Thoracoscopically assisted Ivor-Lewis oesophagectomy potentially combi
nes the pulmonary advantages of transhiatal oesophageal dissection, wi
th the visibility and control permitted by thoracotomy. This study rev
iewed 17 patients who underwent this procedure with an intrathoracic a
nastomosis. Five patients required conversion to thoracotomy, four bec
ause of technical difficulties with the anastomosis. After operation 1
3 patients had radiological evidence of atelectasis, six developed a l
eft pleural effusion and five had clinically significant pneumonia. Th
ree patients developed an anastomotic leak, two of whom died giving an
in-hospital mortality rate of 12 per cent. Median postoperative hospi
tal stay was 12 days. Four patients developed benign anastomotic struc
tures requiring dilatation. The 1- and 2-year survival rates were 73 p
er cent (11 of 15 patients) and 63 per cent (five of eight) respective
ly. The use of minimal access techniques in this context does not appe
ar to reduce the postoperative incidence of either pulmonary or anasto
motic complications.