Reply to Letter: “Tumor Response and Lymph Node Status After Neoadjuvant Chemoradiotherapy for Esophageal Cancer”
We read with some interest the correspondence of Drs Ecker and Dempsey following our recent publication of the impact of neoadjuvant chemoradiotherapy (nCRT) on lymph node status in early stage I and II esophageal cancer.1 Our published data are derived from a post hoc analysis of FFCD 9901, a multicenter randomized control trial comparing surgery alone with nCRT followed by surgery.2 The authors of this correspondence rightly observe that we found that when tumor regression grade was dichotomized as a complete/good response (TRG 1/2) or neutral/poor response (TRG3–5) to nCRT, then no difference was found in the number of lymph nodes resected. A favorable tumor response did, however, correlate with favorable nodal response, with significantly fewer positive nodes in the resected specimen.
The authors relate their experience from as yet unpublished data taken from 50 consecutive patients with stage II and III esophageal tumors—predominantly adenocarcinomas (n = 46)—treated with nCRT followed by transhiatal resection and compared with patients treated with surgery alone at their own institution. In that cohort, pathological complete response (pCR) (n = 7) was associated with a reduced lymph node yield (5.0 vs 15.0, P = 0.02) whereas an incomplete tumoral response (n = 43) was not. It seems that many differences—tumor stage, tumor histology, population size, and surgical approach—exist between their population and the randomized population on which we have reported. The authors suggest that nCRT alone decreases lymph node yield after pCR. Such conclusions are difficult to make on the basis of a small population and surgery by a technique without formal 2-field lymphadenectomy and commonly associated with fewer lymph nodes being resected.3
Our experience is different. Patients in our cohort with a pCR (ypT0N0) did not have fewer lymph nodes found in the resected specimen [median number of resected nodes of 14.0 (2.0–39.0) in group nCRT ypT0N0 (n = 28) vs 16.5 (0–47.0) in group nCRT non-ypT0N0 (n = 52), P = 0.160, data not included in the article]. Although it is still uncertain whether the number of lymph nodes resected after nCRT is of prognostic value (as it is in the absence of neoadjuvant treatment), our own data1 and those from the CROSS group4 make this increasingly unlikely. As the correspondence points out, this would be consistent with recent findings in the setting of rectal cancers.5
Even if reduced lymph node yield were to be associated only with good/complete response after nCRT and if the number of retrieved nodes after nCRT has no prognostic value, we strongly advocate that neither of these facts should change the surgical strategy. First, at present, pCR is a postoperative diagnosis with no accurate means of its prediction; second, a cooperative survival benefit seems to exist between nCRT and adequate lymphadenectomy6; and third, extended lymphadenectomy gives maximal prognostic information and avoids positive nodes being missed, with patients being erroneously classified as pN0.
Finally, it must be realized that our results pertain to early-stage esophageal cancers, although we would expect our results to be amplified in more advanced disease. As our article emphasizes, positive lymph node status, and not the number of nodes resected, is the stronger prognostic indicator for patients after nCRT. It is too early to advocate a change in nodal clearance based on perceived tumor response to treatment. Extensive lymphadenectomy should be routinely performed to rule out missed nodal disease.