PROGRESSIVE IMPROVEMENTS IN POSTOPERATIVE OUTCOMES WITH OVER 500 ROBOTIC ASSISTED MINIMALLY INVASIVE ESOPHAGECTOMIES AT A SINGLE INSTITUTION

Background and aims: Several small studies reported high risk of progression to high grade dysplasia (HGD) and esophageal adenocarcinoma (EAC) in Barrett’s esophagus (BE) patients who undergo solid organ transplantation (SOT). However, the major shortcoming of these reports is the lack of a control population. Therefore, we aim to determine the rates of neoplastic progression in SOT patients with BE compared to controls and also the risk factors associated with progression.
Methods: A retrospective case-control study of all patients with a confirmed diagnosis of BE (age ≥18 years) seen between Jan 2000 and Dec 2021 was conducted. Cases were all BE patients who had SOT; controls were all BE patients with no SOT and no immunosuppressant use. Demographics, endoscopic and histological findings, duration of follow-up (in years), history of fundoplication and immunosuppressant use were abstracted. Patients with ≥1 surveillance endoscopies were included to calculate incidence rates of HGD/EAC. Multivariate logistic regression was done to identify the risk factors associated with progression of BE patients to HGD/EAC.
Results: There were 118 cases with SOT (lung =35, liver=35, kidney=35, heart=14, pancreas=2) and 756 controls with no prior history of SOT or immunosuppressant use. Patients with SOT were predominantly younger (p=0.022); male, had lower body mass index (BMI) and active smokers (p<0.001 for all) (table 1). On multivariate analysis, older age, male gender, BE segment length, higher BMI and hiatal hernia were more likely to be associated with progression of BE to HGD/EAC (table 2). BE patients with immunosuppressant medication use were twice more likely to progress to HGD/EAC (p<0.001) (table 2).
Conclusion: Immunosuppressant use increases the risk of neoplastic progression in patients with BE. Therefore, these patients may need more aggressive surveillance. This study has important clinical implications for surveillance in BE patients with SOT.

INTRODUCTION
Screening colonoscopy (SC) is widely accepted and has been shown to decrease the rate of colorectal cancer death. Guidelines and acceptance of screening for Barrett’s esophagus (BE) are less established despite the fact that esophageal adenocarcinoma (EA) remains the fastest increasing cancer in the United States. The aim of this study was to assess the frequency of SC in patients ultimately found to have EA, and to evaluate the presence of symptoms that might have prompted an esophagogastroduodenoscopy (EGD) and potentially earlier diagnosis of the EA.
METHODS
A retrospective chart review was performed to identify all patients who were referred to a single center with esophageal cancer between July 2016 and November 2022. Patients with any histology other than adenocarcinoma were excluded.
RESULTS
There were 221 patients referred with EA. Of these, a SC had been done prior to the diagnosis of EA in 108 patients (49%), 96 men and 12 women. The median age was 66.4 years. A total of 203 SC had been done (range 1-7 per patient). The median interval from SC to the diagnosis of EA was 2.88 years. The highest stage lesion found on SC was colorectal carcinoma in 2 patients and tubulovillous adenoma in 3 patients. There were 36 patients (33.3%) with no findings on SC. At the time of SC, gastroesophageal reflux disease (GERD) symptoms or regular acid suppression medication use was documented in 48.1% of patients (Table). In those with GERD symptoms, the symptoms had been present either life-long or for many years in 69% of patients. Only 19 patients (17.6%) that had a SC had an EGD at any time prior to the diagnosis of EA. In 8 patients, the EGD that found EA was done at the time of SC for anemia, regurgitation symptoms or new-onset dysphagia. A T1 lesion was found in 25% of these patients compared to only 7% of patients that had an EGD separate from the SC (p=0.059). Dysphagia was present at the time of SC in 9 patients, and in 3 patients a SC without an EGD was performed a median of 3 months prior to the EGD that showed EA.
CONCLUSIONS
Nearly one-half of patients ultimately diagnosed with EA had GERD symptoms for many years or were using acid suppression medications regularly at the time of SC, but did not undergo EGD. In 40% of the patients the SC was done within 2 years of the diagnosis of EA. The addition of an EGD at the time of SC in these patients may have allowed early detection of BE or EA.

Background
Transthoracic esophagectomy has historically been characterized as an operation with substantial postoperative morbidity. The application of the robotic surgical platform to enable robotic-assisted minimally invasive esophagectomy (RAMIE) has been prospectively shown to reduce postoperative complications when compared to open resection with thoracotomy. However, RAMIE requires significant institutional investment to fully realize postoperative benefits of minimally invasive resection. We sought to describe improvements in postoperative outcomes over time in our high-volume single-center experience of over 500 RAMIEs.

Methods
Patients undergoing robotic-assisted transthoracic two-field esophagectomy were identified from a prospectively-maintained institutional database (2010-2021). Patients were included if the abdominal portion of the operation was performed open or minimally invasive; all thoracic portions were performed robotically. Primary postoperative outcomes of interest included length of stay (LOS), 30-day pulmonary complication, and 30-day cardiac complication. Cases were separated into cumulative volume quintile (CVQ) by surgical date. Associations between outcomes and CVQ were assessed using regression analysis, as appropriate, with adjustment for clinical factors (age, sex, receipt of neoadjuvant therapy), tumor factors (site, histology, clinical stage), and open vs minimally invasive abdominal portion.

Results
In all, there were 504 RAMIEs identified for study. Median patient age was 66 years (IQR 58-72) and the majority were male (81.9%), were performed for adenocarcinoma (88.1%), and received neoadjuvant therapy (83.9%). Median operative time was 415 minutes, which did not vary by CVQ (p=0.24). Median operative blood loss decreased with increasing CVQ, but not after adjustment for other factors (p=0.79).
The rate of 30-day respiratory complication was lowest (7.0%) in the highest CVQ (cases 404-504), compared to CVQ1 27.7%, CVQ2 16.8%, CVQ3 27.7%, and CVQ4 17.8%, (p=0.001). The association between higher CVQ and reduced rate of respiratory complication remained significant after adjustment for other factors (OR 0.71, p<0.001). Exploratory analysis revealed LOS decreased steadily with cumulative volume (Figure). Median LOS was 10, 10, 9, 9, and 7 days by increasing CVQ (p<0.001) which remained significant in multivariable adjustment (mean -1.2 days per CVQ, p<0.001). No significant association between rate of 30-day cardiac complication and CVQ was observed (CVQ1 28.7%, CVQ2 24.8%, CVQ3 38.6%, CVQ4 27.7, CVQ5 25.0%, p=0.20).

Conclusions
LOS and postoperative respiratory complication decreased significantly over our experience with 500 RAMIEs. The well-established benefits of minimally invasive resection are likely maximized when performed in a high-volume center with experience in optimization of postoperative care.