https://orcid.org/0009-0005-0942-8745
Figures
Abstract
The prognosis of non-distant metastatic early-onset esophageal cancer (EC) patients undergoing surgical treatment remains unclear, this study aims to compare the prognosis of early-onset and late-onset EC. Information on non-distant metastatic EC patients who underwent surgical treatment and were initially diagnosed between 2004 and 2015 was collected from the Surveillance, Epidemiology, and End Results (SEER) database. Propensity score matching (PSM) was used to balance the baseline differences between early-onset and late-onset EC patients. Univariate and multivariate Cox regression analyses were used to calculate hazard ratio (HR) and 95% confidence interval (CI). The Kaplan-Meier method and log-rank test were used to compare the overall survival (OS) between the two groups of patients. Atotal of 5320 EC patients were included, with 571 in the early-onset group and 4749 in the late-onset group. Multivariate Cox regression analysis showed that early-onset EC patients had better OS (HR = 0.732, 95% CI: 0.655–0.819, p<0.001). Using PSM analysis at a 1:1 ratio, we matched 557 early-onset EC patients with 557 late-onset EC patients. After matching, the multivariate Cox regression model still showed a favorable prognosis for early-onset EC (HR = 0.728, 95% CI: 0.630–0.842, p<0.001). Additionally, subgroup analysis indicated that early-onset EC patients had better long-term prognosis. Non-distant metastatic early-onset EC patients undergoing surgical treatment demonstrated better OS outcomes, confirmed by regression analysis and subgroup analysis in the matched cohort.
Citation: Qin L, Tian J, Zhang Y, Yin Y, Dou Z, Chen J, et al. (2024) Comparison of postoperative survival prognosis between early-onset and late-onset esophageal cancer: A Population-based study. PLoS ONE 19(12): e0315391. https://doi.org/10.1371/journal.pone.0315391
Editor: Paolo Aurello, Universitá Sapienza di Roma, ITALY
Received: October 30, 2024; Accepted: November 25, 2024; Published: December 12, 2024
Copyright: © 2024 Qin et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting Information files.
Funding: The author(s) received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Esophageal cancer (EC) is one of the leading causes of cancer mortality worldwide, ranking seventh in incidence and sixth in mortality, it is estimated that in 2020, 1 in every 18 cancer deaths was due to EC [1]. The two most common histological subtypes of EC are esophageal squamous cell carcinoma (ESCC) and esophageal adenocarcinoma (EAC), ESCC is more prevalent in Asian populations, with alcohol consumption and smoking being significant risk factors, while EAC is more common in Caucasians and is closely associated with obesity and smoking [2, 3]. Currently, surgery is considered one of the most important treatments for EC. Despite significant advancements in modern surgical techniques, radiotherapy, and chemotherapy, the 5-year overall survival (OS) rate after surgery has gradually improved but remains below 40% [4, 5].
EC can be classified into early-onset EC and late-onset EC based on the age of onset. Most studies define early-onset EC as a malignant tumor diagnosed before the age of 50 [6–9]. According to international reports, the incidence of malignant tumors in young people, including early-onset EC, has been on the rise globally, since the 1990s, the incidence of early-onset EC in the United States has increased by 50% [10, 11]. However, whether prognosis differs between early-onset and late-onset EC is still debated. A Dutch study reported that although early-onset EC had higher rates of lymph node positivity and distant metastasis, the proportion of patients undergoing surgery was higher compared to older patients, and there was no significant difference in 5-year OS rates [12]. Markar et al study found that early-onset EC patients undergoing esophagectomy had similar staging to late-onset EC, fewer postoperative complications, but no significant difference in OS [13]. Kolb et al found that early-onset EAC had more advanced staging but higher survival rates compared to older patients. However, they did not further analyze whether the differences in survival were due to variations in treatment methods [14].
The factors influencing the survival differences between early-onset and late-onset EC are multifaceted. Compared to late-onset EC, early-onset EC may present at a more advanced stage. Still, patients with early-onset EC typically have better overall health and can tolerate more aggressive treatment regimens [14, 15]. Previous studies included patients with significant baseline differences and the survival differences between early-onset and late-onset EC patients were influenced not only by age but also by confounding factors such as tumor stage and treatment modalities, raising questions about the conclusions [12–14]. Considering that the current treatment for EC is primarily surgical, we used propensity score matching (PSM) to adjust for potential confounding factors and specifically analyze the survival differences between early-onset and late-onset non-distant metastatic EC patients who underwent surgical treatment.
