Continuous Glucose Monitoring and Long-Term Assessment of Islet Function in Autologous Islet Transplantation after Total Pancreatectomy for Neoplasm: Preliminary Insights from a Prospective Study.

Targets for continuous glucose monitoring (CGM) are well established for type 1 and type 2 diabetes. In total pancreatectomy with islet auto-transplant (TPIAT), stricter glycemic targets are needed to avoid metabolic stress on transplanted islets, but no guidelines exist for CGM targets. We aimed to determine CGM targets for TPIAT clinical management by associating CGM metrics with goal hemoglobin A1c (HbA1c) ≤6.5%. Targets for time in range (TIR) 70-140, TIR 70-180, mean CGM glucose, and time in hyperglycemia (>140, >180, >250 mg/dL) were chosen to give good sensitivity and specificity for identifying HbA1c ≤6.5%. We included 256 episodes of 14-day CGM metrics with a concurrent HbA1c value (n=82 patients, age 32.8 [SD 16.6] years at surgery, 70% female) who were ≥0.5 years post TPIAT (median 4.1 years) and wearing Dexcom G6. Most patients had more than 1 HbA1c and corresponding CGM available (median 3.1 per patient). As expected, CGM metrics correlated highly with each other. Table 1 shows “best” (Youden criterion) CGM thresholds and sensitivity for HbA1c ≤6.5%. TIR 70-140 mg/dL ≥50% and TIR 70-180 mg/dL ≥75% may be reasonable targets for patients and providers using CGM data to manage diabetes long-term after TPIAT. Also, time in hyperglycemia should be kept to <20% above 180 mg/dL and <3% above 250 mg/dL. Failure of these goals should prompt re-starting or adjusting insulin therapy. Disclosure Z.Somani: None. J.S.Hodges: None. K.Ramanathan: None. S.Chinnakotla: None. G.Beilman: None. M.Bellin: Consultant; Insulet Corporation, Vertex Pharmaceuticals Incorporated, Research Support; Dexcom, Inc., ViaCyte, Inc.

ABSTRACTIntroductionTargets for continuous glucose monitoring (CGM) are well established for type 1 and type 2 diabetes. In total pancreatectomy with islet autotransplantation (TPIAT), stricter glycemic targets are needed to avoid metabolic stress on transplanted islets, but no guidelines exist for CGM targets.MethodsWe aimed to determine CGM targets for TPIAT clinical management by associating CGM metrics with goal hemoglobin A1c (HbA1c) ≤ 6.5%. Targets for time in range (TIR) 70–140, TIR 70–180, mean CGM glucose, and time in hyperglycemia (>140, >180, >250 mg/dL) were chosen to give good sensitivity and specificity for identifying HbA1c ≤6.5%.ResultsWe included 256 pairs of 14‐day CGM metrics with a concurrent HbA1c value (n = 82 patients, age 35 [IQR 19–46] years at surgery, 70% female) who were ≥0.5 years post TPIAT (median 4.1 years) and wearing Dexcom G6. Most patients had more than one HbA1c and corresponding CGM available (median 2 [IQR 1–4] per patient).ConclusionWe found that TIR 70–140 ≥ 50% and TIR 70–180 mg/dL ≥ 75% may be reasonable minimum targets for patients and providers using CGM data to manage diabetes long‐term after TPIAT. Failure to meet these targets should prompt starting or adjusting insulin therapy, especially if hypoglycemia is not a concern.

Children undergoing total pancreatectomy with islet autotransplantation (TPIAT) for chronic pancreatitis require intensive insulin therapy early after TPIAT with narrow glycemic targets, which can a present significant care burden. Outpatient use of continuous glucose monitoring (CGM) systems by children and caregivers early after TPIAT is inadequately studied. In this open-label study, we randomized 14 children and adolescents (mean age 15.4 years) after hospital discharge for TPIAT to Dexcom G6 CGM (n = 7) or standard care with a glucometer (n = 7) to assess acceptability and glycemic control with use of CGM versus usual care (glucometer). Participants in the control arm also wore a blinded CGM for 1 week. Children randomized to real-time CGM had lower mean sensor glucose values compared with controls (p = 0.002), and high overall satisfaction with CGM. Our data indicate that CGM is a useful adjunct to diabetes management for children who have recently undergone TPIAT.

