Archive for the ‘Diagnostic’ Category

La tumor registry gets $794,000 pediatric grant

Comments Off
Posted 03 Oct 2011 — by James Street
Category Diagnostic, Finance and Politics of cancer research and treatment, Foundations, genetic research, genetic research, Osteosarcoma, Politics and Finance of Child Cancer research
JANET McCONNAUGHEY, Associated Press
Published 02:55 p.m., Monday, October 3, 2011

NEW ORLEANS (AP) — The Centers for Disease Control and Prevention is giving the Louisiana Tumor Registry a three-year, $794,000 grant to develop a system to quickly collect and report children’s cancers.

Hospitals often take six months or more to report cancer cases because they want to include information about treatment, said Dr. Vivien Chen, director of the registry at LSU Health Sciences Center New Orleans.

She said the grant will let the registry work with pathology laboratories, which diagnose cancers, and get that information within a couple of months. Regional registry workers will go to hospitals in their areas each month to get more information, she said.

Chen said the tumor registry will collaborate with state pediatric organizations and with doctors and hospitals treating children with cancer. Key partners include Lafayette, Baton Rouge and Shreveport clinics affiliated with St. Jude Research Hospital, large out-of-state children’s hospitals, and the LSUHSC-New Orleans pediatric oncology program at Children’s Hospital in New Orleans, where about half of the new pediatric cases in Louisiana are diagnosed or treated.

LSUHSC’s registry will also link to birth records, since a baby’s birth weight and any other abnormalities noted on the birth certificate, and even the parents’ ages may be linked to cancer, she said.

“As we move on, we might explore some other information. Medicare might be another thing we might link on,” she said.

This grant is the second awarded to the tumor registry since December and brings its federal support to about $3.5 million a year, according to LSU.

Scientists Work on Blood Test for Early Lung Cancer Detection

Comments Off
Posted 17 Sep 2011 — by James Street
Category Diagnostic, Lung Cancer

September 13, 2011

Jessica Berman

Scientists are developing a blood test to detect lung cancer, one of the most common and deadly cancers in the world.  The test, which looks for certain proteins in the blood, is designed to find tumors at their earliest, most treatable stage.

According to the World Health Organization, lung cancer is the leading cause of cancer death worldwide, claiming an estimated 1.5 million lives each year.  The disease is caused mainly by cigarette smoking.  Early detection followed by prompt treatment is essential to surviving this deadly, fast-growing cancer.

Researchers at the Fred Hutchinson Cancer Research Center in the northwestern U.S city of Seattle, Washington, report they have developed a new blood test for lung cancer proteins.  Those proteins are produced by tumor tissue early in the development of lung cancer and can be detected in plasma, a blood component that’s rich in proteins.

The scientists say the cancer test is so sensitive, it can detect the presence of markers or signatures that suggest tumor activity before they can be seen by advanced imaging devices such as a CT scan, which can spot tumors only a few millimeters across.

According to Sam Hanash, a scientist at Fred Hutchinson. and a lead researcher on the lung cancer blood test, using CT scans to detect tiny tumors can save the lives of patients at risk of lung cancer. But he says CT screening has a down-side: a high percentage of its images reveal nodules that appear as potentially malignant tumors.

“…That necessitate surgery, that turns out to be benign and a lot of other potential complications. So there’s a need for a blood test so that we can make CT scans more reliable,” Hanash said.

Hanash says the lung cancer blood test looks for protein signatures of the disease similar to the way other cancer blood tests work, including the CA 125 test for ovarian cancer and the prostate specific antigen, or PSA, test for prostate cancer.

In initial experiments with mice, Hanash and his colleagues discovered protein markers by switching on genes that gave the animals lung cancer, and then switching off the cancer-causing genes.

Hanash says scientists next looked to see whether they could find the same cancer protein signatures in human lung cancer cells.

“And the answer was “Yes!”  So that was pretty satisfying that in fact we’re not dealing with a curiosity type of finding that only mice seem to display, but we are dealing with a real feature of cancer cells whether mice-derived or human-derived,” Hanash said.

Hanash says researchers detected protein biomarkers unique to a number of different lung tumors, as well as some of the molecular networks of genes that drive tumor development.

He says the next step is to develop a test that doctors can use with patients at risk for lung cancer, probably in about two years.

An article describing the development of a new diagnostic test for lung cancer protein signatures is published in the journal, Cancer Cell.

Blood test to detect colon cancer gains traction, radiologists remain unconcerned

Comments Off
Posted 04 Aug 2011 — by James Street
Category Colon Cancer, CT Colonoscopy, Diagnostic
By Rebekah Moan | December 9, 2010

 

Despite the increasing popularity of blood testing for colorectal cancer, radiologists don’t have to worry CT colonography will be replaced just yet, according to experts.

A new test from Epigenomics and Warnex Medical Laboratories is infiltrating the ranks of the colon cancer screening world. On Dec. 6 the companies announced the Canadian launch of their diagnostic blood testing service. The test is derived from a blood sample and detects cell-free methylated DNA of the Septin9 gene shed into the bloodstream by colorectal tumors.

Diagnostic Imaging first reported on the Septin9 test in October. But researchers remain unconvinced Septin9 will edge out CT colonography.

“The Septin9 blood test is meant to identify colorectal cancer after it has developed and potentially already spread outside of the bowel wall,” said Dr. Judy Yee, a professor and vice chair of radiology and biomedical imaging at the University of California, San Francisco and a CT colonography expert. “It is not meant to prevent colorectal cancer.”

The purpose of CT colonography, on the other hand, is to identify precursor polyps before cancer develops, she said.

“The goals of the two tests are very different,” she said.

The Septin9 blood test may not be mature enough yet for use, according to Dr. Perry Pickhardt, a professor of radiology at the University of Wisconsin, Madison and also a CT colonography expert.

In the PRESEPT study, Septin9 detected two-thirds of colorectal cancers with a specificity of only 88%.

That means one in every three cancers would be missed, Pickhardt said.

“If this were applied to a screening population, where cancer is present in about one in every 500 adults, there would be more than 60 false-positive tests for every cancer detected,” Pickhardt said. “I don’t believe these are acceptable results, especially when you compare other noninvasive options, such as CT colonography, where the sensitivity and specificity for cancer detection is >95%.”

An attractive option might be following the Septin9 blood test with CT colonography because the risk of cancer is too low to justify colonoscopy in all cases, he said.

In any case, CT colonography remains viable in the radiology world for the detection of colon cancer.

Tumor Marker Testing Fact Sheet

Comments Off
Posted 03 Aug 2011 — by James Street
Category Biomarkers, Circulating Tumor Cells, Diagnostic, Personalized, Tumor biomarkers
By Megan McDowell | August 3, 2011

  • Tumor marker tests include a variety of tests for cancer that can be performed on cells of a tissue sample from a newly biopsied or stored tumor. Tumor marker testing provides the patient and oncologist with vital information about the tumor at the cellular level, expanding traditional pathology reports that are based on tumor size, appearance and staging of the disease.

