Session Item

In stage III non-small cell lung cancer (NSCLC), prophylactic cranial irradiation (PCI) reduces the brain metastases incidence, prolongs the progression-free and brain metastases-free survival, but has no significant effect on overall survival. It increases the risk of neurocognitive toxicity and is currently not adopted in routine care. Our objective was to assess the cost-effectiveness of PCI added to current practice compared with current practice without PCI in stage III NSCLC from a Dutch societal perspective.

A cohort partitioned survival model (Figure 1) including five health states (progression-free, brain metastases, extracranial metastases, brain and extracranial metastases and death) was developed based on individual patient data from three randomized phase III trials (RTOG0214, Guangzhou2005 and NVALT-11, N=670). Quality-adjusted life years (QALYs) and costs per health state were estimated over a lifetime time horizon (one month cycle time). Future effects and costs were discounted by rates of 1.5% and 4.0% respectively (Dutch pharmaco-economic guideline). A willingness-to-pay (WTP) threshold of €80,000 per QALY was adopted. Deterministic and probabilistic sensitivity analyses, as well as scenario analyses, were performed to address parameter uncertainty and to explore what parameters had the greatest impact on the cost-effectiveness results.

The probability of PCI gaining three and six additional months of life were 76% and 56% respectively, and PCI gaining three and six months of life in perfect health were 73% and 42% (Figure 2A). Total discounted costs and QALYs were €118,896 (95% CI €102,205- €139,255) and 3.466 (95% CI [2.986-3.960]) for current practice with PCI and €108,773 (95% CI 92,683–126,630) and 3.023 (95% CI 2.612-3.444) for current practice without PCI. Mean incremental costs and QALYs amounted to €10,123 and 0.443 respectively, resulting in an incremental cost-effectiveness ratio (ICER) of €22,843 per QALY gained (Figure 2B). Sensitivity analyses showed that the ICER was most sensitive to the utility value of the progression-free health state and the number of administered PCI fractions. The probability of PCI being cost-effective at a WTP threshold of €80,000 per QALY was 93% (Figure 2C). Scenario analyses showed that using alternative survival distributions had little impact on the ICER. The scenario analyses assuming fewer PCI fractions (10 instead of 15 fractions of 3 Gy) and excluding indirect costs decreased the ICER to €18,263 and €5,554 per QALY gained, respectively.

Adding PCI to current practice is cost-effective compared to current practice without PCI in stage III NSCLC. From a cost-effectiveness perspective, PCI could be considered in routine care, but patients should be informed about the potential benefits and risks (neurocognitive toxicity) of PCI in a shared decision making process.