Yesterday, I attended a half day symposium on companion diagnostics (CDx) efforts in San Diego which was organized by the SABPA Science & Technology Forum. CDx is the identification and detection of biomarkers to predict whether a drug will work or not in a given patient. Two successful marketed products are HercepTest (marketed by Dako) for Herceptin and KRAS tests for Erbitux and Vectibix. San Diego biotech has a rich history and deep investments in diagnostics. This was reflected in the lineup of various talks.
CLARiENT (a GE company) is one of the major diagnostic companies in SoCal. They service over 1200 hospital pathology units a year and perform tests for virtually all standard biomarkers, including, Her2/neu, BRAF, EGFR mutations, etc. Their website lists over 300 tests. John Glassco, COO of Clarient, who gave the first talk of the day, said that Clarient also works with biopharma to identify, validate and incorporate CDx biomarkers in clinical trials. AltheaDx of San Diego, in partnership with Compendia Biosciences (Ann Arbor, Mich.) has investigated cancer genomic signatures and developed disease specific arrays. AltheaDx has developed a breast cancer panel which captures therapy and disease-stratification information; the results are expressed as an OncoscoreTM. Breast cancer panel is under review at FDA and should come in the market soon. Other OncoscoreTM panels in development are for colorectal, lung and ovarian cancers. BioNanomatrix is developing a novel microfluidic technology (see U.S. Patents 7,217,562 and 7,670,770) which uses a nanochannel fluidic chip to separate long genomic DNAs and thus providing disease-specific structural information such as Copy Number Variants (CNVs) and balanced lesions. This is another way of performing genetic analysis. Epic Sciences is developing methods for enumeration and molecular characterization of circulating tumor cells (CTCs). Quantitation of rare cells in blood can be useful in routine lab analysis as well as in situations where biopsy tissue is unavailable or is of poor quality. David Nelson of Epic showed data comparing CTCs (cytokeratin+ cells) detection by Epic’s method to the gold standard, Cell Search assay; Epic’s method was three logs more sensitive than the Cell Search assay. Illumina, the pioneer in San Diego diagnostic arena, is at the forefront of nextgen sequencing technology. Their benchtop sequencers are affordable and have brought the cost of complete genome sequencing to a few thousands. Complete genome sequencing is expected to revolutionize personalized medicine (PM) in coming years and radically change health management paradigm of an individual from birth to death. There are many more San Diego diagnostic companies who were not present. These include GenProbe and Life Technologies.
There was also an interesting presentation by Laura Shawver, the founder of Clearity Foundation. Clearity is an ovarian cancer advocacy group that helps ovarian cancer patients (most of these are late stage or recurrent cases) obtain a "tumor profile" based on a panel of biomarkers developed by Clearity. The results are obtained as an "H-score" from diagnostic testing firms (such as Clarient) which help guide treatment decisions. One future goal of Clearity is to bring such CDx and pharmacogenomic (PGx) testing to early stage ovarian cancer.
(Clarity's Ovarian Cancer Test Panel, wwwLink)
Current state of CDx and diagnostics (Lecture notes):
- There are only three FDA approved tests in market today: HercepTest (Her2 for Herceptin), K-RAS mutation kits (for Erbitux and Vectibix), and OncotypeDx (an array of 21 genes to predict breast cancer occurrence.)
- Most currently used biomarkers are good in excluding patients (and not including,) for example, Her2+ tumors are candidates for Herceptin treatment, but EGFR mutations in lung cancer exclude Erbitux or Vectibix (both require WT allele for anti-tumor activity.) Furthermore, these biomarkers may or may not be prognostic. Her2 status has marginal predictive value in breast cancer, whereas, falling levels of EGFR mutations or EML4/ALK fusion in lung adenocarcinoma are good prognostic predictors.
- Target may not be the best prognostic predictor. For example, estrogen receptor (ER) is the target of tamoxifen; however, it is the progesterone receptor (PR) status and not ER which has predictive value.
- Presence or absence of a mutation alone is not sufficient for CDx. For example, non-small cell lung cancer(NSCLC) tumors with wild-type K-RAS, but not with G12V mutation, respond to EGFR-targeting drugs. However, tumors with G13D also respond as well as those with wild-type K-RAS. Reliance on WT status only would have excluded about 8% of the NSCLC patients with G13D genotype from the benefit of EGFR-targeting drugs.
- CDx markers help in further expanding a pathologist’s definition of a cancer type. Case in point is lung adenocarcinoma. It is not a single disease but is better defined by its molecular drivers. Different EGFR mutations respond differently to anti-EGFR MAbs or EGFR-TKIs. Three major groups of mutations have been described, which are in the codons 19, 20 and 21. Tumors with mutations in codon 19 are sensitive, whereas those in codon 21 are resistant to EGFR therapy (IPASS clinical trial.) Armed with this type of molecular Dx data, the role of today’s pathologist is changing from providing tumor classification information to prediction of response and helping oncologists in the selection of optimal drug or regimen.
- For pharma, incorporation of CDx very early in clinical trials makes a huge difference in the success of a trial. James Christensen, Director of Pfizer Oncology Translational Research talked about the discovery, preclinical and clinical development of crizotinib (PF2341066) for ALK-positive lung adenocarcinoma. He pointed out that by incorporating CDx (for EML4/ALK fusion) very early in trials, Pfizer was able to focus on a tiny percent of responders. This ALK-positive patient group (aka, Molecularly Enriched Cohort) does not appear to respond to EGFR-TKI and platinum-based chemotherapy. By performing trials exclusively with this Molecularly Enriched Cohort, Pfizer was able to develop an option for this unmet need. [NEJM report and comments]
- Economics of CDx development. What happens if the companion drug fails to jump through the FDA hoops; all CDx efforts will also be lost.
- Several issues remain regarding methods. How do we compare FISH, IHC, PCR, or other methods for testing a biomarker. FISH has the advantage of testing at the cellular level; IHC allows assignment of a score, e.g., 0-300%, could be represented in a simple 0 to 3+ score. Both FISH and IHC are dependent on good validated antibodies. In general, if a right antibody is available, there is good concordance between FISH and IHC (e.g., ALK-fusion assays.) PCR methods (RT-PCR or multiplex) are sensitive to sample quality.
- What is the best sample? Tissue blocks are considered more desirable than tissue slides. All assay antibodies must be validated on FFPE. Another question is, what should be done if no biopsy specimen are available.
- Ownership and patent issues surrounding biomarkers or methods may be stumbling blocks in the development of these assays.
- Diagnostic community is also aware of current hurdles when dealing with FDA. Dx are handled as medical devices requiring separate clinical trials, cost of which is not always justifiable. There is a move to change this. But, as of now, Erbitux or Vectibix cannot be prescribed before testing for the presence of wild-type EGFR genotype; however, there is no FDA-cleared test.
- Finally, patients and people involved in oncology care must be aware of the variability between various labs. PTEN is the poster child for this - only one or two good antibody are known to work well; thus, different diagnostic lab come up with different results.
John Glassco of Clarient in his talk showed as slide with a quote from the 19th century Canadian physician Sir William Osler (1892) - “The practice of medicine is an art, not a trade; a calling, not a business; a calling in which your heart will be exercised equally with your head. Often the best part of your work will have nothing to do with potions and powders, but with the exercise of an influence of the strong upon the weak, of the righteous upon the wicked, of the wise upon the foolish.” In near future, the development of CDx, PGx and other PM tools will help make the medical decision process in cancer care more objective rather than in-part subjective.
Oncology Focused Pharmacogenomic "predictive" Biomarkers. February 6, 2011