Preoperative Chemoradiotherapy Versus Perioperative Chemotherapy for Patients With Resectable Esophageal or Gastroesophageal Junction Adenocarcinoma

BACKGROUND: This study compares neoadjuvant chemoradiotherapy (nCRT) with perioperative chemotherapy (pCT) for patients with resectable esophageal or gastroesophageal junction (GEJ) adenocarcinoma in terms of toxicity, postoperative complications, pathologic response, and survival. METHODS: This study retrospectively analyzed and compared 313 patients with resectable esophageal or GEJ adenocarcinoma treated with either nCRT (carboplatin/paclitaxel 41.4 Gy, n = 176) or pCT (epirubicin, cisplatin and capecitabine, n = 137). RESULTS: The baseline and tumor characteristics were similar in both groups. The ability to deliver all planned preoperative cycles was greater in the nCRT group (92.0 vs. 76.6%). Whereas nCRT was associated with a higher rate of grades 3 and 4 esophagitis, pCT was associated with a higher rate of grades 3 and 4 thromboembolic events, febrile neutropenia, nausea, vomiting, diarrhea, hand–foot syndrome, mucositis, cardiac complications, and electrolyte imbalances. Two patients in the pCT group died during neoadjuvant treatment due to febrile neutropenia. More postoperative cardiac complications occurred in the nCRT group. All other postoperative complications and the in-hospital mortality rate (nCRT, 4.7%; pCT, 2.3%) were comparable. The pathologic complete response (pCR) rate was 15.1% after nCRT and 6.9% after pCT. Radicality of surgery was comparable (R0: 93.0 vs. 91.6%). The median overall survival was 35 months after nCRT versus 36 months after pCT. CONCLUSION: For patients with esophageal or GEJ adenocarcinoma, chemoradiotherapy with paclitaxel, carboplatin and concurrent radiotherapy, and perioperative chemotherapy with epirubicin, cisplatin, and capecitabin lead to equal oncologic outcomes in terms of radical resection rates, lymphadenectomy, patterns of recurrent disease, and (disease-free) survival. However, neoadjuvant chemoradiotherapy is associated with a considerably lower level of severe adverse events and should therefore be the preferred protocol until a well-powered randomized controlled trial provides different insights. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi:10.1245/s10434-017-5827-1) contains supplementary material, which is available to authorized users.