  • Tumor markers are substances often detected in higher than normal amounts in the blood, urine or body tissues of some patients with certain types of cancer.

  • By providing insight into the genetic mechanisms driving tumor growth, tumor marker tests help guide decisions for timing and treatment choices.

  • All women diagnosed with breast cancershould have their tumors tested to help determine how their tumors will behave. Consideration should be given to repeating tumor marker tests on newly obtained tumor tissue if a woman’s breast cancer recurs after treatment. Changes that occur on a cellular level in the tumor during the course of treatment may have significant implications for disease management.

  • Women who have already had a biopsy and know they didn’t receive a tumor marker test may want to ask their physician to contact the lab where the sample was sent and request a tumor marker test be performed on the stored tissue sample.

  • Tumor marker tests are performed in the hospital’s pathologylaboratory or sent out by the hospital or oncologist to an independent laboratory. The tests use very specific techniques to reveal the presence or absence of markers such as hormone receptors for estrogen and progesterone on the surface of the cell or its nucleus.

  • Tumor marker tests include:

    -HER2 – The newest of the tests determines the presence of excessive amounts of the HER2 gene or protein. Alteration of the HER2 gene or protein in normal cells may lead to overexpression or overproduction of HER2 protein. Overexpression of HER2 contributes to an aggressive growth of the cancer and its spread to other parts of the body. HER2 overexpression occurs in 25-30 percent of women with breast cancer.

    -Estrogen Receptors (ER) – Studies have shown that estrogen, one of the female sex hormones, often regulates the growth of breast cancer. Knowledge of whether a tumor is positive or negative for the presence of estrogen receptors is used for prognosis and patient selection for anti-hormonal therapy

    -Progesterone Receptors (PR) – To help determine the response to hormonal therapy, the presence of the estrogen-regulated progesterone receptor is now determined routinely. The rate of response is much higher when estrogen and progesterone receptors are positive compared to estrogen receptors alone.

    -p53 – p53 is a tumor suppressor gene. Normally, the p53 protein, coded for by the p53 gene stops cells with DNA damage from multiplying until the DNA is repaired naturally or sends the defective cell into programmed cell death. When the p53 gene becomes damaged or mutated, the protein becomes nonfunctional and loses its checkpoint control, allowing cancerous cells to replicate more readily.

    -S-phase – In the process of cell replication, a cell cycles through a number of stages. After a cell has duplicated its genetic material and divided through the process of mitosis, it may become inactive or it can start another replicative cycle, beginning with the “S,” or synthesis phase during which genetic material duplicates again . Using a special technique, the number of cells in the S phase can be detected. A higher proportion of S phase than normal is a measure of how actively a tumor is proliferating.

  • Other tests, used primarily in research, conducted on breast cancer cells include DNA, cytometric evaluation to measure S-phase and/or DNA ploidy and tests of other genes such as Cathespin–D, CEA, and CA15–3.

  • Generally health insurance will cover tumor marker testing, however you should consult with your health insurance carrier or your physician.

(Reuters Health) – CT scans to measure lung tumors can be unreliable, potentially leading patients and doctors to believe the cancer is growing when it’s not, a new study suggests.

Comments Off
Posted 09 Jul 2011 — by James Street
Category Diagnostic, Lung Cancer, Lung Metastases, Osteosarcoma Diagnosis, Osteosarcoma Outcomes, Relapse

By Frederik Joelving

NEW YORK | Fri Jul 8, 2011 5:27pm EDT

In principle, that could mean stopping a treatment that is actually keeping the tumor in check, researchers say.

“The patient and the doctor both need to understand that small changes don’t necessarily mean much,” Dr. Gregory Riely, a lung cancer specialist at Memorial Sloan-Kettering Cancer Center in New York, told Reuters Health.

“Changes of up to 10 percent can happen simply as a result of the inherent variability of CT imaging,” he added.

Riely’s study, published in the Journal of Clinical Oncology, is the first to test how reliable lung cancer scans are — work that’s long overdue, experts say, because CT scans have already become the gold standard for measuring cancer growth and treatment response.

“It’s the sense of, ‘Really? Is this first happening now?’” said Dr. Michael Maitland, an oncologist at the University of Chicago, who wrote an editorial about the findings.

“This is telling us scientifically how much noise is naturally there without any treatment or the cancer getting worse,” he told Reuters Health. “It’s an important thing to do whenever you are going to use any kind of marker for a disease.”

For the study, the Sloan-Kettering team asked patients with late-stage lung cancer if they’d be willing to have two chest CT scans done within minutes — 33 said yes.

Doctors normally scan such patients every few months to see if their tumor is growing, which might be a signal to try a new drug.

Then the researchers gave the images to three radiologists who had no idea the scans had been repeated before the tumors could have grown or shrunk appreciably.

According to the radiologists’ measurements, however, many tumors had changed, ranging from 23-percent shrinkage to 31-percent growth.

Overall, three percent of the tumors appeared to have grown so much that doctors would diagnose disease progression according to common criteria. And the smaller the tumor, the bigger the variation.

Riely said some doctors will make treatment decisions based on tiny changes seen on scans, although that might be a costly mistake, according to the new findings.

“We begin to put more and more stock in the data without really understanding the true variability of those measurements,” he said. “Small changes are not clinically meaningful and we should not alter clinical care based on them.”

Riely stressed, however, that his results don’t mean patients should get repeat scans, which would increase their radiation exposure.

Most likely, the results also apply outside of lung cancer, although patients’ breathing could make the chest scans extra variable.

Maitland said the findings will also help drug developers, who look at increasingly small changes in tumor size during drug tests, forgetting that the scans might be unreliable at that scale.

“Many of the individuals analyzing data that way perhaps are not aware of that limitation,” he explained.

With the new data, scientists can build better models of cancer progression that might save both time and money in clinical trials.

“There is a real opportunity here to update our systems and take advantage of the new technology,” said Maitland.

SOURCE: bit.ly/puOY5x Journal of Clinical Oncology, online July 5, 2011.

Common Cancer Link May Unleash Potential of Antibodies

Comments Off
Posted 29 Jun 2011 — by James Street
Category Diagnostic, FSH receptor, Imaging, Understanding Cancer

By Greg Freiherr | November 29, 2010

The search for a magic bullet against cancer historically has glowed bright then dimmed, depending on the stage of discovery. Developments surrounding monoclonal antibodies and angiogenesis inhibitors have followed this cycle, as exuberance for their potential has bowed to the nuances that underlie the complex mechanisms on which they depend.

Now a new possibility promises to light the cancer scene, one that might allow oncologists to finally realize the full potential of monoclonal antibodies and angiogenesis inhibitors.

 

An international research team has found a link among common types and grades of cancer: extraordinarily high concentrations of follicle-stimulating hormone (FSH) receptors in the blood vessels feeding these cancers. The only normal blood vessels on which FSH receptors appear are those in reproductive organs and then only in much lower concentrations than the investigators found on the blood vessels that feed tumors.

