Randomized Phase III Trial ion Locally Advanced Squamous Cell Carcinoma of the Esophagus: Neoadjuvant Radiochemotherapy Followed by Surgery vs. Definitive Radiochemotherapy

Reviewer: S. Jack Wei, MD
Abramson Cancer Center of the University of Pennsylvania
Last Modified: October 13, 2004

Presenter: M. Stuschke
Presenter's Affiliation: University of Essen, Essen, Germany (German Multicenter Trial)
Type of Session: Scientific


  • The optimal treatment for locally advanced squamous cell carcinoma (SCCA) of the esophagus is unclear
  • While the addition of neoadjuvant radiochemotherapy (RTCT) followed by surgery has been shown to improve outcomes compared to surgery alone, the question of whether these patients benefit from surgery at all is in question.
  • This study was designed to determine whether neoadjuvant RTCT followed by surgery improves survival over definitive RTCT alone in high-risk patients with locally advanced SCCA of the esophagus.

Materials and Methods

  • All patients had histologically-confirmed SCCA of the upper or mid-esophagus.
  • Tumors were stage T3-4N0-1M0 by EUS and CT scan
  • All patients had performance status of 0-1 and were deemed medically operable.
  • All patients received chemotherapy upfront consisting of 5-FU, leucovorin, etoposide, and cisplatin q 3wk x 3 cycles followed by CTRT consisting of cisplatin and etoposide and concurrent radiation therapy (RT) to 40 Gy.
  • Patients were then randomized to:
  • Arm A: Surgical resection including a two-field lymphadenectomy via a transthoracic approach 2-4 weeks after CTRT
  • Arm B: Completion of definitive RTCT to 65 Gy, or for transversable T3 tumors, external beam RT to 60 Gy followed by high dose rate brachytherapy boost of 4 Gy x 2.


  • From 6/94 to 5/02, 172 patients were randomized to the study
  • The two arms were well-balanced for age, stage, gender, and amount of weight loss.
  • From Arm A, only 57/86 patients actually underwent surgery.  The majority of patients who did not receive surgery had progressive disease during the induction chemotherapy phase and were deemed inoperable.
  • From Arm B, 5/86 patients elected to undergo surgery.
  • The overall response rate was 36% for Arm A and 35% for Arm B.
  • 8% of patients on Arm A had progressive disease compared to 14% on Arm B (p=NS)
  • In Arm A, of the 57 patients who underwent surgery, 51 received an R0 resection (complete pathologic resection) and 18% had a pathologic complete response to the preoperative RTCT.
  • Mortality from treatment was higher in Arm A compared to Arm B (10.5% vs. 2.3%).
  • 3-year overall survival (OS): Arm A 31% vs. Arm B 24%. p for equivalence = 0.006
  • 3-year freedom from local recurrence: Arm A 64% vs. Arm B 41%, p<0.001
  • 3-year OS for patients who did not respond to induction CTRT: 35% vs. 18%

Author's Conclusions

  • Neoadjuvant RTCT followed by surgical resection results in improved local control compared to definitive RTCT for locally advanced SCCA of the esophagus.
  • Peri-operative mortality and the risks of distant metastases negate the benfit of improved local control with surgery, resulting in equivalent OS.

Clinical/Scientific Implications

The current study demonstrates that in locally advanced SCCA of the esophagus, surgical resection does not provide a survival benefit over definitive radiochemotherapy. There was however, an improvement in local control with surgery. This benefit was largely offset by the increased treatment-related mortality due to surgery. It should be noted that although more patients died initially in the surgery arm, there was a trend towards an increased rate of long-term survival with surgery. In addition, a steep learning curve with regards to surgery was seen: the majority of surgery-related mortality occurred in the first half of the study duration. Presumably, if patients are carefully selected with regards to their ability to withstand surgical resection, and surgical techniques and post-operative care improves, the trend towards improved long-term survival with surgery may translate into a significant overall survival benefit. This may be particularly true for patients who do not show any response to the induction portion of RTCT.

The results of this study are in contrast to a recently reported French study also comparing neoadjuvant RTCT followed by surgery with definitive RTCT. The results of that trial showed improved median and 2-year overall survival for patients receiving definitive RTCT (17.7 mo vs. 19.3 mo and 34% vs. 40%, respectively). However, patients in that study who received definitive RTCT had a higher rate of dilation and stent requirement compared to those undergoing surgery. The French study and the current study demonstrate that the decision to treat with surgery or definitive RTCT should be individualized, and surgical resection should be reserved for carefully selected patients.