Table 3:

Recommendations for reporting results of monogenic diabetes testing across a range of different testing and reporting scenarios.

Reporting scenario	Headline Result/Summary finding	Implications for patient	Implications for family members	Additional comments	References
Pathogenic or likely pathogenic GCK variant	Consistent with/Confirms a genetic diagnosis of monogenic diabetes, subtype GCK	hyperglycaemia is not altered by treatment. No increased risk of micro and macrovascular complications. May discontinue treatment and not require follow up/monitoring/screening. Implications for management of a GCK MODY pregnancy; no risk of macrosomia for a newborn that inherits the variant and insulin therefore not required. Non-invasive prenatal testing may be possible.	50% risk to offspring. Testing affected relatives is possible to confirm diagnosis. Unaffected relatives requesting testing should have FBG and HbA1c measured first, followed by genetic testing. Unaffected females planning pregnancy are advised to have testing to aid management.	Parents of infants with NDM due to a biallelic GCK variant that are heterozygous for the variant should be reported as carriers of GCK NDM in addition to the genetic diagnosis of GCK diabetes (particularly in consanguineous families). Identifying a pathogenic GCK variant in a patient with a more severe hyperglycaemia not consistent with GCK is suggestive of another aetiology such as type 1 DM, and guidance for clinical management so take this into account.	85
Pathogenic or likely pathogenic HNF1A variant	Consistent with/Confirms a genetic diagnosis of monogenic diabetes, subtype HNF1A	Patients with HNF1A diabetes respond well to sulphonylurea treatment.	50% risk to offspring. Testing affected relatives is possible to confirm diagnosis. Predictive testing should be offered with appropriate counselling regarding penetrance, especially for young children (rarely diagnosed <10 years).	HNF1A variants are also associated with hepatic adenomas and rarely with CHI. The penetrance of these conditions is low and there is no requirement to advise on risk to patient or family members. Their risk can be included in the report at the discretion of the testing laboratory.	101,72,102,103
Pathogenic or likely pathogenic HNF4A variant	Consistent with/Confirms a genetic diagnosis of monogenic diabetes, subtype HNF4A	Patients with HNF4A diabetes respond well to sulphonylurea treatment. Risk of macrosomia and CHI for a newborn that inherits the variant from either parent; non-invasive prenatal testing may be available	50% risk to offspring. Testing affected relatives is possible to confirm diagnosis. Predictive testing should be offered with appropriate counselling regarding penetrance, especially for young children (rarely diagnosed <10 years).	The HNF4A p.Arg114Trp variant is associated with a significantly reduced penetrance for diabetes and does not increase risk of macrosomia. Predictive testing for this variant should take this into consideration. The p.Arg63Trp variant causes a Fanconi renal tubulopathy and renal assessment is recommended for these patients.50% risk to offspring of inheriting variant and developing a tubulopathy.	3, 91, 104
Pathogenic or likely pathogenic HNF1B variant	Consistent with/Confirms a genetic diagnosis of monogenic diabetes, subtype HNF1B or RCAD Consistent with/Confirms a genetic diagnosis of HNF1B-related renal disease (for a family member with renal disease but no diabetes).	If diabetes only, an assessment for HNF1B related disease (renal, urogenital, hepatic and pancreatic exocrine) should be recommended. If a whole gene deletion is identified, implications for possible neuropsychiatric problems should be stated. If the diagnosis is made in a relative with renal disease but no diabetes, recommendations for routine HbA1c screening should be given.	50% risk to offspring of developing diabetes and HNF1B related conditions. Testing affected relatives is possible to confirm diagnosis. Predictive testing should be offered with appropriate counselling regarding penetrance, especially for young children (rarely diagnosed <10 years). Whole gene deletions frequently occur de novo and both parents should be tested to confirm this.	If a pathogenic HNF1B variant is identified unexpectedly in a patient with isolated diabetes, a simple request for additional clinical information may reveal undisclosed renal disease which will enable the reporting of the syndrome.	70, 105