Methods
Patient selection and study variables
The SEER (http://seer.cancer.gov/seerstat/) database is an important cancer data resource maintained by the National Cancer Institute (NCI) of the United States. The SEER database covers approximately 34.6% of the population and represents a wide distribution of different racial, gender, and age groups nationwide. It collects and provides detailed information on cancer incidence and survival rates in the U.S. population. Data can be accessed online using SEER*Stat software version 8.4.0.1.
Inclusion criteria include: (1) Clear pathological diagnosis of esophageal tumor location; (2) First primary tumor; (3) Initial diagnosis between 2004 and 2015; (4) Received surgical treatment. Exclusion criteria include: (1) Demographic characteristics such as age, race, and unknown marital status; (2) Clinical and pathological characteristics such as unknown primary tumor site, tumor size, histological type, grade, and stage; (3) Treatment information including unknown surgery, radiotherapy, and chemotherapy.
The risk factors included in the analysis are demographic characteristics such as age, sex, race, and marital status; tumor characteristics such as histological type, primary tumor site, tumor length, differentiation grade, T stage, and N stage; treatment information such as surgery, radiotherapy, and chemotherapy; and follow-up data such as survival status and survival time. The study endpoint is OS, defined as the time from the beginning of the study until death from any cause. To obtain reliable follow-up data, patients included were diagnosed between 2004 and 2015, with the last follow-up conducted in November 2020, ensuring a minimum follow-up of five years for patients still alive at entry. For histological type, ICD-O-3 codes 8140–8145, 8210, 8211, 8255, 8260, 8261, 8263, 8310, 8480, 8481, 8490, 8574 are defined as EAC; ICD-O-3 codes 8050–8052, 8070–8076, 8083–8084 are defined as ESCC, and all other codes are defined as other types. For tumor site, "Cervical esophagus" and "Upper third of esophagus" are defined as upper segment, "Middle third of esophagus" is defined as middle segment, "Lower third of esophagus" and "Abdominal esophagus" are defined as lower segment, "Overlapping lesion of esophagus" and patients with unspecified tumor site "Esophagus, NOS" and "Thoracic esophagus" are defined as other.
PSM
PSM is a method used to address distribution bias in observational data, aiming to reduce bias caused by confounding factors in observational studies and emulate the effects of randomized controlled trials. PSM calculates a propensity score for each participant to receive a specific treatment or exposure, then matches participants with similar scores to balance confounding factors between groups. By using PSM, differences in variables other than age between early-onset EC and late-onset EC are eliminated, focusing observation on the impact of age on long-term prognosis, and minimizing bias in statistical analysis caused by confounding factors, thereby enhancing the reliability of the final results.
Statistical analysis
Pearson’s chi-square test was used to analyze differences in variable distribution between early-onset EC patients and late-onset EC patients. PSM was used to match early-onset EC patients and late-onset EC patients in a 1:1 ratio. The univariate and multivariate Cox regression analysis were used to examine the relationship between variables and OS, and calculate hazard ratios (HR) and 95% confidence intervals (CI). Kaplan-Meier method was used to visually analyze the OS of different subgroups of patients, and the presence of statistical differences was assessed using the log-rank test. A p-value < 0.05 is considered to indicate statistical significance.
X-tile 3.6.1 (Yale University School of Medicine, USA) was used to determine the optimal cutoff value for tumor size. Statistical analyses in this study were performed using R statistical software (version 4.2.1, http://www.R-project.org). Specifically, the "Matching" package and "tableone" package were used for PSM; the "rio" package was used for data output; the "survival" package and "survminer" package were used for Kaplan-Meier survival curve plotting and log-rank test; the "survival" package, "forestplot" package, and "survminer" package were used for univariate and multivariate Cox regression analysis.
