The Fix We Need to Break the Impasse in Precision Oncology

H. Jack West, MD


December 15, 2022

My patient with advanced lung cancer calls with new headaches. I schedule him for an expedited brain MRI followed by a clinic visit so we can review the results together. Between these steps, however, is a process that provides critical value: Hundreds of images are reviewed by an experienced neuro-radiologist who then provides me with an expert interpretation. Sometimes I can detect obvious large brain metastases, but often the images show tiny metastases of 1-3 mm or displaying an ambiguous, ragged appearance suggestive of leptomeningeal carcinomatosis.

The reports that accompany these images serve as a powerful reminder that I am not a neuro-radiologist and that I benefit profoundly from the support of someone with years of training and expertise.

This stands in stark contrast with how we handle next-generation sequencing (NGS) testing results, where I order the test and receive a lengthy report, accompanied by a bewildering list of genomic abnormalities and 50 or more pages of unhelpful boilerplate language that offers recommendations discordant with expert recommendations.

As a subspecialist in thoracic oncology, my focus is narrow enough that I am aware of all clinically relevant molecular markers for my lung cancer patients, but this may not be the case for my general oncology colleagues who may treat patients with a dozen different types of cancer every day. But even with my clinical expertise in lung cancer, every week brings one or more molecular testing results that require further investigation from an online molecular database, Google search, or potentially a colleague with a stronger background in precision oncology.

So, how are general oncologists handling this flood of genetic testing information?

Not surprisingly, the limited evidence we have indicates that they're struggling. For example, a recent study in JCO Precision Oncology showed that nearly 1 in 2 patients newly diagnosed with non–small cell lung cancer (NSCLC) weren't able to take advantage of precision medicine tools widely available today. From problems with the initial biopsy to a failure to act on genetic test results, these hurdles had real and potentially enormous consequences: Nearly 1 in 3 patients (29.2%) did not receive appropriate targeted treatments.

While molecular oncology has revolutionized our field over the past few decades, reviewing the execution of it in practice reveals a gauntlet of challenges. In addition to simply determining whether a patient should receive testing, there is a long, time-sensitive process from testing to deciding on and administering an appropriate treatment. This process involves collecting sufficient tissue or plasma, deciding whether to initiate treatment while waiting for the genetic test results, reviewing these results when they arrive, accurately identifying relevant targets and circumventing red herrings, and finally getting an expensive treatment covered and the patient started.

Each of these steps represents a distinct bottleneck that various committees, professional societies, healthcare organizations, and patient advocacy groups are diligently working to overcome.

But the problem remains: Many patients who receive molecular testing and have a finding that indicates an optimal, targeted therapy still don't receive it.

In a study of over 1200 patients with stage IIIB or IV NSCLC who received genetic testing, more than half with a detected driver mutation failed to receive the indicated targeted therapy. In addition, more than one third of patients with EGFR mutation- or ALK-positive NSCLC received immunotherapy before a tyrosine kinase inhibitor; this was not usually due to molecular test results being unavailable at the time that treatment was initiated.

Similarly, data from Flatiron Health revealed that among more than 3500 patients with advanced NSCLC who underwent NGS testing, only about two thirds with panels that showed an EGFR mutation or ALK rearrangement received the indicated targeted therapy, and only one third with less common markers designated in NCCN guidelines.

These patients completed testing and had an actionable finding, yet a huge fraction still didn't get the benefit of the right treatment because the test probably wasn't interpreted properly.

Improper interpretation represents a fundamental impasse that undermines the concept of molecular oncology. The prevailing mindset is that we need to educate oncologists to recognize and act on every relevant mutation, and that is where efforts have been directed thus far. But there are tens of thousands of oncologists in the world — well over 10,000 in the United States alone — and the volume of new information to keep up with is tremendous. Put simply, it outpaces the rate at which oncologists can learn it.

Moreover, the complexity of this information is staggering: Even dedicated specialists need to chase down esoteric NGS findings regularly. In fact, how to interpret these genetic reports warrants its own specialization.

Regardless of the level of complexity, we need an intervention. No matter how much effort is directed to educating oncologists, NGS interpretation by oncologists is a process that can never scale. This means that every new development in the field of molecular oncology will only serve to widen disparities between the best care achievable in an idealized world and what is actually accomplished with real-world constraints.

But this is a fixable problem.

Molecular tumor boards can serve as a mechanism for reviewing complex molecular panel results. In certain healthcare systems, these boards can potentially process hundreds, even thousands of cases per year. However, this option comes with some big limitations — most notably, this approach could never be scaled broadly to all practices. Molecular tumor boards also don't occur frequently enough or have the capacity to handle all the cases they would need to.

What we need is a strategy that mimics what happens when ordering a brain MRI: a timely, expert interpretation of the data condensed into an easily identifiable and readily actionable format. This format would highlight any targets with a compelling standard treatment, options that are not standard of care but are supported by clinical data, as well as "pertinent negative" results that highlight treatment paths to avoid on the basis of common mistakes in interpreting specific findings.

Implementing this strategy will require developing a narrow community of true experts, ideally a combination of molecular pathologists and subspecialists in specific cancer subtypes, to review every NGS report within a day or two, as an asynchronous real-time process. These specialists would dedicate a major fraction of their time to providing these interpretations that could be issued to every oncologist who orders these tests.

Would this service cost money? Yes, but this process would reclaim value for the NGS testing that is now widely practiced in an erratic way. This step could be included by certain molecular diagnostics companies to differentiate their high-value product from lower-value options.

Of note, this strategy could be offered to everyone, as it's not geographically limited. Like "nighthawk" radiology interpretation, the expertise can be delivered by any qualified service working from anywhere.

Currently, precision medicine is not living up to its potential. And that needs to change.

Implementing such infrastructure must be a universal expectation of the NGS testing process that, in the end, will help translate the promise of precision medicine into reliable benefits rather than an inefficient scattershot approach.

What do others think? Do people think this strategy is feasible?

H. Jack West, MD, is an associate professor at City of Hope Comprehensive Cancer Center in Duarte, California, and vice president of network strategy at AccessHope in Los Angeles. West serves as web editor for JAMA Oncology, edits and writes several sections on lung cancer for UpToDate, and leads a wide range of continuing medical education and other educational programs.

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.