The research done at the Mount Sinai School of Medicine in New York, in collaboration with France’s National Institute of Health and Medical Research, documented these high concentrations on blood vessel walls accompanying cancers of the prostate, breast, colon, pancreas, lung, liver, and ovaries.

Ultimately, the discovery could lead to the development of new diagnostics in MR, PET, or even ultrasound imaging, the researchers said. There is also potential for developing highly specific anticancer drugs.

Built around antibodies specific to FSH receptors, tumor imaging agents might be injected into the bloodstream, where they would selectively bind to the new marker to visualize early tumors, according to Aurelian Radu, PhD, an assistant professor of developmental and regenerative biology at Mount Sinai. These antibodies might also carry therapeutic agents to the tumors. One of the chief active components of these therapeutic agents is an antiangiogenesis agent.

The concept underlying antiangiogenesis as a cancer therapy is to slow or stop the growth of blood vessels that feed new tumor growth, thereby starving the tumor. Such efforts, however, have been complicated by the general nature of and the body’s normal dependence on angiogenesis. The presence of FSH receptors promises to simplify oncology’s attack plan (N Engl J Med 363:1621-1630, 2010).

New therapeutic agents might be tagged with antibodies specific to FSH receptors. Once injected into the bloodstream, they might bind to the FSH receptor in such a way as to block release of the vascular endothelial growth factor that stimulates the growth of blood vessels. Antibodies specific to the FSH receptor might even carry coagulants that clog the vascular beds that surround existing tumors or destroy these blood vessels, Dr. Radu said. The U.S.-French team evaluated tissue samples from the tumors of 1,336 people with any of the 11 most common cancer types and discovered high concentrations of FSH receptors on the blood vessels associated with all of these tissues, raising hopes that specific diagnostic and therapeutic agents can be developed.

Cancer cell tests may help predict drug reaction

Comments Off
Posted 04 Jun 2011 — by James Street
Category Biomarkers, Circulating Tumor Cells, genetic research, Targeted Cancer Therapy
Sat, Jun 4 2011

By Julie Steenhuysen

CHICAGO (Reuters) – Counting bits of prostate cancer found in a patient’s bloodstream may help doctors better predict which drugs work, U.S. researchers said on Saturday.

They said men in a large clinical trial of Johnson & Johnson’s newly approved drug Zytiga who had the lowest levels of circulating tumor cells — cells that break off from the main tumor and float around in the bloodstream — were more likely to survive than men who had more of these cells.

In a separate study, a consortium of lung cancer experts studied tumor samples from 830 patients and found that about half of them had at least one of 10 genetic mutations known to drive lung cancer and could be used to guide doctors’ treatment decisions.

The studies, presented at American Society of Clinical Oncology meeting in Chicago, are part of the search for biological signals known as biomarkers that can predict which patients are the best match for drugs or give an early indication that treatments are working.

The hope is these types of tests will speed effective treatments to patients and shorten clinical trials by giving researchers earlier signals that experimental drugs are working.

In the prostate cancer study, a team led by Dr. Howard Scher of Memorial Sloan-Kettering Cancer Center in New York looked to see if tests that trap circulating tumor cells — tiny bits of cancer cells — could show whether patients are responding to a drug.

So far, Johnson & Johnson’s Veridex unit has the only approved system for capturing these rare cells, but other teams are working on them.

CIRCULATING TUMOR CELLS

Scher’s team analyzed a study of 1,195 men with advanced prostate cancer that showed Zytiga, known chemically as abiraterone, extended patients’ lives.

Scher’s team looked to see if there was a correlation between men who fared best in the trial and levels of prostate tumor circulating in their blood.

The team tested the blood of 972 patients at the beginning of the trial, and 723 patients after three months.

They found that abiraterone cut the number of circulating tumor cells, and the men who had lower levels of tumor in their blood were more likely to survive.

Scher said the study will be used to develop tests to predict whether prostate cancer drugs are working.

In the lung cancer study, researchers from a consortium of U.S. cancer centers set out to collect 1,000 tumor samples from patients with a type of lung cancer known as adenocarcinoma.

Tests of the first 830 patients found that 54 percent had a mutation that drives tumor growth.

“Driver mutations control cancer growth,” Dr. Mark Kris, chief of the Thoracic Oncology Service at Memorial Sloan-Kettering Cancer Center, told the press conference. “If you negate the effects of those mutations in lab experiments, the cancer dies.”

In the study, 97 percent of the driver mutations were the primary driver of the cancer, making them useful for selecting specific drugs, he said.

Kris said many centers do not routinely test patients to see which type of mutation is driving their lung cancer, partially because of cost. But he said having this information can help doctors make treatment decisions and may even serve as a model for treating lung and other cancers.

(Reporting by Julie Steenhuysen; Additional reporting by Deena Beasley; Editing by Xavier Briand)

CT Colonography Equal to Ocular Colonoscopy, says Study

Comments Off
Posted 13 May 2011 — by James Street
Category Colon Cancer, Cost, CT Colonoscopy, Diagnostic, Ethics of Science, Finance and Politics of cancer research and treatment
By Deborah Abrams Kaplan | May 5, 2011

 

CT colonography is a better screening test than optical colonoscopy (OC), according to a new study published in the MayRadiology print issue. Using meta-analysis of studies done over a 15 year period, authors found that the sensitivity of CT colonography for colorectal cancer detection was 96.1 percent, compared with 94.7 percent for OC.

Authors noted that their study supported the clinical equivalence of CTC and OC for screening invasive cancers. They felt the CTC was mature enough to be seen as a universal procedure.

Aside from the sensitivity rates being roughly equal, there are other reasons that the CTC can be better for screening, including cost and the minimal invasiveness. One potential downside, however, is that CTC exposes the patient to radiation, though lead author Perry J. Pickhardt, MD, a professor of radiology at the University of Wisconsin School of Medicine doesn’t think the amount is concerning. “CTC is a low-dose exam applied to adult patients,” he said. “There are no meaningful implications related to the radiation exposure.”

The meta-analysis ultimately included 49 studies covering 11,551 patients, done from January 1994 (the year that CT colonography was first mentioned), through 2009. All the CTC studies were for screening, to diagnose colorectal polyps and cancer, and all positive results were proven histologically. Six studies, encompassing 42 percent of the participants, focused on asymptomatic patients typical in a screening setting. The remaining 43 studies included symptomatic patients and/or a disease-enriched population.

Authors note the low prevalence of invasive colon cancer in screenings, citing a cumulative 3.6 percent rate from the meta-analysis. A total of 414 colorectal cancers were found in the studies, including 20 in the screening group (less than a 0.5 percent prevalence rate) and 394 in the disease-enriched group (almost a 6 percent prevalence rate).

As for the cost, Pickhardt says that using CTC is less expensive. “Our work has shown that CTC is considerably more cost-effective as a primary screening test compared with colonoscopy.”

One reported reason that patients are less likely to get a colonscopy is because the evacuatory and uncomfortable nature of the bowel preparation. Plus, OC requires the use of some anesthesia or medication for patient comfort. With CTC, those issues are diminished, according to Pickhardt. “CTC is much less invasive, requires no IV for sedation or pain medication, and requires no recovery time,” said Pickhardt, adding that patients don’t need a driver to take them home. “In addition, our low-volume bowel prep is much better tolerated than the typical colonoscopy preps in use.”

