Pembrolizumab in advanced recurrent endometrial cancer: A cost-effectiveness analysis

To determine the cost-effectiveness of pembrolizumab in patients with recurrent endometrial cancer that have failed first-line chemotherapy.

We created a model to evaluate the cost-effectiveness of pembrolizumab compared to pegylated liposomal doxorubicin (PLD) or bevacizumab for the treatment of women with recurrent endometrial cancer who have failed carboplatin and paclitaxel. Microsatellite instability-high (MSI-H) and non-microsatellite instability-high (non-MSI-H) tumors were evaluated. We included 4400 patients in the model; 800 patients were assumed to have MSI-H tumors. Drug costs were calculated using 2016-2017 wholesale acquisition costs, and cost of Grade III-IV toxicities was estimated from clinical experience. Effectiveness was calculated as 2-year overall survival (OS). We calculated incremental cost-effectiveness ratios (ICERs) to determine the cost per 2-year survivor. Univariate sensitivity analyses were performed. The willingness to pay threshold was $100,000 per year of OS.

The cost of therapy with PLD and bevacizumab were $33.2 million (M) and $167.9 M, respectively. The cost of pembrolizumab therapy was $318.3 M for non-MSI-H patients compared to $57.9 M for MSI-H patients. For non-MSI-H patients, bevacizumab was cost-effective relative to PLD with an ICER of $153,028, while pembrolizumab was not cost-effective relative to bevacizumab with an ICER of $341,830. For MSI-H patients, pembrolizumab was cost-effective compared to PLD with an ICER of $147,249, while bevacizumab was subjected to extended dominance. Sensitivity analysis revealed that for non-MSI-H patients, one cycle of pembrolizumab would need to cost $7253 or less to be cost-effective.

For patients with MSI-H recurrent endometrial cancers who have failed first-line chemotherapy, pembrolizumab is cost-effective relative to other single agent drugs. To be cost-effective in non-MSI-H patients, the cost of pembrolizumab should decrease substantially.