Predicting Cancer Recurrence
Carolyn Vachani, RN, MSN, AOCN
The Abramson Cancer Center of the University of Pennsylvania
Last Modified: February 7, 2009
The American Cancer Society estimates 178,000 new cases of breast cancer in the United States in 2007. The large majority of these patients will have localized or lymph node negative, locally advanced disease. These patients are frequently treated with adjuvant chemotherapy (chemotherapy given before or after surgery), although only a small portion of these patients will ultimately derive any benefit from this treatment, both in terms of living longer and living without breast cancer recurrence. This means a significant number of patients are subjected to the toxicities of therapy for little or no gain. Determining which patients will benefit from chemotherapy has long been the million dollar question facing clinicians.
Studies have shown that women with early stage, node negative, estrogen receptor- positive (ER+) disease will derive a 4% benefit in distant recurrence over 10 years. Yet, we know that 85% of these women would be cured with surgery (+/- radiation therapy) and tamoxifen alone. So, which women are getting the benefit, and how much more benefit are they getting? There are a few tools available to clinicians to better define the risk of relapse and benefit of chemotherapy for their patients.
The Old Standards Utilized in New Ways
While prognostic indicators have been used for some time, factors such as age, tumor size, histologic grade, menopausal status, ER/PR (progesterone receptor) status, and HER2 are still the building blocks of the decision tree. Several indices have been developed to predict clinical risk utilizing these factors. Using the results of these indices, decisions are made as to which women may benefit from chemotherapy. The Nottingham Prognostic Index assigns a point value to the tumor size, grade and lymph node status, combining them to arrive at a total score. This correlates with a risk of recurrence over 15 years, classified as low, intermediate, or high risk. The Consensus Guidelines of St. Gallen uses several indicators to predict risk of recurrence, but it classifies very few patients as low-risk (about 15%), which results in many women being treated unnecessarily.
A free online tool available to healthcare professionals, called Adjuvant! Online , helps clinicians estimate risk of recurrence or death without adjuvant therapy and the reduction to this risk with adjuvant therapy. The results are based on the meta-analysis of many clinical trials, SEER data, and available literature. The clinician enters information about the case, including the patient's age, tumor grade, size, ER/PR status, and HER2 status. The program generates a chart describing the risk to a patient with these characteristics. This is not designed to give an individualized prediction, but that of a similar population group. Adjuvant! Online was shown to accurately predict overall survival and recurrence rates in a group of 4083 women diagnosed with early breast cancer in British Columbia from 1989-1993, validating the clinical use of this tool. The tool includes sections for determining colon and lung cancer treatment benefit as well.
Gene Expression Profiling
Until now, clinicians have used the previously described prognostic indicators to predict the benefit of therapy to a particular type of breast cancer patient. We know that this method still results in treatment of patients who are gaining little in survival and risking the toxicities associated with therapy. The gene expression of a tumor could be used to better classify the tumor and predict recurrence and response to therapy. Gene expression profiling is achieved through the use of microarray technology or with reverse transcriptase-polymerase chain reaction (RT- PCR ). Both of these technologies can provide information of the expression of many genes from a patient's tumor sample. The varying levels of gene expression are combined to identify the tumor's genetic signature or gene expression profile.
This method of testing has enabled researchers to identify and describe various subtypes of breast cancer, including basal-like and luminal A and B. From this, doctors are able to tailor treatment more specifically. Various gene expressions have been identified and are being studied as a way to predict recurrence or response to therapy. Two of these are commercially available, MammaPrint ® and OncoType Dx®, and we will discuss these further.
MammaPrint®, a 70 gene signature, uses extracted RNA from a tumor to perform microarray analysis, classifying patients as having a low, medium, or high risk of recurrence. The 70 genes examined are responsible for regulating cell cycle, invasion, metastasis and angiogenesis. Several of the examined genes have no function and act as a quality control. MammaPrint® is an individual report, specific to the patient's tumor, unlike the previously available tests, which address a population. The test can be utilized in women with node negative, stage I or II, ER-positive or negative breast cancer. Researchers believe that those classified as highest risk will derive the most benefit from chemotherapy, while the low-risk group would have little gain, but there is less agreement on the benefit for the medium-risk group. While the result may be used to evaluate the need for treatment with chemotherapy, it has not yet been validated for this use and a large trial is ongoing.
To conduct the test, tumor tissue collected during surgery is taken to the pathologist; a punch biopsy is taken within one hour of collection, placed in a preservative and shipped to Agendia laboratories to perform the test. Upon arrival the sample is checked for quality (>50% of the sample must contain tumor cells with the RNA intact), and the gene expression profile is performed. Results are available within 10 working days and are sent to the ordering physician, who can also access the results online. The tumor will be classified as low, medium or high risk of recurrence. Some physicians have expressed concerns with submitting fresh tissue for analysis, as this could result in a less accurate pathology report. At the time of this article, MammaPrint® is not available in the United States, but is available in Europe.
