Treatment Outcome in T-Cell Lymphoblastic Lymphoma in Adults - a Population-Based Study From the Swedish Lymphoma Registry.

Abstract 2727

T-cell lymphoblastic lymphoma (T-LBL) is a rare disease mainly affecting children and young adults. Commonly presenting with a large mass in the anterior mediastinum it represents the lymphoma end of the T-cell acute lymphoblastic leukemia (T-ALL) spectrum and as such has a high risk of CNS involvement. If present, cases with bone marrow infiltration < 25% are classified as T-LBL.

The optimal standard treatment for adult T-LBL is not yet defined due to the rarity of the disorder and the consequent lack of randomized prospective trials. Hence, adults that are diagnosed with T-LBL are exposed to a range of chemotherapy regimens, according to a variety of local therapeutic guidelines. Most published data originates from clinical trials in the pediatric population. Here we present the treatment outcome and clinical prognostic factors from a population-based cohort of adult T-LBL patients.

Using the Swedish Lymphoma Registry (SLR) all adult T-LBL patients diagnosed between January 2000 and December 2009 were identified in this study. Basic data was collected from the registry and further data was extracted from patient records after obtaining informed consent.

Through the SLR 46 cases were identified. Seven cases were re-classified as T-ALL. Median age at diagnosis was 40 years (range 18–77 yrs) with a male:female ratio of 5:3. Bulky disease (>10cm) was present in 67%. Only 2 patients had CNS involvement at diagnosis while 18 (46%) presented with stage IV disease, 10 because of bone marrow involvement and 8 due to extensive extranodal disease. Median follow up for surviving patients was 6.5 years (range 2.0 – 11 yrs). The estimated 5-year OS was 42% for the entire cohort.

Of 39 patients 1 declined participation, 1 died after the first cycle and 4 records could not be accessed, leaving 33 patients evaluable for treatment response. FDG-PET was performed in 13 patients after induction treatment. The ORR was 91% (33% CR, 58% PR, 9% SD). In many cases PR was due to PET negative residual masses. Thirty patients (age 18–66 yrs) received intensive ALL-like treatment. For this group treatments used were hyper-CVAD (n=19), LSA2L2 (n=4), VSTB -95 (n=2), NOPHO-ALL-HR (n=2), Euro LB -02 (n=1), GMALL 99/06 (n=1), ABCDV (n=1) resulting in an ORR of 97% (40% CR, 57% PR). No toxic death occurred. Consolidation consisted of radiation therapy against mediastinum followed by maintenance treatment in 3 patients, maintenance treatment in 19 patients, 2 underwent an autologous transplantation and 2 an allogeneic transplantation. Five-year PFS was estimated to 49% with 12 patients experiencing a relapse or progression including 7 patients with normal FDG-PET. Three patients with relapsed/progressive disease had an allogeneic transplant in second remission but subsequently died. Twelve patients died from disease-related causes and 3 from therapy-related causes, including 2 patients that developed secondary hematologic malignancies. Most relapses/progressions occurred in bone marrow, mediastinum or CNS with 3 patients with an isolated CNS relapse/progression. For the 7 patients receiving low intensity treatment (age 69–77 yrs) no patient survived beyond 30 months (median survival was 6.2 months), with 4 available for response evaluation, 2 achieving PR and 2 SD.

For the whole cohort, age and female gender were identified as risk factors for shorter OS (HR 1.04, p=0.007 respectively HR 4.67, 0.001). Females were significantly older than males and for patients with intensive treatment, neither female gender nor age were any longer significant risk factors.

This study is one of few to provide population-based data on treatment outcome of adult T-LBL. Age as a risk factor is likely to reflect intolerance to intensive treatment since age was not a risk factor for adverse OS or PFS for intensively treated patients. PET scan at time of response evaluation seems to be unreliable as predictor for the risk of relapse. Response rates in this material are comparable with published data while long term OS was slightly inferior, reflecting the population based nature of this data set.