Background
Phase III RCT (German Trial — "Stahl 2005"). 172 eligible patients with locally advanced squamous cell carcinoma (SCC) of the esophagus from Germany. All patients received induction chemotherapy (etoposide, leucovorin, 5-FU, and cisplatin). Randomized to surgery (with lower-dose CRT 40 Gy) vs definitive chemoradiotherapy (with higher-dose CRT ≥65 Gy). Primary outcome was overall survival; secondary outcome was local progression-free survival.
Interventions and follow up
Arm A: Induction chemotherapy → CRT (40 Gy in 16 fx + concurrent cisplatin/etoposide/leucovorin/5-FU) → esophagectomy
Arm B: Induction chemotherapy → definitive CRT (≥65 Gy with same concurrent chemotherapy) — no surgery
Primary endpoint: Overall survival
mFollow up: Median 6 year
Arm B: Induction chemotherapy → definitive CRT (≥65 Gy with same concurrent chemotherapy) — no surgery
Primary endpoint: Overall survival
mFollow up: Median 6 year
Results
Overall survival: Equivalent (log-rank P=.002 for equivalence met); median OS not reported as significantly different
2-year local progression-free survival: 64.3% (surgery) vs 40.7% (CRT alone), HR 2.1, P=.003 — surgery improved local control
Treatment-related mortality: 12.8% (surgery) vs 3.5% (CRT), P=.03
Independent prognostic factor: Clinical tumor response to induction chemotherapy
2-year local progression-free survival: 64.3% (surgery) vs 40.7% (CRT alone), HR 2.1, P=.003 — surgery improved local control
Treatment-related mortality: 12.8% (surgery) vs 3.5% (CRT), P=.03
Independent prognostic factor: Clinical tumor response to induction chemotherapy
Adverse events
Main adverse events: Surgery arm: 12.8% treatment-related mortality — significantly higher than CRT alone (3.5%). This difference in early mortality partially offsets the local control advantage of surgery. Definitive CRT arm: higher rate of local recurrence and requirement for salvage interventions (stenting).
Conclusions
Adding surgery to chemoradiotherapy improved local tumor control in locally advanced esophageal SCC but did not improve overall survival, compared to definitive high-dose CRT. The survival equivalence was achieved at the cost of higher treatment-related mortality in the surgery arm. Tumor response to induction chemotherapy was the single independent prognostic factor for OS in both arms.
Key Limitations
Key Limitations: Restricted to SCC only — results not applicable to adenocarcinoma. Induction chemotherapy used before CRT is not standard in current neoadjuvant CRT protocols (CROSS uses concomitant regimen without prior induction). Higher-dose RT (≥65 Gy) in the definitive arm may be suboptimal compared to concurrent lower-dose CRT with surgery (40 Gy). The treatment-related mortality of 12.8% in the surgery arm reflects older surgical practices (likely pre-modern enhanced recovery protocols). Design confounds comparison of surgery vs RT dose differences.
Clinical Context
Together with FFCD 9102, this trial established that surgery does not improve survival over definitive high-dose CRT for SCC after a clinical response to induction therapy — establishing that surgery can be omitted in responding SCC patients. This conceptual framework underlies organ-preservation and watch-and-wait strategies in SCC. For patients who do not achieve complete clinical response, salvage esophagectomy remains an option.
References