The long‐term survival of esophageal cancer in elderly patients: A multi‐center, retrospective study from China

Abstract Background Esophageal cancer (EC) often occurs in the elderly, and approximately 40% of patients are 70 years or older. To investigate the long‐term survival of EC in elderly patients, to provide a theoretical direction for better management and predicting survival of EC based on the hospital‐based multi‐center study in China. Methods The study was conducted in 18 hospitals including 6 provincial hospitals, 8 municipal hospitals, and 4 county hospitals. We extracted information from medical record homepage, records of admission and discharge, and pathological diagnosis reports from the medical record department of the elderly patients at 70–84 years old to obtain the 3‐year and 5‐year overall survival (OS), and main associated factors, and to analyze the current therapeutic effect of different treatment options for elderly patients. Results The 3‐year and 5‐year OS rate of the 1013 elderly patients was 44.8% and 32.8%, respectively. Their median survival was 28.00 months. The median survival of patients with squamous cell carcinoma was longer than that of other pathological type (squamous vs. other types: 31.00 vs. 20.00 months, p = 0.018). The median survival of patients with surgery only or combined therapy was longer than that of radiotherapy, chemotherapy, and no therapy (surgery only vs. combined therapy vs. radiotherapy vs. chemotherapy vs. no therapy: 56.00 vs. 33.00 vs. 26.00 vs.18.00 vs. 16.00 months, p < 0.001). The 5‐year OS rate of patients with highly differentiated cancer was higher than that of medium differentiated and poor differentiation/undifferentiated. In multivariate analysis, the older ages, pathological stage, were independent prognostic risk factors for poor EC survival. Treatment method was independent protective factors predictive of a good EC OS. Conclusions The survival rate of the elderly EC patients was still low in China. Age, therapy method, and pathological stage were mainly associated with the survival rate of EC in elderly patients.


| INTRODUCTION
Esophageal cancer (EC) is one of the high incidences of digestive tract malignant tumor, which ranks the sixth in the mortality rate in the world 1,2 and the fourth in China. 3 Although the medical technology has made great progress and the study about EC is more and more accurate, the status quo in the treatment of EC is still unsatisfactory in developed and developing countries. In addition, the incidence trend of the disease is getting older and older, reaching the peak between 75 and 80 years old. 4 It often occurs in the elderly, and approximately 40% of patients are 70 years or older.
Esophageal cancer is a worldwide health problem with a high mortality rate due to its natural history and the common diagnosis in advanced stages, especially on elderly patients. 5 Therefore, it is vital to early detect EC in lower stage, which significantly improve survival. Esophageal squamous cell carcinoma is the most common histologic type worldwide including China. However, currently, adenocarcinoma subtype is dominant in the United States, Australia, the United Kingdom, and Western Europe (Finland, France, Norway), and squamous carcinoma has converted to the second place. Risk factors for squamous carcinoma mainly included gender and race, smoking, alcohol, genetics, obesity, gastroesophageal reflux disease and Barrett's Esophagus, diet, and nutrients such as chewers of areca nut and hot drinks. [6][7][8][9][10][11][12] At present, there are great challenges in management and treatment of elderly patients with EC. A large amount of evidence recommended surgically based neoadjuvant chemoradiotherapy or radical concurrent chemoradiotherapy as the main treatment for advanced EC. 13 For elderly patients, various functions are far inferior to those of young patients. Whether the above treatment methods are completely suitable for elderly patients still needs to be explored and discussed. The survival prognosis of patients after treatment is an important index for evaluating the therapeutic effect.
This study was based on elderly patients with EC were analyzed retrospectively, and studies the main related factors influencing the prognosis of survival in elderly patients, as well as the analysis of the current for elderly patients with treatment effects of different choice. We aimed to investigate influencing factors of the survival of EC in elderly patients, in order to provide a theoretical direction for better management and predicting survival of EC based on the hospital-based multi-center study in China.

