The required extent of lymphadenectomy during esophagectomy for cancer remains a controversial topic [17, 18]. In current literature the benefit of an increased lymph node yield on overall survival, with at least 15–23 resected lymph nodes, has been proven [3]. However, the explicit location of affected lymph nodes is still under debate. A distribution pattern of metastatic lymphatic spread in both adenocarcinoma and squamous cell carcinoma identified by Hagens et al. in a systematic review, found metastases to cervical, thoracic, and abdominal lymph node stations, regardless of the primary tumor location [11]. Even though higher accumulations of lymph node metastases are located nearby the primary tumor (depending on tumor location), distribution percentages differ and distant nodal metastasis as well as skip metastasis (metastasis infiltrating more distant lymph nodes without affecting adjacent nodes) are frequently seen in both esophageal adenocarcinoma and squamous cell carcinoma [19, 20].
Resection of these distant lymph node metastases, e.g. in the proximal mediastinal field, is not only beneficial for oncological radicality. Pathologically, systematic lymph node resection enables an exact postoperative tumor staging, important for further therapy decisions [21,22,23]. Additionally, different studies described reduced hazard of death with an increasing number of resected and examined nodes [24, 25]. Despite these benefits, possible complications and harms should also be considered. In order to find an accurate statement for necessary lymph node resection, a separate assessment of each lymph node station is required. The lower paratracheal lymph node station (station 7 according to TIGER-study) turned out to be the most striking station of our data. The resection of the paratracheal lymph nodes is considered as the extension of standard 2FD, which consist of the abdominal lymph node stations as well as a complete dissection of the middle and lower mediastinal nodes, including the paraesophageal, pulmonary ligament, subcarinal, and aortopulmonary window nodes.
We found no positive lymph node in the lower paratracheal region of all 101 patients with adenocarcinoma or squamous cell carcinoma of the esophagus undergoing LPL resection. In fact, the only two patients with positivity in this region suffered from different histopathological entities (melanoma and neuroendocrine tumor). This is the main finding of this study.
The resection of LPL had no significant effect on either 30- and 90-day mortality or tumor recurrence. However, overall survival (Fig. 3) showed a trend towards better survival for LPL resected patients. Since tumor infiltration of the LPL region was only found in two cases the resected LPL region might not be the underlying cause. Interestingly, total lymph node harvest was significantly higher in the LPL group with 35 vs. 25 lymph nodes, with only five resected lymph nodes in the paratracheal region on average. One could argue, when LPL resection was conducted, an overall more thorough LAD was performed, leading to improved survival. Additionally, the lower complications rate of the patients receiving LPL resection might be an important factor for an improvement of long-term survival independent of oncological reasons. These arguments show that LPL resection can be performed without increased morbidity or mortality.
A large cohort study by Harada et al., conducted in the United States, was the first to investigate paratracheal lymph node metastasis from adenocarcinoma of the esophagus [26]. Excluding the cases with initial lymph node metastases in the paratracheal region, 6,5% of the analyzed patients who did not have received LPL suffered from positive paratracheal lymph node recurrence later. However, their definition of the paratracheal regions differed strongly from ours. According to the Japanese Classification of Esophageal Cancer, 11th Edition [27], they included upper thoracic paraesophageal lymph nodes, cervical paraesophageal lymph nodes, recurrent nerve lymph nodes and left tracheobronchial lymph nodes.
There are limitations to this study. Firstly, it is a retrospective study, lacking a randomization. The decision for or against paratracheal resection was made by the surgeon. Thus, pretreated patients (chemotherapy or chemoradiation) tended to receive paratracheal lymph node resection more frequently than patients not pretreated (without reaching significance p = 0.051) – most likely due to the surgeons’ expectations for a more extended underlying disease and the higher probability for a locally advanced tumor (T3–4). However, there was found no difference between both groups regarding occurrence of metastasis and tumor recurrence. Additionally, during operations with robotic assistance, paratracheal stations were resected at a significantly higher rate. The simplification of meticulous dissection between delicate structures allows greater radicality. In fact, the trend towards faster procedures when including paratracheal lymph node resection hints at the same point: easier procedures are more often combined with extended lymph node resection. Supporting this theory, patients not receiving paratracheal lymph node resection had significantly more complicated procedures (p = 0.026) and more major complications according to Clavien-Dindo-classification. Additionally, in 2017 only 22% of the patients received LPL resection, the same year robotic assistance was introduced in our clinic. Some part of the higher complication rate in the non-dissecting group might be due to the learning curve after introduction of robotic assistance. In order to eliminate these bias mistakes, a prospective study with randomization is required.
Based on the presented data, we do not perform standardized resection of the lower paratracheal lymph node station when operating adenocarcinoma and squamous cell carcinoma of the distal esophagus in our institution. Since modern technologies enable safe dissection and no increase of morbidity due to the paratracheal dissection, LPL should always be considered in rare cancer entities or on demand.