Value of Real-World Evidence for Treatment Selection: A Case Study in Colon Cancer

Area of Science:

• Oncology
• Health Services Research
• Biostatistics

Background:

• Real-world data (RWD) analysis generates real-world evidence (RWE) for personalized treatment decisions.
• Generating valid RWE is challenging due to potential confounding factors.
• This study addresses the responsible generation of RWE for clinical decision-making.

Purpose of the Study:

• To demonstrate a method for responsibly generating RWE to guide treatment decisions.
• To validate this RWE generation approach by replicating an existing adjuvant chemotherapy (ACT) guideline for stage II and III colon cancer (CC).

Main Methods:

• Analysis of 27,056 patients with stage II and III CC from the Netherlands Cancer Registry.
• Estimation of ACT's 5-year overall survival (OS) benefit using G-computation, adjusting for patient and tumor characteristics.
• Construction of an ACT decision tree based on estimated patient-specific OS benefits.

Main Results:

• The derived decision tree aligns with the current Dutch guideline for ACT in stage II and III CC.
• The guideline recommends surgery and ACT for stage III or stage II (T4) CC, and surgery alone for stage II (T3) CC.
• No sufficient RWE was found to contradict ACT for stage II (T4) CC with microsatellite instability-high (MSI-H).

Conclusions:

• RWE, when carefully generated, can supplement evidence for clinical decision-making and inform treatment guidelines.
• The study validates current ACT recommendations for CC and suggests further investigation into MSI-H status.
• Future guideline iterations should consider MSI-H status for stage II CC treatment decisions.