ABSTRACT
Objectives
To evaluate the accuracy of routinely available parameters in screening for GCK maturity-onset diabetes of the young (MODY), leveraging data from two large cohorts – one of patients with GCK-MODY and the other of patients with type 1 diabetes (T1D).
Materials and methods
The study included 2,687 patients with T1D, 202 patients with clinical features of MODY but without associated genetic variants (NoVar), and 100 patients with GCK-MODY (GCK). Area under the receiver-operating characteristic curve (ROC-AUC) analyses were used to assess the performance of each parameter – both alone and incorporated into regression models – in discriminating between groups.
Results
The best parameter discriminating between GCK-MODY and T1D was a multivariable model comprising glycated hemoglobin (HbA1c), fasting plasma glucose, age at diagnosis, hypertension, microvascular complications, previous diabetic ketoacidosis, and family history of diabetes. This model had a ROC-AUC value of 0.980 (95% confidence interval [CI] 0.974-0.985) and positive (PPV) and negative (NPV) predictive values of 43.74% and 100%, respectively. The best model discriminating between GCK and NoVar included HbA1c, age at diagnosis, hypertension, and triglycerides and had a ROC-AUC value of 0.850 (95% CI 0.783-0.916), PPV of 88.36%, and NPV of 97.7%; however, this model was not significantly different from the others. A novel GCK variant was also described in one individual with MODY (7-44192948-T-C, p.Ser54Gly), which showed evidence of pathogenicity on in silico prediction tools.
Conclusions
This study identified a highly accurate (98%) composite model for differentiating GCK-MODY and T1D. This model may help clinicians select patients for genetic evaluation of monogenic diabetes, enabling them to implement correct treatment without overusing limited resources.
MODY; type 1 diabetes; glucokinase; diabetes mellitus; genetic techniques
INTRODUCTION
Maturity-onset diabetes of the young (MODY) encompasses various types of monogenic diabetes that account for 1%-5% of all diabetes cases (11. Kavvoura FK, Owen KR. Maturity onset diabetes of the young: clinical characteristics, diagnosis and management. Pediatr Endocrinol Rev. 2012;10(2):234-42.). It is characterized by autosomal dominant inheritance, young-onset hyperglycemia, evidence of residual pancreatic function, and absence of beta-cell autoimmunity or insulin resistance (22. McDonald TJ, Colclough K, Brown R, Shields B, Shepherd M, Bingley P, et al. Islet autoantibodies can discriminate maturity-onset diabetes of the young (MODY) from Type 1 diabetes. Diabet Med. 2011;28(9):1028-33. doi: 10.1111/j.1464-5491.2011.03287.x.
https://doi.org/10.1111/j.1464-5491.2011...
). Variants associated with MODY have been reported in at least 11 genes to date (33. Bonnefond A, Unnikrishnan R, Doria A, Vaxillaire M, Kulkarni RN, Mohan V, et al. Monogenic diabetes. Nat Rev Dis Prim. 2023;9(1):12. doi: 10.1038/s41572-023-00421-w.
https://doi.org/10.1038/s41572-023-00421...
,44. Laver TW, Wakeling MN, Knox O, Colclough K, Wright CF, Ellard S, et al. Evaluation of Evidence for Pathogenicity Demonstrates That BLK, KLF11, and PAX4 Should Not Be Included in Diagnostic Testing for MODY. Diabetes. 2022;71(5):1128-36. doi: 10.2337/db21-0844.
https://doi.org/10.2337/db21-0844...
).
Variants in glucokinase (GCK) and hepatocyte nuclear factor (HNF1A/4A) genes are the most common causes of MODY. Patients with GCK-MODY have mild, nonprogressive fasting hyperglycemia from birth, are typically asymptomatic, and their diagnosis is often incidental (55. Froguel P, Zouali H, Vionnet N, Velho G, Vaxillaire M, Sun F, et al. Familial hyperglycemia due to mutations in glucokinase. Definition of a subtype of diabetes mellitus. N Engl J Med. 1993;328(10):697-702. doi: 10.1056/NEJM199303113281005.
https://doi.org/10.1056/NEJM199303113281...
).
Despite its distinctive clinical features, GCK-MODY is often mistaken by clinicians for other common types of diabetes. On the one hand, patients with GCK-MODY share characteristics with those with type 2 diabetes, including insulin independence and absence of pancreatic autoimmunity. On the other hand, they resemble patients with type 1 diabetes (T1D) regarding age at diagnosis and lack of clinical insulin resistance. As a result, patients with GCK-MODY are often inadvertently treated with insulin or oral antidiabetic agents in routine clinical care, especially when diagnostic tools such as measurement of pancreatic islet antibodies and C-peptide levels are unavailable. Thus, MODY is frequently mislabeled, and it is estimated that more than 80% of patients potentially living with MODY remain without a correct molecular diagnosis (66. Shields BM, Hicks S, Shepherd MH, Colclough K, Hattersley AT, Ellard S. Maturity-onset diabetes of the young (MODY): how many cases are we missing? Diabetologia. 2010;53(12):2504-8. doi: 10.1007/s00125-010-1799-4.
https://doi.org/10.1007/s00125-010-1799-...
).
In this context, two different precision medicine approaches to MODY should be considered: precision treatment and precision diagnosis (77. Chung WK, Erion K, Florez JC, Hattersley AT, Hivert MF, Lee CG, et al. Precision Medicine in Diabetes: A Consensus Report from the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD). Diabetes Care. 2020;43(7):1617-35. doi: 10.2337/dci20-0022.
https://doi.org/10.2337/dci20-0022...
). Precision treatment is a prototypical case of success in this field. Once a molecular diagnosis has been ascertained, individuals with GCK-MODY can have unnecessary treatments discontinued (88. Stride A, Shields B, Gill-Carey O, Chakera AJ, Colclough K, Ellard S, et al. Cross-sectional and longitudinal studies suggest pharmacological treatment used in patients with glucokinase mutations does not alter glycaemia. Diabetologia. 2014;57(1):54-6. doi: 10.1007/s00125-013-3075-x.
https://doi.org/10.1007/s00125-013-3075-...
). Precision diagnosis, in contrast, still needs some fine-tuning. Despite the more widespread availability of molecular diagnosis, genetic testing is still expensive and, currently, universal testing is probably neither feasible nor cost-effective (99. Naylor RN, John PM, Winn AN, Carmody D, Greeley SAW, Philipson LH, et al. Cost-effectiveness of MODY genetic testing: translating genomic advances into practical health applications. Diabetes Care. 2014;37(1):202-9. doi: 10.2337/dc13-0410.
https://doi.org/10.2337/dc13-0410...
).
A widely used approach to determine the probability of MODY is a calculator tool proposed in a British study by Shields and cols. (1010. Shields BM, McDonald TJ, Ellard S, Campbell MJ, Hyde C, Hattersley AT. The development and validation of a clinical prediction model to determine the probability of MODY in patients with young-onset diabetes. Diabetologia. 2012;55(5):1265-72. doi: 10.1007/s00125-011-2418-8.
https://doi.org/10.1007/s00125-011-2418-...
). Different groups, including those studying a Brazilian population, have sought to validate this tool (1111. Tarantino RM, Abreu GM, Fonseca ACP, Kupfer R, Pereira MFC, Campos Júnior M, et al. MODY probability calculator for GCK and HNF1A screening in a multiethnic background population. Arch Endocrinol Metab. 2020;64(1):17-23. doi: 10.20945/2359-3997000000173.
https://doi.org/10.20945/2359-3997000000...
