ALK-negative anaplastic large cell lymphoma is a genetically heterogeneous disease with widely disparate clinical outcomes

Abstract

Anaplastic lymphoma kinase (ALK)-negative anaplastic large cell lymphoma (ALCL) is a CD30-positive T-cell non-Hodgkin lymphoma that morphologically resembles ALK-positive ALCL but lacks chromosomal rearrangements of the ALK gene. The genetic and clinical heterogeneity of ALK-negative ALCL has not been delineated. We performed immunohistochemistry and fluorescence in situ hybridization on 73 ALK-negative ALCLs and 32 ALK-positive ALCLs and evaluated the associations among pathology, genetics, and clinical outcome. Chromosomal rearrangements of DUSP22 and TP63 were identified in 30% and 8% of ALK-negative ALCLs, respectively. These rearrangements were mutually exclusive and were absent in ALK-positive ALCLs. Five-year overall survival rates were 85% for ALK-positive ALCLs, 90% for DUSP22-rearranged ALCLs, 17% for TP63-rearranged ALCLs, and 42% for cases lacking all 3 genetic markers (P < .0001). Hazard ratios for death in these 4 groups after adjusting for International Prognostic Index and age were 1.0 (reference group), 0.58, 8.63, and 4.16, respectively (P = 7.10 × 10(-5)). These results were similar when restricted to patients receiving anthracycline-based chemotherapy, as well as to patients not receiving stem cell transplantation. Thus, ALK-negative ALCL is a genetically heterogeneous disease with widely disparate outcomes following standard therapy. DUSP22 and TP63 rearrangements may serve as predictive biomarkers to help guide patient management.

Figure 1

Outcomes in patients with ALCL based on genetic subtype. (A) OS rates in patients with ALCL, stratified by ALK status only (ALK positive, N = 29; ALK negative, N = 67). (B) OS rates in patients with ALCL, stratified by rearrangements of ALK (N = 29), DUSP22 (N = 21), and TP63 (N = 6). ^−/−/−, triple-negative cases lacking all 3 rearrangements (N = 40). (C) OS rates in patients with ALCL who did not undergo transplantation, stratified by rearrangements of ALK (N = 21), DUSP22 (N = 15), and TP63 (N = 5). ^−/−/−, N = 34.

Figure 2

Representative cases of genetic subtypes of ALCL. (A) ALK-negative ALCL with DUSP22 rearrangement. The tumor cells are positive for CD30 and are negative for ALK, TIA-1, and p63. FISH using a breakapart probe to the DUSP22-IRF4 locus on 6p25.3 shows abnormal separation of the red and green signals on 1 allele (pair of arrows). (B) ALK-negative ALCL with TP63 rearrangement. The tumor cells are positive for CD30, TIA-1, and p63 and are negative for ALK. FISH using a dual-fusion probe to TBL1XR1 on 3q26 and TP63 on 3q28 demonstrates 2 pairs of abnormal fusion signals, indicating 2 copies of inv(3)(q26q28) (TBL1XR1/TP63 fusion; 2 pairs of arrows). (C) ALK-negative ALCL lacking DUSP22 and TP63 rearrangements (triple-negative ALCL). The tumor cells are positive for CD30 and TIA-1 and are negative for ALK and p63. (D) ALK-positive ALCL. The tumor cells are positive for CD30, ALK, and TIA-1 and are negative for p63. Photomicrographs were taken using an Olympus DP71 camera, Olympus BX51 microscope, and Olympus DP Manager image acquisition software at an original magnification of ×400 (insets, ×1000). Original magnification of FISH images, ×600.

Figure 3

Comparative pathologic features of genetic subtypes of ALCL. (A) Percentage of cases with unanimous consensus on a diagnosis of ALCL, based on morphology, phenotype, and clinical data. (B) Percentage of cases with unanimous consensus on the presence of classic histologic features of ALCL, as assessed without knowledge of phenotype (except CD30 expression) or clinical data. (C) Immunophenotypic markers that differed significantly among genetic subtypes. As a defining feature of ALK-positive ALCL, immunohistochemistry for ALK is not included here.