Androgen Deprivation No Better than Conservative Approach in Localized Prostate Cancer

Using androgen deprivation therapy alone in older men with localized prostate cancer does not improve survival outcomes when compared with a conservative management approach, according to a new analysis in the July 8 Journal of the American Medical Association. For localized prostate cancer, androgen deprivation therapy has become a common treatment, despite a lack of definitive evidence that it is more effective than other options, noted the study authors.

To conduct the study, the researchers, from The Cancer Institute of New Jersey, used data from NCI's Surveillance, Epidemiology, and End Results (SEER) program and linked Medicare files to compare the outcomes in men aged 66 and older, who were diagnosed with localized prostate cancer between 1992 and 2002 and had not undergone surgery. Of these men, nearly 8,000 received androgen deprivation therapy and approximately 11,400 were treated with conservative management, often called "watchful waiting."

Patients treated with androgen deprivation therapy - half of whom were treated for more than 30 months - had a lower 10-year, cancer-specific survival rate than those who underwent conservative management (80.1 percent vs. 82.6 percent) and equivalent 10-year overall survival rates. Androgen deprivation therapy did slightly improve cancer-specific survival at 10 years in men with poorly differentiated cancer, the authors noted, but there was no significant difference in overall survival in that patient subset.

"There have been more reports of health risks such as fractures, diabetes, heart disease, and other adverse effects associated with chronic use of this therapy," said the study's lead author, Dr. Grace L. Lu-Yao. "Therefore, it is imperative that more exploration is done on the appropriate application of this treatment."

Testing SNPs Improves Breast Cancer Risk Tool Modestly

In recent years, at least seven common genetic variants - single nucleotide polymorphisms, or SNPs - have been definitively linked to breast cancer. Researchers have begun to ask whether this knowledge can improve current tools for assessing breast cancer risk. The answer is yes, but only modestly, according to an analysis by Dr. Mitchell Gail of NCI's Division of Cancer Epidemiology and Genetics in today's Journal of the National Cancer Institute.

Dr. Gail led the team that developed NCI's Breast Cancer Risk Assessment Tool (BCRAT), which is based on questions about a woman's family and reproductive histories. The tool has helped women make decisions such as whether to take a drug to prevent breast cancer.

The new study found that a panel of seven SNPs was not as informative as the BCRAT in assessing a woman's breast cancer risk and therefore should not replace it as a screening tool. When the SNPs were combined with the BCRAT, they modestly improved the "discriminatory accuracy" of the BCRAT, which is the ability to distinguish women who will develop breast cancer from women who will not.

"These SNPs can improve the discriminatory accuracy of the BCRAT, but it is a small incremental change," Dr. Gail notes. "One would do better by asking women the questions [about their health and family histories] than by looking at these SNPs alone."

The SNPs, reported by two genome-wide association studies and a study of candidate genes, are located in FGFR2, TNRC9, MAP3K1, LSP1, CASP8, chromosomal region 8q, and chromosomal region 2q35.

Dr. Gail estimates that if many additional risk SNPs are identified, it could improve the risk prediction models substantially, but this will need to be evaluated once the true scope of genetic susceptibility factors for breast cancer are known.

Noninvasive Approach to Monitoring NSCLC Shows Promise

In a small study involving patients with metastatic non-small-cell lung cancer (NSCLC), researchers have shown that rare, circulating tumor cells (CTCs) in the blood can be molecularly analyzed before and after treatment to identify mutations that can affect treatment outcomes and influence decision making.

To conduct the study, published online June 2 in the New England Journal of Medicine (NEJM), researchers from Massachusetts General Hospital used an investigational CTC-chip device to capture the CTCs. They then used the Scorpion Amplification Refractory Mutation System (SARMS) assay to analyze the cells for mutations in the epidermal growth factor receptor (EGFR) gene that have been associated with response and eventual resistance to EGFR-targeted agents.

"Together with the CTC-chip, the SARMS assay may allow for noninvasive genotyping in patients with [NSCLC], which could then be repeated at therapeutic decision-making points during a patient's course of therapy," wrote study leader Dr. Shyamala Maheswaran and colleagues.