Mitochondrial DNA variant m.3243A>G	Confirms a genetic diagnosis of mitochondrial diabetes or MIDD due to m.3243A>G variant. Inconclusive result; further testing required if heteroplasmy level is 1-5%.	If diabetes only, recommend audiological assessment. Variant is associated with other clinical conditions but cannot predict disease progression due to heteroplasmy in different tissues. Reproductive options may be mentioned with possible referral to a specialist mitochondrial disease clinic if available.	All offspring of affected females are at risk of inheriting the variant. Males will not pass on the variant. Diagnostic testing offered to affected relatives. Predictive testing should be offered to at-risk offspring, siblings and maternal relatives through a clinical genetics service.	Heteroplasmy level should be provided if measured. For low level heteroplasmy in peripheral blood (1-5%), confirmatory testing from a urine epithelial sample should be advised.	106, 107, 108
Pathogenic activating variant in ABCC8 or KCNJ11 in a patient with suspected MODY	Consistent with/Confirms a genetic diagnosis of monogenic diabetes, subtype ABCC8/KCNJ11	Patients with activating variants in ABCC8 and KCNJ11 respond well to sulphonylurea treatment.	50% risk to offspring of developing neonatal diabetes and/or diabetes in early adulthood. Testing affected relatives is possible to confirm diagnosis. Predictive testing for TNDM causing variants outside of infancy should be offered with appropriate counselling, especially for young children since the penetrance is variable..	Patient has a variant that causes remitting TNDM with relapsing diabetes in adulthood. In some patients, the TNDM is non-penetrant and the diabetes presents clinically as MODY and is referred for MODY testing.	109, 110, 111
Pathogenic activating variant in ABCC8 or KCNJ11 in a patient with NDM	Consistent with/Confirms a genetic diagnosis of permanent or transient neonatal diabetes, subtype ABCC8/KCNJ11.	Patients with activating variants in ABCC8 and KCNJ11 respond well to sulphonylurea treatment. Specific variants in KCNJ11 cause DEND syndrome requiring higher doses of sulphonylurea. For patients with TNDM, the remitting and relapsing disease course should be described. Non-invasive prenatal testing may be available since early treatment with sulphonylurea (including trans-placental if mother has NDM) improves outcomes.	50% risk to offspring of developing neonatal diabetes, or neonatal diabetes and/or diabetes in early adulthood if TNDM is diagnosed. Testing affected relatives is possible to confirm diagnosis. Predictive testing for a TNDM variant in older children and adults is possible with appropriate counselling regarding penetrance.	Sulphonylurea transfer protocols are available at https://www.diabetesgenes.org/about-neonatal-diabetes/. Rarely, variants may be recessively acting and should be reported as for a recessively inherited condition. A recessive variant may cause neonatal diabetes if occurring in trans with a pathogenic inactivating (loss of function) variant.	110, 109, 6, 112, 113, 114
Pathogenic dominant inactivating variant in ABCC8 or KCNJ11 in a patient with suspected MODY	Genetic diagnosis of monogenic diabetes due to a heterozygous inactivating ABCC8 or KCNJ11 variant	Dominantly acting inactivating variants cause CHI and/or adult-onset monogenic diabetes. In contrast in activating variants, these inactivating variants are not associated with increased sensitivity to sulphonylureas.	50% risk to offspring of developing CHI and/or adult -onset diabetes. Testing relatives with CHI or diabetes is possible. Predictive testing is not recommended given the significantly reduced penetrance for diabetes due to these inactivating variants.	The report must clearly distinguish these inactivating variants from the more common activating variants causing diabetes, since the treatment protocols are different. Patients with dominant inactivating variants may respond to other OHAs similar to slim patients with type 2 DM.	115
Pathogenic variant in a syndromic monogenic diabetes gene	Consistent with/Confirms a genetic diagnosis of a monogenic diabetes syndrome. If the patient does not have characteristic features of the syndrome, report as gene X-related monogenic diabetes.	Assessment for clinical features of the syndrome is recommended through referral to a disease specialist if available or through a clinical genetics service.	Testing affected relatives is possible to confirm diagnosis. Carrier testing offered to parents of patients with recessively inherited disorders.	If the diagnosis is unexpected, the laboratory should contact the clinician before reporting to establish if the patient has an undisclosed clinical diagnosis of the syndrome. Patients heterozygous for a pathogenic variant in a recessively inherited disorder will be (with very rare exceptions) a carrier and do not have monogenic diabetes. This is an incidental finding and should not be reported but consider reporting if there is consanguinity in the family.	4, 116, 34
