Personalized Care for Ovarian Cancer Dr. Parvin Tajik

In pursuit of new therapies and personalized treatments, a large number of studies over the past two decades has focused on biomarkers for ovarian cancer. These are clinical characteristics associated with the disease that could indicate which treatment will be most effective for an individual patient. Yet, despite all this effort, the survival rate has not much improved in the last decades. In contrast, breast cancer research has made great strides in successfully using biomarkers to predict outcomes and improve treatments. So, why has progress on ovarian cancer stalled?
Dr. Parvin Tajik, an AXA Research Fund fellow at the University of Amsterdam Academic Medical Center, is working to change this situation, to help medicine move past the roadblock and provide doctors and patients with the necessary tools to make more informed and personalized decisions about cancer treatment.

Approaching a Multifaceted Disease from Multiple Angles

The field of ovarian cancer research and treatment is coming around to the idea that this is not just a single disease, but a heterogeneous condition. By simultaneously exploring multiple aspects with an interdisciplinary team of gynecologists, oncologists, epidemiologists, biostatisticians and pathologists, she is painting a more complete picture of ovarian cancer. Dr. Tajik’s strategy relies, first, on the power of statistics. She is developing methods to assess a host of biomarkers for their ability to predict the benefit and harm for individual patients of the treatment alternatives available to them. Any number of characteristics have the potential to serve as a useful marker, like patient age; tumor size and stage of the disease at diagnosis; presence of genetic changes in the tumor (like the BRCA mutation, well-known in breast cancer and also involved in ovarian cancer); the presence of certain proteins in tumor cells or in the blood; and more.

The multi-marker assessment system developed by the researchers is a statistical program in which the user defines different biomarkers and treatment options to consider. Taking advantage of pre-existing data from large (e.g. 800-patient) clinical trials and the statistical power these numbers bring, her method is able to seek out correlations between the markers and the eventual outcome of patients: people with these clinical characteristics, receiving this treatment, most often sees this much benefit—or, on the contrary, these undesirable side effects. The model can then be used to predict the outcome for future patients.

Meaningful Markers

Dr. Tajik’s team in Amsterdam has performed a systematic review of research published on biomarkers in ovarian cancer over the last five years, which yielded a full 4,000 abstracts. Yet, despite all this research, only two biomarkers have been approved for use by the U.S. Food and Drug Administration in the last decade. In collaboration with researchers in Paris and Toronto, they are also trying to understand why most biomarker studies in the field of ovarian cancer have not been successful in translating findings into clinical applications.

After evaluating the entire process undertaken in these studies, from sample preparation to analysis, the results showed that most claims about markers in ovarian cancer are likely based on false discoveries (i.e. the markers identified can’t truly predict patients’ response to specific treatments), or markers with insufficient performance to be useful. This shows there are methodological problems that future biomarker studies need to address, in order to improve care for women with ovarian cancer, Dr. Tajik says. In the same vein, the researchers are compiling a list of promising ovarian cancer biomarkers—those with the most evidence backing them up—to be used in their comprehensive, multi-marker model.

Personalized Medicine: More Than a Patient’s Genes

Having created the tools to analyze the predictive power of any given biomarker and revealed the ones most likely to be meaningful, Dr. Tajik is now moving on to the prediction of different treatment outcomes of options for individual patients. While her methods will be relevant to any decisions around choosing a medical treatment, she remains focused on oncology. When a diagnosis of late-stage ovarian cancer arrives, both doctors and patients are faced with difficult decisions about which treatment or combination of treatments to use, without knowing the likely outcome of any option for a given individual. Dr. Tajik’s work has the potential to make their care more personalized, and not only by identifying genetic determinants for treatment selection.

The tools developed will put the patient’s diverse needs at the center, many of which are often overlooked. For example, patients have different perceptions of the burden of treatment. When cancer patients are debating surgery versus chemotherapy, some have a strong psychological need to get rid of the tumor and want to remove it before undergoing chemo. On the other hand, certain patients want to avoid the possible complications of surgery, at all costs. By showing them the objective benefits of a treatment—as assessed for their particular case by Dr. Tajik’s model—they can evaluate those relative to their perception, assess the trade-off, and feel as comfortable as possible with their decision. Personalized medicine should be about much more than one’s genes and Dr. Tajik’s tool should help keep it focused on benefits for patients and reducing the burden of treatments.

This tailoring of medical treatment should also improve the efficiency of the health care system, by identifying the most appropriate therapy before treatment even begins. Dr. Tajik’s tools could also aid decisions about whether a new drug can be marketed, or if it should carry the stipulation that the patient must exhibit a specific marker. Professional guidelines could be developed for doctors, as well, advising them on how to help patients choose the best treatment, in a shared decision-making scenario.

Dr. Tajik is pushing for a view of personalized medicine that is much broader than the hunt for genetic indicators. Her message is that it can and should apply to all treatment selection questions in medicine. This could mean mother/child health, for example, and choices about when to opt for a caesarian, when to induce labor, and more. For the research experience it enabled, Dr. Tajik credits the support of the AXA Research Fund with essentially defining her professional future, which will see her pursuing the goal of more patient-centered medical care. AXA’s backing has allowed her to take strides toward a better understanding of markers of ovarian cancer, and she intends to continue advancing down this path.