The American Society of Clinical Oncology held its annual meeting June 1 - 5 in Chicago. The following are abstracts presented at the meeting focused on the application of molecular diagnostics to cancer prognosis.
A prospective clinical utility study of the impact of the 21-gene recurrence score assay (Oncotype DX) in estrogen receptor positive (ER+) node negative (pN0) breast cancer in academic Canadian centers
Researchers from Genomic Health and the British Columbia Cancer Agency, led by James Ashley Davidson, presented data showing that use of Genomic Health's Oncotype DX 21-gene recurrence score test can impact physicians' decision-making in treating patients with stage I-II, node-negative, ER+ breast cancer.
The study measured the impact of the Oncotype DX RS on physicians' treatment recommendations for breast cancer patients in British Columbia's publicly funded health care system with stage I-II, node-negative, ER+ disease who were to be given a standard treatment with tamoxifen for 5 years.
The study found the Oncotype DX RS results "significantly affected adjuvant treatment recommendations in ER+ node-negative breast cancer, the researchers reported in their abstract.
Tracking 150 patients from two participating BCCA centers between May 2010 and July 2011, the group found that physicians changed their chemotherapy recommendation — either adding or omitting adjuvant chemotherapy — after obtaining recurrence scores in 45 cases, or about 30 percent of patients.
The researchers reported that there was also an "overall significant improvement in physician confidence" and a decrease in "patient decisional conflict" after obtaining an Oncotype DX recurrence score, suggesting the test positively affected both patients' and physicians' confidence in their course of treatment.
The group reported that it will present results of ongoing health economic analysis from the study separately at a later date.
Evaluation of variables that may impact chemotherapy (CT) administration after determination of Oncotype DX (ODX) recurrence score (RS)
Researchers from Genomic Health and Cardinal Health Specialty Solutions, led by Kristina Bowen of the oncologic specialty practice Georgia Cancer Specialists, presented study results that suggest that lower-than-expected use of chemotherapy among early-stage breast cancer patients with high Oncotype DX recurrence scores appears to be related to physician preference and patient age, and not tumor size or grade.
According to the group's abstract, the study was prompted by a finding of lower-than-expected use of chemotherapy in Oncotype DX "high-risk" patients in a Cardinal Health Specialty Solutions and CareFirst BlueCross Blue Shield "clinical care pathway."
To examine the variables that most impacted the physicians' decisions to prescribe or not prescribe chemotherapy to such patients, the researchers conducted a study of patterns of care using GCS patients' electronic medical records from 2009 to 2011.
The group retrospectively identified those patients with early-stage, node negative, ER+ breast cancer who also underwent Oncotype DX testing and grouped patients by their Oncotype recurrence score, tracking the number who subsequently received chemotherapy in each category. The team analyzed use of chemo in high-risk patients by "age, tumor size, tumor grade, and physician prescribing patterns."
Of 1,908 patients with early-stage cancer, 288 underwent testing, with fifty percent scoring in the low-risk group, 34 percent in the intermediate group, and 16 percent in the high-risk group, the researchers reported. In a review of patients' electronic records, the team found that six percent of the low-, 41 percent of intermediate-, and 83 percent of high-risk patients received chemotherapy.
100 percent of patients 45 and younger ended up receiving chemo, while only 86 percent of those between 46 and 65, and 25 percent of those over 66 were treated with chemotherapy. Patients with tumors smaller than two centimeters used chemo 82 percent of the time, and those with tumors between two and five centimeters had chemo 86 percent of the time. Grade 1, 2, and 3 tumor patients had chemo 100, 60, and 91 percent of the time, respectively. There were no significant differences in patterns of chemotherapy use among the physicians evaluated.
Overall, the researchers concluded, the lower-than-expected use of chemotherapy in patients with high Oncotype DX recurrence scores appeared to be related to physician preference and patient age, but not to either tumor size or grade.
Comparison of molecular (BluePrint+MammaPrint) and pathological subtypes for breast cancer among the first 800 patients from the EORTC 10041/BIG 3-04 (MINDACT) trial
Researchers from Agendia and a number of European cancer institutes led by Giuseppe Viale of the European Institute of Oncology in Milan, Italy, presented data on the first 800 patients from Agendia's MINDACT trial. The initial results showed some discrepancies between breast cancer subtypes defined by the company's MammaPrint test and those measured by pathological surrogates based on Ki67 and other measurements.
The MINDACT trial is a prospective randomized study comparing the 70-gene MammaPrint signature with the common clinical-pathological criteria in selecting patients for adjuvant chemotherapy in breast cancer with up to three positive nodes. In the trial, researchers are randomizing discordant cases to receive either chemotherapy or hormonal therapy based on clinical-pathological risk assessment or MammaPrint testing.
