The 70-gene signature (MammaPrint) is the only microarray-based assay approved by the US Food and Drug Administration (FDA). This signature was designed to predict distant disease-free survival and overall survival for lymph node–negative patients. MammaPrint has subsequently been shown to have strong prognostic value in patients with node-positive tumors. The utility of this assay has been documented in several independent studies.[32-34] The value of MammaPrint is its ability to identify patients in the “low-risk” group who show a greater than 90% chance of being disease-free for a minimum of 5 years—and who might thus be spared unnecessary adjuvant therapy. Comparative studies have shown efficacy of the signature in selecting patients for adjuvant therapy compared with the St. Gallen criteria. A significant benefit of adding chemotherapy to endocrine therapy in patients classified as “high risk” by the MammaPrint assay has been documented; no significant benefit for the addition of chemotherapy in patients classified as “low risk” has also been documented. The signature was shown to improve quality-adjusted survival and had the highest probability of being cost-effective. It is being evaluated in a prospective, phase III clinical trial (MINDACT trial). The trial has recently been expanded to allow patients with one to two positive nodes.
Until recently, the MammaPrint assay was only able to be used with frozen tissues. The stability of the assay has been documented in tissues collected in RNA-preserving solutions such as RNAlater (Life Technologies, Carlsbad, Calif.). MammaPrint identifies patients with a high or low risk of recurrence and does not directly assess ER, PR, or HER2 mRNA levels. Nearly all ER-negative patients fall into the poor prognosis category; thus, the utility of the assay in ER-negative patients is unclear. Within the ER-positive subtypes, it is heavily dependent on proliferation-related genes, much like the other existing signatures. A qRT-PCR version of the MammaPrint assay, which can be used with formalin-fixed paraffin-embedded tissue, recently became available. However, only a few independent studies of this version of the test are available as yet.
The 70-gene signature was chosen from a list of 5000 differentially expressed genes. In an attempt to identify cellular pathways for targeted therapeutics, Ein-Dor et al analyzed the importance of the 70-gene set by selecting distinct sets of 70 genes from the original 5000 genes. Their analyses showed that many equally predictive lists could have been produced from the same original list. Thus, the genes included in the signature may not always have biological relevance and therefore cannot be used as candidates for targeted therapeutics. In addition, a recent study reported that random gene signatures, unrelated to cancer, are highly associated with breast cancer outcome. This is due to the fact that more than 50% of the breast cancer transcriptome is correlated with proliferation-related genes. Indeed, removal of the proliferation metagene, meta-PCNA, abrogated the prognostic value of the signature in breast cancer patients, thereby calling into question the prognostic relevance of published breast cancer signatures that are heavily dependent on proliferation.
Oncotype DX 21-Gene Assay
Oncotype DX is a Clinical Laboratory Improvement Amendments (CLIA)-approved RT-PCR assay that measures the expression of 21 genes, including 16 cancer-related genes associated with proliferation, invasion, and the HER2 and ER pathways, as well as 5 reference genes. The recurrence score (RS) is used to predict patient prognosis (ie, the likelihood of recurrence at 10 years); scores range from 0 to 100. The assay also categorizes patients into three risk groups:
1) A low-risk group (RS, 0–18), in which the score correlates with a risk of distant recurrence of less than 10%.
2) An intermediate-risk group (18 < RS < 31), in which the risk score correlates with a risk of distant recurrence of between 10% and 20%.
3) A high-risk group (RS ≥ 31), in which the score correlates with a risk of distant recurrence of greater than 20%.
This test can be used in women of all ages with newly diagnosed ER-positive stage I or II breast cancer. In addition to providing prognostic information, it predicts benefit from hormonal therapy (tamoxifen and/or aromatase inhibitors) as well from chemotherapy.[14,41-44]
The key clinical studies for the development and validation of Oncotype DX include National Surgical Adjuvant Breast and Bowel Project (NSABP) B-14, NSABP B-20, and Southwest Oncology Group (SWOG) 8814.[41,44,45] The assay was developed using the placebo-treated arm of B-20 and validated in the B-14 study. The RS predicts the likelihood of distant recurrence in tamoxifen(Drug information on tamoxifen)-treated patients with node-negative, ER-positive breast cancer. The hazard ratio for poor histological grade, used as a prognostic tool, was comparable to that of RS. However, histological grade is associated with significant subjectivity. Further analysis in the B-20 study showed differences in the response of patients to chemotherapy that correlated with the RS; a high RS was associated with response, while low and intermediate RS were associated with minimal response or benefit. The analysis of node-positive patients in the SWOG 8814 study showed that RS was also predictive in this population, although the overall risks were higher for any given score.
