Survival trends in the era of novel therapies identify unmet needs in peripheral T-cell lymphoma: a national cancer database analysis Authors: Julia Wang, Yajuan Gao, N Nora Bennani, Yun Kyoung Tiger, Arushi Khurana, Urshila Durani, Jonas Paludo, Gita Thanarajasingam, Steven R Hwang, Yucai Wang, Grzegorz S Nowakowski, Thomas M Habermann, Thomas E Witzig, Stephen M Ansell, Jithma P Abeykoon Published: 2026 Introduction Peripheral T-cell lymphomas (PTCL) are a heterogeneous group of lymphoproliferative diseases comprising approximately 10% of all non-Hodgkin lymphomas (NHL) [1, 2]. PTCL-not otherwise specified (PTCL-NOS) represents the most common subtype, followed by angioimmunoblastic T-cell lymphoma (AITL) and anaplastic large cell lymphoma (ALCL) [3]. Except for anaplastic lymphoma kinase (ALK)-positive ALCL, these diseases have generally been associated with poor outcomes [4]. Their rarity and biological heterogeneity pose challenges for conducting prospective clinical trials and developing effective new therapies. The frontline treatment of PTCL has been based on CHOP or CHOP-like regimens historically. In the relapsed/refractory setting, several agents have been approved by the U.S. FDA, including pralatrexate (2009), the histone deacetylase (HDAC) inhibitor romidepsin (approved in 2011; withdrawn in August 2021), brentuximab vedotin (BV) for relapsed or refractory ALCL in August 2011, and the HDAC inhibitor belinostat (approved in 2014). Subsequently, BV combined with cyclophosphamide, doxorubicin, and prednisone (CHP) demonstrated significant improvements in survival outcomes compared with CHOP in the ECHELON-2 trial, leading to its approval by FDA in November 2018 for previously untreated ALCL or CD30-expressing PTCL [5]. Despite these therapeutic advances, it remains unclear how the introduction of these agents has influenced survival outcomes in real-world clinical practice. To address this gap of knowledge and to determine unmet needs, we conducted a retrospective analysis using de-identified data from the National Cancer Database (NCDB) to evaluate survival outcomes among patients diagnosed with PTCL before and after the introduction of novel agents. Patients diagnosed with PTCL-NOS, AITL, or ALCL from 2004 to 2020 were identified in the NCDB. Two eras were defined based on the timing of romidepsin and brentuximab vedotin’s FDA approval: Era 1 (2004–2011) and Era 2 (2012–2020). PTCL-NOS and AITL were analyzed as a combined group due to similar survival patterns observed in preliminary analyses. ALK-positive and ALK-negative ALCL were not analyzed separately because this distinction is not available in the NCDB. This study was exempt from institutional review board oversight and did not require ethics approval due to the use of publicly available, de-identified data. Chi-square test was used for categorical variables, and t-test and Mann-Whitney U test for continuous variables. Overall survival (OS) estimates were calculated using the Kaplan-Meier method and compared with log-rank tests. A multivariate Cox regression model identified factors associated with OS. Hazard ratio (HR) and 95% confidence interval (CI) were calculated. A P-value < 0.05 was considered statistically significant. All analyses were performed using SPSS 28.0. A total of 29,075 patients with PTCL were identified, including 21,079 with PTCL-NOS or AITL and 7996 with ALCL. In Era 1, 8,378 patients with PTCL-NOS or AITL and 3966 with ALCL were diagnosed, whereas Era 2 included 12,701 PTCL-NOS or AITL and 4030 ALCL cases. Baseline characteristics of patients with PTCL-NOS/AITL and ALCL across the two treatment eras are summarized in Supplementary Tables 1 and 2. Overall, patients diagnosed in Era 2 differed modestly from those in Era 1. Across subtypes, Era 2 patients were more likely to present with stage III–IV disease and to be treated at academic centers. Differences in age, race/ethnicity, and insurance distribution were also observed between eras, although patterns varied by subtype. PTCL patients diagnosed in Era 2 had better OS compared with Era 1, with a more significant improvement seen in ALCL. For PTCL-NOS or AITL patients, the median OS was 18.5 months (95% CI 17.4–19.6) in Era 2 compared with 15.5 months (95% CI 14.5–16.5) in Era 1 (P = 0.004), and the 3-year and 5-year OS rates were 38.3% and 31.8% in Era 1, compared with 40.0% and 33.5% in Era 2, respectively (Fig. 1A). The age- and sex-adjusted hazard ratio (HR) was 0.924 (95% CI, 0.893–0.956, P < 0.001). For ALCL patients, the median OS was 114.5 months in Era 2 compared with 65.5 months (95% CI 55.9–75.1) in Era 1 (P < 0.001). The 3-year OS rate was 55.1% in Era 1 vs. 65.6% in Era 2, while the 5-year OS rate was 50.9% in Era 1 vs 61.2% in Era 2 (Fig. 1B). The age- and sex-adjusted HR was 0.767 (95% CI, 0.716–0.821, P < 0.001). Survival trends in Era 1 (2004-2011) and Era 2 (2012-2020) among patients with peripheral T cell lymphoma. A Survival of patients with peripheral T cell lymphoma-not otherwise specified (PTCL-NOS) and angioimmunoblastic T-cell lymphoma (AITL). B Survival of patients with anaplastic large cell lymphoma (ALCL). In multivariate Cox regression analysis (Table 1) of the PTCL-NOS and AITL group, diagnosis in Era 2 was associated