Prospective evaluation of preoperative lung ultrasound for prediction of perioperative outcome and myocardial injury in adult patients undergoing vascular surgery (LUPPO study).

Abstract

Myocardial injury after non-cardiac surgery (MINS) is a frequent perioperative event in vascular surgery, associated both with worse outcome and subsequent cardiovascular events. Current guidelines advocate troponin (hs-cTnT) and NT-proBNP measurements in selected patients before surgery, but accurate preoperative identification of patients at risk for MINS is an unmet clinical need. Focused lung ultrasound (LUS) might help to select patients at increased risk for MINS, because it can visualize B-line artifacts correlating to cardiopulmonary disease. Therefore, we investigated whether quantification of B-line artifacts improves perioperative risk predictive accuracy for MINS. In this prospective single-center observational study, 136 consecutive open vascular surgery patients underwent conventional preoperative assessment expanded by lung ultrasound. Lung ultrasound B-lines were counted in each of 28 bilateral scan fields of the anterior and lateral chest. Improvement of risk predictive accuracy was quantified with area under receiver operating characteristic (ROC) curve analysis and net reclassification improvement (NRI). We included 118 patients into the final analysis. Twenty-three (19%) patients fulfilled the criteria for the primary endpoint MINS. Three or more bilateral positive B-line fields were calculated as the best ROC-derived cutoff associated with an increased incidence of MINS (odds ratio: 4.4; 95% confidence interval [CI]: 1.5 to 12.7; P=0.007). Adding LUS to hs-cTnT measurements improved risk predictive accuracy for MINS (NRI: 0.36, P=0.043). Lung ultrasound in combination with hs-cTnT showed a better test accuracy than hs-cTnT alone and might guide clinicians to identify vascular patients at increased risk for MINS.