Abstract
Objective:
To improve understanding of sex differences in clinicopathologic characteristics, treatment and outcomes between male and female patients undergoing esophagectomy for esophageal cancer.
Summary Background Data:
Esophageal cancer is a male predominant disease, and sex has not been considered in previous studies as an important factor in diagnosis or management. Sex differences in demographics, clinicopathologic characteristics and post-operative outcomes remain largely undefined.
Methods:
Retrospective review of 1,958 patients (21% female) with esophageal cancer that underwent esophagectomy at a single institution between 1995–2017.
Results:
Most patients had adenocarcinoma (83%); however, the rate of squamous cell carcinoma was significantly higher in females (35% vs 11%, respectively; p<.0001). Females had a lower rate of smoking (62 vs 73%) and heavy alcohol use (12 vs 19%) but a higher rate of previous mediastinal radiation (8.4 vs 1.8%) (p<.001). Postoperative mortality and overall survival (OS) were similar between sexes. However, sub-analysis of patients with locoregional disease (clinical stage II/III) demonstrated that females received neoadjuvant therapy less frequently than males and had worse OS (median OS 2.56 years vs 2.08; p=0.034). This difference remained significant on adjusted analysis (HR 1.24, 95% CI 1.06–1.46).
Conclusions:
Female patients had higher incidence of squamous cell carcinoma despite lower prevalence of behavioral risk factors. Among patients with locoregional disease, undertreatment in females may reflect treatment bias and history of previous mediastinal radiation. Esophageal cancer in females should be considered a unique entity as compared to the presentation and treatment of males.
INTRODUCTION
Biologic and behavioral differences between males and females affect the incidence, presenting features and prognosis of disease as well as treatment response and outcomes. Despite these variations, the traditional “one-size-fits all” approach to disease prevention and management has often failed to account for sex. (1) Studies on sex bias have demonstrated that this pervasive problem is also present in surgical research. (2,3) Improved understanding of sex-specific differences in disease and treatment is needed in order to ensure that all patients receive high-quality care.
Esophageal cancer is a highly fatal, predominantly male disease. In the United States, a male-to-female ratio of 4.2:1 difference in esophageal cancer incidence has recently been reported. (4,5) The etiology of observed gender differences in the incidence of esophageal cancer are likely multifactorial and remain mostly undefined. Higher prevalence of high risk behavior, such as drinking and smoking, has been demonstrated among male patients. (6, 7) Additionally, recent data has suggested that estrogen exposure in females may have a protective role. (8, 9)
Previous evaluations of sex-based differences in esophageal cancer outcomes have suggested that females have fewer medical comorbidities, are diagnosed at an earlier clinical stage and have less postoperative complications as compared to males. (6, 7, 10) This may explain observed differences in survival, with some studies showing a more favorable prognosis in females after esophagectomy. (6,7,10–12) However, current available studies comparing the clinicopathologic characteristics and prognosis between genders are limited by a small sample size or inclusion of only one histologic subtype.
Individualized care has had an increasingly important role in the discussion on how to improve prognosis for esophageal cancer. (13) Sex-based distinctions may also play an important role. The purpose of this study was to improve understanding of sex-specific differences in the demographics, clinicopathologic characteristics, and prognosis of patients with esophageal cancer treated with surgical resection in a large single-institution cohort.
METHODS
A retrospective review was performed to identify patients that underwent esophagectomy between 1995 and 2017 from a prospectively maintained institutional database. Only patients with adenocarcinoma and squamous cell carcinoma were included. Patients with stage IV disease or unresectable at surgery were excluded. Medical records were reviewed to identify patient demographics, clinicopathologic data and perioperative outcomes. Characteristics and outcomes were compared by sex.
