This story has been updated to include additional information from a US survey on people's attitudes about cost of genetic testing. Originally published June 4.
CHICAGO – When presented with hypothetical scenarios under which they could receive pharmacogenetic testing to gauge whether they were likely to respond to chemotherapy or experience toxicities to treatment, the majority of cancer patients involved in a Canadian study wanted the diagnostic evaluations and were willing to make an out-of-pocket investment to learn the results.
At the American Society of Clinical Oncology annual meeting this week, researchers led by Henrique Hon of the Ontario Cancer Institute at Princess Margaret Hospital reported data from a study in which they surveyed 278 Canadian cancer patients about their willingness to receive and pay for PGx testing under theoretical scenarios. Of the interviewed patients, 153 were deemed by their physicians at the time of the survey to be potentially curable and were prescribed adjuvant chemotherapy, while 125 patients were diagnosed as having metastatic or incurable disease.
Among curative patients, 70 percent were willing to receive a hypothetical chemotherapy that had a five percent improvement in the cure rate and on which less than five percent of patients would experience severe toxicities due to treatment. Of these patients, 99 percent wanted to get a PGx test that could inform them of the likelihood that they would respond to chemotherapy and the same number were willing to pay a median $2,000 for the test and wait a median of 21 days for the results.
For those study participants in the metastatic arm, 90 percent were willing to accept a hypothetical palliative chemotherapy that could shrink tumors in 80 percent of patients and cause severe side effects in five percent. In this arm, 98 percent wanted PGx testing that could gauge which patients would be at risk of experiencing a severe treatment-related adverse event. For such a test, they were willing to pay a median $1,000 and wait a median turnaround time of 14 days.
"Among cancer patients willing to undergo chemo, almost all wanted pharmacogenomic testing] and were willing to pay for it, waiting several weeks for results," the researchers concluded in an abstract presented at the meeting.
In an idealized scenario – where the testing was free, the turnaround time was between one and two days, and the prevalence of the gene variation associated with lack of response to treatment was 50 percent – 98.6 percent of patients in the adjuvant arm accepted PGx testing. In the metastatic group, 97.5 percent accepted PGx testing in an idealized scenario where there was no test cost, a one- to two-day turnaround time, and the prevalence of the marker associated with treatment-related side effects was 75 percent.
"We found that patients' acceptance of testing declined with advanced age and among chemo-naive patients," Hon said during his presentation of the data at the meeting.
The researchers asked patients what they felt was a "reasonable" price for PGx testing, and curable patients cited a median cost of $200, while incurable patients said a median cost of $100.
"We felt that when push came to shove, [the patients] would more likely gravitate toward that reasonable value over the maximum they were willing to pay," Geoffrey Liu, study author and Alan B. Brown chair of molecular genomics at the Ontario Cancer Institute, told PGx Reporter. Liu cautioned that since this study was conducted in a Canadian patient population that receives care in a single-payor system, the results may not be applicable in a US healthcare setting.
Separately, researchers from the Fox Chase Cancer Center surveyed 406 people whose doctors thought they might have cancer-causing gene mutations based on their medical and family history. They reported this week that 82 people, or around 21 percent of study participants, said they would take a genetic test only if it was covered by insurance. Of the patients who were willing to pay out-of-pocket for such testing, the majority limited the amount to $500 or less.
Although study author Jennifer Matro of Fox Chase believes that some expensive tests for cancer-associated genes are worth the personal investment, she advised researchers against broadly administering testing to those who don't really need it. "We need to discover more risk factors for genetic mutations, so we can spare those patients who really don't need to pay for genetic testing," Matro said in a statement.
Findings from the US survey may not be directly comparable to the Canadian survey due to the differences in the countries' healthcare systems. In the Canadian system, services and interventions the government finds to be effective, safe, and meet cost thresholds are covered for residents. But this doesn't mean that all interventions are covered.
According to Liu, the study's finding that Canadian patients are willing to pay for PGx testing is significant since the government, after conducting stringent cost-effectiveness analyses, currently does not fund many PGx tests and there is uncertainty about which molecular diagnostics will be covered in the future.
"So, willingness to pay has real potential meaning, even in a universal healthcare system," Liu told PGx Reporter via e-mail. "The fact that [Canadian patients] have not likely had to [pay] out of pocket (unlike a substantial proportion of US patients) [means they] can serve as a naive population unaffected by prior experiences, and can serve as a good baseline assessment."
