High-Grade B-Cell Lymphoma, Not Otherwise Specified (HGBL, NOS): Characteristics, Treatment, and Outcomes from 17 Academic US Centers

Background: The term HGBL, NOS was introduced by the World Health Organization (WHO) in 2016 for aggressive B-cell lymphomas with Burkitt lymphoma-like (BLL) or blastoid cytomorphology that lack double-hit genetics and do not meet criteria for other entities. Diagnostic patterns and prognosis of these rare tumors are not well understood. We examined the characteristics and outcomes of patients (pts) with HGBL, NOS diagnosed in 17 academic centers across the United States.

Methods: We collected retrospective data on HGBL, NOS cases diagnosed by academic hematopathologists in 2017-2021; 8 centers performed a local review by lymphoma pathology experts to confirm fulfillment of the WHO criteria; pathology reports were reviewed centrally. We excluded pts not tested for MYC rearrangement (MYC-R), any double/triple-hit, diffuse large B-cell, or lymphoblastic lymphomas. Immunohistochemistry (IHC) and cytogenetic tests were done locally. Outcomes included rates of complete response (CR), progression-free (PFS) and overall survival (OS) estimated with 95% confidence intervals (CI).

Results: Among 126 pts with HGBL, NOS, median age was 64 years (range 18-91), 67% were male, and 3 were HIV+. Advanced stage was present in 68%, poor performance status (PS, ECOG ≥2) in 21%, high serum lactate dehydrogenase (LDH) in 68%, extranodal (EN) sites in 79%, central nervous system (CNS) involvement in 6%, and International Prognostic Index (IPI) ≥ 3 in 55%.

Cytomorphology was reported as BLL in 59 (47%) cases, blastoid in 28 (22%), and unspecified in 39 (31%). By IHC, 83% had germinal center B-cell (GCB) phenotype. Using cases with available data, CD10 was expressed in 79%, BCL6 in 81%, MUM1/IRF4 in 48%, MYC in 73%, BCL2 in 55% (dual MYC/BCL2 expressor [DEL]: 37%), CD5 in 13%, and median Ki-67 was 95%. MYC-R (single-hit) was detected in 27% (Fig A), MYC extra copies (EC) in 9%, BCL2-R in 13%, and BCL6-R in 10%. MYC-EC were present in 16% of cases with BCL2-R or BCL6-R, and BCL2/BCL6-EC in 12% of those with MYC-R. Blastoid tumors were more likely than BLL to involve >1 EN site or to have BCL2-R (Fig B). 9 cases were assessed by next generation sequencing and 5 (56%) had a TP53 mutation. Cases which underwent confirmatory pathology review (N=74) did not differ from others clinically but more often had a well-defined HGBL morphology (77% vs 58%, P=.031) and less often MYC-R (20% vs. 37%, P=.004).

The most common first-line regimens (among treated pts, N=121) were DA-EPOCH-R (50%) and RCHOP (35%), with few pts receiving HyperCVAD/MA (5%) or CODOX-M±IVAC (2%); 97% received rituximab, and 44% CNS prophylaxis. Pts selected for DA-EPOCH-R vs. RCHOP were younger (median 61 vs. 68 years, P=.006), more often had stage 3/4 (P=.04), BLL morphology (56% vs. 29%, P=.009) or MYC-R (31% vs. 14%, P=.06). CR was attained in 62% of pts, whereas 20% had progressive disease. The most frequent salvage regimens (± rituximab) included ICE (N=12), DHAP (N=6), and GemOx (N=5). 3 pts underwent autologous, and 3 allogeneic transplant (2/3 subsequently relapsed). 13 received chimeric antigen receptor (CAR) T-cells, with response noted in 7 (54%) and CR in 4 (31%); HGBL relapsed in 3/7 (43%) responders.

With median follow-up of 2.7 years, 39% of pts relapsed, and 33% died. Of 49 observed relapses, 13 (27%) involved the CNS. PFS estimate at 2 years was 51% (95% CI, 42-60%) and OS was 68% (95% CI, 58-76%; Fig C). PFS and OS were not significantly associated with age or PS, but stage and LDH were prognostic (Fig D-G). Furthermore, PFS did not differ by BLL/blastoid morphology, MYC-R status or DEL status, but non-GCB tumors had somewhat worse PFS (Fig H-J). We observed no significant PFS (or OS) difference between pts selected for RCHOP vs. DA-EPOCH-R (P=.83 for PFS, Fig K; P=.55 for OS) in aggregate or in any subset, except for de novo tumors with BLL morphology (N=41), where DA-EPOCH-R showed a superior 2-year PFS (73% vs 38% for RCHOP, P=.027; stratified by IPI: P=.040, Fig L).

Conclusions: HGBL, NOS, as diagnosed in current academic practice, is highly heterogeneous, highlighting the need to classify high-grade lymphomas using molecular rather than morphologic features. Considering poor survival in all age groups (except for few pts with early stage and normal LDH), lack of prognostic significance of MYC-R, DEL status, or cytomorphology, HGBL, NOS needs prospective trials to delineate prognostic biomarkers, the role of intensified chemotherapy, and novel therapeutic approaches.

Figure 1

View largeDownload slide

Figure 1

View largeDownload slide

Close modal