Differential long-term impact of primary glomerular diseases on major outcomes: All are not equal!

Abstract

The landscape of nephrology, particularly that of glomerular diseases, is an evolving dynamic that has witnessed incredible advancements in diagnostic and therapeutic interventions to improve kidney and patient survival [1-3]. These achievements have been further bolstered by emerging clinical prediction models to enable better and faster clinical decision-making in order to maximize clinical benefit and minimize harm [4-6]. The more precise we are at defining the primary cause of disease and its epidemiological profile, including clinical course, trajectory pattern and downstream long-term outcomes, the more likely we will improve universal efforts that target its prevention. Greater precision in the classification of disease, its clinical phenotype, targeted therapeutic interventions and outcome measurement is the hallmark of precision medicine [7]. Glomerular diseases of the kidney are key exemplars where major advances have occurred in elucidating pathogenesis, molecular signatures, diagnostic techniques and novel treatment strategies, all of which have contributed to greater characterization of unique disease phenotypes and clinical outcomes [1-7]. However, despite these efforts, the scientific community has fallen short in mapping the long-term outcomes of common primary glomerular diseases [8, 9]. Published comparisons to date have suffered from relatively small samples and short follow-up periods. Many have considered glomerular diseases at the aggregate level rather than at the individual level [1-10].

In this issue of the Journal of Internal Medicine, Faucon et al. provide new data describing the long-term outcomes of four common primary glomerular diseases from analysis of the Swedish Renal Registry [10]. In this nationally representative observational study, they profile the trajectory patterns of IgA nephropathy (IgA), focal segmental glomerulosclerosis (FSGS), minimal change disease (MCD) and membranous nephropathy (MN). They reveal significant variation in major long-term outcomes across disease groups and demonstrate that the risks of hospitalization, death and kidney replacement therapy (KRT) for these primary glomerular diseases differ substantially from those of a control CKD population that included patients with diabetes (23.7%), hypertension (39.5%), tubulointerstitial nephritis (3.4%), medication-induced kidney disease (3.0%) and other/unknown causes (30.5%).

Their observations highlight strikingly different patterns of risk for estimated glomerular filtration rate (eGFR) decline, risk of cardiovascular events and mortality associated with each primary glomerular disease. They reveal that the cumulative mortality was highest for patients diagnosed with FSGS and MN and lowest for those with MCD. Similarly, patients with FSGS had cardiovascular event rates that were similar to patients in the control CKD population but significantly higher than those with MCD. Their observations were further corroborated in time-to-event analyses where adjusted mortality risks were significantly lower for each primary glomerular disease compared to the patients in the control CKD group. The pattern of risk for major adverse cardiovascular events followed a similar trend. These observations would suggest that primary glomerular diseases generate unique eGFR trajectories that can be tracked over time within a defined population. Moreover, their presence confers measurable impact on major outcomes that differ quite substantially depending on the underlying cause. Although it is possible that the differential risks presented here might be influenced by baseline health status at cohort inception, such as demographic factors, cumulative comorbidity burden or prescribed medications, the comprehensive adjustment for a wide range of confounding factors in multivariable models suggests the presence of a true measurable effect.

The differential impact of each primary glomerular disease on the risk of progression to kidney failure was noteworthy. Patients with underlying IgA disease or FSGS experienced the greatest risk of progression to kidney failure, whereas those with MN or MCD had the lowest risks of progression compared to the control CKD population. These observations were further supported by analysis of eGFR trajectories, whereupon patients with IgA disease and FSGS experienced the steepest declines in annual GFR slopes compared to the control CKD population, whereas patients with MCD experienced the least or no decline. Both FSGS and IgA were characterized by rapid progression trajectories with the highest risks of progression to kidney failure. The analysis also emphasizes a further observation that while all four primary glomerular diseases incurred mortality risks lower than the average CKD cohort, patients with specific primary glomerular diseases, namely, IgA and FSGS, experienced significantly higher risks of progression to KRT.

One might speculate that the current study does not adequately capture the true magnitude of the differences between individual glomerular diseases and individuals in the control CKD population [10]. Differences in baseline characteristics, including demographic profiles, comorbid burden and severity, standard kidney protective medications and specific immunosuppressive therapies, may have led to imprecision in risk estimation. Patient age is a major risk factor for GFR progression and mortality, and in the present study, the mean age difference between IgA and control CKD patients was a notable 21 years. There were also other marked differences in the distribution of clinical variables across groups, which may have influenced the magnitude of effect on each of the assessed outcomes. Furthermore, given that this is a real-world study, it is possible that treatment changes during the study period may have further influenced patterns of disease progression and longer term outcomes. Despite these challenges, the authors provide evidence of a robust analysis that took into consideration the challenges of the observational design, the differential distribution of baseline characteristics among groups and key confounding variables. Moreover, the study also benefited from very high rates of precision in the diagnosis of primary glomerular disease, with 97% of cases confirmed with a kidney biopsy, robust data linkage with national registries that tracked major outcomes, and minimal loss to follow-up.

Although other investigators have compared the outcomes of primary glomerular disease to that of the general population [11], the present study relied on controls derived from routine clinical nephrology practice with CKD attributed to common non-glomerular causes. This choice of control population is useful as it provides a contextual framework upon which to base comparisons with other uniquely classified CKD cohorts. It may be argued that the control population was itself quite heterogeneous as it included a wide range of different kidney disorders and may not represent a ‘immunophenotypically distinct disease’ with a unique disease trajectory and clinical outcome. Restricting to a single specific disease would have provided a more precise landmark on which to conduct comparisons. In sensitivity analysis, the authors found that the direction and magnitude of effects on long-term outcomes using more restricted definitions for the control population were broadly similar to those of the original study.

Glomerular diseases are a heterogeneous group of diseases with significant biological variability [1-3]. It is, therefore, likely that their clinical courses and outcomes will also vary substantially. The current study confirms this heterogeneity of effect on long-term outcomes of kidney failure, hospitalization and death. Moreover, the study underscores the importance of registry-based platforms to examine and characterize the longer term outcomes of primary glomerular diseases and their treatments in real-world settings [8-11]. The establishment of comprehensive disease registries with rigorous attention to disease diagnosis, immunophenotyping, disease trajectories and medication utilization serves to fill an important knowledge gap in our understanding of glomerular disease. These registry-based studies facilitate prognostic studies, development and validation of glomerular-specific risk prediction models and permit comparative effectiveness research on therapeutic interventions. Although clinical trials remain the gold standard for assessment of treatment effect, they have generally been small in size, with short follow-up periods and lack generalizability. In contrast, the registry-based approach fills several important gaps by facilitating larger-scale studies, broadening generalizability and characterizing long-term major outcomes that develop over longer time periods, including the visualization of long-term disease trajectories. The adoption of these real-world observatories into contemporary research programmes is a tremendous addition to the evolving landscape of nephrology and the science of precision medicine.

Austin G. Stack: Conceptualization; writing—review and editing; writing—original draft.

The author declares no conflicts of interest.