Editorial Comment: Cancer of the oesophagus and gastrooesophageal junction: neoadjuvant therapy should not be a surrogate for suboptimal lymphadenectomy

Abstract

In preparation for the 7th edition of the UICC/AJCC cancer staging manual for cancers of the oesophagus and gastrooesophageal junction [1], data were collected from centres world-wide, creating the World-Wide Esophageal Cancer Collaboration (WECC) in 2006 [2]. Thirteen institutions from five countries and three continents submitted 7884 desindentified patient files; within this dataset, 4625 patients who had undergone primary oesophagectomy without neoadjuvant or adjuvant therapy were identified [3]. This database compilation permitted a data-driven proposal for the 7th edition of the AJCC cancer staging manual, and additionally permitted a characterization of the relationship between survival and extent of lymphadenectomy. This effort has resulted in a definition of the optimum extent of lymphadenectomy in patients treated by primary surgery [4]. In this issue of the journal, Stiles et al .[ 5] hypothesized that optimal lymphadenectomy, as defined by the WECC, may also be relevant to cases of surgery following neoadjuvant therapy in patients with cancer of the oesophagus and gastro-oesophageal junction. They conclude that the WECC definitions for lymphadenectomy are especially relevant in cases where induction therapy results in persistent nodal disease, or a lack of tumour downstaging by T-stage classification. The survival figures in their series are impressive, with the overall 5-year survival of 47.8%; a 3-year survival of 50.8% in ypT3 patients who were not downstaged; and 55.4% survival at 3 years in patients with persistent nodal disease. Although a comparison with suboptimal lymphadenectomy did not reach statistical significance due to small patient numbers, these figures underline the value of an optimal lymphadenectomy after neoadjuvant therapy. That is a very important message. Besides this statistical limitation, a few other caveats may be noted: these excellent outcome figures are somewhat skewed by the selection process employed to define the study cohort. Indeed, patients with R1 and R2 resections, as well as patients with seven or more positive nodes at final pathological examination were excluded from the study, given their poor survival and the inability to demonstrate a benefit from extended lymphadenectomy. Furthermore, it is unclear why 41 patients (representing 48% of ypT3/T4 patients) appeared to have a suboptimal lymphadenectomy. Does this represent technical difficulty in achieving optimal lymph node clearance intraoperatively, or is it related to concerns of an increased risk of morbidity with an extensive lymphadenectomy in frail or elderly patients? The WECC guidelines for optimum lymphadenectomy are based on the data obtained after primary surgery. In the vast majority of these patients, lymphadenectomy was limited to a standard two-field lymphadenectomy, with the upper border of the lymph node dissection being the aortic arch. However, in this study, as many as 94 patients underwent a three-field lymphadenectomy, which includes dissection along both recurrent nerves above the level of the aortic arch, a bilateral lymphadenectomy along the remainder of both recurrent nerves in the neck and the removal of the lower deep cervical nodes located posterior and lateral to the carotid sheath [6]. Lymphadenectomy performed in this manner typically adds another 10–15 nodes to the total node count. That is to say, this subset of patients may have had an inadequate thoracic and abdominal lymph node dissection, according to WECC guidelines for lymphadenectomy. The question arises therefore, whether the claimed survival benefit is related to the addition of the third field, rather than the total number of dissected lymph nodes and as such satisfies the definition of optimum lymphadenectomy. As stated by the authors, the most important drawback of the study is the rather limited number of patients within each group (in particular the group with suboptimal lymphadenectomy). Differences in therapeutic modalities may further confound the study outcomes, such as the presence or the absence of radiotherapy in the induction protocol, variations in radiation doses over the study period and the use of different combinations of chemotherapeutic agents, all of which have a potential to modulate the treatment effect. Additionally, outcomes could potentially be ascribed to differences in histological tumour subtypes, the anatomic location of the primary tumour, the preinduction burden of nodal disease, the location of nodal basins within anatomic compartments (single versus multiple compartments) and extracapsular lymph node involvement. Patient factors relating to