MediRama CEO Moon Hanlim Shares Strategies for Companion Diagnostics Development in Oncology
“Cutoff setting” in companion diagnostics is critical — success or failure can hinge on a fine line
The blockbuster anticancer drug Keytruda cannot be prescribed simply because a patient requests it. Since tumor characteristics vary from patient to patient, the drug may be life-saving for some but ineffective or even harmful for others. Therefore, prior to treatment, companion diagnostics (CDx) are used to determine whether a patient’s tumor is likely to respond to a specific therapy.
On the 26th, Moon Hanlim, CEO of clinical development specialist MediRama, held an online seminar titled “Changes in Companion Diagnostics in Oncology Drug Development: Assays, Cutoffs, Samples, and Regulatory Considerations.” He shared trends and development strategies in companion diagnostics, focusing on real U.S. Food and Drug Administration (FDA) approval cases. In particular, he introduced key strategies in “cutoff setting,” a decisive factor in CDx development success.
Biomarkers vs. Companion Diagnostics: What’s the Difference?
Although biomarkers and companion diagnostics are often used interchangeably in practice, they are clearly distinguished from a regulatory perspective.
A biomarker is a broad concept referring to biological indicators used to diagnose disease, predict prognosis, or monitor treatment response.
In contrast, a companion diagnostic (CDx) is an in vitro diagnostic (IVD) medical device approved under regulatory standards to identify the specific patient population eligible for a particular therapy. In other words, when a specific biomarker becomes an essential determinant for selecting a particular treatment, it must receive regulatory approval as a CDx.
Companion diagnostics are approved in tandem with a specific drug, and the drug label specifies that the associated test is mandatory. A representative example is the PD-L1 test required for prescribing Keytruda.
The “Magic” of Cutoffs That Determines Clinical Success or Failure
The core of CDx development lies in setting the cutoff (the threshold for a positive result). This serves as the final gatekeeper in determining which patients are eligible for treatment and ultimately influences market size and approved indications.
Moon explained, “In the early stages, provisional criteria are established based on preclinical or early clinical data, which can lead to trial and error. As clinical development progresses and regulatory review continues, the biomarker definition and cutoff are continuously optimized.” However, once a cutoff is fixed in the approved label, it is directly tied to clinical evidence and becomes extremely difficult to modify afterward.
Keytruda is a prime example of how cutoff strategy can reshape the market landscape. In the early days of immuno-oncology development, Merck was about four years behind its competitor Bristol Myers Squibb (BMS). To overcome this gap, Merck adopted a patient selection strategy based on PD-L1 expression. At the time, there were concerns that narrowing the patient population would be commercially disadvantageous, and PD-L1 was criticized as an imperfect biomarker due to tumor heterogeneity and differences between primary and metastatic tumors. Nevertheless, Merck made the bold decision to select only patients with high PD-L1 expression (TPS ≥50%). Keytruda demonstrated overwhelming response rates and became the standard of care for non-small cell lung cancer (NSCLC). Clear patient selection ultimately strengthened clinical persuasiveness.
In contrast, Elahere is an example of a drug saved by adjusting its cutoff. In the Phase 3 FORWARD I trial, the drug initially failed due to an inappropriate cutoff setting (expression ≥50%). However, a post hoc analysis raised the threshold to “high expression ≥75%,” and after re-challenging with the refined strategy, the drug ultimately received full approval in 2024. Moon emphasized, “If development had been abandoned after the failure at 50%, this drug would never have reached patients,” underscoring the importance of strategic judgment.
Expansion from Tissue to Blood-Based “Liquid Biopsy”
The use of specimens is also evolving. While tumor tissue testing was the standard in the early stages, blood-based “liquid biopsy” (ctDNA NGS) is rapidly expanding to overcome the limitations of tissue sampling and to better reflect the dynamic nature of tumors.
The cases of capmatinib and tepotinib targeting MET exon 14 mutations illustrate this shift. Capmatinib adopted a tissue-based strategy, while tepotinib obtained approval using an integrated approach combining tissue and plasma testing. The FDA has shown flexibility by recommending additional tissue testing if plasma results are negative.
In principle, the FDA requires contemporaneous approval of therapeutic drugs and their companion diagnostic devices. However, in cases involving serious rare diseases or extremely small biomarker-defined populations, exceptions may be made, allowing approval of the drug first with subsequent completion of the diagnostic device requirements.
No “After-the-Fact” Fixes — A Proactive Biomarker Strategy Is Essential
Moon stressed that “companion diagnostics must be developed hand-in-hand with the drug.” Failure to understand the regulatory concept of simultaneous approval may force strategic changes in the later stages of pipeline development.
He added, “Reanalyzing failed clinical trial data post hoc to narrow the patient population based on biomarkers has a low probability of success.” Instead, a proactive strategy is required — designing cutoffs from the planning stage based on deep understanding of target biology and co-developing CDx devices alongside new drugs.
He further noted that as the FDA plans to reduce reliance on laboratory-developed tests (LDTs), efforts should be made from early clinical stages to secure commercially available diagnostic assays that have completed analytical validation.
Source: Hit News (https://www.hitnews.co.kr)