Background
Cost-effectiveness analysis using decision-analytic modeling. Evaluated the cost-effectiveness of proton beam therapy (PBT) vs IMRT for various cancer types including prostate, HNC, non-small cell lung cancer, and pediatric medulloblastoma. Used incremental cost-effectiveness ratios (ICERs) in cost per quality-adjusted life year (QALY) gained to assess which tumor types justify proton investment at the $100,000/QALY willingness-to-pay threshold. Vega RB et al, Cancer 2013.
Interventions and follow up
Arm A: Proton beam therapy (estimated cost: $30,000–$60,000 per course in 2013 USD)
Arm B: IMRT (estimated cost: $15,000–$30,000 per course in 2013 USD)
Primary endpoint: ICER (cost/QALY) by tumor type
mFollow up: N/A (modeling study)
Arm B: IMRT (estimated cost: $15,000–$30,000 per course in 2013 USD)
Primary endpoint: ICER (cost/QALY) by tumor type
mFollow up: N/A (modeling study)
Results
Pediatric medulloblastoma ICER: $25,000–$40,000/QALY — cost-effective at standard threshold
HNC ICER: $50,000–$80,000/QALY — borderline cost-effective
Prostate cancer ICER: >$250,000/QALY — not cost-effective at standard threshold
NSCLC ICER: Variable ($60,000–$150,000/QALY depending on stage and HRQOL gain)
HNC ICER: $50,000–$80,000/QALY — borderline cost-effective
Prostate cancer ICER: >$250,000/QALY — not cost-effective at standard threshold
NSCLC ICER: Variable ($60,000–$150,000/QALY depending on stage and HRQOL gain)
Adverse events
Main adverse events: N/A — modeling study. Late effects differences between proton and photon (secondary malignancies, cardiac events) are incorporated into QALY calculations and are the primary driver of cost-effectiveness in pediatric and young adult populations.
Conclusions
Proton therapy is most cost-effective for pediatric malignancies (medulloblastoma, ependymoma) where long life expectancy makes late effects prevention most valuable. For prostate cancer, proton therapy is not cost-effective compared to IMRT given equivalent oncologic outcomes and similar HRQOL. HNC is borderline cost-effective and requires further clinical data.
Key Limitations
Key Limitations: Modeling assumptions based on limited comparative effectiveness data. Proton costs have declined since 2013. QALY estimates for late effects are extrapolated from non-randomized data. Does not account for emerging pencil-beam scanning technology with lower capital and operating costs. Insurance reimbursement realities not modeled.
Clinical Context
Cost-effectiveness data support proton therapy for pediatric CNS tumors and HNC as the most defensible indications. Coverage decisions by CMS and private insurers often cite these data. Many US insurance carriers cover proton for pediatric CNS tumors but may deny coverage for prostate cancer unless enrolled in a clinical registry (PCPT registry). Proton capacity expansion has shifted the cost-benefit calculus since 2013.
References