Introduction: The majority of PTCL cases are pathologically heterogeneous, aggressive and chemorefractory; nearly 70% of patients who undergo induction chemotherapy develop RR disease. Data on how RR-PTCL patients are managed outside of clinical trials and academic settings have not been adequately studied. Understanding how PTCL patients are treated in the real world allows better design of future clinical trials and refined development of novel therapies. This study aimed to describe contemporary treatment patterns, clinical characteristics, and overall survival (OS) among RR-PTCL patients treated in the US community oncology setting.Methods: We conducted a retrospective, observational cohort study of RR-PTCL patients treated between 1/1/2010-12/31/2016. Data were collected from patient medical records abstracted from 30 community oncologists across all geographic areas in the US (38% South, 22% West, 7% Midwest, 26% Northeast, and 7% unknown). Patient characteristics, treatment patterns, adverse events (AEs), physician-reported response to first- and second-line (1L, 2L) treatments, and date of death were collected. Descriptive statistics were calculated to summarize patient characteristics and study outcomes. Median (95% confidence interval [CI]) OS from RR disease was calculated by the Kaplan-Meier method.Results: We analyzed 200 RR PTCL patients (71% received CHOP, 20% received CHOEP, and 9% received another regimen as 1L systemic therapy). The majority (58%) were PTCL not otherwise specified (NOS), 61% were males, and median age was 62 years (range: 20-90) (Table). Overall response rate (ORR) to 1L treatment was 80% (CR, 54%; PR, 26%), and ORR to 2L treatment was 55% (CR, 31%, PR, 24%). Among all PTCL patients and among those with PTCL NOS, the OS (95% CI) from RR disease was 13.0 months (95% 7.0-not estimable) and 9.0 months (95% 6.0-not estimable), respectively (Figure). Median time to RR disease from 1L initiation was 12.9 months. Median durations of 1L and 2L treatments were 4.4 months and 3.7 months, respectively. Among those who continued to 2L therapy, 27% received brentuximab vedotin (2% in combination), 19% received ICE (carboplatin/cisplatin, ifosfamide, etoposide), 13% received romidepsin (2% in combination with pralatrexate), 10% received pralatrexate monotherapy, 7% received GEMOX (gemcitabine, oxaliplatin), and the rest (22%) received other 2L regimens or an unknown regimen (2%). After 1L initiation, only 14/200 (7%) received hematopoietic stem cell transplant (HSCT) and 28/151 (19%) after 2L. AEs resulting in a dose reduction, discontinuation, or hospitalization during 1L treatment and occurring among ≥10% included anemia (39%), thrombocytopenia (27%), neutropenia (23%), mouth sores (16%), diarrhea (15%), neuropathy (14%), febrile neutropenia (13%), and vomiting (10%).Conclusions: While response to 1L treatment in this cohort was high (80%), all developed RR disease (or died), for whom the prognosis remains poor. Very few patients underwent HSCT in any line of therapy. Our data demonstrate the unmet medical need for RR-PTCL patients and the need for novel therapies in the 1L setting. It also underscores the need to better understand factors leading to treatment selection and optimal treatment sequencing in the real world. [Display omitted] DisclosuresNabhan:Cardinal Health: Employment, Equity Ownership. Laney:Cardinal Health: Employment, Equity Ownership. Feliciano:Seattle Genetics: Employment, Equity Ownership. Lee:Cardinal Health: Employment. Klink:Cardinal Health: Employment, Equity Ownership.