Neo-AEGIS: Preliminary results of phase III RCT of CROSS versus perioperative chemotherapy

Bookmark and Share
Published: 29 Jun 2021
Views: 469
Prof John Reynolds - Trinity College Dublin, Dublin, Republic of Ireland

Prof John Reynolds speaks to ecancer about Neo-AEGIS (Neoadjuvant trial in Adenocarcinoma of the Esophagus and Esophago-Gastric Junction International Study).

He discusses the preliminary results of phase III RCT of CROSS versus perioperative chemotherapy (Modified MAGIC or FLOT protocol). Initially, he mentions the rationale and methodology of this study.

Prof Reynolds then shares the results of this study. This RCT reveals no evidence that perioperative chemotherapy is unacceptably inferior to multimodal therapy, notwithstanding greater proxy markers of local tumour response in the CROSS arm.

In the end, he talks about the future of this study and its impact on the treatment of oesophageal cancer.

The trial is in adenocarcinoma of the oesophagus and the esophago-gastric junction which is a increasingly common cancer in the Western world. The background for the study’s genesis was the publication of the CROSS trial in 2012 and the question back then was should that be the standard of care for this cancer or should pre- or perioperative chemotherapy, which was used in many countries at that time, based on good evidence, should that be tested in comparison with the very high impact CROSS regimen which was combination chemotherapy and radiation therapy prior to surgery for patients presenting with locally advanced adenocarcinoma of the oesophagus or junction.

What was the methodology of your study?

A phase III randomised clinical trial, just patients with predicted clinical T2 or T3, node negative or node positive locally advanced disease. All patients had uniform staging with CT-PET scanning; all patients who were undergoing multimodal CROSS regimen chemoradiation had RT quality assurance built in to the methodology. Complications, surgical complications which were an important secondary endpoint in the study, given some concerns about the impact of preoperative radiation and operative complications, very detailed prospective recording of operative complications as per recently defined consensus definitions as per the Oesophageal Complications Consensus Group, were utilised. So they were three relatively rare or unique elements within the design of the trial.

The primary endpoint was overall survival and there were lots of secondary endpoints including clinical response, disease free survival, operative complications, as mentioned, pathological response rates and many others.

What were the key findings?

The key findings were… the other element of the design that was important, it began as a superiority design with the control arm being predicted to be superior than the perioperative chemotherapy arm. With the first futility analysis that was done and with a new trial, the FLOT4 trial, with a new regimen, the FLOT regimen, emerging during the trial period and a protocol amendment being made, the trial design ultimately became non-inferiority. The main findings, basically, based on the first and second futility or interim analysis was that there was no differences in the primary outcome of overall survival at this time point between the experimental arm, perioperative chemotherapy which was either the MAGIC regimen or more latterly the FLOT regimen, or the control CROSS regimen. Both arms did three year survival was very good – 56% for CROSS and 57% for the perioperative chemotherapy – and the hazard ratio was 1.02.

In terms of secondary endpoints, not surprisingly but importantly, the pathological response rates were higher for the combination chemotherapy and radiation therapy – 16% versus 5% complete pathological response rate, 42% versus 12% major pathological response rate. The R0 resection status was significantly higher – 97% versus 83% - arm B versus arm 4. So those proxies of local regional impact that you would associate with radiation therapy as part of the treatment regimen were proven in this trial. However, interestingly, this did not translate into an advantage with respect to the primary outcome measure, the overall survival.

The other secondary outcome measure I mentioned was operative complications and, again, absolutely no difference in the mortality or major morbidity or types of complications or severity of complications between both treatment arms which is important because this is an area of debate and controversy from previous studies.

How can these results impact the future treatment of gastric cancer?

For oesophageal cancer, so for oesophageal and junctional cancer it does a few things. Firstly it provides good outcome data for either arm. So if somebody is…. It provides clinical equipoise for the use of both arms. It provides support for groups, multidisciplinary groups, who practise one of these treatment arms. So for the control arm, the CROSS data, the pathological response rates, the outcome data, the operative data, the survival data, is very consistent with the original CROSS trial and very affirming of the value of that approach in terms of what it achieves for the oesophageal cancer patient. That might even be more supported going forward by the recent emergence of the CheckMate-577 trial which shows adjuvant nivolumab, so immunotherapy, post this type of regimen resulting in a dramatically reduced recurrence compared to a placebo control. So you can combine those two. For many groups they prefer not to give radiation therapy or their practice, both regionally or nationally, works much better giving pre- or postoperative chemotherapy or they have a natural preference for that for some other reasons, particularly seen in the UK and France and some other quarters and parts of North America. Again you will find good data within this randomised phase III trial, very good outcome data, three years, no difference in primary outcome survival data at this time point. So providing a clinical equipoise in the first phase III, large phase III, trial that has addressed this question, provides important data. However, it does not provide a winner, a clear winner, based on the primary endpoint analysis.

We will be continuing this survival analysis until at least the middle of 2022 but it’s very unlikely given the absolutely matching first and second interim analyses, the second being after 143 patient deaths out of the 377 patients randomised, it’s highly unlikely that this will change once the trial analysis is closed. What will be very interesting, however, will be the patterns of recurrence to see if there is any clear way as to why, notwithstanding much better pathologic data for the control arm, that the overall survival rates are the same. So it’s the systemic effect of chemotherapy counterbalancing the more local regional impact of the radiation therapy. That will only emerge, can only be analysed, once we have all the data looked at when the trial is completed sometime next year.