Current best practices and rationalistic perspectives in causation-based prevention, early detection and multidisciplinary treatment of breast and gastric cancer


Volume 2 - Number 1 - Jan/May 2003

PERSPECTIVE
Gene-Expression Profile: The Future in the Outcome Prediction and Treatment of Breast Cancer


Dimitrios H. Roukos, MD, Nikolaos Pavlidis, MD and Niki J. Agnantis, MD

From the Departments of Surgery (DHR), Medical Oncology (NP) and Pathology (NJA) Ioannina University Medical School,
451 10 Greece , droukos@cc.uoi.gr

ABSTRACT

The malfunctioning of specific genes drives probably the development and progression of neoplastic disease. Indeed, recent studies using microarray analysis provide evidence that breast cancer tumors can be classified according to their gene-expression profiling into subgroups with poor prognosis and good prognosis. These works represent a considerable advance in the clinical application of microarray technology.

The fascinating of these studies is that the DNA-microarray data predict the outcome –overall survival and distant metastases- much better than such classic prognostic indicators as nodal status, grade, stage, and estrogen-receptor status.

The finding that the genetic signature of a breast cancer is strongly associated with outcome independently of nodal status has important biologic implications and may change the current therapeutic strategy regarding clinical decision making about adjuvant chemotherapy in some individuals who at present either overtreated or undertreated.



Although the incidence of breast cancer in the western world still remains high, mortality has been reduced over the last decades. This improvement is partially attributable to an effective postoperative adjuvant treatment -radiation, chemotherapy, hormone therapy.[1-6] Decision making is based on a combination of clinical and tumor characteristics: age, tumor size, lymph-node and estrogen-receptor status, and histology (type, grade).[3-6] However, the ability of these criteria is imperfect and as result some patients undertreated and others overtreated with respect to adjuvant treatment.[7] So some individuals more harm than benefit from such a conventional risk prediction-based adjuvant treatment. For example, about one third of women among those with node-negative breast cancer will develop recurrence[1,2] and might benefit from an adjuvant systemic therapy, and about one third among those with node-positive disease will remain free of recurrence 10 years after treatment[1,2] and thus would eventually benefit avoiding a toxic adjuvant chemotherapy.

Why do some individuals with identical clinical, histological, and therapeutic criteria recur after treatment and die of the disease, while others not? It is apparent that still unidentified molecular factors should determine the ability of the primary tumor to recur or metastasize. Conventional staging systems currently used have limited predictive power for an individual patient. The rational for the search of new molecular markers is reliably receiving increasingly considerable attention because this is the most promising way towards a highly accurate outcome prediction. Predicting who individual is really at very high or low risk for metastasis we can best design an adjuvant treatment highly effective targeting to an individual patient resulting in substantial benefits in both survival and quality of life.

Molecular analysis of the primary tumor raises strong hope towards an accurate prediction of tumor behavior and represents the current major focus of cancer research. However, until now thousands of studies have failed to identify new factors with predictive power,[8] with the exception of estrogen and progesterone receptors and the HER/neu gene that are the only molecular targets routinely used in breast cancer.

The Human Genome Project revealing detailed information about the structure of all human genes, makes now feasible the search for such predictive markers. Powerful new high-performance screening techniques have been developed for molecular analysis. One of these new methods is the DNA microarray, which allows simultaneous analysis of the expression of thousands of genes in a tissue in a single experiment. The use of DNA microarrays has shown distinct patterns in different kinds of tumors.

