Microarrays Extend Promise of Precision Medicine to Pediatric Nephrology: Testing for CNVs Refines Diagnosis of CKD in Children

/kidneynews/7_6/1/graphic/1f1.jpg

Many children with kidney disease have rare “genomic imbalances” as the cause of their kidney dysfunction, often as part of neurodevelopmental syndromes. A new study finds that many unsuspected genetic diagnoses can be made using chromosomal microarrays to identify copy number variants (CNVs)—a “precision medicine” approach with major ramifications for treatment selection, family counseling, and long-term patient management.

The prospective study by a team of pediatric nephrologists and other specialists from seven centers found diagnostic copy number disorders in more than 7 percent of a large cohort of children with chronic kidney disease (CKD). “Detection of pathogenic imbalances has practical implications for personalized diagnosis and health monitoring in this population,” according to the report in the May issue of The Journal of Clinical Investigation (Verbitsky M, et al: J Clin Invest 2015; 125:2171–2178). The senior author was Ali G. Gharavi, MD, of Columbia University).

Chromosomal microarrays in pediatric nephrology

The researchers analyzed the genetic findings of patients enrolled in the Chronic Kidney Disease in Children (CKiD) study—an ongoing, long-term follow-up study of risk factors and outcomes in children with kidney disease. Using patient DNA derived from stored samples, the investigators assessed CNVs using high-density microarrays.

“Chromosomal DNA microarray is a relatively new technology, which essentially looks at the entire genome of an individual and tries to identify gain or loss of DNA material that may cause a genetic disease,” Gharavi said. Microarrays represent a major advance over the microscopic technique of karyotyping—classicially used to diagnose major chromosomal abnormalities such as Down syndrome.

“With karyotyping, we could only look at deletions or duplications larger than 1 to 2 million base pairs. Whereas with [chromosomal microarrays], we are able to detect much smaller gain or loss of DNA material—as small as 100,000 base pairs,” Gharavi said. “As a result of this, we are able to make diagnoses that we weren’t able to make by karyotyping.”

That’s important, because small variations in copy number make up a large part of genomic variation. Even in healthy individuals, up to 10 percent of the genome may be subject to this type of variation. Previous studies using microarray techniques have identified CNV disorders associated with a broad range of congenital and neurodevelopmental defects. Genomic imbalances can affect neurologic, cardiac, and skeletal development (an effect called “pleiotropy”), suggesting that common developmental pathways are involved.

The new study applied chromosomal microarrays to understanding the role of CNVs in a well-characterized group of children with clinically diagnosed kidney disease. A previous report found “pathogenic genomic imbalances” in about 10.5 percent of children and young adults with kidney malformations (Sanna-Cherchi S, et al: Am J Hum Genet 2012; 91:987–997).

These imbalances were clinically unsuspected and “overlapped significantly with CNV disorders implicated in neurodevelopmental disorders.” In the previous study, most known CNV disorders in patients with renal hypodysplasia (RHD) had previously been linked to developmental delay or neuropsychiatric diagnoses.

Matthew Sampson, MD, a pediatric nephrologist and genetic epidemiologist at the University of Michigan, was one of the investigators in the previous study. “Understanding the underlying mechanism always helps in terms of explaining to parents why their child is ill,” he said. “From a clinical perspective it often can help in terms of providing more precise prognoses or suggesting to us the optimal therapeutic regimen or further diagnostic tests. And, particularly important in children, it can help in terms of family counseling.”

While “personalized medicine” doesn’t always mean genetic testing, Sampson pointed out, “Genomic inquiry reveals more molecularly based, fundamental information about the pathogenesis of a child’s condition.” In regard to the new work by Verbitsky et al., he added, “This study is taking an approach that has really only recently been actualized on a clinical basis, or a research basis—to uncover some potentially causative genetic changes that could be responsible for a small but substantial percentage of the cases we see.”

Higher rate of CNV abnormalities in children with CKD

The analysis included 419 unrelated children from CKiD. The patients were being followed up for a wide range of clinical diagnoses, including RHD, obstructive uropathy, reflux nephropathy, and focal segmental glomerulosclerosis (FSGS), among others.

Even though CNVs account for a large part of overall variation in the genome, the population frequency of genomic disorders is very low. To be able to tell apart those low frequency genomic disorders from likely benign common variants, the researchers assembled a multiethnic database of 21,575 children and adults undergoing microarray genotyping for research—either healthy controls or individuals without kidney-related conditions. “By comparing the DNA of children who had chronic kidney disease to the results from these other individuals, we were able to detect rare events that could be disease-causing,” Gharavi said.