Results
Patient characteristics
A total of 5,320 patients with EC who underwent surgical treatment between 2004 and 2015 were evaluated, with 571 patients in the early-onset EC group and 4,749 patients in the late-onset EC group (Table 1). Before PSM, significant differences (p<0.05) were observed between the two groups in terms of race, marital status, primary tumor site, tumor size, T stage, N stage, chemotherapy, and radiotherapy. Specifically, compared to late-onset EC patients, early-onset EC patients had a higher proportion of Black individuals (7.53% vs. 5.01%, p = 0.001) and a lower proportion of White individuals (85.29% vs. 90.38%, p = 0.001). They also had a higher proportion of unmarried individuals (39.75% vs. 31.65%, p<0.001), more tumors located in the lower esophagus (81.44% vs. 75.70%, p = 0.001), more tumors larger than 50 mm (30.82% vs. 24.32%, p<0.001), higher proportions of T3 (50.79% vs. 57.53%, p<0.001) and T4 stages (8.93% vs. 4.82%, p<0.001), and a higher rate of positive lymph nodes (56.22% vs. 48.64%, p = 0.001). Additionally, early-onset EC patients who underwent surgery were more likely to receive chemotherapy (75.13% vs. 61.91%, p<0.001) and radiotherapy (68.83% vs. 57.04%, p<0.001). There were no statistically significant differences (p>0.05) between early-onset and late-onset EC groups in terms of sex, histological type, and tumor grade. To reduce bias, we performed PSM at a 1:1 ratio, creating a matched cohort of 1,114 patients. After matching, there were no significant differences in the variables between the two groups (p>0.05).
Comparison of survival outcomes between the early-onset and late-onset groups
Before PSM, the 1-year, 3-year, and 5-year OS rates were 84.7%, 60.9%, and 47.4% for the early-onset group, and 79.6%, 53.2%, and 42.4% for the late-onset EC patient group, respectively. Univariate Cox analysis showed significant associations of age, sex, race, marital status, histological type, tumor size, T stage, N stage, chemotherapy, and radiotherapy with OS(p<0.05). Further multivariate Cox regression including these variables identified age as an independent prognostic factor for surgical-treated EC patients, with early-onset EC demonstrating better OS (HR = 0.732, 95% CI: 0.655–0.819, p<0.001), as detailed in Table 2.
To mitigate bias due to differences in variable distribution, we conducted further analysis after PSM. Post-PSM, the 1-year, 3-year, and 5-year OS rates were 85.2%, 58.4%, and 47.7% for the early-onset group, and 78.2%, 49.8%, and 39.1% for the late-onset EC patient group, respectively. Univariate Cox regression analysis indicated significant associations of age, tumor size, tumor grade, T stage, N stage, chemotherapy, and radiotherapy with OS in EC patients(p<0.05). Further multivariate Cox regression confirmed age as an independent prognostic factor for surgically treated EC patients, showing better OS for early-onset EC compared to late-onset EC (HR = 0.728, 95% CI: 0.630–0.842, p<0.001), as detailed in Table 3.
Subgroup analysis after the PSM
Because the sample size of T2 and T4 stage patients was small after PSM, we combined patients into T1-2 and T3-4 groups to avoid bias. After PSM, the baseline characteristics of all subgroups were balanced. Compared to patients with late-onset EC, patients with early-onset EC had better OS whether their T stage was T1-2 or T3-4 (p<0.05), as shown in Fig 1. Stratifying by different N stages, patients with early-onset EC had better OS when the N stage was N0 (p<0.001). When the N stage was N1, there was no significant difference in prognosis between the two groups (p = 0.061), but the early-onset group showed a trend towards improved long-term survival, as detailed in Fig 2. For different histological types, patients with early-onset EAC had better prognosis than those with late-onset EAC (p<0.001). However, when the histological type was ESCC, there was no statistical difference in OS between the two groups (p = 0.078), although early-onset ESCC showed a trend towards improved long-term survival, as shown in Fig 3. Additionally, compared to late-onset EC, early-onset EC showed better OS when receiving chemotherapy (p<0.001) or radiotherapy (p = 0.004), as detailed in Fig 4. Although the survival differences between early-onset and late-onset EC patients were not statistically significant in the N1 (p = 0.061) and ESCC (p = 0.078) subgroups, their p-values were close to 0.05. Kaplan-Meier curves showed that the early-onset group had better OS than the late-onset group over the long term.
A.T1-2, B.T3-4.
A.N0, B.N1.