Total Pancreatectomy With Islet Autotransplantation ImprovesQuality of Life in Patients With Refractory RecurrentAcute Pancreatitis.

We investigated six indices based on a single fasting blood sample for evaluation of the beta-cell function after total pancreatectomy with islet autotransplantation (TP-IAT). The Secretory Unit of Islet Transplant Objects (SUITO), transplant estimated function (TEF), homeostasis model assessment (HOMA-2B%), C-peptide/glucose ratio (CP/G), C-peptide/glucose creatinine ratio (CP/GCr) and BETA-2 score were compared against a 90-min serum glucose level, weighted mean C-peptide in mixed meal tolerance test (MMTT), beta score and the Igls score adjusted for islet function in the setting of IAT. We analyzed values from 32 MMTTs in 15 patients after TP-IAT with a follow-up of up to 3 years. Four (27%) individuals had discontinued insulin completely prior to day 75, while 6 out of 12 patients (50%) did not require insulin support at 1-year follow-up with HbA1c 6.0% (5.5-6.8). BETA-2 was the most consistent among indices strongly correlating with all reference measures of beta-cell function (r = 0.62-0.68). In addition, it identified insulin independence (cut-off = 16.2) and optimal/good versus marginal islet function in the Igls score well, with AUROC of 0.85 and 0.96, respectively. Based on a single fasting blood sample, BETA-2 score has the most reliable discriminant value for the assessment of graft function in patients undergoing TP-IAT.

The increasing use of continuous glucose monitor (CGM) necessitates a review of variables that impact accuracy and interrupt use. Manufacturer recommendations include removing CGMs before diagnostic imaging, such as X-ray and computed tomography (CT). Early removal and replacement of CGM components present financial, clinical, and psychosocial burdens to the wearer and interrupt optimal management of diabetes for pediatric patients who receive a total pancreatectomy with islet autotransplantation (TPIAT). The study's aim was to evaluate the effect of scatter dose exposure during X-ray or CT if the CGM remained intact but outside the field of view (FoV). Participants were followed through the first 3 months after TPIAT surgery, managed diabetes with an insulin pump and CGM, and were routinely exposed to diagnostic imaging. Participants' CGMs were unshielded by a protective apron during any X-ray or CT procedures for the duration of the study period, and the transmitter was collected after expiration or removal. Glucometer data was collected from hospital records and home glucometer downloads. Mixed models were used to analyze absolute differences between matched CGM and glucometer values, and Clarke error grid analyses (EGA) were performed. Scatter dose exposure was derived using anthropomorphic phantoms and calculated retrospectively. A total of 14 patients (median 12.2 years, 64% female) received a median of five diagnostic imaging procedures with a median cumulative scatter dose of 559 µGy. The absolute difference between the CGM and glucometer values was not significantly associated with the cumulative scatter dose (p=0.17) or time from TPIAT (p=0.24) when analyzed in a mixed model. Regardless of scatter dose exposure, time from TPIAT, or glucometer, ≥98% of glucose values fell within zones A and B on EGA. Scatter dose exposure from diagnostic imaging did not affect the clinical accuracy of CGM values for the duration of transmitter use. Leaving CGM components in place when not in the FoV during diagnostic imaging successfully mitigated interruptions to use and undue burden or cost to participants.

