Preliminary Results of BÂRRX Ablation Trial in Patients with Non-Dysplastic Intestinal Metaplasia Versus Low Or High Grade Dysplasia

Preliminary Results of BÂRRX Ablation Trial in Patients with Non-Dysplastic Intestinal Metaplasia Versus Low Or High Grade Dysplasia

Abstracts S1560 ˆ RRX Ablation Trial in Patients Preliminary Results of BA with Non-Dysplastic Intestinal Metaplasia Versus Low Or High Grade Dysplas...

57KB Sizes 0 Downloads 10 Views

Abstracts

S1560 ˆ RRX Ablation Trial in Patients Preliminary Results of BA with Non-Dysplastic Intestinal Metaplasia Versus Low Or High Grade Dysplasia Franklin Tsai, Mehrdad Vosoghi, Reza Khoshini, Sonya Reicher, David Chung, Samuel French, Viktor Eysselein ˆ RRX ablation of Barrett’s esophagus Aims: Assess the response associated with BA (BE) in patients with no dysplasia (ND) vs low grade dysplasia (LGD) or high grade dysplasia (HGD), as well as short segment (SSBE, !3 cm) vs long segment (LSBE, O3 cm). Methods: Prospective trial in patients with BE confirmed by bx on prior EGD. Patients receive 2 ablations at 10 J/cm2 in the ND group or 12 J/cm2 in the LGD group during each treatment. Tolerability is recorded using a 7-day, 10-pt scale patient symptom diary. All patients receive lansoprazole 30 mg bid and have 3 & 12 mo follow-up EGD with 4-quadrant bx every 1 cm from original BE region. Any patients with all bx negative are deemed to have complete response (CR). Patients with intestinal metaplasia present on any bx or incomplete response (IR) return at 6 mo for repeat treatment. Patients with known HGD who are not surgical candidates are evaluated with EUS to rule out invasion. Those with HGD limited to the mucosa will undergo EMR if not already done, then return in 2 mo to proceed with the ˆ RRX ablation at 12 J/cm2. Results: 11 of 75 planned patients (5 male, mean age BA 57.6, 9 ND, 1 LGD, 1 HGD, 8 LSBE, mean length 5.7 cm, range 1 to 11 cm, 3 SSBE) have been treated. One patient refused follow-up EGD & dropped out of the study. 3-mo follow-up EGD is pending for 4 patients. Amongst the 6 patients (5 ND, 1 HGD, 4 LSBE) who have already returned for 3-mo follow-up EGD, all had complete healing with no evidence of BE on EGD, except 2 patients with !2 cm patches of residual BE. In the ND group, all 4 LSBE patients had IR while the one SSBE patient had CR. The patient with HGD & SSBE had CR after one treatment. The mean percentage of negative bx at 3 mo was 54% (range 20 to 88%) for LSBE versus 100% for SSBE, with a mean of 69.3% for all 6 patients. Amongst all 11 treated patients, the procedure was better tolerated by those with SSBE (mean 2/10 symptom score) vs LSBE (6.6/10). The most frequently reported symptoms were odynophagia & nausea. There were no strictures, perforations, or other serious ˆRRX ablation is safe, complications. Conclusions: Preliminary data suggests that BA well-tolerated, & effective at clearing SSBE (!3 cm). Our first patient with HGD 2 had CR with ablation at 12 J/cm . Data are pending for patients with LGD. LSBE was most associated with IR, but given the high percentage of negative bx (54%), we expect significant rates of CR after these patients return for repeat ablation. 3 MO EGD Results BY PT GROUP

