[Reporter's View] New drugs for intractable cancers

An approach towards interpreting Progression-Free Survival (PFS)

Although there are diverse types of intractable cancer, different diseases and varying prognoses coexist. The common feature of platinum-resistant ovarian cancer, small cell lung cancer (SCLC), and bile duct cancer is clear. They are characterized by limited treatment options, a short average survival period, and quickly diminishing next-line options.

Drug discovery is fundamentally challenging in this field. The number of patients is limited, biological heterogeneity is significant, and designing a clinical trial scheme is difficult because many patients have worsened systemic conditions at diagnosis. Furthermore, it is difficult to set up a control group and clearly demonstrate a difference in overall survival (OS) due to crossover or follow-up treatments.

As a result, clinical outcomes are mostly reported as improvements in Progression-Free Survival (PFS) or Hazard Zatios (HRs). There have been studies in which the difference in PFC between the control and treatment groups is only 1 month. The outcomes, such as HR 0.73, a 27% reduction in disease progression or risk of death, and an HR below 1, indicate effective treatment and statistical significance.

However, what patients experience is not the relative risk reduction or the time they have saved. Studies with PFS extended from 3.8 months to 5.2 months translate to HR 0.73 and a 1.4-month extension. It shows progress in terms of statistics; however, the clinical significance is much more complicated.

Nevertheless, a tiny improvement in PFS cannot be taken lightly. For patients with few treatment options remaining, disease control of 6 to 8 weeks can represent an opportunity to bridge to the next treatment and time to preserve quality of life. If Objective Response Rate (ORR) or PFS2 appear meaningful, and a patient group with a long Duration of Response (DoR) exists, the 'long tail' behind the median is by no means small. In the HR 0.72 value, the varying times of different patients are contained.

The problem arises when these figures assume identical expectations for all patients. Outcomes such as a 1 to 2-month improvement in PFS must be interpreted alongside toxicity burden, treatment discontinuation rates, and Patient-Reported Outcomes (PRO). Particularly in situations where OS data is not sufficiently mature, the criteria for 'meaningful time' become more complex as PFS-centered evaluations are repeated.

This concern extends beyond the clinical field into institutional judgment. Within limited budgets, which drugs should be permitted for which patient groups and at what point? The fact that the risk of death was reduced by 20% does not necessarily translate into the same reduction for all patients. It is more important to determine how precisely patients with a high likelihood of response can be selected, and how to verify and enhance efficacy in clinical practice.

The approach must be more careful in the field of intractable cancers. Rather than opening the door uniformly to all patients, a structure is needed that applies treatments stepwise, starting with patient groups where evidence is confirmed. A system must be operated in parallel, with access initially permitted on a limited basis while accumulating Real-World Data (RWD), followed by a re-evaluation of OS, PFS2, treatment durability, and quality-of-life improvements after a certain period.

The issue is not the size of the number, but how it is interpreted. Between the desperation of intractable cancer and statistical significance, judgment must be more delicate and transparent.

However, if OS improvement is used as the sole absolute criterion without sufficient consideration of the characteristics of intractable diseases and the difficulties of drug development, patient access to new drugs will inevitably remain low. It is difficult to justify a structure where treatment opportunities are narrowed simply because a patient has a specific type of cancer.