| Data source
Data were collected from 18 hospitals of five regions including Beijing, provinces of Hebei, Henan, Zhejiang, Hubei, and Guangdong. These hospitals consisted of six provincial hospitals, eight municipal hospitals, and four county hospitals. All these hospitals are covered by population-based cancer registries, and there are complete data of mortality and survival. Trained investigators excerpted information from homepage of medical record, records of admission and discharge, and pathological diagnosis reports from the department of medical records. Information included baseline characteristics: gender, age at diagnosis, marriage status, occupations, height and weight at admission, blood type, history of smoking, drinking, and family history of cancer (FHC); tumorrelated information: pathology, subsite, differentiation, and tumor node metastasis (TNM) stage of EC patients; treatment-related information: surgery, radiotherapy, chemotherapy, chemoradiotherapy, combined therapy. For our study, we extracted the information of 1013 EC with highly differentiated cancer was higher than that of medium differentiated and poor differentiation/undifferentiated. In multivariate analysis, the older ages, pathological stage, were independent prognostic risk factors for poor EC survival. Treatment method was independent protective factors predictive of a good EC OS.

Conclusions:
The survival rate of the elderly EC patients was still low in China.
Age, therapy method, and pathological stage were mainly associated with the survival rate of EC in elderly patients.

K E Y W O R D S
clinical risk factors, elderly patient, epidemiology, esophageal cancer, survival patients aged from 70 to 84 years old and diagnosed between 1 January 2011 and 31 December 2013 for survival analysis in this study.

| Criteria of inclusion and exclusion
All included subjects met the following criteria: (1) all registered citizens diagnosed as EC; (2) patients diagnosed between 1 January 2011 and 31 December 2013; (3) patients who got first diagnosis and first treatment in these hospitals; (4) availability of complete clinical records.
Patients meeting the following criteria were excluded: (1) multiple primary or metastatic cancer; (2) nonlocal household records; (3) patients who underwent therapy (Surgery, radiotherapy, or chemotherapy) prior to admission to the above-mentioned hospitals.

| Exposure ascertainment and pathologic evaluation
According to the occupation information on the homepage of the medical records, occupations were divided into three categories, enterprise personnel, production and service personnel and farmers.
Body mass index (BMI) was calculated using the height and weight at admission, and according to Asian standard, BMI was categorized into three groups: underweight (<18.5 kg/m 2 ), normal weight (18.5-23 kg/m 2 ), and overweight or obese (≥23 kg/m 2 ).
According to the admission records, patients were defined as smokers if smoking ≥100 cigarettes in one's lifetime. Drinkers were defined as individuals who drinking any alcoholic beverage one or more times a week. FHC was defined as any cancer among relatives with first-degree relatives or second-degree relatives. The first-degree relatives included father, mother, and siblings; the seconddegree relatives were restricted in grandparents, uncle, and aunt. The pathological types and differentiation of all patients were confirmed by pathological examination. EC patients were staged according to the 2007 American Joint Committee on Cancer (AJCC) TNM classification. Classification of tumor sites and pathological types were according to ICDO-3 coding.

| Follow-up
Patients were diagnosed between 1 January 2011 and 31 December 2013, and followed up until 31 December 2017. Both passive and active follow-up methods to ascertain the vital status of patients were used in every hospital. The staff of hospital linked the patient records and death records of the population-based cancer registries on the basis of identifiable information. Also, patients were followed up every 3 months by active methods in every hospital.
Survival time of patients who dead of EC was calculated from the time of diagnosis of EC until patients' death. Survival time of patients who were alive and censors was calculated from the time of diagnosis of EC until the last contact. Eventually, a total of 5283 patients with EC were included in the study. Up to 31 December 2017, there were 1333 survival cases, 3157 death cases, and 2575 cases of death due to EC. Seven hundred ninety-three cases lost to follow-up. Follow-up rate was 85.0%. The study procedures have been described in detail elsewhere. 14 We extracted the information of 1013 EC patients aged from 70 to 84 years old for survival analysis in this study.

| Quality control
We standardized the format of the variables and applied automatic checking procedures to assess the data quality of each hospital.

| Checking of medical records
After investigation, we randomly selected 10% cases of all the 18 hospitals according to the table of random number methods. The medical records were checked by two other staff members, and the accuracy was all above 99.5%.

| Sampling of medical records
Medical records meeting the inclusion criteria were ranked by months, and these records were numbered according to the table of random number. Starting from any number and reading it in the same direction. The direction can be downward or upward, or left or right. In these numbers, the repeated number was excluded. Evennumbered medical records were included in this cohort study, and odd numbers were excluded.

