Predictive Value of PET-CT in Patients with T-Cell Lymphoma Undergoing Autologous and Allogeneic Stem Cell Transplant

Abstract

Introduction T-NHL are associated with an aggressive clinical course and poor outcomes. The role of PET-CT at the time of SCT (pre-SCT), BM biopsy, and peripheral blood (PB) flow cytometry have not been well investigated. We retrospectively analyzed 83 pts with T-NHL undergoing auto or allo SCT to determine the predictive value of pre-SCT PET-CT scan, BM and PB flow for relapse-free survival (RFS) and OS. Methods PET CR was defined as the absence of any radiographic dz by PET-CT (5-pt score). Overall CR was defined as negative PET, BM and PB flow. Pre-SCT variables were assessed for their effects on RFS and OS: age, number of previous lines of therapy, CIBMTR dz risk, pre-SCT BM, pre-SCT PB flow, and pre-SCT PET-CT. Results Of 83 pts studied, 49 were autoSCT and 34 were alloSCT. Six pts who underwent alloSCT had received a prior autoSCT. Diagnosis was PTCL NOS 34, AITL 23, CTCL/Sezary syndrome 11, nasal NK-T 6, EATL 3, SPTCL 2, hepatosplenic 1, ALK-ALCL 1, HTLV-2 ATL 1, T-lymphoblastic 1. The majority of pts with AITL (16/23) underwent autoSCT upfront. Conditioning for autoSCT was BEAM. Conditioning for alloSCT was Pentostatin/TBI in all but 6 SCT. Median lines of therapy prior to SCT were 1 for auto and 4 (r 2-13) for alloSCT. Median age at SCT was 60 (r 20-75). A total of 6 out of 49 (12%) autoSCT and 8 out of 34 (24%) alloSCT had positive PET-CT (score 4-5) at the time of SCT. BM was positive in 8 out of 49 (16%) at the time of autoSCT and 7 out of 34 (21%) at the time of alloSCT. PB flow was positive in 9 out of 49 (18%) at the time of auto SCT and 6 out of 34 (12%) at the time of alloSCT. CIBMTR dz index was low in 57(69%), intermediate in 15(18%), and high in 11 (13%) of 83 SCT. The majority of allo-SCT were intermediate-high risk. The median follow-up 56 months (r 6 - 122). Twenty (24%) pts had expired. Univariate analysis of pre-autoSCT variables did not reveal statistically significant (SS) results for BM or PET positivity. Univariate analysis of pre-alloSCT variables identified positive BM and positive PET/CT scan as being SS. There was a SS lower incidence of relapse, longer RFS and OS in alloSCT pts with either negative pre-SCT PET-CT scan or negative pre-SCT BM. Other pre-SCT variables tested were not SS. In multivariate analysis, pre-alloSCT PET-CT alone retained SS. Pre-alloSCT PET-CT scan was a strong predictor of 2-yr RFS and 2-yr OS. Conclusion This data indicates that in heavily pretreated pts with T- NHL undergoing alloSCT, a negative pre-SCT PET-CT is a statistically significant predictor of long-term RFS and OS regardless of dz risk and number of lines of therapy. Due to the small number of autoSCT with positive PET-CT scans, BM biopsy, or PB flow at the time of SCT, no association between the above-mentioned variables and relapse was identified. Further studies with larger numbers of pts are warranted to determine the role of pre-autoSCT PET-CT, BM and PB flow as prognostic factors in T-NHL.