Impact of body mass index on perioperative and oncological outcomes in elderly patients undergoing minimally invasive McKeown esophagectomy for esophageal squamous cell carcinoma

Abstract Background The association between elevated body mass index (BMI) and perioperative and oncological outcomes among elderly patients undergoing minimally invasive McKeown esophagectomy (MIE) remains unclear. Methods We performed a single‐center retrospective analysis of 526 consecutive patients aged 65 years or older who underwent MIE for esophageal squamous cell carcinoma (SCC) between January 2016 and December 2019. Two groups were stratified by BMI: normal (18.5 ≤ BMI < 24 kg/m2) and elevated groups (BMI ≥ 24 kg/m2). A 1:1 propensity score matching (PSM) analysis was used to compare perioperative and oncological outcomes between the two groups. Results A total of 480 elderly patients were eventually enrolled, with a mean age of 70.2 years (range: 65–87), and 185 patients were eligible for elevated BMI, with a mean BMI of 26.3 ± 1.9 kg/m2. Compared with the normal BMI group, the elevated BMI group had prolonged operation time (261.7 ± 57.2 vs. 278.9 ± 62.7 mins, p = 0.002) and increased incidence of intraoperative hypoxemia (12.2% vs. 21.6%, p = 0.006). The differences in intraoperative estimated blood loss, transfusion, new‐onset arrhythmia, and conversion rates and postoperative outcomes regarding pulmonary and surgical complications, intensive care unit and 30‐day readmissions, the length of hospital stay, and oncological outcomes regarding R0 dissection, and the number of dissected lymph nodes between two groups were comparable. After a 1:1 PSM analysis, there was no significant difference in both perioperative and oncological outcomes between two groups. Conclusions Among elderly patients undergoing MIE for esophageal SCC, there was insufficient evidence to demonstrate that elevated BMI could increase perioperative and oncological adverse outcomes.


| INTRODUCTION
In recent decades, the prevalence of overweight and obesity has shifted dramatically, 1,2 and metabolic diseases associated with elevated weight such as hypertension, diabetes mellitus, and hyperlipidemia have also increased gradually, [3][4][5] leading to a worldwide health problem. In addition, with the aging of society, the special attributes of elderly surgical patients, such as frailty, cognitive decline, impaired preoperative lung function, and tissue fragility, make the effect of elevated weight on perioperative outcomes more complicated. [6][7][8] Indeed, the rate of weight gain in older patients receiving esophageal cancer surgery has increased over the past few decades. 6 And the ultimate goal of esophageal cancer surgery is to be accompanied by therapeutic resection to ensure short -and long-term prognosis. 9,10 However, the majority of patients undergoing radical resection of esophageal cancer have a high rate of adverse perioperative complications and associated longer hospital stay. 11,12 As an emerging surgical approach, minimally invasive esophagectomy (MIE) is designed to improve poor outcomes compared to traditional open esophagectomy (OE). Several previous studies have demonstrated that MIE could reduce estimated blood loss (EBL) and postoperative complications, shorten the length of hospital stay, and have similar early oncologic outcomes. 13,14 Up to now, little work is being done exploring the effect of elevated body mass index (BMI) on perioperative and oncological outcomes in elderly patients treated with MIE. In the current study, by reviewing a large sample of prospectively collected data, we attempted to evaluate the association between elevated BMI and perioperative and oncological outcomes in MIE.

| Study design and patients
From January 2016 to December 2019, we performed a monocentric retrospective analysis based on a prospectively collected database, including 526 consecutive patients aged 65 years or older who underwent minimally invasive McKeown esophagectomy (MIE) for esophageal squamous cell carcinoma (SCC). Excluded patients were listed in the flow diagram ( Figure 1). A total of 480 elderly patients were eventually enrolled. The Ethics Committee of Shanghai Chest Hospital approved this study (IS21121) and waived the need for informed consent.

| Anesthesia protocol
All patients were routinely monitored by electrocardiogram, pulse oximetry, and capnography. Catheterization of the radial artery and right internal jugular central venous were used to monitor invasive blood pressure (IBP). Patients received lung-protective ventilation (LPV) strategies, including low-tide ventilation based on ideal body weight (≤8 mL/kg), PEEP = 5 cmH 2 O, lung recruitment, and maintenance of airway pressure < 30 cmH 2 O, and were located in the lateral position required for surgery. After the operation, all patients received patient-controlled analgesia (PCA) pump, including sufentanil1.0 μg/kg + desoxocin 0.4 mg/kg.

| Technique of operation
All patients underwent robot-assisted MIE (RAMIE) or thoracoscopic-assisted MIE (TAMIE) with two-or three-field lymphadenectomy by the same group of thoracic surgeons. The procedure of MIE involved the thoracoscopic movement of the esophagus followed by laparoscopic or laparotomy to establish a gastric tube and cervical esophagogastric anastomosis. The choice of the two approaches was completely based on the chief surgeon's preference after consideration of preoperative evaluation, operative planning, patient benefit, and surgical experience. Patients received either hand-sewn or circular anastomosis esophagogastric anastomosis in this study.

| Data collection
Perioperative data were prospectively pooled from our medical record system, including patient's baseline and intraoperative characteristics, intra-and postoperative complications regarding hypoxemia, pulmonary complications, surgical complications (EBL, transfusion, new-onset arrhythmia, conversion to thoracotomy, anastomotic leakage, chylothorax, and reoperation), intensive care unit (ICU) and 30-day readmissions, the length of hospital stay (LOS), and oncological outcomes (R0 dissection and the number of dissected lymph nodes).

