MON-LB114 CGM in Cystic Fibrosis Patients to Predict Cystic Fibrosis-Related Diabetes Onset

Background: Therapeutic progress and improvement on resources enabled the emergence of new comorbidities in cystic fibrosis (CF), such as cystic fibrosis-related diabetes (CFRD). About 20% of adolescents and 40-50% of adults are affected. CFRD and glucose intolerance reduce life expectancy in this population, highlighting the importance of early diagnosis and treatment. Up to 15% of CF patients have hypoglycemia during OGTT and its etiology remains unclear. Some authors associate hypoglycemia with CFRD onset, while others do not agree with this association. Objective: To determine whether abnormal CGM (hypo/hyperglycemia) can predict CFRD onset, pulmonary function and BMI decline in CF patients. Methods: Prospective single center study. All CF patients between 10-19yo from our outpatient clinic were screened for CFRD through OGTT following the World Health Organization (WHO) protocol. The enzymatic colorimetric method was used to classify them as per the ADA. Non-diabetic CF patients performed 3-day CGM, forced expiratory volume in the first second (FEV1), BMI and OGTT. All tests except for CGM were then reassessed after a long follow-up. The WHO’s 2006 curve was used to calculate the z scores for individuals ≤19yo and WHO cut-off values for >19yo. Oral corticoid use during data collection, pregnancy and solid organ transplantation were exclusion criteria. Results: Thirty-nine patients were recruited and 34 completed an average of 3.1 years (±0.51) follow-up. No clinical or laboratory variables could predict diabetes progression or hypoglycemic events. The cohort had an increase in mean BMI (17.80±3.65 vs 18.36±3.49; p=0.025) and a reduction in mean FEV1 (66.91±25.79% vs 56.32±29.57%; p=0.001) between the two evaluations. Patients who developed diabetes showed statistically significant worse FEV1 in the end of the follow-up (22.67±5 vs 59.58±28.9; p=0.041), and lower BMI at both start (14.37±1.23 vs 18.13±3.65; p=0.049) and end (14.81±0.66 vs 18.71±3.46; p=0.029) of follow-up. A logistic regression of the effect of time adjusted for independent variables for progression to CFRD was conducted. A higher possibility of evolution among participants with IGT (odds ratio [OR] 21.67; 95% confidence interval [CI] 7.03-67.36; p<0.01), and a lower possibility among participants with NGT (OR 1.84; 95% CI 1.06-3.19; p=0.031). Conclusion: CGM was not a useful tool to predict early diabetes onset in this population with the current cut-off values. However, the IGT group seems to be the riskiest group. The CF population has particular characteristics and may not have the same diagnostic criteria for DM as the non-CF population. More studies are necessary to determine the appropriate CGM cut-off values for CFRD.