Furthermore, the researchers believe that willingness to pay is a way to quantify how strongly patients want access to PGx tests. "For instance, if people want a test but are not willing to pay for it, then exactly how strong a conviction or how strong the desire is it really of this patient?" Liu posited. "In contrast, if patients were willing to pay [between] 4 percent and 5 percent of their household income, then that would tend to mean that people are not superficially interested, and have a strong desire to adopt a new technology and a strong desire for the test."
Oncologists at the Ontario Cancer Institute currently offer patients PGx tests to gauge EGFR and ALK mutations and HER2 expression, as well as prognostic tests such as Agendia's breast cancer recurrence test MammaPrint.
An ASCO attendee from MD Anderson critiqued the study for focusing on patients' willingness to pay, however, noting that industry usually doesn't use such measures to set prices for healthcare products , considering them an "inaccurate way of determining an appropriate price."
Approximately 30 percent of surveyed patients were from households earning under $50,000 per year; another third came from households with annual income between $50,000 and $100,000; and the remaining third had a household income of above $100,000. The majority of study participants had received a college education.
"We found that patients were willing to pay more for PGx testing if they had higher household incomes [and] if they better understood the scenario defined in pharmacogenomic testing," Hon said. Additionally, patients with higher education and those who said they valued knowing PGx test results before moving ahead with treatment were more willing to wait longer for results.
In an effort to limit biasing the survey by interviewing patients who were already interested in personalized medicine, Liu and his colleagues randomly approached patients in clinic waiting rooms throughout the Ontario Cancer Institute in the summer of 2010. They had a 91 percent participation rate.
The metastatic group was older than the adjuvant group and had a higher proportion of men, a greater incidence of lung and head and neck cancer patients, and a smaller number of breast cancer patients. Patients with metastatic disease tended to be more likely to have had previous chemotherapy, previous clinical trial experience, and genetic testing. Around 10 percent of metastatic patients and 3 percent of adjuvant patients had received at least one genomic test.
Although certain scenarios in the survey, such as chemotherapy efficacy and toxicity, were hypothetical, participants' cancer prognoses status were real. "This was an ethics issue," Liu reflected. "We did not want to give a scenario that could give false hope to metastatic patients, or alternatively scare the curative patient with a metastatic scenario."
The researchers also interviewed patients about whether the prevalence of the tested biomarker changed their desire to get tested. This question was intended to assess whether the likelihood that the marker would influence a patient's treatment impacted the study participants' preferences.
In the curative population, testing preferences "were insensitive to variation of fractions of individuals carrying the genetics associated with lack of [chemotherapy] benefit," the researchers reported. Hon highlighted that in this group, 40 percent of patients would accept PGx testing, even if the prevalence of the gene varied between 5 percent and 95 percent. Similarly, metastatic patients' preferences were not impacted by the prevalence of the genes associated with toxicities in the population.
Hon and his colleagues also found that 76 percent of patients with early-stage cancer and 87 percent of patients with advanced disease wanted to be involved in the decision-making process for whether they should receive PGx testing. However, around 20 percent of patients acknowledged during the interview that they lacked sufficient understanding of PGx testing and the medical implications of the results.
"While patients had a strong desire to be involved in decision-making for PGT, a considerable proportion lacked the necessary knowledge to make informed choices," the researchers concluded in the study.
One audience member from Duke University noted that the fact that so many patients were willing to accept testing, regardless of how prevalent the marker was, suggested that the researchers may have underestimated how well study participants understood the PGx clinical scenarios presented to them. This attendee suggested that the researchers should have also tested patients' understanding of the genomic information in the study.
Liu told PGx Reporter that he and his colleagues have developed an iPad version of the survey and are administering it to patients this summer to determine whether a visual presentation of this type of information improves patients' understanding. With these new educational modalities, researchers are hoping to improve patients' "self-described understanding" by 10 percent compared to the earlier survey.
"The major problem with patients was the ability to understand what an X percent improvement meant in the context of the scenario and of their cancer," Liu noted. "So, we have worked with our department of biomedical communications at the University of Toronto to develop improved communication tools." The PGx clinical scenarios in the survey will be presented with animation and visual cues.
With this new study, "our focus is to assess the educational tools that we use, rather than just assume that they work," Liu said. "After assessment, we will adopt the strategy or strategies that fared the best, and will further present that data when we have received it."