UF researchers find quiet protein speaks loudly in fight against cancer

Comments Off
Posted 03 May 2011 — by James Street
Category Biomarkers, genetic research, Lung Cancer, Proteomics
Filed under Health, Research on Tuesday, May 3, 2011.

GAINESVILLE, Fla. — When a movie character says, “It’s too quiet,” that’s usually a sign something bad may happen.

Now, University of Florida researchers have discovered that when variations of a certain protein in our cells are too quiet, it may add to the risk that someone will develop lung cancer. When scientists restored the protein to its normal, active self, its cancer-inhibiting properties reappeared.

These discoveries, published in two reports in the online version of Oncogene, provide evidence that drugs can potentially suppress tumor growth by restoring cellular processes rather than inhibiting cancer-causing genes known as oncogenes.

“It’s a well-accepted fact that you can inhibit things, particularly oncogenes, that drive cancer. Oncogenes are the cancer’s gas pedal,” said principal investigator Dr. David Reisman, a UF associate professor of medicine and a member of the UF Shands Cancer Center. “What we’ve done is demonstrate the feasibility of reconstituting the cancer brake.”

The protein, known as Brahma, or BRM, is involved in the regulation of cellular functions like gene expression, DNA repair, cell adhesion and telling cells whether to divide and grow or stop dividing and die. Other studies have found “silenced” BRM is present in 10 to 20 percent of all solid tumors. Reisman knew from his own research in mouse models that silencing the BRM gene alone did not cause tumor growth, but when carcinogens were introduced, 10 times as many tumors appeared compared with mice with normal BRM expression.

“The gene was not a tumor suppressor in the classical definition but a tumor susceptibility gene, and when the expression is lost, it primes you to other events that potentiate the development of tumors, such as tobacco carcinogens,” Reisman said.

More people die of lung cancer every year than of cancers of the breast, colon, prostate or lymphoma combined, according to the National Cancer Institute. However, only 10 percent of smokers develop lung cancer and as many as 15 percent of those diagnosed with lung cancer have never smoked.

Reisman’s work suggests the presence of two variations within the BRM gene — known as polymorphisms — could potentially be biomarkers for lung cancer and assist doctors in identifying individuals at higher risk, which could lead to more cost-effective screening practices and lifesaving early detection.

Study investigators sequenced the genes of 160 people and learned that roughly 20 percent carry the gene variants. With collaborator Dr. Geoffrey Liu, a research scientist at the Ontario Cancer Institute at the University of Toronto, the team then verified the presence of the silenced BRM variants in human lung tumors.

Reisman and Lui also conducted case control studies on 1,199 people who were matched for age, gender and smoking history but in whom 484 individuals had lung cancer and 715 were healthy and cancer free.

“We found these polymorphic sites were greatly enriched in the population that had developed lung cancer,” Reisman said. “The chance that you would develop lung cancer if you had both polymorphic sites was 220 percent higher. Our analysis demonstrated those odds to be independent of smoking history, sex, race and cancer type.”

Reisman’s team also studied whether it would be possible to restore the normal expression of the BRM protein. Certain compounds, called histone deacetylase — or HDAC — inhibitors, had been demonstrated by other researchers to reactivate the BRM gene, but did not restore the normal, cancer-suppressing function of the BRM protein.

By introducing the healthy protein alongside the reactivated gene, the researchers were able to stop the growth of cancer cells. That makes the process a potential target for drug therapies to use in suppressing many tumor types.

“We know there are a lot of genes that are silenced in cancer, and it’s believed that gene silencing is necessary in order for the cancer to grow and thrive. This research demonstrates — and is really the first example of — an approach that’s led to the reactivation of a specific tumor-suppressing gene,” said Aubrey Thompson, a professor of cancer biology at Mayo Clinic Comprehensive Cancer Center in Jacksonville, Fla., who was not involved in the research.

“That’s a really big deal,” he said. “It’s an approach that is widely applicable to a lot of genes and a lot of different types of cancer. I think it’s going to be met with a great deal of enthusiasm and interest from researchers in human cancer therapy.”

Cancer Biomarker Technical Guide

Comments Off
Posted 01 May 2011 — by James Street
Category Biomarkers, Diagnostic, genetic research

Cancer Biomarker Technical Guide

Table of Contents

Letter from the Editor
Index of Experts
Breast cancer: Harold Garner, Dong-Young Noh, and David Rimm
Ovarian Cancer: Andrei Drabovich
Leukemia: Charles Mullighan and Richard Wilson
Lymphoma: Izidore Lossos
Circulating Tumor Cells: Paul Hofman
List of Resources

Letter from the Editor

Forty years ago, US President Richard Nixon declared a “war on cancer,” signing into law the National Cancer Act in 1971, which strengthened the National Cancer Institute and made the search for treatments a federal priority as well as scientific one.

Since then, researchers have worked to develop new diagnostic and prognostic tools, new treatments, find out more about various cancer subtypes, and create individualized drug regimens to help cancer patients. New studies have shown that these efforts have resulted in longer lives, lower mortality rates, and more efficient treatment of patients.

But cancer is a pernicious disease — it finds ways to hide, grow, migrate, develop, and mutate. It affects adults and children alike, and doesn’t discriminate. So, many researchers have made it their life’s work to find the hallmarks of cancer wherever they may be and figure out what these hallmarks indicate about how, where, and when the cancer is going to develop, so they can head it off at the pass.

The researchers featured in this technical guide all have different specialties, from breast and ovarian cancer to cancers that affect children and to the relatively new field of cancer stem cells. But the thing they all have in common is their determination to find biomarkers that could lead to earlier detection of cancer, better treatments, longer survival times, and lower mortality rates for patients. They utilize various approaches — searching for low-abundance proteins, rare variants, new gene pathways that lead to oncogenesis — in order to find biomarkers that will be of use to clinicians and possibly to biotech and pharmaceutical companies looking to create new therapies.

And take a look at the list of resources, Web tools, and upcoming conferences at the end of the guide for more information, and suggested reading.

— Christie Rizk

Index of Experts

Many thanks to our experts for taking the time to contribute to this technical guide, which would not be possible without them.

Andrei Drabovich
Mount Sinai Hospital, Toronto

Harold Garner
Virginia Bioinformatics Institute, Virginia Tech

Paul Hofman
Centre Hospitalier Universitaire de Nice

Izidore Lossos
Sylvester Comprehensive Cancer Center, University of Miami

Charles Mullighan
St. Jude Children’s Research Hospital

Dong-Young Noh
Korea Institute of Science and Technology

David Rimm
Yale University School of Medicine

Richard Wilson
The Genome Institute, Washington University School of Medicine

Breast Cancer: Harold Garner, Dong-Young Noh, and David Rimm

Genome Technology: What method or technology do you use to discover breast cancer biomarkers? Why?