Ongoing Research with MammaPrint®
The MINDACT (Microarray in Node negative Disease may Avoid ChemoTherapy) trial is designed to assess the ability of MammaPrint® to predict response to treatment, comparing its ability to the Adjuvant! Online tool. The trial will assign high-risk women to receive chemotherapy and, if ER+, hormonal therapy. Low-risk women will receive only hormone therapy, if ER+. Medium-risk women will be randomized to one of three arms: no therapy, anthracycline-based therapy, or docetaxel and capecitabine, then hormone therapy for ER+ tumors. The results of this trial will not be known for many years.
OncoType Dx® can be used to predict a patient's risk of recurrence over 10 years, when treated with tamoxifen, and the benefit that would be derived from adjuvant chemotherapy. The test is specifically designed for patients with stage I or II, estrogen receptor positive breast cancer that has not spread to the lymph nodes. The test assumes that the patient will be taking tamoxifen as part of her therapy, so it cannot be used to predict risk without tamoxifen. The test has not been validated for patients with tumors that do not fit these criteria (including DCIS), but studies are evaluating its use in a variety of situations.
Tumor tissue can be obtained for the test through core biopsy, lumpectomy, or mastectomy. If the tumor meets the previously mentioned criteria, samples of the tumor are submitted to Genomic Health, Inc. as formalin fixed, paraffin embedded tissue. Scientists determine the levels of expression of 21 specific genes in the tumor tissue utilizing RT- PCR. Sixteen of the genes used are cancer-related; the other 5 are used as "reference" genes. Based on the level of expression of each of these genes, a score is assigned. This is called the Recurrence Score TM and is measured on a scale of 0-100, with higher scores indicating a greater risk of recurrence.
The Recurrence Score TM and predicting recurrence
To demonstrate how the recurrence score (RS) could predict risk of distant recurrence, scientists used tumor samples from women enrolled in clinical trials previously done by the National Surgical Adjuvant Breast and Bowel Project (NSABP) (study B14) and Kaiser Permanente to validate the OncoType Dx® test. These samples were tested and correlated with the development of recurrence in those patients. This was possible since these trials had been performed over 10 years ago, and thus researchers could easily determine which patients had developed a distant recurrence. These studies demonstrated that the RS correlates with the likelihood of distant recurrence at 10 years (given in a percentage value). For example, if a woman's RS was found to have an RS of 10, the report would read "patients with a RS of 10 in the clinical validation study had an average rate of distant recurrence at 10 years of 7%". Although this test compares the tumor to a population, it is based on an individual tumor's genetic signature, making it patient specific.
The Recurrence Score TM is further broken down, based on risk of recurrence, into low-risk (RS < 18), intermediate-risk (RS between 18 and 30), and high-risk (RS > 31) groups. Researchers found that 51% of the tumor samples from the NSABP and Kaiser studies fell into the low-risk group, 22% into the intermediate-risk group, and 27% into the high-risk group. The rate of distant recurrence at 10 years was 6.8%, 14.3%, and 30.5%, for the groups respectively. The groups were also found to differ significantly in relapse rates (local and regional) and overall survival. This demonstrated that more than half of these women were at a relatively low risk of developing a recurrence.
The Recurrence Score TM and chemotherapy benefit
Although this information is valuable, doctors wanted to know if the test could be used to predict how much benefit a woman would derive from receiving adjuvant chemotherapy. Utilizing tumor samples from another NSABP trial (study B20), researchers showed that the RS could predict which patients would and would not benefit from chemotherapy. In the NSABP B20 trial, women in the chemotherapy arm received either CMF or MF ( methotrexate , fluorouracil +/- cytoxan ), leading researchers to believe the results may only be valid with those regimens. The study did find that, when separated, the CMF and MF groups had similar results. In addition, smaller studies looking at other regimens have found that the relationship between the RS and benefit from chemotherapy is likely not regimen specific.
To evaluate chemotherapy benefit, women were divided into low-, intermediate-, and high-risk groups. Women in the high-risk group (RS > 31) were found to have the greatest benefit, with an increase in the 10-year disease-free survival from 60.5% with tamoxifen alone to 88.1% with chemotherapy and tamoxifen. Women in the low-risk group (RS < 18) actually fared slightly worse when they received chemotherapy, with a 10-year disease-free survival of 96.8% with tamoxifen alone and 95.6% with chemotherapy and tamoxifen. In the intermediate group (RS 18-30), the benefit of chemotherapy is not as clear and is being studied further. As the RS rises, so do the absolute and relative benefits from chemotherapy, but it is not clear where the cut-off is for improving recurrence risk with chemotherapy.
Ongoing Research with OncoType Dx®
An ongoing trial, called TAILORx, aims to clarify the benefit of adjuvant therapy for women with intermediate Recurrence Scores. In this study, tumors of participants with estrogen receptor-positive, lymph node-negative breast cancer will be tested using Oncotype DX®. Patients with a low RS who are predicted to have good prognoses without chemotherapy will receive hormonal therapy alone. Women with a high RS will receive chemotherapy and hormonal therapy. Those participants with intermediate RS will receive hormonal therapy and be randomly assigned to chemotherapy or not, but it will be about 10 years before these results are available. In the mean time, women with intermediate scores will need to discuss their individual risk with their physicians, taking in to consideration the old standards; age, tumor size, grade and other pathologic features.
