Improving precision diagnosis in unexplained CKD

Clinical Kidney Journal, vol. 19, pp. sfag006

Abstract

Chronic kidney disease (CKD) is a significant global health challenge, impacting millions of individuals worldwide. Traditionally, CKD classification has focused on identifiable causes such as diabetes, hypertension and glomerulonephritis, as well as glomerular filtration rate (GFR) and degree of albuminuria [1]. Yet, a subset of CKD patients exists for whom no clear underlying reason can be discerned despite comprehensive evaluations. In a recent publication, Halbritter and co-authors, on behalf of the ERA Genes & Kidney Working Group, present a consensus statement on chronic kidney disease of unexplained cause (CKDx) [2]. It is important to point out that CKDx is different from the previously described CKD of unknown origin (CKDu). CKDu is a term used for endemic forms of CKD with rapid progression reported in specific areas in the world, mainly Central America and Sri Lanka, and with a potential link to a high degree of occupational and environmental hazards [3]. CKDu is most frequently diagnosed in male individuals between the ages of 20 and 60 years living in rural or agricultural settings. This important differentiation between CKDx and CKDu may require substantial education of the community, as the names are semantically similar. Halbritter et al. estimate that a classification of CKDx would be applicable in approximately 16%–20% of the overall CKD population [2], representing a substantial proportion of affected individuals. The lack of clarity in diagnosis not only impairs patient management but also leads to uncertainties in clinical and epidemiological research. The Kidney Disease: Improving Global Outcomes (KDIGO) guidelines underscore the importance of a structured assessment framework for CKD, focusing on the ‘cause’ component in CKD classification in addition to GFR and albuminuria [1]. However, in CKDx, the ‘cause’ is unknown. Halbritter and colleagues stress the importance of clearly referring to a potential cause in medical reports after a thorough diagnostic workup. For most patients, this workup will include laboratory testing (blood and urine), imaging studies and often a kidney biopsy, which remains a powerful diagnostic tool. In addition, genetic testing is becoming increasingly important. Genetic testing was historically cumbersome, resulting in high costs and long delays until results were received, and is still not globally accessible. However, novel sequencing techniques allow more rapid analysis at reasonable costs. New computational approaches improve the analytical and interpretational part of genetic testing, albeit the required expertise in kidney genetics is still not widely available. To enhance the clarity and efficiency of CKDx management and to facilitate a comprehensive overview of each patient’s diagnostic journey, Halbritter et al. propose a new reporting system. This system is meant to standardize the documentation of diagnostic efforts, including the results and availability of histological (‘h +/–’) and genetic (‘g +/–’) tests performed during the evaluation process. Such standardization could potentially promote transparency and communication among healthcare providers, ensuring that every diagnostic attempt is recorded and accessible for future reference. An aspect to be evaluated is whether the suggested binary letter code will be sufficient and unequivocal in daily clinical life. Including the full terms ‘histology’ and ‘genetic testing’ and adding the precise results and techniques of diagnostic measures may turn out to be a required additional layer of information. Electron microscopy is not routinely performed in all pathology laboratories. Similarly, the extent of genetic investigations can vary substantially. It clearly makes a difference whether genetic testing was done as a single-gene assay, a small specific panel versus whole-genome sequencing (WGS). All three techniques could be marked as ‘g+’ in the binary code, but the limited testing is only sufficient for disorders with a clearly identifiable phenotype and single genetic aetiology, such as cystinosis [4]. CKDx, in contrast, is characterized by the inability to identify a specific aetiology and broader testing techniques are thus desirable. Yet, as most genomic variants are evolutionary recent and thus rare or even ‘private’, assessment of pathogenicity of variants identified in the genome of an individual with CKDx is often challenging. As pointed out by Halbritter et al., regular re-evaluation of findings is recommended, but of the three techniques mentioned above only WGS could easily be reanalyzed with extension to newly classified disease genes and variants. Novel sequencing technologies also increasingly allow detection of copy number variations, as well as chromosomal rearrangements, but this is not routinely done in all laboratories. Thus, knowledge of technical details of genetic testing is relevant and might become a helpful part of a reporting system. In addition, including information on ‘family history’ as a standardized item could also be considered as a very straightforward, but frequently neglected diagnostic item in the reporting system. Listing ‘family history’ (‘obtained’ vs ‘not-obtained’, ‘positive for kidney disease’ vs ‘negative for kidney disease’) as an item may serve as a straightforward reminder. Special relevance is given to the discussion of genetics in the publication. The benefits of genetic testing may be more obvious in paediatric nephrology with a higher diagnostic yield of genetic testing in children, as most inherited kidney diseases tend to present early in life and children obviously had less time to acquire non-Mendelian disease [5]. However, recent publications have highlighted the incidence of patients with undiagnosed genetic diseases in cohorts of adult patients with kidney diseases. Identifying a genetic cause in an individual with CKDx has broader implications for patient management, including treatment, risk assessment and potential extrarenal manifestations, as well as genetic counseling to understand implications for family members, allowing informed decision-making, and the development of targeted therapies or lifestyle changes. Yet, based on our experience of genetic testing in children, we also raise a note of caution about the potential untargeted use of genetic testing. Importantly, the authors correctly state (statement 4.4): ‘Identification of a pathogenic or likely pathogenic variant does not necessarily indicate that the variant is causal for the patient’s kidney disease.’ As with any diagnostic test used, the positive predictive value of genetic testing indicates the probability that individuals tested positive really are affected. This depends on the specificity of the test and the a priori likelihood of the tested person to be affected by the respective diagnosis. For genetic testing, this likelihood is determined by the prevalence of a Mendelian disorder in the tested population. The prevalence of identifiable causative variants in unselected cohorts of adults with CKD is at least 10%. The specificity of a genetic test is determined by the likelihood of pathogenicity and the commonly used criteria for variant assessment define the threshold for a clinical diagnosis as a likelihood of at least 90% that the identified variant is causative [6]. Still, a specificity of 90% with a prevalence of 10% of identifiable Mendelian disease translates into a positive predictive value of just over 50%—slightly better than flipping a coin (Fig. 1) [7]. Moreover, even pathogenic variants can be found in the general population with no apparent phenotype, such as heterozygous variants in the genes associated with Alport syndrome or polycystic kidney disease (reviewed in [8]). Thus, the identification of such pathogenic variant must be carefully correlated with the clinical phenotype to determine whether it is actually the explanation for the patient’s disease, contributory or simply coincidental. Initiation of genetic testing in a patient with a diagnosis associated with a low prevalence of identifiable causative variants should be very carefully considered and, if undertaken, even more careful assessment of the results is mandatory to avoid labelling the patient with an erroneous diagnosis. In many parts of the world this raises logistical questions, e.g. on the access to certified laboratories or the availability of qualified personnel to appropriately counsel patients in such a situation. Potential costs of a genetic diagnosis and the consequences of a misdiagnosis are obviously highly relevant for patients. This highlights the need for shared-decision making between patients and healthcare professionals based on informed consent and the overarching ethical aspects of—genetic and non-genetic—diagnostics. Thus, many challenging questions for broader implementation of genetic diagnostics in adult nephrology remain to be addressed. There is an ongoing need for guidance in this field that should be coordinated by scientific societies. The development of guidance documents should include the patient’s view and should also discuss the need for ongoing specific education of patients and doctors in different healthcare settings. Positive predictive value of identified genetic variants depends on the a priori likelihood of an underlying genetic cause: posterior probability that a patient’s disease is caused by a detected variant (y-axis) depends both on the probability of that variant being pathogenic (x-axis) and the prevalence of detectable monogenic disease in the population tested (coloured lines). Variants classified as ‘likely pathogenic’ according to the American College of Medical Genetics and Genomics and the Association for Molecular Pathology criteria are deemed to have ≥90% certainty of being pathogenic (vertical dotted line). In tubulopathies, the prevalence of identifiable Mendelian disorders is typically at least 50% and thus, the positive predictive value is >90% (upper horizontal line). In contrast, in disorders, such as CAKUT (congenital anomalies of the kidney and urinary tract), where the reported diagnostic yield of genetic testing is as low as 5%, the positive predictive value is only around 35% (lower horizontal line). Modified from [7]. Obviously, the odds of finding a genetic explanation dramatically increase if there are features suggesting a Mendelian disorder, such as a positive family history or a specific symptom constellation. The six examples provided by Halbritter et al. of patients with CKDx in whom genetic testing was able to provide a specific diagnosis are all characterized by a positive family and/or signs or symptoms suggestive of a specific Mendelian diagnosis. This again highlights the importance of a careful family history and phenotype assessment as proposed by Halbritter et al., yet systematic testing of all organ systems may prove challenging to incorporate into routine clinical practice. In summary, the recognition and management of CKDx represents a critical frontier in nephrology. By employing a systematic approach to diagnosis, Halbritter et al. suggest that healthcare providers can enhance diagnostic precision and tailor individual patient care by straightforward integration of results from diagnostic workup, including biopsies and genetic testing. We highly welcome this effort. The standardized reporting system suggested by Halbritter et al. could be an important step forward. Yet ongoing review is critical to assess the feasibility of implementing this systematic approach in daily clinical life, to validate it in different healthcare systems and to identify those patients most likely to benefit from genetic testing. Through continued collaboration and a focus on personalized healthcare, we can aspire to improve outcomes for patients grappling with CKDx, ultimately providing them with hope and a clearer path forward in their health journey. We acknowledge the ongoing support of projects on genetic kidney diseases by the Working Group ‘Inherited Kidney Diseases’ of the European Society of Paediatric Nephrology. M.C.L. is supported by the German Federal Ministry of Research and Education (NEOCYST consortium, BMBF grant 01GM2203B) and by the EU Horizon program (TheRaCil consortium, EU grant agreement 101080717) and the German Research Foundation (DFG, Li-2397/8-1). P.B. is supported by the German Research Foundation (DFG, Project IDs 322900939 & 445703531 & INST 222/1582–1), European Research Council (ERC Consolidator Grant No. 101001791), the Federal Ministry of Education and Research (BMBF, STOP-FSGS-01GM2202C) and the Innovation Fund of the Federal Joint Committee (Transplant.KI, No. 01VSF21048). None declared. Comment on Halbritter et al. ‘Chronic kidney disease of unexplained cause (CKDx): a consensus statement by the Genes & Kidney Working Group of the ERA’ https://doi.org/10.1093/ndt/gfaf092

Authors 3

1. Max C. Liebau corresponding Aachen

   RWTH Aachen University · Universitätsklinikum Aachen · University of Cologne · University Hospital Cologne

   Affiliation as printed

   University Hospital Aachen, RWTH Aachen University Department of Pediatrics, , Aachen ,

   University Hospital and Medical Faculty, University of Cologne Department of Pediatrics, , Cologne ,

2. Detlef Böckenhauer

   KU Leuven · Universitair Ziekenhuis Leuven

   Affiliation as printed

   KUL Department of Paediatric Nephrology, UZ Leuven and Department of Cellular and Molecular Physiology, , Leuven ,

3. Peter Böör Aachen Institute for Pathology

   RWTH Aachen University · Universitätsklinikum Aachen

   Affiliation as printed

   Institute for Pathology, University Hospital Aachen, RWTH Aachen University , Aachen ,