Debunking the Delusion That Precision Oncology Is an Illusion Free

Oncology is at the forefront of implementing personalized/precision medicine, at least in part because cancer is a genomic disease. With unprecedented advances in the understanding of aberrant molecular activation pathways implicated in tumorigenesis coupled with the ever‐increasing availability of cognate agents, we have a growing capability to inflict an assault on malignancies. Innovations in personalized medicine (including genomic and immunologically targeted therapies) have bestowed the gift of time to numerous patients, with documented increases in survival in deadly diseases including, but not limited to, melanoma and lung cancer.

Many of the major advances in oncology over the past two decades are attributable to precision medicine, defined as biomarker‐driven treatment. Chronic myeloid leukemia is a poster child for the precision medicine paradigm. With the discovery of imatinib as a targeted therapeutic for the BCR‐ABL gene product and its administration to newly diagnosed patients, median survival increased from about five to over 20 years. Other examples of precision medicine‐based transformative advances in lethal, previously mostly untreatable cancers include imatinib as a c‐KIT inhibitor in gastrointestinal stromal tumors, crizotinib in ALK‐rearranged non‐small cell lung cancer, epidermal growth factor receptor (EGFR) inhibitors in EGFR‐mutant lung cancer, ROS1 inhibitors in lung cancer, BRAF/MEK inhibitors in melanoma, HER2/neu‐targeted therapies in breast cancer, anti‐CD30 antibody drug conjugates in Hodgkin disease and anaplastic lymphoma, sonic hedgehog inhibitors in basal cell cancer, and RET inhibitors in medullary thyroid cancer. Most or all of these advances were defined on the basis of randomized clinical trials

Despite these advances, a recent perspective article in a major journal argued that precision medicine is an illusion. However, it cited only selected negative studies [1]. The importance of genomic testing and matching patients to the right drug is readily apparent from each of the above breakthroughs [2]. In that perspective, the studies cited did indeed show minimal if any improvement in the primary outcomes. In light of the wealth of positive studies, a critical analysis of the negative trials is required. As an example, the negative SHIVA randomized trial is often mentioned as an example to bolster the argument that precision medicine is a failure [3]. SHIVA is important, as it demonstrated that a randomized precision medicine trial could be conducted. However, approximately 80% of the patients in SHIVA were matched to single‐agent mTOR or hormone modulators. Hence, it is reasonable to conclude that matched monotherapy with these agents in the advanced cancer setting is not effective. The corollary that all precision medicine is a failure extrapolates the finite observations in this trial to settings that were not adequately explored in the SHIVA trial and is, hence, not justifiable.

The article also quotes an MD Anderson study that showed that only 6.4% of patients who were sequenced could be paired with an agent [4]. However, more recent data from the same institution and others demonstrate that about 25% of patients tested could be matched to a drug , with the higher percentages in the latter studies at least partially due to the greater yield of potentially actionable alterations with the use of larger, more robust next‐generation sequencing gene panels. Other factors that limit the utility of genomic testing need to be acknowledged, most prominently the fact that profiling is often applied to heavily pretreated, end‐stage patients . Finally, despite these limitations, three meta‐analyses totaling approximately 85,000 patients demonstrated that the precision paradigm, that is, biomarker‐driven matching, was safe and independently associated with improvement in all outcome variables . Furthermore, the response rate was a remarkable 42% in phase I studies that used a genomic biomarker. Additionally, these meta‐analyses demonstrated the futility of not using precision medicine, that is, of targeted therapeutics applied without a biomarker. In the latter types of studies, median response rates were only about 5% across trials, and outcome parameters were significantly worse than with any other type of study, including those of trials with traditional cytotoxics.

Another major emerging element that must be considered in the context of precision therapy is immunology‐based treatment and its “marriage” with genomics. It is becoming clear that the immune system recognizes the mutanome. Furthermore, molecular anomalies such as PDL1 amplification in Hodgkin disease, mismatch repair gene defects in colorectal cancer, and high tumor mutational burden serve as biomarkers that predict striking and durable response rates.

The opponent of precision medicine [1] also commented on the futility of in‐depth analyses of exceptional responders, with the small number of case reports of such responders taken as evidence for their rarity [11]. We have personally taken care of numerous exceptional responders , many of whom were on genomically guided precision treatment. The paucity of individual reports is likely due, in large part, to the difficulty in publishing case reports and to the fact that many of these exceptional responders are part of clinical trials and are hence included in the trial publications [19].

Molecular technology has advanced at a pace unparalleled in human history, with the initial sequencing of the first genome at the turn of the century, a feat that took over a decade and almost 3 billion dollars, to the current ability to sequence genomes in a matter of hours for just a few thousand dollars. Drug availability is also rapidly improving, as diverse stakeholders collaborate to deliver a decisive assault on cancer. Science mandates that perspectives be based on more than a few cherry‐picked negative studies . Furthermore, it is important to learn from both the negative and positive studies. Indeed, there may be traditional study models that do not fit well with the precision medicine paradigm. Several parameters that we believe are crucial to developing precision trials include experimental drug and other medication availability, investigator experience, study design, algorithm flexibility, the robustness of the molecular technology, and the type of patient enrolled (e.g., number of prior treatments). The extent to which genomically informed treatment has not met with universal success is not an argument for abandoning this practice. Rather, failures of the current precision approaches only underscore the necessity for ever more precision‐guided clinical research. Clearly, the complexity of human cancer is such that single‐gene markers are unlikely to be sufficient to guide ideal therapy for many or most patients. Thus, it is only through an iterative process that integrates ever‐more complex personalized data into treatment paradigms that progress in precision oncology will accelerate. Furthermore, acknowledging the need to move to patient‐centered trial design for precision medicine is important, as is the emerging [20], but unsurprising, concept that customized combinations, rather than monotherapy, are needed in order to prosecute complex, advanced cancers.

In summary, there are compelling scientific data that indicate that precision medicine represents the future of oncology. It is a delusion to consider precision medicine an illusion.

Disclosures

Vivek Subbiah: Novartis, Bayer, Roche/Genentech, Nanocarrier, Fujifilm, Incyte, Northwest Biotherapeutics, D3, ABBVIE, Berghealth (RF); Razelle Kurzrock: X‐Biotech, Actuate Therapeutics (C/A), Genentech, Pfizer, Sequenom, Guardant, Foundation Medicine, Merck Serono (RF), CureMatch Inc. (OI), Actuate Therapeutics and XBiotech (SAB).

(C/A) Consulting/advisory relationship; (RF) Research funding; (E) Employment; (ET) Expert testimony; (H) Honoraria received; (OI) Ownership interests; (IP) Intellectual property rights/inventor/patent holder; (SAB) Scientific advisory board

References

Author notes