NOVEL BIOMARKERS FOR CANCER CHEMOTHERAPY

November 23, 2009 — Advances in genomics, biotechnology, and molecular pathology have generated many candidate biomarkers that might help predict response to cancer therapy. The hope is that biomarkers will enhance the effectiveness and safety of cancer care by allowing physicians to tailor treatment to individual patients.

"This is an exciting time, in that we see that the past 30 years of basic research into understanding what actually causes cancer is beginning to pay off," said Lewis Cantley, PhD, director of the Cancer Center at Beth Israel Deaconess Medical Center in Boston, Massachusetts. "That knowledge is leading to therapies that are actually working."

Dr. Cantley, who moderated a press briefing during the AACR-NCI-EORTC 2009 International Conference on Molecular Targets and Cancer Therapeutics, noted that there is a lot of excitement in this area, and that "personalized cancer treatment and precision medicine are 'buzzwords'."

During the briefing at the meeting, researchers presented new data on markers of prognosis associated with metastatic and HER2-positive breast cancers, renal cell carcinoma, and other advanced solid tumors.

PK13K: A Novel Therapeutic Approach?

Phosphatidylinositol 3-kinase (PI3K) is thought to play an important role in tumorigenesis, and the PI3K signaling pathway is frequently dysregulated in a large spectrum of human tumors. Early data suggest that targeting PIK3CA, a mutation of the p110 subunit of PI3K, holds promise as a novel therapeutic approach.

In a small phase 1 trial, study author Filip Janku, MD, PhD, a clinical research fellow at the University of Texas M.D. Anderson Cancer Center in Houston, explained that 4 of 10 patients (40%) treated with a protocol that included a drug targeting the PI3K-AKT-mTOR pathway experienced a partial response.

When broken down by individual disease, there were 2 responses in 3 endometrial cancers (67%), 1 in 4 ovarian cancers (25%), and 1 in 1 breast cancer (100%) .

Dr. Janku and colleagues analyzed 117 tumor samples from patients with advanced cancers of various types and observed that 12% had PIK3CA mutations. The mutations appeared to be most frequent in endometrial, ovarian, head and neck, breast, and colon cancers; no mutations were found in patients with melanoma or cervical cancer.

"The numbers are very small," said Dr. Janku, pointing out that although the results were intriguing, they needed to be interpreted with caution and confirmed in a larger population of patients.

Targeting Metastatic and HER2 Breast Cancers

The inherited variation of the gene cytochrome P450 2D6 (CYP2D6) and the use of CYP2D6-inhibiting medications can negatively affect overall survival in patients with metastatic breast cancer, said study author Laureen Lammers, PharmD, from Erasmus MC, in Rotterdam, the Netherlands.

"The CYP2D6 phenotype was an important predictor of treatment outcome in women with hormone-receptor-positive metastatic breast cancer who were treated with tamoxifen," said Dr. Lammers, PharmD. The use of drugs that inhibit CYP2D6 also worsened the outcome of tamoxifen therapy, and "should therefore be discouraged," she explained. One class of drugs that inhibit CYP2D6 are the selective serotonin reuptake inhibitors, and advice to avoid SSRIs in patients receiving tamoxifen was highlighted earlier this year. "We recommend that doctors avoid the combination of CYP2D6 inhibitors together with tamoxifen, not only in the adjuvant setting but also for patients who are treated for metastatic breast cancer," said Dr. Lammers.

The metabolism of tamoxifen is complex, but it is known that CYP2D6 is necessary to metabolize it into the active metabolite endoxifen. In their study, Dr. Lammers and colleagues reviewed patient charts for 99 women who received tamoxifen treatment for hormone-receptor-positive metastatic breast cancer, and analyzed the data to assess patient and tumor characteristics, time to progression, the use of CYP2D6 inhibitors, and overall survival.

They found that patients with a poor CYP2D6 metabolizer phenotype had worse overall survival than those who were extensive metabolizers (hazard ratio [HR], 2.11; P = .031). The time to progression was also shorter in poor metabolizers than in extensive/intermediate metabolizers (1.7 vs 2.9 years; P = .089). Overall survival was also shorter (5.4 vs 9.9 years).

These findings were also true for patients taking tamoxifen who were also taking drugs that inhibit CYP2D6, the researchers noted.

A second presentation, also focused on breast cancer, suggested that a protein known as Trim62 could be a potential biomarker to predict the biologic response of breast cancer cells to anti-HER2 therapeutics such as lapatinib (Tykerb).

