Endocrine Abstracts

JOINT2831

Objective: There is increasing data that sodium glucose cotransporter 2 inhibitors (SGLT-2i) may increase the risk of euglycemic diabetic ketoacidosis (DKA). This study aims to retrospectively evaluate clinical and metabolic differences observed during the course of DKA among type 2 diabetic patients (T2DM) using or not using SGLT-2i and type 1 diabetic patients (T1DM). The differences in treatment modalities and intensive care processes among the groups were also assesed.

Materials and Methods: Demographic data, treatment modalities, laboratory results (plasma glucose, HbA1c, C-peptide, blood gas analysis, electrolytes, inflammation markers), clinical parameters [Glasgow Coma Scale (GCS), vital signs], micro- and macrovascular complications and DKA management parameters (ICU admissions, time to resolution of ketosis, the duration of insulin infusion, the amount of HCO3 and saline requirement) were obtained by reviewing electronic medical records and patient files.

Results: Among included 124 patients (61F, 63M), 30,6% were diagnosed with T1DM (n:38), while 69.4% had T2DM (n:86). Among patients with T2DM, 37.2% (n:32) of them were receiving SGLT-2i. The average diabetes duration was similiar among the groups (13.5 years in SGLT-2i users vs. 13.6 years in non-SGLT-2i users vs.12.2 years in T1DM; P>0.05). Infection was the main precipitating factor for DKA in (13 out of 32) SGLT-2i users. There was no statistically significant difference among the groups regarding GCS, other vital signs, or ICU admission rates. Initial glucose levels, HbA1c, C-peptide, lipid profile, blood gas ph, lactate, electrolyte levels were similiar among the groups. But, all patients with euglycemic DKA with plasma glucose<300 mg/dL (n =9, 7.2%) were found in the SGLT-2i users group. On the otherhand, SGLT-2i users had higher degree of ketosis compared to patients non-SGLT-2i users (keton 3 positivity, 65.6% vs. 35.2%, P<0.05). Moreover, ketonemia resolution time and insulin infusion duration were significantly higher in patients using SGLT-2i compared to other groups (P<0.05 for all). In addition, more amount of HCO3 (6.59±8.68 mmol/l in SGLT-2i users vs. 2.41±4.67 mmol/l in non-SGLT-2i users vs. 3.11±5.71 mmol/l in T1DM; P<0.05) and saline infusion (6078.1±1739.9 mL in SGLT-2i users vs. 3894.4±965.3 mL in non-SGLT-2i users vs. 4894.7±1732.3 mL in T1DM; P<0.05) was required in SGLT-2i users compared to other groups during the treatment.

Conclusion: Considering the association of SGLT-2i use with prolonged ketone negativity and insulin infusion time and increased HCO3 and fluid requirements, intensive care treatment protocols should be personalized and more stringent control and follow-up protocols should be developed for these patients.