|
By Jorge Leon
10/09/07
Biomarkers: Ready for Primetime?
Which biomarkers are ready for implementation and how can you evaluate and
predict market acceptance? This article will compare the old molecular diagnostics
test menu with a new one, as well as explain the elements of a successful test
and where future opportunities exist.
A Changing Landscape
“The landscape of diagnostics is changing fundamentally,” says
Jorge Leon, PhD, President of Leomics Associates Consulting, Inc. (Emerson,
NJ). “It’s not just about diagnostics screening, it’s about
genetic risk and therapeutic intervention, and the trilogy of molecular diagnostics
with clinical proteomics and molecular pathology that are the key drivers of
this process.”
From a science and technology perspective, Leon notes that
there is an abundance of leads for biomarkers. “We have never had in
the history of diagnostics so many new leads, so many new markers that have
been propelled and catalyzed by the development of genomic technologies. We
can therefore perform larger and longer technical and clinical validation trials.
The average cost of developing a diagnostics test today is around $8 million,
but it can go as high as $50 million.”
In addition, Leon notes that the
availability of lower-cost platforms and chemistries—including
miniaturization, homogenous reactions, and horizontal automation—are
going to revolutionize the way global molecular diagnostics functions. On a
global scale, Leon notes that there is even less access to molecular diagnostics
because the costs are significantly higher than they are in the United States. “Fortunately,
we have new technologies coming in that are going to be key in enabling these
large populations to access molecular diagnostics and better medicine.
“Finally,
another key is that molecular pathology and molecular tissue pathology is evolving
dramatically and is incorporating new tests, which will add clinical and anatomic
pathologies to the picture of molecular diagnostics, Leon explains. “Molecular
diagnostics is not going to displace physicians and pathologists. On the contrary,
it is going to accelerate and increase their rate of efficiency and efficacy
in their diagnostics tools.”
The Five Fields of Clinical Diagnostics
According to Leon,
there are five major fields in molecular diagnostics today.
1. Predisposition
testing for the general population as well as the population at high risk. “This
is important because there is a new wave of pharmacology called chemoprevention,” he
notes. “Chemoprevention is the way to delay or actually block the onset
of the disease. For instance, in breast cancer, patients that have BRCA 1 and
BRCA 2 genes have an 85 percent risk of developing breast cancer. If they take
Tamoxifen or Raloxifen, the risk of breast cancer is reduced somewhere between
50 and 72 percent. That is a clear intervention of reducing the incidence of
disease with a genetic test. The same thing is happening with macular degeneration,
colon cancer, and other diseases, and we’re going to see a large wave
of new chemoprevention trials in the next 10 years that are based on a genetic
test that pre-selects patients that are at high risk.”
2. Screening. “Screening
by imaging is very cost-ineffective unless you have a test that identifies
a group at a very high risk of developing that disease,” Leon says. Such
is the case with breast cancer—if you have a patient at high risk you
can do MRIs or two mammograms per year, but that’s expensive and will
not detect the cancer. Even for a base MRI, it takes about 0.3 to 0.5 centimeters
for that mass to be detected. We need better tests to screen for earlier cancers.”
3.
Diagnosis. “The
important aspect of differential diagnosis that molecular diagnostics is enabling
today is based on the provision of better prognostic information, which is
important to better classify the diseases and make intervention decisions,
he says. “In addition, the diagnosis of pre-malignant lesions with a
high risk for cancer is one of the new modalities of molecular diagnostics,
especially epigenetic testing (DNA methylation).”
4. Companion
diagnostics for therapeutic purposes.
5. Disease
monitoring for progression, recurrence, and response to treatment.
Main Targets for Biomarkers
According to Leon, the main targets when companies start looking for new biomarkers
are as follows:
- Responders
versus non-responders—i.e., herceptin, which is an antibody for treating
metastatic breast cancer.
- Toxicity
prediction
- Dosage
- Risk
of disease to justify or discover uses of chemoprevention. “This is a
very important situation for cancer, and one of the best examples is cervical
cancer,” Leon illustrates. “We know that the human papillomavirus
(HPV) causes cervical cancer. Before the vaccine actually eradicates cervical
cancer, which will not happen in the next 20 years, we need to intervene with
patients that are infected and identify which ones are actually developing
lesions or dysplasias and block that process.”
- Detection
of early disease or pre-existing conditions that justify a drug intervention.
Tests
that provide additional opportunities for biomarkers that detect and rationalize
early intervention and/or chemoprevention, include osteoporosis, Alzheimer’s,
macular degeneration, breast cancer, ovarian cancer, RA/gastrointestinal disease,
RA/cardiovascular disease, pre-eclampsia, and premature labor.
