Table IV. Summary of studies examining the prognostic role of BCL6 in PCNSL.

Reference	Patients (n)	Treatment	Median OS of all patients (months)	Antibody	Cut-off for positivity (%)	% Positive	Association between BCL6 positivity and PFS (p=)	Association between BCL6 positivity and OS (p=)
Chang, et al (2003)	14	CT*	NR	Santa Cruz Biotechnology (Dallas, TX, USA)	20	57	NR	0.16#
Camilleri-Broet, et al (2006)	83	HDMTX	42	Clone PG-B6p; Dako (Carpinteria, CA, USA)	30	56	NS	NS
Momota, et al (2010)	27	HDMTX	Not reached	Polyclonal; Dako (Carpinteria, CA, USA)	30	48	0.038#	0.124#
Rubenstein, et al (2013)	26	HDMTX	Not reached	Clone PG-B6p; Dako (Carpinteria, CA, USA)	60	59	0.019#&	0.009#
Lin, et al (2006)	29	HDMTX**	19.8	Clone PG-B6p; Dako (Carpinteria, CA, USA)	20	61	NR	0.073##
Braaten, et al (2003)	33	HDMTX	101	Clone PG-B6p; Dako (Carpinteria, CA, USA)	10	79	NR	0.002##
Levy, et al (2008)	48	HDMTX	34.6	Novocastra (Buffalo Grove, IL, USA)	50	46	0.02##	0.18##
Present study	49	HDMTX***	23.8	Clone PG-B6p; Dako (Carpinteria, CA, USA)	30	56	0.004##	0.05##

OS, overall survival; PFS, progression-free survival; CT, chemotherapy; NR, not reported; NS, not statistically significant; HDMTX, high dose methotrexate-based chemotherapy.

^*Chemotherapy not specified

^**Bomes regimen (Cheng, et al 1998) in majority of patients

^***HDMTX with rituximab, temozolomide, etoposide and cytarabine

^#Shorter OS or PFS with BCL6 positivity;

^##Longer OS or PFS with BCL6 positivity;

^&- analysed as continuous variable