Clinical stage was determined in accordance with the standard institutional approach using computed tomography, endoscopic ultrasound and positron emission tomography (PET-CT), when it became available. Neoadjuvant chemotherapy or chemoradiation was administered to patients with locoregionally advanced disease, unless contraindicated. Operation type was selected by tumor location and stage. Behavioral risk factors of interest included previous or current history of tobacco use, any alcohol use, and heavy alcohol use (≥ 14 drinks/week). Previous history of mediastinal radiation was identified. Radiation-associated esophageal cancer has previously been described as occurring with a latency period ≥ 5 years after radiation treatment.(14, 15)
Outcomes of interest included 30-day mortality, complications and serious complications (grade ≥ 3) as well as hospital length of stay (LOS), in-hospital mortality and overall survival (OS). (16, 17)
Categorical variables were compared using chi-square or Fisher’s exact test and are presented as percentages. Continuous data was compared using the Wilcoxon rank sum test and expressed as median and interquartile range (IQR). OS from the time of surgery was evaluated using Kaplan-Meier and compared between groups using the log rank test. Cumulative incidence of recurrence (CIR) was calculated from the date of surgery until the date of recurrence or last follow-up, and compared using Gray’s test. Death without recurrence was considered a competing risk on analysis. A sub-analysis comparison of OS was performed in patients with locoregional disease since this group of patients faces peculiar treatment challenges as a result of more extensive disease requiring multimodality approaches. Cox regression analysis was used to adjust for other factors associated with OS including surgery year, age, pulmonary disease, cardiac disease, diabetes and pathologic stage. Hazard Ratios (HR) with 95% confidence intervals (CI) are presented. A p-value <0.05 was considered significant. Analysis was performed using SAS v 9.4 (SAS Institute, Cary, North Carolina).
RESULTS
There were 1958 patients that met study criteria; 1542 (79%) male and 416 (21%) female. Most patients were white (90%) with adenocarcinoma (84%) with median age of 63 at time of surgery (IQR 56–70).
Demographic and Clinicopathologic Characteristics
A comparison of demographic characteristics between groups is presented in Table 1. Females had a lower rate of cardiac disease and diabetes, heavy alcohol use and smoking but more frequent history of previous mediastinal radiation (p<.001). Female patients had lower median BMI and were less likely to be obese (25% vs 31%, p=0.013).
Table 1.
Comparison of demographic characteristics between 1,958 male and female patients that underwent esophagectomy
| Variable | Male (N=1542) | Female (N=416) | P Value |
|---|---|---|---|
| Median Age, years (IQR) | 63 (56–70) | 64 (56–71) | 0.202 |
| Median BMI (IQR) | 27.6 (24.7–31.1) | 24.4 (20.9–30.0) | <.0001 |
| Comorbidities, N (%) | |||
| Pulmonary | 163 (11) | 47 (11) | 0.656 |
| Cardiac | 485 (32) | 77 (19) | <.0001 |
| Diabetes | 268 (17) | 36 (9) | <.0001 |
| Race, N (%) | |||
| White | 1401 (91) | 366 (88) | 0.014 |
| Black/Hispanic | 54 (3) | 28 (7) | |
| Asian/Other | 87 (6) | 22 (5) | |
| Any Alcohol Consumption, N (%) | 1182 (77) | 240 (58) | <.0001 |
| Heavy Alcohol Consumption, N (%) | 296 (19) | 50 (12) | 0.0005 |
| Any Tobacco Use, N (%) | 1123 (73) | 257 (62) | <.0001 |
| Median Pack Year History (IQR) | 30 (15–45) | 30 (15–44) | 0.650 |
| Mediastinal Radiation, N (%) | 28 (1.8) | 35 (8.4) | <.0001 |
| Married, N (%) | 1257 (82) | 243 (58) | <.0001 |
Clinicopathologic and treatment characteristics are compared between groups in Table 2. Female patients had a 24% higher incidence of squamous cell carcinoma (35% vs 11%, p<.0001). Trends in esophagectomy by sex and year overall and by histology type are presented in Figure 1A, B.
Table 2.