The researchers looked at the first 800 patients in the trial, comparing breast cancer subtypes measured with Agendia's BluePrint and MammaPrint tests against pathological surrogates based on ER, PR, HER2, and Ki67 measurement, according to the group's abstract.
In 621 tumors analyzed, the researchers found that all but one of the pathological cases deemed basal-type were also basal according to BluePrint. Among all the BluePrint-assigned basal cases, 15 were not pathologically defined as basal. One was luminal A, 11 were luminal B and three were HER2.
Overall, the researchers found that most discordant cases among these first 800 patients were seen within the luminal subtype, indicating that Ki67 discriminates luminal A and B differently than MammaPrint does.
This suggests "potential important impact for treatment decision-making," which the researchers hope the continuing MINDACT trial will help to better describe.
Validation of a genomic classifier (ColoPrint) for predicting outcome in the T3-MSS subgroup of stage II colon cancer patients
Researchers from Agendia and a group of European institutes led by Ramon Salazar of the Institut Català D´Oncologia in Barcelona, Spain, shared data from a study validating Agendia's 18-gene ColoPrint signature for identifying early-stage colon cancer patients at higher risk of relapse. The group concluded in its abstract that combining the test with clinical and pathological measures "significantly improves prognostic accuracy" for stage II patients in general and for the T3 and MSS phenotype subgroups.
"ColoPrint combined with established clinicopathological factors and MSI, significantly improves prognostic accuracy, thereby facilitating the identification of patients at higher risk who might be considered for additional treatment," the authors wrote.
According to the group's abstract, "Adjuvant therapy is not indicated for [stage II colorectal cancer] patients with MSI-H status who are considered of being at low risk of disease relapse. However, this leaves the majority of patients with an undetermined risk."
The group conducted univariate and multivariate analyses on 320 stage II patients and a 227-patient subset with the T3/MSS phenotype.
According to the researchers, ColoPrint classified two-thirds of the stage II patients as being at lower risk. The data showed that three-year relapse-free-survival was 91 percent for low-risk patients and 74 percent for patients at higher risk.
Clinicopathological parameters recommended under American Society of Clinical Oncology and National Comprehensive Cancer Network guidelines, meanwhile, "did not predict a differential outcome for high-risk patients," according to the group.
In the subgroup of patients with a T3/MSS phenotype, ColoPrint classified 61 percent as having lower risk with a three-year RFS of 91 percent, and 39 percent of patients as having higher risk with a three-year RFS of 73 percent. No clinical parameter was significantly prognostic in this subgroup, the researchers reported.
The PARSC trial, a prospective study for the assessment of recurrence risk in stage II colon cancer patients using ColoPrint
At the meeting, researchers from Agendia and multiple institutions highlighted the launch of a blinded, prospective trial to validate ColoPrint's ability to assess whether stage II colon cancer patients will relapse within a three-year time frame.
In the Prospective study for the Assessment of Recurrence risk in Stage II colon cancer patients, or PARSC, researchers led by Ramon Salazar of the Institut Catala d'Oncologia in Barcelona, Spain, plain to compare the risk assessment by ColoPrint against investigators' determination of patients' disease risk using ASCO criteria. Researchers will also look at whether patients' treatments confound ColoPrint's ability to gauge relapse risk and whether the test can assess relapse risk for Stage III colon cancer patients.
PARSC, which was launched in September 2008, is enrolling Stage II and III colon cancer patients with adenocarcinomas, no prior neo-adjuvant therapy, and no synchronous tumors. Following informed consent, participants will submit a fresh tumor sample. Physicians are free to treat the patients as they see fit with National Comprehensive Cancer Network-approved regimens or an acceptable alternative.
So far there are 340 Stage II and 280 Stage III patients enrolled in the trial. Researchers are aiming to enroll 575 stage II patients.
Use of the PRO Onc Assay (Prometheus) to measure HER2 overexpression/activation in circulating tumor cells (CTCs) in women with HER2-negative metastatic breast cancer (MBC): A Sarah Cannon Research Institute (SCRI) trial
Researchers led by John Hainsworth of the Sarah Cannon Research Institute in Tennessee used a multiplexed immunoassay to measure the expression and activation of HER2 proteins and other signal transduction molecules in circulating tumor cells in 57 tumor samples from women with HER2-negative breast cancer. Using the PRO Onc assay, the investigators are hoping to improve the accuracy of determining HER2 status in breast cancer patients, which is critical in determining the right treatment strategy.
According to an abstract presented at the meeting, the PRO Onc assay "can detect overexpression and activation (phosphorylation) of HER2, with sensitivity at a single CTC level." Tumor samples were taken from HER2-negative women who were being treated with standard chemotherapy and had had their HER2 status assessed by fluorescence in situ hybridization (32 women) or by immunohistochemistry (25 women). It has been estimated that discordance may be as high as 20 percent between IHC and FISH-based HER2 results at low-volume labs, though conflicting test results may be less prevalent at central labs handling larger sample volumes.