Harold Garner: This technology is not just for breast cancer — breast cancer is our largest and most effective data, set but it appears to have potential to be more universally applicable to a lot of cancers. The basic technology is a unique microarray approach that looks at global changes in the repetitive or microsatellite content in the genome of the cancer patients versus non-cancer patients. The other important thing is that we’re finding this global change not only in the tumors, but also in the germline of the patients, so they’re born with a predisposition in their genomes that shows up in their tumors as well. In the human genome, which is 3 billion bases long, there are about 2 million loci that have repeats — repetitive motifs like CAG CAG CAG CAG CAG — in which case they’ll have some variable number of copies of that motif. However, these microsatellites are dramatically under-studied or under-appreciated as a causative agent in disease, in phenotype, even in speciation, and this is largely because there hasn’t been a technology to be able to look at them efficiently en mass. We developed a unique microarray that allows us to look at all 2 million microsatellites at one time — not at individual places, but the sum of all the contributions of them — and so what this does is the microarray reads out if there are any particular motifs that are highly different in one population versus another. For example, in the paper in Genes, Chromosomes, and Cancer (Galindo et al., 2011. See resource list), there’s a number of motifs, especially AT-rich motifs like the motif TTA, where in the germline or the blood of the cancer patients as well as in the tumor of the cancer patient, there are elevated amounts of that in the genome. The entire genome has changed in all the positions that include TTA motif, and so our array records that.

Therefore, we really create two kinds of biomarkers. One biomarker is simply the pattern on the array, and that pattern can be used to distinguish the people who are at elevated risk for cancer or not. The other thing it does is it provides us with leads for looking at individual loci — for example, the TTA motif may exist in thousands of places spread across the genome, and we know where they are because we’ve sequenced the genome. The array simply reads out that these things have changed, but we don’t know exactly which ones have changed, and the array tells us that the TTA-containing motifs are expanded. Then we use bioinformatics and the Human Genome Project to go and look where these TTA motifs are. And then we can select to look at those that happen to be in genes that might be implicated in cancer, and find individual places in the genome where an expansion could cause for you to have a cancer phenotype. In other words, the array is both a biosignature or a biomarker [detector], and also a lead generator. There’s a companion paper that came out in Breast Cancer Research and Treatment (Galindo et al., 2010. See resource list) — what we did is we found this motif to be changed on the array as well and then we traced it down and looked at this individual place and found out that it correlated and is probably implicated in causing cancer itself.

Dong-Young Noh: We use both genomic and proteomic approaches to identify novel biomarkers. Each method has its own advantages and disadvantages. We are currently focusing on the next-generation sequencing for the genomic approach and multiple reaction monitoring for the proteomic approach.

David Rimm: We use the AQUA method of quantitative immunofluorescence for biomarker discovery in breast cancer. QIF maintains spatial information (nuclear versus cytoplasmic, stromal versus epithelial, et cetera) but unlike immunohistochemistry can be easily multiplexed to optimize information obtained from colocalization. The biggest disadvantage of this method is that it requires a candidate target approach, and does not have the discovery potential of genomic-based methods.

Genome Technology: Do you use a multiplexed approach? Why or why not?

DYN: We have some experience in multiplex approach such as Luminex (Kim, Lee, et al., 2009. See resource list), but as mentioned before, our major interest these days is the MRM method to identify a novel set of biomarkers for breast cancer diagnosis.

DR: AQUA is based on multiplexing and co-localization. Thus, multiplexing is a critical underpinning of the approach. Multiplexing using AQUA can be achieved in two ways. True multiplexing is limited by the number of distinguishable fluorophores (about five using conventional methods), but we also do virtual or serial section multiplexing. We have shown that the reproducibility of AQUA scores between serial sections has an R value between 0.97 and 0.99. Thus, for practical purposes serial sections can be used to extend multiplexed queries. For example, we have serially multiplexed more than 40 biomarkers on a tissue section by using serial sections (Dolled-Filhart et al., 2006; Giltnane et al., 2009; Rothberg et al. , 2009. See resource list). However, when true multiplexing is used, it is a critical step in the validation of the approach to show that the multiplexed result is equivalent to the single-plexed result (Harigopal et al , 2010. See resource list).

GT: How do you validate those putative biomarkers?

HG: The way we do that is we create an expanded cohort. In the 2011 paper, basically, we expand the study from looking at what we discover on the arrays to looking at 500 or so patients. So really we validate in two ways — we do an expanded study, to be assured of the statistics for a given risk or pharmacogenomic informative marker. And the second thing we do is we undertake a mechanistic study so the best biomarkers are not only ones that are informative but also so you know why. What we wanted to do in this particular case is we found out this repeat expansion is in the promoter of this gene, and indeed the repeat itself is the promoter and so therefore when the repeat expands, the promoter activity expands. We understand more about the mechanism by which changes in the repeat length can change the performance of this gene leading to cancer.

DYN: The method of validation differs on the purpose of the biomarker we are looking at. For example, we commonly use a cohort of breast cancer patients and normal healthy control to validate diagnostic biomarkers. Mostly, we prefer blood samples such as plasma for early detection biomarkers, and usually we use the plasma after we deplete the abundant proteins in the blood by using commercial filters. For the validation of predictive markers, the platform for validation differs as the source of the biomarker, and mostly we use tissue samples such as fresh frozen tissues or FFPE for predictive biomarkers.

DR: Validation of biomarkers is critical and should be based on Hayes’ Levels of Evidence (Hayes et al., 1996, Simon et al., 2009. See resource list). It is challenging to reach level 1 or 2 evidence, but that is the best validation of a biomarker. To reach that level, typically assessment of two or three cohorts using the putative biomarker will provide sufficient data to apply for cohorts that can achieve level 2 (or 1) evidence. Single marker, small cohort studies are level 4 evidence and can often be misleading. They should be considered preliminary data and not validation data.

GT: What kind of samples, such as fresh, fresh-frozen, FFPE, or other, ensure an optimal screen for breast cancer biomarkers? Why?

HG: Any of them – basically we’re looking at genomic material and not transcriptomic material, so we’re looking at DNA instead of RNA, and we also are principally looking for markers in the germline. Tumors are unstable and their genome’s continuously shifting, and they’re particularly shifting in microsatellites, so of the most value here is to find something that is a predisposition marker that will predispose you to increased risk of cancer, predispose you to particular therapeutics that might be effective or not effective, things like that. Therefore, we need very little material depending on which assay — for the array we need 10 micrograms and for the follow-on looking at individual loci, as little as you can start with any PCR reaction.

DYN: We prefer fresh frozen tissues for the validation of gene signatures discovered by the genomic methods, however, in cases where there are limited numbers of fresh frozen tissues, we prefer RT-PCR using FFPE. Our opinion is that despite what tissue we are using, the most important thing is to have a standardized method of collection and storage and this principle applies to even FFPE where the time interval between resection and fixation is important for the antigen preservation.

