Delaying Esophageal Cancer Surgery After Neoadjuvant CRT Fails in Trial

Prolonging the time to surgery following neoadjuvant chemoradiotherapy (CRT) for esophageal cancer failed to improve histological complete response rates and may even worsen overall survival (OS), according to results from the randomized NeoRes II study.

Among the subset with adenocarcinoma, complete response rates were not significantly better with the prolonged surgical approach (10-12 weeks) compared with the standard time frame (4-6 weeks) after neoadjuvant CRT (26% vs 21%; P=0.429), reported Magnus Nilsson, MD, PhD, of the Karolinska Institute in Stockholm, and colleagues.

Moreover, a "strong trend" toward worse OS was seen with the prolonged approach (HR 1.35, 95% CI 0.94-1.95, P=0.107), they detailed in Annals of Oncology.

In explaining the reason for the study, the researchers pointed out that the optimal time interval between neoadjuvant CRT and curative-intent surgical resection has never been established in esophageal cancer. Plus, but a shift toward delayed surgery has been observed in clinical practice as observational studies have suggested an increased probability of histological complete response with the approach.

Nilsson's group said the new study findings suggest "caution in routinely delaying surgery for more than 6 weeks" in this setting.

In a related editorial, Sarah Derks, MD, PhD, and Hanneke W.M. van Laarhoven, MD, PhD, both of Amsterdam UMC in the Netherlands, pointed out that the optimal time to surgery could depend on several factors, including the amount of time needed to both recover from neoadjuvant CRT and achieve maximum therapeutic efficacy from neoadjuvant CRT, as well as the potential for cancer cell regrowth and disease progression over time.

"Timing matters," they wrote, adding that "the best timing for surgery is pivotal in securing optimal patient outcomes during the perioperative period, and at the same time secure long-term survival." Thus, considering the survival outcomes demonstrated in the trial, they agreed with the authors that "we should be cautious in delaying surgery in this patient population."

From 2015 to 2019, the NeoRes II trial enrolled 249 patients with esophageal cancer (clinical stage T1N1-3M0 or T2-4aN0-3M0) from 10 participating centers in Sweden, Norway, and Germany. Participants were randomized 1:1 to either the prolonged (10-12 weeks after CRT) or standard (4-6 weeks after CRT) surgery groups following CROSS-type neoadjuvant CRT.

Patients had an average age of about 65 years, and more than 80% were men. Adenocarcinomas (80%) comprised most tumors, with the remaining being squamous cell carcinoma. The study's primary endpoint was complete histological response in the adenocarcinoma subgroup.

Ultimately, 223 of the patients underwent tumor resection: 106 with the prolonged approach and 117 with the standard approach. Ten patients allocated to the prolonged group were diagnosed with distant metastases after 18-FDG-PET-CT performed following completion of neoadjuvant CRT, which precluded resective surgery, compared with just one patient in the standard surgery arm.

Median time to surgery was 75 days (10.7 weeks) in the prolonged arm and 39.5 days (5.6 weeks) in the standard arm.

Over a median follow-up of 36.4 months, 50% of patients in the prolonged surgery group remained alive versus 58% of those in the standard group. First-quartile OS was 14 months in the prolonged arm as compared with 26.5 months in the standard arm (P=0.003), and patients allocated to the delayed approach had an increased risk of mortality after 7 months that declined over time:

• 10 months: HR 3.09 (95% CI 1.44-6.61)
• 20 months: HR 1.92 (95% CI 1.25-2.93)
• 30 months: HR 1.08 (95% CI 0.67-1.74)
• 40 months: HR 0.49 (95% CI 0.22-1.11)

Worse OS in the prolonged group appeared to be driven by histological non-responders with tumor regression grade 4 (>50% remaining tumor cells), Nilsson and colleagues observed, a subset with a 2.5-fold (95% CI 1.1-5.8) higher risk for death compared with non-responders in the standard surgery arm.

No significant differences were seen between the two surgery timing arms for overall tumor regression grade, tumor-free resections margins, or for the number of resected or metastatic lymph nodes in either of the two histologic types or across all patients.

A limitation, the study authors noted, is that not all patients were operated on per protocol. For example, surgery was delayed more than 7 weeks for eight of the patients (7%) in the standard timing arm as they needed extra time to recover from treatment. However, postponing surgery to give patients time to recover from therapy "is in accordance with how these patients would have been managed in clinical practice, which strengthens the external validity of the trial," the researchers wrote.

Comment