The PRO Onc Assay was performed in 31 out of 54 samples found to have CTCs. Of these, PRO Onc found HER2 overexpression in six samples, while four samples had HER2 activation without overexpression.
However, seven out of the 10 patients identified as having HER2 activation or overexpression by PRO Onc had been deemed HER2 negative by FISH, and three samples were HER2-negative by IHC. Based on the results of the IHC and FISH tests, these 10 patients had received a median of three chemotherapy regimens.
"When CTCs were present, the PRO Onc Assay identified HER2 overexpression or activation in 31 percent of women with HER2- negative MBC. The therapeutic significance of this finding is unknown," the researchers concluded in the abstract. "Since the predetermined threshold of [greater than or equal to] 10 percent HER2- positive assay results was exceeded, this trial will proceed to a planned second portion, in which women with HER2- negative metastatic breast cancer (by FISH testing) and HER2 overexpression or activation in CTCs will receive a trial of HER2- targeted therapy."
San Diego-based Prometheus Labs developed the PRO Onc Assay. Hainsworth told PGx Reporter that the test needs to be further evaluated "to prove that the findings have clinical relevance." Such studies are currently ongoing and will be completed in the next 12 to 24 months.
Use of a proliferation-based mRNA signature to predict outcome in early-stage non-small cell lung adenocarcinoma
Researchers from Myriad Genetics and several other institutions investigated the prognostic ability of a 46-gene mRNA expression profile in two sample sets: a public microarray data set of surgically resected lung adenocarcinomas and formalin-fixed paraffin-embedded samples collected from lung cancer patients at the MD Anderson Cancer Center and the European Institute for Oncology.
"Adjuvant treatment of patients with early-stage lung adenocarcinoma is based on post-surgical pathological staging and patient performance status," the researchers, led by Carmen Behrens of the MD Anderson Cancer Center, wrote in the abstract presented at the meeting. "Disparate outcomes within each staging group suggest that additional prognostic markers could improve our understanding of risk-benefit and potentially lead to better treatment decisions."
Myriad has said it is in the process of clinically validating a lung cancer test to guide therapeutic decisions for patients diagnosed with stage I or stage II adenocarcinomas (PGx Reporter 5/2/2012).
The public expression data were derived using Affymetrix HG-U133A arrays, while quantitative PCR was used to analyze the clinical FFPE samples. Researchers calculated a cell cycle progression score using 31 cell cycle genes and 15 housekeeper genes, and then gauged the score's ability to assess outcomes of Stage I and Stage II lung cancer patients.
In 256 cases from the public expression data cohort, "the CCP score was a significant predictor of death in univariate and multivariate analysis using age, stage, gender, smoking status and treatment as covariates," the investigators reported in the abstract. In a second data set of 204 samples, the "CCP score was [also] highly associated with death."
When the same signature was applied to the 381 FFPE samples from patients with lung adenocarcinomas who were followed for a median of five years, the CCP score remained "the most significant predictor of death" in the presence of clinical covariates such as tumor size and pleural invasion.
"A 46 gene mRNA signature is a significant predictor of disease-related death in early-stage lung adenocarcinoma, providing independent prognostic value in the presence of clinical variables," the researchers concluded, noting that this same signature will be further evaluated to gauge its impact on clinical decision making.
Behrens disclosed having received research funding from Myriad. Several other investigators, who are employees of Myriad, disclosed owning stock in the company.
Predicting early relapse in patients with BRAF V600E melanoma with a highly sensitive blood BRAF assay
Researchers from the Massachusetts General Hospital Cancer Center, Beth Israel Deaconess Medical Center, and Dana-Farber Cancer Institute have developed an assay that can detect as little as one BRAF mutant melanoma cell among 400,000 peripheral blood lymphocytes. They are hoping that this molecular diagnostic will be able to identify which patients are likely to experience progressive disease despite BRAF inhibitor treatment earlier than imaging allows.
"Many patients with metastatic melanoma who progress on BRAF inhibitors develop rapidly progressive disease, which is difficult to manage," the researchers, led by Ryan Sullivan from the Massachusetts General Hospital Cancer Center, said in the abstract presented at the meeting. "Identification of an early marker of progressive disease may allow for therapeutic intervention prior to clinical deterioration."
Sullivan and colleagues aimed to investigate the utility of their assay in following patients with BRAF-mutated melanoma who are receiving either single-agent vemurafenib (marketed by Roche's Genentech as Zelboraf) or the combination of dabrafenib and trametinib, two investigational drugs under development at GlaxoSmithKline. Patients enrolled in the trial had their BRAF V600E mutation status established by a lab test before starting BRAF inhibitor therapy. BRAF mutation status was also assessed by the researchers' proprietary assay at different time points post-treatment and correlated with their treatment response and disease progression.