,1212. Santomauro AC Jr, Magalhães ÁLF, Motta FT, de Santana LS, Franco PC, de Freitas SM, et al. The performance of the MODY calculator in a non-Caucasian, mixed-race population diagnosed with diabetes mellitus before 35 years of age. Diabetol Metab Syndr. 2023;15(1):15. doi: 10.1186/s13098-023-00985-3.
https://doi.org/10.1186/s13098-023-00985...
). However, predictive models employing population-specific data are important for better screening individuals who must undergo molecular diagnosis. Therefore, this study aimed to evaluate the accuracy of routinely available clinical and laboratory parameters in discriminating GCK-MODY from T1D and clinical MODY without associated genetic variants (NoVar), leveraging data from two large Brazilian cohorts – one of patients with GCK-MODY and the other of patients with T1D.
MATERIALS AND METHODS
Patient recruitment
This study analyzed patients from the Brazilian Type 1 Diabetes Study Group (BrazDiab1SG) and the Brazilian Monogenic Diabetes Study Group (BRASMOD). The BrazDiab1SG is an observational, cross-sectional, multicenter study of 3,591 patients with T1D showing typical clinical presentation of overt diabetes (unequivocal clinical features of insulin deficiency or diabetic ketoacidosis at diagnosis) and continuous requirement of insulin since diagnosis. Pancreatic autoantibodies were not systematically measured for diagnosis since they are not widely available in participating study centers. The recruitment strategy of BrazDiab1SG has been detailed elsewhere (1313. Gomes MB, Coral M, Cobas RA, Dib SA, Canani LH, Nery M, et al. Prevalence of adults with type 1 diabetes who meet the goals of care in daily clinical practice: A nationwide multicenter study in Brazil. Diabetes Res Clin Pr. 2012;97(1):63-70. doi: 10.1016/j.diabres.2012.02.008.
https://doi.org/10.1016/j.diabres.2012.0...
). For the present study, we included 2,687 individuals from the BrazDiab1SG study in whom glycated hemoglobin (HbA1c) was assessed using methods certified by the National Glycohemoglobin Standardization Program (NGSP). The BRASMOD is a Brazilian national registry of patients with MODY with continuous and ongoing recruitment. Most patients included in the present study have been included in two previous studies (1414. Giuffrida FMA, Moises RS, Weinert LS, Calliari LE, Manna TD, Dotto RP, et al. Maturity-onset diabetes of the young (MODY) in Brazil: Establishment of a national registry and appraisal of available genetic and clinical data. Diabetes Res Clin Pract. 2017;123:134-42. doi: 10.1016/j.diabres.2016.10.017.
https://doi.org/10.1016/j.diabres.2016.1...
,1515. Franco LF, Szarf G, Dotto RP, Dib SA, Moises RS, Giuffrida FMA, et al. Cardiovascular risk assessment by coronary artery calcium score in subjects with maturity-onset diabetes of the young caused by glucokinase mutations. Diabetes Res Clin Pract. 2021;(176):108867. doi: 10.1016/j.diabres.2021.108867.
https://doi.org/10.1016/j.diabres.2021.1...
), except for 12 cases, which are described here for the first time. All patients provided written consent (ethics committee protocol number CAAE 32784514.4.0000.5505). The methodology of the BRASMOD study has been described in the study’s original publications. Both BrazDiab1SG and BRASMOD were approved by local ethics committees at each participating institution.
The present study included 2,687 patients with T1D (T1D group), 202 patients with clinical features of MODY but without GCK or HNF1A variants by Sanger sequencing (NoVar group), and 100 patients with a genetic diagnosis of GCK-MODY (GCK group).
Clinical and laboratory measurements
The following parameters, which were available for analysis, were obtained from the patients’ medical records: sex, age at diabetes diagnosis, age at recruitment, body mass index (BMI), BMI z-score (calculated using the LMS method) (1616. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ. 2000;320(7244):1240-3. doi: 10.1136/bmj.320.7244.1240.
https://doi.org/10.1136/bmj.320.7244.124...
), systolic blood pressure, diastolic blood pressure, fasting plasma glucose, HbA1c, total cholesterol, high-density lipoprotein (HDL) cholesterol, low-density lipoprotein (LDL) cholesterol, triglycerides, fasting C-peptide, microalbuminuria, creatinine, presence of hypertension, previous episodes of diabetic ketoacidosis, and diabetes-related complications such as retinopathy, nephropathy, and neuropathy. Given the cross-sectional design of both datasets employed, the data were collected at unspecified points in the patients’ disease history.
Genetic analyses
All individuals with GCK-MODY were diagnosed by Sanger sequencing as described in the original studies and were included only if they had either pathogenic (P) or likely pathogenic (LP) variants according to the American College of Medical Genetics and Genomics (ACMG) and Association for Molecular Pathology (AMP) (1717. Richards S, Aziz N, Bale S, Bick D, Das S, Gastier-Foster J, et al. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med. 2015;17(5):405-24. doi: 10.1038/gim.2015.30.
https://doi.org/10.1038/gim.2015.30...
), or variants of uncertain significance (VUS) according to the same criteria but with evidence of pathogenicity based on in silico prediction tools (1818. Adzhubei IA, Schmidt S, Peshkin L, Ramensky VE, Gerasimova A, Bork P, et al. A method and server for predicting damaging missense mutations. Nat Methods. 2010;7(4):248-9. doi: 10.1038/nmeth0410-248.
https://doi.org/10.1038/nmeth0410-248...
,1919. Ioannidis NM, Rothstein JH, Pejaver V, Middha S, McDonnell SK, Baheti S, et al. REVEL: An Ensemble Method for Predicting the Pathogenicity of Rare Missense Variants. Am J Hum Genet. 2016;99(4):877-85. doi: 10.1016/j.ajhg.2016.08.016.
https://doi.org/10.1016/j.ajhg.2016.08.0...
). Although the 202 patients in the NoVar group had previously only been tested for GCK and HNF1A variants, 24 of them took part in another study (2020. Santana LS, Caetano LA, Costa-Riquetto AD, Franco PC, Dotto RP, Reis AF, et al. Targeted sequencing identifies novel variants in common and rare MODY genes. Mol Genet Genomic Med. 2019;7(12):e962. doi: 10.1002/mgg3.962.
https://doi.org/10.1002/mgg3.962...
) where they underwent evaluation with a next-generation sequencing targeted panel of 11 genes (HNF4A, GCK, HNF1A, PDX1, HNF1B, NEUROD1, CEL, INS, ABCC8, KCNJ11, and APPL1). None of the patients had pathogenic or likely pathogenic variants, leading us to infer that these evaluated genes were not a major cause of diabetes in our dataset.
Statistical analysis
Most variables showed a skewed distribution in at least one of the groups and we opted to employ nonparametric methods throughout the univariate analysis. Continuous variables were reported as median (interquartile range) and categorical variables were reported as counts (percentages). Since the only comparisons of interest were between GCK versus T1D and GCK versus NoVar, a global comparison among all three groups was not performed and we proceeded straight to pairwise comparisons, although all analyses have been corrected for three multiple comparisons. Continuous variables were analyzed by Dunn’s test and categorical variables were compared using pairwise Fisher’s test, both with Bonferroni correction. P values were capped at 1.0 whenever the corrected value was higher. All p values displayed were corrected, with p < 0.05 considered significant throughout the analyses. The logistic regression models were devised with either GCK versus T1D or GCK versus NoVar as binary outcomes, as described in the Supplementary Methods.
