Precision Cancer Medicine: Large Studies Indicate Steady Progress

In this issue of Cancer Discovery, Horak and colleagues reported on clinical outcomes of 1,310 patients (75.5% with rare cancers) enrolled in the biology-driven, multicenter, prospective, observational NCT/DKTK MASTER trial (1). It applied whole-genome/whole-exome and RNA sequencing to inform the care of adults with advanced cancer across histologies who are younger than 51 and patients with rare tumors, including rare subtypes of more common entities, regardless of age. Evaluations of biomarkers' clinical actionability (based on 472 single and six composite biomarkers) and assignments of molecularly informed therapies were performed in 88% of cases following semiweekly, multicenter molecular tumor board conferences. Recommended therapies were administered in 362 of 1,138 patients (31.8%) and resulted in significantly improved overall response and disease control rates (23.9% and 55.3%) compared with previous therapies, translating into a progression-free survival ratio >1.3 in 35.7% of patients. The therapeutic decisions were assigned on the basis of seven biomarker baskets based on the cellular pathways or processes involved [tyrosine kinases (TK), PI3K-AKT–mTOR, RAF–MEK–ERK, cell cycle, developmental regulation, DNA damage repair (DDR), and immune evasion (IE)]. When exploring the distribution of clinically actionable biomarkers across the different baskets and histologic entities, the authors observed that most biomarkers based on which treatments were recommended fell into the DDR (26.8%), TK (25.8%), and IE (10.7%) baskets. Looking into the data from a clinical perspective, how many patients in the MASTER trial received a treatment allocation with a clinically meaningful chance of affecting outcome? Investigators observed that recommendations based on clinical data collected in the same histologic entity could be made in 200 of 1,138 patients (17.6%) and that only 8.2% of all recommendations fell into this group, reflecting the lack of molecularly stratified clinical trials in patients with rare cancers. Applying the European Society for Medical Oncology Scale for Clinical Actionability of Molecular Targets (ESCAT), the authors found that only 4.6% and 3.7% of recommendations fell into ESCAT tiers I and II, 25.8% into ESCAT tier III, 30.8% into ESCAT tier IV, 0.7% into ESCAT tier V, and 17.7% into ESCAT tier X. In the context of the MASTER trial, only 4.6% of patients received a molecular alteration–drug match treatment allocation with a potential clinical benefit and 3.7% a molecular alteration–drug match treatment allocation with a potential antitumor activity. Other basket trials in the past had this approach with conflicting results (2, 3), and a previous meta-analysis demonstrated that, across malignancies, a personalized strategy was an independent predictor of better outcomes and fewer toxic deaths (4). We believe the most important observation of the MASTER trial is related to the 30.8% of patients with rare cancer who had ESCAT tier III molecular alterations, for which we may predict potential clinical benefit because they harbor specific alteration (as tiers I and II) but in a different tumor type. The MASTER trial is a real-world trial validating the concept of pan-cancer basket trials (tumor-agnostic), supporting the need for N-of-one treatment strategies with personalized drug combinations. We are moving from tumor-specific to genomically agnostic clinical trials. The traditional approach of cancer care based on site/histology, rather than genomic alterations, would be more appropriate for rare cancers. An expanded genome-sequencing approach will provide more evidence that cancers are driven by their genomic road map, rather than just the tissue of origin, and this will allow for improved outcomes across rare cancers. A seminal basket trial of the NTRK inhibitor larotrectinib included 55 patients with NTRK fusions across several rare tumors, with a very impressive sustained long-term response rate of 75%, leading to the second FDA agnostic approval (5, 6).

In this issue of Cancer Discovery, we have another very interesting report where Van Tilburg and colleagues present the results on the long-term clinical outcome of the first 519 patients included in the pediatric precision oncology INFORM registry (7). Patients with relapsed, progressive, or high-risk pediatric cancer from 72 centers in eight European countries were molecularly defined using low-coverage whole-genome sequencing (lcWGS), whole-exome sequencing, RNA sequencing, and methylation array. Progression-free survival benefit was reported for 3.8% patients with very high evidence targets that received matched targeted treatment. Hereditary cancer predisposition syndromes were identified in 7.5%, a change or refinement of diagnosis in 8.2% of patients.

The report confirms the feasibility of cancer profiling with an outstanding mean turnaround time of 25.4 days, but it also highlights several main challenges of these programs.

Highest target priority level alterations are observed only in a small subgroup of cancer, and targeted treatments have resulted in relevant clinical activity in patients with driver events like NTRK/ALK/RET/ROS gene fusions or BRAF^V600 mutations, including in pediatric cancers (8). However, most malignancies exhibit various and multiple oncogenic events that underline the cancer complexity. Therefore, priority scales for other alterations seem helpful to simplify decision making on treatment strategies, whereas it will be important to analyze these multiple occurrences in their context to be able to develop relevant combination strategies.

The pediatric INFORM used a seven-scale target prioritization algorithm that is based on biological findings of the alterations. In contrast, most adult precision medicine programs, as well as the pediatric European MAPPYACTS trial (NCT02613962), also refer to clinical evidence for therapy suggestions. The next step will need to incorporate increasing knowledge of biological and clinical research to provide algorithms using artificial intelligence tools that address the 95% patient population for whom optimal treatment strategies are still lacking.

Limited access to cancer complexity inappropriate treatment options is the main flaw of current cancer precision medicine. In the absence of yet clearly defined biomarkers, innovative clinical proof-of-concept trials like the AcSé-ESMART platform trial (NCT02813135) allow the exploration of targeted treatment combinations, hypothesis-driven, based on knowledge in basic and translational research, in molecular enriched study populations (9, 10).

Integrating molecular profiling at an early treatment failure is crucial, and high-level alterations are now introduced in first salvage or first-line therapies. Nevertheless, most advanced patients will necessitate comprehensive clinical discussions with the treating physicians to share knowledge on tumor biology and expertise on novel anticancer agents to accompany suggesting experimental therapeutics in patients without curative treatment options.

So where are we going with these large studies indicating steady progress in the implementation of precision cancer medicine?

Real-world data/real-world evidence is currently used by regulatory agencies in terms of monitoring post-market drug adverse events and may generate evidence for drug approval. The MASTER trial is a large-scale, real-world trial that offers insights into rare diseases. The uptake of the agnostic paradigm has introduced new complexities in regulatory pathways. Health technology assessment bodies face serious challenges to uptake the innovation in oncology, being required to estimate the clinical impact of drugs and calculate their cost-effectiveness, that is, the magnitude of clinical benefit. The naïve conception of abandoning the histologic paradigm in favor of biomarker-based strategies is now evolving into a renewed consciousness of the inextricable connections between the two entities. The spectrum of actionability reflects our knowledge on those connections and is therefore dynamic by definition. A consequent dynamism is concurrently required in drug approval and reimbursement paradigms, which will need to adapt to the new type of evidence produced by precision medicine trials.

Finally, the DKTK initiative in Germany, the creation of a high performance, well-funded, collaborative network for development of innovative translational cancer research infrastructures and projects, is an impressive success story that sets an example to be followed!

No disclosures were reported.