Esophageal cancer is the eighth most common cancer worldwide. It is endemic in many parts of the world, particularly in developing nations,1 and accounts for more than 200 000 deaths every year in China. An effective and rational staging system of esophageal cancer is the essential prerequisite to determine the appropriate treatments and predict long-term survival. Although the confirmed prognostic indicators include histological grading, tumour differentiation, tumour (T), nodal metastases (N) and metastasis (M) classification,2 two studies3,4 observed that tumour length was an independent prognostic predictor. On the other hand, in the latest version (7th edition) of the American Joint Committee on Cancer (AJCC) TNM system, tumour length is not listed as a risk factor in esophageal cancer.5 To better understand the impact of tumour length on survival, we reviewed a cohort of esophageal cancer patients who had undergone esophageal resection only and without preoperative chemotherapy or radiotherapy from 2004 to 2008 in our hospital. Univariate and multivariate analyses were used to determine the prognostic value of tumour length, and showed that esophageal tumour length was a remarkable independent prognostic factor in long-term survival. Even so, few articles discuss the relationship between tumour length and other clinicopathological characteristics as well as long-term survival in patients with esophageal squamous cell carcinoma (ESCC).
Of the 340 patients with esophageal cancer operated on at the Department of Cardiothoracic Surgery of the Huashan Hospital, Fudan University from January 1, 2004 to December 31, 2008, 202 patients with squamous cell carcinoma who received oesophagectomy were included in this study. All these patients were classified as R0 resection (with no residual tumour according to AJCC classification). The 12 patients who had palliative or R1/R2 resections and the 126 patients who had histology other than ESCC were excluded.
All patients were evaluated for preoperative staging by means of barium meal, fibrogastroscopy, computed tomography of the chest and abdomen, ultrasound of the neck and retroperitoneal lymph nodes (LNs). Some patients received PET-CT. Pulmonary and cardiac function studies and other blood tests were done for assessment of surgical tolerance.
Either the left or the right chest was incised in this study. Patients with tumours in middle or lower thoracic oesophagus with no evidence of LN involvement in the superior mediastinum or in the neck (n=150) received oesophagectomy via left thoracotomy (single incision). Patients with tumours in the middle or upper thoracic oesophagus or with possible LN metastasis in the superior mediastinum or neck (n=52) were operated via cervico-right thoracic-abdominal (triple incision).
Histopathological examination included evaluations of tumour length (measured on the 10% formaldehyde fixed specimens), differentiation, T classification, number of positive and removed LNs. TNM staging was performed according to the AJCC 7th edition guidelines.5 Patients were followed up every six months for the first three years and then annually. Survival time was measured from the date of surgery to the date of death or the latest follow-up time.
All statistical analyses were performed with SPSS 16.0 for Windows (SPSS Inc., USA). Chi-square analysis was used to study the association between tumour length and other clinicopathological characteristics. Univariate analysis of survival was performed using Kaplan-Meier method and log rank test to estimate the prognostic value of clinicopathological characteristics. Multivariate analysis of survival was performed using Cox regression model to identify independent prognostic factors. The level of significance was set to P <0.05.
Association of tumour length with other clinicopathological characteristics and analyses of prognostic factors
The clinicopathological characteristics of the 202 patients, results of chi-square analyses and univariate analyses are shown in Table 1. There were 154 males and 48 females, with median age 60 years (36 to 81 years). The median tumour length was 3 cm (0.3 to 9.0 cm). The AJCC stage distribution was as follows: stage I, n=30 (14.9%); stage II, n=88 (43.5%) and stage III, n=84 (41.6%).
Chi-square analyses found patients with tumour length ≥3 cm had increasing T stage (χ2=55.864, P <0.001), number of positive LNs (χ2=14.625, P <0.001), metastatic LN ratio (χ2=16.072, P <0.001), and overall TNM stage (χ2=48.134, P <0.001), all highly statistically significant. There was no relation between tumour length and age, differentiation, tumour location or surgical procedure.