| Following up of these medical records
All hospitals followed the patients according to the above follow-up methods and to minimize the loss of follow-up and ensured the reliability of the data. The follow-up rate remained above 85%.

| Statistical analyses
The statistical analyses were performed using the IBM SPSS Statistics software (version 25.0; SPSS, Inc.). Qualitative variables were described by frequency and percentage. The forest map of associated factors with survival was drawn by the R software (version 4.1.1; The R Foundation for Statistical Computing). Rate of 3-year and 5-year overall survival (OS) and median survival time were analyzed by Kaplan-Meier method. The Cox's proportional hazard regression model was used to calculate hazard ratios (HR) and 95% confidence intervals (95% CI) of the univariate and multivariate survival analyses for mortality risk. Variables that were associated with the outcome variables at p less than 0.10 in the univariate logistic analysis model were included in the multivariate Cox's proportional hazard regression model. The statistical significance threshold was p < 0.05.
Tumor characteristics include subsite, pathological types, differentiation, and TNM stage. As shown in Table 1, subsite was mainly in thoracic part (579, 57.2%) and the main pathological type was squamous cell carcinoma (826, 81.5%). 508 (50.1%) patients had stage III/IV at diagnosis. Moderately differentiated status was the most common histological type which accounted for 19.2% (194/1013) in the elder EC patients.

| OS rate and median survival time in elderly EC patients
The 3-year and 5-year OS rate of the 1013 elderly patients was 44.8% and 32.8%, respectively (

| Univariate and multivariate analysis of EC survival in elderly patients
The factors associated on univariate analysis with elderly EC survival are shown in