The optimal method of monitoring glycemia in pregnant women with type 1 diabetes remains controversial. This study aimed to assess the predictive performance of HbA1c, continuous glucose monitoring (CGM) metrics, and alternative biochemical markers of glycemia to predict obstetric and neonatal outcomes. One hundred fifty-seven women from the Continuous Glucose Monitoring in Women With Type 1 Diabetes in Pregnancy Trial (CONCEPTT) were included in this prespecified secondary analysis. HbA1c, CGM data, and alternative biochemical markers (glycated CD59, 1,5-anhydroglucitol, fructosamine, glycated albumin) were compared at ∼12, 24, and 34 weeks' gestation using logistic regression and receiver operating characteristic (ROC) curves to predict pregnancy complications (preeclampsia, preterm delivery, large for gestational age, neonatal hypoglycemia, admission to neonatal intensive care unit). HbA1c, CGM metrics, and alternative laboratory markers were all significantly associated with obstetric and neonatal outcomes at 24 weeks' gestation. More outcomes were associated with CGM metrics during the first trimester and with laboratory markers (area under the ROC curve generally <0.7) during the third trimester. Time in range (TIR) (63-140 mg/dL [3.5-7.8 mmol/L]) and time above range (TAR) (>140 mg/dL [>7.8 mmol/L]) were the most consistently predictive CGM metrics. HbA1c was also a consistent predictor of suboptimal pregnancy outcomes. Some alternative laboratory markers showed promise, but overall, they had lower predictive ability than HbA1c. HbA1c is still an important biomarker for obstetric and neonatal outcomes in type 1 diabetes pregnancy. Alternative biochemical markers of glycemia and other CGM metrics did not substantially increase the prediction of pregnancy outcomes compared with widely available HbA1c and increasingly available CGM metrics (TIR and TAR).

Two weeks of continuous glucose monitoring (CGM) sampling with >70% CGM use is recommended to accurately reflect 90 days of glycemic metrics. However, minimum sampling duration for CGM use <70% is not well studied. We investigated the minimum duration of CGM sampling required for each CGM metric to achieve representative glycemic outcomes for <70% CGM use over 90 days. Ninety days of CGM data were collected in 336 real-life CGM users with type 1 diabetes. CGM data were grouped in 5% increments of CGM use (45%-95%) over 90 days. For each CGM metric and each CGM use category, the correlation between the summary statistic calculated using each sampling period and all 90 days of data was determined using the squared value of the Spearmen correlation coefficient (R2). For CGM use 45% to 95% over 90 days, minimum sampling period is 14 days for mean glucose, time in range (70-180 mg/dL), time >180 mg/dL, and time >250 mg/dL; 28 days for coefficient of variation, and 35 days for time <54 mg/dL. For time <70 mg/dL, 28 days is sufficient between 45 and 80% CGM use, while 21 days is required >80% CGM use. We defined minimum sampling durations for all CGM metrics in suboptimal CGM use. CGM sampling of at least 14 days is required for >45% CGM use over 90 days to sufficiently reflect most of the CGM metrics. Assessment of hypoglycemia and coefficient of variation require a longer sampling period regardless of CGM use duration.

Background: Total pancreatectomy with islet autotransplantation (TPIAT) requires strict glycemic management for islet survival using insulin pumps and continuous glucose monitors (CGMs). Hydroxyurea prevents reactive thrombocytosis but interferes with the accuracy of the Dexcom CGM. Hydroxyurea is reported to not interfere with the Libre CGM but has not been studied after TPIAT. Methods: Seven patients wore both Dexcom and Libre starting approximately a week after TPIAT. Dexcom and Libre values were obtained with point-of-care testing blood glucose (POCT BG) at 560 unique time points. Descriptive statistics included median, interquartile range (IQR), absolute difference between CGM and POCT, and mean absolute relative difference (MARD) for each Dexcom and Libre. Wilcoxon-Mann-Whitney tests were performed to compare parameters between Dexcom and Libre, with two-sided significance of P < 0.05. Clarke error grids and boxplots were constructed. Results: In the 9 h after hydroxyurea, median POCT BG was 110 mg/dL (IQR 88-143), median Dexcom BG was 172 mg/dL (135-219), and median Libre BG was 106 mg/dL (76-138). MARD for Dexcom was 59.5% and for Libre was 14.8% (P < 0.001). Median absolute difference between Dexcom and POCT BG (56 mg/dL [32-88]) was greater than that for Libre (12 mg/dL [6-23]; P < 0.001). In Clarke error grids, 98.3% of values fell within clinically acceptable Zones A/B for Libre; 77.9% of values fell within these zones for Dexcom. At all other times, median POCT BG was 110 mg/dL (86-133), median Dexcom BG was 124 mg/dL (97-154), and median Libre BG was 104 mg/dL (76-128). MARD for Dexcom was 19.8% and for Libre was 14.7% (P < 0.001). Median absolute difference between Dexcom and POCT BG (18 mg/dL [9-30]) was clinically similar to that for Libre (13 mg/dL [6-23], P < 0.001). Conclusion: Hydroxyurea does not seem to interfere with the accuracy of Libre in contrast to Dexcom. Use of Libre after TPIAT could facilitate improved glycemic management.

