High Dose Chemotherapy with Autograft in First Relapse May Overcome the Poor Prognosis of Diffuse Large B-Cell Lymphoma Patients with MYC/BCL2 Co-Expression

The immunohistochemical (ICH) co-expression of MYC and BCL2 (double expressors, DE) has been emerging as a strong and reproducible risk factor in diffuse large B cell lymphoma (DLBCL) patients (pts) treated with R-CHOP. Our aim was to analyze the prognostic value of the co-expression in DLBCL pts treated with high-dose chemotherapy followed by autologous stem cell transplantation at first relapse.

In this retrospective study, we analyzed a cohort of 64 young [median age 53, (range18-70)] and/or fit DLBCL pts treated between 2003 and 2015 at the Istituto Nazionale dei Tumori Milano. All pts relapsed at median time of 4.9 months after a Rituximab-based chemotherapy (48 R-CHOP-like and 16 other intensive regimens). All pts were considered fit and eligible for high-dose chemotherapy salvage treatment. Tumor samples were analyzed by ICH for MYC and BCL2 (40%/70% threshold). Germinal B-Cell (GBC) and Activated B-Cell (ABC) DLBCL classification was done with ICH as established by the Hans algorithm. Pts characteristics at relapse were: refractory: 37/64 (58%), Ann Arbor stage≥3: 42/64 (65%), IPI score≥2: 26/60 (43%); Extranodal disease: 37/64 (58%); Bulky (defined as max diameter ≥5 cm): 29/64 (45%); CNS involvement: 8/64 (12.5%).

Twenty-nine of 64 (45%) pts responded to salvage high-dose therapy (CR,PR) and 25 of them underwent autologous stem cell transplantation (ASCT). Among 39 (61%) refractory pts to first salvage intensive chemotherapy approaches, 11 (28%) responded after third or fourth chemo-immunotherapy salvage lines and underwent ASCT as well. Overall the reasons for not receiving ASCT were: refractory disease (n=20), poor mobilization (n=3), the occurrence of significant infection or toxic complications during the salvage therapy (n=4) or patient decision (n=1). ICH MYC and BCL2 positive expression was observed in 31 (48%) and 39 (61%) pts respectively. Among them, 17 (26%) were characterized by BCL2 and MYC co-expression. According with Hans algorithm, 23 (56%) and 18 (44%) DLBCL pts were classified as GBC and ABC-DLBCL. DE pts did not show any significant differences compared to others in terms of age, IPI score, presence of extranodal disease, BM infiltration, CNS involvement at relapse, refractory disease or relapse occurrence less than 1 year after first line therapy, and proportion of ABC-DLBCL subtype. Considering the whole pts cohort 5-years progression free survival (PFS) and overall survival (OS) were 27.3% (95% CI, 21.1%-33,4%) and 40.6% (95% CI, 33.2%-48%) respectively. In univariate analysis, pts with disease relapsed less than 1 year after first line therapy were characterized by a significant worse outcome in terms of 5-years PFS [11.5% (95% IC 6.1%-16.9%) vs 72.2% (95% IC 60%-84.4%] and OS [24.5% (95% IC 16.9%-32.1%) vs 84.4% (95% IC 74%-94.8%)] (p=0.0001 and p=0.0009 respectively). Similar results were observed comparing pts with refractory disease to first line therapy with others [5-years PFS 11.7% (95% IC, 5.9%-17.5%) vs 48.2% (95% IC, 37.1%-59.3%)] and 5-years OS [19.9% (95% IC, 11.8%-28%) vs 67.9% (95% IC, 57.6%-78.2%)] (p=0.0003 and p=0.001 respectively). Other variables at first relapse associated with worse outcome in terms of PFS and OS were IPI score ≥2 and presence of bulky disease. In contrast, DE patients did not shown any significant differences compared to other patents in term of 5-years PFS [39.7 (95% IC, 27.5-51.9) vs 25.1 (95% IC, 18.4-31.8), p=0.5] and 5-years OS [40.8 (95% IC, 32.3-49.3) vs 22.2 (95% IC, 5-39.4), p=0.8]. Furthermore singularly ICH MYC or BCL2 expression did not influence neither PFS nor OS.

These data suggest that salvage programs with high dose chemotherapy with autograft in first relapse may overcome the previously reported poor prognosis associated with ICH MYC/BCL2 co-expression. The most significant and robust prognostic factor among DLBCL pts in first relapse is still represented by refractoriness and the time of relapse.