R0 surgery represents the “gold standard” multimodal treatment of tumors of the esophagus because it offers the best chance of cure even though esophagectomy (despite the significant technical improvements and advances in surgical technique and perioperative management) is associated with significant morbidity and mortality. Minimally invasive surgery has been developed to reduce the complications related to esophagectomy, especially respiratory diseases (which represent the main cause of mortality). Although minimally invasive surgery for esophageal cancer started in the early 1990s, debate continues regarding its safety, efficacy, and benefits (contrary to the situation, for example, with colorectal surgery). Thus, in recent years, minimally invasive surgery for esophageal cancer has spread worldwide. However, this spread has been slower compared with other laparoscopic and thoracoscopic procedures, mainly due to the technical difficulties that this surgery entails and the lack of consensus in the literature. The reason for this is multifactorial and based on the relative rarity of esophageal tumors (which limits randomized studies) and the great variety of minimally invasive surgical approaches more or less associated with traditional surgery. A recent international survey involving 269 surgeons indicated that 78% of them continued to favor open approaches, 14% indicated a preference for minimally invasive resection, and 8% had no preference [42]. What emerges from numerous studies is that MIE is definitely a time-consuming process, as confirmed by the meta-analyses of Butler et al. [43] and Watanabe et al., [44] with a steep learning curve, but with a significant reduction in blood loss. In these studies, the percentage of conversion differs widely and is closely dependent on the experience of the surgeon, so in high-volume centers the rate of conversion is 0–7.3% [45–48], whereas in low-volume centers it is 10–36% [43,49,50]. In the meta-analysis of Butler et al., [43] the median mortality in total minimally invasive transthoracic esophagectomy was 1% (range, 0–6.5%), and in minimally invasive transhiatal esophagectomy (THE) it was 0% (range, 0–4.6%). The rate of mortality of MIE was similar to that for open surgery in the meta-analysis of Nagpal et al. [51] (p = 0.26) and in the meta-analysis of Uttley et al. [52], in which mortality was 2.4% for MIE and 3.8% for open surgery. The possible role of MIE in the reduction of morbidity in general and for respiratory complications in particular has been investigated by many retrospective and prospective studies as well as meta-analyses (Table 1.1). In particular, the meta-analysis of Nagpal et al. [51], which took into account 12 studies involving 672 patients (MIE and hybrid mininvasive esophagectomy [hMIE]) compared with 612 patients (open esophagectomy [OE]), showed a reduction of morbidity, including respiratory complications (p = 0.04). However, the same authors pointed out that there may be a bias in the analysis related to the inclusion of studies involving THE. Conversely, in the meta-analysis of Watanabe et al. [44], 10 of the 17 retrospective cohort studies did not show substantial differences in respiratory complications. Many authors [46,53–55] believe that the prone position (PP) could reduce pulmonary complications and have technical and physiological advantages. Regarding the former, certainly the visualization of anatomical structures is better because the lungs do not obscure the surgical field, even with one-lung ventilation. Moreover, the esophagous does not lie in the most declivous portion of the chest and is not obscured by overlying blood. Trauma to the lung is also reduced because it does not need to be retracted; the surgeon operates in a plane parallel to the camera with a similar view to that enabled by abdominal laparoscopic surgery. Finally, mobilization of the esophagus and lymphadenectomy become easier, especially at the level of the aortopulmonary window and close to the recurrent nerve, also on the left side. The undoubted physiological advantage in the prone position is the ability to operate without the excluded lung or a partially desufflated lung, thereby avoiding the pulmonary insufflation and desufflation that causes the release of mediators of inflammation and, even if not clinically tested [56], could be responsible for respiratory complications. The limits of the PP are an increase in operating time, and the difficulty in conversion to thoracotomy in case of massive bleeding. Only one randomized trial between MIE carried out in the PP and open surgery has been published, by Biere et al. [57]. They studied 115 patients randomized to an open esophagectomy or total MIE in the PP and esophagogastric anastomosis in the neck. Pulmonary infections in the first 2 weeks were 29% in the open group and 9% in the minimally invasive group (p = 0.005%). The rate of anastomotic leaks and re-operation was greater for minimally invasive surgery (7% vs. 4%, p = 0.390 and 14% vs. 11%, p = 0.641 respectively), whereas the rate of vocal-cord paralysis was higher in the open group (14% vs. 2%, p = 0.012). The duration of hospital stay was shorter in the MIE group compared with the open group (14% vs. 11 days, p = 0.044).

