|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Originally published as JCO Early Release 10.1200/JCO.2004.08.965 on October 25 2004 © 2004 American Society of Clinical Oncology.
Profiling the PerpetratorCedars Sinai Medical Center, Los Angeles, CA In the classic Western movie genre, the bad guys wear black hats. Identifying all bad guys isn't always as easy, so the US Federal Bureau of Investigation's Violent Criminal Apprehension Program (ViCAP) assembles a psychological profile to catch serial killers. More recently, television shows like CSI feature sophisticated profiling techniques (DNA analysis, gas chromatography, and so on) to supplement conventional ballistics and common sense detective work to trap nefarious villains. Molecular profiling of epithelial ovarian cancers holds promise that goes beyond identifying the most aggressive tumors. Gene expression signatures may also be a cornerstone to understanding the root causes of ovarian cancer, and to designing pathway-specific targeted therapies. Clinicians and scientists have long searched for the fundamental molecular alterations that differentiate an ovarian cancer cell from its normal counterpart. Until recently, our predictions have relied on descriptive clinical and tumor characteristics such as patient age, tumor histology and stage, and extent of surgical cytoreduction.1 In vivo surrogates of chemosensitivity, such as time to CA125 normalization,2 and genetic alterations, such as mutations in BRCA1 or BRCA2 genes, have recently factored into our discussions of prognosis with patients and their families.3,4 In this issue of the Journal of Clinical Oncology, Spentzos and colleagues5 describe an Ovarian Cancer Prognostic Profile (OCPP), a 115-gene signature that seems to be an independent prognostic determinant of outcome for women with epithelial ovarian cancer. Their data imply that the OCPP is a better predictor of patients' overall survival than any of the currently available clinical criteria, and they suggest that gene profiling may have an important role in identifying patients for future investigational therapies. Using 68 frozen tumor specimens collected at two institutions, the authors performed microarray hybridization using Affymetrix U95A2 GeneChips (Affymetrix, Santa Clara, CA). The OCPP was derived from gene expression analysis of one-half the samples (the training set) using supervised statistical methods of pattern recognition, class prediction, and leave-one-out cross validation. Importantly, survival was analyzed as a continuous variable rather than dichotomizing the cohort around an arbitrary time point. As such, samples from women at the extremes of overall survival in the training set were used in the first step of developing the prognostic profile. The shortest survival was 4 months, and the longest was 73 months; a minimum of 32 months separated these groups. After optimizing the class predictors, the expression profile was refined, and the final predictive signature was developed using the multigene patterns with the highest accuracy for predicting survival. The final 115-gene predictor profile, the OCPP, was validated using the remaining half of the patient samples. Labels of "favorable" or "unfavorable" were assigned to each of the tumors in this independent validation set based on their OCPP. Kaplan-Meier analysis demonstrated statistically different overall survival rates between these two groups that also seem to be clinically important. In the favorable group, the median survival had not yet been reached at 47 months, while in the unfavorable group, the median survival was only 30 months, almost a year and a half shorter. Furthermore, there was a definite plateau in the favorable survival curve, suggesting that there is a cohort of ovarian cancer patients with a more indolent disease course and a 70% 5-year survival. When patients were segregated by traditional clinical prognosticators, the OCPP continued to discriminate between favorable and unfavorable outcomes. Examining only stage III/IV patients, only patients with grade 3 tumors, or only those who had optimal surgical cytoreduction, the molecular signature was able to further refine our ability to accurately predict patient survival. In multivariate analysis, controlling for age, stage, grade, and debulking status, the OCPP demonstrated independent prognostic value, with a hazard ratio for death of 3.6 when all 68 patients were included in the calculation. So, how can we best use this information at the current time? Eighty-one percent of patients with an unfavorable profile achieved a complete clinical remission after first-line platinum/taxane chemotherapy, and in the subset that underwent second-look laparoscopy, 55% were pathologically free of disease. It would be difficult to recommend altering our standard of care first-line chemotherapy prescription based on these response rates. Yet, this observation does challenge our belief that clinical or pathological complete response after first-line chemotherapy is a good in vivo biomarker for long-term survival. In the future, molecular profiling may allow us to more rationally select patients for investigational studies of consolidation therapy or upfront regimens that add novel agents to platinum-based combinations, or new molecular imaging modalities that detect microscopic recurrent disease. The most exciting insights from this work, however, may come from the functional classification of the genes that make up the OCPP. It's unlikely that 115 specific molecular alterations are required to transform a cell in the ovary's surface epithelium to a cancer cell. Rather, a handful of seminal events affecting key nodes in regulatory and signaling