Reliability | Feature | Figures | Comments | EL | References |
---|---|---|---|---|---|
1. Highly reliable | Basal plasmacytosis | 2G | Strongest predictor of IBD Highly consistent between studies May be focal or diffuse Basal plasma cells are normal in caecum/ascending colon | EL1a RGA | 4 18 32 |
Crypt distortion/crypt branching/abnormal crypt architecture | 2E,F | Highly consistent between studies Some interobserver variability | EL1a RGA | 1 4 18 25 26 32 33 58 | |
Crypt atrophy | 2D | <5 crypts per mm unique to IBD in one study Some interobserver variability | EL1a RGA | 1 4 18 25 33 58 | |
Irregular/villous mucosal surface | 2D | Some interobserver variability | EL1a RGA | 1 4 18 25 26 33 | |
2. Fairly reliable | Granulomas | 3A | Exclude cryptolytic granulomas; only valid for Crohn's disease; infective vs ulcerative colitis: not discriminant | EL1a RGA | 4 18 25 26 33 |
Basal giant cells | Not assessed in some studies | EL1b RGB | 25 26 33 | ||
Basal lymphoid aggregates | 2H | Can be difficult to distinguish from normal lymphoid aggregates | EL1b RGB | 18 25 26 33 58 | |
3. Less reliable | Lamina proprial chronic inflammation/hypercellularity | Variably defined; some overlap with the more precise terms above Low reproducibility | EL1b RGD | 26 33 58 | |
Paneth cell metaplasia (NOS/distal to the splenic flexure) | 3B | Weak association in some studies Utility may be restricted to longstanding disease Some interobserver variability | EL1b RGD | 4 21 26 30 | |
4. Limited data | Deep (rather than superficial) crypt abscesses | EL2b RGC | 25 | ||
Variation in crypt diameter | EL2b RGC | 4 | |||
High counts of crypt intraepithelial neutrophils/crypt lumenal neutrophils | >10 per crypt or per crypt lumen, assessed in 10 crypts | EL2b RGC | 58 |
EL, evidence levels; IBD, inflammatory bowel disease; NOS, not otherwise specified.