The NIH defines precision medicine as “an emerging approach for disease treatment and prevention that takes into account individual variability in genes, environment, and lifestyle for each person1.” In cancer patients, we can rephrase the definition to “through detailed understanding of a cancer’s biology, providing the right drug, for the right patient, at the right time.”
In order to identify the correct drug, biomarkers are used to identify patients that can be treated with the appropriate therapy for their cancer. The FDA defines biomarkers as “a defined characteristic that is measured as an indicator of normal biological processes, pathogenic processes, or responses to an exposure or intervention, including therapeutic interventions2.” Great strides have been made in the discovery and validation of biomarkers in drug development.
Two of the most common biomarkers in clinical trials are predictive and pharmacodynamic biomarkers, critical for the drug development process. Predictive biomarkers are those that will identify the patients most likely to respond or to be resistant to the drug. An example of a predictive sensitive biomarker is the staining of breast or gastric cancers for HER2 using immunohistochemistry (IHC). Patients with a breast tumor that over-expresses the HER2 protein (a score of 2+ or 3+) will be considered for treatment with Herceptin®. Resistant biomarkers are also important. For example, Panitumumab (an EGFR antibody approved in colorectal cancer) is contraindicated if the person’s tumor tissue has a Kras mutation3. These types of biomarkers are selected most often by drug companies in the early phases of development. Millions of dollars are spent using animal and cell models to identify and initially characterize such biomarkers. Subsequently, if they are validated in clinical trials, these biomarkers could potentially become a companion diagnostic for the drug in development. These biomarkers are very useful for drug companies since they help to develop the drug faster in the right group of patients by keeping both safety and efficacy of the compound high.
Finding predictive biomarkers that can be developed into companion diagnostics is important in precision medicine, but is not an easy task to accomplish. Two things need to happen simultaneously: the correct biomarker must be found and the associated drug must get approved. However, drug approval is not easy. Of all the drugs that enter clinical testing, only about 10 percent will receive approval from federal agencies. In addition, there are thousands of biomarkers that are tested, but only a few enter clinical trials4, and fewer become companion diagnostics. A total of only 32 companion diagnostics have been approved by the FDA for 16 drugs, or roughly only 9 percent of all approved drugs since 1988 have a companion diagnostic (figure 1)5.These numbers clearly indicate the difficulty of finding and validating biomarkers that result in a companion diagnostic. Yet, finding them can help save hundreds of millions of dollars for a company that is developing a drug.
Figure 1.

Pharmacodynamic biomarkers are those markers that change upon compound treatment. They are important because they provide clues on the mechanism of action (MOA) of the molecule. Pharmacodynamic biomarkers are very useful in the early stages of human testing since they can validate the proposed pathway of activity of the drug. These are most often discovered in pre-clinical development by looking at the pathway that the drug is inhibiting, or by relaying in a secondary effect observed in the models tested. While pharmacodynamics biomarkers are heavily used in pre-clinical and Phase I clinical trials, their use falls dramatically in later stages of the drug development process. These biomarkers tend not to be used as companion diagnostics.
Applying precision medicine to immuno-oncology (IO) is not a straightforward application of the methods used in the past. Several factors need to be considered since the patient’s immune system is being treated to act on the cancer. Therefore, both the immune system and tumor need to be examined. When developing pharmacodynamic biomarkers, researchers may look at changes in the overall number of immune cells, proliferation, or activation status of specific immune cells as an indicator of drug activity. One example is the presence of all lymphocytic counts in peripheral blood of patients treated with Yervoy®, a CTLA-4 inhibitor. Patients treated with increasing amounts of the inhibitor were found to have higher levels of ALC, suggesting that the drug had the desired effect of down regulating the activity of regulatory T cells, and thus observing more lymphocytes in circulation (figure 2)7.,)
Figure 2.
Also, one can look at the overall number and clonality of T cells by sequencing the T cell receptor (TCR) of those cells associated with the tumor8. The clonality of the T cells associated with the tumor is an indicator of the immune response to the tumor. In addition to providing clues to the exact mechanism of action, these biomarkers can also point to the therapeutic choice for the appropriate IO drugs. For example, if a patient shows that the immune system can recognize his or her own tumor, then using a check point inhibitor, like PD-1 or CTLA-4 inhibitors may be an appropriate course to follow. On the other hand, if the immune system has not recognized the tumor, then stimulation with a vaccine may be the appropriate therapy to follow. Biomarkers can help figure out the mechanism used by the immune system to eradicate the tumor and give clues to what type of medication will be more appropriate for each patient.
Predictive biomarkers are also found in IO, especially those associated with the tumor itself. Several methods have been used to discover them: pathological staining to look at the type of cells present in the tumor, including immune cells; comparing the ratio of effector T cells vs T regulatory cells; and the presence of immune cells in the leading part as compared to those outside of the tumor. Interestingly, one biomarker has already been developed as a companion diagnostic in IO, PD-L1 IHC. Patients that are positive for this marker can be treated with Keytruda (pembrolizumab) a PD-1 inhibitor (figure 3)9, thus providing the first example that precision medicine can help increase the response in immunotherapy. One word of caution, while the selection of patients increases the overall response rate, there are about 10-25 percent of patients that are low expressers or negative for this test and are able to respond to the drug (compare green and orange lines with the black line of patients with higher than 50 percent of staining levels). More work is needed with this drug to increase patient response rates.
The ongoing advances in cancer immunotherapy together with precision medicine may promise a bright future for patients. Immunotherapy is starting to prolong the life of some patients up to 10 years10. New drugs to treat different parts of the immune system are being developed, from novel checkpoint inhibitors to vaccines to peptides and neo-antigens. By merging these two approaches, immunotherapy and precision medicine, the goal is to increase the percentage of patients that respond to the drugs, with hope for even more durable and lasting responses. While optimistic, it is not unreasonable to envision a world where cancer can be relegated to a chronic condition or even be eliminated.
Suso Platero, Ph.D., is Executive Director, Precision Medicine Oncology Leader, Clinical Development Services.
1 https://www.nih.gov/precision-medicine-initiative-cohort-program
3 Amado et al., J Clin Oncol., 2008
4 Kern, S Cancer Research, 2012
5 Dracopoli et al, Trends in Pharmacological Sciences 2017; 38: 41-54
6 Wolchok et al., The Lancet Oncology 2010; 11: Issue 2,155–64
7 Robins et al., Science Translational Medicine, December 2013
8 Garon E et al., N Engl J Med 2015; 372:2018-2028
9 Schadendorf et al., JCO doi:10.1200/JCO.2014.56.2736