DR: Breast cancer biomarkers can, in theory, be derived from fresh, frozen, or FFPE tissue. Historically, we have seen that to be translated to common usage in the US, biomarkers must work on FFPE tissue. However, in the EU, frozen tissue is more readily available. Often fresh and frozen tissue can be used for discovery experiments, but once candidate biomarkers are identified, the assays must be converted to use FFPE.

GT: What method or technology is best to detect low-abundance biomarkers?

HG: Microarray or sequencing.

DR: Low-abundance biomarkers, for example rare circulating tumor cells, or low-abundance message or peptide levels in serum, generally depend on specialized methods. While mass spec is promising for low-abundance serum markers, it has yet to be broadly validated in breast cancer.

GT: How do you determine which biomarkers are both sensitive and specific for use in the clinic?

HG: Out of our studies on expanded cohorts, the main goal is to understand what is the statistical sensitivity or specificity for this marker. Again, in your genome the markers are not dependent on the local environment of the tumor or what you happen to have eaten earlier today — things like that, that tends to change your transcriptome. It really comes down to larger extended studies.

DYN: We commonly use ROC curves to determine the usefulness of diagnostic biomarkers. To do so, it is very important to have a high-quality healthy control because while it is easy to have correct cancer patients, it is really difficult to have purely healthy control. Therefore, we use healthy control from the women who visit our comprehensive health care system and the overall health screening results show no evidence of cancer.

DR: Sensitivity and specificity required for clinical usage is a function of the clinical question. For example, in a screening setting, a test must have very high sensitivity and specificity. This has been a barrier for new serum-based screening in breast cancer since low specificity would result in significant numbers of unnecessary and expensive workups for false-positive tests. For companion diagnostic tests, the sensitivity must be very high, but specificity can be sacrificed and is often shockingly low. For example, a positive FISH or IHC test for HER2 in determining who should get trastuzumab must have high sensitivity since the clinician does not want even a single patient to miss the opportunity to benefit from this drug. However, the number of positive patients that do not benefit from the drug is probably close to 50 percent or more. Therefore, the current tests are designed to maximize the sensitivity at the expense of specificity. Once a more specific test is validated, it will be interesting to see whether or not it is adopted. One can imagine clinicians who want to give their patients “every chance” who may not be willing to withhold trastuzumab, even if some new test (with both high sensitivity and specificity) suggests that a HER2+ patient will not respond.

Ovarian Cancer: Andrei Drabovich

GT: What method or technology do you use to discover ovarian cancer biomarkers? Why?

Andrei Drabovich: To discover cancer biomarkers, we use an integrated proteomic platform to analyze multiple types of biological samples. We also employ gene expression and literature data mining to confirm our findings. Mass spectrometry-based proteomic approaches allow us to identify in biological samples as many as several thousand proteins, some of which may be putative cancer biomarkers. Typically, we use bottom-up proteomics and tandem mass spectrometry to identify proteins, while ELISA or immuno-mass spectrometry assays are used for accurate quantification of candidate biomarkers in blood serum. To narrow the candidate list to a manageable size, we consider proteins discovered in all types of biological samples such as human cancer cell lines, proximal fluids, and cancer tissues.

GT: Do you use a multiplexed approach? Why or why not?

AD: Absolutely. We use multiplex selected reaction monitoring assays to verify long lists of candidate biomarkers in proximal fluids and select candidates for further verification and validation. ELISA and immuno-mass spectrometry assays allow multiplexing only a few proteins in a single analysis, but are superior when verification in blood serum is required. Taking into account very high heterogeneity of cancer, there is a lot of potential for multiplex assays and panels of biomarkers to provide accurate cancer diagnosis and prognosis.

GT: How do you validate those putative biomarkers?

AD: Ultimate validation of biomarkers requires accurate protein quantification in blood serum and is typically done by ELISA. Validation of cancer-specific biomarkers by mass spectrometry is limited due to its insufficient sensitivity and low nanogram-picogram per milliliter concentration range of candidate biomarkers, in blood serum. Additional purification or enrichment steps increase sensitivity but decrease sample throughput. However, most medium-to-high abundance proteins are amenable to validation by targeted mass spectrometry in biological fluids and even in blood serum without additional purification. Hopefully, future advances in mass spectrometry instrumentation will increase its sensitivity to measure even low-abundance proteins.

GT: What kind of samples, such as fresh, fresh-frozen, FFPE, or other, ensure an optimal screen for ovarian cancer biomarkers? Why?

AD: We typically use fresh frozen samples for biomarker discovery, verification and validation. Formalin-fixed paraffin-embedded samples may not be suitable for mass spectrometry-based proteomics, but are a great source to verify tissue biomarkers by immunohistochemistry. Standardized sample collection and handling is crucial for accurate evaluation of biomarker performance. Accurate clinical information for each sample should be available and should include as many parameters — stage, grade, demographics — as possible to allow not only for biomarker verification but also for prospective or retrospective clinical studies.

GT: What method or technology is best to detect low-abundance biomarkers?

AD: ELISA is still a superior method to quantify low-abundance biomarkers in a large number of clinical samples. However, development of two highly specific antibodies suitable for ELISA is a long and challenging process that may not be practical if quantification of dozens of proteins is required. There is a lot of hope that the following alternative technologies will mature to improve or complement ELISA: (i) high-throughput development of engineered affinity ligands (aptamers, antibody mimetics); (ii) targeted mass spectrometry with increased sensitivity and automated sample preparation; (iii) immuno-mass spectrometry approaches such as SISCAPA.

GT: How do you determine which biomarkers are both sensitive and specific for use in the clinic?

AD: To determine biomarker specificity and sensitivity for use in the clinic, the proper validation with a very large number of samples should be performed. Such validation would first require development of a specific, sensitive, and high-throughput analytical assay. When such an assay is available, proper validation will require asking a very specific clinical question (biomarker for diagnosis, prognosis, or treatment selection), collecting a large cohort of samples (blood, proximal fluids, tissues) with clear and accurate clinical diagnosis, and following standardized study designs (double-blind, PRoBE). Extensive national and international collaborations between hospitals are crucial to collect the sufficient number of clinical samples and thus facilitate proper biomarker validation.

Leukemia: Charles Mullighan and Richard Wilson

GT: What method or technology do you use to discover leukemia biomarkers? Why?

Charles Mullighan: My work uses genomic profiling — SNP arrays, gene expression arrays, methylation, and DNA sequencing. The goal is to find new genetic alterations that drive the development of leukemia, and influence outcome. We study large cohorts of well-annotated ALL samples, so we can correlate new genetic changes with outcome. Once we find new genetic changes of clinical relevance from genome-wide approaches, we investigate the most appropriate platform for diagnostic testing. Other investigators at St. Jude use complementary approaches, such as flow cytometry to detect low levels of leukemic cells in patient samples during treatment (minimal residual disease measurement).

Richard Wilson: Whole genome sequencing. At the time we began, sequencing of selected genes was not telling us anything we didn’t already know about leukemia. We knew we had to somehow sequence all genes in the genome. It was about this time that next-gen sequencing (Solexa) came along. We presently use the Illumina HiSeq technology, and also utilize exome hybrid capture as an ancillary approach to WGS.

