4.1 IRD-ID diagnostic algorithm

We present an algorithm to aid clinicians in evaluating and classifying individuals presenting with both an IRD and ID (Figure 2). When assessing a patient with IRD-ID, the presence or absence of additional systemic features such as microcephaly, neurologic abnormalities, renal disease, or skeletal differences will help the clinician narrow the differential to a group of phenotypically related disorders. In doing so, the clinician may focus on elements of the history, physical examination, and investigations that may be most relevant in confirming or excluding a diagnosis. Furthermore, adopting a systematic approach to a patient with IRD-ID may allow for more targeted, gene-specific, or panel testing. This strategy reduces the likelihood of incidental findings, is less expensive, and provides a higher coverage (number of sequence reads) compared to broader, hypothesis-independent testing such as exome or genome sequencing (Sun et al., 2015). On the other hand, it is recognized that for some patients or in some jurisdictions, broader testing such as exome sequencing may be more appropriate as an early investigation, in which case this framework may also help the laboratory and clinician interpret patient variants.

The following sections will highlight the various components of this diagnostic algorithm and present an overview of the diverse spectrum of genetic syndromes currently associated with IRD-ID. The rapid pace of gene discovery in recent years has led to several emerging candidate genes for IRD-ID disorders being described. We have included these genes for consideration as well, indicating where this is the case. Additionally, we have included a small number of syndromic IRD disorders in which ID is not a classic feature but may occasionally occur; for example, Alström syndrome.

4.2 IRD-ID syndromes with microcephaly

For an individual with IRD-ID and microcephaly, a handful of diagnoses may be considered. A classic recognizable disorder, Cohen syndrome, fits this clinical presentation. Additionally, several microtubular and centriolar disorders appear within this group. Microtubules and centrioles play important roles in cell proliferation, migration, and polarity (Kuijpers & Hoogenraad, 2011; Nigg & Raff, 2009). Impairment in their formation, structure, or function leads to abnormalities in brain development including microcephaly (Naveed et al., 2018). Microtubules and centrioles also form the building blocks of cilia (Nigg & Raff, 2009). Retinal photoreceptor cells have modified primary cilia which are important in facilitating protein trafficking between the inner and outer segments (Wheway, Parry, & Johnson, 2014).

In differentiating between the disorders in this group, consider the onset, trajectory, and severity of the microcephaly. The microcephaly seen in Cohen syndrome has its onset postnatally and is variable in severity (H. Wang et al., 1993). The same applies to ACBD5- and ACO2-related disorders, both of which are characterized by progressive, neurodegenerative clinical courses (Ferdinandusse et al., 2017; Metodiev et al., 2014). On the other hand, the microtubular and centriolar disorders within this category tend to manifest with primary or congenital microcephaly which can often be quite pronounced (Naveed et al., 2018). Finally, it is important to distinguish between absolute and relative microcephaly, an example of the latter being PLK4-related disorder in which microcephaly is present in the context of primordial dwarfism (Martin et al., 2014). Table 2 outlines the features of the IRD-ID syndromes associated with microcephaly.

4.3 IRD-ID syndromes with neurologic or structural brain abnormalities

For an individual with IRD-ID and associated neurologic or structural brain anomalies, one neuroradiologic finding is particularly important. The molar tooth sign, formed by the combination of a deep interpeduncular fossa, thickened and horizontal superior cerebellar peduncles, and cerebellar vermis hypoplasia, is characteristic of Joubert syndrome, a prototypic ciliopathy (Maria, 1999). Over 35 genes have now been associated with Joubert syndrome, though some cases are more likely to present with IRD-ID than others (Bachmann-Gagescu et al., 2020). In the absence of a molar tooth sign, there are many additional disorders on the differential for a patient presenting with IRD-ID and associated neurologic or structural brain abnormalities (Table 3).

