Cost-effectiveness of a circulating tumor fraction molecular biomarker for treatment response monitoring

Authors Zachary Rivers, Charu Aggarwal, Marc Ryan Matrana, Josephine Louella Feliciano, Akash Mitra, Halla Nimeiri, Rotem Ben-Shachar, Cathy Eng, Sheetal Mehta Kircher

Background: Clinical validation studies have demonstrated that molecular biomarkers quantifying ctDNA changes in circulating tumor fraction (TF) predict survival outcomes and may be used for treatment response monitoring (TRM). While clinical utility studies to determine the impact on outcomes of molecular biomarker-driven treatment decisions versus standard of care imaging are ongoing, cost-effectiveness has not been evaluated. Here we evaluate the cost-effectiveness of a molecular biomarker, Tempus xM, used for TRM.

Methods: xM quantifies changes in TF from baseline and on-treatment liquid biopsies. We use a patient-level Markov simulation to compare xM-guided treatment (intervention) to diagnostic imaging-guided treatment (control) over 24 weeks of therapy. In both arms xM and imaging is assessed at 12 weeks and treatment decisions are made based on xM (intervention) or diagnostic imaging (control). Imaging and xM concordance (Table) was based on a retrospective, real-world (RW) study of 51 patients tested with xM that received rw-imaging that evaluated the association of xM with rw-outcomes in advanced pan-cancer patients treated with immune checkpoint inhibitors (ICIs) +/- chemotherapy (CT). We assume non-responders discontinue ICIs and switch to CT and responders remain on ICIs. Appropriate (inappropriate) therapy was defined as treatment decisions concordant (discordant) with xM results. Costs were calculated from Medicare’s perspective in 2023 USD. Control patients do not accrue the cost of xM. Sensitivity analyses were conducted.

Results: Intervention patients saved $4,400 ($63,424 vs. $67,824 in controls) by preventing use of ineffective therapy (4.3 weeks avoided vs 0.2 weeks in controls), a savings of $1076 per week of ineffective therapy avoided. In all sensitivity analyses, xM-guided therapy remained cost effective (Table). The timing of therapy intervention and the cost of therapy had the largest impact on savings.

Conclusions: This model demonstrates that xM-guided treatment is cost-saving compared to imaging alone during 24 weeks of treatment. Future work will incorporate variable clinical uptake and long-term outcomes by expanding the time horizon and including treatment discontinuation due to toxicities and mortality.

Parameter Base Value Range Range of Savings per Week of Ineffective Therapy Avoided, USD
xM Sensitivity 0.98 0.96 – 1 $1,087-$1,063
xM specificity 0.99 0.96 – 1 $1,122-$1,061
xM cost $2,000 $1,000 – $3,000 $1,320-$831
Annual ICI Cost $180,187 $144,150- $216,224 $763-$1,389
Week of therapy intervention 12 12 – 8 $1,076-$2,863
% of patients that are non-molecular responders and have CR/PR/SD 25.5 20.4 – 30.6 $670-$1,390
% of patients that are molecular responders and have PD 9.8 7.8 – 11.8 $1,342-$832