of disease progression to end-stage kidney disease (ESKD).7 Although individuals of Asian ethnicity have been reported to have an increased risk of ESKD,8 it was not known if ethnicity per se contributed to this risk or whether it was confounded by other measures of disease activity, such as MEST-C histology scores.9-11 Until recently there was no validated method to predict individual risk of disease progression in multiple ethnic groups worldwide. In 2019, a large collaboration of researchers through the International IgA Nephropathy Network pooled datasets of adults with IgAN from North and South America, China, Japan and Europe to derive and externally validate the International IgAN Prediction Tool.12 Two models were developed to predict the risk of a 50% decline in estimated glomerular filtration rate (eGFR) or ESKD over a median follow up of 5 years. One model included Chinese, Japanese and White ethnicity, and the other model did not. Other predictor variables included in both models were age, blood pressure, eGFR, proteinuria, use of immunosuppression or medications that block the renin-angiotensin system, and MEST histology scores. There are now different versions of the Prediction Tool that have been validated to work in both adults and children, and either at the time of biopsy or 1- or 2-years post-biopsy to allow re-evaluation of risk after supportive therapy or early immunosuppression treatment.13, 14 The Prediction Tool models for adults have undergone additional validation in other cohorts from China, Argentina, India, Korean, Greece and France.15-20 All versions of the Prediction Tool are available for use in mobile-app and web-based calculators through Calculate by QxMD for iOS, Android and the web at https://qxmd.com/calculate-by-qxmd. In the original derivation of the Prediction Tool models, Chinese and Japanese ethnicity were associated with an increased risk of a 50% decline in eGFR or ESKD (HR 2.27, 95% CI 1.73, 2.96 and HR 1.50, 95% 1.07, 2.11, respectively). Although this may imply that Chinese and Japanese ethnicity may increase the risk of disease progression, several caveats exist. The Prediction Tool models without ethnicity had similar prediction performance as the models that included ethnicity. Therefore, it is not necessary to consider ethnicity to accurately predict individual risk of disease progression. This is fortunate, because it is not known if these results can be applied to patients of Chinese or Japanese ethnicity in regions outside of those that contributed data to the analytic cohorts. It is also not known if the models require updating to apply to other ethnic groups or to patients of mixed ethnicity. As a result of these uncertainties, it would be preferable to primarily use the Prediction Tool models without ethnicity outside Japan or China to avoid confusion regarding the validity or generalizability of assessing ethnicity in individual patients. In summary, although both the incidence of disease and the risk of disease progression may be associated with ethnicity, substantial variability remains that precludes accurate generalization to specific geographic regions or to individual patients. Instead, data infrastructure should be developed to identify regions with an increased burden of disease, and all available risk factors should be considered together in the International IgAN Prediction Tool when evaluating individual risk of disease progression.