Microarrays are small glass plates or nylon membranes to which specific sequences of thousands of genes adhere. Two different kinds of target DNA sequences are widely used: complementary DNA (cDNA) sequences or oligonucleotide sequences representing small but highly specific segments of the target genes. As probes, complementary DNA or complementary RNA (cRNA) prepared from RNA derived from tumor tissue and the appropriate control tissue are used. There are two main ways of using data from a DNA microarray: systematic search for single genes that may associated with prognosis and the use of the entire set of expressed genes to classify tumors according to the microarray derived gene-expression profiles.[9]

Histologically similar tumors can be classified, on the basis of these patterns, into specific subtypes.[9] Using complementary DNA (cDNA) microarrays to analyze breast-cancer tissue, Perou et al. identified two subgroups with distinct gene-expression profiling.[11] These two subgroups had different clinical outcome.[12] In addition, microarray analysis has been used to distinguish breast cancers associated with BRCA1 or BRCA2 mutations[13,14] and to determine estrogen-receptor status[11,14,15] and lymph-node status.[16,17]

Recently, van 't Veer et al. identified a specific gene-expression profile that is associated with prognosis in patients with breast cancer.[14] Using oligonucleotide microarrays the expression patterns of 98 women younger than 55 years with primary tumors less than 5 cm in diameter and with lymph-node–negative disease were analyzed. Within a set of 25,000 genes, they could identify a set of 70 genes only with an expression pattern that allowed highly accurate classification of the patients into those with a poor prognosis and those with a good prognosis.

More recently the same research team reports an extension of their original study that includes not only patients with lymph-node–negative but also with lymph-node–positive breast cancer.[18] The previous established 70-gene prognosis profile[14] was tested in this study. The 295 stage I or II breast cancer patients were classified\on the basis of their expression prognosis profile into two groups; 180 patients had a poor-prognosis signature and 115 patients had a good-prognosis signature.

At 10 years, overall survival and disease-free from distant metastases rates were significantly lower in the poor-prognosis (54% and 50%) than in the good-prognosis signature group (94%, and 85%) respectively. The estimated hazard ratio for distant metastases in the group with a poor-prognosis signature, as compared with the group with the good-prognosis signature, was 5.1 (95% CI, 2.9-9.0; P<0.001). This ratio remained significant when the groups were analyzed according to lymph-node status. The hazard ratio for distant metastases in the group of 151 patients with lymph-node–negative disease and the group of 144 patients with lymph-node–positive disease was 5.5 (95% CI, 2.5 to 12.2; P<0.001) and 4.5 (95% CI, 2.0 to 10.2; P<0.001) respectively among those with a poor-prognosis signature as compared with those with a good-prognosis signature.

These data clearly indicate the prognostic power of gene-expression profiling, but the fascinating result with potential important clinical implication derived from the multivariable analysis. The gene-expression signature was a stronger independent predictor of the outcome than any other currently used prognostic indicators as nodal status, stage, grade, and estrogen-receptor status. The poor-prognosis signature was by far the strongest predictor of the likelihood of distant metastases, with an overall hazard ratio of 4.6 (95% CI, 2.3 to 9.2; P<0.001). The finding that the development of hematogenous distant metastases are independent from the presence or absence of lymph node metastases may suggest that hematogenous spread and lymphatic spread of the disease are events that represent two distinct pathways in the genetic evolution of cancer.[7]

But the most important clinical implication of the findings by van 't Veer et al.[14] and van de Vijver et al.[18] is the ability of the genetic signature to identify patients at high-risk or low risk within the groups with node-negative or node-positive cancer.

The ability of prognosis signature in the study by van de Vijver et al. to predict the risk of distant metastases in node-positive patients is clinically important. Currently all node-positive patients receive adjuvant chemotherapy since the presence of node metastases is by itself a strong predictor of poor survival. The importance of the prognosis profile is that node-positive patients with poor-prognosis profile could eventually benefit from a more aggressive chemotherapy. However, no estimation can be made for the prognostic value of the profile in patients with untreated lymph-node–positive disease, since most of these patients in this study received adjuvant chemotherapy or hormonal therapy (120 of 144 patients).[18] Most patients with node-negative disease can be effectively treated with local therapy consisted of surgery and radiation since there is no evidence that cancer cells have spread beyond the primary tumor. However, a proportion of node-negative patients develops recurrence. On the basis of various clinical and histological characteristics, the St. Gallen criteria[4] and the National Institutes of Health (NIH) consensus criteria[5] have been developed to classify node-negative patients into low-risk and high-risk subgroups. Decision about adjuvant treatment is currently based on this classification. van de Vijver et al. compared the predictive value of the gene-expression profile, the St. Gallen and the NIH subgroups. This comparison shows a more accurate classification of node-negative patients by prognosis profile. Since the St. Gallen and the NIH criteria misclassify patients who would be overtreated or undertreated in current clinical practice, the prognosis profile-based classification into low-risk and high-risk subgroups would result in a more effective adjuvant treatment targeting to the appropriate individual patient.

