There’s a new genomic test in town — but can it really provide the payoff it claims for women with breast cancer?
The test, called MammaPrint, uses patterns of gene expression to assess a woman’s risk of having a recurrence of breast cancer following surgery. She and her doctor can then use this information to decide whether chemotherapy is needed after the operation, a treatment known as adjuvant chemotherapy. MammaPrint’s utility was evaluated in a large phase III trial, the results of which were presented in April at the annual meeting of the American Association for Cancer Research (AACR); a paper on the study is being published today in the New England Journal of Medicine (NEJM).
According to the study, women with early-stage breast cancer deemed to have a high risk of recurrence by standard clinical measures — such as tumor size and grade — but a low risk by MammaPrint did about equally well whether they received adjuvant chemotherapy or not. Just under 95% of women whose tumors fell into this category and who received no chemotherapy were still alive and free of distant metastases five years later.
Because chemotherapy can be associated with significant risk of toxicity — including future cancers — there is great interest among oncologists and patients in avoiding unnecessary treatment. Based on these new findings, the investigators of this study (called MINDACT, for Microarray In Node negative Disease may Avoid ChemoTherapy) conclude that about 46% of clinically defined high-risk patients who are currently prescribed chemotherapy may not really need it.
This was big news at the AACR conference, and garnered a lot of attention in the press. Memorial Sloan Kettering’s Physician-in-Chief, José Baselga, who is on the board of the group that ran the study, called the results “practice-changing,” adding, “This study is telling us in a very clear way we can spare many women chemotherapy.”
MammaPrint is not the only genomic test on the market: Oncotype DX, the one most commonly used in the US — as well as at MSK — also aims to spare women from unnecessary chemotherapy. So how can patients make sense of the pros and cons of each? To find out, we did some digging.
A Tale of Two Tests
Oncotype DX is made by California-based Genomic Health. To date, it is the only test officially recommended by the National Comprehensive Cancer Network, a consortium of US-based cancer centers that develops evidence-based treatment guidelines.
MammaPrint was developed in the Netherlands by a company called Agendia, but is approved by the FDA for use in the US and is covered by most insurance.
Both tests use gene-expression signatures to assess recurrence risk, but there are important differences between them. Oncotype uses a signature of 21 genes that relate primarily to estrogen signaling, which is known to influence cancer aggressiveness. MammaPrint uses a signature of 70 genes that were identified “agnostically” by comparing tumors that had a good outcome to those that did not.
Oncotype gives a graded scale of risk broken up into low, intermediate, and high risk categories, whereas MammaPrint is either low or high risk. Oncotype is applicable only to estrogen receptor (ER)–positive tumors; MammaPrint can be used for both ER-positive and ER-negative tumors.
Perhaps the biggest difference between the two tests is the percentage of women whom they classify as low risk. With Oncotype, about 15% of women with early-stage breast cancers are categorized as such, versus about 45% of women with MammaPrint.
“That’s a big difference,” says Dr. Baselga. “With MammaPrint, you identify more patients who are candidates for avoiding chemotherapy.”
And yet, because of the way the MINDACT study was designed, there are some important caveats for doctors and patients to consider when drawing conclusions about MammaPrint — and when interpreting who is considered low risk by either test.
Prognostic versus Predictive
The MINDACT study was a phase III, prospective randomized trial that enrolled nearly 7,000 women from 111 cancer centers in nine countries in Europe. Each participant had her tumor evaluated in two ways: by a clinical tool called Adjuvant! Online, which assesses standard clinical measures like size, lymph node involvement, grade, and age, and by MammaPrint. Those women whose tumors were scored as low risk of recurrence by both measures were given no chemotherapy. Those whose tumors were scored as high risk by both measures were given chemotherapy. Those whose tumors were discordant between the two measures were randomized to either follow the clinical tool or the MammaPrint test when determining whether to receive chemotherapy. The investigators then followed these groups for five years and assessed distant metastatic disease-free survival.
What the study team found was that the group of clinically high risk/genomically low risk women who did not receive chemotherapy had excellent survival: 94.7 of them had no evidence of distant metastases after five years. The other group, which received chemotherapy, also did very well: 95.9% of them had the same result. But the study was not designed to determine whether this difference between the two arms was statistically significant — one of its limitations.
“You can’t rule out a benefit [of chemotherapy],” says Maura Dickler, Interim Chief of MSK’s Breast Medicine Service. Dr. Dickler co-authored an editorial on the MammaPrint study with Clifford Hudis, former Chief of MSK’s Breast Medicine Service and current CEO of the American Society of Clinical Oncology, that was also published today in NEJM. Though both groups did very well, she says, it remains theoretically possible that chemotherapy gives some extra benefit. And, indeed, the trend was for chemotherapy to add a benefit (albeit a small one, of about 1.5%).
Another way of saying this is that MammaPrint has prognostic but not predictive value. This is in contrast to Oncotype DX, which has both.
