Contributions
Abstract: PB1765
Type: Publication Only
Background
Diagnosis of lymphomas with DLBCL morphology and MYC, BCL2 and/or BCL6 breaks (double hit, DH) is rare but relevant given the poor outcome with conventional treatment and possible treatment-escalation benefits suggested by retrospective studies.
Aims
To characterize the impact of MYC, BCL2 and BCL6 breaks in treatment decisions in DLBCL patients (pts) enriched for immunophenotypically defined germinal center type (GC) and to evaluate the relationship between breaks and MYC and BCL2 expression in a single center.
Methods
All DLBCL biopsies tested by Fluorescence in situ hybridization (FISH) for MYC, BCL2 and BCL6 breaks between 2010-2017 were reviewed. Cell of origin was defined according to Hans algorithm. Positivity cut-offs for MYC and BCL2 were ≥40% and ≥50%, respectively. FISH was mostly done in pre-selected GC cases and in additional pts with unusual morphological and/or clinical features. Pts characteristics and outcomes were reviewed and groups compared by Fisher exact test. The impact of FISH on clinical decisions was assessed as the number of pts needed to screen (NNS) in order to intensify treatment in one, calculated in two scenarios using the prevalence of DH and the number of treatment changes informed by FISH. The NNS to avoid one progression/death was calculated based on published data in DH pts comparing PFS with aggressive treatments versus RCHOP according to Altman et al (B Med J 1999)
Results
76 biopsies (73 diagnosis, 3 relapse) (9% of all DLBCL) underwent FISH; the proportion of pts analyzed increased over time, the majority (58%) after the publication of the revised WHO classification draft. The population (49% male, median age 57, 66% stage III/IV, 39% intermediate-high and high-risk IPI) included 43% pts older than 60 years, 16% transformed lymphomas and 30% pts with significant co-morbidities, including 9 previously exposed to anthracyclines. 85% were GC and 16% were double MYC/BCL2 expressors (DE). 16% (95%CI 8%>26%) had DH (6 MYC/BCL2, 5 MYC/BCL6 and 1 MYC/BCL2/BCL6), more frequently in the DE group (45% versus 9% in non-DE, p=0.01). 75% pts received RCHOP and 12% had aggressive regimens (mostly LMB96) based on clinical risk factors, age, fitness and FISH. Out of 12 DH cases 3 had aggressive regimens, only in one due to the FISH result; reasons for not escalating treatment in others included age, comorbidities and prior anthracyclines. Only 35/76 pts were fit for aggressive regimens; of those 3 were DH (9%, 95%CI 2-23%) and treatment was escalated in 2. Assuming a 9% prevalence of DH in fit pts and that 2/3 would change treatment based on FISH, the NNS for one therapeutic change would decrease from 76 to 17. For a median 8-month PFS with RCHOP (Petrich et al, Blood 2014), a 47% reduction of PFS events with aggressive regimens (HR=0.53; 95% CI 0.29-0.98) (Howlett et al, Br J Haematol 2015), a 16% prevalence of DH and 8% therapeutic changes, 469 pts (95% CI 313-11250) would need testing to avoid one progression/death at 8 months. If selection criteria for FISH included only fit pts, this would decrease to 102 (95% CI 68-2425).
Conclusion
This retrospective single center study suggests that from the therapeutic perspective DLBCL candidates for FISH can be restricted considering clinical information added to cell-of-origin and/or protein expression. This impacts on costs and feasibility of large scale testing while new targeted therapies with acceptable toxicity profiles for a predominantly elderly/co-morbid population are unavailable.
Session topic: 21. Aggressive Non-Hodgkin lymphoma - Clinical
Keyword(s): FISH, lymphoma, Treatment
Abstract: PB1765
Type: Publication Only
Background
Diagnosis of lymphomas with DLBCL morphology and MYC, BCL2 and/or BCL6 breaks (double hit, DH) is rare but relevant given the poor outcome with conventional treatment and possible treatment-escalation benefits suggested by retrospective studies.
Aims
To characterize the impact of MYC, BCL2 and BCL6 breaks in treatment decisions in DLBCL patients (pts) enriched for immunophenotypically defined germinal center type (GC) and to evaluate the relationship between breaks and MYC and BCL2 expression in a single center.
Methods
All DLBCL biopsies tested by Fluorescence in situ hybridization (FISH) for MYC, BCL2 and BCL6 breaks between 2010-2017 were reviewed. Cell of origin was defined according to Hans algorithm. Positivity cut-offs for MYC and BCL2 were ≥40% and ≥50%, respectively. FISH was mostly done in pre-selected GC cases and in additional pts with unusual morphological and/or clinical features. Pts characteristics and outcomes were reviewed and groups compared by Fisher exact test. The impact of FISH on clinical decisions was assessed as the number of pts needed to screen (NNS) in order to intensify treatment in one, calculated in two scenarios using the prevalence of DH and the number of treatment changes informed by FISH. The NNS to avoid one progression/death was calculated based on published data in DH pts comparing PFS with aggressive treatments versus RCHOP according to Altman et al (B Med J 1999)
Results
76 biopsies (73 diagnosis, 3 relapse) (9% of all DLBCL) underwent FISH; the proportion of pts analyzed increased over time, the majority (58%) after the publication of the revised WHO classification draft. The population (49% male, median age 57, 66% stage III/IV, 39% intermediate-high and high-risk IPI) included 43% pts older than 60 years, 16% transformed lymphomas and 30% pts with significant co-morbidities, including 9 previously exposed to anthracyclines. 85% were GC and 16% were double MYC/BCL2 expressors (DE). 16% (95%CI 8%>26%) had DH (6 MYC/BCL2, 5 MYC/BCL6 and 1 MYC/BCL2/BCL6), more frequently in the DE group (45% versus 9% in non-DE, p=0.01). 75% pts received RCHOP and 12% had aggressive regimens (mostly LMB96) based on clinical risk factors, age, fitness and FISH. Out of 12 DH cases 3 had aggressive regimens, only in one due to the FISH result; reasons for not escalating treatment in others included age, comorbidities and prior anthracyclines. Only 35/76 pts were fit for aggressive regimens; of those 3 were DH (9%, 95%CI 2-23%) and treatment was escalated in 2. Assuming a 9% prevalence of DH in fit pts and that 2/3 would change treatment based on FISH, the NNS for one therapeutic change would decrease from 76 to 17. For a median 8-month PFS with RCHOP (Petrich et al, Blood 2014), a 47% reduction of PFS events with aggressive regimens (HR=0.53; 95% CI 0.29-0.98) (Howlett et al, Br J Haematol 2015), a 16% prevalence of DH and 8% therapeutic changes, 469 pts (95% CI 313-11250) would need testing to avoid one progression/death at 8 months. If selection criteria for FISH included only fit pts, this would decrease to 102 (95% CI 68-2425).
Conclusion
This retrospective single center study suggests that from the therapeutic perspective DLBCL candidates for FISH can be restricted considering clinical information added to cell-of-origin and/or protein expression. This impacts on costs and feasibility of large scale testing while new targeted therapies with acceptable toxicity profiles for a predominantly elderly/co-morbid population are unavailable.
Session topic: 21. Aggressive Non-Hodgkin lymphoma - Clinical
Keyword(s): FISH, lymphoma, Treatment