Clinicopathologic and treatment comparison between 1,958 female and male patients that underwent esophagectomy
| Variables | Male (N=1542) | Female (N=416) | P Value |
|---|---|---|---|
| Surgery Year | |||
| 1995–2002 | 472 (31) | 167 (40) | 0.001 |
| 2003–2010 | 548 (35) | 129 (31) | |
| 2011–2017 | 522 (34) | 120 (29) | |
| Clinical Stage | |||
| 0 | 23 (1) | 7 (2) | 0.169 |
| 1 | 252 (16) | 87 (21) | |
| 2 | 457 (30) | 128 (31) | |
| 3 | 673 (44) | 160 (38) | |
| NA | 137 (9) | 34 (8) | |
| Histology | |||
| Adenocarcinoma | 1371 (89) | 271 (65) | <.0001 |
| Squamous Cell Carcinoma | 171 (11) | 145 (35) | |
| Grade | |||
| Well | 85 (5) | 23 (6) | 0.718 |
| Moderate | 780 (51) | 221 (53) | |
| Poor | 621 (40) | 155 (37) | |
| Tumor Location | |||
| Middle/Proximal | 92 (6) | 68 (16) | <.0001 |
| Distal/GEJ | 1434 (94) | 346 (84) | |
| Neoadjuvant Type1 (n=1418) | |||
| None | 157 (14) | 66 (23) | 0.0004 |
| Chemotherapy Alone | 94 (8) | 37 (13) | 0.022 |
| Chemoradiation | 861 (77) | 179 (63) | <.0001 |
| Procedure Type | |||
| Ivor Lewis | 1222 (79) | 265 (64) | <.0001 |
| 3-Hole | 104 (7) | 61 (15) | |
| Transhiatal | 101 (7) | 51 (12) | |
| Other | 115 (7) | 39 (9) | |
| pT Stage | |||
| 0 | 261 (17) | 71 (17) | 0.871 |
| In Situ/1 | 423 (27) | 105 (25) | |
| 2 | 280 (18) | 81 (19) | |
| 3 | 558 (36) | 153 (37) | |
| 4 | 17 (1) | 6 (1) | |
| pN Stage | |||
| 0 | 913 (59) | 260 (62) | 0.412 |
| 1 | 338 (22) | 92 (22) | |
| 2 | 187 (12) | 44 (11) | |
| 3 | 101 (7) | 20 (5) | |
| Pathologic Stage | |||
| 0 | 233 (15) | 68 (16) | 0.166 |
| 1 | 416 (27) | 103 (25) | |
| 2 | 430 (28) | 136 (33) | |
| 3 | 461 (30) | 109 (26) | |
| Total Nodes Harvested | 20 (15–27) | 21 (16–27) | 0.642 |
| Vascular Invasion | 466 (30) | 102 (25) | 0.067 |
| Positive Margin | 84 (5) | 15 (4) | 0.164 |
| Any Adjuvant | 152 (10) | 24 (6) | 0.033 |
Pairwise treatment comparison by sex among 1418 patients with locoregional (Stage II/III disease)
Figure 1A, B.
Esophagectomy by year and by, A) sex B) sex and histology (histology presented as proportion of total esophagectomies in a given year)
While the proportion of squamous cell carcinoma versus adenocarcinoma decreased significantly over time in males (p=0.0002), the rate was constant in females (p=0.501).
Female patients had a non-significant trend towards presentation with earlier clinical stage. Pathologic T, N and overall stage were similar between groups. Among patients with locoregional disease, females received neoadjuvant therapy less often than males (p=0.0004) and were more likely to receive chemotherapy alone than males (p=0.022).
Postoperative Outcomes
Overall, 49% of patients had a postoperative complication and 18% had a serious complication. There was no difference in overall or serious complication rates between groups. (Table 3) Female patients had a longer median length of stay (LOS) compared to males (p=0.002), and a higher rate of in-hospital mortality (p=0.046). There was non-significant difference in 30-day mortality between sexes.
Table 3.
