TaLLy-ing Up Prognostic Features for T-Lymphoblastic Lymphoma

T-lymphoblastic lymphoma (T-LLy) is a neoplasm of immature T-cell precursors that represents approximately 30 percent of childhood non-Hodgkin lymphoma (NHL) and less than 2 percent of adult NHL.¹  It presents typically in a young male patient as a mediastinal mass or bulky lymphadenopathy with minimal or no marrow involvement. The morphology and immunophenotype of T-LLy are nearly identical to T-cell acute lymphoblastic leukemia (T-ALL), and thus the World Health Organization Classification of Lymphoid Neoplasms has combined these two cancers into one category: precursor T-cell lymphoblastic leukemia/lymphoma.²  The only distinction noted is the extent of bone marrow infiltration by the lymphoblasts, namely bone marrow involvement of 25 percent or more in T-ALL, and bone marrow involvement of less than 25 percent in T-LLy.

The prognosis for T-LLy in children and young adult patients has improved significantly with the use of high-dose multiagent ALL chemotherapy regimens. Currently, T-LLy is curable for approximately 85 percent of children and nearly 70 percent of young adults.^3,4  Nevertheless, relapsed T-LLy is rarely cured in children or adults, with the overall survival (OS) rate for children being less than 30 percent and less than 10 percent for adults.^5-7  Despite the progress of genomic analysis, molecular classification, and minimal residual disease analysis for T-ALL, there is a paucity of genomic data and independent prognostic indicators for T-LLy to identify patients at high risk of relapse and guide alternative therapies for cure. Thus, the discovery of prognostic factors for early identification of poor-risk T-LLy would be highly beneficial to the clinical care of this patient population.

In pediatric T-LLy, some genomic data are emerging that may help define high-risk patients. Retrospective molecular studies have shown association with outcome with NOTCH1 and/or FBXW7 mutations,^8,9  biallelic T-cell receptor γ deletions,⁹ and loss of heterozygosity at chromosome 6q.¹⁰  Recently, Dr. Tasneem Khanam and colleagues reported that for patients without NOTCH1/FBXW7 mutations, KMT2D and PTEN mutations define a group with high risk of relapse.¹¹  Also, other investigators are correlating high-level minimal disseminated disease (MDD) assessments with NOTCH1/FBXW7 status to identify high-risk patients.^12,13 

In an alternative approach, investigators from the Children's Oncology Group (COG) analyzed the immunophenotype of 180 children and adolescents enrolled on the COG T-LLy study (A5971) to understand whether immunophenotypic subgroups provide a useful basis for classification and prediction of clinical outcome.¹⁴  The authors reported that the immunoprofile of T-LLy is generally similar to that of T-ALL, but interestingly identified a subset of pediatric patients (~14%) suspected to have the early T-cell precursor (ETP) phenotype, a unique subtype known for T-ALL with high-risk features¹⁵  but not described in T-LLy at the time.

The ETP subtype of T-ALL has a specific immunophenotype (lack of expression of CD1a and CD8, CD5 expression 1 log lower than the mature T cells or < 75% positive, and expression of one or more of the following myeloid antigens: CD13, CD33, CD34, CD117, or HLA-DR), distinctive genetic abnormalities as compared to non-ETP ALL, higher risk of chemotherapy treatment failure (induction failure and higher rates of minimal residual disease detection) and, thus, inferior outcomes as compared to non-ETP ALL.^15-18  As a result, in the clinical setting, it is important to identify patients with an ETP phenotype as they may need intensification of chemotherapy or allogeneic hematopoietic transplantation (allo HSCT) for cure. Whether there is an equivalent of ETP ALL in T-LLy remains uncertain because conclusive data of incidence, clinical, and genomic characteristics of the ETP phenotype in T-LLy are lacking for pediatric and adult patients.

In the current study, Dr. Xinjie Xu and colleagues expanded on the COG's earlier report of ETP-LLy in a larger cohort of pediatric patients with T-LLy. They analyzed the incidence and clinical characteristics of ETP subtype, explored genome copy number profiles, and compared patient outcomes with non-ETP-LLy. The authors reviewed the immunophenotyping data on 218 patients with T-LLy enrolled on the COG AALL0434 phase III clinical trial⁴  and identified nine cases (4%) with the definitive ETP phenotype. Fifteen non-ETP cases were randomly selected for comparison, and these tumor specimens were required to express CD1a, CD5, and CD8. Of note, all patients enrolled on AALL0434 had the diagnosis confirmed by central review at the University of Utah, and the MDD status was performed using multiparameter flow cytometry of bone marrow specimens and analyzed at the University of Washington to ensure reliability. In the prior COG study (A5971), the flow cytometry of patient specimens was performed at various reference laboratories rather than in a central laboratory. The mean age of the T-LLy population was 10.7 years (range, 1-23), and there was no difference in the age of patients with ETP phenotype versus non-ETP phenotype (10.3 vs. 10.9 years). The male-to-female ratio was 2.0 for the 24 patients in the study, and there was no difference between ETP and non-ETP in male-to-female ratio (1.25 vs. 2.75, p=4.0).

Genomic microarray hybridization was performed on all 24 specimens using a molecular inversion single nucleotide polymorphism assay and identified copy number alterations (CNAs) and copy-neutral loss-of-heterozygosity (CN-LOH). CNAs and CN-LOH were identified in eight of nine EPT LLy cases and in 15 of 15 non-ETP LLy cases. The CNAs that were observed only in ETP-LLy were deletions of chromosome 12p including the ETV6 gene (3/9 patients; 33% in ETP-LLy vs. 0/15, 0% in non-ETP-LLy; p=.042), and deletions of chromosome 1p including the RPL22 gene (4/9 patients; 44% in ETP-LLy vs. 0/15, 0% in non-ETP-LLy; p=.012). The RPL22 gene is part of the T-cell receptor signaling pathway and is essential during T-cell development, and the deletion of this gene has not previously been associated with T-LLy. ETP-LLy showed significantly less frequent 9p deletion/CN-LOH as compared to non-ETP-LLy (4/9 patients; 44% vs 14/15, 93%; p=.015. All the 9p deletion/CN-LOH cases except for one ETP-LLy specimen included both the CDKN2A and CDKN2B genes.

The clinical outcomes of the 24 patients whose tumor specimens were included in this study were analyzed after a median follow up of 6.3 years. The survival rate for the nine patients with ETP-LLy was excellent with no chemotherapy failures or treatment-related deaths (4-year event-free survival rate [EFS] of 100%). Patients with non-ETP-LLy had an EFS of 79 percent. Of note, the four-year EFS and OS for the entire T-LLy cohort enrolled in AALL0434 (n=288) was 84.7 percent and 89.0 percent, respectively.