Cost-Effectiveness Analysis of Routine Surveillance Imaging of Patients With Diffuse Large B-Cell Lymphoma in First Remission

Abstract

PURPOSE: Surveillance imaging of asymptomatic patients with diffuse large B-cell lymphoma (DLBCL) in first remission remains controversial. A decision-analytic Markov model was developed to evaluate the cost-effectiveness of follow-up strategies following first-line immunochemotherapy. PATIENTS AND METHODS: Three strategies were compared in 55-year-old patient cohorts: routine clinical follow-up without serial imaging, routine follow-up with biannual computed tomography (CT) scans for 2 years, or routine follow-up with biannual [(18)F]-fluorodeoxyglucose positron emission tomography-computed tomography (PET/CT) for 2 years. The baseline model favored imaging-based strategies by associating asymptomatic imaging-detected relapses with improved clinical outcomes. Lifetime costs, quality-adjusted life years (QALYs), and incremental cost-effectiveness ratios (ICERs) were calculated for each surveillance strategy. RESULTS: Surveillance strategies utilizing 2 years of routine CT or PET/CT scans were associated with minimal survival benefit when compared with clinical follow-up without routine imaging (life-years gained: CT, 0.03 years; PET/CT, 0.04 years). The benefit of imaging-based follow-up remained small after quality-of-life adjustments (CT, 0.020 QALYs; PET/CT, 0.025 QALYs). Costs associated with imaging-based surveillance strategies are considerable; ICERs for imaging strategies compared with clinical follow-up were $164,960/QALY (95% CI, $116,510 to $766,930/QALY) and $168,750/QALY (95% CI, $117,440 to 853,550/QALY) for CT and PET/CT, respectively. Model conclusions were robust and remained stable on one-way and probabilistic sensitivity analyses. CONCLUSION: Our cost-effectiveness analysis suggests surveillance imaging of asymptomatic DLBCL patients in remission offers little clinical benefit at substantial economic costs.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 69 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.