A. Esophageal adenocarcinoma, B. Esophageal squamous cell carcinoma.
A. receiving chemotherapy, B. receiving radiotherapy.
Discussion
To our knowledge, this is the first large-sample study analyzing clinical characteristics and survival differences between surgically treated non-metastatic early-onset EC and late-onset EC. We found that early-onset EC had more severe tumor infiltration and more lymph node involvement compared to late-onset EC, but they received more aggressive radiotherapy and chemotherapy. Considering that differences in variables such as staging and treatment methods may lead to inaccurate prognostic analysis, we used PSM for precise assessment. The results showed that whether before or after PSM, patients with early-onset EC had better OS. Subgroup analysis after PSM revealed that early-onset EC patients generally had better outcomes compared to late-onset EC patients. In subgroups where there was no statistical difference in OS, the early-onset group also showed a trend toward improved long-term survival.
According to an analysis based on the SEER database, the annual incidence of early-onset EAC increased more than threefold from 1975 to 2015 [15]. A recent study based on the Global Burden of Disease (GBD) found that from 2010 to 2019, the incidence, death rates, and disability-adjusted life years of early-onset EC showed a declining trend in most regions globally [8]. This decline could be due to more focus on risk screening and advances in treatment technologies, though more research is needed. Nevertheless, early-onset EC remains a cancer with high mortality rates. Our study found that although the 5-year OS rate is higher than that of late-onset EC, it remains below 50%.
Our study found that early-onset EC patients present with more aggressive characteristics at initial diagnosis, attributed to several factors. Delayed diagnosis may be one reason. A study found that warning symptoms such as weight loss (40% vs. 57%, P = 0.0003) and dysphagia (29% vs. 42%, P = 0.0067) were less commonly observed in early-onset esophageal cancer patients [6]. For EC or gastroesophageal junction cancer patients, the median time to diagnosis was longer for early-onset cases (100 days vs. 64 days, p = 0.02), which may lead to later staging at diagnosis for early-onset patients [6]. Guidelines recommend endoscopic screening for Barrett’s esophagus (BE) patients with chronic reflux symptoms and three or more risk factors, including male gender, age >50 years, Caucasian ethnicity, smoking, obesity, and a family history of BE or EAC [16]. While the precursors of ESCC are complex, it is certain that both EAC and ESCC screening lack emphasis on younger populations [17].
Marital status is one of the socioeconomic factors, and being married can serve as a surrogate indicator of social support. Our study found that fewer early-onset EC patients were married compared to late-onset EC patients (60.25% vs. 68.35%, p<0.001). Research by Krajc et al demonstrated that married cancer patients are less likely to have metastases and have a lower likelihood of cancer-related mortality compared to unmarried patients [18]. Aizer et al similarly found that married patients have a lower likelihood of distant tumor metastasis and are more likely to receive curative treatment [19]. Additionally, research by Torrejon et al. found that early-onset EC patients have a higher proportion of non-white individuals and lower income, which aligns with our study findings. However, further research is needed to determine whether these differences are due to racial disparities or differences in socioeconomic status [20].
Our study found that early-onset EC patients who had surgery had better survival outcomes than late-onset EC patients, which is consistent with previous studies [12, 14, 21]. In contrast, Markar et al suggest that there is no difference in OS among EC patients of different age groups, possibly due to insufficient sample size [13]. However, prognosis in EC is not only associated with demographic and tumor characteristics but also closely linked to treatment modalities. Our data shows that early-onset EC patients are more likely to receive chemotherapy (75.13% vs. 61.91%, p<0.001) and radiotherapy (68.83% vs. 57.04%, p<0.001). Similarly, previous studies have shown that early-onset EC and late-onset EC patients receive different treatments, including radiotherapy and chemotherapy. These treatment differences may explain the variation in prognosis between the two groups. Additionally, our study found significant differences in race, marital status, tumor size, T stage, N stage, and the distribution of radiotherapy and chemotherapy between early-onset and late-onset EC patients (p<0.05). Cox regression analysis showed that these variables, as risk factors, influence prognosis to varying degrees. To determine whether differences in prognosis between early-onset and late-onset EC are due to the type of EC itself (early or late-onset), we applied PSM. After PSM, there were no significant differences in demographic, tumor characteristics, or treatment distribution between the two groups. Even post-PSM, early-onset EC patients still had a better prognosis than late-onset EC patients. This suggests that after accounting for staging, chemotherapy, radiotherapy, and other factors, early-onset EC patients still have better overall survival.
