For all the talk about personalized medicine, in most cases the tools we have in oncology are more often disease site–focused than personalized. Each tumor type has a list of potentially active agents. For example, in kidney cancer, I can talk to my patients about sunitinib (Sutent), pazopanib (Votrient), bevacizumab (Avastin) and interferon, sorafenib (Nexavar), temsirolimus (Torisel), and everolimus (Afinitor).
However, I cannot tell an individual patient which agent has the best probability of working for him or her. And when I see patients who have unexpected outcomes, I am often left wondering, what makes this person different? Why did he/she have such a dramatic response? Or, why has the disease progressed so quickly?
In a perfect world, we would submit a piece of tumor tissue or a blood sample to the laboratory and receive a report that outlines the best treatment for that specific patient. Rather than thinking of a patient’s tumor as lung cancer or breast cancer, we would think about tumors in terms of their molecular markers and their potential for response to a particular targeted therapy.
This would allow us to improve the value of the targeted therapy. Patients would only be treated with medications that were expected to help them, and clinicians would not waste valuable time treating them with agents that have no benefit and that can cause toxicity (both physical and financial).
Notable Examples Two notable examples illustrate how we can achieve our goal of practicing personalized medicine. The Oncotype DX assay measures the expression of 21 genes to calculate a recurrence score for women with early-stage breast cancer, allowing oncologists to identify who is most likely to experience a recurrence and who is most likely to benefit from chemotherapy, in an effort to reduce the risk of cancer recurrence. The assay is also a valuable tool in identifying women who are destined to do well without chemotherapy and can be spared its toxicity.
In patients with advanced colorectal cancer, KRAS testing has helped to identify patients with advanced colorectal cancer who are unlikely to respond to agents targeting the epidermal growth factor receptor, such as panitumumab (Vectibix) or cetuximab (Erbitux).
What the Oncotype DX assay and KRAS testing have in common is that their clinical utility was established using banked specimens from clinical trials. When used in the appropriate patient population, we now know how to interpret these test results, and how to use the information to guide clinical practice, by identifying appropriate candidates for therapy.
Novel molecular tests have most value when validated data demonstrate that they improve outcomes through better patient selection compared with established clinical criteria.
The Value of Clinical Trials To optimize the potential value of personalized medicine, physicians and payers need to encourage patients to enroll in clinical trials. Currently only 5% of American patients enroll in clinical trials; this slow pace delays the completion of studies and the translation of new knowledge into practice.
Providers should encourage eligible patients to explore clinical trials early in their disease course, instead of as a last-resort option. Payers should eliminate barriers to clinical trial coverage. Otherwise, countless patients will continue to be treated with standard therapy.
When treatment is given outside of clinical trials, no primary data are collected or specimens banked. This information is critical to the development of new molecular tests to further personalize medicine and improve the value of treatment for future patients
