Discussion
Here we demonstrated that the presence of RLN node metastasis was not a prognostic predicator in node-positive patients with squamous cell carcinoma of the middle thoracic esophagus. A previous report including 55 patients with esophageal squamous cell carcinoma who underwent esophagectomy with 2-field lymphdenectomy showed that RLN node metastasis was the strongest prognostic predicator. That report was more heterogeneous in term of tumor site: tumors were located below and above the carina in 40 and 15 patients, respectively. Different tumor sites might lead to different frequencies of lymph node metastasis. In that report, frequencies of RLN node metastasis was 18% in all patients (10/55) and 26% (10/34) in node-positive patients. While in our study, the frequencies of RLN node metastasis were 34% (81/235) in all patients and 62% (81/133) in node-positive patients. More importantly, Authors only performed univariate analysis, but did not perform multivariate analysis in that cohort. Dealing with data this way may cause confounding effect that influenced the interpretation of results. There was another report on clinical outcomes of 106 patients with esophageal squamous cell carcinoma who underwent 3-field lymphadenectomy. Univariate and multivariate analyses indicated that RLN node metastasis was the most unfavorable prognostic factor. In that series, 10, 67 and 29 patients had tumors located in the upper, middle and lower thoracic esophagus, respectively. Although RLN node metastasis occurred in 60 of 78 (77%) node-positive patients, the report did not state how many patients with lesions in the middle thoracic esophagus had RLN node metastasis. Furthermore, the factor of the number of metastatic lymph nodes was not included in the analysis.
Among various possible prognostic predicator of esophageal carcinoma, the importance of number of metastatic nodes has been widely recognized. Patients with a large number of metastatic nodes had a lower average survival rate than those with less metastatic nodes. Stratification of the number of metastatic nodes varied in different reports (for example, 1–3 vs ≥ 4, 1–4 vs ≥ 5, 1–5 vs ≥ 6, and 1–7 vs ≥ 8). Our report showed that the survival rate decreased with an increasing number of metastatic nodes, and that the optimal cutoff value was between 1–6 and ≥ 7 metastatic nodes. On the other hand, there was little evidence supporting that the site of metastatic nodes influenced the prognosis of esophageal carcinoma. For example, celiac node metastasis, which was regarded as M1 disease in the past, did not mean poor prognosis in node-positive patients with esophageal cancer. It was found that for middle and lower thoracic esophageal carcinoma, survival of patients with celiac node metastasis did not differ from those with left gastric node metastasis. The 7th edition of TNM staging system also has redefined a regional node of esophageal cancer as any periesophageal lymph nodes from cervical nodes to celiac nodes; yet N staging has already been subclassfied according to the number of metastatic nodes.
The frequency of RLN node metastasis was reported between 20% and 50% in patients with squamous cell carcinoma of the upper and middle thoracic esophagus. In our institution, upper thoracic tumor is routinely treated with radiotherapy-dominated multidisciplinary therapy. Some authors pointed out that RLN was the initial metastatic site (including micrometastatic site) in esophageal squamous cell carcinoma. Others found that the histology of RLN node was characterized by large cortical area without anthracosis and hyalinization, which suggests a high filtration activity. All these features of RLN nodes need to be further investigated. Some authors found that the prognoses of patients with RLN node metastasis was better in the three-field lymphadenectomy group than in the two-field lymphadenectomy group, while in patients without RLN node metastasis, there was no significant differences in survival between these two groups. Their results could not be duplicated in this study. It should be noted that the features of patients in that study including age, tumor location and disease stage, differed between patients with RLN node metastasis and those without RLN node metastasis. These differences between patients groups could cause biased results. Frequency of cervical nodal metastasis (30%) in this report was similar to previous reported. Significant associations between RLN node metastasis and cervical node metastasis in esophageal squamous cell carcinoma were emphasized by many authors, and they firmly believed that 3-field lymphadenectomy was indicated if RLN node metastasis happens. But there is lack of high-level evidence supporting 3-field lymphadenectomy in terms of long-term survival. Instead it is certain that increased postoperative morbidity and impaired long-term quality of life are associated with 3-field lymphdenectomy. Although 3-field lymphadenectomy might offer survival benefit for selected patients with esophageal cancer, the controversy over the optimal extent of lymphadenectomy still exists. For a majority of patients there would be no arguments about performing two-field lymphadenectomy to offer a balance between benefits and risks. In addition, the emphasis of three-field lymphadenectomy lies more in RLN lymphadenectomy than in cervical lymphadenectomy. In this study, 3-filed lymphadenetomy did not show its survival benefits compared with 2-field lymphadenectomy, but RLN node metastasis also did not portend a worse prognosis in node-positive patients. Thus lymphadenectomy including dissection of RLN nodes is strongly supported.
Several potential shortcomings of the present study are worth mentioning. This retrospective study from a single institution suffers from the typical biases associated with such studies. The choice of surgical procedures depended on surgeons' preference without strict criteria. It is likewise unavoidable that lymphadenectomy was performed in more or less different extent by different surgeons. In addition, there was no set standard for patients to receive adjuvant therapy. As shown in the result of the univariate analysis, patients with adjuvant therapy had worse survival than those without adjuvant therapy. The majority of patients with adjuvant therapy had a large number of metastatic nodes (data not shown). However, this series was proved to be homogenous in clinical variables including tumor site and pathologic type. Further multi-institutional studies with larger sample size are needed to confirm these results.