MADRID – , findings from the Dutch SANO trial suggest.
After 2 years, researchers found no significant differences in overall and disease-free survival between patients on active surveillance and those who received surgery either immediately following neoadjuvant chemoradiotherapy or who switched from active surveillance to surgery.
Overall, patients who underwent active surveillance had “noninferior overall survival at 2 years,” said Berend J. Van der Wilk, PhD candidate, Erasmus University, Rotterdam, the Netherlands, who presented the findings at the annual meeting of the European Society for Medical Oncology.
Over the 2-year follow-up, at least 35% of patients on active surveillance were spared surgery. Patients on active surveillance who experienced locoregional regrowth could still undergo surgery, Mr. Van der Wilk said.
Magnus Nilsson, MD, PhD, the invited discussant, who was not involved in the research, said performing such a conceptually important and complex trial was a “huge achievement.” However, Dr. Nilsson highlighted some “major concerns” with the trial design, which could affect the generalizability of the findings.
Avoiding surgery?
Esophagectomy remains the “keystone of curative treatment for esophageal cancer,” Mr. Van der Wilks explained. However, this operation is a “major surgical procedure” that comes with a mortality rate of up to 5%. As many as 59% of patients experience complications.
The CROSS trial, which included more than 360 patients with esophageal or esophagogastric junction cancer, found that neoadjuvant chemoradiotherapy improved survival among patients with potentially curable disease; 29% of patients achieved a pathologic complete response.
Mr. Van der Wilk said those strong outcomes create some uncertainty as to whether all patients need standard surgery after chemoradiotherapy.
In other words, Mr. Van der Wilk asked, “Should we be willing to follow an active surveillance, organ-sparing strategy for patients with a clinical response?”
An active surveillance strategy, he said, would require frequent evaluations of the patient’s clinical response. Surgery would be performed only in cases of proven residual tumor in which there were no distant metastases. The potential pitfall of an active surveillance approach is that patients may develop unresectable tumor regrowths, “possibly resulting in inferior overall survival.”
To compare active surveillance with standard surgery, the team conducted a phase 3 noninferiority stepped-wedge cluster randomized trial involving patients with locally advanced esophageal cancer.
Patients received neoadjuvant chemoradiotherapy with carboplatin and paclitaxel for 5 weeks. Concurrent radiotherapy was delivered at 41.4 Gy in 23 fractions, 5 days per week, as in the CROSS trial.
More than 300 patients who achieved a complete clinical response 12 weeks after completing chemoradiotherapy were randomly assigned to undergo standard surgery or active surveillance. Surgery was performed for those with subsequent tumor regrowth.
Overall, 198 patients underwent active surveillance, and 111 patients underwent standard surgery. The two groups were well balanced in terms of median age, sex distribution, proportion of adenocarcinomas, and World Health Organization performance scores. At the last patient assessment, on July 6, 2023, the median follow-up was 38 months.
Overall, 101 of 111 patients in the standard surgery arm and 83 of 198 (42%) in the active surveillance group had surgery. The time to surgery in the active surveillance arm was 5.9 months, compared with 0.7 months with standard surgery. For both groups, the R0 resection rate was 98%.
Mr. Van der Wilk reported no significant difference in overall survival between the active surveillance and standard surgery groups (hazard ratio for death, 1.14; 95% confidence interval, 0.74-0.78; P = .55). Overall survival in the active surveillance group was noninferior to that in the standard surgery group at 2 years. Noninferiority was defined as an overall survival difference between the two arms of less than 15%.
Mr. Van der Wilk also reported no significant difference in disease-free survival between the active surveillance and the standard surgery groups – 35 months with active surveillance, and 49 months with surgery (HR, 1.35; P = .15). At 30 months following neoadjuvant chemoradiotherapy, 43% of patients on active surveillance and 34% with standard surgery developed distant metastases, but the difference was not significant (odds ratio, 1.45; P = .18).
Among the patients in the active surveillance arm who had a complete response, 35% (n = 69) had a persistent clinical response, while 17% (33 patients) developed distant metastases, and 48% (n = 96) experienced locoregional growth. The postoperative 90-day mortality was 4% in the active surveillance group and 5% in the surgery group.
Health-related quality of life was significantly better at 6 and 9 months in the active surveillance group, Mr. Van der Wilk noted.
Although Dr. Nilsson, the invited discussant, highlighted the importance of the trial, he also expressed concern over the trial design.
The intention-to-treat analysis was contaminated, Dr. Nilsson said, because the trial design allowed for one crossover, but patients in the trial crossed over at two time points – 35 patients who were initially assigned to standard surgery crossed over to the active surveillance arm, and later, seven patients from a preSANO trial were included in the active surveillance arm.
Dr. Nilsson also expressed concern about mixing squamous cell carcinoma and adenocarcinoma histologies in the study. If the authors had distinguished patients with squamous cell carcinoma and those with adenocarcinoma in each arm, there may have a difference in overall survival, given that squamous cell carcinoma is much easier to treat.
The study also included some patients who did not have a complete clinical response and whose surgery was delayed by more than 10 weeks – a practice that, Dr. Nilsson said, “does not really seem to be safe.” A recent study led by Dr. Nilsson found that delaying surgery for 10-12 weeks in comparison with 4-6 weeks did not improve histologic complete response or other pathologic endpoints and may have led to worse survival.
“I’m afraid it’s not really certain that it’s safe to prolong surgery more than 10 weeks or longer in the clinical noncomplete responders,” said Dr. Nilsson, from the department of clinical science, intervention, and technology, Karolinska Institute, Stockholm.
Overall, he said, the study “suggests that survival may be noninferior” among patients on active surveillance in comparison with those who undergo immediate surgery, but the findings need to be confirmed in a trial with a more stringent intention-to-treat analysis that is stratified by histologic subtypes.
The study was funded by the Dutch Cancer Society and the Netherlands Organisation for Health Research and Development (ZonMw). Mr. Van der Wilk has disclosed no relevant financial relationships. Dr. Nilsson has relationships with Medtronic, Intuitive Surgical, Bristol-Myers Squibb, and Merck Sharp & Dohme, from which he received no personal financial benefit.
A version of this article first appeared on Medscape.com.