Evidence Brief October 9, 2026

Oncology Evidence Brief | October 9, 2026

This week’s strongest evidence focuses on three foundational questions for oncology evidence generation:

Are we measuring the right burden? Are we studying the right population? And are we managing uncertainty appropriately?

New work on treatment-related time, clinical trial participation, and accelerated approval shows why these questions need to be addressed before sophisticated analysis begins.

Peer-Reviewed Evidence

1. Chow et al. Development and Validation of a Tool to Measure Documented Health Care Encounter Time. JAMA Network Open.

The investigators developed and validated the MASCC TIME tool to standardize measurement of health care encounter duration from medical records.

The approach improves on simpler measures such as health care contact days, which can treat a brief blood draw and a prolonged infusion as equivalent forms of treatment burden.

Validation demonstrated improvements in abstraction accuracy, interrater reliability, and efficiency.

Why it matters

Time spent receiving health care is increasingly recognized as an important consequence of cancer treatment.

Measuring that burden in hours rather than simply counting contact days provides a more meaningful representation of what treatment requires from patients.

For patient-preference studies, this may support more realistic attributes describing treatment time and care-setting burden.

For health-economic evaluation, it creates opportunities to represent patient time more explicitly. However, documented encounter time still captures only part of the burden. Travel, waiting, recovery, and caregiver time may remain outside the measure.

The broader lesson is that the measurement unit itself can determine whether an important treatment consequence becomes visible.

2. Jiang et al. Physician- and Practice-Level Variation in Clinical Trial Participation Among Patients With Advanced Cancer. JAMA Network Open.

Using a large US real-world oncology database, the investigators examined clinical trial participation among more than 200,000 patients with advanced cancer.

Trial enrollment was highly concentrated. A relatively small proportion of oncology practices and physicians accounted for a large share of participating patients.

Physician- and practice-level characteristics explained substantial variation in participation beyond measured patient characteristics.

Why it matters

Clinical trial representativeness is often discussed primarily in terms of eligibility criteria and patient recruitment.

This study highlights an additional layer of selection: where patients receive care and which physicians treat them can influence whether they ever have the opportunity to enter a trial.

That has implications for both clinical development and real-world evidence.

Differences between trial populations and routine-care populations may reflect research infrastructure, physician engagement, referral patterns, and access to clinical studies as well as biological or demographic differences between patients.

For external comparator studies, transportability assessments should therefore consider health-system and provider-level selection, not simply patient-level eligibility.

3. Jazowski and Dusetzina. Drug Spending on Accelerated Approval Indications Subsequently Withdrawn From the Market. JAMA Health Forum.

The authors examined US spending on drugs granted accelerated approval for indications that were later withdrawn.

Many of the withdrawn indications were oncology-related, and failure to verify anticipated clinical benefit in confirmatory studies was a common reason for withdrawal.

The analysis estimated billions of dollars in health-plan and patient spending during periods when the ultimate clinical benefit of these indications remained uncertain.

Why it matters

Accelerated approval deliberately permits access before all uncertainty has been resolved.

That can be clinically appropriate, but it also transfers part of the evidentiary risk to patients and payers.

The study makes the financial consequences of that uncertainty more tangible.

For evidence strategy, accelerated approval should therefore be viewed as the beginning of an evidence-development obligation rather than the end of the regulatory process.

Confirmatory evidence, reimbursement arrangements, and plans for managing residual uncertainty need to be considered together.

Published Viewpoint

4. Nabhan and Subbiah. Beyond Tumor Size, Toward Meaningful Patient Outcomes—Resisting RECIST. JAMA Oncology.

This viewpoint questions whether conventional radiographic response criteria remain sufficient as oncology treatment moves beyond traditional cytotoxic therapies.

RECIST provides a standardized and reproducible framework for assessing tumor response and progression, but the authors argue that tumor shrinkage may incompletely represent meaningful patient benefit in some contemporary treatment settings.

Why it matters

The issue extends beyond RECIST.

An endpoint can be reliable and operationally useful without capturing the full value of treatment to patients.

Evidence-generation programs increasingly need to consider the relationship between tumor response, survival, symptoms, functioning, treatment burden, and quality of life.

This is an argument for reconsidering endpoint selection rather than evidence that conventional response criteria should simply be abandoned.

Preprints and Emerging Evidence

5. Ruiz Borrego et al. Multimodal Artificial Intelligence Predicts Pathological Response and Prognosis in HER2-Positive Early Breast Cancer Treated With Chemotherapy De-escalation Strategies. Preprint.

The investigators evaluated multimodal artificial-intelligence models combining pathology images and clinical information to predict pathological response and prognosis in HER2-positive early breast cancer.

Higher model scores were associated with greater pathological complete response rates, and a separate prognostic model identified groups with differing event-free survival.

Why it matters

The potentially important application is treatment de-escalation.

If a model could reliably identify patients likely to achieve good outcomes with less intensive treatment, it could reduce toxicity and treatment burden.

However, prediction is not the same as clinical utility.

Before model-guided de-escalation can inform care, evidence would need to demonstrate that acting on the prediction produces acceptable clinical outcomes and improves the overall benefit-burden balance.

Evidence status: Preprint. Not peer-reviewed.

DCE and DCE-TTO Watch

No newly published oncology DCE or DCE-TTO identified this week met our threshold for inclusion.

The MASCC TIME study is nevertheless highly relevant to future preference research because it provides a more rigorous empirical basis for representing treatment-related time burden as a preference attribute.

What This Means for Evidence Strategy

Three questions connect this week’s publications.

What are we measuring? Treatment burden can disappear when the measurement unit is too coarse.

Who contributes to the evidence? Trial populations are shaped not only by eligibility criteria, but also by physicians, practices, and access to research infrastructure.

How is uncertainty handled? Accelerated approval creates value through earlier access, but it also creates evidentiary and financial exposure when anticipated benefits remain unverified.

Together, these findings support a broader principle:

The credibility of evidence depends on what is measured, who is represented, and how uncertainty is managed before decisions are made.

Statistical sophistication comes later.

Worth Reading First

For patient-centered methodology, Chow et al. is the strongest paper this week. The MASCC TIME framework provides a practical way to make treatment-related time burden more visible.

For RWE and trial interpretation, Jiang et al. is particularly useful because it demonstrates that selection into clinical research occurs through health care systems as well as patient eligibility.

For payer strategy, Jazowski and Dusetzina provides the clearest illustration of the financial consequences of unresolved clinical uncertainty.