Estimating the loss of lifetime function using flexible parametric relative survival models

Area of Science:

• Biostatistics
• Epidemiology
• Oncology

Background:

• Dynamic evaluation of loss in expectation of life (LEL) offers critical insights for cancer patients and clinicians.
• LEL estimation is complicated by time-to-event data censoring, requiring survival function extrapolation.
• The accuracy of various extrapolation methods for LEL has not been previously assessed.

Purpose of the Study:

• To evaluate the accuracy of different extrapolation approaches for estimating the loss of lifetime function in cancer patients.
• To introduce and assess a novel mixture cure model for survival data extrapolation.
• To provide guidance on selecting appropriate extrapolation methods for LEL calculations.

Main Methods:

• Calculated loss of lifetime function by decomposing the all-cause survival function using relative and general population survival functions.
• Fitted the relative survival function using parametric relative survival models and a novel mixture cure model for extrapolation.
• Assessed extrapolation accuracy via simulation studies and data from the Danish Cancer Registry and Lymphoma Registry.

Main Results:

• No single extrapolation method demonstrated uniform superiority across all scenarios.
• Flexible parametric mixture cure models and flexible parametric relative survival models proved suitable in various situations.
• The choice of extrapolation method impacts LEL estimation accuracy.

Conclusions:

• Extrapolating the relative survival function beyond the follow-up period requires careful consideration.
• Extensive sensitivity analyses are recommended when estimating the loss of lifetime function.
• Flexible parametric models offer viable options for LEL extrapolation in cancer research.