Receiver-operating characteristic (ROC) curve analysis was carried out first with each of the three most significant predictors from the univariate analysis and then with fitted values from the logistic models, employing the same binary outcomes as multivariable models. Areas under the ROC curve (ROC-AUCs) with 95% confidence intervals (CIs) were computed for each variable, and the optimal cutoff point was estimated using Youden’s J statistic (sensitivity + specificity - 1). Using only raw values from two-by-two tables to calculate predictive values would overestimate prevalence and thus falsely inflate calculations. Therefore, positive predictive values (PPVs) and negative predictive values (NPVs) were calculated using the equations described in the Supplementary Methods, employing prevalence (2121. Casscells W, Schoenberger A, Graboys TB. Interpretation by Physicians of Clinical Laboratory Results. N Engl J Med. 1978;299(18):999-1001. doi: 10.1056/NEJM197811022991808.
https://doi.org/10.1056/NEJM197811022991...
). Since the prevalence of GCK-MODY is not entirely known, the following imputed prevalence values based on existing literature were employed: 0.5%, 1%, and 5% for GCK relative to T1D (2222. Irgens HU, Molnes J, Johansson BB, Ringdal M, Skrivarhaug T, Undlien DE, et al. Prevalence of monogenic diabetes in the population-based Norwegian Childhood Diabetes Registry. Diabetologia. 2013;56(7):1512-9. doi: 10.1007/s00125-013-2916-y.
https://doi.org/10.1007/s00125-013-2916-...
,2323. Mozzillo E, Salzano G, Barbetti F, Maffeis C, Lombardo F, Franzese A, et al. Survey on etiological diagnosis of diabetes in 1244 Italian diabetic children and adolescents: Impact of access to genetic testing. Diabetes Res Clin Pract. 2015;107(3):e15-8. doi: 10.1016/j.diabres.2015.01.003.
https://doi.org/10.1016/j.diabres.2015.0...
) and 10%, 40%, and 70% for GCK relative to NoVar (2424. Thanabalasingham G, Owen KR. Diagnosis and management of maturity onset diabetes of the young (MODY). BMJ. 2011;343:d6044. doi: 10.1136/bmj.d6044.
https://doi.org/10.1136/bmj.d6044...
).
Since all available GCK-MODY cases were included, a replication dataset was not available for external validation of cutoff and predictive values obtained in the analyses. Therefore, we proceeded with internal validation by bootstrapping. In brief, bootstrapping consists of drawing multiple simulated subsamples from a larger dataset and analyzing the distribution of the results from each simulation. The distribution of simulations usually approximates a normal distribution due to the central limit theorem and can, therefore, be used to compute CIs. This strategy was accomplished by the following simulation: drawing a random subsample of 50 individuals with GCK-MODY and 500 individuals with T1D, plotting a ROC curve for each parameter or multivariable model, and calculating ROC-AUC values. After 5,000 simulations, 95% bootstrap CIs were obtained utilizing the percentiles 50, 2.5, and 97.5 of the distribution of simulated ROC-AUC values. The same strategy was repeated to compare the GCK and NoVar groups, which included simulated samples of 50 and 100 individuals, respectively. All statistical analyses were performed using R, version 4.1.0 (R Foundation for Statistical Sciences, Vienna, Austria).
RESULTS
Univariate analysis
The study population’s clinical and laboratory data are shown in Table 1. The GCK group, compared with the T1D group, had a higher age at diagnosis (14 [9-27] years versus 10 [6-15] years, respectively, p = 2.98 x 10-6), higher age at recruitment (28 [10-41] years versus 19 [13-27] years, respectively, p = 0.0065), lower fasting plasma glucose level (118 [111-128] mg/dL versus 155 [100-248] mg/dL, p = 2.63 x 10-5), lower HbA1c level (6.3 [6.1-6.6]% versus 8.9 [7.7-10.7]%, p = 3.95 x 10-32), and lower prevalence rates of retinopathy (1% versus 10.1%, respectively, p = 0.0024), nephropathy (0% versus 11.6%, respectively, p = 3.56 x 10-5) and neuropathy (0% versus 5%, respectively, p = 0.043). A previous episode of diabetes ketoacidosis was not recorded in any of the patients with GCK-MODY but occurred in 161 (6.0%) of those with T1D (p = 0.0114).
The GCK group, compared with the NoVar group, had lower age at diagnosis (14 [9-27] years versus 25 [18-35] years, respectively, p = 9.52 x 10-7), lower age at recruitment (28 [10-41] years versus 32 [21-49] years, respectively, p = 0.0035), lower BMI z-score (0.3 [-0.3-1] versus 0.8 [0.2-1.4], p = 0.0063), and lower HbA1c (6.3 [6.1-6.6]% versus 7.0 [5.8-9.2]%, p = 3.18 x10-5) and triglyceride (68 [50-104] mg/dL versus 104 [72-170] mg/dL, respectively, p = 0.0001) levels.
Genetic analyses
Previously unpublished genetic data on 12 patients from the GCK group are described in this report. One novel variant was found in one of these patients: a T>C nucleotide substitution at position chr7:44,192,948, resulting in a p.Ser54Gly amino acid substitution. This variant was deemed a VUS according to the ACMG guidelines; however, it was considered possibly damaging according to in silico prediction by PolyPhen-2 (score 0.886), showed a REVEL score of 0.591 (denoting approximately 68% sensitivity, 92% specificity, and 9.0 likelihood ratio for being a true positive pathogenic variant), and was not found in gnomAD. The GCK variants are detailed in Table 2.
Multivariable models
Logistic regression models with GCK versus T1D as the binary outcome are described in Supplementary Table 1. Only HbA1c and age at diagnosis were independent predictors of GCK in all models. Model 2 had high p values due to categorical variables forced into the model, but demonstrated the best fit of all models. It had the lowest (268.65) Akaike Information Criterion (AIC; lower AIC values indicate a better model fit) and the highest R2 (0.71, i.e., explaining 71% of the variability in the outcome).
Models with GCK versus NoVar as the binary outcome are shown in Supplementary Table 2. Age at diagnosis, HbA1c, and triglycerides remained as independent predictors in Models 4, 5, and 6 after adjustment. Model 5 showed the best fit (AIC of 128.29) and explained the highest variability in the outcome (R2 of 0.70).
Receiver-operating characteristic curve analysis and bootstrap internal validation
The ROC curves analyzing the performance of the predictors in discriminating between GCK-MODY and T1D are shown in Figure 1A. Model 2 (which included HbA1c, fasting plasma glucose, age at diagnosis, hypertension, presence of microvascular complications, history of diabetic ketoacidosis, and presence of a first-degree relative with diabetes) stood out as the best model. This was confirmed by the ROC-AUC 95% CIs depicted in the graph in Figure 1C, which shows that the ROC-AUC value for Model 2 (0.980, 95% CI 0.974-0.985) does not overlap with the ROC-ACU values of the other predictors, as the maximum upper boundary of their 95% CIs was 0.954. The 95% CIs obtained by bootstrapping, shown in Figure 1E, follow the same pattern seen in Figure 1C, confirming the findings.
(A) Receiver-operating characteristic (ROC) curves of predictors significantly discriminating between GCK maturity-onset diabetes of the young (GCK-MODY) and type 1 diabetes and (B) between GCK-MODY and clinical maturity-onset diabetes of the young (MODY) without detected variants (NoVar). The dots plotted inside each curve in Panels A and B represent J points (maximum distance from the diagonal line). (C) Area under the ROC curve (ROC-AUC) and 95% confidence intervals (CIs) for each predictor shown in Panels A and (D) B. (E) Bootstrap ROC-AUC and 95% CIs for each predictor shown in Panels A and (F) B. The dots in Panels C to F represent ROC-AUC values and the horizontal colored bars represent their 95% CIs, enabling a visual comparison of significance among the predictors.