Univariate analyses showed no relation between gender, tumour location, type of surgical procedure or survival. The strongest prognostic factors were tumour length (χ2=14.402, P <0.001), and T stage (χ2=23.968, P <0.001), number of positive LNs (χ2=48.781, P <0.001), overall TNM stage (χ2=45.990, P <0.001) and metastatic LN ratio (χ2=50.660, P <0.001), all highly statistically significant. Differentiation and age were also related to survival.
Influence of tumour length on overall survival and survival in staging system subgroups
The survival rates, assessed against different tumour lengths in 1 cm increments, became worse as tumour length increased. Patients with tumour length <3 cm had significantly better survival rates (Figure 1).
To assess the impact of tumour length on different tumour stages, we divided the patients into subgroups by T stage (T1–2 vs. T3–4), LN stage (LN negative vs. LN positive) and overall TNM stage (I+II vs. III). Compared with tumour length ≥3 cm, patients with tumour length <3 cm had significantly better survival in subgroups of T1-T2 (n=81, χ2=4.538, P=0.033, Figure 2), LN negative (n=115, χ2=7.093, P=0.008, Figure 3) and TNM stage I+II (n=118, χ2=4.843, P=0.028, Figure 4). However no significant influence of tumour length was found in subgroups of T3-T4 (n=121), metastatic LNs (n=87) or TNM stage III (n=84). Therefore, tumour length appeared to have evident impact on patients with lower T stage, LN negative and lower TNM stage.
Multivariate survival analyses
The multivariate analyses were performed with the Cox proportional hazards model, which included age, differentiation, T stage, metastatic LN ratio, number of positive LNs and tumour length. The results indicated that T stage, differentiation, metastatic LN ratio, number of positive LNs and tumour length were significant risk factors (Table 2). Patients with tumour length ≥3 cm had a worse prognosis compared to those <3 cm.
The TNM Classification of Malignant Tumours (T for tumour, N for node and M for metastasis) was originally developed by Pierre Denoix in the 1940s and adopted by the AJCC and has now become the worldwide accepted basis of cancer staging. The latest edition of the AJCC TNM staging system of esophageal cancer was published in 2009, in which adenocarcinoma and squamous cell carcinoma are staged as two different types. Histological grade, depth of tumour invasion and number of metastatic lymph nodes are independent staging factors for esophageal cancer. However, the prognostic role of tumour length in patients with esophageal cancer especially ESCCs was not mentioned.5,6
Few studies have focused on the prognostic impact of tumour length on esophageal cancer. Eloubeidi et al7 retrospectively evaluated 10 441 cases of esophageal cancer with data collected from the National Cancer Institute. The treatments included surgery, radiotherapy and chemotherapy. Tumour length was assessed with surgical resection specimens, preoperative endoscopy and radiological investigations. Although data of' tumour length were missing from half of the cases, they held that tumour length was an independent prognostic factor in patients with esophageal cancer.
Welch et al4 retrospectively evaluated 309 patients (72 squamous cell carcinomas, 225 adenocarcinomas and 12 other tumours) and concluded that tumour length was an independent predictor of survival for adenocarcinomas but not for squamous cell carcinomas. To minimize the biases resulting from the heterogeneity of data sources, Wu et al8 retrospectively reviewed a total of 582 patients with esophageal squamous cell carcinoma who received surgery as the primary treatment and concluded that tumour length served as an independent predictor of long-term survival.
In this study, 202 squamous cell carcinoma patients were enrolled. Tumour length had a significant correlation with other clinicopathalogical characteristics. Patients with tumour length ≥3 cm had increasing T stage, worse N stage, increasing metastatic LN ratio and increasing overall TNM stage. It is widely agreed that lymph node status, depth of tumour invasion and overall TNM stage are strong, independent prognostic factors in esophageal cancers.9–11 It may well be that the influence of tumour length on the subgroup with different T and N stages is important for the understanding of its role in overall survival of postoperative patients with esophageal cancer. The results of our study showed tumour length should be considered as a prognostic grouping factor in early stage ESCC, which is quite consistent with the previous studies12,13 showing that tumour length had a greater prognostic value for localized esophageal cancer than for cancer with metastases.
