PURPOSE OF THE REPORT: The usefulness of brain 18 F-FDG PET/CT in primary central nervous system lymphoma (PCNSL) remains underexplored. This study investigated whether early metabolic responses in interim brain FDG PET/CT serve as a prognostic indicator of PCNSL treatment outcomes. PATIENTS AND METHODS: This prospective study included 53 patients with PCNSL who underwent a high-dose methotrexate-based treatment. Brain FDG PET was performed at diagnosis (baseline PET) and after induction chemotherapy (interim PET), assessing interim PET parameters such as the highest maximum standardized uptake value (hSUV max ), sum of SUV max (sumSUV max ), highest tumor-to-normal ratio (hTNR max ), sum of TNR max (sumTNR max ), highest metabolic tumor volume (MTV) (hMTV), and sum of MTV (sumMTV) across all PET-positive lesions. RESULTS: High interim hTNR max (hazards ratio: 9.76, 95% confidence interval: 1.90-50.11, P = 0.01) was an independently significant predictor of poor progression-free survival in multivariate analysis. Patients with low interim hTNR max (≤1.0) had a significantly longer median progression-free survival than those with high interim hTNR max (>
1.0) (25.0 vs 3.6 months, P <
0.002). Incorporating interim MRI-based clinical response assessments and hTNR max allowed the classification of partial response subgroups with markedly different prognoses ( P <
0.002). High interim hTNR max (hazards ratio: 2.76, 95% confidence interval: 1.39-5.48, P = 0.004) was an independently significant predictor of poor overall survival in multivariate analysis. CONCLUSIONS: The hTNR max measurement from interim brain FDG PET scans emerges as an important prognostic marker in PCNSL. These findings underscore the potential of interim FDG PET evaluations to refine response assessments and inform tailored therapeutic strategies.