Regarding models for discriminating GCK-MODY versus NoVar, none stood out as more accurate than the others. The ROC curves are depicted in Figure 1B. Despite Model 6 showing the highest ROC-AUC value (0.850, 95% CI 0.783-0.916), the 95% CIs of most predictors overlapped (Figure 1D). The same pattern is depicted in Figure 1F, which shows the 95% CIs obtained on bootstrapping.
Predictive value analysis
Table 3 shows the cutoff values of predictors discriminating GCK-MODY versus T1D, with sensitivity and specificity estimated using the J statistic. The optimal cutoff value for Model 2, derived from the regression equation, was a fitted value of 0.043, which yielded a sensitivity of 100% and specificity of 93.23%. This model had an NPV of 100% for all imputed prevalences of GCK-MODY and PPVs that ranged from 6.91% to 43.74%. Since Model 2 had a significantly higher ROC-AUC value compared with the other predictors, it performed better in all settings.
Table 4 shows analyses similar to those in Table 3, but with GCK-MODY versus NoVar as the outcome. Model 4 showed a high specificity (82.67%) at a cutoff value of 0.384 (fitted value), with a PPV ranging from 32.47% to 90.99%, depending on the imputed frequency of GCK-MODY. The highest sensitivity (98.53%) was for HbA1c alone at a cutoff value of 7.65%, which resulted in the highest NPVs of all predictors (92.98-99.64%). Model 6, in turn, showed the highest ROC-AUC value, with PPVs ranging from 26.54% to 88.36% and NPVs from 66.93% to 97.70%, depending on the imputed prevalence of GCK-MODY.
DISCUSSION
We presented herein a robust sample of 100 patients with confirmed variants in the GCK gene, 202 patients with clinical characteristics of MODY but no variants in either GCK or HNF1A, and 2,687 patients with T1D, obtained from data leveraged from two Brazilian multicenter cohorts.
Patients with GCK-MODY were older at diagnosis compared with those with T1D. However, age at diagnosis alone was not an accurate predictor in discriminating between diabetes subtypes. Patients with GCK-MODY have dysglycemia since birth and usually remain without a correct diagnosis for extended periods (2525. Peixoto-Barbosa R, Reis AF, Giuffrida FMA. Update on clinical screening of maturity-onset diabetes of the young (MODY). Diabetol Metab Syndr. 2020;12(1):50. doi: 10.1186/s13098-020-00557-9.
https://doi.org/10.1186/s13098-020-00557...
). Therefore, their higher age at diagnosis reflects this gap between identifying asymptomatic hyperglycemia and the correct determination of the etiology of diabetes. Gloyn and cols. have found that children and adolescents with hyperglycemia are more likely to have MODY than older individuals (>35 years); therefore, early screening of individuals younger than 25 years would facilitate the identification of MODY cases (2626. Gloyn AL, Bunt M van de, Stratton IM, Lonie L, Tucker L, Ellard S, et al. Prevalence of GCK mutations in individuals screened for fasting hyperglycaemia. Diabetologia. 2009;52(1):172-4. doi: 10.1007/s00125-008-1188-4.
https://doi.org/10.1007/s00125-008-1188-...
).
Probands with GCK-MODY may have a family history of clinical type 2 diabetes with no complications, or parents without known diabetes who have mildly increased fasting blood glucose levels (2525. Peixoto-Barbosa R, Reis AF, Giuffrida FMA. Update on clinical screening of maturity-onset diabetes of the young (MODY). Diabetol Metab Syndr. 2020;12(1):50. doi: 10.1186/s13098-020-00557-9.
https://doi.org/10.1186/s13098-020-00557...
). A family history of diabetes is an inclusion criterion in the BRASMOD study, and we considered it relevant to assess the potential role of this predictor in discriminating between GCK-MODY and T1D in the Brazilian population, given the high proportion (32.9%) of patients with T1D who had a first-degree relative with diabetes. A British study found that a family history of diabetes was the strongest predictor discriminating between T1D and MODY (2727. Shields BM, Peters JL, Cooper C, Powell RJ, Knight BA, Hyde C, et al. Identifying clinical criteria to predict Type 1 diabetes, as defined by absolute insulin deficiency: a systematic review protocol. BMJ Open. 2012;2(6):e002309. doi: 10.1136/bmjopen-2012-002309.
https://doi.org/10.1136/bmjopen-2012-002...
). Individuals with at least one parent affected with diabetes had approximately 23 times higher odds of having MODY than those with unaffected parents. Therefore, when GCK-MODY is suspected, assessing the parents’ fasting glucose and HbA1c levels can be a helpful strategy (2828. Hattersley AT, Greeley SAW, Polak M, Rubio-Cabezas O, Njølstad PR, Mlynarski W, et al. ISPAD Clinical Practice Consensus Guidelines 2018: The diagnosis and management of monogenic diabetes in children and adolescents. Pediatr Diabetes. 2018;19 Suppl 27:47-63. doi: 10.1111/pedi.12772.
https://doi.org/10.1111/pedi.12772...
). Although most patients with GCK-MODY have inherited the variant from one of their parents, de novo GCK inactivating variants have been reported in the literature (2929. Hulín J, Škopková M, Valkovicová T, Mikulajová S, Rosolanková M, Papcun P, et al. Clinical implications of the glucokinase impaired function - GCK MODY today. Physiol Res. 2020;69(6):995-1011. doi: 10.33549/physiolres.934487.
https://doi.org/10.33549/physiolres.9344...
). Nevertheless, this rare finding has potentially little implication in the present setting.
Fasting glucose and HbA1c levels were lower in the GCK-MODY group compared with the T1D group. This finding is consistent with the predicted phenotype of patients with GCK-MODY, in whom fasting glucose level varies between 100-153 mg/dL and HbA1c level is usually between 5.6%-7.6%. Steele and cols. found that HbA1c effectively identified individuals with a GCK variant among those with clinical T1D (ROC-AUC value of 0.94) (3030. Steele AM, Wensley KJ, Ellard S, Murphy R, Shepherd M, Colclough K, et al. Use of HbA1c in the Identification of Patients with Hyperglycaemia Caused by a Glucokinase Mutation: Observational Case Control Studies. PLoS One. 2013;8(6):e65326. doi: 10.1371/journal.pone.0065326.
https://doi.org/10.1371/journal.pone.006...
). In our model, HbA1c alone showed comparable discrimination between GCK-MODY and T1D at the threshold of 7.35%, with a ROC-AUC value of 0.93 (95% CI 0.91-0.94), 94.12% sensitivity, 81.35% specificity, and NPV greater than 99%.
The frequency of acute and chronic complications was extremely low among patients with GCK-MODY. In addition, there were no cases of diabetic ketoacidosis in this group, reflecting a preserved pancreatic function. This finding is consistent with other case series (3131. Chambers C, Fouts A, Dong F, Colclough K, Wang Z, Batish SD, et al. Characteristics of maturity onset diabetes of the young in a large diabetes center. Pediatr Diabetes. 2016;17(5):360-7. doi: 10.1111/pedi.12289.
https://doi.org/10.1111/pedi.12289...
