Abstract Primary Central Nervous System Lymphoma (PCNSL) is a rare aggressive non-Hodgkin lymphoma involving exclusively the central nervous system (CNS). The majority of PCNSLs are diffuse large B-cell lymphomas (DLBCL), but treatment and prognosis differ from systemic DLBCL due to differences in biology and the difficulty of delivering effective therapies with high penetration across the blood-brain barrier (BBB). While PCNSL often responds to initial therapy, relapses are common even after achieving a complete remission. The aims of this study were to estimate the risk of death or relapse and the loss of life expectancy in PCNSL after primary treatment with high-dose methotrexate (HD-MTX) containing regimens. Outcomes were assessed at baseline and for patients reaching pre-defined milestones of progression-free survival (PFS). Data on PCNSL patients were extracted from the nationwide Danish lymphoma register. The inclusion criteria were I) histologically-proven DLBCL morphology, II) involvement restricted to parenchymal or leptomeningeal CNS involvement without ocular involvement, III) treatment protocols containing HD-MTX, and IV) diagnosis between 2000-2017. PFS was defined as the time from diagnosis until death, relapse/progression, or end-of-treatment response assessment for patients with stable or progressive disease at the response assessment. The five-year PFS event probability risk was estimated for all patients and conditional on patients reaching different PFS milestones. The five-year restricted loss of lifetime (5y-RLEL) was defined as the numeric difference in the number of days patients and individuals from a background population are expected to live in the following five year period. This was estimated for all patients and for subsets of patients free of PFS events after one (PFS1), two (PFS2), or three (PFS3) years. Additionally, the results were stratified according to gender, ECOG performance status 0-1/> 1, elevated LDH status, treatment with/without rituximab, and age at diagnosis ≤60/>60 years. The survival of an age- and gender-matched general population was calculated by using life tables from the Human Mortality Database. In total 253 patients were included in the analyses; 60% were male, median age at diagnosis was 66 (range 27 - 85), 46% had an ECOG performance status > 1, and 33% had elevated LDH levels. Consolidation therapy (radiotherapy and/or high-dose therapy with autologous stem cell transplantation) was used in 23% of patients and 36% received rituximab in first line. The median follow-up was 6.9 years (range 0.7 - 17.7), the 5-year overall survival was 35% (95% CI 29-42), and the five-year PFS was 28% (95% CI 22-34). Patients reaching PFS1 had a 51% (95% CI 41-61) probability of a PFS event in the following five years (Figure 1A). After the PFS1 milestone, the five-year probability of a PFS event did not change substantially (Figure 1A) and the event probability remained high even after three years of PFS. On average, the PCNSL patients lost 2.2 living years (95% CI 1.9 - 2.4) in the five years after first pathologic diagnosis of PCNSL (Figure 1B). At PFS1, the 5y-RLEL decreased to 1.0 years (95% CI 0.7 - 1.3) (Figure 1B). The achievement of later PFS milestones only led to minor additional decreases in 5y-RLEL (PFS3: 0.7 years [95% CI 0.3 - 1.1]) (Figure 1B). The 5y-RLEL estimates were substantially larger for patients with an ECOG performance status > 1 vs patients with an ECOG performance status ≤ 1 (Figure 1B). Outcome differences between risk factor defined subgroups decreased after PFS1 and later PFS milestones (Figure 1B). The outlook of PCNSL patients treated with HD-MTX-based therapy improves significantly given a progression-free survival of one year, after which baseline adverse risk factors lose prognostic impact over time. However, in contrast to systemic DLBCL, survival does not normalize to the background population even after several years without PFS event. By the time of the ASH, updated results that include patients from the population-based lymphoma database in British Columbia (Canada) will be presented. Disclosures No relevant conflicts of interest to declare.