Comparison of postoperative outcomes between 1,958 males and females after esophagectomy
| Variable | Male (N=1542) | Female (N=416) | P Value |
|---|---|---|---|
| Median Length of Stay, Days (IQR) | 11 (10–16) | 12 (10–18) | 0.002 |
| 30 Day Mortality, N (%) | 53 (3) | 21 (5) | 0.145 |
| In-Hospital Mortality, N (%) | 62 (4) | 27 (7) | 0.045 |
| Any Complications, N (%) | 750 (49) | 206 (50) | 0.782 |
| Any Serious Complications, N (%) | 270 (18) | 79 (19) | 0.472 |
Overall Survival
Median follow-up among survivors was 2.48 years (range, 0.05–22.91). The OS for the study cohort was 2.98 years (95% CI 2.73–3.39). At the end of the study period, 892 males and 267 females had died. There was no difference in median OS between males (3.04 years, 95% CI 2.76–3.46) and females (2.64, 95% CI 2.16–3.84) (p=0.34). Similarly, the overall 5-year cumulative incidence of recurrence did not differ between groups (46% male, 95% CI 0.43–0.49 vs 44% female, 95% CI 0.38–0.49; p=0.14).
However, stratification by gender and clinical disease stage demonstrated significant differences. In patients with local disease (clinical stage 0/I), there was no significant different in OS between females (median 7.18, 95% CI 4.26–10.45) vs males (5.50, 4.31–6.66) (p=0.38). However, among those with locoregional disease (clinical stage II/III), OS was worse in females (median 2.08, 95% CI 1.63–2.47) than males (2.56, 2.24–2.83) (p=0.034). (Figure 2A,B) These remained significant on multivariable analysis (HR 1.24, 95% CI 1.06–1.46). (Supplementary Table 1)
Figure 2A, B.
Overall survival by sex and clinical tumor stage among 1.958 patients that underwent esophagectomy, A) local (clinical stage 0/I) B) locoregional (clinical stage II/III)
DISCUSSION
In an era of increased focus on sex bias in research and individualization of care, this study identified several important differences between females and males undergoing surgical resection for esophageal cancer. Female patients had a higher rate of squamous cell carcinoma despite lower prevalence of behavioral risk factors. Additionally, although no overall differences in postoperative mortality or OS were observed between male and female patients, females with locoregional disease received neoadjuvant therapy less often, had more positive margins, and worse OS. Understanding sex-specific differences in the presentation and prognosis of esophageal cancer may help improve outcomes for this highly lethal disease.
It is known a well-known fact that the incidence of adenocarcinoma is rising in the United States while squamous cell carcinoma is decreasing; we demonstrated this is not true for women where the proportion of squamous cell carcinoma and adenocarcinoma have remained constant over the years in our data. (18, 19) The high rate overall of squamous cell carcinoma in this cohort may reflect changes in incidence that occurred over the long study period, with 48% of cases undergoing surgery between 1995–2002. While a male-to-female incidence ratio up to 9:1 has been reported for adenocarcinoma, it is lower (2.7:1) for squamous cell carcinoma even with reduced prevalence of behavioral risk factors for squamous cell carcinoma, such as alcohol and tobacco use, in females. (5–7) A possible contributing explanation may be the higher history of previous mediastinal radiation in females. In this series, the rate of previous mediastinal radiation among females was nearly 5x greater than males. Radiation exposure has previously been associated with increased risk of esophageal squamous cell carcinoma, and in a previous study we showed that such patients were more likely to be female than those with primary disease (65% vs 38%). (14, 20–22)
Histology may also account for the demonstrated differences in postoperative outcomes in this study. Females overall had higher in-hospital mortality and longer length of stay despite fewer comorbid medical conditions. This contrasts with other studies that showed fewer complications and better outcomes in female patients. (6, 7) However, those observations are limited by small sample sizes and study cohorts predominantly comprised of one histologic subtype, essentially eliminating the impact of tumor location on postoperative outcomes. The higher rate of proximal and mid-thoracic tumors in females, which creates a more challenging resection, likely explains the observed differences.
