In this issue of ONCOLOGY, Drs. Schuchert and Luketich present an excellent review of the management of Barrett's esophagus (BE).[1] In general, this review accurately summarizes current management strategies. However, two points bear further examination. The first is the role of a surgical antireflux procedure in averting dysplasia or cancer; the second is the role of esophagectomy for patients with BE and high-grade dysplasia (HGD).
Does a Wrap Protect Against Cancer or Cause Regression of Dysplasia?
The literature regarding the fate of dysplasia and the risk of cancer following a surgical antireflux procedure in BE has multiple conflicting studies with heterogeneous methodologies. As Drs. Schuchert and Luketich note, we have performed a meta-analysis of the effect of surgical antireflux procedures on the incidence of esophageal adenocarcinoma in BE.[2] Thirty-four articles, representing more than 10,000 patient-years of follow-up, were analyzed. In sum, there was no significant difference in the cancer incidence rate between the surgical antireflux procedure group and the medical management group.
Some case series demonstrate high percentages of reversion of dysplasia among subjects treated with a surgical antireflux procedure. Such data must be viewed with some suspicion. First, the rate of sampling error in BE can be quite high, and the surface area involved in dysplasia may be only a small fraction of the overall BE. Second, spontaneous reversion of dysplasia in BE appears to be common, and indeed, in the recent randomized controlled trial comparing proton pump inhibitor (PPI) therapy to endoscopic ablation with photodynamic therapy (PDT), 39% of those in the control (PPI only) arm had reversion of the HGD to lower forms of dysplasia.[3]
Although the authors of the present review discuss several suggestive surgical case series, we believe that it is the cancer data from the meta-analysis that should be considered most compelling. As the authors note, randomized data do not demonstrate an antireflux wrap to be any better than medical therapy in averting cancer. Surgical antireflux procedures should not be held out to our patients as an antineoplastic measure based on currently available data.
What Should Be Done for the Patient With HGD?
Drs. Schuchert and Luketich note that "fit patients with Barrett's esophagus and high-grade dysplasia should undergo esophagectomy to prevent the risk of developing esophageal adenocarcinoma." While we agree that the management options include intensive endoscopic surveillance, endoscopic ablative therapy, or esophagectomy, the relative efficacy of these options remains the subject of ongoing debate[4-6] and the optimal management strategy is not yet decided.
Dr. Luketich and colleagues previously reported their results in a series of 222 patients undergoing minimally invasive esophagectomy at their center, with a 30-day mortality rate of only 1.4% (3 of 222).[7] However, data from the national level, rather than the single center level, provide a more accurate assessment of outcomes for the average patient. Birkmeyer and colleagues' seminal article on surgical mortality as a function of hospital volume using national Medicare data illustrates the difficulty with extrapolating results from high-volume centers into recommendations at the community level.[8]
Three of the findings of Birkmeyer et al are germane here: First, esophagectomy was the highest risk procedure of the 14 surgeries studied. Second, mortality varied substantially based on hospital volume. Adjusted 30-day mortality rates were 20.3%, 17.8%, 16.2%, 11.4%, and 8.4% for centers that performed fewer than 2, between 2 and 4, 5 to 7, 8 to 19, and greater than 19 esophagectomy procedures annually, respectively. Third, of the 1,575 hospitals studied, only 31 (2%) performed more than 19 esophagectomies per year. Fully 618 (39%) hospitals performed fewer than 2 esophagectomies per year, and the same proportion performed between 2 and 4 of these procedures. The reality, then, is that the majority of esophagectomies in the United States are performed at very low volume centers where mortality rates are unacceptably high.
In addition, data raise the question of whether esophagectomy confers a survival benefit over endoscopic ablation of HGD. While there are no randomized clinical trials comparing the two strategies, a recent investigation by Prasad and colleagues sheds light on this question.[9] These researchers compiled data for a large series of patients presenting to Mayo Clinic with HGD who either had endoscopic ablation with PDT or underwent esophagectomy. The patients treated endoscopically were significantly older and had more comorbidity than the surgical group. Despite these baseline differences, over the mean follow-up period of approximately 60 months, there was no difference in overall survival between the patients treated by the endoscopic vs surgical approach.
The 'Bottom Line' in Management of BE With HGD
Where do these data leave us with respect to the optimal approach for BE with HGD? The bottom line is that no global recommendation for proceeding to esophagectomy, even with a minimally invasive technique, can be made at this time. Instead, local expertise must play a critical role in the treatment algorithm for both surgeons and gastroenterologists. A young, fit patient at a center of excellence where esophagectomy volume is high may be best served by esophagectomy. Outcomes for the same patient in a low volume center may be unacceptably poor.
As the authors note, endoscopic ablative techniques such as circumferential balloon-based radiofrequency ablation[10] and endoscopic mucosal resection[11] are rapidly maturing. These may prove to be a superior approach. Until further data are available, it is incumbent on us as clinicians to make a frank assessment of our local expertise in delivering these modalities, and include this information in our discussions with the patient.
Schuchert and Luketich have provided a service by reviewing the current state of BE management. Given the rapidly evolving nature of this field, management algorithms will continue to morph as new data become available.
—Evan S. Dellon, MD
—Nicholas J. Shaheen, MD, MPH
Financial Disclosure: Dr. Shaheen has received research funds and speaker's fees from AstraZeneca and TAP, and research funds from CSA Medical and Procter & Gamble.
