Introduction: Patients with aggressive peripheral T-cell lymphoma (PTCL) often undergo induction with anthracycline-based chemotherapy followed by autologous stem cell transplantation in fit patients. Up to 40% of patients have primary refractory disease, and even for those who respond to induction, 5-year overall survival (OS) is 32% with the exception of ALCL. This results in a significant number of patients who need treatment in the relapsed and refractory (R/R) setting. While there is no standard approach for this group, multiagent chemotherapy regimens are often used, particularly in those who may be transplant candidates. Several novel single agents have shown activity in R/R PTCL and can often be given in a more continuous fashion than combination chemotherapy. In order to explore the potential benefits of single agents in comparison to combination chemotherapy in R/R PTCL, we examined treatment outcomes in a prospectively enrolled cohort of PTCL patients. Methods: Patients were enrolled in the Comprehensive Oncology Measures for Peripheral T-cell Lymphoma Treatment (COMPLETE), a multinational prospective cohort of 500 newly diagnosed PTCL patients. Demographics, treatments, and outcomes were recorded until death or for at least 5 years. Analysis was restricted to 6 subtypes of PTCL: PTCL-NOS, AITL, ALCL, EATL, NKT, HSTCL. Primary refractory was defined as lack of a complete response (CR) to initial treatment. Relapsed was defined as a CR to initial treatment with progression at a later date. Patients in whom initial treatment was ≤4 days, unreported, or began >30 days before or after informed consent were excluded. Single agents were pralatrexate, romidepsin, brentuximab vedotin (BV), bendamustine, alisertib, denileukin diftitox, and lenalidomide. Combinations were any multiagent chemotherapy regimen excluding the above agents. Results: Of 462 patients with locked records, 57 met the above eligibility criteria and had treatment and followup data available for analysis. As first treatment in R/R disease, 26 patients received combination therapy (by subtype: PTCL-NOS [10], AITL [6], ALCL [3], EATL [1], NKT [5], HSTCL [1]) and 31 received single agents (by subtype: PTCL-NOS [13], AITL [5], ALCL [7], EATL [2], NKT [2], HSTCL [2]). The most common combination regimens had an ifosfamide-, gemcitabine-, or platinum-based backbone. Single agents used were: BV (12), romidepsin (8), pralatrexate (5), alisertib (3), bendamustine (1), denileukin diftitox (1), and lenalidomide (1). The objective response rate (ORR) for single agents was higher than for combination therapy (59% [17/29] vs. 46% [12/26], P=0.36), as was the CR rate (41% [12/29] vs. 19% [5/26], P=0.02). In those with R/R ALCL who received BV, ORR was 83% (5/6) and CR was 67% (4/6). Given these expected high response rates of BV in R/R ALCL, a subtype analysis excluding ALCL patients who received BV (ALCL-BV) was conducted and showed similar ORR between single and combination therapy (52% [12/23] vs. 46% [12/26], P=0.26), though CR still favored single agents (35% [8/23] vs. 19% [5/26], P=0.07). Excluding the ALCL-BV group, CRs were observed with BV (3/6), romidepsin (2/8), pralatrexate (1/5), bendamustine (1/1), and alisertib (1/3). Median progression-free survival (PFS) favored single over combination therapy (11.2 vs. 6.7 months, P=0.02). PFS in the ALCL-BV group was comparable at 11 months. Stem-cell transplantation occurred more frequently after single agent use than combination therapy (25% [8/31] vs. 7% [2/26]). Conclusion: Our analysis confirms the unmet need for better therapies in R/R PTCL. This data shows a trend towards increased CR and PFS with single agents in comparison to combination therapy, while maintaining the potential to bridge to transplantation. It is unclear whether the differences were driven by reduced toxicity and longer treatment duration with single agents, increased efficacy from BV in CD30-expressing cancers, or patient factors that led to therapy selection. However, despite the small size and exploratory nature of this analysis, single agents do not appear inferior to combination therapy with respect to disease control or ability to bridge to transplant with intent to cure. Our data can serve as a foundation for larger datasets or randomized trials which are needed to identify the truly superior strategy. In the meantime, the optimal approach for R/R PTCL remains unclear. Disclosures Schwartz: MS Biostatistics, LLC: Employment. Acosta:Spectrum: Employment. Pro:Seattle Genetics: Consultancy, Other: Travel expenses, Research Funding; kiowa: Honoraria; portola: Honoraria; Takeda Pharmaceuticals: Honoraria, Other: Travel expenses. Shustov:Seattle Genetics: Research Funding. Horwitz:Aileron Therapeutics: Consultancy, Research Funding; Millennium/Takeda: Consultancy, Research Funding; Mundipharma: Consultancy; ADC Therapeutics: Consultancy, Research Funding; Spectrum: Research Funding; Trillium: Consultancy; Corvus: Consultancy; Forty Seven: Consultancy, Research Funding; Seattle Genetics: Consultancy, Research Funding; Celgene: Consultancy, Research Funding; Innate Pharma: Consultancy; Infinity/Verastem: Consultancy, Research Funding; Portola: Consultancy; Kyowa-Hakka-Kirin: Consultancy, Research Funding. Foss:Seattle genetics: Consultancy; Miragen: Consultancy, Speakers Bureau; Mallinkrodt: Consultancy; Spectrum: Consultancy.