The Clinical Impact and Cost-Effectiveness of PET-Based and Ctdna-Based Evaluation at End-of-Therapy after Frontline Treatment of DLBCL

Background While the majority of diffuse large B-cell lymphoma (DLBCL) patients are cured with frontline therapy, up to 30-40% of patients will have refractory or relapsed (R/R) disease with primary refractory/early relapse patients often considered for second-line chimeric antigen receptor (CAR) T cell therapy. At the end of therapy (EOT), a positron emission tomography-computed tomography scan (PET-CT) is the standard to determine the response to treatment and remission status defined by Lugano response criteria. Several studies have demonstrated that a complete response (CR) at EOT by PET-CT is predictive of long-term survival outcomes after first-line treatment. However, PET-CT is associated with high false positivity, which may lead to over-treatment of patients and unnecessary treatment-related adverse events and financial toxicities. Guidelines recommend confirmation of positive PET-CT with tissue biopsy before salvage therapy, though it is only sometimes followed in real-world practice or required by clinical trials. Likewise, invasive biopsies are not always possible given anatomic location or other factors. Circulating tumor DNA (ctDNA) is an emerging biomarker for predicting outcomes in DLBCL that may improve the EOT response determination as a marker of measurable residual disease (MRD). This study aims to evaluate the cost-effectiveness of different approaches to evaluate response to frontline DLBCL treatment. Methods We modeled a hypothetical cohort of US adults (mean age, 62 years) with treatment-naïve DLBCL by developing a Markov model (lifetime horizon) to model the clinical impact and cost-effectiveness of three different strategies to assess EOT response assessment (PET-CT alone, PET-CT with confirmatory biopsy for positive results, and PET-CT with confirmatory ctDNA for positive results). We used sensitivity and specificity estimates from the published literature for PET-CT, ctDNA, and biopsy for primary refractory disease. Using the 12-month PFS as a surrogate for primary refractory disease, we used a sensitivity (Sn) of 46.3% and specificity (Sp) of 89.7% for PET-CT based on the GOYA study. We used a Sn of 89% and a Sp of 98% for invasive biopsy based on a large cohort study of EOT biopsy. As ctDNA is an emerging technology, we assumed a baseline theoretical Sn and Sp of 60% and 97%, respectively. We also performed sensitivity analyses for the Sn and Sp rates of PET-CT and ctDNA, ranging from 20% to 99%. Progression rates and OS for front-line therapy with R-CHOP were estimated from the literature. Patients with primary refractory disease or relapse within 12 months received 2nd-line CAR-T using 3-year outcomes modeled from the ZUMA-7 study. Patients with relapse after 12 months received chemoimmunotherapy followed by autologous stem cell transplant if they achieved CR or third-line CAR T if they had ongoing disease. Outcome measures were reported in incremental cost-effectiveness ratios (ICERs), with a willingness-to-pay (WTP) threshold of $150,000/quality-adjusted life-year (QALY), as well as 5-year OS, PFS, and the proportion of patients who receive 2nd-line CAR-T. Results PET/CT with confirmatory testing using either tissue biopsy or ctDNA was less expensive and more effective) than PET-CT at EOT alone. (i.e. absolutely dominated) Thus, this strategy was removed from the analysis. Confirmatory testing with ctDNA (for a positive PET/CT) with the baseline assumed Sn and Sp was the dominant strategy (i.e., less costly and more effective than PET/CT with confirmatory biopsy). PET-CT with confirmatory ctDNA remained the cost-effective strategy as long as the the specificity of ctDNA was greater than 95.2%, indicating that the cost-effectiveness of confirmatory ctDNA is primarily sensitive to test specificity. Confirmatory testing with ctDNA remained cost-effective at PET/CT sensitivities ranging from 20 to 90%. Conclusion Confirmatory tissue biopsy testing for positive PET-CT reduces unnecessary use of 2nd-line CART and lowers healthcare costs. Indeed, confirmatory biopsy for positive PET-CT for the 32,000 new cases of DLBCL in the US would save $3.6 billion annually. Confirmatory ctDNA testing for positive PET-CT may be cost-effective compared to tissue biopsy if high specificity of ctDNA is maintained. This highlights that test specificity is critical for confirmatory testing for positive PET-CTs at EOT.