Small amounts of the cell-cycle inhibitory protein p27 are predictive of a poor outcome in breast cancers. Trim62 appears to be responsible for the misregulation of the regulatory protein p27 in HER2-positive breast cancer tumors. Understanding the mechanism of p27 mislocalization in breast cancers might offer insight into pathways of oncogenic transformation, reported researchers from the Fred Hutchinson Cancer Research Center in Seattle, Washington. It might also provide new markers for predicting clinical responses to specific therapies.

In tumor cells, levels of p27 are typically lower within the cell's nucleus while, at the same time, they are retained or increased in the cytoplasm, explained lead author Claire M. Faltermeier, a research assistant in the lab of Jim Roberts, MD, PhD, at the Hutchinson Center's Basic Sciences Division. In the nucleus, p27 functions as a tumor suppressor by inhibiting cell proliferation, but new evidence suggests that cytoplasmic p27 has an oncogenic action and might interact with the GTPase RhoA to promote cellular migration and motility.

"Just like Dr. Jekyll and Mr. Hyde, p27 has a good side and a bad side," she said at the briefing.

Cytoplasmic p27 could be a marker of tumor cell invasiveness and metastatic potential and, to further investigate the mechanism of p27 mislocalization in breast cancers, Ms. Faltermeier and colleagues developed antibodies that reliably detected cytoplasmic p27. Using the antibodies, they also identified a subset of breast cancers positive for HER2 that exhibited increased amounts of cytoplasmic p27.

In parallel experiments, they discovered that Trim62, an E3 ubiquitin ligase, regulates the stability of p27, and that is it overexpressed in HER2-positive breast cancers.

The researchers investigated the role of Trim62 in modulating the response of breast cancer cells to anti-HER2 therapeutics. Lapatinib, which is an HER2/epidermal growth-factor receptor inhibitor, increased levels of p27; this increase in p27 was observed exclusively in the nucleus. Reduction of Trim62 increased cellular sensitivity to lapatinib, and when Trim62 levels were reduced in HER2-positive cancer cells, p27 that was in the cytoplasm moved to the nucleus and the cells stopped proliferating. Conversely, overexpression of Trim62 had the opposite effect.

"Collectively, our results suggest that Trim62 underlies the misregulation of p27 in HER2-positive breast cancer cell lines," the authors write.

Plasma Biomarkers Predict Outcomes in Renal Cancer

Although the findings are preliminary, researchers have identified several plasma biomarkers that appear to be predictive of overall survival in advanced renal cell carcinoma (RCC).

A component of the phase 3 Treatment Approaches in Renal Cancer Global Evaluation Trial (TARGET) was to identify plasma biomarkers for possible prognostic value. TARGET demonstrated sorafenib (Nexavar) to be effective and safe for patients with advanced RCC in whom first-line therapy had failed. The study consisted of 903 patients with advanced clear cell RCC who were randomized to receive either sorafenib or placebo.

Carol Peña, PhD, associate director for clinical cancer biomarkers at Bayer HealthCare Pharmaceuticals, and colleagues conducted an analysis on a subset of the patients enrolled in TARGET to evaluate the relation between biomarker levels and outcomes.

"We looked at biomarkers for prognosis of RCC in the absence of treatment and also looked at biomarkers that predict response to sorafenib," said Dr. Peña.

Plasma levels of vascular endothelial-growth factor (VEGF), soluble VEGF receptor (VEGFR)-2, CAIX, TIMP-1, and p21 Ras were measured at baseline and after 3 and 12 weeks of treatment. Using the median baseline level of each biomarker to define high and low levels, univariate analysis of patients from the placebo group showed that elevated VEGF (HR, 1.65; P = .002), CAIX (HR, 2.26; P = .034), TIMP-1 (HR, 3.34; P = .001), and p21 Ras (HR, 2.49; P = .016) correlated with poor prognosis.

However, multivariate analysis of CAIX, TIMP-1, Ras, and VEGF together with clinical variables in patients in the placebo group showed that TIMP-1 was independently prognostic for survival. In this exploratory analysis, explained Dr. Peña, there was also no correlation between baseline levels of plasma VEGFR-2, CAIX, TIMP-1, or Ras and the efficacy of sorafenib.

The sample was small and the study was performed retrospectively, so further trials are needed, she said. "But it suggests that TIMP-1 may have value as a prognostic marker for renal cell carcinoma."

AACR-NCI-EORTC 2009 International Conference on Molecular Targets and Cancer Therapeutics: Abstracts B134, C118, A36, A64. Presented November 9, 2009.