| DISCUSSION
This study found that age, tumor stage, and treatment were independent risk factors affecting survival time of patients. The survival time of patients aged over 75 years is significantly shorter than that of patients aged under 75 years old. Elderly patients are often accompanied by organ dysfunction, decreased immunity, and various chronic diseases. When clinicians choose treatment strategies according to pathological characteristics and clinical stages of tumors and so on, the basic physiological status of patients gave enough attention and the significance of different treatment methods to the prognosis, survival, and quality of life of patients should also be considered. In the clinical treatment, it is common to choose relatively conservative treatment strategies when patients are too old, which may cause that some older patients do not receive more suitable treatment, thus affecting the survival time of patients. In this study, age is negatively correlated with survival time. Firstly, the patient's own physical condition is taken into account, and secondly, the choice of clinical treatment also affects the survival time of patients because they are too old.
The best treatment for elderly EC patients has always been controversial. Most scholars believe that surgery has great trauma to elderly patients and chemotherapy has many toxic and side effects, so radiotherapy is recommended for elderly patients with EC. 15,16 With the rapid development of minimally invasive surgical techniques and anesthesia techniques, the feasibility and safety of  surgical treatment for elderly EC patients have been confirmed. 17,18 Lymph node dissection is an important step in EC surgery. Previous studies showed that the more lymph nodes dissected, the more accurate the postoperative staging was, and the better long-term prognosis of patients had. 19,20 However, extensive lymph node dissection lengthens the operative time and increases intraoperative bleeding, resulting in increased perioperative complications and mortality. 21 Elderly patients with EC have short life expectancy, poor postoperative recovery ability and high risk of perioperative death. 22 Therefore, surgeons should consider the pros and cons of lymph node dissection in elderly patients.
Our study also suggests that surgery and combination therapy may have a significant survival benefit in elderly patients with EC. A large number of studies have basically confirmed that surgically based combination of radiotherapy and chemotherapy has significant efficacy for young patients with EC or patients with good physical fitness. 23,24 For advanced EC, some studies have advocated the use of radiotherapy alone or combination of radiotherapy and chemotherapy. 25 A recent study in Japan showed that the 5-year OS rate after esophageal resection for cancer in patients older than 70 years was similar to that in nonelderly patients, 26 better than those of chemoradiotherapy alone. 27 The surgery compared to nonsurgical treatment increased median OS among elder patients from 5 to 15 months in a European study. 18 However, for elderly advanced EC patients, the poor physical quality, declined organ function, reduced tolerance of treatment and late-onset characteristics of EC can affect the choice of treatment methods and treatment effect, which need more sample size and the research of higher quality to explore the suitable treatment methods in elderly EC patients.
The 5-year survival rate of early EC is improved by more than 90% after the comprehensive therapy of surgery, radiotherapy, and chemotherapy, but the 5-year survival rate of middle and advanced EC is not optimistic. EC in the elderly is mostly in the middle and late stage, with many complications, poor physique, and poor efficacy. Most patients seek medical treatment after obvious symptoms. After comprehensive treatment, the efficacy is also poor, and most patients die within 1 year. In this way, it is expected to strengthen the elderly tumor survey work in the future, early detection, early diagnosis, early reasonable treatment, in order to improve the survival period of the elderly EC.
In our study, it was found that the 5-year survival rate of elderly patients with EC stage I was 53.4%, and the median survival time was 64 months while the 5-year survival rate of EC stage IV was 26.5%, and the median survival time was 22 months. In our study, the 3-year OS rate was 71.7% among Stage I patients, highly similar to 71.6% in metropolitan areas in Japan, which implies the importance for the detection and timely treatment for EC. 28 TNM staging of EC is a general comprehensive judgment of the stage in the development of EC, which integrates three indicators (Tumor size, lymph node, metastasis) that have an important impact on the prognosis of EC patients, so it is an important factor affecting the treatment and prognosis of patients with EC. 5,29 The earlier the tumor stage is, the higher the surgical resection rate will be, the better the treatment effect will be. Therefore, early detection, early surgical resection, and comprehensive treatment are the main measures to improve the prognosis of esophageal carcinoma. The prognosis of EC is highly correlated with tumor stage, which suggests the importance of early detection of EC with lower stage and early diagnosis and treatment to prolong the survival time.
The occurrence of EC is related to environment, diet and lifestyle, poverty. Low nutrient, high-calorie drinks, and food can increase the incidence of EC, [30][31][32][33][34] and smoking and excessive drinking are also one of the important risk factors of EC. 35,36 The occurrence of EC is also related to the polymorphism of genetic loci, and thus, EC has genetic susceptibility. 37 EC in China has obvious regional distribution characteristics of high incidence, mainly distributed in Taihang  areas. 38,39 The experience in the prevention and treatment of EC shows that the key strategy to EC prevention and treatment is early detection, early diagnosis, and early treatment. The combined operation techniques of endoscopic screening, iodine staining of esophageal mucosa and indicative biopsy are the most practical and effective methods at present. 40 Wei et al. 41 conducted a 10-year follow-up study on the study cohort of more than 45,000 people in Cixian County, Hebei province, a high-incidence site of EC, and found that endoscopy as an early screening method for EC can effectively reduce the morbidity and mortality of esophageal squamous cell carcinoma. Since 2005, China's Ministry of Health has launched a population-based endoscopic screening program in 11 high-risk areas, which has been expanded to cover more than 110 cities and counties in 29 provinces in China. The implementation results of the program show that EC screening has a high detection rate, early diagnosis rate, and treatment rate, with good performance. 40 Therefore, it is very necessary to strengthen the early detection, early diagnosis, and early reasonable treatment of EC, in order to improve the survival rate of EC in the elderly, especially on elderly population. However, our study also had some limitations. Firstly, in the study, different levels of diagnostic evaluation and treatment in different hospitals may have influenced the results. In future studies, we will investigate more cases in city and county hospitals. Secondly, smoking, alcohol consumption status, and other variables associated with EC are self-reported, which led to recall bias. As a result, people who have never smoked may be misclassified, especially those who have been ever smokers and stopped smoking for less than 15 years. In addition, the proportion of the missing data of some variables such as differentiation status was relatively high, but our previous study had verified that it was within a reasonable range. 14 Further more detailed data are required to increase the precision of the effect for these exposures and explored more factors related to EC survival.

| CONCLUSIONS
The survival rate of the elderly EC patients was still low. Age, therapy method, and pathological stage were mainly associated with the survival rate of EC in elderly patients. This study revealed important survival information of elderly EC patients in China and provided helpful information for their clinical management and early detection.