Precision Medicine Carries Significant Challenges but Will (Eventually) Lower Cancer Care Costs and Is the Way to Go (Precisely)

Andre Goy, MD

Editor-in-Chief of Oncology & Biotech News

Chairman and Director Chief of Lymphoma Director, Clinical and Translational Cancer Research Chief Science Officer and Director of Research and Innovation Regional Cancer Care Associates John Theurer Cancer Center at Hackensack University Medical Center

No one would deny that cancer care and healthcare at large are at a crossroads and need a change. Sky-high medication prices are emerging as a political issue. As I have discussed in previous columns, some of the novel cancer therapies are truly transformative and might replace chemotherapy in the future, though they have to be taken for years (until progression), while costing patients up to several thousands dollars a month. On the other hand, most approved drugs only work in a fraction of patients, with up to 30% to 70% of patients who do not benefit from them. This has huge implications for patients’ lives first, but also in terms of cost. Meanwhile, the biotech field continues with more promising new tools, drugs, and speculation. Lots—if not most—of these novel therapies result from a better understanding of cancer cell biology, which has also shed light on the enormous molecular diversity of cancer.

Personalized therapy, renamed precision or precise medicine (PM), has emerged as the answer to help improve patients outcomes. Thanks to high throughput technology, including genomic testing of each tumor, and each patient’s inherited DNA— along with proteomics in the future—oncologists will be able to tailor regimens for each case.

The field of emerging liquid biopsies will allow us to monitor patients’ responses in vivo and detect disease and/or relapse earlier. Having the ability to adjust therapy to a given patient is not new and is already being used in many ways including in oncology. For decades, detection of ER (estrogen receptors) and/ or PR (progesterone receptors) in breast cancer served as basis for hormone therapy and prognostic evaluation. Nowadays, a growing panel of defined mutations helps choose a TKI (tyrosine kinase inhibitor) in CML or in lung cancer, for example.

On the other hand, critics argue PM has become just a “buzzword” and we should keep with the statue quo and continue the current drug development paradigm. To their point, two recent papers reporting on “picking a treatment” based on molecular features were rather disappointing. In the first one, the BASKET trial (Lopez- Chavez A, et al. J Clin Oncol. 2015;33[9]:1000-1007) in lung cancer, the authors concluded that their trial was not feasible for many of the study arms that had patients with rare mutations. The second one— the SHIVA trial (Le Tourneau C, et al. Lancet Oncol. 2015;16[13]:1324- 1334.)—looked at 741 patients with advanced refractory cancer, who were given either a molecular targeted agent outside the known approved indication (based on molecular profiling) versus investigators’ choice and showed no benefit in outcome.

Skeptics claim PM has not delivered overall and point to a tendency of both public and care providers to imbue emerging innovation (not only DNA sequencing but also other new technologies) with almost mystical powers. The only way to escape the expectation/disillusionment cycle is for genomics and PM to demonstrate a tangible impact. This will happen when sequencing is embraced not because it’s interesting or “edgy,” but because it’s so obviously useful.

When enough patients will be sequenced, more reliable signatures and patterns will likely emerge (as opposed to statistical noise). Biomarker-based trials need to be embraced and supported both by pharma but also by regulatory authorities. Too often, such efforts toward correlative science in early trials are either simply ignored, or not supported based on extra cost or risk (additional procedures) and too frequently hence recommended as optional by local IRBs. Finally, mutation-based trials are frequently slow in turn around leading to difficult enrollment and/or biased (easiest cases) populations.

That being said, the cost of cancer care in the United States is estimated at over $200 billion a year and will continue to rise as our population expands and lives longer (with a projected expected increase of over 40% in cancer incidence in the next 30 years). Global oncology statistics also show that drug-related expenses surpassed $100 billion in 2014. Those are huge numbers, without even accounting for the hidden cost of lost productive years, in addition to the emotional toll on friends and family.

Because a high proportion of these costs are incurred at the end of life and/or have no real impact on outcome, some believe that palliative care with hospice should be an option offered much earlier for many patients…This is debatable for several reasons, including the impact we see with some novel therapies at this point, particularly in immuno-oncology.

The problem is, we are not there yet in PM and more education and changes of behavior are needed for all stakeholders. This sounds simple but is definitely not easy. Even in cases with strong evidence-based medicine, changes occur slowly for multiple reasons including human nature among other things.

Paying for a test that in lung or breast cancer might spare a patient chemotherapy is obvious but is still not done routinely or in some cases, it’s done but not subsequently used. Current pricing of molecular testing is still very high (several thousands to have one or maybe two “actionable” items), but this will go down as competition increases and testing is used more frequently. However, it is likely too simplistic to think—as illustrated by the two papers cited above—that one mutation will really help in practice given the diversity and complexity of cancer cells.

The often breathless excitement around the promise of PM has generated a pointed response from skeptics, who appropriately ask whether the science has been oversold. I would argue that the optimal treatment of cancer should and hopefully will become, like in HIV, a rationale combination of targeted agents based on “actionable” profiles of tumors. This is obviously every oncologist’s dream and I think that only then will we make true progress. This also implies dramatic changes in drug development including codevelopment of drugs early on.

PM is and will be good for everyone. It already can and will help cancer patients by sparing them the toxicity—and time—of trying ineffective drugs. It will cost much more before it costs less. However, the extra short-term cost should be put in perspective with the huge waste seen in oncology, particularly later in the disease course with no impact on outcome. The rapidly changing landscape in the pipeline calls for new ways to develop trials and speed up approval process, which in itself might reduce drugs costs. The integration of big data in this revolution is essential, although to not fall again into the hype, digital data needs to be meaningful and reflect the patterns seen in the real world. It also needs to engage all stakeholders, including patients that need to be part of monitoring their clinical benefit (which will also help compliance for mostly oral drugs at this point).

Our goal should be to use PM as a way to deliver value-based care. This critical endeavor needs to be embraced by everyone in oncology, while impassionate PM champions keep the fire going…