| Definition
Postoperative pulmonary complications (PPCs) were defined based on the European Perioperative Clinical Outcomes (EPCO), 15 including atelectasis, pulmonary infection, and respiratory failure. Perioperative new-onset arrhythmia included incidents of atrial fibrillation (AF) and atrial flutter based on the 2014 Guidelines of the American Association of Thoracic Surgeons (AATS). 16 By referring to the Guidelines for Prevention and Control of Overweight and Obesity in Chinese Adults, 17 two groups were stratified by BMI status: normal (18.5 ≤ BMI < 24 kg/ m 2 ) and elevated groups (BMI ≥ 24 kg/m 2 ).

| Statistical analysis
Continuous variables were compared between normal and elevated groups using Two independent sample ttest or Mann-Whitney U test. Chi-square test or Fisher exact test, depending on the sample size, were used to compare categorical variables. A 1:1 propensity score matching (PSM) 18 analysis with a caliper size of 0.05 was used to lessen the selection bias and other potential confounding effects. All pre-, intraoperative variables, and surgical years were included in the PSM. Standardized mean difference (SMD) between two cohorts on all covariables before and after matching was calculated, with differences of <10% indicating adequate balance in the cohort. Statistical analysis was conducted using the SPSS 26.0 software (IBM Corp.  (Figure 1). The BMI distribution of all enrolled patients were depicted in Figure 2. Patients with elevated BMI had a higher incidence of hypertension (24.9% vs. 14.9%, p = 0.007) and a lower incidence of chemoradiotherapy (8.1% vs. 15.6%, p = 0.017), and better preoperative pulmonary function (FEV 1 / FVC, 101.0 ± 10.8 vs. 98.5 ± 11.0, p = 0.016; DLCO%, 98.2 ± 22.0 vs. 94.4 ± 19.8, p = 0.054) when compared with their counterparts (Table 1). Additionally, patients with elevated BMI required prolonged operative time (278.9 ± 62.7 vs. 261.7 ± 57.2 mins, p = 0.002) compared with those with normal BMI (Table 2).

| DISCUSSION
A total of 185 elderly patients who underwent MIE for SCC were eligible for elevated BMI. This study found that elderly patients with elevated BMI had similar rates of perioperative complications and comparable oncological outcomes compared to patients with normal BMI. Thus, elevated BMI in elderly patients should not be a hindrance to preoperative evaluation, risk stratification, and surgical planning during MIE.
The World Health Organization has recommended BMI thresholds for underweight (<18.5 kg/m 2 ), normal weight (18.5-24.9 kg/m 2 ), overweight (25-29.9 kg/m 2 ), and obesity (>30 kg/m 2 ) to predict risk for all cancer types and non-cancer diseases. However, whether the above criteria applied to Asian populations remains controversial. 17,19,20 The elevated BMI classification in this investigation by referring to the Guidelines for Prevention and Control of Overweight and Obesity in Chinese Adults, which may be more suitable for the Chinese population. 17 MIE has been established to improve perioperative poor outcomes with regard to the standard open approaches. 13,14 Presumably, the proportion of elderly patients with elevated BMI undergoing MIE will constantly increase in the future. 6 Intuitively, it seems that  preoperative comorbidities related to elevated BMI, impaired pulmonary function, prolonged operative time, tissue fragility and reduced mobility should be associated with an increased risk of complications among elderly patients. Therefore, it is mandatory to fully understand the impact of elevated BMI on perioperative outcomes in elderly patients.
In terms of intraoperative complications, before a 1:1 PSM, patients with elevated BMI had a high rate of hypoxemia and prolonged operative time and developed comparable EBL, transfusion, new-onset arrhythmia, and conversion rates compared to normal-BMI patients, but none of these differences were significant after matching. Our previously published literature echoed these results and showed that elevated BMI was not associated with high rates of intraoperative conversions and new-onset arrhythmia. 21,22 Similarly, Salem 23 and Kilic 24 have evaluated the impact of BMI on perioperative clinical outcomes after MIE and open esophagectomy, respectively, and found that elevated BMI was associated with longer operative time but not with a significant increase in EBL, despite differences in baseline characteristics among these studies.
In this study, no difference was observed between patients with normal and elevated BMI for pulmonary and surgical complications, ICU and 30-day readmissions, and LOS. Several other studies also showed no higher incidence of complications following esophagectomy with increased BMI. [23][24][25][26] An analysis of the Nationwide Inpatient Database in Japan demonstrated that BMI showed U-shaped dose-response associations with mortality, major complications, and multiple complications. 27  T A B L E 6 Perioperative complications and oncological outcomes stratified by BMI after a 1:1 PSM Our investigation also assessed the effects of elevated BMI on oncological outcomes regarding R0 dissection and the number of dissected lymph nodes after MIE in elderly patients and showed that the difference was not significant between patients with normal and elevated BMI. And the results were consistent with other published studies. 26,29 Besides, Sachdeva and colleagues' research using STS General Thoracic Surgery Database also indicated that R0 resection or lymphadenectomy did not differ among five BMI groups. 30 Given the consistency of these findings, we tend to believe that neither older age nor elevated BMI may be independent factors affecting oncological outcomes.
Potential defects of our study include as follows. First, as a retrospective study based on a prospectively collected database, it has inherent design biases. Besides, this study did not further subdivide elevated BMI into overweight and obesity, as only 30 patients enrolled were considered obese. Second, due to the limited granularity of postoperative care data, some poor outcomes such as pain control and other surgical complications could not be pooled in this study. Third, the relationship between elevated BMI and long-term prognosis following MIE in elderly patients needs further investigation.

| CONCLUSIONS
By performing a single-center retrospective study of 480 elderly patients receiving MIE for esophageal SCC, our study found that elevated BMI did not increase perioperative adverse complications and oncological outcomes. These data contribute to the increasing body of evidence that elevated BMI in older patients should not exclude candidates for MIE for esophageal SCC.