Apart from demographic characteristics, tumor features, and treatment methods, many factors contribute to the better prognosis of early-onset EC patients who undergo surgical treatment compared to late-onset EC patients. Early mortality and postoperative complications are important factors that must be considered. Markar et al.’s review found that elderly EC patients have higher risks after esophagectomy, compared to younger patients undergoing esophagectomy, elderly patients have increased in-hospital mortality (7.83% vs. 4.21%), along with higher rates of respiratory (21.77% vs. 19.49%) and cardiovascular complications (18.7% vs. 13.17%) [22]. The 5-year OS rate (21.23% vs. 29.01%, p<0.05) and 5-year disease-free survival rate (34.4% vs. 41.8%, p<0.05) are lower in the elderly group than in younger patients [22]. Mantziari et al conducted a more precise analysis, which also showed that compared to younger patients undergoing esophagectomy, elderly patients had higher rates of respiratory complications (20% vs. 16%), cardiovascular complications (15.6% vs. 7.0%), and postoperative mortality (7.9% vs. 3.4%) [23]. Additionally, compared to late-onset EC patients, early-onset EC patients have fewer comorbidities, better Performance Status (PS), and stronger physiological reserves and tolerance to stress [24, 25]. In some studies, age is considered a significant factor in competitive mortality, with advancing age, elderly patients are more likely than younger patients to die from heart failure, respiratory failure, stroke, falls, and other causes [26–28].
This study has limitations. Firstly, while this study included adjuvant therapies in patient prognosis analysis, the SEER database lacks specific chemotherapy regimens and radiotherapy doses. Additionally, information on prognosis-related factors such as PS, postoperative complications, and nutritional status is also lacking. Secondly, this study included patients diagnosed between 2004 and 2015, during which the American Joint Committee on Cancer (AJCC) sixth edition TNM staging system was used. Consequently, patients were categorized only as N0 and N1, and the guidelines at that time differed from recent advancements. The effect of these factors on our results is unclear and needs further investigation. Lastly, this study is based on large-scale retrospective analysis and lacks randomization. Although PSM was employed to minimize confounding factors, prospective data are still needed to further validate our conclusions.
In conclusion, early-onset EC patients without distant metastasis who undergo surgical treatment exhibit better OS outcomes, as confirmed by regression analysis and subgroup analysis conducted in matched cohorts.
References
- 1. Sung H. et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 71, 209–249, (2021). pmid:33538338
- 2. Pennathur A., Gibson M. K., Jobe B. A. & Luketich J. D. Oesophageal carcinoma. Lancet 381, 400–412, (2013). pmid:23374478
- 3. Wang R. H. From reflux esophagitis to Barrett’s esophagus and esophageal adenocarcinoma. World J Gastroenterol 21, 5210–5219, (2015). pmid:25954094
- 4. Cummings L. C., Kou T. D., Schluchter M. D., Chak A. & Cooper G. S. Outcomes after endoscopic versus surgical therapy for early esophageal cancers in an older population. Gastrointest Endosc 84, 232–240.e231, (2016). pmid:26801375
- 5. Markham M. J. et al. Clinical Cancer Advances 2020: Annual Report on Progress Against Cancer From the American Society of Clinical Oncology. J Clin Oncol 38, 1081, (2020). pmid:32013670
- 6. Lumish M. A. et al. Clinical and molecular characteristics of early-onset vs average-onset esophagogastric cancer. J Natl Cancer Inst 116, 299–308, (2024). pmid:37699004
- 7. Vishwanath A., Krishna S., Manudhane A. P., Hart P. A. & Krishna S. G. Early-Onset Gastrointestinal Malignancies: An Investigation into a Rising Concern. Cancers (Basel) 16, (2024). pmid:38672634
