104 The utility of multiparametric MRI in reducing diagnostic uncertainty for CNS lymphoma

Abstract

Diagnostic uncertainty represents a significant barrier to the timely initiation of treatment for central nervous system lymphoma (CNSL). Our previous 10-year retrospective service evaluation project demonstrated that concordance on multiparametric MRI (MPMRI) could be used as a radiological predictor of CNSL in uncertain cases. University Hospitals Birmingham (UHB) subsequently introduced the Lymphoma Pathway, which states that MPMRI should be requested for uncertain radiological CNSL cases. We stated that radiological certainty using MPMRI might make biopsy redundant in selected cases – those with known biopsy-confirmed systemic high-grade lymphoma and a typical brain lesion, those unfit for biopsy, and those with previous biopsy-confirmed primary CNSL with suspected relapse. We present a prospective evaluation of the utility of MPMRI in patients with an uncertain diagnosis of CNSL following introduction of this pathway. All cases of MPMRI conducted between December 2024 and September 2025 at UHB following the introduction of the UHB Lymphoma Pathway were prospectively reviewed. In nine recorded cases, CNSL was diagnosed based on MPMRI or histologically following MPMRI. Four cases proceeded to biopsy, two of them histologically conclusive for lymphoma. Two cases were empirically treated for primary CNSL following MP MRI without histological diagnosis due to poor surgical performance status and infeasible biopsy; three cases with known systemic lymphoma were treated for CNSL following MPMRI without biopsy. Uptake of MPMRI has increased following the introduction of the UHB Lymphoma Pathway. Although low patient numbers do not support definitive statistical analysis, more patients underwent empirical treatment for CNSL following MP MRI, suggesting increased reliance on advanced radiological modalities. Ongoing prospective data should be collected to assess the accuracy, sensitivity and specificity of MPMRI in the diagnosis of PCNSL.