ND

ND

ND

ND

ND

HGD

LENGTH (cm) BE on EGD? % NEGATIVE BX

2 N 100

3 Y 33

4 N 75

5 Y 20

8 N 88

1 N 100

S1561 Endoscopic Mucosal Resection (EMR) in Barrett’s Esophagus: ‘‘Suck and Cut’’ Versus ‘‘Band and Snare’’ Paolo Fedi, Julian A. Abrams, Efsevia Vakiani, Helen E. Remotti, Charles J. Lightdale Background: EMR has been increasingly used for the therapy of high grade dysplasia and intramucosal carcinoma in patients with Barrett’s esophagus. Two techniques are currently available: a cap assisted device (suck and cut) and a band mucosectomy device (band and snare). We compared our experience using these techniques. Method: A retrospective analysis of 20 cases of EMR carried out with each technique was reviewed. The suck and cut technique was performed using an Olympus EMR kit (K-001 and K-002) with diluted epinephrine injection before each tissue resection. The band and snare technique was performed using the Duette kit from Wilson Cook. No injection was required for this technique. The distal attachment used with each device was comparable in size and shape. Two pathologists determined the size and depth of each specimen blindly. Results: No complications occurred in any of the cases. Patients tolerated both procedures well. Significant differences were noted in the size of the specimens, with a mean of 1.26 cm and 1.06 cm in the banding versus suck and cut (p ! 0.05). No differences were noted in the depth of the tissue specimens, although injection, by disruption of the submucosal layer may induce changes in the tissue thickness (see Table 1). While

AB142 GASTROINTESTINAL ENDOSCOPY Volume 63, No. 5 : 2006

the suck and cut technique does not allow visualization of the sample while performing the resection, with the banding technique optical view is always maintained. Moreover, after placement of the band the target area can be inspected for proper placement and the band can be potentially removed in case of misplacement. Performing multiple resections by banding resulted in consistently contiguous resection margins and clean resection bases, whereas the snare and cut technique was frequently associated with visible small tissue areas left in between resections. The band and snare technique was subjectively easier and more efficient because of no requirement for injection and constant visualization of the area to be resected. In conclusion, both EMR techniques are safe and can provide comparable results. However, the band and snare technique was consistently easier and more efficient, providing larger and contiguous tissue resection without the need for injection. Table 1. Resected Specimen Comparison. Tissue sample Size Depth

Olympus Kit

Wilson Cook Duette

p value

1.06 0.553

1.26 0.573

p ! 0.05 p Z 0.6

S1562 Cost-Benefit Analysis of Capsule Endoscopy Compared to Standard Upper Endoscopy for the Detection of Barrett’s Esophagus Lauren B. Gerson, Otto Lin Background: Esophageal capsule endoscopy is a promising new technology for the detection of esophageal pathology. Potential advantages for Barrett’s esophagus screening using capsule endoscopy include absence of need for intravenous sedation, and time lost from work. Methods: We performed a cost-benefit analysis comparing the initial usage of capsule endoscopy (CE) versus conventional upper endoscopy (EGD) for the diagnosis of BE in patients with chronic GERD. Procedure and physician cost estimates were obtained from the 2005 AMA Current Procedural Terminology coding website and averages from five different geographical areas in the US were calculated. We assumed that 1% of patients would be unable to swallow the capsule and would require EGD, but that no patients would experience a complication due to CE. (Table 1) We estimated that 0.01% of patients would experience a complication post-EGD with a mortality rate of 0.005% and a mean complication cost (average of GI hemorrhage and perforation) based on DRG estimates of $10,730. We assumed that capsule endoscopy would detect 82% of BE cases (Lin et al DDW 2005) and initial EGD 82% (Eloubedi 1999). Once CE identified potential BE, however, we assumed that the sensitivity of EGD for BE diagnosis would approach 95%. Results: Using the base-case probabilities, initial EGD was less expensive and more effective compared to initial CE. (Table 2) Assuming a theoretical cohort of 10,000 patients with GERD, BE would be present in 8% or 800 patients. Capsule endoscopy followed by EGD if BE was suspected cost $1267 and initial standard endoscopy cost $571 per case of BE diagnosed. Threshold values on sensitivity analysis that would lead to CE as the preferred strategy included a probability of BE detection on capsule endoscopy of 86% or greater, probability of BE detection on EGD of 78% or less, a cost of $424 or less for CE, or a cost of $1507 or more for EGD. Conclusion: Initial EGD is a more cost effective approach for BE screening compared to initial capsule endoscopy.

Table 1. Baseline Probabilities for the Decision Model Probability or Cost BE present, CE detected BE present, EGD detected BE detected on CE after positive CE Cost, CE Cost, EGD with biopsy)

Base Case

Range

82% 82% 95% \$1120 \$570

50-100% 70-100% 75-100% \$200-\$1500 \$200-\$1200

)Includes surgical pathology costs

Table 2. Results from Analysis for Cohort of 10,000 Patients with GERD Arm Capsule Endoscopy Upper Endoscopy

Average Cost

# EGDs

# BE Detected

Complications

\$1267 \$571

2571 10,000

624 (78%) 656 (82%)

0.9 1.0

www.giejournal.org