pathways likely lead to downstream disruption of normal gene transcription resulting in the malignant phenotype. Molecular profiling of tumors can potentially direct us to these key nodes and the most relevant pathways to target as we refine and individualize our future treatments and preventative strategies.6-8 Some of the gene families implicated by the OCPP that may be therapeutic targets in the near future include growth factor receptors (platelet-derived growth factor receptor), tumor invasion genes (PAI-1), angiogenesis genes (VEGF-C), and hormone-receptor associated genes (estrogen receptor binding site-associated antigen 9). While further validation studies of the OCPP are clearly required, it is reassuring that many of these genes and pathways have been previously implicated in the pathogenesis of ovarian cancer.9-14 Additionally, it is worth noting that these authors did not microdissect the tumor cells in their specimens. The OCPP therefore includes the stromal and infiltrative immune cell response. As a consequence, the impact of an intratumoral immune response on clinical outcome was factored into the molecular signature.15 Many mesenchymal genes such as fibronectin and vimentin were overexpressed in the unfavorable group and may reflect the roles of the host response and the tumor stroma in the tumors' aggressive behavior. This is another example of the robust nature of gene profiling and how its findings can help elucidate subtle aspects of ovarian cancer's in vivo biology. Gene profiling of ovarian cancer may ultimately provide a rational approach toward crafting individualized treatments combining pathway-specific targeted drugs. The well-executed studies in this article help to lay the groundwork for future investigations. Next logical steps will include cross-validating these results with those of other microarray studies and enriching the sample set for early-stage tumors. Then, not only will we be able to modify our ability to counsel patients regarding their prognosis and craft more specific therapies, but we may also glean insight into the optimal targets for early detection and prevention.16 The enormous complexities and intricacies of the human genome have inspired us toward invention and discovery. Now, less than a half decade into the 21st century, and only 50 years since the initial description of the double helix, we are beginning to harness the promise of genomics, proteomics, and informatics to improve not only our understanding of the human existence, but also to affect how we, as physicians, treat human disease. Author's Disclosures of Potential Conflicts of Interest The author indicated no potential conflicts of interest. REFERENCES
1. Bristow RE, Tomacruz RS, Armstrong DK, et al: Survival effect of maximal cytoredutive surgery for advanced ovarian carcinoma during the platinum era: A meta-analysis. J Clin Oncol 20:1248-1259, 2002 2. Hawkins RE, Roberts K, Wiltshaw E, et al: The prognostic significance of the half-life of serum CA125 in patients responding to chemotherapy for epithelial ovarian carcinoma. Br J Obstet Gynaecol 96:1395-1399, 1989[Medline] 3. Cass I, Baldwin RL, Varkey T, et al: Improved survival in women with BRCA-association ovarian carcinoma. Cancer 97:2187-2195, 2003[CrossRef][Medline]
4. Boyd J, Sonoda Y, Federici MG, et al: Clinicopathologic features of BRCA-linked and sporadic ovarian cancer. JAMA 283:2260-2265, 2000
5. Spentzos D, Levine DA, Ramoni MF, et al: A gene expression signature with independent prognostic significance in epithelial ovarian cancer. J Clin Oncol 22:4700-4710, 2004 6. Liotta L, Petricoin E: Molecular profiling of human cancer. Nat Rev Genet 1:48-56, 2000[CrossRef][Medline] 7. Huang E, Ishida S, Pittman J, et al: Gene expression phenotypic models that predict the activity of oncogenic pathways. Nat Genet 34:226-230, 2003[CrossRef][Medline]
8. Sauter G, Simon R: Predictive molecular pathology. N Engl J Med 347:1995-1996, 2002
9. Henriksen R, Funa K, Wilander E, et al: Expression and prognostic significance of platelet-derived growth factor and its receptors in epithelial ovarian neoplasms. Cancer Res 53:4550-4554, 1993
10. Konecny G, Untch M, Pilhan A, et al: Association of urokinase-type plasminogen activator and its inhibitor with disease progression and prognosis in ovarian cancer. Clin Cancer Res 7:1743-1749, 2001 11. Chambers SK, Ivins CM, Carcangiu ML: Plasminogen activator inhibitor-1 is an independent poor prognostic factor for survival in advanced stage epithelial ovarian cancer patients. Int J Cancer 79:449-454, 1998[CrossRef][Medline] 12. Kodama J, Hashimoto I, Seki N, et al: Thrombospondin-1 and -2 messenger RNA expression in epithelial ovarian tumor. Anticancer Res 21:2983-2987, 2001[Medline] 13. Hsieh CY, Chen CA, Chou CH, et al: Overexpression of HER-2/neu in epithelial ovarian carcinoma induces vascular endothelial growth factor C by activating NF-kappa B: Implications for malignant ascites formation and tumor lymphangiogenesis. J Biomed Sci 11:249-259, 2004[Medline] 14. Ueda M, Terai Y, Kumagai K, et al: Vascular endothelial growth factor C gene expression is closely related to invasion phenotype in gynecological tumor cells. Gynecol Oncol 82:162-166, 2001[CrossRef][Medline]
15. Zhang L, Conejo-Garcia JR, Katasaros D, et al: Intratumoral T cells, recurrence, and survival in epithelial ovarian cancer. N Engl J Med 348:203-213, 2003
16. Lu KH, Patterson AP, Wang L, et al: Selection of potential markers for epithelial ovarian cancer with gene expression arrays and recursive descent partition analysis. Clin Cancer Res 10:3291-3300, 2004
Related Article
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2004 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