GT: Do you use a multiplexed approach? Why or why not?

CM: The genome-wide approaches incorporate assays for millions of markers. Specific assays tend not to be multiplexed, but it highly depends on the nature of the alteration: e.g. deletions, gains, or sequence alterations may be detected by qPCR, FISH, DHPLC, sequencing etc.

RW: Not for WGS. As the run yield of the Illumina HiSeq has increased, we have begun to multiplex exome sequencing. For capture-based targeted sequencing, we make extensive use of multiplexing.

GT: How do you validate those putative biomarkers?

CM: By studying large cohorts of patients in clinical trials.

RW: First by additional DNA or RNA sequencing, then using any one of a number of laboratory assays, expression in tissue culture cells, and mouse models to understand function and mechanism.

GT: What kind of samples, such as fresh, fresh-frozen, FFPE, or other, ensure an optimal screen for leukemia biomarkers? Why?

CM: We typically use any sample that can yield non-degraded material suitable for DNA and RNA analyses. Most commonly, these are fresh or cryo-preserved leukemic cells. Flow cytometry assays require viable cells.

RW: Fresh-frozen samples typically yield DNA and RNA of the highest quality and quantity. FFPE blocks are fine, although there is quite a bit of variability depending on how the samples are prepared and/or stored.

GT: What method or technology is best to detect low-abundance biomarkers?

CM: It depends on the marker. QPCR may be suitable for some variants. Flow cytometry is suitable for biomarkers expressed on the leukemic cell surface, and this approach is widely used for monitoring of levels of minimal residual disease in leukemia. This is well established in management of patients with ALL.

RW: We are primarily focused on discovering somatic mutations in tumor genomes. We find that Illumina sequencing allows us to work with as little as 10 nanograms of DNA. If whole genome amplification or PCR (for selected genes) were added to the front end, fewer input tumor cells would be required. Once mutations have been discovered and validated, there are several methods and technologies that one could employ to detect their presence in patient samples.

GT: How do you determine which biomarkers are both sensitive and specific for use in the clinic?

CM: This primarily depends on which biomarkers are shown to have independent prognostic or diagnostic value that complements the wide range of existing molecular and related tests that are currently in use. We have detected many new genetic alterations, many of which are of great interest in leukemia biology, but not all are clinically useful. To determine which may be used in the clinic, careful trials assessing the clinical utility (association with outcome, and the ability to measure robustly in a CLIA environment) are required.

RW: Our goal is to discover and characterize the driver mutations in leukemia, and other cancers. Key characteristics include genes that are frequently mutated in patients (7 percent) and appear to give some clue as to outcome. In acute myeloid leukemia, for example, mutations in DNMT3A and IDH1 are correlated with poor outcomes. In fact, presence of DNMT3A mutations may be the best indicator of which AML patients will require bone marrow transplant. The assay for IDH1 mutation is relatively straightforward as nearly all mutations occur at the same site in the gene; furthermore, the gene product can be detected by biochemical methods. In contrast, detection of DNMT3A mutations in patient samples will require sequencing all exons and splice sites in the gene.

Lymphoma: Izidore Lossos>

GT: What method or technology do you use to discover lymphoma biomarkers? Why?

Izidore Lossos: We are personally using three technologies. We are using microarrays, we are using real-time PCR for genes and microRNAs, and we are using immunohistochemistry mostly for validation of the results from the various other techniques. There are different kinds of biomarkers — when you want to have a global picture of the tumor, it’s very easy to do gene expression arrays because you are analyzing the whole genome, and once you have all the data from that, you can decide which specific biomarkers you want to use or to validate or to try to eventually translate to the clinic. Now, real-time PCR allows you quicker, cheaper, more precise with more quantities of analysis than gene expression, so once you know your targets, it’s definitely significantly easier to analyze them by real-time PCR, it’s more quantitative if you want to transform it to clinical applications. It would be easier to use real-time PCR than gene expression array [in this case] so that’s why we take it there.

GT: How do you validate those putative biomarkers?

IL: To validate it we use immunohistochemistry. You need to validate it several times. One validation is usually not enough so you need to validate in several settings. … Many people ask why biomarkers are stuck at the investigator level and are not getting to the clinic. And one of the reasons for this, in my opinion, is that many of these biomarkers are developed by academic institutions and we unfortunately don’t have the capabilities — we can develop the assay, we can develop the biomarkers, but we cannot make it available for routine use for whoever wants it. At a certain point in time, companies need to pick up the ownership of these biomarkers and make the assay available and unfortunately, very frequently, the companies’ considerations are how frequent the disease is, how frequently the test will be used for a specific biomarker. If you are not dealing with breast cancer or with something that is very, very common, many companies will be reluctant to establish the methodology or make the methodology for routine use and frequently many of the biomarkers are stuck at this transitional step.

GT: What kind of samples, such as fresh, fresh-frozen, FFPE, or other, ensure an optimal screen for lymphoma biomarkers? Why?

IL: It depends on what you’re using it for. If you have a serum biomarker, you use serum. But definitely your ability to retrieve expression of genes at RNA levels are frozen or fresh specimens is the best approach, but the problems is that unfortunately, this is not very practical. The majority of the patients are treated not in academic centers but in private practice and nobody will deal with preservation of frozen samples or fresh samples. You eventually will need to use paraffin samples. We’ve developed techniques, and there are commercial techniques, that allow you to measure RNA expression in paraffin samples without any problem and quite reproducibly. You can go both ways but it’s more practical to go with paraffin samples.

GT: What method or technology is best to detect low-abundance biomarkers?

IL: Once again, real-time PCR. There are new methodologies — I don’t have personal experience with them, but Nanostring seems to be a possible technology that will allow you to detect low-abundance biomarkers.

GT: How do you determine which biomarkers are both sensitive and specific for use in the clinic?

IL: Once again, validation, and validation, and validation. Usually the way to do it, you discover the biomarker and you need to independently validate it. Discovery is not usually a big problem, but when you discover you always have a problem that you may over-feed your data. You can decide that you are using this cutoff or this cutoff, because that’s how you’re getting the best result, and that’s correct, and that’s the first step to do it. But eventually you need to make sure with 100 percent integrity that that’s the best cutoff, so you need to have a sample in which you have a different sample of patients or different cohorts of patients, in which you need to apply it independently and see if your cutoff or your methods reproduces your results. For me, for example, if your P values are better in your validation set, I will be very, very surprised. Usually you have a lower P value and less clear separation in the validation set. But even then, there is always a possible bias in sample collections, so we usually try to go into prospective trials and to test specimens routinely with specific criteria and uniform protocols at every step and then to validate the biomarker in such a prospective trial. That’s why it takes some time.

Circulating Tumor Cells: Paul Hofman

GT: What method or technology do you use to discover ovarian circulating tumor cell biomarkers? Why?