Using the CTC-chip, the researchers isolated CTCs in all 27 patients in the study in sufficient quantities and with sufficient purity for molecular analyses. So-called "activating" mutations in the EGFR gene were detected in the CTCs from all patients; a specific mutation that is associated with resistance to EGFR-targeted agents was detected in 55 percent of the 20 patients who had CTC and/or tumor and plasma specimens available for molecular analysis.

Reductions or increases in CTCs were associated with response to treatment and progression of disease, respectively, the authors wrote.

In the CTCs from several patients, they also identified the emergence of additional EGFR-activating mutations during treatment, which "we therefore presume… represents the emergence of different tumor clones," they wrote.

In a related NEJM commentary, Dr. Joan Schiller from the Harold C. Simmons Comprehensive Cancer Center in Texas said the study represents a proof of principle. "The ease with which this approach can be used by others remains to be determined," she wrote.

Glioblastoma Stem Cell Signature Identified

Researchers from the University of Lausanne have identified a "stem-cell like" gene signature associated with poorer survival outcomes in patients with glioblastoma treated with radiation therapy and temozolomide.

The researchers analyzed 80 tumors taken from patients in two other prospective clinical trials investigating glioblastoma (one phase II and one phase III trial), they reported in the June 20 Journal of Clinical Oncology. Patients whose tumors expressed a signature dominated by homeobox (HOX) genes, which play an important role in development, had outcomes nearly three times worse than patients without that signature.

The researchers suggest that the HOX signature belongs to a stem-cell-like phenotype that responds poorly to the temozolomide-radiotherapy regimen that has shown some success in treating glioblastoma. The results support the idea that tumors with the signature may include mutated self-renewing progenitor cells which give rise to malignant gliomas that are resistant to radiotherapy and chemotherapy, compared with other glioblastoma cells.

This work underlines "the need for multimodality treatments" aimed at the factors that determine outcomes in glioblastoma, the researchers suggest. Their data also show that tumors overexpressing the epidermal growth factor receptor gene conferred a comparably poor outcome, though this effect was independent of the HOX association.

"These findings provide the first clinical evidence to our knowledge for the relevance of a stem-like cell phenotype in treatment resistance of glioblastoma," the researchers write. The signature, they note, "is evocative of a self-renewal signature recently described for leukemia."

Yeast-Based Vaccine Triggers Immune Response in Mouse Models of Cancers

Researchers in NCI's Center for Cancer Research have tested a new approach for developing cancer immunotherapy; genetically engineering a harmless strain of yeast to produce a tumor protein called carcinoembryonic antigen (CEA) that triggers an immune response against certain types of cancer but not against healthy tissue. Their study results appeared July 1 in Clinical Cancer Research.

The scientists, led by Dr. Jeffrey Schlom, genetically engineered Saccharomyces cerevisiae, a strain of yeast that causes no disease in humans and has been used safely in other clinical trials, to express CEA, a surface protein that is normally expressed in the blood during fetal development but rarely appears again in healthy individuals. In patients with some carcinomas, however, CEA is overexpressed. Using mouse models for colon and pancreatic cancer types that were also engineered to express CEA, the researchers injected either the CEA-yeast vaccine or a control vaccine, a non-genetically engineered form of the yeast that was heat-killed prior to injection.

In mice that received the CEA-yeast vaccine, CD4+ and CD8+ T-cells responded to the stimulus and began to attack the tumors. Giving the vaccine as a series of injections (up to four times) increased the T-cell response. Higher vaccine doses and multiple injection sites also increased the response, which resulted in tumor shrinkage and increased overall survival. These results were also seen in a CEA-mouse-model in which tumors had metastasized to the lung. In all cases, the mice that received the CEA-yeast vaccine showed no toxicity or autoimmune disease.

S. cerevisiae is easy to genetically engineer and culture, it triggers a strong immune response, and it is safe. "All of these qualities make [it] an attractive vehicle for cancer immunotherapy," the authors conclude.