Table 2:
Model/Variable | HR | 95% Cl | p-value | HR | 95% Cl | p-value |
---|---|---|---|---|---|---|
Univariate Analysis | Tamoxifen Treated (n=476) | Tamoxifen Untreated (n=481) | ||||
CYP2D6 non-PM Status (AS>0) | 0.68 | 0.33–1.40 | 0.29 | 0.44 | 0.22–0.89 | 0.023 |
CYP2D6 AS (continuous) | 1.16 | 0.84–1.62 | 0.37 | 0.72 | 0.52–1.00 | 0.051 |
Age (continuous) | 0.98 | 0.96–1.00 | 0.07 | 0.98 | 0.97–1.00 | 0.044 |
Age (>50 vs. <50) | 0.67 | 0.34–1.32 | 0.24 | 0.71 | 0.41–1.23 | 0.22 |
PR status (Positive vs. negative) | 0.83 | 0.53–1.31 | 0.43 | 0.66 | 0.41–1.09 | 0.10 |
Tumor Size (vs. <2 cm) | ||||||
>2 and <5 cm | 2.01 | 1.28–3.14 | 0.0003 | 2.43 | 1.50–3.94 | 0.0002 |
>5 cm | 4.03 | 1.95–8.35 | 4.12 | 1.70–9.99 | ||
Nodes (vs. negative) | ||||||
1–3 | 1.29 | 0.80–2.07 | <0.0001 | 1.85 | 1.04–3.28 | 0.010 |
4–9 | 3.77 | 2.24–6.36 | 1.47 | 0.46–4.70 | ||
>10 | 6.52 | 3.39–12.54 | 4.41 | 1.60–12.20 | ||
Database (P01 vs. SPORE)a | 2.06 | 1.40–3.05 | 0.0003 | 4.46 | 2.70–7.36 | <0.0001 |
CYP2D6 PM Status Multivariable Modelb | Tamoxifen Treated | (n=469)d | Tamoxifen Untreated (n=476)d | |||
Age (continuous) | 0.98 | 0.96–1.00 | 0.059 | 0.99 | 0.97–1.01 | 0.19 |
Tumor Size | ||||||
>2 and <5 cm | 1.77 | 1.12–2.79 | 0.034 | 1.84 | 1.10–3.09 | 0.014 |
>5 cm | 2.07 | 0.96–4.46 | 3.20 | 1.27–8.06 | ||
Nodes | ||||||
1–3 | 1.12 | 0.69–1.83 | <0.0001 | 1.33 | 0.72–2.47 | 0.44 |
4–9 | 3.11 | 1.82–5.34 | 1.19 | 0.36–3.93 | ||
>10 | 5.14 | 2.49–10.62 | 2.26 | 0.76–6.74 | ||
CYP2D6 non-PM Status | 1.11 | 0.50–2.44 | 0.80 | 0.41 | 0.20–0.84 | 0.015 |
CYP2D6 AS Multivariable Modelc | 1.43 | 1.00–2.04 | 0.050 | 0.66 | 0.47–0.92 | 0.015 |
Abbreviations: AS: Activity score, Cl: Confidence Interval, HR: Hazard Ratio, PM: Poor Metabolizer
Violated proportional hazards assumption
Analysis stratified by database (P01 and SPORE)
HR, 95% Cl, and p-values reported are for the covariates in the final multivariable that includes CYP2D6 non- PM status. Covariate values for the models with CYP2D6 AS were not meaningfully different.
Tamoxifen treated models were stratified by database, untreated and overall models were stratified by database and PR status.
12 Patients missing age data were excluded from multivariable analyses