Gene testing of cancer patients is now so complicated and full of uncertainties that even some Harvard oncologists are unprepared for its clinical application, according to a studypublished online March 24 in the Journal of Clinical Oncology.
Researchers at the Dana-Farber Cancer Institute in Boston conducted a survey of their colleagues, who are all on the faculty of Harvard Medical School, to assess their use of cancer gene testing and attitudes and knowledge about it.
The testing can help guide treatment recommendations and patient management. For example, testing patients with lung cancer for EGFR mutations can guide the use of targeted therapies erlotinib (Tarceva,Astellas, Roche) and gefitinib (Iressa, AstraZeneca).
Such tests “hold the potential to transform care,” write the study authors, led by medical oncologist Stacy Gray, MD.
She and her team found that a “sizeable minority” (22%) of the 160 responding clinicians at Dana-Farber reported having “low confidence” in their “genomic knowledge.”
Furthermore, the anticipated use of sophisticated “multiplex” cancer gene testing at Dana-Farber varied widely in the survey sample, which included medical, surgical, and radiation oncologists.
Multiplex testing allows for the simultaneous analysis of many genes and their variants.
Of the survey respondents, 18% anticipated using multiplex testing infrequently (in 10% of patients or less) and 25% anticipated using it in most patients (more than 90%).
The findings on the multiplex testing surprised an expert not involved with the study.
“Among the many intriguing findings of this study, wide variability in interest in [multiplex testing] was identified,” writes Michael J. Hall, MD, a medical oncologist from the Fox Chase Cancer Center in Philadelphia, in an accompanying editorial.
“Strikingly, less than 50% of physicians planned to view the [multiplex testing] results routinely,” he adds.
Dr. Hall was also surprised that the majority of survey participants (58%) planned to only “rarely” or “sometimes” use the “clinically relevant” results.
Clinically relevant results are where genomic variants are found in a patient that have been previously validated and are considered actionable or are approved by the US Food and Drug Administration, such as EGFR mutations and lung cancer.
Dr. Hall believes that these “seemingly paradoxical responses” by the Boston physicians can, in part, be explained by the “timing of the study” and the “composition of the population surveyed.”
The survey was administered from September 2010 to January 2012, when testing was less common. Currently, all patients at Dana-Farber routinely undergo testing with OncoMap, a multiplex somatic genomic test for 471 potential alterations in 41 cancer-related genes.
Also, Dr. Hall emphasizes that 48% of the sample was composed of surgical and radiation oncologists, who are known to use gene testing less frequently. In fact, about half of these types of oncologists reported never having ordered a genomic test at baseline.
Genomic data are highly uncertain in nature.
Dr. Gray and colleagues note themselves that it “may not be surprising” that there is wide range of confidence levels in the use of genomics.
“Genomic data are highly uncertain in nature,” they explain.
Areas of uncertainty include, according to the researchers, the role that specific variants play in disease pathogenesis, the magnitude of effect associated with alterations, and the clinical utility of genomic information.
In an email to Medscape Medical News, Dr. Gray provided a concrete example of the trickiness of genomic information.
A patient with lung cancer might have a “mutation in an unusual position in the EGFR gene,” she said. If the mutation had been in the typical location, treatment with an EGFR inhibitor would be indicated. However, “we may not understand if, or how,” this unusual mutation can guide treatment, she explained.
In his editorial, Dr. Hall suggests that some of the confusion about genomics testing is rooted in the abundance of tests and data.
There are “growing numbers” of multiplex somatic genomic tests that examine “tens to hundreds” of genes, but only a “minority” of those “genetic loci,” such as EGFR and BRAF, harbor clinically relevant variants, he writes.
It is a “daunting task” for the physician to keep up, says Dr. Hall.
Both Dr. Hall and the researchers believe that professional education is a key to improving physician competence with genomics testing.
Medical Oncologists More Likely to Be Confident
Of the 276 physicians who were contacted, 160 completed the survey, for a response rate of 61%. The response was “robust,” says Dr. Hall, because the survey was nonincentivized and Web-based.
Participants reported that, before the introduction of OncoMap, they ordered tumor genomic testing for an average of 24% of patients. The most common tests ordered were for KRAS(25%), EGFR (24%), BRAF (24%), C-KIT (13%), BRC-ABL(13%), and JAK2 (9%).
A considerable number of participants were “not very confident” or “not confident at all” in their knowledge of genomics (22%), ability to explain genomic concepts to patients (14%), and ability to make treatment recommendations on the basis of genomic data (26%).
The average score on the genomic confidence scale was 3, which corresponds to a response of “somewhat confident,” the researchers report.
After potential confounders were controlled for, high genomic confidence was associated with being a medical oncologist, being a researcher, and high baseline test use.
Logistics Can Influence Testing Choices
Dr. Gray explained that multiplex tests — such as OncoMap or OncoPanel, another test that has recently come into use at Dana-Farber — have not yet fully removed the need to do, for example, an individual test for BRAF mutations in melanoma.
All of the “currently clinically useful [individual] tests” are being integrated into these larger multiplex tests, she said.
However, the turnaround time for multiplex testing can be 2 to 3 weeks. That can be a problem, she acknowledged.
“At this time, clinical tests such as BRAF and EGFR continue to be ordered if a quicker turnaround is needed,” she said.
The goal is to make multiplex testing the only test required. “In the future, we hope to use OncoPanel as a clinical test across all diseases for which there may be clinically relevant mutations,” she said.
The research was supported by the Dana-Farber Cancer Institute. Dr. Gray reports receiving support from the American Cancer Society and the National Human Genome Research Institute. The other authors and Dr. Hall have disclosed no relevant financial relationships.