Survival and Risk of Central Nervous System Recurrence in Burkitt or High-Grade B-Cell Lymphoma in the DA-EPOCH-R Era

Background: The 2017 World Health Organization (WHO) classification distinguished new categories of high-grade B-cell lymphoma (HGBCL). Treatment of these lymphomas is in flux, as some were historically classified as DLBCL and treated with the RCHOP regimen, while others, akin to Burkitt lymphoma (BL), were treated using high-intensity regimens (e.g. CODOX-M/IVAC or hyper-CVAD) that include systemic high-dose methotrexate (HDMTX) as central nervous system (CNS) prophylaxis . Recently, the less intensive DA-EPOCH-R regimen has been increasingly applied for BL or HGBCL with concurrent MYC and BCL2 and/or BCL6 rearrangements based on phase 2 data (Dunleavy et al., NEJM 2013). We examined progression-free survival (PFS) and risk of CNS relapse among HGBCL/BL patients treated in our institution.

Methods: In this retrospective series from an academic center, we integrated cancer registry and electronic medical records for all patients treated for BL or HGBCL at Lifespan Cancer Institute in 2005-2017. We designated as "HGBCL" all cases with concurrent MYC and BCL2/BCL6 rearrangements, or those previously diagnosed as "B-cell lymphoma, unclassifiable, with features intermediate between DLBCL and BL" (per WHO 2008). We compared characteristics of patients treated with intensive regimens or with DA-EPOCH-R, as well as their PFS (using log-rank test) and cumulative incidence function (CIF) of CNS relapse (using Gray's test).

Results: Among 64 patients with BL (n=38) and HGBCL (n=26), those with BL were somewhat younger (median age 52 versus [vs.] 61 years), more often male (84% vs. 58%), HIV-positive (29% vs. 8%), or with CNS involvement at baseline (21% vs. 4%). Among HGBCLs, 58% had a MYC rearrangement, whereas 31% had concurrent MYC and BCL2/BCL6 rearrangements. Eight patients who did not receive chemotherapy (median age, 78 years) were excluded from outcome analysis. Compared with patients with BL, those with HGBCL more often received DA-EPOCH-R (41% vs. 15%) or R-CHOP (27% vs. 12%), and less often high-intensity regimens (32% vs. 73%, P=.027).

Compared with patients treated using high-intensity regimens, those treated with DA-EPOCH-R were significantly older (61 vs. 49 years, P=.023), with high/high-intermediate International Prognostic Index (IPI, 86% vs. 50%, P=.027), or diagnosis after 2010 (86% vs. 53%, P=.049). There was no difference in baseline CNS involvement (P=.68) or receipt of intrathecal prophylaxis (P=.16) between DA-EPOCH-R and high-intensity regimens.

After median follow-up of 5.7 years, we observed 12 recurrences, including 5 (42%) in the CNS. Median PFS was not reached, whereas 3-year PFS was 56% (95% confidence interval [CI], 42-68%), numerically better in BL than in HGBCL (63% vs. 45%, P=.33, Fig. A). Overall survival at 3 years was also 56% (95%CI, 41-68%). Factors associated with shorter PFS included age >60 years (log-rank P=.017), poor performance status (P<.001), high/high-intermediate IPI (P=.0003), and lack of CNS prophylaxis (P=.021). Treatment with DA-EPOCH-R rather than a high-intensity regimen was also associated with worse PFS (P=.001), but not when stratified by histology and age (P=.14). HIV status (P=.53) or CNS involvement at baseline (P=.15) were not prognostic. Survival after recurrence was dismal (median, 1 month, 95%CI, 0.2-3.4), despite 58% of patients receiving salvage therapy.

The 3-year CIF of CNS recurrence was 9% (95%CI, 3-18%), and higher in patients with CNS involvement at baseline (P=.002). All CNS recurrences occurred during the first year of follow-up and were among patients receiving DA-EPOCH-R (35.7% vs. 0% for other regimens, P=.0004). Administration of HDMTX for CNS prophylaxis was associated with a numerically lower risk of CNS recurrence (3% vs. 16% without, P=.09, Fig. B). Conversely, we observed no difference in CNS relapse with or without intrathecal prophylaxis (9% vs. 8%, P=.84, Fig. C).

Conclusions: The high proportion of CNS recurrences despite prophylaxis, and very poor outcomes at relapse, indicate persistent major areas of need in BL/HGBCL. Outcomes of DA-EPOCH-R were heavily influenced by selection bias (as evidenced by unfavorable characteristics of patients selected for this regimen), so evaluation of this regimen in comparison with high-intensity approaches is warranted in a larger sample. However, our data suggest that in BL/HGBCL systemic HDMTX may be essential for effective CNS prophylaxis.

View largeDownload slide

View largeDownload slide

Close modal