Study design and population
Oesophageal center of the First Affiliated Hospital of Nanjing Medical University is a regional provincial professional esophageal cancer treatment center, in which esophageal patients underwent surgical and/or multidisciplinary therapy. Between January 1, 2018 and December 31, 2018, consecutive patients from the esophageal center with esophageal cancer were retrospectively included if they underwent R0 transthoracic oesophagectomy. The Institutional Review Board of the First Affiliated Hospital of Nanjing Medical University approved the study protocol. This study followed the Declaration of Helsinki. Due to the nature of this retrospective study, the Institutional Review Board waived the need for patient consent.
We identified retrospectively 1-year disease-free survivors, and excluded patients who had an early recurrence of disease, postoperative death at the first 12 months after oesophageal cancer surgery [9]. We also excluded the patient with incomplete resection (R1–R2). Additional exclusion criteria were incomplete surgical records and missing follow-up outcome data for identification of postoperative nutrition dysfunction.
At last, 145 patients were included in this study with 12-month disease-free survivors. The patient was regard as free of disease when postoperative comprehensive examination results were favorable at least including a full-body scan and esophageal endoscopy.
Data Collection
According to the case information system data of the First Affiliated Hospital of Nanjing Medical University, patient demographic data (age, gender, height, weight), and tumor tumour characteristics (pathological type, clinical stage, tumor location), and surgical procedures (surgical approach) were all collected. Laboratory, radiology, and anesthesiologists and follow-up staffs also assisted in supplementing and perfecting the data. Follow-up data was collected through patient outpatient and inpatient medical records or telephone Internet. The primary endpoints were death, recurrence, metastasis, and weight changes. Clinical and pathological staging was estimated following the system of the International Union Against Cancer [10].
Procedure
According to the protocol of our institute, the standard oesophagectomy included en bloc excision with 2- or 3-field lymph node removals at their discretion of surgeons, gastric tube reconstruction and oesophagogastric anastomosis using the stomach in all patients as described previously [11,12]. Specific surgical procedures were performed according to our previously reported articles [13].
Body mass index
In the present study, BMI was calculated four times during the study period :(1) at the onset of any pre-treatment symptoms prior to treatment (pre-treatment BMI); (2) Before and during surgery (preoperative BMI); (3) Six months after surgery (6-month BMI) and (4) 12 months after surgery (1 year-BMI) [9].
Weight loss assessment
Denutrition was defined as a weight loss ≥15% measured as (weight at the onset of symptoms before any treatment in kilograms - weight at 1 year after oesophageal cancer surgery in kilograms)/ weight at the start of symptoms without any intervention in kilograms. Patients were divided into groups who experienced a weight loss < 15% of the pretreatment body weight and those with a weight loss ≥ 15% of the pre-treatment body weight [9].
Statistical analysis
Continuous data were presented as mean (SD) or median (IQRs) and compared using a t-test or Kruskal-Wallis testing according to distributed characteristics, and categorical data were presented as percentages (%) and compared using χ2 testing. The odd ratio (ORs) of factors of 1-year body weight loss ≥ 15% and 95% confidence intervals (CIs) were analyzed by binary logistic regression models. Baseline adjustments for age at surgery, sex, BMI, histology, tumour stage, tumour location, neoadjuvant treatment, jejunostomy, complications, and surgery mode. All covariates showing relative strong associations (p value <0.1) with 1-year body weight loss ≥ 15% in univariate analysis were modelled together to investigate independent risk factors of1-year body weight loss ≥ 15% using multivariate logistic regression. Statistical analyses were performed by R software (version 3.2.0). P <0.05 were considered statistically significant.