Genetic Testing for the Child With Short Stature—Has the Time Come To Change Our Diagnostic Paradigm?

Why is my child short? Is there something wrong? These are the questions that parents of children with short stature often ask their child's pediatrician and pediatric endocrinologist. Traditionally, children with short stature have undergone extensive medical evaluation, often focused on identifying hormonal etiologies of short stature or other chronic medical conditions that affect growth. Studies have shown that the standard medical evaluation rarely leads to a diagnosis in an otherwise healthy child with isolated short stature (1).

Over the last decade, we have gained a tremendous amount of insight into the underlying genetic factors that influence one’s height. Genome-wide association studies have taught us that literally thousands of genes play a role in determining an individual's height (2). In most cases, common genetic variants, each of which may have millimeter effect on height, in aggregate are the major determinants of one’s height. On the other end of the spectrum, there are individuals with rare genetic mutations that lead to severe growth disorders such as achondroplasia, Laron syndrome, and many others. We are now finding that there are many individuals who have genetic variants with more moderate effects; there is a range of short stature among this group. These individuals would previously have been classified as having idiopathic short stature (ISS) or perhaps a mild skeletal dysplasia. In the largest study to date, Hauer et al. (3) studied 565 individuals with unexplained short stature. Two hundred of these children underwent whole-exome sequencing, from which a genetic cause could be identified in 21% of syndromic cases and in approximately 14% of those with isolated short stature. By exome sequencing, many causes of short stature were identified that were missed on routine clinical evaluation, most which were milder presentations of known genetic growth disorders.

Using a similar approach, in a recent issue of Journal of Clinical Endocrinology & Metabolism, Freire et al. (4) performed next-generation sequencing studies of 55 children born small for gestational age (SGA) with isolated short stature. Approximately 85% of children who are born SGA will achieve spontaneous catch-up growth and reach a normal height (5, 6). The majority of these children have in utero environmental factors that limited their prenatal growth. However, the 15% of children who do not achieve adequate catch-up growth present a clinical conundrum for physicians. Some of these children may have syndromic features or other associated medical comorbidities, whereas other cases likely represent milder forms of genetic growth disorders with nonspecific clinical features similar to the ISS population. Similar to ISS, routine laboratory evaluation is often nondiagnostic in SGA cases. Not surprisingly, Freire et al. (4) identified a likely genetic etiology in eight of 55 patients (15%), a similar yield to that of the Hauer et al. study (3).

Several interesting themes emerge when looking at these two studies together as well as a number of similar, preceding smaller studies of genetic testing in growth disorders. First, most pathologic mutations found in these studies are in genes associated with growth plate function or known skeletal dysplasias. This is in contrast to very few mutations being found in classic endocrine hormonal pathways and suggests that pediatric endocrinologists need to focus more attention on the growth plate as the primary site of growth disturbances (7). Second, an overwhelming majority of these mutations are heterozygous and were inherited from an affected parent. In the Freire et al. paper (4), 69% of the studied patients had familial short stature (4). In all cases where parental DNA was available (five of eight cases), the identified mutations were inherited from an affected parent. Therefore, when one is evaluating a patient for short stature or for being SGA, it is critically important to measure the parents’ heights and consider whether there is possibly a dominantly inherited disorder in the family. Third, there is no single gene that stands out as the predominant cause of ISS or SGA. Rather, there is likely a host of causative genes, each of which contributes 1% to 2% or an even smaller percentage to the etiology of patients previously classified as ISS or SGA. This is not a surprising finding given the results of the genome-wide association studies that have shown the multitude of genes involved in growth. In specific subgroups with particular clinical features, the individual gene contributions may be higher. Fourth, the yield of genetic testing far exceeds that of traditional hormonal testing in this patient population. This raises the question of when should genetic testing become the first-line evaluation for children with growth disorders who are otherwise healthy and do not manifest any other clinical signs of a hormonal deficiency.

