Outcome and pathologic classification of children and adolescents with mediastinal large B-cell lymphoma treated with FAB/LMB96 mature B-NHL therapy

Abstract

Mediastinal large B-cell lymphoma (MLBL) represents 2% of mature B-cell non-Hodgkin lymphoma in patients ≤ 18 years of age. We analyzed data from childhood and adolescent patients with stage III MLBL (n = 42) and non-MLBL DLBCL (n = 69) treated with Group B therapy in the French-American-British/Lymphome Malins de Burkitt (FAB/LMB) 96 study. MLBL patients had a male/female 26/16; median age, 15.7 years (range, 12.5-19.7); and LDH < 2 versus ≥ 2 × the upper limit of normal, 23:19. Six MLBL patients (14%) had < a 20% response to initial COP (cyclophosphamide, vincristine, and prednisone) therapy. Central pathology revealed approximately 50% with classical features of primary MLBL. Five-year event-free survival for the stage III MLBL and non-MLBL DLBCL groups was 66% (95% confidence interval [CI], 49%-78%) and 85% (95% CI, 71%-92%), respectively (P < .001; 14%). The 5-year overall survival in the 42 MLBL patients was 73% (95% CI, 56%-84%). We conclude that MLBL in adolescent patients is associated with significantly inferior event-free survival compared with stage III non-MLBL DLBCL and can be of multiple histologies. Alternate treatment strategies should be investigated in the future taking into account both adult MLBL approaches and more recent biologic findings in adult MLBL.

Figure 1

Main deregulated signaling pathways in PMBCL. Shown are the main activation cascades of JAK-STAT and NF-κB signaling. Alternative pathway activation exists. Known gene alterations leading to constitutive pathway activity in PMBCL are shown in color. Reprinted from Steidl and Gascoyne by permission of the American Society of Hematology.

Figure 2

Pathologic features of mediastinal large-cell lymphoma. (A) H&E histologic section of a MLBL tumor showing the characteristic large-cell infiltrate with intervening bands of compartmentalizing fibrosis. The tumor cells form individual cells and small groups (400× magnification). (B) CD20 immunoperoxidase staining demonstrating the mature MLBL infiltrate (1000× magnification). (C) IRF4/MUM 1 staining of a MLBL tumor showing characteristic nuclear staining (100× magnification; Olympus BX40 microscope).

Figure 3

Treatment course for each MLBL patient treated in the FAB/LMB 96 study. All MLBL patients received Group B therapy initially. After the COP reduction phase, patients with a CR or intermediate response (defined as a 20%-99% reduction in the product of the 2 largest diameters of evaluable lesions) continued on Group B therapy. At the third evaluation, a CR included residual masses as long as adequate biopsy demonstrated no viable tumor. Patients with residual viable tumors (persistent disease) were switched to Group C therapy beginning with CYVE1 consolidation. Nine patients in CR continued onto Group C M1-M4 therapy, 1 patient from Group B and 8 patients from Group C. Patients with disease progression (> 25% increase in the product of 2 largest diameters) were taken off study. Disease status at last follow-up for all patients completing therapy is included.

Figure 4

Probability of 5-year EFS and OS calculated using the Kaplan-Meier method for MLBL patients treated on Group B therapy in the FAB/LMB 96 study.

Figure 5

Probability of 5-year EFS calculated using the Kaplan-Meier method for patients with MLBL compared with stage III non-MLBL DLBCL patients treated on Group B therapy in the FAB/LMB 96 study. The 5-year EFS for MLBL patients was 66% (95% CI, 49%-78%) and for stage III non-MLBL DLBCL patients, it was 85% (95% CI, 71%-92%; P < .001).