Furthermore, our cohort demonstrated that females with locoregional disease received neoadjuvant therapy less often and had higher incidence of previous mediastinal radiation, which have both been associated with worse postoperative and long-term outcomes. The reason for the lower rate of neoadjuvant therapy administration among females with locoregional disease may be multifactorial. First, the treatment of patients with locoregional disease that have received previous mediastinal radiation is challenged by toxicity associated with reirradiation and complex surgery in a previously radiated field. We previously demonstrated that patients with radiation-associated esophageal cancer that received neoadjuvant therapy had lower disease recurrence but higher in-hospital mortality after esophagectomy.(14) In another recent study, Markar et al similarly demonstrated that patients with previous radiation history had higher risk of serious postoperative complications and worse overall survival. (15)
The second reason for undertreatment among females with locoregional disease may be bias in treatment selection. Previous studies utilizing Surveillance, Epidemiology and End Results data have demonstrated that females with esophageal cancer are less likely to be married than males, and that unmarried patients are significantly less likely to receive aggressive treatment and have worse survival. (23, 24) As males may be more likely to be working, females may serve as caregivers more often. As such, the observed difference in neoadjuvant treatment may reflect lack of support for female patients given that 58% were married as compared to 82% of men.
Although this study is limited by its single-institution retrospective nature, it represents an important addition to the current literature on sex differences in patients undergoing esophageal resection. The large study population and inclusion of both adenocarcinoma and squamous cell carcinoma support the strength of our findings, and suggest that observed differences in outcomes after esophagectomy are the reflection of tumor histology and location rather than sex-specific causes. It is not possible to draw conclusions regarding changes in incidence of esophageal cancer among females from this study.
Females undergoing surgical resection for esophageal cancer have a different profile than males. Higher in-hospital mortality and worse OS may reflect the observed higher proportion of females with squamous cell carcinoma, history of previous mediastinal radiation and undertreatment of females with locoregional disease. Although esophageal cancer is traditionally perceived as a disease of males, the presentation and management of disease in females should be considered a unique entity.
Supplementary Material
Acknowledgments
FUNDING SOURCES: This study was supported, in part, by the National Institutes of Health/National Cancer Institute Cancer Support Grant P30 CA008748. Tamar Nobel is supported, in part, by a grant from the American Cancer Society.
REFERENCES
- 1.Sex Regitz-Zagrosek V. and gender differences in health. EMBO reports 2012; 13:596–603. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Kibbe MR. Reporting of sex as a variable in research published in surgical journals. JAMA Surgery 2018; 153:983. [DOI] [PubMed] [Google Scholar]
- 3.Mansukhani NA, Yoon DY, Teter KA, et al. Determining If Sex Bias Exists in Human Surgical Clinical Research. JAMA Surgery 2016; 151:1022–1030. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Noone AM, Howlader N, Krapcho M, et al. (eds). SEER Cancer Statistics Review, 1975–2015, National Cancer Institute; Bethesda, MD, https://seer.cancer.gov/csr/1975_2015/, based on November 2017 SEER data submission, posted to the SEER web site, April 2018. [Google Scholar]
- 5.Xie S, Lagergren J. The male predominance in esophageal adenocarcinoma. Clin Gastroenterol Hepatol. 2016;14:338–347. [DOI] [PubMed] [Google Scholar]
- 6.Morita M, Otsu H, Kwano H, et al. Gender differences in prognosis after esophagectomy for esophageal cancer. Surg Today. 2014;44:505–512. [DOI] [PubMed] [Google Scholar]
