Table 2.
Comparison of Cancer Detection between Groups.*
| Outcome | MRI-Targeted Biopsy Group (N = 252) |
Standard-Biopsy Group (N = 248) |
Difference† | P Value |
|---|---|---|---|---|
| Biopsy outcome — no. (%) | — | — | ||
| No biopsy because of negative result on MRI | 71 (28) | 0 | ||
| Benign tissue | 52 (21) | 98 (40) | ||
| Atypical small acinar proliferation | 0 | 5 (2) | ||
| High-grade prostatic intraepithelial neoplasia | 4 (2) | 10 (4) | ||
| Gleason score | ||||
| 3+3 | 23 (9) | 55 (22) | ||
| 3+4 | 52 (21) | 35 (14) | ||
| 3+5 | 2 (1) | 1 (<1) | ||
| 4+3 | 18 (7) | 19 (8) | ||
| 4+4 | 13 (5) | 6 (2) | ||
| 4+5 | 7 (3) | 2 (1) | ||
| 5+5 | 3 (1) | 1 (<1) | ||
| No biopsy‡ | 4 (2) | 3 (1) | ||
| Withdrawal from trial§ | 3 (1) | 13 (5) | ||
| Clinically significant cancer¶ | ||||
| Intention-to-treat analysis — no. (%) | 95 (38) | 64 (26) | 12 (4 to 20) | 0.005 |
| Modified intention-to-treat analysis — no./total no. (%) | 95/245 (39) | 64/235 (27) | 12 (3 to 20) | 0.007 |
| Per-protocol analysis — no./total no. (%) | 92/235 (39) | 62/227 (27) | 12 (3 to 20) | 0.007 |
| Clinically insignificant cancer — no. (%) | 23 (9) | 55 (22) | −13 (−19 to −7) | <0.001 |
| Maximum cancer core length — mm | 7.8±4.1 | 6.5±4.5 | 1.0 (0.0 to 2.1) | 0.053 |
| Core positive for cancer — no./total no. of cores (%) | 422/967 (44) | 515/2788 (18) | — | — |
| Men who did not undergo biopsy — no. (%)‖ | 78 (31) | 16 (6) | — | — |
Clinically significant cancer was defined as the presence of a single biopsy core indicating disease of Gleason score 3+4 (Gleason sum of 7) or greater, and clinically insignificant cancer as a biopsy sample with a Gleason score of 3+3 (Gleason sum of 6). The Gleason score is composed of a primary (most predominant) grade plus a secondary (highest nonpredominant) grade; the range for a primary or secondary grade is from 3 to 5, with the Gleason sum ranging from 6 to 10, and with higher scores indicating a more aggressive form of prostate cancer.
Differences between rates are shown in percentage points, and the difference in maximum cancer core length is shown in millimeters. Differences in the percentages of men with clinically significant cancer detected and men with clinically insignificant cancer were calculated with a generalized linear mixed model (with the use of an identity link function with a binomial distribution) that included trial center as a random effect. The between-center variance estimates for the intention-to-treat analysis of the proportion of men with clinically significant cancer was 0.002 and for the proportion of men with clinically insignificant cancer was 0; the 95% prediction intervals for the detection rates of clinically significant and clinically insignificant cancer, incorporating between-center variation, were 14 to 39% and 17 to 28%, respectively, for standard biopsy, and 26 to 51% and 4 to 11%, respectively, for MRI-targeted biopsy. The difference in the maximum cancer core length was calculated with the use of a linear mixed model with trial center as a random effect. The between-center estimate of variance was 2.14; the 95% prediction interval for the maximum cancer core length, incorporating between-center variation, was 3.3 to 9.8 mm for standard biopsy and 4.4 to 10.8 mm for MRI-targeted biopsy.
In four participants in the MRI-targeted biopsy group, MRI identified at least one area with a score on the Prostate Imaging–Reporting and Data System, version 2, of 3 or greater (on a scale from 1 to 5, with higher numbers indicating a greater likelihood of clinically significant cancers), but targeted biopsy was not performed. In the standard-biopsy group, three participants declined transrectal ultrasonography–guided biopsy and underwent an MRI. The MRI revealed no areas that were suggestive of prostate cancer, and the participants did not undergo biopsy.
These participants did not complete any diagnostic test.
The intention-to-treat analysis included all the participants who underwent randomization, the modified intention-to-treat analysis excluded participants who did not complete a diagnostic test strategy, and the per-protocol analysis included only participants who underwent the randomly assigned testing procedure as specified in the protocol.
Data include men who did not undergo biopsy because they withdrew before undergoing any diagnostic test or because they did not complete the diagnostic strategy.