4.4 IRD-ID syndromes with obesity

Although not uncommon in the general population, obesity in an individual with IRD-ID may suggest certain syndromes on the differential such as Bardet–Biedl syndrome. Bardet–Biedl syndrome is a genetically heterogeneous ciliopathy associated with at least 21 different genes (Forsythe et al., 2018). If Bardet–Biedl syndrome is not deemed to be a fitting diagnosis, both Alström syndrome and Cohen syndrome where childhood-onset central obesity is common may be considered as alternate possibilities in the presence of obesity (Alvarez-Satta et al., 2015; H. Wang et al., 1993). However, distinguishing features early on in their presentation help clinically separate the diagnosis of Bardet–Biedl syndrome with nyctalopia (rod-cone dystrophy) from Alström syndrome with photophobia (cone-rod dystrophy). Table 4 highlights all three of these disorders, along with a few others to consider in the individual with IRD-ID and obesity.

4.5 IRD-ID syndromes with renal involvement

Renal involvement is common in several IRD-ID syndromes. Nephronophthisis and cystic renal disease, in particular, are characteristic features shared by many ciliopathies (Gascue, Katsanis, & Badano, 2011). Symptoms of renal disease in this setting can include polyuria and polydipsia. Individuals may also develop failure to thrive, anemia, and hypertension as signs of renal dysfunction, with renal hyperechogenicity or cystic disease evident on ultrasound (Tory et al., 2009). When evaluating an individual with IRD-ID and associated nephronophthisis, Joubert syndrome should be considered as a leading possibility (Bachmann-Gagescu et al., 2020). Bardet–Biedl syndrome can present with nephronophthisis as well, though unlike Joubert syndrome, it can also be associated with numerous other structural or functional renal anomalies (Forsythe et al., 2018). Finally, both retinal and renal involvement may not necessarily be simultaneously apparent on initial presentation. For example, among the diseases associated with nephronophthisis, individuals with Senior–Löken syndrome often develop a retinopathy in childhood before manifesting overt signs of renal disease several years later (Hemachandar, 2014). Early recognition of syndromic diseases may lead to earlier medical intervention in these patients; this supports a broad, multigene approach to genetic testing to identify these cases. The features of the IRD-ID syndromes associated with renal involvement are shown in Table 5.

4.6 IRD-ID syndromes with short stature or skeletal abnormalities

Several IRD-ID syndromes are associated with short stature as part of their syndromic presentation, whereas others present with specific skeletal anomalies (Table 6). Determining the age of onset of short stature can be helpful in differentiating some of these conditions. For example, in an individual with IRD-ID and short stature of prenatal onset, PLK4 which is associated with primordial dwarfism is a likely candidate gene (Martin et al., 2014). Patients with Alström syndrome, on the other hand, typically demonstrate normal growth in childhood but have a reduced final adult height due to deficiencies in growth hormone and insulin-like growth factor 1 (Romani et al., 2013). In an individual presenting with IRD-ID and thoracic dysplasia, IFT81 and IFT140 should be considered (Perrault et al., 2012; Perrault et al., 2015). Polydactyly can be seen in patients with pathogenic variants in these intraflagellar transport proteins, and can be an early presenting feature of several other ciliopathies including Bardet–Biedl and Joubert syndromes. Other, more subtle skeletal differences such as brachydactyly are reported in IRD-ID syndromes as well. Finally, short stature may be seen in patients with other medical comorbidities, particularly if not completely treated, such as renal (Table 5), endocrine (Table 7), or metabolic disorders.

4.7 IRD-ID syndromes with endocrine dysfunction

There are a handful of IRD-ID syndromes in which a higher incidence of various endocrine abnormalities can be observed (Table 7). These endocrine findings may be a defining feature of certain disorders, such as Boucher–Neuhauser syndrome where hypogonadism is frequent (Synofzik et al., 2014). The allelic disorder, Oliver–McFarlane syndrome, can present as early as infancy with features of congenital hypothyroidism, which untreated leads to intellectual disability and poor growth (Hufnagel et al., 2015). Depending on the particular testing method implemented, this central hypothyroidism may or may not be detected on a newborn screen. In other circumstances, a diagnosis of other IRD-ID syndromes with associated endocrine complications should prompt screening for these associated co-morbidities. Many of these endocrine features are readily treatable and when treated may lead to considerable improvement in quality of life.

