Goy Taps Into the True Potential of Precision Medicine in Oncology

Current approaches to precision medicine in oncology have been fruitful, but require better integration and utilization of available resources to inform sustainable and effective drug development and clinical care, according to Andre Goy, MD.

“We want to support smarter medicine. If we start to align data [on] patient adherence, treatment compliance, quality of life, genomics, optimization of the point of care, the best option, and the best sequence of care, that will be the ecosystem of precision medicine,” said Goy. “Once we start to do this, we will get the proper momentum for precision medicine because we will rationalize care and reduce waste.”

In an interview with OncLive® during the 3rd Annual Precision Medicine Symposium, an event hosted by Physicians’ Education Resource®, LLC, Goy, chair and physician in chief of the Hackensack Meridian Health Oncology Care Transformation Services, chairman and chief physician officer, and Lymphoma Division Chief, at John Theurer Cancer Center at Hackensack University Medical Center, discussed the evolution of precision medicine in oncology and the steps that should be taken to drive future progress.

OncLive®: How would you define precision medicine? Is there a standardized definition?

Goy: The first thing to do is to make sure that we agree on the definition of precision medicine. The misconception in some ways is that personalized medicine is a treatment that is tailored and customized to everyone. What we are trying to do by integrating all the information of the progress of molecular medicine and other sorts of data is to refine the decision to give the best option to a patient. We have to really organize all the progress and knowledge and apply it as soon as possible to refine treatment decisions and develop ways to use modern medicine because that’s the best way to derive the longest mileage for a patient.

What is interesting is that precision medicine has really gained momentum in the public, and also, because of the progress in medicine that is unprecedented over the past 2 or 3 decades from the Human Genome Project, a better understanding of cancer and the immune system— particularly the diversity of cancer, 1000s of novel therapies in the oncology pipeline, and 50 to 60 drugs approved by the FDA every year––drugs that are game changing, many of which now also have a companion diagnostic test. Also, big data, and access to more diagnostics with next-generation sequencing [NGS]. All of these [things] create the illusion that you can have 1 gene and 1 test, and then you have the magic treatment. It’s a little more complicated than that.

The expectations have perhaps been too high in some ways because you can see the emergence of skepticism now [implying that] precision medicine has not delivered. We’ve heard about the Human Genome Project impact for 20 years, and patients are still waiting.

In what ways have current approaches to precision medicine failed?

There was a large study called the SHIVA trial [NCT01771458] thatlooked at 800 patients and randomized them to the standard of care decided by the physician, or 1 of 3 targeted therapies outside of their standard indication to try to take advantage of 1 biomarker to give a better treatment. [However,] the trial was negative, which is not surprising. The idea to simplify a treatment decision just based on 1 biomarker, 1 gene, and 1 treatment is very simplistic.

Although, we do have models like this. If you look at chronic myeloid leukemia, which is a rare form of chronic leukemia, this is a perfect example of a targeted therapy and precision medicine––an ideal marriage. BCR-ABL is the target [for which] we have imatinib [Gleevec], which became the first TKI. Then we understood the second-, third-, and fourth-generation TKIs and their resistance profiles over time, and we could match a treatment based on the genotype of the tumor and the characteristics of the tumor. This has been a huge success. There are very few cancers that have a predominant driver that you can control like that. We have made a lot of progress though. We could not have had a conference dedicated to precision medicine 10 years ago.

At last year’s meeting, John Marshall, MD, of Georgetown-Lombardi Comprehensive Cancer Center had a great slide comparing cancer 10 years ago and now and all the changes in cancer being monoclonal or more polyclonal, immunotherapy, NGS, diagnostic capabilities, targeted therapy––a lot of new things have happened that are really changing and accelerating the field of oncology.

Why is precision medicine so important to pursue?

Although there’s clear momentum, with huge investment from companies, and a global precision medicine market driven by oncology that is very impressive across the world with double digit growth annually, it’s an important [question] to take into consideration because: How do we reconcile these efforts? We have all the tools in our ecosystem. How do we make this better?

We have made some progress. If you take aggressive lymphoma, diffuse large B-cell lymphoma, the most common lymphoma with 30,000 to 40,000 new cases a year, we cure more than half of patients with just 6 cycles of R-CHOP. However, there are still a lot of patients who don’t do well, and most relapses occur early. The [truth is that], with 6 cycles of R-CHOP, if a patient has early-stage, localized disease without high-risk features, they are going to do really well. There’s even an opportunity in those patients to reduce the number of treatment cycles.

All patients with a high-risk International Prognostic Index, intermediate- and high-risk features, high-risk molecular features, a non–germinal center B-cell–like subtype, double hit and double expressors, all these patients do very poorly. Some of them can be salvaged now with CAR T cells. Autologous transplant was not a great salvage to chemoimmunotherapy until we had CAR T cells. The goal is to treat these patients the best the first time around.

