Alexander Black, M.D., J.W. Li, M.D., B. Florentine, M.D., and John Barstis, M.D.

The use of adjuvant chemotherapy in early stage breast cancer has increased steadily over the last 3 decades, as randomized clinical trials have shown at least an incremental disease free survival advantage with the use of chemotherapy in progressively lower risk women. In axillary lymph node negative (N0) and estrogen receptor positive (ER+) cases, the National Comprehensive Cancer Network (NCCN) guidelines recommend chemotherapy in cancers larger than 2 cm (T2 or higher) and consideration in cancers of 1.0 to 2.0 cm. (T1c) with high risk features.^1,2 Due to the increased public awareness of breast cancer risk and the widespread use of screening mammography, a significant proportion of breast cancer at diagnosis is N0 and ER+. While in aggregate, this group of patients benefits modestly from adjuvant chemotherapy, it is important to determine which subgroup(s) derive so little relative advantage that they can avoid the toxicity and expense of chemotherapy.

The 21 gene recurrence score (RS; Oncotype DX) measures both the level of mRNA expression of 16 genes identified in gene array experiments on patient breast cancer samples that are relevant to ER+ breast cancer biology, and the level of mRNA expression of 5 control genes obtained from formalin fixed tissue.¹ The RS can identify N0 and ER+ breast cancer patients who appear not to benefit from adjuvant chemotherapy. The RS was validated initially in a retrospective analysis of 2 large, randomized National Surgical Adjuvant Breast Cancer Program (NSABP) in ER+ early stage breast cancer. In the B14 trial, 668 N0, ER+ tamoxifen-treated patients could be separated into 3 groups: low risk (LR: 1-18), intermediate risk (IR: 19-30) and high risk (HR: >30), with a 10 year distant recurrence rate of 6.8%, 14.3%, and 30.5% respectively. In addition, using samples from the B20 trial of similar women, the RS could divide women into the HR group that derived a substantial (28% relative recurrence free survival improvement) benefit from cyclophosphamide/methotrexate/ fluorouracil (CMF) chemotherapy and the LR/ IR group that had little benefit from adding CMF to tamoxifen as adjuvant therapy.¹ The RS is currently being studied in a prospective, multicenter United States trial in N0, ER+ patients (TAILORx) in which all HR patients receive chemotherapy, all LR patients receive anti-estrogen therapy alone, and all IR patients are randomized to added chemotherapy or anti-estrogen therapy alone. Hence, use of the RS in clinical decision-making has the potential to reverse the ever-increasing trend of using chemotherapy in early stage breast cancer.

To test whether the RS would reduce the use of adjuvant chemotherapy in early stage breast cancer patients seen in a University of California Los Angeles (UCLA) outpatient oncology office, we reviewed the records of 26 consecutive patients whose cancers were tested for a RS and compared the indication for using adjuvant chemotherapy with those obtained using current clinical guidelines from the NCCN. We used conservative NCCN guidelines for fit women with early stage breast cancer.² Specifically, any T2 (>2.0 to 5 cm.) or larger (T3 > 5 cm) cancer should and any T1c (>1.0 to 2.0cm) cancer patient might reasonably receive chemotherapy by NCCN criteria. The RS also separates cases into those that should (HR), should not (LR) or might (IR) be candidates for chemotherapy. The 26 cases analyzed are shown in Table 1. The clinical criteria include age, co-morbidities (other diagnoses), tumor size (T), ER and progesterone receptor (PR) status, Her2 neu expression (overexpressed defined as 3+ positive by immunohistochemical staining or amplified as defined by an increase in Her2 gene copy number by fluorescence in situ hybridization (FISH)), RS, and whether or not chemotherapy was used.

Table 1.

Our 26-case review demonstrated a significant reduction in the use of adjuvant chemotherapy by employing RS instead of just NCCN criteria. The RS confirmed a need for chemotherapy in only 3 cases, 2 of which were also found to overexpress Her2neu (ErbB2), a well-characterized high risk feature that requires chemotherapy and the anti-Her2 monoclonal antibody, traztuzumab (herceptin), by current consensus guidelines. Since the RS score weights Her2 and associated genes heavily, there appears to be no point in testing for RS if the cancer is Her2+. In those 2 patients, the RS was ordered before the Her2 results were available. Using either the RS or NCCN criteria, 9 cases were inconclusive, but both assays were in agreement as debatable chemotherapy use in only 4 cases. These situations clearly depend strongly on patient preference. Overall, using the RS, a minimum of 3 (HR only) to a maximum of 12 (HR and all IR) would receive chemotherapy. In contrast, the NCCN criteria recommended chemotherapy in a minimum of 11 (no T1c) to a maximum of 22 (all T1c) patients. Hence, the RS compared to NCCN guidelines spared an additional 8 to 10 patients of the 26 from receiving chemotherapy. In our series, 9 patients received chemotherapy (Table 1).