The OncoType Dx® test costs about $3,500 and is covered by Medicare and many private insurance companies. Genomic Health will assist patients and physicians' offices with insurance claims, and offers financial assistance to women who qualify for financial need. The test could determine the necessity of 6 or more months of chemotherapy treatments and all the associated costs, making it beneficial for the patient and her insurers.
In today's healthcare market, we must consider the financial implications of a new test and several investigators have performed cost analysis of the OncoType Dx® test. One group compared the cost of treating based on the RS to utilizing the National Comprehensive Cancer Network (NCCN) guidelines. Using the population of the NSABP B-14 study, the NCCN guidelines classified 615 of the 668 patients (92%) as high risk for recurrence and recommend treating with chemotherapy. The RS classified only 181 of these patients as high risk, 149 as intermediate risk, and 338 as low risk.
The investigators then evaluated a hypothetical group of 100 patients, 7-9% of whom were NCCN classified as low-risk. The RS would reclassify 2 patients from low- to intermediate- or high-risk and 45 patients from intermediate- or high-risk to low- risk. The testing will cost $350,000, but will in turn reduce chemotherapy costs (and associated costs) by $754,000, assuming the low-risk group did not receive chemotherapy and the intermediate and high groups did receive chemotherapy.
Concerns and Future Research
Though both available tests have been shown to predict recurrence, the 2 available gene expressions have only one gene in common. Another group is studying a 76- gene signature, which overlaps the MammaPrint® by only 3 genes. Is it possible that different gene sets can predict similar outcomes? Some investigators say yes, while others feel a comparison study would help clarify this concern. It has been argued that one cannot expect these signatures to replicate each other, given the varying inclusion criteria (lymph node status, ER status, etc), platform ( DNA array, PCR ) and data analysis methods used, but the debate continues. Many cancer centers have adopted utilizing the testing in the setting of a clinical trial, while others prefer to wait for more study results.
Additional research hopes to utilize genetic signatures to predict benefit for neoadjuvant therapy, response to particular chemotherapy agents and survival and response to therapy for other types of cancer.
Nurses play a vital role in assisting patients with complex decisions regarding cancer treatment. Gene expression profiling can be helpful in this decision making process and it is imperative that nurses understand the clinical utility of these tests and be able to help patients interpret results. As this area of cancer care continues to develop and grow, nurses must keep abreast of the clinical applications and the evidence supporting them.
American Cancer Society , Facts and Figures 2007.
Andre, F., C. Mazouni, et al. (2006). " DNA arrays as predictors of efficacy of adjuvant/neoadjuvant chemotherapy in breast cancer patients: current data and issues on study design." Biochim Biophys Acta 1766(2): 197-204.
Buyse, M., S. Loi, et al. (2006). "Validation and clinical utility of a 70-gene prognostic signature for women with node-negative breast cancer." Journal of the National Cancer Institute 98(17): 1183-92.
Creighton, C. J. and J. M. Rae (2006). "When will tumor gene expression profiling be incorporated into clinical breast cancer decision making?" Breast Cancer Research 8(4): 302.
Habel, L. A., S. Shak, et al. (2006). "A population-based study of tumor gene expression and risk of breast cancer death among lymph node-negative patients." Breast Cancer Research 8(3): R25.
Hornberger, J., L. E. Cosler, et al. (2005). "Economic analysis of targeting chemotherapy using a 21-gene RT- PCR assay in lymph-node-negative, estrogen-receptor-positive, early-stage breast cancer." American Journal of Managed Care 11(5): 313-24.
Lyman, G. H., L. E. Cosler, et al. (2007). "Impact of a 21-gene RT- PCR assay on treatment decisions in early-stage breast cancer: an economic analysis based on prognostic and predictive validation studies." Cancer.
Olivotto, I. A., C. D. Bajdik, et al. (2005). "Population-based validation of the prognostic model ADJUVANT! for early breast cancer." Journal of Clinical Oncology 23(12): 2716-25.
Paik, S., S. Shak, et al. (2004). "A multigene assay to predict recurrence of tamoxifen-treated, node-negative breast cancer." New England Journal of Medicine 351(27): 2817-26.
Paik, S., G. Tang, et al. (2006). "Gene expression and benefit of chemotherapy in women with node-negative, estrogen receptor-positive breast cancer." Journal of Clinical Oncology 24(23): 3726-34.
Sims, A. H., K. R. Ong, et al. (2006). "High-throughput genomic technology in research and clinical management of breast cancer. Exploiting the potential of gene expression profiling: is it ready for the clinic?" Breast Cancer Research 8(5): 214.
Sparano, J. A. (2006). "The TAILORx trial: individualized options for treatment." Community Oncology 3: 494-96.
Swain, S. M. (2006). "A step in the right direction." Journal of Clinical Oncology 24(23): 3717-8.
Twombly, R. (2006). "Breast cancer gene microarrays pass muster." Journal of the National Cancer Institute 98(20): 1438-40.