The aim of this work was to define a unique remission status using glycaemia risk index (GRI) and other continuous glucose monitoring (CGM) metrics in individuals with type 1 diabetes for improved phenotyping. A group of 140 individuals with type 1 diabetes were recruited for a cross-sectional study. The participants were categorised into four groups based on their remission status, which was defined as insulin-dose-adjusted A1c (IDAA1c) <9 or C-peptide ≥300 pmol/l: new-onset (n=24); mid-remission (n=44); post-remission (n=44); and non-remission (individuals who did not experience remission, n=28). Participants in the remission phase were referred to as 'remitters', while those who were not in the remission phase were referred to as 'non-remitters', the latter group including new-onset, post-remission and non-remission participants. Clinical variables such as HbA1c, C-peptide and insulin daily dose, as well as IDAA1C and CGM data, were collected. The patterns of CGM metrics were analysed for each group using generalised estimating equations to investigate the glycaemic variability patterns associated with remission status. Then, unsupervised hierarchical clustering was used to place the participants into subgroups based on GRI and other CGM core metrics. The glycaemic variability patterns associated with remission status were found to be distinct based on the circadian CGM metrics. Remitters showed improved control of blood glucose levels over 14 days within the range of 3.9-10 mmol/l, and lower GRI compared with non-remitters (p<0.001). Moreover, GRI strongly correlated with IDAA1C (r=0.62; p<0.001) and was sufficient to distinguish remitters from non-remitters. Further, four subgroups demonstrating distinct patterns of glycaemic variability associated with different remission status were identified by clustering on CGM metrics: remitters with low risk of dysglycaemia; non-remitters with high risk of hypoglycaemia; non-remitters with high risk of hyperglycaemia; and non-remitters with moderate risk of dysglycaemia. GRI, an integrative index, together with other traditional CGM metrics, helps to identify different glycaemic variability patterns; this might provide specifically tailored monitoring and management strategies for individuals in the various subclusters.

Objective: The optimal method of monitoring glycemia in pregnant women with type 1 diabetes remains controversial. This study aimed to assess the predictive performance of HbA1c, continuous glucose monitoring (CGM) metrics, and alternative biochemical markers of glycemia to predict obstetric and neonatal outcomes. &lt;p&gt;Methods: 157 women from the CGM in pregnant women with type 1 diabetes trial (CONCEPTT) were included in this pre-specified secondary analysis. HbA1c, CGM data, and alternative biochemical markers (glycated CD59, 1,5 anhydroglucitol, fructosamine and glycated albumin) were compared at approximately 12, 24 and 34 weeks gestation using logistic regression and ROC curves to predict pregnancy complications (pre-eclampsia, preterm delivery, large-for-gestational-age, neonatal hypoglycemia, admission to neonatal intensive care unit). &lt;/p&gt; &lt;p&gt;Results: HbA1c, CGM metrics, and alternative laboratory markers were all significantly associated with obstetric and neonatal outcomes at 24 weeks gestation. More outcomes were associated with CGM metrics during the 1&lt;sup&gt;st&lt;/sup&gt; trimester and with laboratory markers (area under ROC generally &lt;0.7) during the third trimester. Time-in-range (TIR; 63-140 mg/dl; 3.5-7.8 mmol/l) and time-above-range (TAR; &gt;140 mg/dl; &gt;7.8 mmol/l) were the most consistently predictive CGM metrics. HbA1c was also a consistent predictor of suboptimal pregnancy outcomes. Some alternative laboratory markers showed promise, but overall, they had lower predictive ability than HbA1c. &lt;/p&gt; &lt;p&gt;Conclusions: HbA1c is still an important biomarker for obstetric and neonatal outcomes in type 1 diabetes pregnancy. Alternative biochemical markers of glycemia and other CGM metrics did not substantially increase the prediction of pregnancy outcomes compared to widely available HbA1c and increasingly available CGM metrics (TIR and TAR).&lt;/p&gt; &lt;br&gt; &lt;p&gt; &lt;/p&gt;