By multivariate analysis, tumour length, which might be the easiest one to know, was found an independent risk factor of survival for ESCC. This does not only provide the surgeon with a more comprehensive view in evaluating the prognosis of esophageal cancer, but also a helpful piece of information for considering the postoperative treatments.
As a retrospective study, limitations are inevitable. As well as the small sample size, mortality from all causes was used instead of disease specific deaths, which are difficult to confirm. In addition, patients with distant metastasis were not enrolled in our study so our results may not apply to all patients. To confirm the prognostic role of tumour length on overall survival of ESCC patients, a prospective randomized multicentre study is necessary.
In conclusion, esophageal tumour length ≥3 cm was significantly associated with increasing tumour stage, worse lymph stage, increasing metastatic LN ratio, increasing overall TNM stage and poor survival. Esophageal tumour length is a predictive factor for long-term survival especially for lower tumour stage, LNs negative and lower TNM stage patients. Therefore, it is advisable that tumour length should be incorporated in the staging system as an important grouping factor for better prognostic evaluation.
1. Kamangar F, Dores GM, Anderson WF. Patterns of cancer incidence, mortality, and prevalence across five continents: defining priorities to reduce cancer disparities in different geographic regions of the world. J Clin Oncol 2006; 24: 2137-2150.
2. Ajani JA, Barthel JS, Bentrem DJ, D'Amico TA, Das P, Denlinger CS, et al. Esophageal and esophagogastric junction cancers. J Natl Compr Canc Netw 2011; 9: 830-887.
3. Swisher SG, Yendamuri S, Correa AM, Hofstetter W, Ajani JA, Francis A, et al. Esophageal tumor length is independently associated with long-term survival. Cancer 2009; 115: 508-516.
4. Welch IM, Griffiths EA, Brummell Z, Gorthi G, Pritchard SA. Tumor length as a prognostic factor in esophageal malignancy: univariate and multivariate survival analyses. J Surg Oncol 2006; 93: 258-267.
5. Edge SB. American Joint Committee on Cancer: AJCC cancer staging manual, 7th ed. New York: Springer; 2009: 103-115.
6. Shields TW. General thoracic surgery, 7th ed. Philadelphia: Wolters Kluwer Health/Lippincott Williams & Wilkins; 2009: 1961-1971.
7. Eloubeidi MA, Desmond R, Arguedas MR, Reed CE, Wilcox CM. Prognostic factors for the survival of patients with esophageal carcinoma in the U.S.: the importance of tumor length and lymph node status. Cancer 2002; 95: 1434-1443.
8. Wu YC, Wang BY, Goan YG, Hsu PK, Hsu WH. Tumor length as a prognostic factor in esophageal squamous cell carcinoma. Ann Thorac Surg 2011; 91: 887-893.
9. Altorki NK, Zhou XK, Stiles B, Port JL, Paul S, Lee PC, et al. Total number of resected lymph nodes predicts survival in esophageal cancer. Ann Surg 2008; 248: 221-226.
10. Peyre CG, Hagen JA, DeMeester SR, Altorki NK, Ancona E, Griffin SM, et al. The number of lymph nodes removed predicts survival in esophageal cancer: an international study on the impact of extent of surgical resection. Ann Surg 2008; 248: 549-556.
11. Wijnhoven BP, Tran KT, Esterman A, Watson DI, Tilanus HW. An evaluation of prognostic factors and tumor staging of resected carcinoma of the esophagus. Ann Surg 2007; 245: 717-725.
12. Eloubeidi MA, Desmond R, Arguedas MR, Reed CE, Wilcox CM. Prognostic factors for the survival of patients with esophageal carcinoma in the US-The importance of tumor length and lymph node status. Cancer 2002; 95: 1434-1443.
13. Swisher SG, Bolton WD, Hofstetter WL, Francis AM, Correa AM, Ajani JA, et al. Impact of tumor length on long-term survival of pT1 esophageal adenocarcinoma. J Thorac Cardiov Sur 2009; 138: 831-836.