Table 1.
Key differences in the three non-endoscopic cell collection devices studied for the detection of Barrett’s oesophagus
| Test | Cytosponge-TFF3 | EsophaCap with 5 gene methylation panel | EsoCheck with EsoGuard (2 gene methylation panel) |
|---|---|---|---|
| Device type | Pan-oesophageal sampling device | Pan oesophageal sampling device | Targeted distal oesophagus sampling device |
| Characteristics | Not operator dependent | Not operator dependent | Operator-dependent balloon inflation and deflation |
| Biomarkers | TFF3–IHC with potential for AI-assisted reporting [45, 46] | MDMs: VAV3, ZNF682, NDRG4, FER1L4, and ZNF568 [47] | MDMs: VIM and CCNA1 [48] |
| Safety record |
•Detachment rate •1/2672 (<0.1%) [49] •1/1654 (<0.1%) [50••] •Minor bleeding: 1/2672 (<0.1%) [49] •Sore throat : 63/1654 (4%) [50••] |
5/268 (2%) reported AE-1 tether detachment and 4 other events [47] | No AE recorded, and data on safety is limited [48] |
| Evidence trial type |
•Multicentre cohort study in primary care (BEST1) [51] •A multicentre case control study—11 UK hospitals (BEST2) [52] •RCT (cluster and individual randomised) in >13,000 individuals in primary care (BEST3) [50••] |
Multisite case-control study—three tertiary care centres and 1 community hospital in the US [47] | Non-randomised observational study—1 tertiary care tertiary care institution in the US [48] |
| Sample size (n), sensitivity, and specificity | BEST2 study: 1110 [52] (463 controls, 647 BO), sensitivity = 79.9%; 87.2% (≥ 3 cm BO), (per protocol including inadequate sampling), and specificity = 92.4% | 295 (89 controls, 112 BO, 67 indeterminate) [47], sensitivity = 92%, and specificity = 94% | 156 (36 controls, 42 BO) [48], sensitivity = 90.3%, and specificity = 91.7% |
| Overall acceptability score (10-scale grading) |
High score most acceptable 6 (IQR 5–8), n= 2418 9 (IQR 8–10), n=1654 |
Low score most acceptable 2 (IQR 0, 4), n=268 |
Low score most acceptable 1–2, n=92 |
| RCT evidence |
Yes Primary endpoint: detection Barrett’s increased with rate ratio 10.6 (95% CI 6.0–18.8) compared with usual care, p<0.0001 [50••] |
No | No |
| Regulatory approval for device and assay |
CE-marked—yes (device only) FDA 510(k)—yes |
CE-marked—yes (device only) FDA 510(k)—not yet |
CE-marked—not yet FDA 510(k)—yes |
| Other trials in progress | Observational, multicentre implementation research study (ISRCTN91655550) | Case-control study, n= 2500 (ClinicalTrials.gov Identifier: NCT04214119) | Multicentre single-arm study, n= 1000 (ClinicalTrials.gov Identifier: NCT04293458) |
IHC immunohistochemistry, MDMs methylated DNA markers, AE adverse effect, OGD oesophagogastroduodenoscopy, BO Barrett’s oesophagus, BEST Barrett’s oesophagus screening trial