,3232. Li X, Ting TH, Sheng H, Liang CL, Shao Y, Jiang M, et al. Genetic and clinical characteristics of Chinese children with Glucokinase-maturity-onset diabetes of the young (GCK-MODY). BMC Pediatr. 2018;18(1):101. doi: 10.1186/s12887-018-1060-8.
https://doi.org/10.1186/s12887-018-1060-...
). Patients with a GCK variant have a low prevalence of microvascular and macrovascular complications despite being exposed to sustained lifelong mild hyperglycemia (1515. Franco LF, Szarf G, Dotto RP, Dib SA, Moises RS, Giuffrida FMA, et al. Cardiovascular risk assessment by coronary artery calcium score in subjects with maturity-onset diabetes of the young caused by glucokinase mutations. Diabetes Res Clin Pract. 2021;(176):108867. doi: 10.1016/j.diabres.2021.108867.
https://doi.org/10.1016/j.diabres.2021.1...
,3333. Steele AM, Shields BM, Wensley KJ, Colclough K, Ellard S, Hattersley AT. Prevalence of vascular complications among patients with glucokinase mutations and prolonged, mild hyperglycemia. JAMA. 2014;311(3):279-86. doi: 10.1001/jama.2013.283980.
https://doi.org/10.1001/jama.2013.283980...
,3434. Gjesing AP, Engelbrechtsen L, Thuesen ACB, Have CT, Hollensted M, Grarup N, et al. 14-fold increased prevalence of rare glucokinase gene variant carriers in unselected Danish patients with newly diagnosed type 2 diabetes. Diabetes Res Clin Pr. 2022;194:110159. doi: 10.1016/j.diabres.2022.110159.
https://doi.org/10.1016/j.diabres.2022.1...
).
No differences in serum lipids were observed between the GCK-MODY and T1D groups. Fendler and cols. demonstrated that triglyceride levels were not helpful in discriminating between GCK-MODY and other types of diabetes. However, HDL cholesterol levels were lower in patients with a GCK variant than in those with T1D (3535. Fendler W, Borowiec M, Antosik K, Szadkowska A, Deja G, Jarosz-Chobot P, et al. HDL cholesterol as a diagnostic tool for clinical differentiation of GCK-MODY from HNF1A-MODY and type 1 diabetes in children and young adults. Clin Endocrinol (Oxf). 2011;75(3):321-7. doi: 10.1111/j.1365-2265.2011.04052.x.
https://doi.org/10.1111/j.1365-2265.2011...
). A Chinese study found that patients with GCK-MODY had lower total cholesterol and LDL cholesterol levels than those with T1D (3636. Fu J, Wang T, Liu J, Wang X, Zhang Q, Li M, et al. Using Clinical Indices to Distinguish MODY2 (GCK Mutation) and MODY3 (HNF1A Mutation) from Type 1 Diabetes in a Young Chinese Population. Diabetes Ther. 2019;10(4):1381-90. doi: 10.1007/s13300-019-0647-x.
https://doi.org/10.1007/s13300-019-0647-...
). These divergent results could be explained by ethnic differences and issues related to sample sizes and therapies.
The most used model to select patients for genetic testing is the one proposed in 2012 by Shields and cols. The model evaluates patients diagnosed with diabetes before the age of 35 years. For MODY, compared with T1D, the most accurate discriminators in their study were lower HbA1c level, having a parent with diabetes, female sex, and older age at diagnosis (2727. Shields BM, Peters JL, Cooper C, Powell RJ, Knight BA, Hyde C, et al. Identifying clinical criteria to predict Type 1 diabetes, as defined by absolute insulin deficiency: a systematic review protocol. BMJ Open. 2012;2(6):e002309. doi: 10.1136/bmjopen-2012-002309.
https://doi.org/10.1136/bmjopen-2012-002...
). Except for sex, the findings of our study are similar to those of the study by Shields and cols. While the current UK standard for MODY testing relies on a PPV > 25%, other studies in non-Caucasian populations have found that higher cutoff values using the MODY probability calculator (above 62.5%) demonstrate high specificity and NPV (1111. Tarantino RM, Abreu GM, Fonseca ACP, Kupfer R, Pereira MFC, Campos Júnior M, et al. MODY probability calculator for GCK and HNF1A screening in a multiethnic background population. Arch Endocrinol Metab. 2020;64(1):17-23. doi: 10.20945/2359-3997000000173.
https://doi.org/10.20945/2359-3997000000...
,3737. Ang SF, Lim SC, Tan CS, Fong JC, Kon WY, Lian JX, et al. A preliminary study to evaluate the strategy of combining clinical criteria and next generation sequencing (NGS) for the identification of monogenic diabetes among multi-ethnic Asians. Diabetes Res Clin Pr. 2016;119:13-22. doi: 10.1016/j.diabres.2016.06.008.
https://doi.org/10.1016/j.diabres.2016.0...
). In a more recent Brazilian study, including a robust sample of patients with MODY, similar values (calculated probability above 60%) were found to be associated with good accuracy in selecting suspected MODY cases for molecular diagnosis (1212. Santomauro AC Jr, Magalhães ÁLF, Motta FT, de Santana LS, Franco PC, de Freitas SM, et al. The performance of the MODY calculator in a non-Caucasian, mixed-race population diagnosed with diabetes mellitus before 35 years of age. Diabetol Metab Syndr. 2023;15(1):15. doi: 10.1186/s13098-023-00985-3.
https://doi.org/10.1186/s13098-023-00985...
).
Our study demonstrated that HbA1c was the best isolated parameter to discriminate between GCK-MODY and T1D. Furthermore, a model composed of HbA1c, fasting glucose, age at diagnosis, hypertension, presence of microvascular complications, history of diabetes in a first-degree relative, and a previous episode of diabetic ketoacidosis had excellent accuracy, resulting in a ROC-AUC value of 0.980 (95% CI 0.974-0.985).
In the analysis of predictors of GCK-MODY versus NoVar, no model emerged as superior to the others. However, the models provided a fair ability to distinguish GCK-MODY from NoVar. Although the ROC-AUC values were lower than those from the GCK-MODY versus T1D analysis, the higher imputed prevalence of GCK relative to NoVar resulted in higher predictive values. This strategy could still be useful in deciding among different sequencing strategies. While beyond the scope of the present study, other biomarkers could be incorporated into future studies to improve the discrimination between these diabetes subtypes (2525. Peixoto-Barbosa R, Reis AF, Giuffrida FMA. Update on clinical screening of maturity-onset diabetes of the young (MODY). Diabetol Metab Syndr. 2020;12(1):50. doi: 10.1186/s13098-020-00557-9.
https://doi.org/10.1186/s13098-020-00557...
).
The correct etiological diagnosis of GCK-MODY has important implications for the patients’ quality of life, treatment, and follow-up and for the allocation of health resources and genetic counseling. A US study found that 49% of patients with GCK-MODY were inadvertently treated and exposed to adverse events related to the use of oral antidiabetics. Although Latinos, African Americans, and Asians have a higher prevalence of diabetes in general, they accounted for only 20.5% of the probands screened in the study and 17.2% of those diagnosed with a GCK variant (3838. Carmody D, Naylor RN, Bell CD, Berry S, Montgomery JT, Tadie EC, et al. GCK-MODY in the US National Monogenic Diabetes Registry: frequently misdiagnosed and unnecessarily treated. Acta Diabetol. 2016;53(5):703-8. doi: 10.1007/s00592-016-0859-8.
https://doi.org/10.1007/s00592-016-0859-...
). This finding reinforces the need for studies including diverse populations.