The ability that we have now is to really stop driving a car by looking at the back mirror. We need to look at this prospectively. We need to look at this at the point of care for patients. We could not drive a car [going backwards]. In medicine, it applies as well. We have now plenty of opportunity. NGS is becoming accessible for routine diagnostics to augment standard pathology to make sure we leapfrog into the future where we have access to enough mutation profiles that allows you to really refine a subset of patients by their similarities.

Once you make this decision, you’re able to sort out a double expressor in large cell lymphoma, for example. You need to go with the real gold standard of classifying these patients to rationalize your decision and improve the chance of the outcome.

We have the opportunity now in large cell lymphoma and other cancers to integrate molecular features, pathology, relevant electronic medical record information, and imaging to really try to optimize how we follow each patient, particularly regarding minimal residual disease [MRD].

[MRD raises the] concept of the dynamic biomarker. Patients who are PET negative and negative by cell-free DNA [cfDNA], the latter ushering in the concept of liquid biopsy, have deep responses, [indicating that] they’re going to do really well. This is where the future is going. If you take an example outside of large cell lymphoma like lung cancer, this is a perfect example. It was one size fits all, and now we have 14 or more mutations that matter because we have an actionable novel therapy for them.

The most well-known mutations are EGFR and ALK, which have been around more than 10 years, and for which we have long term follow-up on. The standard of care for those patients who have those mutations is not to give chemotherapy. [These targeted therapies have] a huge impact as we combine these smaller molecules in the subset of lung cancer, [and if we perhaps] combine it with immunotherapy, we’ve suddenly redefined the landscape on how to treat a microsubset of lung cancer.

We’re also starting to see some patients we can cure with advanced-stage lung cancer, which was not heard of. That’s another form of doing precision medicine. Patients who have localized, early-stage lung cancer, there are data that show, not surprisingly, that many of these patients have cfDNA detectable in the blood. If you give them a targeted therapy in the adjuvant setting, you prevent recurrence. This is really important because this technology does not just apply at the time of diagnosis.

The diagnostic capabilities that we have now allow us to refine our decision from the precancer setting––early detection, potential intervention, genetic risk management––to a post-cancer setting––MRD management, potentially immunoconsolidation and early detection––to cancer care. The two biggest questions that patients want to know, and we need to know as a society is what my best option now is. For us, we need to know what the best sequence of care is, given the complexity of the landscape.

What can be done to realize the full potential of precision medicine in oncology?

We looked at 2000 patients with non–small cell lung cancer [NSCLC], and we looked at the EGFR and ALK mutation profiling. Between 40% to 60% of patients in a real-world setting did not get molecular testing. If you had molecular testing and had a mutation, the median survival was almost 48 months. If you did not have molecular testing and had chemotherapy as a default treatment, the median survival was 10 months, so for patients, [molecular testing] is critical.

Additionally, the cost in 4 years was less than 10 months of survival in managing a patient with NSCLC with chemotherapy because of complications and admissions to the hospital, etc. You might say that if you use a novel therapy, it is going to be more expensive, but cost is relative in medicine. You want to do what’s going to lead to the most mileage for a patient. This is important because one-third of the $3 trillion a year spent in health care in the United States is considered a waste. It comes from redundancies from procedures that are not helpful. In oncology, over one-third of what we do has no impact on outcomes. Most drugs get approved with a response rate of 30% to 70%, which is fantastic, but it also means that these drugs don’t work in 30% to 70% of patients. Our role is to understand, with real-world data, what the real benefit of those drugs is and how we select the patient who can benefit from “X, Y, and Z drugs” and stop just using a drug that is approved under a label.

This whole dynamic evolution of the precision of medicine and rational therapeutics is very important. We should also apply this in drug development. We must do small trials that are molecularly driven by subtype or drive trials that are open, where we monitor the signatures of these patients to try to understand where the signal [of efficacy] is.

Then, we enrich the next subset of patients into that signature. We can’t continue to do drug development the way we do it because it doesn’t work. We don’t have enough patients. There are 10,000 new drugs in the pipeline of medicine. We need to be smarter so that we do not waste time. We’re going to make trials that are very short. Small trials get a conditional approval based on the companion diagnostic and based on supplementation with real-world data to see where the actual benefit is.

I am optimistic that this is the solution to the crisis in health care now. I would push it further by saying this not only applies in cancer but in the prevention of cancer, precision health, and precision survivorship. I’m very excited to hear more about the progress in specific tumors as we dive into this conference, where we can see the unprecedented progress and unexpected progress, with so many sub-classifications of diseases into smaller and smaller buckets. Eventually, we will be able to identify [more] molecular subtypes of cancer and [more] drugs will be approved based on the molecular marker well beyond the organ of origin [than already are].