User login
based on results of a randomized controlled trial.
The Barrett’s Oesophagus Surveillance Versus Endoscopy at Need Study (BOSS) showed that surveillance endoscopy did not significantly improve overall survival (OS) or cancer-specific survival, compared with “at-need” endoscopy requested by patients with symptoms, reported Oliver Old, MD, of Gloucestershire Hospitals NHS Foundation Trust, Gloucester, England, and colleagues.
“Surveillance endoscopy at regular intervals is advocated by major gastroenterology societies and practiced in numerous countries worldwide to detect esophageal adenocarcinoma (EAC) early in these high-risk patients,” investigators wrote in Gastroenterology. However, “there are conflicting observational studies on the benefits of Barrett’s esophagus surveillance.”
To address this knowledge gap, Old and colleagues conducted the first randomized study of its kind. The BOSS trial was a multicenter trial conducted at 109 hospitals across the United Kingdom. Adults with an endoscopic and histologic diagnosis of BE were eligible if they were fit for endoscopy and had no history of high-grade dysplasia, esophageal adenocarcinoma, or prior upper gastrointestinal cancer. Patients with low-grade dysplasia were permitted to enroll, consistent with practice guidelines at the time.
A total of 3,453 participants were randomized between 2009 and 2011 to undergo surveillance endoscopy every 2 years or to receive at-need endoscopy triggered only by symptoms. Patients and clinicians were aware of treatment assignment. In the surveillance group, quadrantic biopsies were taken at 2-cm intervals throughout the Barrett’s segment.
The primary endpoint was OS. Secondary outcomes included cancer-specific survival (deaths from any cancer), time to diagnosis of EAC, stage at diagnosis, number of endoscopies performed, and procedure-related adverse events.
Of the 3,453 patients randomized, 1,733 were assigned to surveillance and 1,719 to symptom-driven, at-need endoscopy. Baseline characteristics were similar between groups; the mean age was 63 years, and about 70% were men.
Over the course of the trial, 25% of patients in the surveillance arm, and 9% in the at-need arm, withdrew from the trial back into clinical care, but allowed data collection of their outcomes. After a median of 12.8 years of follow-up, there was no significant difference in overall survival: 333 deaths occurred in the surveillance group (19.2%) and 356 in the at-need group (20.7%; hazard ratio [HR], 0.95; 95% CI, 0.82-1.10). Cancer-specific survival was also similar across groups, with 108 cancer-related deaths in the surveillance arm and 106 in the at-need arm (HR, 1.01; 95% CI, 0.77-1.33).
EAC was diagnosed in 40 patients in the surveillance arm (2.3%) and 31 in the at-need arm (1.8%), a nonsignificant difference (HR, 1.32; 95% CI, 0.82-2.11). Stage at diagnosis did not differ between the two groups.
Endoscopy use was higher in the surveillance arm, with 6,124 procedures compared with 2,424 in the at-need arm. Serious adverse events were rare, reported in 0.5% of surveillance patients and 0.4% of at-need patients, with most related to complications of endoscopy such as bleeding or perforation.
End-of-trial exit endoscopy, offered only to patients in the at-need group (based on data and safety monitoring committee recommendation), detected an additional eight cases of EAC and eight cases of high-grade dysplasia, but these findings were not included in the primary trial analysis.
“These data challenge current clinical practice where surveillance is advocated for all patients with BE,” the investigators wrote. “These results are likely to influence societal guidelines regarding surveillance for nondysplastic BE and inform decision making for individual patients and clinicians.”
The study was supported by the Health Technology Assessment Programme, United Kingdom. The investigators disclosed no conflicts of interest.
Old et al report the results of a herculean effort to randomize patients with Barrett’s esophagus (BE) to either scheduled endoscopy at 2-year intervals or endoscopy “at need.” While the intent was to understand the protective effect of endoscopic surveillance, this goal was frustrated by the extensive use of endoscopy in the at-need arm. All told, 59% of at-need patients underwent endoscopy at a median of 25.7 month intervals (compared to the 24.8-month median interval in the surveillance group).