Variant of uncertain significance (general)	Inconclusive result; further investigations required.	Report should clearly state that the result does not confirm a genetic diagnosis of monogenic diabetes. Further clinical investigations should be requested as appropriate. The result cannot be used in isolation for clinical decision making, although a trial change in therapy may be possible.	Testing of other affected relatives to determine whether variant segregates with diabetes. Knowledge of clinical phenotype of the relatives tested is essential. The result should not be used for predictive testing.	Class 3 (VUS or VOUS) variants should be reported based on professional judgement, the level of supporting evidence, possible impact on clinical management and on whether additional investigations can be undertaken to change the classification. Consider not reporting the variant if there is very weak supporting evidence (a ‘cold’ VUS) or variant is in a syndromic diabetes gene with no clinical suspicion of a syndrome.	65
Variant of uncertain significance in GCK	Inconclusive result; further investigations required.	The result should not be used in isolation for clinical decision making. Further biochemical testing (FBG, OGTT and HbA1c) to confirm GCK MODY phenotype. A trial off treatment could be performed if safe to do so, with no change in HbA1c consistent with GCK MODY.	Segregation of the variant with fasting hyperglycaemia in the family should be investigated. Testing can be offered to clinically affected and unaffected relatives. All genetically tested relatives must undergo FBG and HbA1c testing.	Functional studies are likely to be beyond the scope of work offered by a diagnostic laboratory and should only be suggested if the laboratory can provide this. RNA studies on variants predicted to affect splicing are possible using mini-gene assays and these may be offered by the laboratory.	117, 118
Variant of uncertain significance in HNF1A or HNF4A	Inconclusive result; further investigations required.	The result should not be used in isolation for clinical decision making. A trial on sulphonylurea could be performed if appropriate with a good response or hypoglycaemia on low doses consistent with HNF1A/HNF4A diabetes. Details of birth weights and history of CHI in proband and relatives should be requested.	Segregation of the variant with diabetes in the family should be investigated. Clinical details of tested relatives should be collected to assess whether they have a MODY phenotype since phenocopies with type 1 and 2 DM can be present in families. Predictive testing should not be offered.	Functional studies are likely to be beyond the scope of work offered by a diagnostic laboratory and should only be suggested if the laboratory can provide this. RNA studies on variants predicted to affect splicing are possible using mini-gene assays and these may be offered by the laboratory.	119
Variant of uncertain significance in ABCC8 or KCNJ11	Inconclusive result; further investigations required.	Ask if there is a history of NDM or CHI in the proband and their family. If strong suspicion of a PNDM or TNDM variant, consider a trial on sulphonylurea with a good response or hypoglycaemia on low doses consistent with a genetic diagnosis.	Parental testing for a heterozygous variant in an NDM proband is recommended to check for de novo occurrence. Segregation of the variant with diabetes in the family should be investigated. Clinical details of tested relatives should be collected to assess whether they have a MODY phenotype since phenocopies with type 1 and 2 DM can be present in families. Predictive testing should not be offered.	Careful consideration needed as to whether the variant should be reported given complexity of interpretation. The variant will typically fall into one of four categories: Pathogenic TNDM variant with non-penetrance of NDM. Pathogenic variant causing dominant CHI with non-penetrance of CHI. Pathogenic recessively acting CHI variant and patient is a carrier of CHI. Benign variant of no clinical significance. Heterozygous null variants in these genes (nonsense, frameshift, splicing) should not be reported (unless there is consanguinity in the family) since they do not cause diabetes and patients are carriers of recessive CHI.	120