Overall, chromosomal microarrays found diagnostic copy number disorders in 31 of the children—representing 7.4 percent of the study cohort. The CKiD cases also had a high prevalence of large, gene-disrupting autosomal CNVs: 37.7 percent, compared to 23.4 percent of the reference cohort. “These data suggest that potentially up to 14.3 percent of the pediatric CKD cases might be attributable to a CNV of 100 kb or larger,” the researchers wrote.

In an analysis focusing on a list of 131 known genomic disorders, 4.5 percent of the CKiD population had a deletion or duplication that was “clearly diagnostic.” These patients had a deletion or duplication with a known association with a specific syndrome. The rate of known genomic disorders rose to 10.5 percent in children clinically diagnosed with RHD.

Further annotation identified another 12 patients with a “likely pathogenic imbalance,” representing 2.9 percent of the CKiD group. These children had very large, very rare chromosomal abnormalities that were predicted to be pathogenic. “These lesions fulfilled very strict criteria for pathogenicity and would be considered reportable in a clinical setting,” the researchers wrote.

Many of the detected CNVs involved genes thought to be involved in kidney development and thus may be “novel candidate genes” for human kidney disease. Although previously unknown, these abnormalities are considered likely to be disease-causing because of their large size and low frequency in the population and because they involve genes important for normal development.

On adjusted analysis, the odds of known genomic disorders were more than 10 times higher in the CKiD cohort overall, and 30 times higher in those with RHD, as compared to controls. Even after exclusion of known disorders (19 cases), a number of “large, rare gene-disrupting CNVs” were found in the CKiD cohort—including 35 cases with CNVs larger than 500 kb.

Close to one-fourth of patients with known or likely pathogenic copy number disorders also had rare, gene-disrupting second-site CNVs. That was consistent with reported series of patients with developmental delay.

Most baseline clinical and demographic characteristics were similar for children with and without pathogenic CNVs. There were “nominal” differences in estimated glomerular filtration rate and proteinuria, consistent with an impact of the genetic changes on kidney function. These differences will need to be validated in longitudinal studies or independent cohorts.

Major effects on diagnosis and clinical management

“If you can diagnose a patient with a known genomic disorder, that might help the treatment of their kidney disease,” Sampson said. “But it also may allow us to provide additional medical care—whether it’s screening for neurodevelopmental problems, diabetes, or other congenital anomalies. It really provides the opportunity at an early stage, presymptomatically, to provide a genetic diagnosis—which may help to reduce the risk of long-term complications or optimize the care of a patient across their lifespan.”

In the CKiD sample, of eight children clinically diagnosed with cystinosis, three were homozygous for a known deletion of the cystinosin lysosomal cystine transporter gene (CTNS). For this group, CNV testing pinpointed the cause of cystinosis and provided information on the exact mutation for family counseling.

In the remaining 28 patients with a diagnostic copy number disorder, the final genomic diagnosis was clinically unsuspected. In these cases, the CNV findings “either resulted in reclassification of the disease or provided additional information that would have warranted genetic counseling, targeted workup, or surveillance.”

Identical genetic abnormalities were found across different clinical categories. Diagnoses of 1q211.1 recurrent microduplication were made in children with a clinical diagnosis of FSGS, hemolytic uremic syndrome, and chronic glomerulonephritis, while XXX syndrome was diagnosed in patients classified as having recessive polycystic kidney disease, reflux nephropathy, and RHD.

“Our ability to clinically differentiate some causes of kidney disease is probably more limited than we’d like to think,” Gharavi said. “Different kidney disorders can present in the same way or in many different ways that overlap with our classical classification.”

He cited the example of a patient with clinically diagnosed glomerular disorder to discuss the deeper insights offered by genetic diagnosis. “We think of glomerular diseases as something that affects just the kidney, and they’re usually due to an inflammatory or an immune-mediated disease. These types of classifications are important, because if we think somebody has an immune-mediated disease, they may be treated for it by immunosuppressive medication.

“Whereas if they have a developmental disorder, then we know those medications are not going to help and will only result in side effects. So by making a precise diagnosis, we can at least try to come up with the right therapy and the right course of action.”

Making the correct genetic diagnosis is also essential for understanding the long-term clinical course and management. “For example, we found patients who have deletions of a gene called HNF-1 beta, which is diagnostic of a disorder called renal cysts and diabetes syndrome,” Gharavi said. “As the name [implies], the kidneys develop cysts, and there are problems with kidney function. In addition, these individuals are prone to developing diabetes later on in life.”

