Topline Presentation Points

Concepts & patterns of CNS relapse 

  • Utility and limitations of the CNS International Prognostic Index (CNS-IPI) 
    • Risk model for CNS relapse in patients with DLBCL treated with R-CHOP 
  • The denominator & predicting events
  • Impact of the intervention 
    • A positive predictive value of 10% equates to a number needed to treat’ (NNT) of 10 patients  
    • A prophylactic intervention may reduce (but won’t eliminate) risk 
    • Assume a given intervention may reduce risk by 50%…. 
  • To inform CNS risk 
    • Cell of origin by GEP (Klanova et al, Blood 2019, Fox CP Blood 2019)
  • DLBCL genomic classification to inform CNS risk (Wright GW et al Cancer Cell 2020)
  • ctDNA as a predictive biomarker? 
    • Emerging data on utility of cell free circulating tumour (ct) DNA in CSF and plasma (Bobillo et al, Haematologica 106; 2021)


Approaches to CNS prophylaxis 

  • Majority of CNS relapses involve the brain parenchyma 
    • Insufficient evidence to support Intrathecal (IT) prophylaxis  
    • Intravenous HD-MTX commonly used but not supported by strong evidence – no RCT data (Ferreri, et al 2004, Boehme, et al 2009, Bernstein, et al 2009, Cheah, et al 2014, Eyre et al, Hematologica 2020)

CNS prophylaxis: a look to the future…..  

  • When prophylaxis fails… 
    • Rationale for the Phase II MARIETTA trial (IELSG42) 
    • MATRIX-RICE followed by ASCT achieved the primary endpoint without major safety concerns 
    • Response to 2x MATRIX was a strong prognostic factor. 
    • Patients with SCNSL refractory to MATRIX did not benefit from RICE + ASCT. 
    • The best survival outcomes were achieved in chemo-naïve patients and in patients with disease limited to a single CNS compartment (parenchyma or CSF/meninges)