- 8. Danpanichkul P. et al. Global epidemiology of early-onset upper gastrointestinal cancer: trend from the Global Burden of Disease Study 2019. J Gastroenterol Hepatol, (2024). pmid:38772743
- 9. Shi M., Tang J. W. & Cao Z. R. Nomograms for predicting survival in early-onset esophageal cancer. Expert Rev Gastroenterol Hepatol 15, 437–446, (2021). pmid:33121317
- 10. Ugai T. et al. Is early-onset cancer an emerging global epidemic? Current evidence and future implications. Nat Rev Clin Oncol 19, 656–673, (2022). pmid:36068272
- 11. Islami F., DeSantis C. E. & Jemal A. Incidence Trends of Esophageal and Gastric Cancer Subtypes by Race, Ethnicity, and Age in the United States, 1997–2014. Clin Gastroenterol Hepatol 17, 429–439, (2019). pmid:29902641
- 12. van Nistelrooij A. M. et al. Treatment and outcome of young patients with esophageal cancer in the Netherlands. J Surg Oncol 109, 561–566, (2014). pmid:24338660
- 13. Markar S. R., Karthikesalingam A. & Low D. E. Outcomes assessment of the surgical management of esophageal cancer in younger and older patients. Ann Thorac Surg 94, 1652–1658, (2012). pmid:23098941
- 14. Kolb J. M. et al. Early-Onset Esophageal Adenocarcinoma Presents With Advanced-Stage Disease But Has Improved Survival Compared With Older Individuals. Gastroenterology 159, 2238–2240.e2234, (2020). pmid:32777286
- 15. Codipilly D. C. et al. Epidemiology and Outcomes of Young-Onset Esophageal Adenocarcinoma: An Analysis from a Population-Based Database. Cancer Epidemiol Biomarkers Prev 30, 142–149, (2021). pmid:33328255
- 16. Shaheen N. J. et al. Diagnosis and Management of Barrett’s Esophagus: An Updated ACG Guideline. Am J Gastroenterol 117, 559–587, (2022). pmid:35354777
- 17. Wong M. C. S. et al. Performance of screening tests for esophageal squamous cell carcinoma: a systematic review and meta-analysis. Gastrointest Endosc 96, 197–207.e134, (2022). pmid:35413332
- 18. Krajc K. et al. Marital status and survival in cancer patients: A systematic review and meta-analysis. Cancer Med 12, 1685–1708, (2023). pmid:35789072
- 19. Aizer A. A. et al. Marital status and survival in patients with cancer. J Clin Oncol 31, 3869–3876, (2013). pmid:24062405
- 20. Torrejon N. V., Deshpande S., Wei W., Tullio K. & Kamath S. D. Proportion of Early-Onset Gastric and Esophagus Cancers Has Changed Over Time With Disproportionate Impact on Black and Hispanic Patients. JCO Oncol Pract 18, e759–e769, (2022). pmid:35544654
- 21. Radkiewicz C., Asplund J. & Lagergren J. Incidence Trends and Survival in Early-Onset Esophagogastric Adenocarcinoma: A Swedish Population-Based Cohort Study. Cancer Epidemiol Biomarkers Prev, Of1-of8, (2023). pmid:37222659
- 22. Markar S. R. et al. Systematic review and pooled analysis assessing the association between elderly age and outcome following surgical resection of esophageal malignancy. Dis Esophagus 26, 250–262, (2013). pmid:22591068
- 23. Mantziari S. et al. Esophageal Cancer in Elderly Patients, Current Treatment Options and Outcomes; A Systematic Review and Pooled Analysis. Cancers (Basel) 13, (2021). pmid:33925512
- 24. Johansson J. & Walther B. Clinical outcome and long-term survival rates after esophagectomy are not determined by age over 70 years. J Gastrointest Surg 4, 55–62, (2000). pmid:10631363
- 25. Lv J. et al. Long-term efficacy of perioperative chemoradiotherapy on esophageal squamous cell carcinoma. World J Gastroenterol 16, 1649–1654, (2010). pmid:20355244
- 26. de Boer A. Z. et al. Impact of Older Age and Comorbidity on Locoregional and Distant Breast Cancer Recurrence: A Large Population-Based Study. Oncologist 25, e24–e30, (2020). pmid:31515242
- 27. Eguchi T. et al. Impact of Increasing Age on Cause-Specific Mortality and Morbidity in Patients With Stage I Non-Small-Cell Lung Cancer: A Competing Risks Analysis. J Clin Oncol 35, 281–290, (2017). pmid:28095268
- 28. Li X., Liu Z., Ye Z., Gou S. & Wang C. Impact of age on survival of patients with pancreatic cancer after surgery: Analysis of SEER data. Pancreatology 18, 133–138, (2018). pmid:29170052