Table 3.
Association between polyp subtypes and incidence of colorectal cancer (CRC) by cancer subsite
| Reference individuals |
Hyperplastic polyps |
SSA/Ps | Tubular adenomas |
Tubulovillous adenomas |
Villous adenomas |
Synchronous serrated polyps and conventional adenomas |
|
|---|---|---|---|---|---|---|---|
| Proximal colon cancer | |||||||
| No. of cases | 5040 | 501 | 40 | 622 | 535 | 52 | 190 |
| HR (95% CI)* | 1 (ref) | 2.14 (1.90-2.42) | 2.77 (1.84-4.18) | 2.08 (1.86-2.33) | 3.27 (2.88-3.71) | 3.69 (2.52-5.42) | 3.21 (2.63-3.91) |
| P* | <0.0001 | <0.0001 | <0.0001 | <0.0001 | <0.0001 | <0.0001 | |
| Distal colon cancer | |||||||
| No. of cases | 4061 | 175 | 12 | 291 | 286 | 29 | 95 |
| HR (95% CI)* | 1 (ref) | 0.81 (0.68-0.97) | 1.11 (0.57-2.18) | 1.18 (1.02-1.36) | 2.35 (2.01-2.74) | 2.70 (1.68-4.34) | 1.79 (1.40-2.30) |
| P* | 0.02 | 0.75 | 0.02 | <0.0001 | <0.0001 | <0.0001 | |
| Rectal cancer | |||||||
| No. of cases | 4687 | 151 | 22 | 350 | 514 | 91 | 69 |
| HR (95% CI)* | 1 (ref) | 0.62 (0.52-0.74) | 1.73 (1.05-2.84) | 1.28 (1.12-1.46) | 3.45 (3.04-3.92) | 7.49 (5.28-10.63) | 1.12 (0.85-1.48) |
| P* | <0.0001 | 0.03 | <0.0001 | <0.0001 | <0.0001 | 0.43 | |
| P for heterogeneity† | <0.0001 | 0.05 | <0.0001 | <0.0001 | <0.0001 | <0.0001 |
Abbreviations: CI, confidence interval; HR, Hazard ratio; SSA/Ps, sessile serrated adenoma/polyps.
Adjusted for family history of CRC (yes, no), income levels (quintiles), education (9 years or less, 10-12 years, >12 years, missing), number of prior clinic visits at baseline (quintiles), and number of prior colonoscopies or sigmoidoscopies at baseline (0, 1, 2 and >2). The matching factors including birth year, age, sex, and county of residence were automatically adjusted for by the stratified Cox regression.
P for heterogeneity was calculated to assess whether there was a trend across the ordinal subtypes in the polyp-CRC association using the meta-regression method with a subtype-specific random effect term.