The clinical phenotype of GCK-MODY is rather homogeneous compared to that of T1D, but many patients with GCK-MODY are misdiagnosed and inadequately treated. Despite advances in gene sequencing technologies, access to molecular testing is still not widely available or financially viable for most patients. A precision diagnosis approach to MODY should consider the prevalence and epidemiology of this type of diabetes, as well as its clinical features and diagnostic tests that should be robust, widely available, easy to perform, cost-effective, and approved by regulatory agencies (77. Chung WK, Erion K, Florez JC, Hattersley AT, Hivert MF, Lee CG, et al. Precision Medicine in Diabetes: A Consensus Report from the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD). Diabetes Care. 2020;43(7):1617-35. doi: 10.2337/dci20-0022.
https://doi.org/10.2337/dci20-0022...
). Therefore, incorporating additional biomarkers along with the clinical and laboratory parameters analyzed in the present study may enhance the selection process for molecular diagnosis, particularly when access is limited. This approach should remain practical and cost-effective.
The sample size and proportion of patients with T1D and GCK-MODY are some of the strengths of the present study. Another strength of the study is the development of potentially helpful screening strategies for diagnosing GCK-MODY using parameters widely available in clinical practice. This approach has enabled us to achieve high accuracy levels and, despite differences among populations, similar strategies could be useful in other groups. Making use of newer statistical tools, such as bootstrap simulation, allowed for internal replication, thus adding to the robustness of the data. However, some limitations of our study must be addressed. First, this was a cross-sectional study that employed random values of HbA1c. Thus, it was not possible to assess the implications of using the patients’ highest HbA1c value during longitudinal follow-up to rule out GCK-MODY. Likewise, other clinical and laboratory measurements were taken at different points in the patients’ disease history, making the cross-sectional design of our study unable to capture the longitudinal effect of therapy on such parameters. Second, only 13.2% of the evaluated patients reached the 7% target of HbA1c level in the BrazDiab1SG study. Although higher HbA1c values in our patients with T1D could potentially overestimate the role of HbA1c in discriminating between GCK-MODY and T1D, other large studies have shown similar HbA1c levels (3939. Nathan DM, Group DR. The Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications Study at 30 Years: Overview. Diabetes Care. 2013;37(1):9-16. doi: 10.2337/dc13-2112.
https://doi.org/10.2337/dc13-2112...
,4040. Foster NC, Beck RW, Miller KM, Clements MA, Rickels MR, DiMeglio LA, et al. State of Type 1 Diabetes Management and Outcomes from the T1D Exchange in 2016-2018. [published correction appears in Diabetes Technol Ther. 2019 Apr;21(4):230]. Diabetes Technol Ther. 2019;21(2):66-72. doi: 10.1089/dia.2018.0384.
https://doi.org/10.1089/dia.2018.0384...
). Therefore, the difference between both groups regarding glycemic control could be seen as reflecting real-world data. Third, our sample was heterogeneous, since MODY patients were recruited using different inclusion criteria in the original studies. However, given the peculiar phenotype of GCK-MODY, we believe the large sample of 100 individuals compensated for any potential clinical heterogeneity. Fourth, the laboratory measurements were performed at different facilities. Still, no batch effect has been observed in MODY and T1D multicenter trials (data not shown). Fifth, we cannot exclude the possibility of MODY in the T1D group, as the inclusion criteria were based solely on clinical parameters, genetic testing was not carried out on all patients, and the diagnosis was not confirmed by pancreatic autoantibodies. However, other studies have also used clinical criteria for diagnosing T1D, reinforcing that, in many centers, the evaluation of pancreatic autoantibodies is not yet readily available on a large scale in clinical practice (3939. Nathan DM, Group DR. The Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications Study at 30 Years: Overview. Diabetes Care. 2013;37(1):9-16. doi: 10.2337/dc13-2112.
https://doi.org/10.2337/dc13-2112...
,4141. Vistisen D, Andersen GS, Hansen CS, Hulman A, Henriksen JE, Bech-Nielsen H, et al. Prediction of First Cardiovascular Disease Event in Type 1 Diabetes Mellitus: The Steno Type 1 Risk Engine. Circulation. 2016;133(11):1058-66. doi: 10.1161/CIRCULATIONAHA.115.018844.
https://doi.org/10.1161/CIRCULATIONAHA.1...
,4242. Beck RW, Tamborlane WV, Bergenstal RM, Miller KM, DuBose SN, Hall CA, et al. The T1D Exchange Clinic Registry. J Clin Endocrinol Metab. 2012;97(12):4383-9. doi: 10.1210/jc.2012-1561.
https://doi.org/10.1210/jc.2012-1561...
). Sixth, our study did not systematically evaluate other forms of MODY through molecular study. However, as described in the Methods section, around 12% of patients in the NoVar group were tested for 10 genes in addition to GCK, and none of them had pathogenic or likely pathogenic variants. Thus, we believe that even the possibility of misclassified patients would not have critically impacted our final model. Last, although we performed internal validation by bootstrapping, all our available GCK-MODY data were analyzed in the study and we were unable to conduct external validation.
In conclusion, our study demonstrated the feasibility of a highly accurate (98%) composite model in differentiating between GCK-MODY and T1D. This model may help clinicians select more appropriately those patients who should undergo genetic evaluation for monogenic diabetes, enabling correct treatment without overusing limited resources. However, an ideal model still needs to be established. There seem to be differences between populations in Brazilian and European studies regarding the accuracy of clinical criteria in GCK-MODY screening. Therefore, developing appropriate models for each population is paramount to optimize the selection of patients for genetic testing, enabling an accurate diagnosis of diabetes.
Acknowledgements
the original research presented in this paper was supported by Fundação de Amparo à Pesquisa do Estado de São Paulo (Fapesp, São Paulo, Brazil; Grant number 2015/05123-9, awarded to AFR).
REFERENCES
-
1Kavvoura FK, Owen KR. Maturity onset diabetes of the young: clinical characteristics, diagnosis and management. Pediatr Endocrinol Rev. 2012;10(2):234-42.
-
2McDonald TJ, Colclough K, Brown R, Shields B, Shepherd M, Bingley P, et al. Islet autoantibodies can discriminate maturity-onset diabetes of the young (MODY) from Type 1 diabetes. Diabet Med. 2011;28(9):1028-33. doi: 10.1111/j.1464-5491.2011.03287.x.
» https://doi.org/10.1111/j.1464-5491.2011.03287.x -
3Bonnefond A, Unnikrishnan R, Doria A, Vaxillaire M, Kulkarni RN, Mohan V, et al. Monogenic diabetes. Nat Rev Dis Prim. 2023;9(1):12. doi: 10.1038/s41572-023-00421-w.
» https://doi.org/10.1038/s41572-023-00421-w -
4Laver TW, Wakeling MN, Knox O, Colclough K, Wright CF, Ellard S, et al. Evaluation of Evidence for Pathogenicity Demonstrates That BLK, KLF11, and PAX4 Should Not Be Included in Diagnostic Testing for MODY. Diabetes. 2022;71(5):1128-36. doi: 10.2337/db21-0844.
» https://doi.org/10.2337/db21-0844 -
5Froguel P, Zouali H, Vionnet N, Velho G, Vaxillaire M, Sun F, et al. Familial hyperglycemia due to mutations in glucokinase. Definition of a subtype of diabetes mellitus. N Engl J Med. 1993;328(10):697-702. doi: 10.1056/NEJM199303113281005.
