The two most commonly employed open techniques of esophagectomy are transhiatal esophagectomy (THE) and Ivor Lewis esophagectomy (ILE). First described in 1936, THE involves laparotomy with blunt dissection of the esophagus and cervical esophagogastric anastomosis [7
]. ILE involves combined laparotomy with right thoracotomy and intrathoracic anastomosis [8
]. Other approaches include resection via left thoracotomy or left thoracoabdominal approach or three-incision McKeown-type esophagectomy [1
]. In a previous report by Pennathur et al. in 2010, the various open techniques were described along with a few comparative studies on open versus MIE [11
]. For the purpose of this paper, open techniques of esophagectomy include any of the above procedures via completely open approach (without any laparoscopic or thoracoscopic component): THE, ILE, esophagectomy via left thoracotomy or left thoracoabdominal approach; MIE techniques include both total laparoscopic/thoracoscopic THE or ILE and also hybrid procedures with at least one of the approaches being done via either laparoscopy or thoracoscopy. Summarized in a review by Herbella, the various techniques of MIE include any combination of laparoscopy instead of laparotomy, thoracoscopy instead of thoracotomy, and either cervical (THE) or intrathoracic (ILE) anastomosis [13
As with many new novel procedures, the initial publications involving minimally invasive esophagectomy were mostly institutional series. Outcomes from these institutional series included anastomotic leak rates of 4% to 11.7%, pneumonia rates of 7.7% to 16.7%, major morbidity rates of 12.5% to 23%, and operative mortality rates of 0.9% to 6% ().
MIE outcomes in institutional series, case-control studies, and systematic reviews.
Luketich, one of the earlier pioneers of MIE, reported his extensive experience from 1996 to 2002 on 222 patients who underwent MIE for either high-grade dysplasia (n
= 47) or invasive cancer (n
= 175). MIE was successfully completed in 206 (92.8%) patients. Operative mortality was 1.4%. Morbidity included anastomotic leak rate of 11.7%, pneumonia incidence of 7.7%, and recurrent laryngeal nerve injury with vocal cord palsy rate of 3.6% [14
]. Their preferred and most commonly employed approach is combined right thoracoscopic and laparoscopic THE. If significant gastric extension of the tumor is encountered, they prefer to resect more stomach and perform an intrathoracic anastomosis, that is minimally invasive ILE. The same group later described their early experience with minimally invasive ILE in 50 patients from 2002 to 2005 with an operative mortality and leak rate of 6% each [15
Rajan et al. also published a large series of 463 patients in India who underwent minimally invasive esophagectomy from 1997 to 2009. Interestingly, 71 percent of patients had squamous cell carcinoma, and 29% had adenocarcinoma of the esophagus. Operative mortality was 0.9%, and overall morbidity was 16% [16
]. Similarly, Nguyen and colleagues reported a series of 104 MIE procedures performed between 1998 and 2007. Most procedures were minimally invasive THE or ILE (98 of 104) utilizing a circular staple technique.
Major complication rate was 12.5%, and minor complication rate was 15.4%. Anastomotic leak rate was 9.6%, and operative mortality was 2.9% [17
]. Consequently, we reported our results in 105 consecutive patients who underwent MIE utilizing a side-to-side 6
cm linear stapled technique from August 2007 to January 2011. Our mortality was 1% (1/105), and morbidity included 7% transient left recurrent laryngeal nerve injury, 9% pneumonia, 1% wound infection, and 4% anastomotic leak rate [18
]. In a separate study, we studied the effect of neoadjuvant chemoradiation on outcomes after MIE and noted that there were no significant differences in operative blood loss, median operative time, total or individual complication rates, pneumonia, atrial fibrillation, recurrent laryngeal nerve injury, or anastomotic leaks between patients who received neoadjuvant chemoradiation and patients who did not [19
]. Finally, due to the proximity of our Veterans Administration Hospital to the University Hospital, we were able to demonstrate the feasibility of an MIE program at the Veterans Hospital in our initial series of 18 consecutive MIE. There was one (5.6%) 30-day mortality, 1 (5.6%) anastomotic leak, and 3 (16.7%) postoperative pneumonias [27
In 2007, Gemmill and McCulloch published one of the earlier systematic reviews of minimally invasive operations for esophageal and gastric cancer based on an electronic search of the literature from 1997 to 2007. From 188 abstracts reviewed, 23 articles were found on minimally invasive esophagectomies (n
= 1398)—the operations spanning any combination of thoracoscopy or thoracotomy with laparoscopy, hand-assisted laparoscopy, or laparotomy (i.e., MIE or hybrid MIE). Twenty-one of the 23 were case studies, and the remaining 2 were case-matched studies; there were no randomized controlled studies of open versus MIE at the time of this systematic review. For MIE or hybrid MIE, 30-day mortality was 2.3%, combined major and minor morbidity was 46.2%, anastomotic leak rate was 7.7%, and respiratory tract infection rate was 13.2%. The authors stated that while there appears to be substantial literature suggesting the feasibility and safety of minimally invasive surgery for esophageal cancer, the quality of the studies was poor [20
]. Flaws included (1) the predominance of case series—low levels of evidence, (2) lack of valid direct comparisons of open versus MIE, (3) heterogeneity of the studies with regard to the type of MIE or hybrid MIE and, thus, lack of generalizability, (4) selection bias—patients selected for minimally invasive surgery are unlikely to have been representative of the general population of esophageal cancer patients (i.e., earlier stage, smaller tumors, and/or less co-morbid), and (5) publication bias—surgeons with unsatisfactory results may have been less inclined to publish their data. Hence, the authors suggested a prospective nonrandomized cooperative study by surgeons interested in first establishing a consensus on both the appropriate question and the appropriate procedure to be tested against open esophagectomy. The study would allow for evaluation of learning curves, power calculations based on observed treatment effect, and development of quality measures; the study would thus serves as an important and less costly preliminary step before a randomized controlled trial.
