Acute haemolysis after transcatheter mitral valve implantation

This commentary refers to ‘Clinical and haemodynamic outcomes of balloon-expandable transcatheter mitral valve implantation: a 7-year experience’, by M. Urena et al., 2018, pages 2679–2689. We carefully read the article recently published by Urena et al. about their experience of transcatheter mitral valve implantation (TMVI) in patients with failed surgical mitral bioprosthesis (valve-in-valve, ViV) or ring (valve-in-ring, ViR). This study confirms that this procedure is safe and could be an excellent alternative therapeutic option in selected patients at very high-risk for a redo surgery.1 The early complications that may occur are described, but we are concerned by the absence of reported cases of acute haemolysis. We report two consecutive cases of acute haemolysis complicated by acute renal failure after ViR procedures in two octogenarian women. Both had a severe mitral regurgitation (MR) related to a failed mitral valve repair using a complete semi-rigid ring (Edwards Physio 2 N°32 and N°26) and were considered by our heart team as contraindicated for cardiac surgery. After a careful cardiac computed tomography analysis, an Edwards Sapien 3 n°26 prosthesis was chosen for the two patients. The two procedures were performed under general anaesthesia with transoesophageal echocardiography monitoring. All the different steps were performed as described by Urena et al.2 After catheters removal, only mild MR were observed and the final mean gradient was respectively 3 mmHg and 5 mmHg for the two patients. Interestingly, the residual mitral valve leaks in these two patients were not located around the Edwards valve, but passed through the uncovered part of the stent (Figures 1 and 2). While no cardiac event or haemodynamic instability were observed after the procedure, an acute renal failure with anuria associated with an acute mechanical haemolysis occurred within the first hours in the two patients. The two patients required blood transfusions during the period of haemolysis: twice the first week after the procedure for the first patient, and once a week until the second procedure in the second patient. The haemolysis rapidly decreased (Figure 1) in the first patient, but dialysis was required during 2 months for the acute renal failure. At 3-month follow-up, renal function was normalized and no trace of MR was observed. The second patient had persistent severe haemolysis and residual intrastent leaks at 2 months (Figure 1). The haemolysis was corrected by the implantation of a new Edwards Sapien 3 n°26 valve deeper into the left ventricle, over the uncovered portion of the first Edwards stent.