Cardiopulmonary complications of ERCP in older patients

Background: Biochemical markers of ERCP-related myocardial injury have not previously been investigated. Objective: To evaluate ERCP-related cardiac troponin I (cTnI) release, myocardial ischemia, hemodynamic changes, and arterial hypoxemia in a series of consecutive patients according to age and to determine their relationship to preexisting cardiovascular risk factors (RF) and the development of post-ERCP pancreatitis. Design: Prospective cohort study. Setting: Tertiary teaching hospital, Canberra, Australia. Patients: Data were collected on 130 consecutive ERCPs performed on 100 unselected patients (aged 18-93 years) by one endoscopist. Patients were divided into two groups: 65 years of age and older (group 1, n = 53; 27 women) and less than 65 years of age (group 2, n = 47; 33 women). Interventions: ERCP Main Outcome Measurements: Cardiovascular RFs were identified, and electrocardiogram (ECG), cTnI, creatine kinase (CK), amylase, and lipase were measured before and 24 hours after ERCP Oxygen saturation (SpO(2)), heart rate (HR), blood pressure (BP), and ECG were monitored continuously during each procedure. Results: New ECG changes (ischemia, arrhythmias) occurred in 24% of procedures in group 1 and in 9.3% in group 2 (p = 0.168), and episodic arterial hypoxemia (SpO(2) < 90%) in 16.2% (group 1) and 21.4% (group 2) (P = 0.596). A post-ERCP rise in cTnI levels was documented in 6 patients in the older group. Two of these patients died: one from acute myocardial infarction and one from undiagnosed ascending aortic aneurysm. A cTnI rise was not related to any comorbid conditions, total number of Us, hemodynamic or ECG changes, or arterial desaturation. In patients with a new cTnI rise, the duration of ERCP was significantly longer (59.5 vs. 26.4 minutes, p = 0.026), being 30 minutes or longer in 5 of 6 patients. Post-ERCP pancreatitis was associated with desaturation (relative risk [RR] = 5.9; 95% confidence interval [CI] [1.2, 32.0], p = 0.027) and myocardial ischemia/injury (RR = 4.4; 95% CI [1.4, 7.8]; p = 0.009). Conclusions: Although the majority of older patients tolerated ERCP well, in 8% of procedures, most of which were prolonged (> 30 minutes), myocardial injury, as defined by the release of cTnI, occurred. Desaturation and myocardial ischemia/injury were associated with post-ERCP pancreatitis.