Temporal deep learning framework for retinopathy prediction in patients with type 1 diabetes.

The adoption of electronic health records in hospitals has ensured the availability of large datasets that can be used to predict medical complications. The trajectories of patients in real-world settings are highly variable, making longitudinal data modeling challenging. In recent years, significant progress has been made in the study of deep learning models applied to time series; however, the application of these models to irregular medical time series (IMTS) remains limited. To address this issue, we developed a generic deep-learning-based framework for modeling IMTS that facilitates the comparative studies of sequential neural networks (transformers and long short-term memory) and irregular time representation techniques. A validation study to predict retinopathy complications was conducted on 1207 patients with type 1 diabetes in a French database using their historical glycosylated hemoglobin measurements, without any data aggregation or imputation. The transformer-based model combined with the soft one-hot representation of time gaps achieved the highest score: an area under the receiver operating characteristic curve of 88.65%, specificity of 85.56%, sensitivity of 83.33% and an improvement of 11.7% over the same architecture without time information. This is the first attempt to predict retinopathy complications in patients with type 1 diabetes using deep learning and longitudinal data collected from patient visits. This study highlighted the significance of modeling time gaps between medical records to improve prediction performance and the utility of a generic framework for conducting extensive comparative studies.

Association Between the Tissue and Circulating Advanced Glycation End-Products and the Micro- and Macrovascular Complications in Type 1 Diabetes: The DIABAGE Study.

INTRODUCTION: Type 1 diabetes is associated with an increased risk of vascular complications. We aimed to investigate the association between serum and tissue advanced glycation end-products (AGEs) and micro- and macrovascular complications in type 1 diabetes (T1D). METHODS: We conducted a cross-sectional study on 196 adults with T1D (mean age 44.53 ± 16, mean duration of diabetes 22 ± 12 years, mean HbA1c 8 ± 1.2%). AGEs were measured in blood serum (i.e., carboxymethyllysine (CML), methylglyoxal-hydroimidazolone-1 (MGH1), and pentosidine) and by measurement of skin autofluorescence (SAF). Associations between AGEs levels and vascular complications were analyzed using binary logistic regression. Correlations between AGEs and pulse wave velocity (PWV) were also assessed by linear regressions. Significant differences were set for p values less than 0.05. RESULTS: We found positive associations between different AGEs and vascular complications. SAF was associated with both microangiopathy (retinopathy: OR = 1.92, p = 0.011; neuropathy: OR = 2.02, p = 0.04; any microangiopathy: OR = 2.83, p < 0.0001) and macroangiopathy (coronaropathy: OR = 3.11, p = 0.009; any macroangiopathy: OR = 2.78, p = 0.003). For circulating AGEs, pentosidine was significantly associated with coronaropathy (OR = 1.61, p = 0.01) and any macroangiopathy (OR = 1.52, p = 0.005) while MGH1 was associated with nephropathy (OR 1.72, p = 0.03). Furthermore, a significant linear correlation was found between PWV and SAF (r = 0.43, p < 0.001), pentosidine (r = 0.28, p < 0.001), and MGH1 (r = 0.16, p = 0.031), but not for CML (r = 0.03, p = 0.598). CONCLUSIONS: Skin autofluorescence appears to be a useful marker for investigating both micro- and macrovascular complications in T1D. In this study, pentosidine was associated with macroangiopathy and MGH1 with nephropathy among the circulating AGEs. Furthermore, the correlations between PWV and AGEs may suggest their value in early prediction of vascular complications in T1D.

Characteristics of newly diagnosed type 1 diabetes in paediatric and adult population from Reims University Hospital, France from 1997 to 2019.

French health insurance data showed that the incidence of type 1 diabetes mellitus (T1DM) in children increased over the years to 2015. The objective of our study was to assess the evolution of the number of incident cases of paediatric and adult type 1 diabetes in our institution, and to describe their clinical presentation and its evolution. All patients with T1DM managed at diagnosis at Reims University Hospital between 1997 and 2019 were included. The clinical and biological data were extracted from the Champagne-Ardenne Diabetes Network database. Included were 847 patients with a median age of 10.3 years. Diagnosis was established in 71% of cases before 15 years, 7.4% after 35 years. The number of newly diagnosed cases was 3.6-times higher in 2019 compared to 1997. Ketoacidosis, the frequency of which decreased with age (P < 0.0001), revealed diabetes in a total of 32% of cases and in 46% of children under 5 years. It was more severe in children than in adults (P = 0.03), and its frequency increased over the study period. Hypotrophy was found in 23% of children under 15 years of age, and was more pronounced before 5 years of age, with no improvement over time. We saw an increase in the frequency of obesity or overweight among adults. Our study showed an increase in incident cases of diabetes in our hospital that continued over time for both children and adults. Clinical features at diagnosis deteriorated during this period for those under 15 years of age with an increase in ketoacidosis frequency.

Early predictors of diabetic retinopathy in type 1 diabetes: The Retinopathy Champagne Ardenne Diabète (ReCAD) study.

