MITRAL-VALVE OPERATION VIA PORT ACCESS VERSUS MEDIAN STERNOTOMY

Objective: The advantages and disadvantages of minimally invasive Port Access mitral valve operation have not been defined relative to stand ard median sternotomy. A study was therefore designed to delineate dif ferences in outcome from mitral operation via Port Access versus stern otomy in comparable patients. Methods: The records of 41 consecutive p atients undergoing isolated mitral valve replacement (n = 14) or repai r (n = 27) were examined. All operations were performed using cardiopl egic arrest through either median sternotomy (n = 20) or a small right anterolateral thoracotomy using an endoaortic clamp and catheter syst em (Heartport, Redwood City, CA) to arrest and decompress the heart (P ort Access, n = 21). Results: Both groups were well matched for age,mi tral pathology, ejection fraction, and comorbidity, except that Port A ccess patients were less likely to be female. Three patients had under gone previous cardiac operations. Surgical procedure time was longer f or Port Access patients (384 +/- 80 vs. 263 +/- 41 min, P < 0.05). Por t Access provided significantly smaller incision length (8 +/- 2 vs. 2 6 +/- 2 cm, P < 0.01) and similar or shorter hospital stay (6 +/- 4 vs . 7 +/- 3 days). Port: Access provided excellent visualization of the mitral valve and subvalvular apparatus, generally better than sternoto my, to allow complex mitral valve repairs. The greatest advantage of P ort Access mitral operation was that Port Access patients returned to normal activity more rapidly (4 +/- 2 vs. 9 +/- 1 weeks, P = 0.01) tha n did patients undergoing standard median sternotomy. Conclusions: By avoiding a sternotomy, Port Access mitral valve operation provided a s maller incision and a dramatically more rapid return to normal activit y than did median sternotomy. Port Access cardioplegic arrest with the Heartport system allowed visualization of the mitral valve superior t o median sternotomy and has become the standard approach at this insti tution. (C) 1998 Elsevier Science B.V. All rights reserved.