Minimally Invasive Esophagectomy Presents Better Long-term Survival Rates than Open Procedure

Patients requiring esophagectomy may have greater long-term survival odds with minimally invasive esophagectomy (MIE) than with open esophagectomy (OE), according to findings presented at the 65th Annual Meeting & Exhibition of the Southern Thoracic Surgical Association.

For their study, researchers collected data on esophagectomy patients from 2010 to 2014 with T1-4N1-3M0, adenocarcinoma, or squamous cell histology, in middle or lower esophagus using the National Cancer Database.

“Patients before this period do not have information on their surgical approach; patients after this period do not have survival outcomes,” explained study author Mickey Ising, a surgery resident at the University of Louisville, during his presentation.

Patients were classified by their surgical procedure: robotic assisted minimally invasive (RAMIE), laparoscopic (LMIE), or open esophagectomy (OE). The team compared postoperative and survival outcomes. In total, 7,435 were eligible for inclusion; a disproportionate percentage underwent OE (67.7%) as opposed to MIE (LMIE, 24.9%; RAMIE, 7.4%). Ising noted that this trend begun to shift during the study period. RAMIE and LMIE patients were propensity matched (1:1) to create a collective MIE group, which was matched to OE, resulting in a 1:1:2 matching of RAMIE-LMIE-OE.

After matching, there were 527 RAMIE, 522 LMIE, and 1,044 OE patients. MIE patients had more nodes sampled than OE patients (16 vs 14, P < 0.001) as well as more R0 resections (96.0% vs 93.6%, P = 0.013). The OE cohort had longer postoperative stays (10 vs 9 days, P = 0.04). Short-term survival rates were not different between the groups (30 days, P = 0.893; 90 days, P = 0.894). But median long-term survival was significantly increased after MIE than OE (4.98 vs 3.43 years, P = 0.001). In subgroup analyses, RAMIE and LMIE patients had similar survival rates (P = 0.698).

The study had several limitations. One limitation was the viability of some data points, “specifically tumor size, which was missing in about a quarter of our patients, which is why we didn’t include it,” Ising said. “There is also no information on recurrence or cancer-related mortality, which may better characterize the overall survival.”

Center bias was cited as another limitation. Ising said that “the majority of minimally-invasive patients were at larger centers,” and that different protocols and better long-term follow-up at larger centers may “ultimately lead to increased survival.”

The researchers concluded in their study that “After propensity matching, patients undergoing MIE had better long but not short-term survival than OE […]. This suggests MIE may have potential oncologic benefits over open esophagectomy. This benefit seems to be independent of the use of robotic technology.”

During the discussion portion of the presentation, a question about the quality of the procedure itself in conjunction with the surgeon’s experience was raised by Mara B. Antonoff, MD, FACS, of the Anderson Cancer Center in Houston, Texas.

“One would typically expect that it takes a certain level of expertise and experience to perform minimally invasive esophagectomies, and I suspect that if the surgeons performing these minimally invasive operations also performed open operations, they’d also have superior outcomes, regardless of their technique,” Antonoff said. “To what extent can we state that the MIE itself results in better long-term survival, and how do we know that this not a surrogate for some other mark of expertise among those surgeons?”

Ising agreed that the question of surgeon specificity is challenging to answer with database information, because most databases do not provide such details.

“Even though we can control for multiple other variables, ultimately the surgeons themselves are probably the ones we have the most trouble doing,” he said. “While we can’t control for that, one thing we could do is ‘control’ for surgical quality. We could have matched for R0 resection rates and number of lymph node sample. We haven’t done that [here], but it would be interesting to see what the survival difference would be.”