Cost-Effectiveness of Prophylactic Cranial Irradiation vs. MRI Surveillance for Extensive Stage Small Cell Lung Cancer

Abstract

Due to conflicting prospective data, controversy exists regarding prophylactic cranial irradiation (PCI) in extensive stage small cell lung cancer (ES-SCLC). We sought to evaluate the cost-effectiveness of PCI versus MRI surveillance alone in the context of modern radiotherapy treatment modalities using available evidence and the CMS alternative payment model (APM).A Markov state transition model was constructed using software to perform a cost-effectiveness analysis comparing MRI surveillance alone with PCI for ES-SCLC. PCI in the base case analysis was assumed to be 25 Gy in 10 fractions delivered with a 3D conformal radiotherapy technique (3D-CRT) without daily image guidance. In the MRI surveillance alone arm, MRI surveillance was assumed to be performed at baseline then every 3 months for the first year and every 6 months for year 2, consistent with the surveillance plan currently being implemented in the SWOG MAVERICK cooperative group trial (NCT04155034). Clinical parameters were obtained from clinical trial data, and cost data were based on 2019 Medicare reimbursement. Strategies were compared using the incremental cost-effectiveness ratio (ICER) with effectiveness in quality-adjusted life years (QALYs) and evaluated with a willingness to pay threshold of $100,000 per QALY gained. One-way and probabilistic sensitivity analyses were performed to consider model uncertainty.In the base case scenario, PCI was not cost-effective with an ICER of $168,456 per QALY gained compared to MRI surveillance alone, assuming that overall survival was 10 months and 12.5 months for MRI surveillance alone and PCI, respectively. One-way sensitivity analysis showed that results were most sensitive to the variation of overall survival and cognitive decline rates between the two groups. In a scenario analysis, when all patients receive hippocampal-avoidance PCI (HA-PCI), the model results become near cost effective with PCI due to presumed lower rate of cognitive decline (ICER $129,307 per QALY gained). If HA-IMRT with memantine was also utilized for salvage treatment in all patients in the MRI surveillance group with the assumption above, PCI was found to be cost effective ($88,096 per QALY gained). In addition, the total radiation therapy reimbursement per episode suggested by APM for radiation therapy for brain metastases ($10,783) is almost identical to the current Medicare reimbursement rate for 10 fractions with HA-IMRT ($10,445). Thus, this scenario analysis also reflected the use of the APM.PCI was not cost effective compared to MRI surveillance alone due to the neurocognition decline impact of PCI based on available evidence. HA-PCI may be a potential cost-effective strategy for ES-SCLC, with expected confirmation following the ongoing SWOG MAVERICK trial, which includes assessments of cognitive function.