In this large, population-based study of lymph node-negative patients not treated with chemotherapy, we found that the Recurrence Score was strongly associated with risk of breast cancer death among ER-positive patients treated with tamoxifen. We also found that the Recurrence Score was strongly associated with risk of breast cancer death among ER-positive patients not treated with tamoxifen and among ER-negative patients. In addition, we found that these associations remained after accounting for tumor size and grade, and that the Recurrence Score was able to identify a larger subset of patients with low risk of breast cancer death than was possible with either of these standard prognostic indicators.
Several limitations should be considered when interpreting our results. We lacked ER status from the medical record for a substantial proportion of patients, and we therefore classified ER status based on gene expression. However, the estimates of relative risk were not materially changed when analyses were restricted to the 84% of patients with ER status from the charts (data not shown). Because of the diagnosis years of the study, only approximately 30% of patients were treated with tamoxifen. Although this is consistent with what has been reported for other patient populations during this period [27
], it limited the numbers of tamoxifen-treated patients for analysis. Given that the cases and controls were matched with respect to tamoxifen treatment, we could not directly examine whether the Recurrence Score is able to identify patients who are likely to respond to tamoxifen therapy. We did find a stronger association between the Recurrence Score and risk of breast cancer death among patients treated with tamoxifen than among those untreated with tamoxifen, suggesting that the Recurrence Score captures response to tamoxifen therapy as well as prognosis. This is most likely explained by the fact that the expression of ER-related genes was more strongly associated with breast cancer-specific mortality in the tamoxifen-treated patients than in those not treated with tamoxifen (Figure ), and is also consistent with the established relationship between ER status of the tumor (by ligand binding or immunohistochemistry assay) and response to tamoxifen [4
]. Our results are also consistent with findings from a study conducted among participants of the NSABP B-14 clinical trial, which randomized patients to tamoxifen versus placebo [29
]. In this population, the tamoxifen benefit varied by Recurrence Score and was greatest for those with low Recurrence Score values. As expected, the strong association between the quantitative expression of the ER gene and tamoxifen benefit was largely responsible for this finding.
Currently, adjuvant hormonal and/or cytotoxic chemotherapy are recommended for most women with early-stage invasive breast cancer. Treatment decisions are based on axillary node status, age, tumor size, histologic tumor type, tumor grade, hormone receptor status (ER, PgR), and coexisting medical conditions [4
]. Hormonal therapy is recommended for nearly all women with ER-positive tumors. Although tamoxifen is generally well tolerated, a significant proportion of women experience hot flashes and leg cramps, and up to 20% do not complete a 5-year course of tamoxifen therapy [30
]. Despite its potential for serious adverse effects [33
], cytotoxic chemotherapy has been recommended for most women with lymph node-positive disease and for node-negative patients with tumors greater than 1 cm or with unfavorable pathology [4
]. Very little information is available to support the use of other clinical or biologic factors in selecting patients for adjuvant chemotherapy [4
]. Most patients with node-negative disease who receive chemotherapy will not derive benefit, because they would not go on to have a recurrence even without such treatment. New prognostic and predictive tests are needed to better individualize therapy and confine systemic treatment, especially cytotoxic chemotherapy, to those patients who are most likely to benefit.
A growing number of studies suggest that multigene expression assays may be able to provide important prognostic and/or predictive information for breast cancer patients [34
]. A variety of approaches and technologies are being used to select genes and characterize expression (for example, cDNA microarray chips, RT-PCR) in different types of pathology specimens (for example, fresh frozen tissue, formalin-fixed paraffin embedded tissue). No matter what the approach or technology, multiple, well conducted confirmatory studies using standardized methodologies will be needed before the clinical usefulness of any of these assays can be established.
This is the third study to evaluate the performance of the Recurrence Score among patients not treated with systemic chemotherapy [12
]. The three studies used identical pre-specified scores and laboratory methods and were conducted among patients who were independent of those used for gene selection and Recurrence Score algorithm development. The results of the first study, conducted by Esteva and coworkers [13
], differ substantially from our results and from those of the NSABP B-14 study [12
]. In that study no association was found between the Recurrence Score and risk for distant recurrence among a series of 149 node-negative patients who were treated without adjuvant hormonal therapy or chemotherapy at the MD Anderson Cancer Center between 1978 and 1995, who had potentially 5 or more years of follow up, and for whom archived tissue was available. In contrast to the NSABP B-14 study, patients in the study conducted by Esteva and coworkers were not treated with tamoxifen. In contrast to our study, the outcome of interest was distant recurrence instead of breast cancer death. Although these may explain some of the differences observed, it is also possible that the study by Esteva and coworkers included a nonrepresentative group of patients. Patients with poorly differentiated tumors had better prognosis than those with well differentiated tumors, and there was a suggestion that patients with ER-negative tumors did better than those with ER-positive tumors.
Our relative risk estimates for the Recurrence Score in ER-positive, tamoxifen-treated patients are generally quite similar to those observed in the NSABP B-14 study. Relative risks associated with expression of individual genes were also very similar. In the NSABP B-14 study, the 10-year risk of breast cancer death was 3.1% (95% CI 1.2–5.0%) in the low risk group, 12.2% (95% CI 6.7–17.6%) in the intermediate risk group, and 27.0% (95% CI 20.4–33.6%) in the high risk group. Although very similar to our findings for the low and intermediate risk groups, the 10-year risk for the NSABP B-14 high risk group was higher than ours. Patients in the two studies had a comparable age distribution, and although those in the NSABP B-14 study were more likely to have larger tumors, their tumors were also more likely to be well differentiated. Therefore, there is uncertainty regarding the extent to which differences in the distribution of prognostic factors in the two study populations may explain the difference observed in absolute risk estimates for the high risk group. In the Kaiser population, tumor size and tumor grade remained statistically significantly associated with risk of breast cancer death in most multivariate models that also included the Recurrence Score, whereas only tumor grade remained independently associated with risk in the NSABP B-14 study. Tumor size was determined by pathology in the Kaiser study and by clinical examination in the NSABP B-14 study. Clinical examination of tumor size is generally less accurate and could have resulted in attenuated relative risk estimates.