- 7.Nishino T, Yoshida T, Inoue S, et al. Gender differences in clinicopathologic features and prognosis of squamous cell carcinoma of the esophagus. Esophagus. 2017;14:122–130. [Google Scholar]
- 8.Yang H, Sukocheva OA, Hussey DJ, et al. Estrogen, male dominance and esophageal adenocarcinoma: Is there a link? World J Gastroenterol. 2012;18:393–400. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Mathieu LN, Kanarek NF, Tsai H, et al. Age and sex differences in the incidence of esophageal adenocarcinoma: Results from the Surveillance, Epidemiology and End Results (SEER) Registry (1973–2008). Dis Esophagus. 2014;27:757–763. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Koide N, Kitazawa M, Komatsu D, et al. Gender differences in clinicopathologic features and outcomes of esophageal cancer patients treated surgically. Esophagus. 2011;8:107–112. [Google Scholar]
- 11.Hidaka H, Hotokezaka M, Nakashima S, et al. Sex difference in survival of patients treated by surgical resection for esophageal cancer. World J Surg. 2007;31:1982–1987. [DOI] [PubMed] [Google Scholar]
- 12.Shimada H, Matsubara H, Okazumi S, et al. Improved surgical results in thoracic esophageal squamous cell carcinoma: a 40-year analysis of 792 patients. J Gastrointest Surg. 2008;12:518–526. [DOI] [PubMed] [Google Scholar]
- 13.Nobel TB, Molena D. From standardization to personalized medicine: Moving beyond cookie-cutter treatment of esophageal cancer. JThorac Cardiovasc Surg 2018; 156:1736–1738. [DOI] [PubMed] [Google Scholar]
- 14.Nobel TB, Barbetta A, Hsu M, et al. Outcomes of Radiation-Associated Esophageal Squamous Cell Carcinoma: The MSKCC Experience. J Gastrointest Surg 2018. [epub ahead of print]. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Markar SR, Gronnier C, Pasquer A, et al. Surgically treated oesophageal cancer developed in a radiated field: Impact on peri-operative and long-term outcomes. European Journal of Cancer 2017; 75:179–189. [DOI] [PubMed] [Google Scholar]
- 16.Low DE, Alderson D, Cecconello I, et al. International Consensus on Standardization of Data Collection for Complications Associated With Esophagectomy: Esophagectomy Complications Consensus Group (ECCG). Ann Surg 2015; 262(2):286–94. [DOI] [PubMed] [Google Scholar]
- 17.U.S. Department of Health and Human Serives. Common Terminology Criteria for Adverse Events (CTCAE) Version 5.0 2017. Available at: https://ctep.cancer.gov/protocolDevelopment/electronic_applications/docs/CTCAE_v5_Quick_Reference_5x7.pdf. Accessed December 10, 2018.
- 18.Hur C, Miller M, Kong CY, et al. Trends in esophageal adenocarcinoma incidence and mortality. Cancer. 2013;119:1149–1158. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Raman R, Deorah S, McDowell BD, et al. Changing incidence of esophageal cancer among white women: analysis of SEER data (1992–2010). Contemp Oncol (Pozn). 2015;19:338–340. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Zablotska LB, Chak A, Das A, Neugut AI. Increased risk of squamous cell esophageal cancer after adjuvant radiation therapy for primary breast cancer. Am J Epidemiol 2005;161:330–337. [DOI] [PubMed] [Google Scholar]
- 21.Morton LM, Gilbert ES, Stovall M, et al. Risk of esophageal cancer following radiotherapy for Hodgkin lymphoma. Haematologica. 2014;99:e193–e196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Grantzau T, Overgaard J. Risk of second non-breast cancer after radiotherapy for breast cancer: a systematic review and metaanalysis of 762,468 patients. Radiother Oncol 2015;114:56–65 [DOI] [PubMed] [Google Scholar]
- 23.Du L, Kim JJ, Chen B, et al. Marital status is associated with superior survival in patients with esophageal cancer: a Surveillance, Epidemiology, and End Results study. Oncotarget 2017;8: 95965–95972. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Zhang Q, Lin X, Zhang C, et al. The influence of marital status on the survival of patients with esophageal cancer: a population-based, propensity-matched study. Oncotarget 2017;8: 62261–62273. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.