» https://doi.org/10.1056/NEJM199303113281005 -
6Shields BM, Hicks S, Shepherd MH, Colclough K, Hattersley AT, Ellard S. Maturity-onset diabetes of the young (MODY): how many cases are we missing? Diabetologia. 2010;53(12):2504-8. doi: 10.1007/s00125-010-1799-4.
» https://doi.org/10.1007/s00125-010-1799-4 -
7Chung WK, Erion K, Florez JC, Hattersley AT, Hivert MF, Lee CG, et al. Precision Medicine in Diabetes: A Consensus Report from the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD). Diabetes Care. 2020;43(7):1617-35. doi: 10.2337/dci20-0022.
» https://doi.org/10.2337/dci20-0022 -
8Stride A, Shields B, Gill-Carey O, Chakera AJ, Colclough K, Ellard S, et al. Cross-sectional and longitudinal studies suggest pharmacological treatment used in patients with glucokinase mutations does not alter glycaemia. Diabetologia. 2014;57(1):54-6. doi: 10.1007/s00125-013-3075-x.
» https://doi.org/10.1007/s00125-013-3075-x -
9Naylor RN, John PM, Winn AN, Carmody D, Greeley SAW, Philipson LH, et al. Cost-effectiveness of MODY genetic testing: translating genomic advances into practical health applications. Diabetes Care. 2014;37(1):202-9. doi: 10.2337/dc13-0410.
» https://doi.org/10.2337/dc13-0410 -
10Shields BM, McDonald TJ, Ellard S, Campbell MJ, Hyde C, Hattersley AT. The development and validation of a clinical prediction model to determine the probability of MODY in patients with young-onset diabetes. Diabetologia. 2012;55(5):1265-72. doi: 10.1007/s00125-011-2418-8.
» https://doi.org/10.1007/s00125-011-2418-8 -
11Tarantino RM, Abreu GM, Fonseca ACP, Kupfer R, Pereira MFC, Campos Júnior M, et al. MODY probability calculator for GCK and HNF1A screening in a multiethnic background population. Arch Endocrinol Metab. 2020;64(1):17-23. doi: 10.20945/2359-3997000000173.
» https://doi.org/10.20945/2359-3997000000173 -
12Santomauro AC Jr, Magalhães ÁLF, Motta FT, de Santana LS, Franco PC, de Freitas SM, et al. The performance of the MODY calculator in a non-Caucasian, mixed-race population diagnosed with diabetes mellitus before 35 years of age. Diabetol Metab Syndr. 2023;15(1):15. doi: 10.1186/s13098-023-00985-3.
» https://doi.org/10.1186/s13098-023-00985-3 -
13Gomes MB, Coral M, Cobas RA, Dib SA, Canani LH, Nery M, et al. Prevalence of adults with type 1 diabetes who meet the goals of care in daily clinical practice: A nationwide multicenter study in Brazil. Diabetes Res Clin Pr. 2012;97(1):63-70. doi: 10.1016/j.diabres.2012.02.008.
» https://doi.org/10.1016/j.diabres.2012.02.008 -
14Giuffrida FMA, Moises RS, Weinert LS, Calliari LE, Manna TD, Dotto RP, et al. Maturity-onset diabetes of the young (MODY) in Brazil: Establishment of a national registry and appraisal of available genetic and clinical data. Diabetes Res Clin Pract. 2017;123:134-42. doi: 10.1016/j.diabres.2016.10.017.
» https://doi.org/10.1016/j.diabres.2016.10.017 -
15Franco LF, Szarf G, Dotto RP, Dib SA, Moises RS, Giuffrida FMA, et al. Cardiovascular risk assessment by coronary artery calcium score in subjects with maturity-onset diabetes of the young caused by glucokinase mutations. Diabetes Res Clin Pract. 2021;(176):108867. doi: 10.1016/j.diabres.2021.108867.
» https://doi.org/10.1016/j.diabres.2021.108867 -
16Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ. 2000;320(7244):1240-3. doi: 10.1136/bmj.320.7244.1240.
» https://doi.org/10.1136/bmj.320.7244.1240 -
17Richards S, Aziz N, Bale S, Bick D, Das S, Gastier-Foster J, et al. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med. 2015;17(5):405-24. doi: 10.1038/gim.2015.30.
» https://doi.org/10.1038/gim.2015.30 -
18Adzhubei IA, Schmidt S, Peshkin L, Ramensky VE, Gerasimova A, Bork P, et al. A method and server for predicting damaging missense mutations. Nat Methods. 2010;7(4):248-9. doi: 10.1038/nmeth0410-248.
» https://doi.org/10.1038/nmeth0410-248 -
19Ioannidis NM, Rothstein JH, Pejaver V, Middha S, McDonnell SK, Baheti S, et al. REVEL: An Ensemble Method for Predicting the Pathogenicity of Rare Missense Variants. Am J Hum Genet. 2016;99(4):877-85. doi: 10.1016/j.ajhg.2016.08.016.
» https://doi.org/10.1016/j.ajhg.2016.08.016 -
20Santana LS, Caetano LA, Costa-Riquetto AD, Franco PC, Dotto RP, Reis AF, et al. Targeted sequencing identifies novel variants in common and rare MODY genes. Mol Genet Genomic Med. 2019;7(12):e962. doi: 10.1002/mgg3.962.
» https://doi.org/10.1002/mgg3.962 -
21Casscells W, Schoenberger A, Graboys TB. Interpretation by Physicians of Clinical Laboratory Results. N Engl J Med. 1978;299(18):999-1001. doi: 10.1056/NEJM197811022991808.
» https://doi.org/10.1056/NEJM197811022991808 -
22Irgens HU, Molnes J, Johansson BB, Ringdal M, Skrivarhaug T, Undlien DE, et al. Prevalence of monogenic diabetes in the population-based Norwegian Childhood Diabetes Registry. Diabetologia. 2013;56(7):1512-9. doi: 10.1007/s00125-013-2916-y.
» https://doi.org/10.1007/s00125-013-2916-y -
23Mozzillo E, Salzano G, Barbetti F, Maffeis C, Lombardo F, Franzese A, et al. Survey on etiological diagnosis of diabetes in 1244 Italian diabetic children and adolescents: Impact of access to genetic testing. Diabetes Res Clin Pract. 2015;107(3):e15-8. doi: 10.1016/j.diabres.2015.01.003.
» https://doi.org/10.1016/j.diabres.2015.01.003 -
24Thanabalasingham G, Owen KR. Diagnosis and management of maturity onset diabetes of the young (MODY). BMJ. 2011;343:d6044. doi: 10.1136/bmj.d6044.
» https://doi.org/10.1136/bmj.d6044 -
25Peixoto-Barbosa R, Reis AF, Giuffrida FMA. Update on clinical screening of maturity-onset diabetes of the young (MODY). Diabetol Metab Syndr. 2020;12(1):50. doi: 10.1186/s13098-020-00557-9.
» https://doi.org/10.1186/s13098-020-00557-9 -
26Gloyn AL, Bunt M van de, Stratton IM, Lonie L, Tucker L, Ellard S, et al. Prevalence of GCK mutations in individuals screened for fasting hyperglycaemia. Diabetologia. 2009;52(1):172-4. doi: 10.1007/s00125-008-1188-4.
» https://doi.org/10.1007/s00125-008-1188-4 -
27Shields BM, Peters JL, Cooper C, Powell RJ, Knight BA, Hyde C, et al. Identifying clinical criteria to predict Type 1 diabetes, as defined by absolute insulin deficiency: a systematic review protocol. BMJ Open. 2012;2(6):e002309. doi: 10.1136/bmjopen-2012-002309.
