It is widely thought that companion diagnostic (CDx) tests could improve the process of developing oncology drugs and become truly useful tools for guiding the correct choice of treatment for the individual patient. In practice, personalized medicine commonly uses measurements of an individual patient’s biomarker profile to determine both the likely outcome of the disease (prognostic), as well as the likelihood of response to certain therapies (predictive). The biomarkers measured are frequently genetic mutations, including EGFR mutations for tyrosine kinase inhibitors (TKI) in non-small cell lung cancer (NSCLC) and KRAS mutations for Epidermal Growth Factor Receptor (EGFR)-antibodies to treat colon cancer.

Biomarkers can also come in the form of expression of specific proteins in the cancer tissue, such as the expression of human epidermal growth factor receptor (HER2), estrogen receptor (ER), and the progesterone receptor (PR) to indicate the likelihood of response or recurrence of breast cancer. Biomarker tests can become quite multivariate and elaborate, such as the Oncotype Dx (Genomic Health) and MammaPrint (Agendia) assays which each profile the expression of a large number of gene sets. The IHC4 biomarker assay utilizes the protein expression of four different proteins in breast cancer. Despite the enormous increases in information regarding the molecular characteristics from each patients using some of these tests, the predictive value remains generally unimpressive, and it seems to remain a mystery as to why certain patients respond to drugs and others do not. In part, this may be due to the fact that these biomarkers are correlative but not causal to drug mechanism. Additionally, with each additional biomarker, there is an increase in the noise of the readout, and the predictive benefit of each biomarker must be very significant. Simply increasing the number of biomarkers will not by itself lead to a more predictive test, which has been highlighted by the overall success of the field.

Within the leukemia field, a variety of biomarkers have been identified including: FLT3, IDH, and p53 mutations for acute myeloid leukemia (AML) and IGHV and p53 mutations for chronic lymphocytic leukemia (CLL). Like the biomarkers in the solid tumor area, the predictive value of these tests requires additional sensitivity. While not commonly used, direct measurement of responses in patient cancer cells ex vivo can also provide predictive value. At Eutropics, we have identified a functional assay that measures the response of leukemia cells to the apoptosis signaling pathway. While not realized, we hypothesized that the mechanism of many common drugs and the patient’s response to those agents depends on the ability of the cancer cells to respond to pro-apoptotic signaling, such that if a cell is unresponsive to these death signals, it will not respond to certain therapeutics.

Our test, Praedicare Dx, measures the functionality of the Bcl-2 family proteins in patient leukemia cells. If the test shows that the cancer cells undergo apoptosis in response to pro-death agents, then that patient is likely to respond to specific pharmacologic agents. These "priming" values have been proven to correlate with patient response to clinical treatment (predictive) and their outcome (prognostic). Thus far, we have identified medical utility of Praedicare Dx in AML 6 and CLL 7. Importantly, the sensitivity of the assay in AML exceeds the best predictive profile that can be obtained using nine separate established prognostic markers (including complex genetic and cytogenetic tests) using a single functional assay, highlighting the central importance of apoptotic signaling for the successful action of therapeutic agents.37 Praedicare Dx is CLIA certified and it is now being used to guide patient selection into AML clinical trials.

The FDA seems poised to require the marriage of the biomarkers to oncology drugs on a larger scale as indicated in their guidance on Cdx issued in July 2014. We believe that extending Praedicare Dx as a CDx test will aid in the goal of personalized medicine. The FDA has an open mind as to what form CDx tests will take, as long as the evidence supports the use of the biomarker test, without increasing the risks to patients. The robust correlations of Praedicare Dx to patient response will not only enable the best treatment for each patient; it will help speed drug development and clinical trials.

 

References

1. Gonzales DeCastro, D, Clarke,P, Al-Lazikani, B, Workman, P (2013) Personalized cancer Medicine:Moleculr Diagnostics, Predicitive Biomarkers, and Drug reisitance Clinical Pharmacolgy and Therpeutics:93:3;252-159.
2. Zoon, CK, Starker, EQ et al (2009)Current molecular diagnostics of breast cancer and the potential incorporation of microRNA Expert Rev Mol Diagn. 9(5): 455-467.
3. Deng J. et al. (2007) BH3 profiling identifies three distinct classes of apoptotic blocks to predict response to ABT-737 and conventional chemotherapeutic agents. Cancer Cell. 12(2):171-85
4. Letai A. (2008) Diagnosing and exploiting cancer’s addiction to blocks in apoptosis. Nat Rev Cancer;8:121-32.
5. Vo TT, Ryan J, Carrasco R, Neuberg D, Rossi DJ, Stone RM, et al. (2012) Relative mitochondrial priming of myeloblasts and normal HSCs determines chemotherapeutic success in AML. Cell;151:344-55.
6. Pierceall WE, Kornblau SM, Carlson NE, Huang X, Blake N, Lena R, Elashoff M, Konopleva M, Cardone MH, Andreeff M. (2013) BH3 Profiling Discriminates Response to Cytarabine-based Treatment of Acute Myeloid Leukemia. Mol Cancer Ther. 12(12):2940-9
7. Pierceall, WE , Warner, SL, RJ Lena RJ, CDoykan,C, Blake, N, Elashoff,M, Von Hoff,D , Bearss,DJ, Cardone, MH, Andritsos,L , Byrd, JC, Lanasa,MC, Grever,MR, Johnson, AJ, (2014) Mitochondrial priming of chronic lymphocytic leukemia patients associates Bcl-xL dependence with alvocidib response Leukemia 28(11):2251-4
8. Pierceall, W. E., R. J. Lena, B. C. Medeiros, N. Blake, C. Doykan, M. Elashoff, et al. 2014. Mcl-1-dependence predicts response to vorinostat and gemtuzumab ozogamicin in acute myeloid leukemia. Leukemia Res, 38(5):564-8.