The chart at
right provides a detailed view of the molecular marker landscape.
Old Versus
New Menu
“The old molecular diagnostics menu was established
in the last 15 years and the new phase of proven, validated tests are beginning
to enter the marker and will be widely accepted in three to five years,” Leon
says. The “old” tests
includes, prenatal cystic fibrosis; genetics factor 2 and 5, hemochromatosis;
gene rearrangements (i.e., BCR ABL, HER-2, and BRCA 1); and infectious diseases
(i.e., HIV, HPV, chlamydia, gonorrhea, and herpes).
“In the new molecular diagnostics prenatal area, we’re going
to see the imminent appearance of new methods to diagnose cystic fibrosis (CF)
and other prenatal conditions, using either circulating DNA or circulating
fetal cells in blood,” Leon says. “In genetics, the Warfarin dosage
is going to be the next homegrown test, which probably will sell half a million
or a million tests per year. In oncology, we’re going to see tests for
predicting which patients will respond to receptor blockers, which is the next
generation of inhibitors that are proving to be very effective in some cancers
when used in combination with chemotherapy or herceptin. There are tests that
can pre-select for those patients.”
Sometimes there are tests that have
been here for 10, 15, or 20 years and then technology improves and so does
the test,” Leon points out. “Estrogen
progesterone receptor is one example. We have been using this test for 20 years,
with immunochemistry at the center of practice. It has been shown that if you
measure and quantify the ER/PR by MRNA, you actually can predict Tamoxifen
response with a much higher degree of accuracy than if you just do immunochemistry.
The test is more reproducible, more reliable, and it has a better predictive
value. There is devastating evidence that this is the case. At the San Antonio
Breast Cancer Symposium this past December, there were over 30 papers showing
exactly the same thing in 30 different places.”
Leon also believes that
the same thing will be true for HER-2—i.e., that
HER-2 measured by MRNA expression is going to be much more reliable than HER-2
measured by immunochemistry.
Elements of a Successful Test
According to Leon, five elements are critical to guaranteeing the success of
a new molecular diagnostics test.
1. Technology.
The technology must be solid, reproducible, accessible, and cost-effective. “Five
years ago we all lived through the fiasco of clinical proteomics,” Leon
relates. “Everyone rushed to market with this ovarian cancer test—it
was even published in journals and endorsed by the government and consumers.
However, the technology was not good—it wasn’t reliable or solid.”
2.
Clinical evaluation. “We have to make sure the clinical evaluation is
done with multiple cohorts, multicentric trials at a minimum of two or three
different sites, transferring technology to the sites, and with at least 1,000
patients,” he
explains.
3. Clinical
economics. According to Leon, you need to determine why payers should cover
the test. In other words, what is the effect of this test in saving money for
the payers or the people that manage that disease?
4. Interpretation. “Physicians
need to be provided with clear interpretation that can enable them to take
clinical action,” he says.
5. Cultural
platform. This is very important because you need to determine if the test
will replace cultures,” Leon notes. “For example, will it replace
colonoscopies and, as a result, upset gastroenterologists.”
“These
factors need to be very carefully considered and analyzed before defining the
probability of success,” Leon points out. “The probability
of clinical success, and the probability of commercial success in a given period
of time, is what really defines the odds of a test making it in the market.”
Companion
Diagnostics and Personalized Medicine
“Companion diagnostics is really the entrance to personalized medicine,” Leon
says.
They are used in the selection and dosage of a specific drug (or combination
of drugs), and to define a therapy or a specific intervention. There are three
types of companion diagnostics tests:
1. Target
diagnostics—for example, herceptin. You look at polymorphorisms in the
receptor of the drug itself and how that changes response.
2. Toxicity
is the metabolism of that drug.
3. Pathways,
where you look at changes in the genes that control a pathway that is independent
of therapy predicting outcomes.
The chart on the facing page illustrates the
types of companion diagnostic tests in detail, along with their applications.
There are many companion diagnostics tests today in the market. Many of them
are rare, low volume. There are a number of current key ones, which are outlined
in the two charts below.
However, while there are a number of companion tests,
there is still a lot of room for improvement, Leon notes. For example, the
negative predictive value of herceptin (HER-2) is very good, but the positive
predictive value is poor at 45 percent. In addition, estrogen and progestin
receptor tests have a relatively poor positive predictive value. At the end
of the day, 40 to 45 percent of these women with breast cancer end up developing
resistance to tomoxophin. A new generation of tests that will predict resistance
to tomoxophin at a higher level, and predict resistance to herceptin at a better
level, are going to be here in the next 12 to 18 months and are going to help
us increase the efficacy of those treatments.”