4.8 IRD-ID syndromes with hearing loss

Hearing loss, generally bilateral, is a feature of many of the IRD-ID syndromes. Depending on how these individuals initially present, the combination of retinal dystrophy and hearing loss may lead to a presumptive diagnosis of Usher syndrome (Yan & Liu, 2010). However, ID is not typically seen in Usher syndrome and should prompt consideration of the disorders in Table 8. Sensorineural hearing loss (SNHL) tends to be the type most often associated with the IRD-ID syndromes. Profound congenital SNHL, as in type 1 Usher syndrome, can be seen in patients with PRPS1-associated Arts syndrome (de Brouwer et al., 2010). Progressive SNHL is frequent in Alström syndrome (Alvarez-Satta et al., 2015), or may represent EXOSC2- or RCBTB1-related disorders (Coppieters et al., 2016; Di Donato et al., 2016). Finally, although infrequent, progressive SNHL with onset in the first or second decade of life (Akizu et al., 2014; Singh, Shrinkhal, Pradhan, & Chakraborty, 2017) and conductive hearing loss secondary to middle ear infections (Beales et al., 1999; Kroes et al., 2010) may be observed in Bardet-Biedl and Joubert syndromes.

4.9 IRD-ID syndromes with hematologic or immunologic abnormalities

A few of the IRD-ID syndromes, highlighted in Table 9, present with associated hematologic or immunologic abnormalities that may have important clinical consequences. Noncyclic neutropenia is a feature of Cohen syndrome along with recurrent infections and aphthous ulcers; although the neutropenia is typically mild to moderate, granulocyte colony-stimulating factor therapy is sometimes administered (H. Wang et al., 1993). B-cell immunodeficiency in an individual with IRD-ID suggests a TRNT1-related disorder, in which periodic fevers and congenital sideroblastic anemia may also be present (Chakraborty et al., 2014). Arts syndrome may cause immune impairment as well, particularly with severe respiratory infections (de Brouwer et al., 2010). Although it should be noted that any of the IRD-ID syndromes with renal involvement may manifest with anemia secondary to renal dysfunction; these syndromes are not included within this section.

4.10 IRD-ID in inborn errors of metabolism

Both IRD and ID are observed in several inborn errors of metabolism (IEMs; Poll-The & Maillette de Buy Wenniger-Prick, 2011; Rajappa, Goyal, & Kaur, 2010). Although a detailed review of metabolic disorders that can manifest with IRD-ID lies outside the scope of this review, the clinician should consider this category in the diagnostic workup of a patient.

IRD and ID can result from peroxisomal disorders, several mitochondrial disorders, lysosomal storage disorders such as the mucopolysaccharidoses and neuronal ceroid lipofuscinoses, congenital disorders of glycosylation, neurodegeneration with brain iron accumulation, abetalipoproteinemia, Menkes disease, cobalamin C deficiency, and ornithine aminotransferase deficiency (Poll-The & Maillette de Buy Wenniger-Prick, 2011; Rajappa et al., 2010; Werdich et al., 2014). These disorders tend not to present with IRD-ID alone, as the biochemical pathways involved are implicated in multiple cellular processes. In general, additional features that should raise suspicion for these IEMs include developmental regression, failure to thrive, a progressive neurodegenerative course, multisystem involvement, or clinical deterioration in times of increased metabolic demand (Saudubray, Berghe, & Walter, 2012). Some disorders manifest with suggestive signs, for example, coarse facial features and corneal clouding in some mucopolysaccharidoses, or kinky and brittle hair in Menkes disease. Others, such as mitochondrial disorders and congenital disorders of glycosylation, can present in any organ system at any age (Saudubray et al., 2012). Prompt diagnosis may in some cases lead to the early implementation of treatment options (Rajappa et al., 2010). Of note, some of these IEMs do also have milder forms which may present with IRD without neurodevelopmental features, for example, PEX1- and PEX6-related Heimler syndrome and certain variants in HGSNAT (Haer-Wigman et al., 2015; Ratbi et al., 2015).