The goal of surgery is to obtain an R0 resection. Since the advent of minimally invasive surgery there has been a debate as to whether this approach could be similar for open surgery for oncological outcomes. Some studies have focused on the margins of resection, lymph-node retrieval as well as short- and long-term survival. In the meta-analysis of Butler et al., [43] the positive resection margins have been reported in 0–14% of cases. Martin et al. [58] reported that 13.9% (5/36 patients) of patients, who underwent a transthoracic three-stage esophagectomy had involved margins. In the study of Smithers et al., [53] there was no difference in the resection margins between open, total minimally invasive, and hMIE (19%, 14% and 20%, respectively); however in patients referred to surgery alone, the lateral margin involvement was greater in the open group compared with the assisted thoracoscopic group (15% vs. 8%).

LN retrieval during esophagectomy correlates directly with long-term survival, and several studies have confirmed this aspect [59–60] with a possible cutoff of 23 LNs [60]. Case series studies show no differences between open, MIE and hMIE in terms of LN retrieval (MIE vs. open, p = 0.83 and hMIE vs. open p = 0.62) [53]. In the meta-analysis of Dantoc et al., [61] the median (range) number of LNs found in the open group, MIE and hMIE groups was 10 (3–32.8) 16 (5.7–33.9) and 17 (17–17.15) respectively. There was a significant difference between the MIE and open groups (p = 0.032) but not between MIE and hMIE (p = 0.25). The explanation provided by several authors is that the increased visualization of LNs by thoracoscopic methods has led to a greater yield of LNs [54,62]. Despite numerous retrospective studies and meta-analyses, few authors have also evaluated the prognosis of patients who underwent MIE and if this technique gives the same oncological outcomes compared with open surgery. In the study of Dantoc et al., [61] there are no significant differences in survival to 5 years between MIE vs. OE and hMIE vs. OE (p = 0.33 and 0.41, respectively). Thus, based on this meta-analysis, minimally invasive surgery seems to have no advantage or disadvantage in terms of long-term survival. Similar results were reported by Osugi et al. [63] who compared 77 patients who underwent video-assisted thoracoscopic (VATS) esophagectomy vs 72 OE patients who underwent three-stage esophagectomy. Survival at 3 years and 5 years showed no significant differences (70% and 55% and 60% and 57%, respectively).

One of the major technical difficulties of minimally invasive esophageal surgery is intra-thoracic anastomoses, which may explain (at least in part) the choice of some authors to carry out three-stage esophagectomy and anastomosis in the neck even in patients with tumors of the distal esophagus and cardia. It is hard to compare the different studies in the literature with regard to the location (thoracic or cervical) and type of anastomoses (manual or mechanical and endto- end, side-to-side, end-to-side). Maas et al. [64] conducted a review of 12 studies reporting on total minimally invasive Ivor Lewis esophagectomy in which the anastomotic leaks ranged from 0% to 10% and anastomotic stenoses from 0% to 27.5% (Table 1.2). Anastomotic stenoses were more common with the transoral technique. Based on these data, we believe that minimally invasive intrathoracic anastomoses should be undertaken only in controlled trials.

In conclusion, although >20 years have passed since the first MIE, and despite numerous studies, there are many unresolved issues. First, the multiple studies in the literature are, for the most part, retrospective and case series, with a limited number of patients, and are hard to compare with each other with respect to stage of disease, surgical technique, and adjuvant treatments. In addition, most studies have methodological limitations that reduce the statistical significance (e.g., authors at the beginning of their surgical experience may have selected patients with early-stage disease or with a better performance status). For these reasons, MIE, although achieving similar results in terms of oncological outcomes, has not demonstrated a clear advantage with respect to traditional surgery and, even if it is a safe alternative, it cannot be considered the procedure of choice. Second, it is very difficult to carry out randomized trials because, in most cases, it is preferable to choose a tailored surgery centered on patient need and not on a surgical technique that some authors consider better than another. Third, minimally invasive surgery should be totally consistent with the open approach (including surgical indications), so tumors of the cardia and distal esophagus should be approached using the Ivor Lewis procedure. In fact, it is well known that anastomoses in the neck have a higher percentage of fistulas and a greater number of recurrent nerve injuries. Moreover, in esophageal adenocarcinoma, laparoscopy may change the management strategy for up to 20% of patients with occult peritoneal or hepatic metastases. For this reason, carrying out the thoracoscopic stage first could make surgery become palliative. Robotic technology, in which three-dimensional vision and articulated arms facilitate surgical dissection, could have, especially in anatomically confined spaces, an important role in this very challenging type of surgery. We must emphasize that MIE is an advanced procedure requiring knowledge of advanced laparoscopic and thoracoscopic techniques and experience in conventional esophageal surgery. Therefore, it should be undertaken only in centers with vast experience in esophageal surgery.

In the last decade, the incidence of superficial esophageal cancer (carcinoma in situ (Tis) and T1 lesions) has increased as a result of advances in endoscopic techniques that can be used to detect high-grade dysplasia and early esophageal carcinoma [65]. Early esophageal cancer is a localized lesion with a low risk of LN metastasis and a high potential for cure after complete resection.