Department of Gastroenterology, Hepatology and Gastrointestinal Oncology, Bogenhausen Academic Teaching Hospital, Technical University of Munich, Munich, Germany.
Background: Diabetes is frequently diagnosed in patients with cirrhosis and represents an important risk factor for morbidity and mortality. Pharmacological therapy is limited due to hepatotoxicity and the risk of hypoglycemia. Investigations on medical practice in this patient population, frequency of diabetes-associated complications and the impact of quality of metabolic control are rare. Aims and Methods: A retrospective analysis was performed to compare the effects of hypoglycemic treatment, the achieved glycemic control under therapy, the prevalence of typical cirrhosis-related or microangiopathic complications, and cardiovascular comorbidities between a group of diabetic patients with cirrhosis (n = 87) and a nondiabetic cirrhotic population (n = 198). Results: The prevalence of diabetes in our cohort was 30.5%. Of all diabetic patients, 39.1% received therapy which might potentially result in serious side effects in patients with end-stage liver disease. The rate of ongoing alcohol abuse (28.7%) and noncompliance under medication (41.4%) was high. Only 28.7% of all diabetic subjects showed satisfactory (as defined by HbA(1c) ≤ 6.5%) glycemic control under therapy. Patients achieving satisfactory control experienced a lower rate of certain cirrhosis-related complications such as hepatic encephalopathy (HE) and hepatocellular carcinoma (HCC), arterial hypertension, and hypercholesterolemia. HE was significantly more frequent in diabetic than nondiabetic cirrhotic patients [HE 36.6% (diabetics) vs. 20.7% (non-diabetics), p = 0.001; OR(adj) = 3.21 (CI: 1.63, 6.28)], whereas no significant difference in the frequency of HCC [18.4% (diabetics) vs. 14.1% (nondiabetics), p = 0.606] was observed. In the majority of our diabetic population (59.7%), no microvascular damage was diagnosed. However, diabetic patients had a borderline significant high prevalence of arterial hypertension [48.3% (diabetics) vs. 26.8% (nondiabetics), p = 0.078; OR(adj) = 1.68 (CI: 0.944, 2.978)] and high cholesterol levels [17.2% (diabetics) vs. 8.6% (nondiabetics), p = 0.120, OR(adj) = 1.93 (CI: 0.842, 4.410)]. Conclusion: Antidiabetic therapy in cirrhotic diabetic patients often seems to be inappropriate in everyday medicine, while glycemic control is frequently not satisfactory, possibly due to incompliance or insufficient metabolic control. HE occurs more often in cirrhotic patients with diabetes than in nondiabetic patients with cirrhosis. The rate of macro- and microangiopathic complications even in the diabetic cohort is low.