Newborn screening, which varies by region, does not detect the majority of the aforementioned IEMs associated with IRD-ID (Therrell et al., 2015). An exception is mucopolysaccharidosis type I, which has recently been added to the Recommended Uniform Screening Panel (RUSP) in the United States, although the implementation remains variable (RUSP, 2018). When suspecting an IEM in an individual with IRD-ID, the Treatable Intellectual Disability Endeavor (TIDE) protocol is one resource which may guide biochemical testing (van Karnebeek, Shevell, Zschocke, Moeschler, & Stockler, 2014). Specifically, useful initial investigations for the metabolic IRD-ID disorders include plasma amino acids, total homocysteine, copper and ceruloplasmin, urine organic acids, and urine glycosaminoglycans. These investigations can assist in the workup of mucopolysaccaridoses, Menkes disease, cobalamin C deficiency, and ornithine aminotransferase deficiency. Additionally, the clinician can consider very long-chain fatty acids if a peroxisomal disorder is suspected, transferrin isoelectric focusing in the case of a congenital disorder of glycosylation, and a lipid profile for possible abetalipoproteinemia. Magnetic resonance imaging of the brain can be a useful adjunct as well, particularly if considering mitochondrial or peroxisomal disorders, neuronal ceroid lipofuscinosis, or neurodegeneration with brain iron accumulation. Broad molecular testing such as exome sequencing, which is increasingly being applied earlier in the diagnostic workup, can of course also be quite helpful in confirming the diagnosis of an IEM.

4.11 IRD-ID as a dual diagnosis

Finally, despite Occam's razor, at times a single unifying genetic diagnosis may not explain an individual's entire phenotype. In such cases, the clinician must keep in mind the possibility of a dual diagnosis (Balci et al., 2017; Posey et al., 2017). This is especially relevant when it comes to the IRD-ID syndromes, as both IRD and ID are independently common. Likely some older reports in the literature of single instances of ID associated with established retinal disorder genes (or vice versa) may have resulted from separate molecular genetic disorders. Finally, a significant proportion of the genes and syndromes reviewed above have a wide phenotypic spectrum and may also cause nonsyndromic ophthalmologic manifestations or systemic disease without ocular features. Thus, even in a scenario where an individual is found to have a pathogenic result in one of the genes highlighted within this review, careful consideration of the variant in question and the individual's clinical phenotype remains invaluable in establishing an accurate diagnosis.

4.12 Emerging evidence for IRD-ID disorders

Although we have provided a summary of the existing evidence for genes associated with IRD-ID phenotypes, the list of IRD-ID disorders will likely continue to grow. We conclude our review by sharing a clinical encounter that highlights evidence for the nonrandom association of IRD with ID.

4.13 KIZ (MIM# 615757)

We examined a 23-year-old male Sudanese refugee who presented with GDD, short stature, kyphoscoliosis, presumed hearing loss (awaiting formal testing), high myopia (−20D), and a retinal dystrophy. A retinal dystrophy panel consisting of 266 genes (Blueprint Genetics, Helsinki, Finland) identified two heterozygous variants in KIZ (MIM# 615757): c.1612 + 2 T > G, a consensus splice donor variant; c.709C > A, p.(Pro237Thr), a missense variant. This panel essentially ruled out a collagenopathy, including Kniest dysplasia. Neither variant in KIZ has been reported before in association with human disease. Using the American College of Medical Genetics and Genomics (ACMG) classification guidelines, the consensus splice donor variant was classified as likely pathogenic and the missense variant as a variant of uncertain significance (VUS). Unfortunately, neither parent was available for segregation analysis.

The KIZ gene encodes a centrosomal protein kizuna located in the basal body of cilia. Mutations in KIZ have been shown to cause rod-cone dystrophy; however, there have been no reports of a syndromic presentation (El Shamieh et al., 2014, 2017; Gustafson et al., 2017). Given the exceptionally rare occurrence of the above variants individually in control databases combined with the known association of KIZ with retinopathy, we believe that these changes explain our patient's ocular phenotype. An alternate genetic cause, however, cannot be entirely excluded. As this disorder is a ciliopathy, it is plausible that mutations in KIZ can also affect other commonly involved systems, providing an explanation for our patient's developmental delay and vertebral anomalies. However, the possibility of a dual diagnosis remains. Examining parents and analyzing the pedigree may help clinicians clarify both single and dual genetic diagnoses and nongenetic contributions to disease. Future functional studies that examine these variants in a model organism, as well as further replication in other clinic patients, may also help to clarify this observation.