Objective: The optimal method of monitoring glycemia in pregnant women with type 1 diabetes remains controversial. This study aimed to assess the predictive performance of HbA1c, continuous glucose monitoring (CGM) metrics, and alternative biochemical markers of glycemia to predict obstetric and neonatal outcomes. &lt;p&gt;Methods: 157 women from the CGM in pregnant women with type 1 diabetes trial (CONCEPTT) were included in this pre-specified secondary analysis. HbA1c, CGM data, and alternative biochemical markers (glycated CD59, 1,5 anhydroglucitol, fructosamine and glycated albumin) were compared at approximately 12, 24 and 34 weeks gestation using logistic regression and ROC curves to predict pregnancy complications (pre-eclampsia, preterm delivery, large-for-gestational-age, neonatal hypoglycemia, admission to neonatal intensive care unit). &lt;/p&gt; &lt;p&gt;Results: HbA1c, CGM metrics, and alternative laboratory markers were all significantly associated with obstetric and neonatal outcomes at 24 weeks gestation. More outcomes were associated with CGM metrics during the 1&lt;sup&gt;st&lt;/sup&gt; trimester and with laboratory markers (area under ROC generally &lt;0.7) during the third trimester. Time-in-range (TIR; 63-140 mg/dl; 3.5-7.8 mmol/l) and time-above-range (TAR; &gt;140 mg/dl; &gt;7.8 mmol/l) were the most consistently predictive CGM metrics. HbA1c was also a consistent predictor of suboptimal pregnancy outcomes. Some alternative laboratory markers showed promise, but overall, they had lower predictive ability than HbA1c. &lt;/p&gt; &lt;p&gt;Conclusions: HbA1c is still an important biomarker for obstetric and neonatal outcomes in type 1 diabetes pregnancy. Alternative biochemical markers of glycemia and other CGM metrics did not substantially increase the prediction of pregnancy outcomes compared to widely available HbA1c and increasingly available CGM metrics (TIR and TAR).&lt;/p&gt; &lt;br&gt; &lt;p&gt; &lt;/p&gt;

Total pancreatectomy with islet autotransplantation (TPIAT) is an established and effective treatment modality for patients diagnosed with intractable chronic pancreatitis (CP) and recurrent acute pancreatitis (RAP). TPIAT primarily aims to manage debilitating pain leading to impaired quality of life among patients with CP or RAP, which can be successfully managed with medical, endoscopic, or surgical interventions. TPIAT is significantly successful in relieving pain associated with CP and improving health‐related quality of life outcomes. Furthermore, the complete loss of pancreatic endocrine function attributed to total pancreatectomy (TP) can be compensated by autologous islet transplantation (IAT). Patients receiving IAT can achieve insulin independence or can be less dependent on exogenous insulin compared with those receiving TP alone. Historically, TPIAT has been mainly used in the United States, and its outcomes have been improving due to technological advancements. Despite some challenges, TPIAT can be a promising treatment for patients with CP‐related intractable pain. Thus far, TPIAT is not commonly performed in Japan. Nevertheless, it may improve health‐related quality of life in Japanese patients with CP, similar to Western patients. This review article aimed to provide an overview of the indications, related procedures, and outcomes of TPIAT and to discuss future prospects in Japan.

