In this issue of Diabetes, Colclough et al. (1) report interesting data suggesting that the current genetic testing strategy for monogenic diabetes should be modified, extending it by default to genes known to cause syndromic forms of the disease.
Section 1.01 Monogenic diabetes presents with a broad and heterogeneous clinical spectrum due to mutations in a single gene (2,3) and affects 1–5% of patients with diabetes (2,4). To date, several distinct subtypes of monogenic diabetes have been identified (2), including neonatal diabetes (diagnosed within the first 6 months of life), syndromic diabetes (very rare forms of hyperglycemia occurring in combination with extrapancreatic abnormalities, which include but are not limited to urogenital morphological abnormalities, hearing, and vision loss), and maturity-onset diabetes of the young (MODY), which is, by far, the most frequent. MODY follows an autosomal dominant inheritance pattern (5) and has been related to mutations in 14 different genes (OMIM [Online Mendelian Inheritance in Man] project no. 606391). Accordingly, outside of very early infancy, the current genetic testing procedure for monogenic diabetes only targets the 14 MODY genes, while the genes responsible for syndromic diabetes, hitherto considered distinct from MODY, are tested solely in the presence of suggestive clinical features (6).
By using targeted next-generation sequencing (tNGS), the U.K. group (1) investigated the role of 27 monogenic diabetes genes, including those known to cause syndromic diabetes, in 1,280 unrelated individuals who were referred for MODY testing with no clear syndromic clinical signs. Molecular evidence of monogenic diabetes was found in only 297 patients (i.e., 23% of the entire sample). The vast majority of them (n = 241, 81% of those with positive genetic testing) carried mutations in nonsyndromic diabetes (i.e., MODY) genes, with the most common, namely, HNF1A, GCK, and HNF4A, playing the main role (4,7). Conversely and quite unexpectedly, 56 patients (19% of all patients with monogenic diabetes) carried mutations in syndromic diabetes genes, including mitochondrial DNA (n = 24, 8%, all m.3243A>G) and HNF1B (n = 18, 6%) as well as WFS1 (n = 6), INSR (n = 4), GATA6 (n = 1), SLC29A3 (n = 1), TRMT10A (n = 1), and PPARG (n = 1). Understandably, the authors conclude that syndromic monogenic diabetes (particularly due to m.3243A>G and HNF1B mutations) should be tested in all patients with suspected MODY, regardless of the presence of typical syndromic features. Of note, in this issue of Diabetes, Saint-Martin et al. report very similar results in 1,676 patients who were referred for MODY testing with no clear evidence of syndromic diabetes (8). By using tNGS comprising 18 monogenic diabetes genes, mutations were found in only 307 patients. Of these, 18% carried mutations in syndromic diabetes genes, mostly being in the mitochondrial DNA (all m.3243A>G), in HNF1B, and in WFS1. Similar to what was reported by Colclough et al. (1), patients with syndromic diabetes gene mutations were also more often treated with insulin and less likely to have a family history of diabetes than their counterparts (8).
The data from both the U.K. and the French studies are very similar, reinforcing each other and allowing us to comment on some aspects in a solid way.
First, despite the use of tNGS comprising many genes, the pick-up rate for monogenic diabetes testing was only about 20% (1,8). This may partly be a consequence of our largely incomplete knowledge of the genetic background of monogenic diabetes (i.e., we probably have not yet discovered many genes that cause monogenic diabetes), but it is also likely a consequence of a suboptimal patient selection by clinicians somehow unable to differentiate monogenic diabetes from type 1 and type 2 diabetes (4). The topic of monogenic diabetes (e.g., higher than expected prevalence, differential diagnosis, etc.) therefore should be better taught to medical students and treated more thoroughly during continuing medical education courses.
Second, ∼20% of mutations found in patients referred to as affected by nonsyndromic monogenic diabetes (i.e., MODY) are, in fact, in syndromic diabetes genes. This finding challenges the assumption that nonsyndromic and syndromic diabetes are totally distinct clinical entities. Rather, we should accept that, as a likely consequence of the variable penetrance (4) and expressivity of mutations in syndromic diabetes genes (9–12), this form of diabetes has very blurred boundaries that contribute to making the monogenic diabetes spectrum heterogeneous (13). On the other hand, the relevant proportion of mutations in the genes of syndromic diabetes also depends on about 10–20% of the suboptimal clinical work preceding the genetic testing (e.g., lack of reporting deafness in m.3243A>G mutation carriers and their relatives) (1,8), recalling once again the need for better training in the scope of monogenic diabetes.
In conclusion, the study from Colclough et al. (1) as well as that from Saint-Martin et al. (8) should be taken very seriously in terms of modifying the current clinical and genetic testing strategy of monogenic diabetes. First, before referring to genetic testing, clinicians should do their best not to overlook important personal and family history, physical features, and biochemical measures, including C-peptide levels and type 1 diabetes antibodies (14), which are all very useful for the diagnosis of monogenic diabetes, including syndromic forms. Second, investigating genes of syndromic diabetes in all patients with monogenic diabetes (i.e., regardless of the presence of syndromic clinical features) 1) helps identify some individuals whose mild syndromic presentation can been missed, 2) urges clinicians to retrospectively consider screening for the presence of additional features using nonroutine medical approaches (renal ultrasonography, hearing test, and ophthalmological evaluation), and 3) provides the correct diagnosis for patients and their relatives who truly lack syndromic features but who may develop them in the future.
Finally, these two studies (1,8) clearly suggest that the current classification of monogenic diabetes into syndromic and nonsyndromic forms is too rigid and does not correspond to clinical reality. It may be time to review this classification, keeping the mind open to new evidence.
Article Information
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