Leon believes that payers,
rather than the pharma industry, will play a critical role in driving testing
improvements because doctors rely on the testing information. As he sees it,
the pharma industry is simply protecting existing sources of income.
“However,
Medco, one of the largest payers for drugs, sells pharmaceuticals to 60 million
people and they endorse personalized medicine,” Leon explains. “They
have indicated that they will identify the top 10 pharmacogenetics tests in
the coming months, endorse them, and tell physicians that are giving prescriptions
to patients to give them those tests. That’s going to catalyze personalized
medicine because they have a captured population, they are endorsing it, they
are paying for the cost of that testing, and they are going to benefit directly
from it. This is going to be a catalyzing event that perhaps is going to play
a critical role in the increased interest in personalized medicine.”
Where
Opportunities Exist
According to Leon, the short-term future of companion diagnostics will break
down as shown in the chart at right.
Leon notes that the potential leaders
in the field will be those that:
- Integrate
techniques and standardize them
- Consolidate
multiple biological and clinical information into a single clinical interpretation
and report
- Excel
in clinical validation and medical and consumer education
- Succeed
in lobbying and forming alliances with pharma, the National Institutes of Health,
and the FDA
However, Leon notes that pharmacogenetics on the pharma side is
going to be a marketing issue. “The marketing people need to start using
pharmacogenetics to start marketing themselves better and to differentiate themselves
from their competitors. There are new players in diagnostics and there is a need
for new innovation,” he says. “We have imaging companies, life science
companies, IT companies, managed care, and pharma. Diagnostics companies that
are good investments are going to get the attention of the investors and, ultimately,
receive more money for research and development.”
The projected market timeline for upcoming tests is reflected in the chart
at right.
In addition, the chart at the top of the facing page outlines historical
growth drivers and projections in molecular diagnostics.
Molecular diagnostics is, as Leon puts it, joining the million-dollar test
club of diagnostics—i.e., tests that are given more than a million times
each year—for instance chlamydia and HIV. However, he notes that there
are some critical tests flying under the “revenue radar,” including:
- B/T
Gene Rearrangements
- BCR
ABL
- Chromosome
FISH
- HNPCC
- RET
Oncogene
- UGT
1 a1
- 6
TPMT
- 5FU
- West
Nile Virus
- Methycillin
Resistant Staph A
- wXDX
Transplant Rejection
- GH
Breast Cancer
- Aviara
Tumor of Unknown Origin
- Veridex
Tumor Micromets
- Bird
Flu
The middle chart provides a snapshot of the potential for different areas
of opportunity for testing.
The chart at the bottom details the potential of
tests among diagnostics applications.
Assessing Opportunities
“Whenever you assess an opportunity, there are three things you need
to determine,” Leon says. First, is it real—do the biology, medicine,
technology, IP, and markets exist. Second, is it worth it—what is the
anatomy of the opportunity, the probability of success (technical, clinical,
and commercial), and what happens if you don’t move forward. Third, can
we win—in terms of how, when, and how much.
“I don’t think
good science is necessarily good business,“ Leon
admits. For good science to be good business, it has to be good medicine,
society has to buy in, you have to save lives, as well as make and save money.
To address all these requirements, you need to have the whole picture in front
of you.”
Finally, Leon provides his “picks” as to where the opportunities
in molecular diagnostics will exist.
Lessons Learned
To summarize, Leon offers the following “lessons
learned” about
the future of molecular diagnostics.
1. We
all read journals and we’re excited about molecular diagnostics—however,
not all physicians have the time to read journals. “Education is a critical
gap,” he notes.
2. Resistance
to change is a phenomenal blockage. “People don’t want to change
and there are cultural barriers,” Leon says.
3. It
always takes longer than planned to commercialize a test—at least twice
as long.
4. About
90 percent of the tests that are published do not materialize or do make it
in the market.
5. No
other segment of health care is growing as fast as or has more product leads
than molecular diagnostics.
6. Technical
and clinical evaluation is becoming more expensive. “Gone are the days
when developing a diagnostics test cost $50,000 or $1 million. Today we have
to invest about $10 million from discovery to study on 1,000 patients (before
approval and marketing). It is expensive, but worthwhile,” Leon says.
7. Walk
the reimbursement and clinical path before you go commercial.
8. Genes
do not work in a vacuum; you need genetic counselors and doctors to manage
that information.
9. Do
not exaggerate the capability of a test.
|