with a lower risk of death compared with Era 1 (HR = 0.930, P < 0.001). Increased age, male sex, Black race and a Charlson–Deyo comorbidity score ≥1 were associated with higher mortality (all P < 0.001). Insurance type and treatment facility were also associated with survival outcomes. Similar results were observed among patients with ALCL. Diagnosis in Era 2 was associated with a significantly lower risk of death (HR = 0.73, P < 0.001). Older age, male sex, Black race, comorbidity burden, and government or uninsured insurance status were associated with worse survival (all P < 0.001). In contrast, treatment at non-academic centers was associated with a lower risk of death compared with academic centers (HR = 0.887, P = 0.001). Table 1Multivariable Cox regression analysis for overall survival in patients with peripheral T-cell lymphoma-not otherwise specified (PTCL-NOS), angioimmunoblastic T-cell lymphoma (AITL) and anaplastic large cell lymphoma (ALCL).PTCL-NOS/AITLALCLHazard Ratio95% CIP-valueHazard Ratio95% CIP-valueEraEra 1 (2004–2011)ReferenceReferenceEra 2 (2012–2020)0.9300.897–0.965< 0.0010.7300.680–0.784< 0.001Age< 65ReferenceReference≥ 65 and < 791.4211.353–1.493< 0.0011.8831.713–2.069< 0.001≥ 802.3282.196–2.468< 0.0013.4853.113–3.902< 0.001SexMaleReferenceReferenceFemale0.8540.824–0.885< 0.0010.7920.738–0.849< 0.001RaceWhiteReferenceReferenceBlack1.3001.237–1.366< 0.0011.3351.216–1.466< 0.001Others1.0050.927–1.0890.9120.9810.809–1.1910.847EthnicityNon-HispanicReferenceReferenceHispanic0.9410.876–1.0110.0960.9970.871–1.1420.971Charlson-Deyo score0ReferenceReference≥ 11.4031.349–1.459< 0.0011.6371.519–1.764< 0.001Insurance typePrivateReferenceReferenceGovernment1.2721.211–1.366< 0.0011.5021.368–1.651< 0.001None1.3841.255–1.526< 0.0011.4111.192–1.669< 0.001Treatment facilityAcademicReferenceReferenceNon-Academic1.0871.048–1.127< 0.0010.8870.825–0.9540.001 Multivariable Cox regression analysis for overall survival in patients with peripheral T-cell lymphoma-not otherwise specified (PTCL-NOS), angioimmunoblastic T-cell lymphoma (AITL) and anaplastic large cell lymphoma (ALCL). To our knowledge, this is the first report on survival trends in PTCL across different treatment eras based on large-scale database analysis. Although OS improved over time, the gain was only pronounced in ALCL and remained limited in PTCL-NOS or AITL. BV (for both relapsed and frontline treatment) likely contributed to the improvement of OS in ALCL. The median OS increased from 65.5 months to 114.5 months and the 5-year survival rate rose from 50.9% to 61.2% in our study. However, the lack of broadly effective therapeutic advances for PTCL-NOS and AITL was the likely reason for limited survival improvement. The median OS only increased by approximately 3 months and the improvement in 5-year survival was only 1.7%. These results highlight the urgent need for breakthrough therapies. Consistent with our findings, the LEO-MER cohort shows persistent heterogeneity and little survival improvement in systemic PTCL over the past two decades [6]. Multiple agents have been approved in the relapsed/refractory setting, but the overall response rates (ORRs) to these agents remain modest at 25–30% with the exception of BV for patients with ALCL [7]. More effective therapies with higher ORRs and more durable responses, such as with new agents and combinations, are critically needed. Biology-driven and individualized targeted therapies are future directions for improving outcomes with further identification of molecular profiling and predictive biomarkers. The definition of nodal T-follicular helper cell lymphomas (nTFHLs) is newly introduced in both the WHO-HAEM5 [1] and the International Consensus Classification (ICC) [2] reflecting overlapping clinical and immunophenotypic features as well as mutations in epigenetic regulators. Epigenetic therapies including new HDAC inhibitors, oral 5-azacitidine, valemetostat and IDH2 inhibitors have shown promise and are under active investigation [7]. Multiple pathway-directed strategies are emerging in PTCL, including inhibition of PI3K pathway, JAK-STAT pathway, Notch pathway, TK::SYK fusion, ALK and PDGFR, supporting biomarker-guided treatment such as duvelisib in PRIMO trial [8]. The recent report of significant activity of the oral farnesyltransferase inhibitor tipifarnib in AITL patients whose tumors have responder mutations such as DNMT3A, RhoA, IDH2 or TET2 demonstrates the importance of using molecular profiling in these patients to select therapy [9]. Immunotherapy is slowly advancing, though challenges include risk of hyperprogression with checkpoint blockades, technical barriers in CAR-T-cell therapy such as T-cell aplasia, fratricide, contamination of CAR T-cell products with malignant cells [10]. New strategies such as novel targets on the cell surface or in the immune microenvironment have been employed to investigate the possibility of immunotherapies in PTCL [10]. Although the initial response rate of PTCL-NOS and AITL to CHOP-like front-line treatment is around 75% (with CR rate of 60%), patients experience high rates of recurrence and outcomes after progression or relapse are poor [11]. CHOP or CHOEP-based approaches have remained the backbone of therapy for decades [7], and better frontline treatments with more durable responses are needed. Multiple attempts to improve outcomes using “CHOP + X” strategies including alemtuzumab, denileukin diftitox, bortezomib and romidepsin have failed to demonstrate benefit in unselected PTCL patients [7] and only addition of etoposide to CHOP appeared to benefit young patients in the retrospective analysis from the German High-Grade Non-Hodgkin Lymphoma Study Group, while the benefit is less clear in older patients [12]. New drugs should continue to be explored in this space, including PI3K inhibitors (A051902; NCT04803201), epigenetic drugs and immunomodulatory drugs [13], given the success of adding BV to frontline chemotherapy in ALCL. Autologous stem cell transplant (ASCT) is offered in first remission for consolidation among fit patients with higher-risk ALK-positive ALCL or other common PTCL subtypes [7], though the benefit of ASCT needs to be confirmed (EA4232; NCT06724237). There are several limitations to this study. Firstly, although we defined the treatment era based on major therapeutic milestones, the NCDB does not provide regimen-level detail, subsequent lines of therapy, and cause-specific mortality, limiting mechanistic interpretation of OS differences. Secondly, other variables that influence prognosis such as performance status, lab tests, transplant eligibility and response depth are not captured and raise the possibility of confounding. Third, histologic classification, staging and coding of PTCL has evolved over time which may introduce misclassification and limit comparability across eras. Fourth, the NCDB underrepresents minority groups and individuals from nonmetropolitan areas, which may introduce selection bias and limit generalizability. Finally, the inability to distinguish ALK-positive from ALK-negative ALCL may obscure subtype-specific differences. In summary, our findings demonstrate that although survival outcomes have improved over time in PTCL overall, these gains remain modest in PTCL-NOS and AITL, in contrast to the more pronounced improvement in ALCL, highlighting unmet needs for more effective therapies, particularly for PTCL-NOS and AITL. Supplementary information Supplementary Tables 1 and 2 Author contributions JW contributed to conceptualization and statistical analysis. YG contributed to drafting the manuscript. JPA contributed to conceptualization and supervision, and revised the manuscript. NNB, YKT, AK, UD, JP, GT, SRH, YW, GSN, TMH, TEW and SMA contributed to supervision and revised the manuscript. All authors discussed the results, commented on the manuscript, and approved the final version before submission. Funding All authors declare no relevant funding for this work. Data availability The data that support the findings of this study are derived from the National Cancer Database (NCDB), a joint project of the American College of Surgeons’ Commission on Cancer and the American Cancer Society. These data are not publicly available but may be obtained from the American College of Surgeons upon reasonable request and with appropriate approvals. Competing interests NNB has received honoraria as a consultant and advisory board member for Acrotech Biopharma and for serving on the Scientific Review Committee for Pfizer. YKT has received honoraria as a consultant for Acrotech, Novartis, Kite, MJH Life Sciences, and Plexus Communications. JP has received research funding from Karyopharm, Biofourmis, and AstraZeneca, and honoraria from AbbVie, all paid to his institution. YW has received research funding to his institution from Incyte, InnoCare, LOXO Oncology, Eli Lilly, MorphoSys, Novartis, Genentech, Genmab, AbbVie, BeiGene, Merck, AstraZeneca, and Bristol Myers Squibb; advisory board compensation to his institution from Eli Lilly, LOXO Oncology, TG Therapeutics, Incyte, InnoCare, Kite, Jansen, BeiGene, AstraZeneca, Genmab, and AbbVie; consultancy compensation to his institution from InnoCare and AbbVie; and honoraria to his institution from Kite. TMH has received honoraria for serving on Data Monitoring Committees for Seagen and Eli Lilly & Co, and research support for the LEO grant from Genentech, Bristol Myers Squibb, and Genmab, as well as additional research support from Sorrento; all payments were made to his institution. SMA has received research funding for clinical trials from Bristol Myers Squibb, Takeda, Pfizer, Regeneron, AstraZeneca, Affimed, StepPharma, and ADC Therapeutics. JPA has received research funding from Qurion, CRISPR Therapeutics, Eli Lilly and BioMed Valley Discoveries for clinical trials, all paid to his institution. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Supplementary information The online version contains supplementary material available at 10.1038/s41408-026-01533-8. References Associated Data Supplementary Materials Supplementary Tables 1 and 2 Data Availability Statement The data that support the findings of this study are derived from the National Cancer Database (NCDB), a joint project of the American College of Surgeons’ Commission on Cancer and the American Cancer Society. These data are not publicly available but may be obtained from the American College of Surgeons upon reasonable request and with appropriate approvals.