No variant detected (NGS)	A genetic diagnosis of monogenic diabetes has not been confirmed	Additional specific testing may be offered/advised e.g. CNV testing, MS-MLPA for 6q24 related TNDM or genotyping for m.3243A>G	N/A.	Gene panels should analyse the common causes of monogenic diabetes and perform CNV analysis. The report should list all genes analysed, horizontal coverage and the types of variants that can be identified. Limitations should be stated and will include genes not sequenced by the NGS assay, not detecting CNVs, imprinting abnormalities, non-coding variants and mtDNA variants, and regions of low sequencing coverage. Sanger in-fill may be used for regions with low coverage. Detailed technical and bioinformatics information may be added as an appendix but not in the main body of the report.	44, 65
No variant detected in gene X (Sanger)	Does not confirm a genetic diagnosis of monogenic diabetes, subtype gene X	Result does not exclude a diagnosis since pathogenic variants in other genes can cause monogenic diabetes. Either reflex to additional testing (e.g. NGS), offer/advise additional genetic testing as appropriate or advise that no further genetic testing is indicated.	N/A.	Report should state the limitations of the method used, i.e. that other monogenic diabetes genes have not been sequenced, but may also state that CNVs or variants in non-coding regions cannot be detected.	121, 122
m.3243A>G variant not detected (genotyping on blood DNA)	Does not confirm a genetic diagnosis of mitochondrial diabetes or MIDD due to m.3243A>G variant.	Does not exclude a diagnosis since the variant may be present at very low heteroplasmy levels or present in other tissues such as muscle or urine epithelia. Urine testing should be offered for not detected tests on peripheral blood DNA in patients with clinical suspicion of MIDD.	N/A.	Report should state the method used and the heteroplasmy limit of detection.	106, 108
Predictive testing	Genetically predisposed to monogenic diabetes, subtype X Not genetically predisposed to monogenic diabetes, subtype X	If variant is not detected, state that the family member’s risk of developing diabetes is reduced to that of the population. If the variant is detected, then state that periodic monitoring for diabetes by HbA1c is recommended. If diabetes penetrance is age-related (e.g. HNF1A) and family member is a young child, then consider delaying monitoring until >10 years.	If variant is detected, state risk to offspring of inheriting the variant. Offer testing to affected relatives to confirm diagnosis. Predictive testing should be offered with appropriate counselling regarding penetrance, especially for young children (rarely diagnosed <10 years).	If predictive testing is undertaken for both parents of a proband, and the variant is not detected in either parent, the variant is likely to have arisen de novo in the proband. Germline mosaicism is a possibility; The parents’ reports should state the lowest level of mosaicism that the assay could theoretically detect (e.g 10% for Sanger), and state that the risk of having another affected child is low but risk due to germline mosaicism cannot be excluded.	72, 123
Relative with diabetes and familial variant not detected	Diagnosis of monogenic diabetes, subtype gene X has not been confirmed	Considerthe clinical features of the relative. If not suggestive of monogenic diabetes, then state that a diagnosis has not been confirmed and that the diabetes is likely to have a different aetiology. If features are consistent with monogenic diabetes and result is unexpected, request repeat samples for genetic testing using a different method (e.g. NGS or Sanger with alternative primers).	N/A	Sanger sequencing is the most likely assay used for familial variant testing but is prone to false negative results due to sequence variation under primer binding sites that may prevent the variant allele being sequenced. If the not detected result is unexpected, the lab should re-test using new primer sequences that have been checked for common variants in binding sites. NGS testing is another option since it is not prone to this issue and also tests for all other causes of monogenic diabetes, but is more expensive.	44, 64

NDM - Neonatal Diabetes Mellitus; CHI - Congenital Hyperinsulinism; RCAD - Renal Cysts And Diabetes syndrome; MIDD - Maternally Inherited Diabetes and Deafness syndrome; TNDM- Transient Neonatal Diabetes Mellitus; VUS or VOUS - Variant Of Uncertain Significance; PNDM - Permanent Neonatal Diabetes Mellitus; CNV - Copy Number Variant; MS-MLPA - Methylation-Specific Multiplex Ligation-Dependent Probe Amplification; NGS - Next Generation Sequencing.