Contradictions in Precision Medicine Research: Imprecision Poses a Threat to Healthcare

In late January 2015, U.S. President Barack Obama announced the launch of the Precision Medicine Initiative (PMI). The PMI is dedicated to providing personalized healthcare services to all individuals, embodying the principle of “delivering the right treatment to the right person at the right time.” The U.S. National Institutes of Health (NIH) defines precision medicine as follows: “An emerging approach for disease treatment and prevention that takes into account individual variability in genes, environment, and lifestyle.” In 2016, the U.S. Food and Drug Administration (FDA) invested $215 million and recruited one million volunteers, aiming to advance precision medicine from concept to clinical application.

Since the announcement of the plan, the technology, healthcare, and industrial sectors have all welcomed it with great enthusiasm. However, in stark contrast to this widespread interest, reimbursement for precision medicine has struggled to keep pace with research advancements. Insurance companies remain uncertain about whether they should cover the cost of genetic testing essential to precision medicine, a hesitation that has profoundly impacted biotechnology companies developing genetic technologies.

Furthermore, the imprecision of research data and results renders precision medicine a misnomer. VCBeat (WeChat ID: VCbeat) has compiled and translated an article published in Science, offering a detailed look at the various threats that imprecise research findings pose to precision medicine. The authors are Spencer Phillips Hey, Ph.D., a fellow in the Program on Regulation, Therapeutics, and Law (PORTAL) at Brigham and Women’s Hospital and a faculty member at the Center for Bioethics at Harvard Medical School; and Aaron S. Kesselheim, M.D., an Associate Professor of Medicine at Brigham and Women’s Hospital and Director of PORTAL.

The goal of precision medicine is to bring about a “transformation,” fundamentally altering the current “one-size-fits-all” model of healthcare. For the average American, particularly cancer patients, precision medicine offers more targeted therapies and is therefore correspondingly more effective. However, often contrary to the aims of precision medicine, the scientific evidence underpinning these targeted treatments is not highly precise, leading to considerable uncertainty in outcomes.

The authors argue that there are three major obstacles hindering the transformation of precision medicine: 1. Researchers often fail to rigorously test the biological theories underlying the efficacy of targeted therapies; 2. Researchers frequently neglect to validate the accuracy of diagnostic tests used to determine patient eligibility for corresponding treatments; 3. A lack of collaboration among researchers leads to inefficiencies in research.

Biological theories are of great importance.In the past, researchers did not need to fully understand the mechanism of action of a drug; it was sufficient that it improved health or saved lives. For example, statins were initially thought to prevent heart disease by reducing levels of harmful low-density lipoprotein (LDL) cholesterol in the blood, but they actually exert their cardioprotective effects by inhibiting inflammation.

Precision medicine is implemented through varying approaches. First, basic research identifies the genes or proteins that underlie specific diseases or conditions. Second, drugs are developed to modify gene products or block the action of these proteins. Finally, diagnostic tests are created to identify individuals who carry the relevant genes or exhibit abnormal protein levels, enabling targeted administration of appropriate therapies. While all these steps are grounded in biological theories, it is crucial to remain cognizant that such theories may well be erroneous or incomplete.

Take cetuximab (Erbitux) as an example. This drug was developed as a precision medicine for the treatment of metastatic colorectal cancer and received FDA approval in 2004 for use in patients with epidermal growth factor receptor (EGFR)-positive tumors. However, the pivotal clinical trials that led to its approval did not include any EGFR-negative patients. This omission is significant because it precluded testing of the drug’s key biological hypothesis—namely, that cetuximab provides no benefit to EGFR-negative patients. Subsequent research revealed that EGFR expression is largely unrelated to whether patients with metastatic colorectal cancer benefit from cetuximab. Instead, a genetic biomarker involving mutations in the Kirsten rat sarcoma viral oncogene homolog (KRAS) has proven to be a better predictor of therapeutic response.

Prior to the approval of cetuximab, inadequate testing of the biological links among abnormal genes or proteins, cancer, and diagnostic tools resulted in some patients who could have benefited from the drug failing to receive it. Although these patients were spared from adverse drug reactions, they also missed the opportunity to derive therapeutic benefit. This is precisely what precision medicine aims to prevent.

Managing Uncertainty.In conventional disease research, a group of subjects is administered a drug believed to treat a specific condition, and the criterion for evaluating its efficacy is whether patients show improvement. In precision medicine research, researchers must consider additional factors, such as which genes, proteins, or biomarkers are involved; which diagnostic tests can be used to identify them; and which drugs are most suitable for treating patients with those biomarkers. Uncertainty regarding these questions can significantly impact treatment decisions.

HER2 is a receptor protein on the surface of breast cancer cells and is recognized as one of the most robust biomarkers. The higher the level of HER2, the more “growth and division” signals are transmitted to cancer cells. Trastuzumab (Herceptin) and lapatinib (Tykerb) are examples of direct HER2-targeted therapies. Standard HER2 testing involves quantifying the amount of protein on the cell surface. Typically, scores of 0 to 1+ are classified as HER2-negative, 2+ as equivocal, and 3+ as HER2-positive. However, criteria for interpreting test results as positive or negative vary slightly across different laboratories and pathologists. Consequently, some patients who should receive anti-HER2 therapy and would benefit from it miss this opportunity, while others who are HER2-negative are unnecessarily treated with these agents. Despite extensive research in this field, the precise threshold for initiating anti-HER2 therapy remains ill-defined.

Wilderness and Chaos.When various research groups, laboratories, and companies collaborate within the same field, the lack of a single party responsible for final result verification or work coordination often makes it difficult to ensure comparability among their respective findings.

For example, in the treatment of lung cancer, many teams are investigating the use of a protein biomarker called Excision Repair Cross-Complementation Group 1 (ERCC1) to enable more precise targeted chemotherapy for lung cancer. Theoretically, the chemotherapy drugs cisplatin and carboplatin are more effective for patients with lower levels of ERCC1 protein in their tumors. In reality, however, the situation is more complex and frustrating.

An evaluation of 33 studies on ERCC1 and chemotherapy for lung cancer revealed that the heterogeneity in diagnostic methods and scoring criteria for ERCC1 was so substantial that the results were largely incomparable. Consequently, despite more than a decade of research, it remains unclear whether ERCC1 measurement is truly useful. Nevertheless, numerous ERCC1 testing kits are being marketed and used to help “guide” chemotherapy for lung cancer.

The path forward.The goals of precision medicine are admirable. However, given the current low research efficiency, achieving these goals will take a considerable amount of time and cost far more money than anticipated. Multiple research groups are exploring the same objectives, with a lack of communication and oversight among them. This type of uncollaborative research does not lead to better or more reliable biomarker tests and treatments; instead, it only creates confusion and undermines interventions aimed at improving patient health.

Precision medicine research should not be conducted in this manner. Drugs for precision medicine must undergo high-quality testing comparable to that required for any other drug, including randomized trials to validate potential mechanisms of action. Furthermore, we should establish a core, advanced, publicly accessible database encompassing studies related to biomarkers and diagnostic tests. This database could be collaboratively developed by major research institutions, such as the National Human Genome Research Institute and the National Cancer Institute. Such an initiative would undoubtedly enhance research transparency in precision medicine, reduce waste, and improve research efficiency.