AIMS: To determine the relationship between early markers of diabetes control and diabetic retinopathy (DR) in type 1 diabetes. METHODS: A historic cohort study was conducted on 712 patients from the CARéDIAB database. HbA1c and usual metabolic parameters were measured one year after diagnosis of diabetes. First occurrences of severe hypoglycemia and ketoacidosis during follow-up were selected as time-dependent markers of diabetes control. Data were analyzed in a Cox model using SPSS software to predict DR with significance level at p-value <0.05. RESULTS: In multivariate regression, any diabetic retinopathy was predicted by HbA1c (HR = 1.38; CI = 1.25-1.52; p < 0.0001), severe hypoglycemia (HR = 3; CI = 1.99-4.52; p < 0.0001), ketoacidosis (HR = 1.96; CI = 1.17-3.22; p = 0.009), and age at diagnosis (HR = 1.016; CI = 1.002-1.031; p = 0.02). Proliferative DR was predicted by HbA1c (HR = 1.67; CI = 1.51-1.79; p < 0.0001), severe hypoglycemia (HR = 3.67; CI = 2.74-5.25; p < 0.0001), and ketoacidosis (HR = 2.37; CI = 1.56-3.18; p < 0.0001). CONCLUSION: This study shows that the failure to achieve diabetes control after the first year of diagnosis as well as early episodes of acute diabetes complications may contribute to the occurrence of diabetic retinopathy in type 1 diabetes patients.

Treatment of pituitary adenomas by fractionated stereotactic radiotherapy: a prospective study of 110 patients.

PURPOSE: To optimize and reduce the toxicity of pituitary adenoma irradiation by assessing the feasibility and effectiveness of fractionated stereotactic radiotherapy (FSR). METHODS AND MATERIALS: Between 1990 and 1999, 110 consecutive patients, 47 with a functioning adenoma, were treated according to a strategy of either early surgery and FSR (n = 89) or FSR only (n = 21). Of the 110 patients, 75 had persistent macroscopic tumor and 47 persistent hormonal secretions; 15 were treated in the prophylactic setting. The linear accelerator-delivered dose was 50.4 Gy (5 x 1.8 Gy weekly), with a 2-mm safety margin. RESULTS: After a minimal follow-up of 48 months, only 1 patient had developed progression. Of the 110 patients, 27 (36%) had a complete tumor response, 67 (89.3%) had an objective tumor response, 20 (42%) had a hormonal complete response, and 47 (100%) had a hormonal objective tumor response. The proportion of patients without a complete tumor response, objective tumor response, complete hormonal response, and objective hormonal response was 85.1%, 62%, 83%, and 59.3% at 4 years and 49.3%, 9%, 59.3%, and 10.6% at 8 years, respectively. The sole unfavorable predictive factor was preoperative SSE >20 mm for tumor response (p = 0.01) and growth hormone adenoma for the hormonal response (p <0.001). No late complications, except for pituitary deficiency, were reported, with a probability of requiring hormonal replacement of 28.5% and 35% at 4 and 8 years, respectively. Nonfunctioning status was the sole unfavorable factor (p = 0.0016). CONCLUSIONS: Surgery plus FSR is safe and effective. FSR focused to the target volume seems more suitable than standard radiotherapy, and standard fractionation reduces the risk of optic neuropathy sometimes observed after single-dose radiosurgery. Therefore, FSR allows us to consider combined transrhinoseptal surgery and early radiotherapy, with a curative goal without patient selection.

Assessment of patient-led or physician-driven continuous glucose monitoring in patients with poorly controlled type 1 diabetes using basal-bolus insulin regimens: a 1-year multicenter study.

OBJECTIVE: The benefits of real-time continuous glucose monitoring (CGM) have been demonstrated in patients with type 1 diabetes. Our aim was to compare the effect of two modes of use of CGM, patient led or physician driven, for 1 year in subjects with poorly controlled type 1 diabetes. RESEARCH DESIGN AND METHODS: Patients with type 1 diabetes aged 8-60 years with HbA(1c) ≥ 8% were randomly assigned to three groups (1:1:1). Outcomes for glucose control were assessed at 1 year for two modes of CGM (group 1: patient led; group 2: physician driven) versus conventional self-monitoring of blood glucose (group 3: control). RESULTS: A total of 257 subjects with type 1 diabetes underwent screening. Of these, 197 were randomized, with 178 patients completing the study (age: 36 ± 14 years; HbA(1c): 8.9 ± 0.9%). HbA(1c) improved similarly in both CGM groups and was reduced compared with the control group (group 1 vs. group 3: -0.52%, P = 0.0006; group 2 vs. group 3: -0.47%, P = 0.0008; groups 1 + 2 vs. group 3: -0.50%, P < 0.0001). The incidence of hypoglycemia was similar in the three groups. Patient SF-36 questionnaire physical health score improved in both experimental CGM groups (P = 0.004). Sensor consumption was 34% lower in group 2 than in group 1 (median [Q1-Q3] consumption: group 1: 3.42/month [2.20-3.91] vs. group 2: 2.25/month [1.27-2.99], P = 0.001). CONCLUSIONS: Both patient-led and physician-driven CGM provide similar long-term improvement in glucose control in patients with poorly controlled type 1 diabetes, but the physician-driven CGM mode used fewer sensors.