» https://doi.org/10.1136/bmjopen-2012-002309 -
28Hattersley AT, Greeley SAW, Polak M, Rubio-Cabezas O, Njølstad PR, Mlynarski W, et al. ISPAD Clinical Practice Consensus Guidelines 2018: The diagnosis and management of monogenic diabetes in children and adolescents. Pediatr Diabetes. 2018;19 Suppl 27:47-63. doi: 10.1111/pedi.12772.
» https://doi.org/10.1111/pedi.12772 -
29Hulín J, Škopková M, Valkovicová T, Mikulajová S, Rosolanková M, Papcun P, et al. Clinical implications of the glucokinase impaired function - GCK MODY today. Physiol Res. 2020;69(6):995-1011. doi: 10.33549/physiolres.934487.
» https://doi.org/10.33549/physiolres.934487 -
30Steele AM, Wensley KJ, Ellard S, Murphy R, Shepherd M, Colclough K, et al. Use of HbA1c in the Identification of Patients with Hyperglycaemia Caused by a Glucokinase Mutation: Observational Case Control Studies. PLoS One. 2013;8(6):e65326. doi: 10.1371/journal.pone.0065326.
» https://doi.org/10.1371/journal.pone.0065326 -
31Chambers C, Fouts A, Dong F, Colclough K, Wang Z, Batish SD, et al. Characteristics of maturity onset diabetes of the young in a large diabetes center. Pediatr Diabetes. 2016;17(5):360-7. doi: 10.1111/pedi.12289.
» https://doi.org/10.1111/pedi.12289 -
32Li X, Ting TH, Sheng H, Liang CL, Shao Y, Jiang M, et al. Genetic and clinical characteristics of Chinese children with Glucokinase-maturity-onset diabetes of the young (GCK-MODY). BMC Pediatr. 2018;18(1):101. doi: 10.1186/s12887-018-1060-8.
» https://doi.org/10.1186/s12887-018-1060-8 -
33Steele AM, Shields BM, Wensley KJ, Colclough K, Ellard S, Hattersley AT. Prevalence of vascular complications among patients with glucokinase mutations and prolonged, mild hyperglycemia. JAMA. 2014;311(3):279-86. doi: 10.1001/jama.2013.283980.
» https://doi.org/10.1001/jama.2013.283980 -
34Gjesing AP, Engelbrechtsen L, Thuesen ACB, Have CT, Hollensted M, Grarup N, et al. 14-fold increased prevalence of rare glucokinase gene variant carriers in unselected Danish patients with newly diagnosed type 2 diabetes. Diabetes Res Clin Pr. 2022;194:110159. doi: 10.1016/j.diabres.2022.110159.
» https://doi.org/10.1016/j.diabres.2022.110159 -
35Fendler W, Borowiec M, Antosik K, Szadkowska A, Deja G, Jarosz-Chobot P, et al. HDL cholesterol as a diagnostic tool for clinical differentiation of GCK-MODY from HNF1A-MODY and type 1 diabetes in children and young adults. Clin Endocrinol (Oxf). 2011;75(3):321-7. doi: 10.1111/j.1365-2265.2011.04052.x.
» https://doi.org/10.1111/j.1365-2265.2011.04052.x -
36Fu J, Wang T, Liu J, Wang X, Zhang Q, Li M, et al. Using Clinical Indices to Distinguish MODY2 (GCK Mutation) and MODY3 (HNF1A Mutation) from Type 1 Diabetes in a Young Chinese Population. Diabetes Ther. 2019;10(4):1381-90. doi: 10.1007/s13300-019-0647-x.
» https://doi.org/10.1007/s13300-019-0647-x -
37Ang SF, Lim SC, Tan CS, Fong JC, Kon WY, Lian JX, et al. A preliminary study to evaluate the strategy of combining clinical criteria and next generation sequencing (NGS) for the identification of monogenic diabetes among multi-ethnic Asians. Diabetes Res Clin Pr. 2016;119:13-22. doi: 10.1016/j.diabres.2016.06.008.
» https://doi.org/10.1016/j.diabres.2016.06.008 -
38Carmody D, Naylor RN, Bell CD, Berry S, Montgomery JT, Tadie EC, et al. GCK-MODY in the US National Monogenic Diabetes Registry: frequently misdiagnosed and unnecessarily treated. Acta Diabetol. 2016;53(5):703-8. doi: 10.1007/s00592-016-0859-8.
» https://doi.org/10.1007/s00592-016-0859-8 -
39Nathan DM, Group DR. The Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications Study at 30 Years: Overview. Diabetes Care. 2013;37(1):9-16. doi: 10.2337/dc13-2112.
» https://doi.org/10.2337/dc13-2112 -
40Foster NC, Beck RW, Miller KM, Clements MA, Rickels MR, DiMeglio LA, et al. State of Type 1 Diabetes Management and Outcomes from the T1D Exchange in 2016-2018. [published correction appears in Diabetes Technol Ther. 2019 Apr;21(4):230]. Diabetes Technol Ther. 2019;21(2):66-72. doi: 10.1089/dia.2018.0384.
» https://doi.org/10.1089/dia.2018.0384 -
41Vistisen D, Andersen GS, Hansen CS, Hulman A, Henriksen JE, Bech-Nielsen H, et al. Prediction of First Cardiovascular Disease Event in Type 1 Diabetes Mellitus: The Steno Type 1 Risk Engine. Circulation. 2016;133(11):1058-66. doi: 10.1161/CIRCULATIONAHA.115.018844.
» https://doi.org/10.1161/CIRCULATIONAHA.115.018844 -
42Beck RW, Tamborlane WV, Bergenstal RM, Miller KM, DuBose SN, Hall CA, et al. The T1D Exchange Clinic Registry. J Clin Endocrinol Metab. 2012;97(12):4383-9. doi: 10.1210/jc.2012-1561.
» https://doi.org/10.1210/jc.2012-1561 -
43Thomson KL, Gloyn AL, Colclough K, Batten M, Allen LIS, Beards F, et al. Identification of 21 novel glucokinase (GCK) mutations in UK and European Caucasians with maturity-onset diabetes of the young (MODY). Hum Mutat. 2003;22(5):417. doi: 10.1002/humu.9186.
» https://doi.org/10.1002/humu.9186 -
44Alkorta-Aranburu G, Carmody D, Cheng YW, Nelakuditi V, Ma L, Dickens JT, et al. Phenotypic heterogeneity in monogenic diabetes: The clinical and diagnostic utility of a gene panel-based next-generation sequencing approach. Mol Genet Metab. 2014;113(4):315-20. doi: 10.1016/j.ymgme.2014.09.007.
» https://doi.org/10.1016/j.ymgme.2014.09.007 -
45Osbak KK, Colclough K, Saint-Martin C, Beer NL, Bellanné-Chantelot C, Ellard S, et al. Update on mutations in glucokinase (GCK), which cause maturity-onset diabetes of the young, permanent neonatal diabetes, and hyperinsulinemic hypoglycemia. Hum Mutat. 2009;30(11):1512-26. doi: 10.1002/humu.21110.
» https://doi.org/10.1002/humu.21110
-
Brazilian Monogenic Diabetes Study Group (BRASMOD), Brazilian Type 1 Diabetes Study Group (BrazDiab1SG)
Publication Dates
-
Publication in this collection
16 Aug 2024 -
Date of issue
2024
History
-
Received
9 Aug 2023 -
Accepted
3 Mar 2024