Paul Hofman: Several methods can be used for this proposal. We use in the Laboratory of Clinical and Experimental Pathology (at the University of Nice Sophia in Nice, France) different methods: A direct method — namely, isolation by size of epithelial tumor cell or ISET technology — and an indirect method — the cell search method using more specifically, the epithelial Cell Search kit. We believe that the combination of different methods can greatly optimize the potential for the detection of circulating tumor cell in blood samples. Some CTC can weakly express the family of cytokeratin antigens since these cells can demonstrate an epithelial to mesenchymal transition phenomenon. Therefore, the cell search method cannot detect these latter cells which do not express cytokeratin but only vimentin. By contrast, the ISET method is able to detect these cells morphologically. Moreover, ISET method can potentially allow better characterizing the CTC by using an associated immunocytochemistry approach and/or by doing molecular biology such as FISH, for example. However, this latter method needs also to be improved in order to better quantify the detected CTC, and thus to increase its sensitivity.

GT: Do you use a multiplexed approach? Why or why not?

PH: Currently, we do not use a multiplexed approach in our laboratory. We believe, of course, that it is a very interesting approach, very challenging, and probably it is a cutting-edge format in different areas for the characterization of different mutations in the same tumor sample, for example. It will be very important to develop this approach in the field of molecular biology, in particular in the health care domain. However, to my point of view, this technology is not totally well developed in the health care field and, other more “classical” methods — such as direct sequencing, pyrosequencing, Taqman, et cetera — currently available in most of the laboratory, are more developed in routine practice.

GT: How do you validate those putative biomarkers?

PH: The putative biomarkers are validated by doing large series of patients, with well characterized cohorts of patients — matched in age, gender, pTNM staging, etc. Correlation with the overall survival, specific survival and disease-free survival parameters is systematically performed in a “training set of patients” then in a “validation set” of patients, which is an independent set of patients. The specificity and sensitivity are systematically evaluated for each biomarker. Ideally the validation of some biomarkers can be made using a tissue microarray approach built with several hundred of tumor samples.

GT: What kind of samples, such as fresh, fresh-frozen, FFPE, or other, ensure an optimal screen for these biomarkers? Why?

PH: There is not ideal “type of sample” for an optimal screening of new biomarker. However, comparison of the same biomarker obtained from different samples — such as for example from the plasma or the sera, and from the tissue fixed in formaldehyde or frozen tumors — is usually of great interest. Ideally, some biomarkers can be expressed in fixed tissue, allowing the pathologist to easily look for its expression. However, it can be also of interest to make xenograft in mice by using fresh sample of tumor and to look for the behavior of cancer cells after tumor implantation in mice. So, the best way is to have different collections of samples in the same laboratory: fixed and frozen tumor samples, plasma, sera, whole blood samples, and primary cell cultures plus mice xenograft.

GT: What method or technology is best to detect low-abundance biomarkers?

PH: Certainly new high throughput technology such as microarrays (DNA, microRNA microarrays, etc.) and new generation of deep sequencing.

GT: How do you determine which biomarkers are both sensitive and specific for use in the clinic?

PH: The correlation with clinical data is essential for this determination. It depends also if this biomarker is a diagnostic, prognostic, or theranostic biomarker. This approach needs to include in the data management a very strong biostatistics approach. It is necessary also to do a meta-analysis consideration to look if the same biomarker is described by different teams and using different technology platforms all around the world. In fact a very few meta-analyses exist for the biomarkers. A strong biomarker can be used in the clinic if the same results are obtained in different countries, with different cohorts of patients, and sometimes by different techniques.

List of Resources

Sometimes you need to know more. Here are more sources that may help you out.

Publications

Dolled-Filhart M, Ryden L, Cregger M, Jirstrom K, Harigopal M, Camp RL, Rimm DL. (2006). Classification of breast cancer using genetic algorithms and tissue microarrays. Clinical Cancer Research. 12, 6459-6468.

Galindo CL, McCormick JF, Bubb VJ, Abid Alkadem DH, Li LS, McIver LJ, George AC, Boothman DA, Quinn JP, Skinner MA, Garner HR. (2010). A long AAAG repeat allele in the 5′ UTR of the ERR-γ gene is correlated with breast cancer predisposition and drives promoter activity in MCF-7 breast cancer cells. Breast Cancer Research and Treatment. E-pub.

Galindo CL, McIver LJ, Tae H, McCormick JF, Skinner MA, Hoeschele I, Lewis CM, Minna JD, Boothman DA, Garner HR. (2011). Sporadic breast cancer patients’ germline DNA exhibit an AT-rich microsatellite signature. Genes, Chromosomes, and Cancer. doi: 10.1002/gcc.20853.

Giltnane JM, Moeder CB, Camp RL, Rimm DL. (2009). Quantitative multiplexed analysis of ErbB family coexpression for primary breast cancer prognosis in a large retrospective cohort. Cancer. 115, 2400-2409.

Harigopal M, Barlow WE, Tedeschi G, Porter PL, Yeh IT, Haskell C, Livingston R, Hortobagyi GN, Sledge G, Shapiro C, et al. (2010). Multiplexed assessment of the Southwest Oncology Group-directed Intergroup Breast Cancer Trial S9313 by AQUA shows that both high and low levels of HER2 are associated with poor outcome. American Journal of Pathology. 176, 1639-1647.

Hayes DF, Bast RC, Desch CE, Fritsche H Jr, Kemeny NE, Jessup JM, Locker GY, Macdonald JS, Mennel RG, Norton L, et al. (1996). Tumor marker utility grading system: a framework to evaluate clinical utility of tumor markers. Journal of the National Cancer Institute. 88, 1456-1466.

Kim BK, Lee JW, Park PJ, Shin YS, Lee WY, Lee KA, Ye S, Hyun H, Kang KN, Yeo D, Kim Y, Ohn SY, Noh DY, Kim CW. (2009). The multiplex bead array approach to identifying serum biomarkers associated with breast cancer. Breast Cancer Research. E-pub.

Rothberg BE, Berger AJ, Molinaro AM, Subtil A, Krauthammer MO, Camp RL, Bradley WR, Ariyan S, Kluger HM, Rimm DL. (2009). Melanoma prognostic model using tissue microarrays and genetic algorithms. Journal of Clinical Oncology. 27, 5772-5780.

Simon RM, Paik S, Hayes DF. (2009). Use of archived specimens in evaluation of prognostic and predictive biomarkers. Journal of the National Cancer Institute. 101, 1446-1452.

Upcoming Conferences

Translational Approaches to Cancer
May 23-27, Suzhou, China
Cold Spring Harbor Laboratory Asia Conferences

2011 ASCO Annual Meeting
June 3-7, Chicago
American Society of Clinical Oncology

Cancer Proteomics
June 20-23, Dublin
Select Biosciences

13th Breast Cancer Milan Conference
June 22-24, Milan, Italy
European Institute of Oncology

Cancer Genomics
September 17-19, Heidelberg, Germany
EMBO-EMBL

Web Resources

Virginia Bioinformatics Institute
http://innovation.vbi.vt.edu/

ChipDX Tumor Analysis Platform
http://www.chipdx.com

← Previous Entries
Next Entries →