“Oncotype gives you a risk assessment of how likely your breast cancer is to recur,” says Tiffany Traina, a medical oncologist at MSK who specializes in breast cancer. This is its prognostic value. “But where it has real value to us is that it has predictive value, meaning not only do I know you have a high-risk tumor, but I know that giving you chemotherapy is going to make a significant impact. Or I see that you have a low-risk cancer and I now have proof that chemo is not going to help, so we can avoid using it.”
For that reason, many doctors who currently use Oncotype DX will likely continue to use it over MammaPrint.
Factoring in Side Effects
Given that metastatic breast cancer is almost universally fatal, there is a great interest in not under-treating patients — which is why most of them currently receive some form of chemotherapy. But the downside of this approach is that it means there is a group of women who are probably over-treated, getting chemotherapy when they don’t really need it and experiencing a decreased quality of life as a result.
Although not all chemotherapy drugs carry serious side effects, many of them do. Dr. Baselga rattles off a list of potential issues, including “an increased risk of developing leukemia, hair loss, menopause, potential risk of heart damage, risk of ending up in the hospital with a low blood count, and infections.”
These potential drawbacks, he says, represent an important counterbalance to the potential benefits of chemotherapy and must be seriously considered by both the doctor and the patient when making a treatment decision.
For some women, these risks may be worth it to improve their overall chances of survival. “There is evidence in the literature that women will do a lot for small benefits,” says Dr. Dickler. “They’ll take chemotherapy for even a 2 to 3 percent improvement in outcome.” For others, quality of life may be more important.
Weighing the Numbers
Everyone agrees that it would be valuable to have a test that provided a clear quantification of the “cost” of forgoing chemotherapy, so that women could weigh the risks and benefits of that choice. Unfortunately, as Drs. Dickler and Hudis point out in their editorial, the MINDACT study does not permit this sort of calculation.
To see why this is so, we need to know a bit more about the study design. The study’s primary goal was to show that the MammaPrint test could yield an outcome, in terms of survival rate, that was better than a predetermined benchmark. That benchmark was a distant metastasis-free survival rate of 92% at five years. The actual study result of 94.7% exceeded that benchmark, and so the test was deemed a success. But the statistical margin of error was plus or minus 2.2% points on either side of that number, meaning it could be as low as 92.5% or as high as 96.2%. And so the decision to forego chemotherapy becomes, in part, a question of how comfortable one is with that level of uncertainty.
“The argument against chemo is a subjective one,” says Dr. Hudis. “Is 92% [the predetermined benchmark of success] good enough to exclude a chemo benefit? My answer is, I don’t know.”
By comparison, the trial evaluating Oncotype DX found with very good statistical evidence that 99% of women classified as low risk (a score of <11) who did not have chemotherapy were alive and free of disease at five years. (Although, again, the number of women who fall into this category is relatively small: 16%.)
Large, randomized studies evaluating the benefit of chemotherapy (or lack thereof) in women with intermediate scores on Oncotype (between 11 and 25) are ongoing and will help to answer the question of whether more women can be spared unnecessary chemotherapy.
Right now, says Dr. Hudis, comparing the number of low-risk patients as judged by Oncotype DX and MammaPrint is like comparing apples to oranges, given the different definitions of low risk across the trials, the relative strength of the evidence evaluating the impact of chemotherapy, and the subjectivity that patients and doctors bring to their assessments of risk and benefit.
Using gene-expression signatures to make treatment decisions represents the latest in a long line of efforts to tailor treatments to individual patients. It used to be that women with early-stage breast cancer were routinely given radical mastectomies, on the theory that cure was possible only by removing all possible places where breast cancer cells may have spread locally — including breast tissue, surrounding lymph nodes, and even muscles of the chest wall. MSK surgeons in particular were vocal supporters of this approach for many decades.
By the mid-1970s, a competing view emerged that favored less-invasive surgical procedures (like lumpectomy) paired with adjuvant therapies like radiation and chemotherapy, which were designed to kill circulating tumor cells that remained in the body after surgery. By the 1990s, virtually all women with early-stage breast cancer were being treated with adjuvant chemotherapy to help prevent recurrence.
Scientists now know that breast tumors differ widely in their aggressiveness. Tumors that overexpress the protein HER2, for example, are much more aggressive than those that do not. Even very small (less than 1 centimeter) HER2-positive tumors can be very dangerous and call for specialized treatment (usually with the targeted drug trastuzumab [Herceptin] plus chemotherapy). But other tumors, even large ones, can be more indolent. Genomic tests help doctors to discriminate between more and less aggressive tumors, and therefore assess risk of recurrence.
A Better Test?
Oncotype DX was the first genomic test available for assessing risk of recurrence. As such, says Dr. Baselga, it has spared many women from unnecessary chemotherapy. The test is performed as a standard part of care at MSK for patients whose tumors are bigger than 5 millimeters and estrogen receptor–positive.
What does MammaPrint add to this picture? “It’s an additional piece of information that might be useful in some cases to discern how a person will fare without chemotherapy,” Dr. Baselga says.
Though there are limitations, the test might help some women feel more comfortable with their decision to forego chemotherapy.
Dr. Baselga adds: “This is not the gospel, but it’s something to show patients, to help them make an informed decision.”