This degree of endoscopy use in at-need patients complicates interpretation, since, as the authors note, such contamination would likely bias the results to the null. Also, only 26% of randomized at-need patients took advantage of the study-end endoscopy. In this group, an additional eight cancers and nine high-grade dysplasia were noted, raising the specter that undiagnosed important disease was present in the at-need group.
Given these concerns, the BOSS results do not provide compelling evidence to change clinical practice. I continue to recommend endoscopic surveillance to my BE patients. However, the trial provides valuable insight. First, it prospectively confirms the low incidence of esophageal adenocarcinoma in low-risk BE, a rate of 0.23%/patient-year. Second, a lot of endoscopy is probably not better than some endoscopy in BE surveillance, and current trends seen in guidelines toward lengthening intervals in low-risk patients are likely merited. Finally, clinicians and patients may overestimate the utility of surveillance, and a patient-centered approach, acknowledging the uncertainties of the utility of endoscopy and the low risk of progression to cancer, is appropriate.
Nicholas J. Shaheen, MD, MPH, AGAF, is the Bozymski-Heizer Distinguished Professor of Medicine and senior associate dean for Clinical Research at the University of North Carolina School of Medicine, Chapel Hill, N.C. He has no conflicts to report.
Old et al report the results of a herculean effort to randomize patients with Barrett’s esophagus (BE) to either scheduled endoscopy at 2-year intervals or endoscopy “at need.” While the intent was to understand the protective effect of endoscopic surveillance, this goal was frustrated by the extensive use of endoscopy in the at-need arm. All told, 59% of at-need patients underwent endoscopy at a median of 25.7 month intervals (compared to the 24.8-month median interval in the surveillance group).
This degree of endoscopy use in at-need patients complicates interpretation, since, as the authors note, such contamination would likely bias the results to the null. Also, only 26% of randomized at-need patients took advantage of the study-end endoscopy. In this group, an additional eight cancers and nine high-grade dysplasia were noted, raising the specter that undiagnosed important disease was present in the at-need group.
Given these concerns, the BOSS results do not provide compelling evidence to change clinical practice. I continue to recommend endoscopic surveillance to my BE patients. However, the trial provides valuable insight. First, it prospectively confirms the low incidence of esophageal adenocarcinoma in low-risk BE, a rate of 0.23%/patient-year. Second, a lot of endoscopy is probably not better than some endoscopy in BE surveillance, and current trends seen in guidelines toward lengthening intervals in low-risk patients are likely merited. Finally, clinicians and patients may overestimate the utility of surveillance, and a patient-centered approach, acknowledging the uncertainties of the utility of endoscopy and the low risk of progression to cancer, is appropriate.
Nicholas J. Shaheen, MD, MPH, AGAF, is the Bozymski-Heizer Distinguished Professor of Medicine and senior associate dean for Clinical Research at the University of North Carolina School of Medicine, Chapel Hill, N.C. He has no conflicts to report.
Old et al report the results of a herculean effort to randomize patients with Barrett’s esophagus (BE) to either scheduled endoscopy at 2-year intervals or endoscopy “at need.” While the intent was to understand the protective effect of endoscopic surveillance, this goal was frustrated by the extensive use of endoscopy in the at-need arm. All told, 59% of at-need patients underwent endoscopy at a median of 25.7 month intervals (compared to the 24.8-month median interval in the surveillance group).
This degree of endoscopy use in at-need patients complicates interpretation, since, as the authors note, such contamination would likely bias the results to the null. Also, only 26% of randomized at-need patients took advantage of the study-end endoscopy. In this group, an additional eight cancers and nine high-grade dysplasia were noted, raising the specter that undiagnosed important disease was present in the at-need group.