As with other signatures, proliferation significantly contributes to the utility of the Oncotype DX signature in ER-positive tumors. A recent study showed that an algorithm based on the IHC4 combination of four markers (ER, PR, HER2, and Ki67) could provide the same prognostic information as RS; however, these data need further validation. It has been observed that there is a high degree of concordance between ER/PR on RT-PCR results and ER/PR status as determined by IHC, and between HER2 on RT-PCR results and HER2 status as determined by FISH.[46,47] This has led to single-gene scores being added to the Oncotype DX reports. Although the concordance between RT-PCR and traditional methods of determining tumor status has been high, there are a small but significant number of cases in which the results are positive with one method but not with the other. To date, there are no data to suggest that patients negative by IHC or FISH (for ER, PR, or HER2) but positive by RT-PCR respond to targeted therapeutics. However, it should be noted that this is an area of active investigation and that several studies are planned to examine this question.
Adjuvant! Online provides information that complements RS.[42,43] Recently, Tang et al showed that the addition of clinicopathological parameters leads to an improvement in the prognostic ability of RS. The Oncotype DX assay was used in the prospective, randomized TAILORx clinical trial in which patients assigned an intermediate RS are being randomized to receive either chemotherapy and hormone therapy or hormone therapy alone. This trial enrolled more than 11,000 patients and was closed in October 2010. A trial for node-positive patients (RxPONDER; SWOG S1007) is being launched.
Emerging Technologies for Future Prognostic and Predictive Assays to Be Used in Clinical Decision Making
Recent advances in gene sequencing technology have led to the introduction of high-throughput, massively parallel sequencing, including cost-efficient whole-exome and whole-transcriptome sequencing. These methods allow for the detection of somatic cancer genome alterations and mRNA quantification. Due to the huge amount of data involved and the complex statistical and computational challenges, this technology is still in its infancy and is not used in diagnostic laboratories. However, these emerging technologies will have a tremendous near-term impact on the practice of pathology, medicine, and the entire discipline of biology. Although the whole-genome technology is expensive, two articles have demonstrated its powerful utility as a diagnostic tool in cancer patients.[51,52] The first article, by Link et al, identified a novel deletion of three exons of the TP53 gene in a woman who was diagnosed with breast cancer at age 37, diagnosed with ovarian cancer at age 39, and who died of treatment-related acute myeloid leukemia (tAML) at age 42. The second paper, by Welch et al, reported the use of whole-genome sequencing in a 39-year-old woman diagnosed with AML; the sequencing identified a cryptic fusion oncogene, a novel insertional translocation on chromosome 17 that created a pathogenic PML-RARA gene fusion. The patient’s diagnosis and treatment were changed based on this information. The patient remains in remission as reported in the article.
Molecular Profiling and Routine Clinical Practice: Guidelines and Challenges
One of the major challenges for clinicians is deciding which patients require adjuvant therapy. Although adjuvant therapy decreases the risk of recurrence and death in some cases (for example, in patients with ER-negative disease), not all patients benefit from it. Overtreatment can also cause adverse effects. Therefore, guidelines have been developed, including the National Institutes of Health (NIH) Consensus Development criteria, the National Comprehensive Cancer Network (NCCN) guidelines, the St. Gallen Expert Opinion criteria, the American Society of Clinical Oncology (ASCO) guidelines, and the computer-based algorithm Adjuvant! Online. These guidelines state that the precise clinical utility and application of MammaPrint is “under investigation.” The Oncotype DX assay is included in the ASCO guidelines on the use of tumor markers in breast cancer; the assay is also included in the NCCN guidelines for breast cancer treatment—as a predictor of recurrence for ER-positive, lymph node–negative breast cancer patients treated with 5 years of hormonal therapy. Bearing in mind that these are expensive assays, it is only natural to ask whether they are ready for prime time. Alternatively, can (almost) the same information be obtained using more traditional means?