To answer that question, well-designed studies are needed that compare the cost-effectiveness of standard medical evaluation vs genetic testing for the evaluation of short stature. However, before embarking on these studies, we must first grapple with the question as to how we will define a causative mutation in a child with short stature. The American College of Medical Genetics and Genomics and the Association for Molecular Pathology have proposed criteria for the interpretation of the potential pathogenicity of genetic variants (8). However, when applied in clinical practice to the setting of short stature, there are many variants that will be classified as having uncertain significance or being likely pathogenic but lack definitive evidence for causality. For example, in the NPR2 gene, a purported cause of 1% to 2% of ISS, there are 454 loss-of-function or missense variants listed in the gnomAD database (9). Of these, 258 of the variants are only found in a single individual. Many of these will be predicted to be deleterious by in silico computer prediction algorithms. Familial segregation data can be tremendously beneficial in interpretation of these variants but are also fraught with difficulty. First, DNA is often not available from all individuals in the family. Second, we know there is considerable variable expressivity in many of these height genes. For example, if a child whose height is −2.5 SD of the mean carries a variant found in a parent whose height is −1.5 SD of the mean, should we consider this evidence for or against segregation with the phenotype? With limited ability to interpret segregation data, it will be difficult to determine whether genetic variants found through routine clinical sequencing are truly pathogenic or simply innocent bystanders.

In addition, the notion that a single genetic variant is actually the cause of an individual’s short stature is overly simplistic. It is more accurate to say that the additive effects of thousands of common variants with small effect sizes, as well as perhaps a handful of rare variants with more prominent effects, together add up to determine an individual’s stature. In most cases, a single variant or two recessive variants in a single gene are not the sole determinants of an individual’s height. Rather, genetic testing is identifying a variant that is contributing to the individual’s short stature, but it is often difficult to say to what degree. Our current knowledge is extremely limited with regard to interactions between multiple variants in the same or different growth-related pathways. In addition, almost nothing is known about gene-environment interactions with regard to short stature, especially on the level of individual variants identified on genetic testing.

Despite all these limitations, I do think the time has come to change the paradigm of how we approach the diagnosis of short stature. For patients who are otherwise healthy and have ISS or idiopathic SGA, especially those with a clear family history of short stature and more substantial short stature (perhaps a height below −3 SD of the mean), I believe genetic testing should be the primary diagnostic modality. As noted, there will be difficulty in interpreting many of these variants, but these difficulties are not unique to this setting. Geneticists, genetic counselors, and molecular diagnostic laboratories have expertise in interpreting these tests and conveying the meaning of the results to patients. Endocrinologists will either need to develop this expertise or partner with these other disciplines. In some cases, the interpretation will be quite clear, and it will be easy to determine causality. In others, the results may raise more questions than provide answers, but over time, additional research or clinical experience may further inform results interpretation. In my opinion, the time has come for our field to grapple with these questions in the clinical realm as opposed to waiting for a time when research studies have identified and interpreted every potential stature related variant—a time that will clearly never come.

The approach to genetic testing in short stature is complex and there is not a single agreed-upon diagnostic algorithm. Factors to be considered include whether the patient has proportionate or disproportionate short stature, the latter being more commonly associated with skeletal dysplasia genes. Additional factors to consider are whether short stature began pre- or postnatally or whether there are associated syndromic features or microcephaly. A detailed discussion of these factors is beyond the scope of this article, but I refer the reader to two review articles that discuss many of these issues in detail (10, 11). The exact testing strategy will depend on the clinical scenario but may include single gene testing, gene panel testing, chromosomal microarray, or whole-exome sequencing. The superiority of one genetic testing strategy for short stature over others has not yet been proven.

If one is able to identify a genetic variant that is likely to be making a major contribution to the patient’s short stature, the next questions become what can and should we do therapeutically. The answer to those questions is quite complex and requires much more research into the natural history of individual genetic subclasses of short stature as well as expected responses to currently available and novel therapeutic agents. However, we can only begin to answer those questions once we remove the word idiopathic from the diagnosis of short stature.

Abbreviations:

Abbreviations:
 
• ISS

  idiopathic short stature

 
• SGA

  small for gestational age

Acknowledgments

Disclosure Summary: A.D. has received consulting fees from Novo Nordisk, Pfizer, Ipsen, Sandoz, and OPKO Biologics as well as research support from Ipsen and Novo Nordisk.

References