The fact that prognosis profile is based on a small number of genes raises strong hope for the development of a new highly accurate, affordable, and commercially available biologic staging system. Indeed, the set of genes that distinguished the good-prognosis and poor-prognosis groups consisted of only 70 genes. The majority of genes do not obviously affect the clinical outcome of breast cancer.

The results of the studies using cDNA or cRNA data are just one example of how gene-expression profiling can provide highly useful prognostic information, point out Sauter and Simon,[9] who believe that it is almost certain that analogous studies will yield similar results for virtually all other types of tumors. Moreover, they anticipate that simultaneous expression analysis of the majority of human genes will identify molecular profiles that are linked to the response to treatment.[9]

It is certain that gene-expression profiling will increasingly be received attention. However, there are several limitations and many questions should be answered before molecular profile staging system can be intergraded into the routine clinical management of breast cancer. First of all, the results obtained by these studies[14,18] should be validated by other research teams in larger groups of patients since we have experienced many times over the last years that initially strongly promising data not to be subsequently reproduced in other laboratories.

An eventually validation will open the way for testing the predictive power of molecular profiling for clinical decision-making treatment. Randomized clinical trials are needed for testing the effectiveness of adjuvant systemic therapy on molecular profile-based subgroups of patients. Only the results from such trials can establish whether currently undertreated or overtreated patients may benefit from a prognosis profile-based adjuvant treatment. Furthermore, new studies are needed to evaluate critically whether the analyzed set of 70 predictor genes needs to be refined or expanded[7] to increase the accuracy of predictive power of gene-expression-based staging systems. There are promises that microarrays data combined with the conventional clinical and histologic criteria and new biologic markers, such cyclin E which is reviewed in this issue of the Journal,[19] will define a more accurate prediction system than the currently used.

On microarrays-based surgical decision-making about the extent of surgery is another future important clinical implication of gene signature. Although the current data are limited to the prediction of distant metastases only, new microarray studies might be focused on local failures prediction. The potential of gene-expression profile to accurately predict lymph node status and local failure makes able a surgical decision between minimally invasive approach, breast-conserving surgery with or without axillary lymphadenectomy, and total mastectomy. Currently, this decision based on classic criteria is imperfect and breast conservation therapy is associated with increased risks of positive margins and local failures [20,21] in this issue of the Journal]. DNA-microarrays is likely to have clinical implications not only for breast cancer, but also for multiple other solid tumors[22] including gastric cancer.[23]

A major obstacle for broad application of gene-expression profiling and other new techniques will be the requirement of fresh tumor tissue for DNA-microarray analysis,8 that is in contrast to the established, inexpensive method of formalin-fixed tissue handling in universal use.

The data of the gene-expression profile studies reflect the rational in the incorporation of microarray technology into clinical use. These exciting DNA-microarrays findings provide evidence that the way towards a highly accurate prediction of the outcome in an individual patient has been opened. A genetic signature-based targeted adjuvant conventional treatment in individuals is the next step. Our great expectation in the new millennium however, is the development of new molecular agents targeting genetic abnormalities that drive cancer evolution, progression and metastatic potential of a tumor.