Given these concerns, the BOSS results do not provide compelling evidence to change clinical practice. I continue to recommend endoscopic surveillance to my BE patients. However, the trial provides valuable insight. First, it prospectively confirms the low incidence of esophageal adenocarcinoma in low-risk BE, a rate of 0.23%/patient-year. Second, a lot of endoscopy is probably not better than some endoscopy in BE surveillance, and current trends seen in guidelines toward lengthening intervals in low-risk patients are likely merited. Finally, clinicians and patients may overestimate the utility of surveillance, and a patient-centered approach, acknowledging the uncertainties of the utility of endoscopy and the low risk of progression to cancer, is appropriate.
Nicholas J. Shaheen, MD, MPH, AGAF, is the Bozymski-Heizer Distinguished Professor of Medicine and senior associate dean for Clinical Research at the University of North Carolina School of Medicine, Chapel Hill, N.C. He has no conflicts to report.
based on results of a randomized controlled trial.
The Barrett’s Oesophagus Surveillance Versus Endoscopy at Need Study (BOSS) showed that surveillance endoscopy did not significantly improve overall survival (OS) or cancer-specific survival, compared with “at-need” endoscopy requested by patients with symptoms, reported Oliver Old, MD, of Gloucestershire Hospitals NHS Foundation Trust, Gloucester, England, and colleagues.
“Surveillance endoscopy at regular intervals is advocated by major gastroenterology societies and practiced in numerous countries worldwide to detect esophageal adenocarcinoma (EAC) early in these high-risk patients,” investigators wrote in Gastroenterology. However, “there are conflicting observational studies on the benefits of Barrett’s esophagus surveillance.”
To address this knowledge gap, Old and colleagues conducted the first randomized study of its kind. The BOSS trial was a multicenter trial conducted at 109 hospitals across the United Kingdom. Adults with an endoscopic and histologic diagnosis of BE were eligible if they were fit for endoscopy and had no history of high-grade dysplasia, esophageal adenocarcinoma, or prior upper gastrointestinal cancer. Patients with low-grade dysplasia were permitted to enroll, consistent with practice guidelines at the time.
A total of 3,453 participants were randomized between 2009 and 2011 to undergo surveillance endoscopy every 2 years or to receive at-need endoscopy triggered only by symptoms. Patients and clinicians were aware of treatment assignment. In the surveillance group, quadrantic biopsies were taken at 2-cm intervals throughout the Barrett’s segment.
The primary endpoint was OS. Secondary outcomes included cancer-specific survival (deaths from any cancer), time to diagnosis of EAC, stage at diagnosis, number of endoscopies performed, and procedure-related adverse events.
Of the 3,453 patients randomized, 1,733 were assigned to surveillance and 1,719 to symptom-driven, at-need endoscopy. Baseline characteristics were similar between groups; the mean age was 63 years, and about 70% were men.
Over the course of the trial, 25% of patients in the surveillance arm, and 9% in the at-need arm, withdrew from the trial back into clinical care, but allowed data collection of their outcomes. After a median of 12.8 years of follow-up, there was no significant difference in overall survival: 333 deaths occurred in the surveillance group (19.2%) and 356 in the at-need group (20.7%; hazard ratio [HR], 0.95; 95% CI, 0.82-1.10). Cancer-specific survival was also similar across groups, with 108 cancer-related deaths in the surveillance arm and 106 in the at-need arm (HR, 1.01; 95% CI, 0.77-1.33).
EAC was diagnosed in 40 patients in the surveillance arm (2.3%) and 31 in the at-need arm (1.8%), a nonsignificant difference (HR, 1.32; 95% CI, 0.82-2.11). Stage at diagnosis did not differ between the two groups.
Endoscopy use was higher in the surveillance arm, with 6,124 procedures compared with 2,424 in the at-need arm. Serious adverse events were rare, reported in 0.5% of surveillance patients and 0.4% of at-need patients, with most related to complications of endoscopy such as bleeding or perforation.
End-of-trial exit endoscopy, offered only to patients in the at-need group (based on data and safety monitoring committee recommendation), detected an additional eight cases of EAC and eight cases of high-grade dysplasia, but these findings were not included in the primary trial analysis.
