Contributions
Abstract: PB1752
Type: Publication Only
Background
A continued risk of relapse in limited-stage diffuse large B-cell lymphoma (DLBCL) has been reported (Stephens et al. J Clin Oncol 2016). However, the characteristics of patients (pts) who develop late relapse (LR) are not well known.
Aims
To investigate the clinico-pathological features of pts with DLBCL who develop LR and its possible mechanisms.
Methods
Of 339 pts with limited-stage (Ann Arbor stage I/II) de novo DLBCL diagnosed between 1997 and 2012, 21 pts who had achieved a complete response to initial therapy but relapsed after >5 years since the initial diagnosis (i.e. LR) underwent clinico-pathological analyses. Immunohistological analyses were performed on pts who relapsed with a DLBCL histology and whose samples were available. Cell of origin (COO) subtypes were classified by Hans algorithm. Programmed death–ligand 1 (PD–L1) expression on tumor cells and tumor-infiltrating cells (TICs) were evaluated by double staining for PAX5 and PD–L1. Cut-off values for the PD–L1(+) stain on tumor cells and TICs were 30% and 20%, respectively (Kiyasu et al. Blood 2015). Fluorescence in situ hybridization (FISH) analysis using probes mapping to 9p24.1 and centromere 9 was performed on PD–L1(+) tumor cells. The cut-off value for MYC(+) was 40%. CD163 or PD–1(+) TIC numbers were counted manually in three representative fields (0.14 mm2 area); the average number of positive cells per unit area in samples at initial diagnosis and LR were then compared.
Results
The characteristics of the 21 pts at initial diagnosis were as follows: 60 years median age (range, 32–77); 13 (62%) had stage I disease; 12 (57%) had extranodal disease; and frequent diseases in gingival, sinonasal/nasal cavities and testes (Table). The majority of pts (n=14, 70%) had a non-germinal–center B-cell–like (non-GCB) subtype. As an initial treatment, CHOP (median, 5 cycles; range, 3–8), with (n=18) or without (n=2) radiation therapy (RT), was performed. One pt received other intensive chemotherapy. Rituximab was administered to eight pts. The median duration from initial diagnosis to LR was 7 years (range, 5–18) and the median follow-up time for survivors was 11 years. Systemic relapses (n=15, 71%) were more common than local ones. Nine pts (43%) had more than one extranodal disease; skin/subcutaneous and bone/bone marrow diseases were frequent. Of 18 pts who had received RT, 12 pts relapsed outside of the RT field only, and six relapsed both within and outside the RT field. One pt relapsed showing a composite histology of DLBCL and a low-grade B-cell lymphoma (LGBCL), and two pts relapsed with only a LGBCL histology. Four pts had a different COO subtype from that at initial diagnosis. At initial diagnosis, MYC was positive in six pts and negative in eight pts; at LR, it turned positive in five pts, all of whom relapsed with stage IV disease, although it turned negative in another five pts. Interestingly, two pts showed PD–L1(+) tumor cells at LR despite being negative at initial diagnosis. FISH analysis revealed a 9p24.1 gain in one of the two pts. More PD–1(+) TICs tended to be observed at LR than at initial diagnosis, although this was not statistically significant (p=0.13). No clear trend in the number of PD–L1(+) or CD163(+) TICs was found.
Conclusion
LR patterns of advanced-stage disease, multiple extranodal diseases and non-GCB subtype were frequently seen in limited-stage DLBCL. This study suggests heterogeneity as a LR mechanism, in which, in addition to the well-known mechanism of relapse with LGBCL components, we speculate functions associated with immune evasion or cell proliferation are acquired.
Session topic: 21. Aggressive Non-Hodgkin lymphoma - Clinical
Keyword(s): DLBCL, Immunity, MYC, Relapsed lymphoma
Abstract: PB1752
Type: Publication Only
Background
A continued risk of relapse in limited-stage diffuse large B-cell lymphoma (DLBCL) has been reported (Stephens et al. J Clin Oncol 2016). However, the characteristics of patients (pts) who develop late relapse (LR) are not well known.
Aims
To investigate the clinico-pathological features of pts with DLBCL who develop LR and its possible mechanisms.
Methods
Of 339 pts with limited-stage (Ann Arbor stage I/II) de novo DLBCL diagnosed between 1997 and 2012, 21 pts who had achieved a complete response to initial therapy but relapsed after >5 years since the initial diagnosis (i.e. LR) underwent clinico-pathological analyses. Immunohistological analyses were performed on pts who relapsed with a DLBCL histology and whose samples were available. Cell of origin (COO) subtypes were classified by Hans algorithm. Programmed death–ligand 1 (PD–L1) expression on tumor cells and tumor-infiltrating cells (TICs) were evaluated by double staining for PAX5 and PD–L1. Cut-off values for the PD–L1(+) stain on tumor cells and TICs were 30% and 20%, respectively (Kiyasu et al. Blood 2015). Fluorescence in situ hybridization (FISH) analysis using probes mapping to 9p24.1 and centromere 9 was performed on PD–L1(+) tumor cells. The cut-off value for MYC(+) was 40%. CD163 or PD–1(+) TIC numbers were counted manually in three representative fields (0.14 mm2 area); the average number of positive cells per unit area in samples at initial diagnosis and LR were then compared.
Results
The characteristics of the 21 pts at initial diagnosis were as follows: 60 years median age (range, 32–77); 13 (62%) had stage I disease; 12 (57%) had extranodal disease; and frequent diseases in gingival, sinonasal/nasal cavities and testes (Table). The majority of pts (n=14, 70%) had a non-germinal–center B-cell–like (non-GCB) subtype. As an initial treatment, CHOP (median, 5 cycles; range, 3–8), with (n=18) or without (n=2) radiation therapy (RT), was performed. One pt received other intensive chemotherapy. Rituximab was administered to eight pts. The median duration from initial diagnosis to LR was 7 years (range, 5–18) and the median follow-up time for survivors was 11 years. Systemic relapses (n=15, 71%) were more common than local ones. Nine pts (43%) had more than one extranodal disease; skin/subcutaneous and bone/bone marrow diseases were frequent. Of 18 pts who had received RT, 12 pts relapsed outside of the RT field only, and six relapsed both within and outside the RT field. One pt relapsed showing a composite histology of DLBCL and a low-grade B-cell lymphoma (LGBCL), and two pts relapsed with only a LGBCL histology. Four pts had a different COO subtype from that at initial diagnosis. At initial diagnosis, MYC was positive in six pts and negative in eight pts; at LR, it turned positive in five pts, all of whom relapsed with stage IV disease, although it turned negative in another five pts. Interestingly, two pts showed PD–L1(+) tumor cells at LR despite being negative at initial diagnosis. FISH analysis revealed a 9p24.1 gain in one of the two pts. More PD–1(+) TICs tended to be observed at LR than at initial diagnosis, although this was not statistically significant (p=0.13). No clear trend in the number of PD–L1(+) or CD163(+) TICs was found.
Conclusion
LR patterns of advanced-stage disease, multiple extranodal diseases and non-GCB subtype were frequently seen in limited-stage DLBCL. This study suggests heterogeneity as a LR mechanism, in which, in addition to the well-known mechanism of relapse with LGBCL components, we speculate functions associated with immune evasion or cell proliferation are acquired.
Session topic: 21. Aggressive Non-Hodgkin lymphoma - Clinical
Keyword(s): DLBCL, Immunity, MYC, Relapsed lymphoma