Table 2.
Estimation of the benefit from adjuvant tamoxifen treatment
| Tumor status | Tamoxifen treatment versus no endocrine treatment | |||||
|---|---|---|---|---|---|---|
| Distant recurrence | Breast cancer death | |||||
| Hazard ratio (95% CI) | P value | Pinteraction* | Hazard ratio (95% CI) | P value | Pinteraction* | |
| ER-positive | ||||||
| Low HOXB13 expression | 0.38 (0.23 to 0.60) | 0.000048 | 0.35 (0.20 to 0.60) | 0.00016 | ||
| High HOXB13 expression | 0.88 (0.47 to 1.65) | 0.69 | 0.035 | 0.84 (0.37 to 1.90) | 0.67 | 0.060 |
| ER-positive, PR-positive | ||||||
| Low HOXB13 expression | 0.26 (0.14 to 0.49) | 0.000022 | 0.24 (0.11 to 0.50) | 0.00015 | ||
| High HOXB13 expression | 0.70 (0.31 to 1.56) | 0.38 | 0.072 | 0.82 (0.27 to 2.56) | 0.74 | 0.059 |
| ER-positive, PR-negative | ||||||
| Low HOXB13 expression | 0.69 (0.31 to 1.56) | 0.37 | 0.64 (0.26 to 1.56) | 0.33 | ||
| High HOXB13 expression | 0.89 (0.31 to 2.54) | 0.83 | 0.51 | 0.60 (0.17 to 2.13) | 0.43 | 0.98 |
Cox regression analysis of distant recurrence rate and breast cancer-related deaths for patients with estrogen receptor (ER)-positive tumors, ER-positive and progesterone receptor (PR)-positive tumors, and ER-positive and PR-negative tumors in relation to HOXB13 protein expression. CI, confidence interval. *Tests whether there is a difference in the treatment response. The models included treatment (tamoxifen vs. no endocrine treatment), HOXB13 expression (high vs. low), an interaction variable, tumor size (>20 mm vs. ≤20 mm) and HER2 status (3+ vs. 0 to 2+).