SAN ANTONIO, TX-Not often does one attend a symposium's "great debate" session in which an audience of clinicians is actually swayed to change their management of a disease. But such was the case here during the Society of Surgical Oncology's Annual Cancer Symposium, where an audience of surgical oncologists indicated they would change their practice in the management of high-grade dysplasia (HDG) of the esophagus.
Not to say that either debater was more persuasive than the other on the topic of "Surgery vs endoscopic ablation for high-grade dysplasia-carcinoma in situ (HDG-CIS) of the esophagus." In fact, both speakers generally agreed on most points, which is so often the case.
Instead, it was a seemingly minor point presented by both speakers, on the matter of patient compliance, that made the audience think twice about how to manage dysplasia and Barrett's esophagus.
While both speakers agreed there was a time and a place for endoscopic mucosal resection and also for esophagectomy, the main point taken was that a patient who cannot commit to the many follow-up endoscopic treatments required might be better off having surgery, even if it does carry a higher risk of complications.
For Endoscopic Ablation
Robert C. Martin, MD, PhD, argued for endoscopic mucosal resection (EMR) and radiofrequency ablation.
"I have the daunting task of convincing a group of surgeons not to operate, so obviously the chips are stacked against me," he said.
As background, Dr. Martin said the prevalence of Barrett's esophagus is far less than that of gastroesophageal reflux disease, with the extent of Barrett's estimated at 1.6% of the general adult US population, and 6.8% of people over age 40.
If high-grade dysplasia is appropriately diagnosed in patients with focal Barrett's esophagus, EMR is appropriate, he said.
He showed data from a study on EMR presented at last year's Digestive Disease Week, which found a five-year survival rate of 98%, overall complication rate of 9%, and a recurrence rate of 11% for EMR in patients with early high-grade dysplasia-"That's fairly good for a simple endoscopic procedure," he said.
For multi-focal or multiglandular types of dysplasia, radio-frequency ablation (RFA) is appropriate.
"There is evidence that with RFA, complete eradication of dysplasia is greater than 90%, and recurrence of the dysplasia is less than 2%," Dr. Martin said.
He cited the ongoing NCT00282672 study on the durability of epithelial reversion after RFA, part of the AIM Dysplasia Trial which compares ablation plus anti-secretory medication with a sham procedure plus anti-secretory medication. He said that trial has established that the degree of complete eradication of low- and high-grade dysplasia was superior with ablation than with observation, even after three years of follow-up.
For each technique, he added, patient factors such as age, extent of disease, comorbities and multiple procedures, and endoscopic factors-namely expertise-must be taken into consideration.
For Surgery
Tom R. DeMeester, MD, spoke in favor of surgical dissection, explaining that there has been a shift in therapy for high-grade dysplasia and intermucosal carcinoma since the 1980s, when surgical treatment was en bloc resection. When it was seen that the rate of nodal disease was less than 2%, transhiatal esophagectomy took its place.
But morbidity and side effects of that procedure were still high for very early disease, and eventually the need to preserve the esophagus led to the two competing therapies today-vagal-sparing esophagectomy and endoscopic mucosal resection and ablation with a possible fundoplication to completion.
To compare the two, Dr. DeMeester cited a recent retrospective study on which he was senior author in which endoscopic treatment was used for 40 patients-22 patients with HGD, 18 with intramucosal carcinoma (IM Ca) were compared with esophagectomy in 61 (13 HGD, 48 IM Ca) (Zehetner et al: J Thorac Cardiovasc Surg 2011;141:39-47).
The study concluded that endoscopic therapy for high-grade dysplasia or intramucosal cancer has lower morbidity than an esophagectomy and similar survival rates during short-term follow-up, but requires multiple procedures in most patients.
"Both therapies are appropriate options, but preservation of the esophagus allows the option of a fundoplication for reflux control, perhaps further improving long-term quality of life," the authors concluded.
Dr. DeMeester listed the important considerations in selecting therapy for that study:
* Tumor factors for endoscopy included: only HGD or IM Ca.
* Visible lesions resected should have clear, deep margins.
* All intestinal metaplasia could be successfully eradicated.