So far the researchers have initial response data in 16 patients,10 treated with Zelbraf and 6 with dabrafenib/trametinib, and 15 patients have experienced some form of disease progression.
"Disease progression, as defined by an increase in blood BRAF V600E level, occurred in advance (by at least 43 days) of radiographic progression in 11 [out of] 12 cases," Sullivan and colleagues reported in a poster presented at ASCO. Furthermore, the data show that a reduction in BRAF V600E blood level is in line with tumor regression as seen on imaging. Furthermore, "serial blood BRAF V600E levels can be readily followed and predict clinical outcome," they reported.
"Our assay is able to quantify changes in BRAF V600E levels in the blood of patients with BRAF mutant metastatic melanoma receiving BRAF inhibitor therapy," the study authors concluded. In these patients, "the BRAF V600E levels are reduced in the setting of initial disease regression and increase well in advance of clinical or radiographic progressive disease."
The researchers described their lab-developed test as "highly sensitive" and "inexpensive." Noting that further studies are necessary to develop the test for commercial use, the study investigators noted that the assay may be used as a tool to detect disease progression earlier than imaging allows. This "may improve the effectiveness of next line therapy (given the lower tumor burden) and mechanism of resistance analysis for patients with BRAF mutant disease progressing on BRAF-directed therapy," researchers said in the abstract.
The study investigators did not disclose any relationships with industry.
Last year, the US Food and Drug Administration approved Roche/Plexxikon's BRAF inhibitor Zelboraf as a treatment for BRAF-mutated metastatic melanoma. The agency simultaneously approved a BRAF mutation test developed by Roche as a tool to gauge best responders to the drug. The FDA has noted in a draft guidance on companion diagnostics that tests that select the intent-to-treat population of a drug require approval for the agency.
GlaxoSmithKline is studying the BRAF inhibitor dabrafenib in combination with the MEK inhibitor trametinib in a Phase III study (PGx Reporter 6/62012).
A molecular diagnostic panel for thyroid cancer disease management
Quest Diagnostics has launched a new molecular test panel to help physicians determine if a thyroid gland is cancerous and should be surgically removed. The test gauges four gene markers – BRAF, RAS, RET/PTC, and PAX8-PPARγ – associated with papillary and follicular thyroid cancer.
In 2009, the American Thyroid Association recommended that thyroid cancer patients who receive indeterminate results from fine-needle aspiration biopsy be molecularly tested for these markers to aid physicians' decision making.
In a study presented at the meeting, researchers from Quest evaluated 149 thyroid FNA samples and tested them for BRAF V600E and K601E mutations by allele-specific PCR; for RAS H, K, and N mutations by pyrosequencing; and for RET/PTC1 and RET/PTC3 rearrangements, as well as PAX8-PPARγ translocations using RT-PCR.
Study investigators found that 90 of 149 FNA specimens, or about 60 percent, harbored mutations of one or more of the four markers tested by the new panel. "The presence of the four markers was generally mutually exclusive, suggesting potential value in a hierarchical screening strategy for samples with limited tissue," Quest said in a statement.
Study investigators identified BRAF V600E mutations in around 54 percent of samples, making them the most prevalent mutations in papillary thyroid cancer. The researchers did not identify any K601E mutations in the samples.
Several papillary, follicular adenoma, and follicular thyroid cancer samples contained RAS mutations, half of which were NRAS Q61 mutations. Investigators reported RET/PTC rearrangements in 3.5 percent of papillary thyroid cancer samples and PAX8/PPARγ translocations in nearly 19 percent of follicular thyroid cancer samples, but not in follicular adenoma samples.
Of the seven indeterminate thyroid nodules, two had BRAF mutations and two had RAS mutations, the study investigators found. The presence of the four markers did not usually overlap, researchers noted, highlighting that one papillary thyroid cancer sample had concurrent RAS and RET/PTC1 alterations.
"The prevalence of BRAF, RAS, RET/PTC, and PAX8/PPARγ alterations in thyroid cancer specimens highlights the potential for targeted therapeutic strategies," the researchers concluded. "The mutually exclusive pattern of alterations also suggests a hierarchical screening strategy for samples with limited availability."
Of the 300,000 thyroid FNA biopsies performed each year, Quest estimates that between 15 percent and 20 percent of these biopsies produce indeterminate results. An unclear result may increase the chance that a person's thyroid gland, suspected of being cancerous, is unnecessarily removed.
"The removal of a healthy thyroid gland is an unfortunate outcome for many patients, due to limitations of current test methods," Richard Reitz, medical director of the endocrine division at the Quest Diagnostics Nichols Institute, said in a statement announcing the new test's launch. The company hopes its test will "enhance the reliability of thyroid biopsy testing," as well as "prevent the unnecessary removal of healthy thyroids for many patients."