Traditionally, oncologists have used the TNM staging system to make therapeutic decisions. In the age of mammographic screening, in which the majority of the tumors diagnosed are small with little or no extramammary involvement, the utility of the TNM system is becoming limited. The topic du jour is “what is the role of biology in small tumors?” and many editorials have discussed the inadequacy of size as the sole criterion for therapeutic decision making.[56,57] This point is forcefully brought home by small triple-negative tumors and HER2-positive tumors, which have a greater propensity for recurrence than typical ER-positive tumors.
It has been strongly argued by some that tumor grade should be included in the decision-making process. Tumor grade is a strong indicator of biology and has been shown to predict outcomes. In addition, grade is strongly associated with ER status, with the vast majority (> 95%) of low- and intermediate-grade tumors being ER-positive. Grade has also been a significant predictor in most molecular profiling studies and remains significant in multivariate analysis. This begs the question—why is grade not used? The two most common answers are 1) that grade is subjective, and 2) that a large number of tumors are grade 2. In most studies, 30% to 60% of tumors are graded as grade 2. This is an impediment to the utility of tumor grade. However, as recently highlighted by Rakha et al, a similar percentage of cases are classified as “intermediate risk” by Oncotype DX. It should also be remembered that many cases are not suitable for molecular analysis because of inadequate tumor or contamination by normal elements (as with the “normal” subtype). Also, histologic grade has limited utility in ER-negative tumors, although so do Oncotype DX and MammaPrint.
A common problem with all the gene signatures described to date is that they are dependent on proliferation. More importantly, if the contribution of this pathway is excluded from the gene signatures, the gene signatures lose their value. This has revived interest in the analysis of proliferation-related markers in archival tissues using immunohistochemistry. The IHC4 (ER, PR, HER2, and Ki67) score was shown to provide information similar to that provided by Oncotype DX. Currently, a number of efforts are directed at standardizing Ki67 immunohistochemistry for routine clinical use. In particular, attention is being paid to which antibodies can be recommended for routine use, how many fields need to be counted, and whether these fields should be taken from the leading edge of the tumor or from the center of the tumor. A multinational task force has been established to address these issues.
Because of the presence in the US market of the Oncotype DX breast cancer assay and the MammaPrint assay, the Evaluation of Genomic Applications in Practice and Prevention (EGAPP) Working Group (EWG) assessed the value of both tests. The EWG found insufficient evidence to make a recommendation for or against the use of tumor gene expression profiles to improve outcomes in defined populations of women with breast cancer. For Oncotype DX, they found preliminary evidence of a potential benefit of testing results in some women with regard to decisions about treatment options (reduced adverse events in low-risk women who avoided chemotherapy), but they could not rule out the potential for harm in others (breast cancer recurrence that might have been prevented). The EWG encouraged further development and evaluation of these technologies.
It is clear that the molecular profiling tests have a great potential to improve clinical decision making, since they address the complexity of breast cancer. We highly favor the combinatorial use of these assays with the existing traditional clinicopathologic parameters, as shown in the algorithm in the Figure. Indeed, a recent study by Tang et al used a similar combinatorial approach in which the Oncotype DX RS was integrated with clinicopathological parameters to develop a tool—the RS-Pathology-Clinical (RSPC) assessment. This model, although in need of validation, might have the greatest predictive and/or prognostic utility in cases classified as “intermediate risk” by Oncotype DX. Studies such as these highlight the difficulties in prognostication in patients with breast cancer and the need to use anatomical, histological, and biological approaches to assist with clinical decision making. It is indisputable that multigene classifiers cannot replace, but rather strengthen, prognostication and prediction in combination with clinicopathological parameters. They do not have a role in cases in which the patient (or the clinician) has already made the decision to proceed with systemic adjuvant therapy. However, these tests do have a role to play in those patients who are undecided or for whom a definite decision cannot be made based on clinicopathological findings. No test should be ordered if its results are not going to influence clinical decisions.
Financial Disclosure: Dr. Badve is a speaker for Genomic Health. Dr. Gökmen-Polar has no significant financial interest or other relationship with the manufacturers of any products or providers of any service mentioned in this article.