Factors encouraging esophageal resection were: long length of Barrett's segment; length of esophagus; length of hiatal hernia; poor esophageal body function that leads to aphasia; multifocal disease; and multiple failures of ablation.
"It makes no sense trying to preserve an esophagus that's not working," Dr. DeMeester said.
Endoscopic mucosal resection and ablation achieved complete eradication of intramucosal increases in 69% of patients at one year, 80% at three years, and 100% at five years.
"There was no recurrence of dysplasia or cancer in patients with complete eradication of Barrett's epithelium," Dr. DeMeester said. And there were no cancer deaths after endoscopic treatment or esophagectomy.
But, he pointed out, it took five years to achieve complete eradication.
Dr. DeMeester said patients considered for EMR must: understand the pros and cons of endotherapy versus vagal-sparing esophagectomy; be willing and committed to return for frequent procedures; and agree to live with the uncertainty that goes with long-term follow-up.
"Endoscopic therapy works, but it is very work intensive and carries risk," he said. The endoscopy study cohort required 102 EMRs and 79 mucosal ablations.
On the other hand, morbidity after endoscopic treatment was 0%, versus 39% with esophagectomy.
"And the patient retains the esophagus after ablation," Dr. DeMeester said.
Dr. DeMeester concluded that endoscopic mucosal resection is very good therapy but it has to be selective, and it has to take time, and the patient has to be very cooperative.
Given that, he said about 80% of patients with HGD and IM Ca can be treated with ablation, but about 20% will still need resection.
Voting
The SSO "Great Debates" audience was asked to respond to three questions before and after the debate via handheld voting keypads:
Question 1: A 61-year old man had been undergoing routine upper GI endoscopies for symptoms of GERD. After a three-year hiatus, endoscopy revealed a 4-cm patch of Barrett's esophagus involving half the circumference of the distal esophagus. Four of six biopsies revealed high-grade dysplasia (HGD). The patient wants to know if he should undergo surgical resection. Should he?
Before Debate:
Yes: 75%
No: 25%
After Debate:
Yes: 27%
No: 73%
Question 2: What is the most compelling reason to consider resection rather than ablation as initial management in this patient:
1. Lack of perspective randomized clinical trial data supporting local ablation as at least equivalent.
Before Debate: 15%
After Debate: 11%
2. The need for ongoing endoscopic surveillance.
Before Debate: 3%
After Debate: 4%
3. Retrospective studies showing a 40 to 50 percent incidence of invasive cancer in patients undergoing resection for HGD in Barrett's esophagus.
Before Debate: 64%
After Debate: 25%
4. Concern that healing after ablation may hide an underlying evolving malignancy.
Before Debate: 14%
After Debate: 9%
5. Uncertainty about patient compliance with scheduled surveillance endoscopies.
Before Debate: 4%
After Debate: 51%
Question 3: The patient underwent RFA of the area. Repeat endoscopy was performed six months later. Which finding would exceed your threshold for recommending surgical resection?
1. Focally recurrent HGD in a background of Barrett's epithelium, in the same area.
Before Debate: 19%
After Debate: 9%
2. Nodularity with HGD in a background of Barrett's epithelium, in the same area.
Before Debate: 11%
After Debate: 13%
3. Multifocal HGD in a 6 cm patch of Barrett's epithelium.
Before Debate: 13%
After Debate: 20%
4. Biopsy showing a focal T1a cancer in the background of Barrett's epithelium.
Before Debate: 57%
After Debate: 58%
Voting Reveals Changes
In an interview after the session, the moderator-organizer of the debate who also chose the speakers, Daniel G. Coit, MD, a surgical oncologist at Memorial Sloan-Kettering Cancer Center, who is now co-leader of the MSKCC Melanoma Disease Management Team, said, "As the initial management for the uncomplicated patient, I think most of us agree that ablation, the non-surgical, much less morbid procedure, is here to stay."
But while the pre-debate audience worried more that an invasive, curable cancer might be missed with ablation, after the debate their concern shifted to patient compliance with follow-up.
"In all the time I've been doing debates, the first two questions from this debate showed about as big a swing in the audience as any I've ever seen," Dr Coit said.
He said the message coming out of this debate could be considered practice-changing if it encourages clinicians to stress to the patient the need to comply with all follow-up visits.