In Western countries, the recent epidemiological shift from squamous cell carcinoma to adenocarcinoma arising in Barrett's metaplasia has led to an increasing referral of patients with early oesophageal tumours detected during endoscopic surveillance [
1]. Squamous cell carcinoma (SCC) is associated with low socioeconomic status [
2], active tobacco and alcohol abuse, malnutrition, liver dysfunction, pulmonary co-morbidities, and second malignancies [
3].
Patients with adenocarcinoma (AC) are characterized by co-morbidities such as coronary heart disease and a higher median age [
4]. AC is predominantly (94%) located in the lower third of the oesophagus, whereas 51% of SCC are found in the middle third and only 36% in the lower third. Moreover, a better prognosis with a significantly higher overall survival after resection of AC than SCC was reported in some studies [
5‐
7] whereas a SEER database review of 4752 patients showed no difference [
8]. However, the majority of patients still present with advanced disease and up to two thirds are inoperable at the time of diagnosis.
Complete resection (R0), N- and T-stage are independent prognostic factors for SCC. Patients are categorised in risk groups by Karnofsky Performance Scale (KPS), cardiac function, liver and lung parameters [
9]. Pre-operative improvement of nutritional status, abstention from tobacco and alcohol can decrease the perioperative risk. Patients with SCC of the cervical oesophagus, T1 - 2, with low surgical risk according to Bartels et al. [
9], can be treated by a limited resection including regional lymphadenectomy and reconstruction using a free jejunal loop with microsurgical vessel anastomoses, whereas T3-4 patients are treated with neoadjuvant radiochemotherapy. Patients with a high perioperative risk get definitive radiochemotherapy regardless of T-stage. In the low risk situation, T1-2 tumours located in the middle and lower third of the oesophagus are treated with transthoracic en-bloc-oesophagectomy with two-field lymphadenectomy and reconstruction with a gastric tube. Use of the colon as an esophageal substitute is reserved to patients with previous gastric resection. In patients with T3-4 tumours the same surgical strategy is chosen, if possible after preoperative radiochemotherapy. Again, for patients with higher perioperative risk definitive radiochemotherapy is the treatment of choice. For AC R0, T- and N-stage are also independent prognostic markers. Grading is more advantageous in carcinoma of the gastro-oesophageal junction (GEJ) I than GEJ II/III, with 80% of intestinal metaplasia (Barrett's oesophagus) being found in GEJ I [
6]. The surgical procedure of choice for GEJ I is subtotal oesophagectomy with proximal gastric resection and a two-field lymphadenectomy, whereas GEJ II/III is treated by transhiatal extended gastric resection and oesophagojejunostomy. For early GEJ I-III a transabdominal limited resection of the distal oesophagus and the proximal stomach with interposition of small intestine (Merendino procedure) can be performed. When transthoracic oesophagectomy (TTE) is compared to the transhiatal oesophagectomy (THE) for adenocarcinoma of the mid and distal oesophagus, no significant difference in overall survival can be observed, but a tendency towards better 5-year survival for TTE in GEJ I and better locoregional control with limited lymphnode invasion have been reported [
10,
11]. Kato et al. showed a significantly higher overall survival in 3-field versus 2-field lymphadenectomy [
12], whereas a randomised trial showed no benefit [
13]. Cervical lymphadenectomy seems to be useful in carcinomas located in the cervical and upper third of the oesophagus [
13,
14]. Transhiatal oesophagectomy is indicated in patients with high pulmonary risk since it decreases early morbidity and mortality but has a trend to worse long term survival. With either a 3-field or a 2-field approach 5-year overall survival rates of 20% can be achieved [
15]. Hence, oesophagectomy is a complex operation that entails a two or three-field approach depending on the site of tumor, clinical staging, and Karnofsky performance status. Although overall postoperative mortality has decreased to less than 5% in high-volume centers [
16], anastomotic and respiratory failures are still frequent [
11]. In the past three decades surgery has developed from transhiatal oesophagectomy [
17] to video-assisted surgery [
18,
19]. Laparoscopy has provided the opportunity of minimally invasive surgical staging [
20] and gastric mobilisation with D2 lymphadenectomy extended to the lower mediastinal compartment [
21,
22]. Furthermore, it was shown that hybrid operations combining laparoscopy and right thoracotomy could be advantageous in regards to respiratory function [
23]. A three-stage thoracoscopic oesophagectomy with cervical anastomosis may represent a better minimally invasive surgical option in SCC patients [
24,
25]. Expected advantages of minimal access techniques include a decrease in postoperative pain, inflammatory cytokine production, cardiopulmonary complications, blood loss, and the length of hospital stay. Although short and medium-term efficacy of these procedures have been proven [
26‐
28], results are still inconclusive. As multicentre studies are not available and because of problems with standardization of such complex procedures, the effectiveness of minimal access oesophageal surgery is difficult to demonstrate.
In summary, from a surgical point of view, AC and SCC need separate therapeutic strategies for which accurate patient selection (staging, evaluation of co-morbidities) is indispensable. Minimally invasive oesophageal surgery is evolving and may become increasingly important. Still, it is hard to imagine that the management of oesophageal cancer will merely be based on improved surgery. Instead, surgeons should be ready for a new scenario, which comprises biological tumour staging and targeted therapies combined with neoadjuvant radiochemotherapy.