“These data challenge current clinical practice where surveillance is advocated for all patients with BE,” the investigators wrote. “These results are likely to influence societal guidelines regarding surveillance for nondysplastic BE and inform decision making for individual patients and clinicians.”
The study was supported by the Health Technology Assessment Programme, United Kingdom. The investigators disclosed no conflicts of interest.
based on results of a randomized controlled trial.
The Barrett’s Oesophagus Surveillance Versus Endoscopy at Need Study (BOSS) showed that surveillance endoscopy did not significantly improve overall survival (OS) or cancer-specific survival, compared with “at-need” endoscopy requested by patients with symptoms, reported Oliver Old, MD, of Gloucestershire Hospitals NHS Foundation Trust, Gloucester, England, and colleagues.
“Surveillance endoscopy at regular intervals is advocated by major gastroenterology societies and practiced in numerous countries worldwide to detect esophageal adenocarcinoma (EAC) early in these high-risk patients,” investigators wrote in Gastroenterology. However, “there are conflicting observational studies on the benefits of Barrett’s esophagus surveillance.”
To address this knowledge gap, Old and colleagues conducted the first randomized study of its kind. The BOSS trial was a multicenter trial conducted at 109 hospitals across the United Kingdom. Adults with an endoscopic and histologic diagnosis of BE were eligible if they were fit for endoscopy and had no history of high-grade dysplasia, esophageal adenocarcinoma, or prior upper gastrointestinal cancer. Patients with low-grade dysplasia were permitted to enroll, consistent with practice guidelines at the time.
A total of 3,453 participants were randomized between 2009 and 2011 to undergo surveillance endoscopy every 2 years or to receive at-need endoscopy triggered only by symptoms. Patients and clinicians were aware of treatment assignment. In the surveillance group, quadrantic biopsies were taken at 2-cm intervals throughout the Barrett’s segment.
The primary endpoint was OS. Secondary outcomes included cancer-specific survival (deaths from any cancer), time to diagnosis of EAC, stage at diagnosis, number of endoscopies performed, and procedure-related adverse events.
Of the 3,453 patients randomized, 1,733 were assigned to surveillance and 1,719 to symptom-driven, at-need endoscopy. Baseline characteristics were similar between groups; the mean age was 63 years, and about 70% were men.
Over the course of the trial, 25% of patients in the surveillance arm, and 9% in the at-need arm, withdrew from the trial back into clinical care, but allowed data collection of their outcomes. After a median of 12.8 years of follow-up, there was no significant difference in overall survival: 333 deaths occurred in the surveillance group (19.2%) and 356 in the at-need group (20.7%; hazard ratio [HR], 0.95; 95% CI, 0.82-1.10). Cancer-specific survival was also similar across groups, with 108 cancer-related deaths in the surveillance arm and 106 in the at-need arm (HR, 1.01; 95% CI, 0.77-1.33).
EAC was diagnosed in 40 patients in the surveillance arm (2.3%) and 31 in the at-need arm (1.8%), a nonsignificant difference (HR, 1.32; 95% CI, 0.82-2.11). Stage at diagnosis did not differ between the two groups.
Endoscopy use was higher in the surveillance arm, with 6,124 procedures compared with 2,424 in the at-need arm. Serious adverse events were rare, reported in 0.5% of surveillance patients and 0.4% of at-need patients, with most related to complications of endoscopy such as bleeding or perforation.
End-of-trial exit endoscopy, offered only to patients in the at-need group (based on data and safety monitoring committee recommendation), detected an additional eight cases of EAC and eight cases of high-grade dysplasia, but these findings were not included in the primary trial analysis.
“These data challenge current clinical practice where surveillance is advocated for all patients with BE,” the investigators wrote. “These results are likely to influence societal guidelines regarding surveillance for nondysplastic BE and inform decision making for individual patients and clinicians.”
The study was supported by the Health Technology Assessment Programme, United Kingdom. The investigators disclosed no conflicts of interest.
FROM GASTROENTEROLOGY