Children with renal cysts and diabetes syndrome may also have other metabolic disorders, such as low magnesium or high uric acid levels, with a risk of developing gout. “The issue is that many of these complications won’t happen all at once,” Gharavi said. “The kidney cysts are evident earlier on in life, sometimes at birth, [while] the diabetes often occurs around the age of 25. Because these individuals are at risk for diabetes, they should receive targeted health monitoring to make sure that their serum glucose levels are monitored regularly.”

Patients need ongoing lifestyle advice to reduce their risk of diabetes, and should avoid medications that can potentially increase blood glucose, including immunosuppressive therapy with steroids. Other issues may arise later in life—for example, female patients should be advised that they are at risk of uterine abnormalities and problems with conception.

Precision medicine in pediatric CKD

By providing this type of information, CNV testing in children with kidney disease may be a prime example of the NIH’s Precision Medicine Initiative. As President Obama stated when announcing the initiative, precision medicine carries the promise of “delivering the right treatments, at the right time, every time to the right person.”

In the case of pediatric CKD, chromosomal microarrays allow nephrologists to define the exact genetic diagnosis, make a profile for specific complications—in some cases, unrelated to the patient’s kidney dysfunction—and plan clinical care accordingly.

Many of the children in the CKiD cohort had pathogenic genomic imbalances associated with neuropsychiatric disorders, such as autism, schizophrenia, intellectual disability, and seizure disorders, Gharavi noted. “That’s important to be aware of, because we know that children with CKD in general have impaired neurocognitive function and behavioral issues.” Children may have problems at school and at home, or may not meet developmental landmarks.

“And many times that’s been attributed to the sequelae of kidney dysfunction and being chronically ill, being on medications, [and] being in the hospital,” he added. “We attribute this to kidney disease.”

Instead, this group of children has a “fundamental neurodevelopmental disorder,” requiring a different approach. In addition to treatment for kidney disease, the clinical plan needs to consider treatment for neurocognitive issues, educational interventions, and appropriate behavioral therapy.

Behavioral issues can also affect compliance and adherence to treatment for renal dysfunction. “You can choose also your therapy for kidney disease better knowing that maybe some medications will affect neurocognitive function,” Gharavi said. “You can get a much better appreciation of the spectrum of problems that may be going on with that individual and tailor the therapy directly to their problem.”

Of course, much work remains to realize the full impact of precision or personalized medicine for children with CKD. But Gharavi emphasized that DNA microarray studies are clinically available now and are recommended as the first-line diagnostic test for children with intellectual disability, neurocognitive disorders, or major congenital abnormalities.

As these tests come into use for diagnosis of children with kidney abnormalities, Gharavi said the main challenges will be related to test indications and interpretation. While many children will have a clear-cut genetic diagnosis, the situation will be less clear for the significant number of patients with other abnormal findings, including “likely” pathogenic variants. “It’s really difficult to interpret what’s causal, and what’s not, and so you need a lot more studies,” he said. Building a national research cohort of a million or more volunteers is a key component of the Precision Medicine Initiative.

Sampson emphasized the importance of saving patient specimens, linked to clinical data, for future research and analysis. “For all us clinicians who are sending patients with kidney anomalies or CKD for chromosomal microarays, I think there needs to be a way to store that information or store that DNA at the same time,” he said. “[Verbitsky et al] showed that there’s an excess burden of large genomic imbalances in cases versus controls, and we don’t know what those mean.”

Building patient databases will document the growing experience with children who do, or do not have contributing genomic imbalances. With a growing body of saved data, Sampson said, “We can go back to the medical record and then say over time, ‘OK, this is actually a harmless [finding], because we actually see it in quite a few controls.’ Or, ‘Now, we’ve seen 10 patients with this same disorder and we’re starting to make inferences on their long-term care.”

For now, Gharavi said RHD is the main indication for chromosomal microarrays in pediatric nephrology. “We think that children who have congenital kidney malformations are really the ones at highest risk for having chromosomal disorders. So there is pretty good evidence now that [DNA microarrays] should be applied to this subset of individuals.”

“And then for the rest of the children with CKD, I think we need to expand the study and see what is the impact,” he added. The question to be answered is, “Does it make a difference to make [a genetic] diagnosis in the care of these patients?”

Sampson agreed with the recommendation to test children with kidney malformations. “With the caveat that [testing] has to be done in conjunction with the appropriate specialist who can interpret the results. Any patient who [is] sent for microarray, there should be a plan in place to also send that patient to a genetic counselor or geneticist for evaluation. Being able to properly counsel patients in terms of their genomic disorder, in terms of their risk for developing future problems or the problems they already have is nuanced and really needs help from experts.”