References
1. Early Breast Cancer Trialists' Collaborative Group. Polychemotherapy for early breast cancer: an overview of the randomised trials. Lancet 1998;352:930-942.
2. Early Breast Cancer Trialists' Collaborative Group. Tamoxifen for early breast cancer: an overview of the randomised trials. Lancet 1998;351:1451-1467.
3. Bergh J, Johnsson PE, Glimelius B, et al. A systematic overview of chemotherapy effects in breast cancer. Acta Oncol 2001; 40: 253-81.
4. Goldhirsch A, Glick JH, Gelber RD, Coates AS, Senn HJ. Meeting highlights: International Consensus Panel on the Treatment of Primary Breast Cancer: Seventh International Conference on Adjuvant Therapy of Primary Breast Cancer. J Clin Oncol 2001;19:3817-3827.
5. Eifel P, Axelson JA, Costa J, et al. National Institutes of Health Consensus Development Conference Statement: adjuvant therapy for breast cancer, November 1-3, 2000. J Natl Cancer Inst 2001;93:979-989.
6. Elisaf M., Roukos DH, Briasoulis E, Michos KM, Agnantis NJ, Kappas AM. Comparing recent advances in gastric and breast cancer. Gastric Breast Cancer 2002; 1(3): 59-64
7. Kalloniemi A. Molecular signatures of breast cancer – predicting the future. N Engl J Med 2002 Dec;347:2067-8.
8. Isaacs C, Stearns V, Hayes DF. New prognostic factors for breast cancer recurrence. Semin Oncol 2001;28:53-67
9. Sauter G, Simon R. Predictive molecular pathology. N Engl J Med 2002 Dec.;347:1995-6.
10. Alizadeh AA, Ross DT, Perou CM, van de Rijn M. Towards a novel classification of human malignancies based on gene expression patterns. J Pathol 2001;195:41-52.
11. Perou CM, Sorlie T, Eisen MB, et al. Molecular portraits of human breast tumours. Nature 2000;406:747-752.
12. Sorlie T, Perou CM, Tibshirani R, et al. Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc Natl Acad Sci U S A 2001;98:10869-10874.
13. Hedenfalk I, Duggan D, Chen Y, et al. Gene-expression profiles in hereditary breast cancer. N Engl J Med 2001;344:539-548.
14. van 't Veer LJ, Dai H, van de Vijver MJ, et al. Gene expression profiling predicts clinical outcome of breast cancer. Nature 2002;415:530-536.
15. Gruvberger S, Ringner M, Chen Y, et al. Estrogen receptor status in breast cancer is associated with remarkably distinct gene expression pe experienced way atterns. Cancer Res 2001;61:5979-5984.
16. West M, Blanchette C, Dressman H, et al. Predicting the clinical status of human breast cancer by using gene expression profiles. Proc Natl Acad Sci U S A 2001 ;98:11462-11467.
17. Ahr A, Karn T, Solbach C, et al. Identification of high risk breast-cancer patients by gene expression profiling. Lancet 2002;359:131-132.
18. van de Vijver MJ, He YD, van 't Veer LJ, et al. A gene-expression signature as a predictor of survival in breast cancer. N Engl J Med 2002;347:1999-2009.
19. Agnantis NJ, Roukos DH, Paraskevaidis E, Elisaf M. Cyclin E: a new biologic prognostic marker for breast cancer? Gastric Breast Cancer 2003; 2(1): 9-11.
20. Roukos DH, Agnantis NJ, Kappas AM. Breast-conserving treatment for breast cancer: rational, current reality, and limitations. Gastric Breast Cancer 2003; 2(1): 12-14.
21. Roukos DH, Agnantis NJ, Batsis H, Paraskevaidis E, Kappas AM. Breast-conserving surgery and risk of positive margins in breast cancer. Gastric Breast Cancer 2003; 2(1): 15-19.
22. Ramaswamy S, Ross KN, Lander ES, Golub TR. A molecular signature of metastasis in primary solid tumors. Nat Gen 2003; 33: 49-54.
23. Roukos DH. Tumor stage-based tailored therapeutic strategy for gastric cancer: rational approach or new trend? Gastric Breast Cancer 2003; 2(1): 1-4.

Online ISSN : 1109 - 7647
   Print ISSN : 1109 - 7655

We subscribe to the HONcode principles of the HON Foundation. Click to verify. We subscribe to the HONcode principles. Verify here.
please, read our policy about privacy and confidentiality of information and transparency of sponsorship

last update: 22 May 2003