Multi-Institution Review of Adult Early T-Cell Precursor Acute Lymphoblastic Leukemia/Lymphoma (ETP-ALL)
Abstract Background: The Early T-cell precursor (ETP) variant of acute lymphoblastic lymphoma/leukemia (ALL) is a recognized high risk variant, recognized by the absence of CD1a, with aberrant myeloid antigen expression (CD13, CD33, CD117, and CD34), and frequent absence of CD4 or CD8. Treatment intensification may improve outcome in this subset. We undertook a multi-center retrospective analysis to explore clinical features, treatment exposure, and outcomes in ETP ALL as compared to non-ETP variants. Methods Adult T-ALL/T-LBL cases were compiled from 3 high volume cancer centers between the years 2003-2015. Data collected included patient demographics, tumor characteristics (white count at diagnosis, flow cytometry, FISH, cytogenetics, bone marrow involvement), treatment regimens and patient outcomes. ETP cases were defined as definite (CD1a-/CD8-/myeloid+) or probable (CD1a unk/CD8-/myeloid+, or CD1a-/myeloid+ with CD4+ and/or CD8+). All other cases were defined as non-ETP. Demographic data were compared using independent t-test assuming non-equal variance. OS and PFS were calculated from diagnosis and compared by Kaplan Meier and log-rank testing. Results Among 95 cases, 33 met criteria for definite/probable ETP (35%). OS and PFS data were indistinguishable between these groups (p=0.24, p=0.34), and were subsequently analyzed as a single group. Within the ETP group, no factors were associated with OS, including histology (CD1a+ vs unk, CD3cyt vs CD3sur, CD5dim vs CD5+, CD1a+/13+ vs CD1a+/13-, or CD13, CD33, CD117, CD34, & TdT status), marrow blast burden, peripheral blast burden, white blood cell count (wbc), hemoglobin (hgb), platelet count (plt), cytogenetics/FISH status, chemotherapy choice, or allogeneic transplant (in CR1 or at any time). With regards to PFS, only the inclusion of asparaginase with induction was associated with outcome (p=0.009), while all other covariates failed to show any significance. The ETP group was compared with the non-ETP subset (table 1). ETP were more likely to abuse marijuana, possibly reflecting unrecognized pesticide exposure, and were more likely to abnormalities of chrom 5 & 7. ETP trended towards lower response and higher rate of relapse, with lower PFS. Comparison of OS was not significant, likely related to small numbers (5y OS 37% vs 22%, figure 1). Non-ETP failed to show PFS benefit with frontline asparaginase, otherwise no treatment differences were apparent. Conclusions In this muti-center cohort we were able to identify and characterize ETP cases, confirming poor outcomes. Improvement in PFS among ETP patients treated with frontline asparaginase warrants attention and prospective confirmation. Unfortunately, OS remains poor independent of treatment or receipt of allogeneic transplant, suggesting a critical need remains for development and study novel therapies. Table 1. ETP Non-ETP p-value Median Age 37.45 34.74 0.42 Male 82% 66% 0.89 FamilyHx of Lymph/Leuk 21% 8% 0.112 FamilyHx of Ca 42% 25% 0.09 THC 24% 5% 0.021 P blasts 40% 28% 0.158 >25% M blasts 30% 55% 0.0571 WBC 78.45 76.55 0.948 wbc>100 24% 24% 0.995 Hgb 10.72 11.78 0.148 hgb<12 67% 47% 0.097 plt 151.59 138.66 0.644 Chrom 5/7 40% 7% 0.005 Remission 61% 79% 0.096 Relapse 76% 58% 0.073 OS 27.00 22.00 0.595 PFS 13.00 17.00 0.048 PFS Asp ETP (asp no vs yes) 12 59 0.009 non-ETP (asp no vs yes) 17 15 0.777 Figure 1. Figure 1. Disclosures Shah: Acetylon: Membership on an entity's Board of Directors or advisory committees; Spectrum: Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Bayer: Honoraria; Celgene: Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; PLexus Communications: Honoraria; Pharmacyclics: Speakers Bureau; Rosetta Genomics: Research Funding; Seattle Genetics: Research Funding. Kota:Pfizer: Membership on an entity's Board of Directors or advisory committees; Leukemia Lymphoma Society: Research Funding. Hathaway:OnQ Health: Research Funding.