Background: Total pancreatectomy with islet autotransplant (TPIAT) is done in patients with chronic pancreatitis to treat intractable pain. In TPIAT, islets are isolated after pancreatectomy and infused into the liver via the portal vein to mitigate post-operative diabetes. Outcomes vary, with ≥60% needing exogenous insulin supplementation to maintain normoglycemia. The current study’s aim was to determine if pre-surgical body composition is associated with islet function and insulin sensitivity after TPIAT.Methods: We characterized body weight and composition as related to insulin sensitivity and dependence and diabetes outcome in 88 adults who underwent TPIAT for chronic pancreatitis at the University of Minnesota. At baseline, 12 and 18 months after TPIAT, insulin independence was assessed; metabolic testing used mixed meal tolerance testing and frequent sample intravenous glucose tolerance testing. Body composition was measured by Dual X-ray absorptiometry (DXA). Statistical analyses used linear and logistic regression.Results: At baseline, mean age was 39.9 (SD 11.1) years. 9.1% were underweight (BMI<18.5 kg/m2), 45.5% normal weight (BMI=18.5–24.9), 22.7% overweight (BMI=25–29.9) and 22.7% obese (BMI≥30). Islet equivalent per kg did not differ between body weight categories (p=0.17). Overweight/obese patients had higher peak and AUC c-peptide and lower insulin sensitivity index, as expected. Compared to baseline, android to gynoid fat ratio was lower at 12 (0.80 vs 0.88; p=0.012) and 18 months (0.81 vs 0.88; p=0.041), and lean mass was lower at 18 months (38848 vs 42338 kg; p=0.029). Baseline body weight was positively associated with acute insulin response to glucose (AIRg) at 12 months (effect size 38.5, SE 17.1 mU/L/min; p=0.029) and 18 months (38.3, SE 18.5 mU/L/min; p=0.045), while baseline lean mass was inversely associated with AIRg at 12 (p=0.01) and 18 months (p=0.033). Baseline body weight was positively associated, and fat mass inversely associated with disposition index (Di; islets’ ability to secrete insulin normalized to insulin resistance) at 18 months (p=0.019 for both).Percent body fat and percent gynoid fat predicted Sg (glucose effectiveness index, i.e., ability of glucose to promote its own disposal and inhibit hepatic glucose production absent an incremental insulin effect) at 18 months (p=0.042 and p=0.019, respectively). Insulin independence at 12 and 18 months was not significantly associated with baseline body weight or body composition.Conclusions: Overweight/obesity is common in patients with chronic pancreatitis. After TPIAT, patients had lower muscle mass and A/G ratio. Preoperative body weight and composition were associated with islet function but not insulin independence after TPIAT surgery.

A selective therapy for pancreatitis is total pancreatectomy and islet autotransplantation. Outcomes and geographical variability of patients who had total pancreatectomy (TP) alone or total pancreatectomy with islet autotransplantation (TPIAT) were assessed. Data were obtained from the Healthcare Cost and Utilization Project National Inpatient Sample database. Weighed univariate and multivariate analyses were performed to determine the effect of measured variables on outcomes. Between 2002 and 2013, there were 1006 TP and 825 TPIAT in patients with a diagnosis of chronic pancreatitis, and 1705 TP and 830 TPIAT for any diagnosis of pancreatitis. The majority of the TP and TPIAT were performed in larger urban hospitals. Costs were similar for TP and TPIAT for chronic pancreatitis but were lower for TPIAT compared with TP for any type of pancreatitis. The trend for TP and TPIAT was significant in all geographical areas during the study period. There is an increasing trend of both TP and TPIAT. Certain groups are more likely to be offered TPIAT compared with TP alone. More data are needed to understand disparities and barriers to TPIAT, and long-term outcomes of TPIAT such as pain control and glucose intolerance need further study.