FDG-PET Is An Independent Predictor for Survival in Primary Central Nervous System Lymphoma

Abstract 2687 Background: Primary central nervous system lymphoma (PCNSL) is an aggressive non-Hodgkin lymphoma confined to the CNS compartment at presentation. Treatment strategies mainly comprise high-dose methotrexate (HD-MTX) based protocols. The value of [18F] fluorodeoxyglucose positron emission tomography (FDG-PET) for the evaluation of PCNSL has only sparsely been investigated so far. Aim of this retrospective study was to investigate the prognostic value of pretreatment FDG-PET in PCNSL patients regarding tumor response and patient survival. Patients and Methods: Immunocompetent patients with biopsy proven PCNSL who received an FDG-PET before start of treatment were included. All patients had contrast-enhanced brain magnet resonance imaging scans for baseline and tumour response evaluation as recommended by the International Primary CNS Lymphoma Collaborative Group (Abrey et al. 2005). PET examinations were carried out using a Siemens ECAT EXACT 922/47 scanner after iv-injection of 366+55 MBq FDG. Tumor mean and maximum standardized uptake values (SUV) were assessed by volume of interest (VOI) analyses employing an automatic isocontour definition (80% of VOI maximum; PMOD software, PMOD Technologies Ltd.). In addition, a visual grading system was used to linearly grade tumor uptake by means of a simple, custom-made 6-step colour-scale: The maximum threshold of this colour-scale was individually adjusted to display physiological cerebellar uptake (reference region) as white (range 10%-20% of maximum; grade 1). FDG uptake below cerebellar uptake is colour-coded as black (<10% of maximum; grade 0), whereas FDG uptake above cerebellar uptake was colour-coded in four discrete steps (20–40%, 40–60%, 60–80% and >80% of maximum, grade 2–5). We used the chi-square test for contingency table tests, the log-rank test to compare survival probabilities, and multivariable Cox regression models to investigate the prognostic impact of pretreatment tumor grades on overall and progression free survival (OS and PFS). Results: 42 immunocompetent patients (23/19 female/male, median age 65 [range 38–83]) were included. All patients were treated according to HD-MTX based protocols. Median maximum SUV of the lymphomas was 10.48 (range 3.1–22.8). Distribution of patients according to FDG uptake: grade 1 N=13 (31%); grade 2 N=17 (40.5%); grade 3 N=8 (19%); grade 4 N=3 (7.1%); grade 5 N=1 (2.4%). 90% of patients responded to therapy; distribution of response according to FDG uptake: grade 1 31%, grade 2 33%; grade 3 19%; grade 4 7.1% and grade 5 0%. Overall, response was inversely associated with FDG uptake (p=0.012). Median OS and PFS of the entire cohort were 59 and 22.5 months, respectively. Patients with FDG uptake grade ≥3 had a significantly shorter OS and PFS (median both 11 months) compared to patients with lower FDG uptake (median OS and PFS not reached; p= 0.0441 and p=0.0219, respectively). After adjustment for age and performance status, FDG uptake was still independently associated with decreased OS (Hazard ratio [HR] 1.77, p=0.028) and PFS (HR 1.83, p=0.016). Conclusions: The present study demonstrates that higher pretreatment PCNSL glucose metabolism is inversely correlated with treatment response and, thus, represents an independent negative outcome predictor with respect to OS and PFS. Disclosures: No relevant conflicts of interest to declare.