Today we are speaking with Donna Niedzwiecki, PhD, of the Alliance Statistics and Data Center at Duke University Medical Center in Durham, North Carolina, about a new gene expression assay that may be helpful to parse patients with stage II colorectal cancer into low- and high-risk recurrence groups. The biomarker test, called GeneFx Colon (or the ColDx assay), could be helpful in identifying patients who need more or less adjuvant therapy. The team’s results on this expression assay were published last month in the Journal of Clinical Oncology.
— Interviewed by Anna Azvolinsky
Cancer Network: What exactly does the GeneFx Colon test do, and what does it measure?
Dr. Niedzwiecki: The GeneFx Colon is a microarray-based prognostic assay that uses formalin-fixed paraffin-embedded samples. The genes identified by the GeneFx Colon signature are those involved in a broad range of processes related to tumor biology. This genetic signature was developed to quantify patients’ risk of colon cancer following surgery with curative intent. About 25% of colon cancer patients present with stage II disease. The 5-year survival rate in this population is 75% to 80%, and about 15% to 20% of these patients will eventually experience disease recurrence.
Currently, there is no optimal strategy for clinical management of patients with stage II colon cancer. After surgical resection, adjuvant chemotherapy may be recommended for patients with high-risk features—for example if they have T stage of 4, a high tumor grade, inadequate number of node samples, and other features.
Molecular markers can potentially provide additional information on patient prognosis. For example, assessment of tumor microsatellite instability (MSI), sometimes called mismatched repair (MMR), is used in clinical practice to classify patients as having a higher or lower risk of recurrence. Other markers such as loss of heterozygosity and p53 status are being investigated as prognostic markers. In addition, several gene signatures such as GeneFx Colon have been shown to be independently associated with clinical outcome.
It is of interest to identify patients with a higher risk of recurrence since these patients could potentially comprise a subgroup that may benefit from adjuvant therapy. Whether patients identified as having a higher risk of recurrence do benefit from adjuvant therapy is a separate question and was not addressed in our paper.
In terms of assay development, the gene expression signature was developed in two stages. First a colorectal cancer disease-specific array, including genes specifically expressed in human colorectal cancers, was created. The content, 61,528 total probe sets, was generated by a combination of high-throughput sequencing, public database mining, and experimental investigation. A set of 215 human colorectal cancers with known outcome data were then identified as a training set and used to develop a 634 probe set specific for high risk of recurrence. The related probe sets, statistical methods, and pathway and analytical properties of the assay were described in prior publications and are referenced in the paper.
Cancer Network: Can you talk about the validation study you and your colleagues published. What were the main results?
Dr. Niedzwiecki: So the GeneFx Colon assay was previously validated in an independent dataset comprising 144 patient samples that had associated clinical data. Our current article presents the results of the second independent validation that uses tumor specimens and data from a large phase III clinical trial in stage II colon cancer patients without high risk features. The trial, Alliance CALGB 9581, was conducted by the Cancer and Leukemia Group B, which is now the Alliance for Clinical Trials in Oncology, or Alliance.
CALGB 9581 was a randomized trial of 1,738 patients, randomized to observation or adjuvant treatment with edrecolomab, which is a monoclonal antibody, after surgical resection of their primary tumor. The primary endpoint of the trial was overall survival. Tissue samples from patients enrolled on the trial were collected and archived for biomarker analysis and served as a resource for the validation study. Of 1,454 patients meeting the eligibility requirements for the validation study, formalin-fixed tissue was available for 901 patients. A case-cohort trial design was used, and the primary endpoint was a recurrence-free interval defined as distant recurrence of primary disease or death due to primary disease.
The final sample size for the study was 393 patients. Based on a prespecified cutoff point for the GeneFx Colon assay, 55% of patients were categorize as having a high risk of recurrence at 5 years since study entry, and 45% were categorized as low risk. The GeneFx Colon signature was significantly associated with sex, race, and MMR status. Higher proportions of men (57% vs 49%), non-white race (73% vs 52%), and MMR-intact tumors (56% vs 45%) were classified as high risk. The primary analysis tested the association of the categorized GeneFx Colon signature and recurrence-free interval when selected known prognostic variables were considered. These prognostic variables included age, sex, race, assigned treatment arm, T stage, number of nodes examined, perineural invasion, lymphovascular invasion, tumor grade, obstruction of perforation, tumor location, and MMR.
In our results, the signature remained statistically significant after adjustment for these conventional prognostic factors. The adjusted recurrence-free hazard ratio estimate was 2.13 for high vs low risk (95% CI, 1.3–3.5). Thus, patients classified as high risk were at a two-times greater risk of recurrence than the patients classified as low risk. At 5 years, the recurrence-free interval for high-risk patients was 82% compared with a 91% risk at 5 years for patients classified as low risk.
Although the assay was not developed as a prognostic for overall survival, there is a marginally significant association between the GeneFx Colon signature and overall survival with a hazard ratio of 1.4. As mentioned, the study is the second independent validation of the GeneFx Colon assay in patients with stage II colon cancer. The first study was published in the Journal of Clinical Oncology in 2011 by Kennedy et al. The current study further substantiates the assay as an effective prognostic biomarker for stage II colon cancer.
Cancer Network: Just briefly, you mentioned prior studies using this gene signature. So, is this test commercially available for clinicians to use and is there other validation testing that is needed or ongoing?
Dr. Niedzwiecki: Actually, no further validation studies are required for this test. So far, including our study, there are two robust clinical validations of the assay and that is typically sufficient for clinical use of the assay and for reimbursement to be provided. We are conducting additional analyses to incorporate the GeneFx Colon signature with conventional tumor characteristics and other biomarkers into a more refined predictive model to assess patients’ risk of recurrence. Also, studies to determine whether high-risk patients receiving adjuvant therapy would benefit from treatment would further establish its usefulness.
The company that discovered and developed the assay is Almac Diagnostics in Northern Ireland. The test is commercially available in the United States via a partnership that Almac has with Helomics Corporation located in Pittsburg, Pennsylvania. Helomics sells the assay under the name GeneFx Colon in the United States. Almac is currently finalizing its strategy to make the assay available in Europe and the rest of the world.
Cancer Network: Thank you so much, Dr. Niedzwiecki, for joining us today.
Dr. Niedzwiecki: You’re very welcome, thank you for asking.