Verhage et al. published their results of a systematic review consisting of 10 case-control studies comparing open to MIE. Blood loss for MIE (compared to open esophagectomy) was uniformly lower in all studies, whereas hospital and ICU length of stay, total complication rate, and pulmonary complications were significantly lower with MIE in most studies [21
]. This paper was limited by the heterogeneity of the studies with regards to technique of MIE and both selection and publication bias. A meta-analysis by Nagpal et al., consisting of 12 studies comparing open esophagectomy (n
= 612) and MIE or hybrid MIE (n
= 672), concluded similar findings as noted by Verhage's group. There were no significant differences in 30-day mortality or anastomotic leak rates. Blood loss, ICU length of stay, overall hospital stay, and total morbidity were significantly lower in the MIE group [22
A more recent systematic review by Dantoc et al. comparing open to MIE consisted of 17 case-control studies, and the review showed no significant differences in 30-day survival or 5-year survival rates. Median number of lymph nodes retrieved was significantly higher with MIE versus open esophagectomy (16 versus 10) attributed to better visualization with MIE [23
]. Furthermore, in a meta-analysis by Sgourakis et al., 8 out of 71 screened trials comparing open versus MIE or hybrid MIE were included in the final study (n
= 1,008). It was found that total complications were lower with MIE (odds ratio or OR 1.93, 95% confidence interval or CI 1.08–3.43 for open versus MIE). It should be noted, however, that this comparison was performed with only 3 studies. Anastomotic stricture rates were lower with open esophagectomy (OR 0.11, 95% CI 0.04–0.31), but this comparison was performed with only 2 studies [24
]. Additionally, Biere et al. published their findings of a meta-analysis in which 1 controlled clinical trial and 9 case-control studies were included in the final study (n
= 1,061). Trends were observed in favor of MIE for the following outcomes: major morbidity, pulmonary complications, anastomotic leakage, mortality, length of stay, operating time, and blood loss, but statistical significance was not reached [25
]. The obvious limitation of these meta-analyses is that they consist of primarily nonrandomized and retrospective case-control studies. Ultimately, it was concluded in all three meta-analyses that prospective randomized controlled trials comparing open versus MIE were needed.
Lastly, a large United Kingdom population-based study by Mamidanna et al. analyzed the Hospital Episode Statistics data from April 2005 to March 2010 and included 7,502 esophagectomies, 1,155 (15.4%) of which were MIE—with marked increase in the proportion of MIE (24.7%) performed from 2009-2010. There was no difference between open and MIE groups, respectively, in 30-day mortality (4.3% versus 4.0%) and overall morbidity (38.0% versus 39.2%). Reintervention rate was higher with MIE compared to open (21% versus 17.6%, P
= 0.006) [26
]. Despite the seeming equipoise in outcomes between open and MIE in the above study, there were significant limitations of this population-based study based on an administrative database, as commented in editorials by Rice and Blackstone [28
] and Pennathur and Luketich [29
Based on these large studies on MIE, total complication rates range from 38% to 46.2% and operative mortality rates range from 1.3% to 4.3% (). The numerous studies comparing open versus MIE (case-control studies and meta-analyses or systematic reviews) suggest that oncologic outcome and survival are not significantly different, whereas overall morbidity might be similar or possibly improved with MIE. However, the ultimate message is that better data is necessary to claim the benefits of MIE over open esophagectomy. Consequently, Biere et al. published the study protocol on the “TIME” trial or traditional invasive versus minimally invasive esophagectomy, which will be the first prospective, multicenter, randomized study comparing open versus MIE [30
]. In this study proposal, patients will be randomized to either traditional open Ivor Lewis esophagectomy or MIE. For the MIE group, THE is the preferred method while the Ivor Lewis esophagectomy (i.e., an intrathoracic anastomosis) will be performed if more of a gastric resection margin is necessary based on location of the tumor.