Current and upcoming treatment approaches to common subtypes of PTCL (PTCL, NOS; ALCL; and TFHs) Free

The “common” nodal peripheral T-cell lymphomas (PTCLs), which include PTCL, not otherwise specified (PTCL, NOS), the T-follicular helper (TFH) lymphomas, and anaplastic large-cell lymphomas (ALCLs), make up only ∼10% of the non-Hodgkin lymphomas.^1,2 With the exception of anaplastic lymphoma kinase (ALK)-positive ALCL, these diseases have typically been associated with poor outcomes, and their rarity poses challenges for conducting clinical trials and understanding their biology.² Recent studies have highlighted important differences among these diseases with respect to the cell of origin or cell state, genetic profile, and drug sensitivity; however, until recently, these diseases had been treated similarly.³ With improved classification, development of new drugs, and identification of predictive biomarkers, treatment approaches are beginning to and will continue to diverge.

PTCL, NOS is the most common subtype of T-cell lymphoma; however, it is a diagnosis of exclusion and thus represents a vastly heterogeneous group of diseases. With improved biologic understanding and the increased sophistication of molecular tools, some entities previously lumped under this label have now been reclassified into more specific diagnoses. Beginning in 2016 with the fourth edition of the World Health Organization classification, entities originally classified as PTCL, NOS, but with features overlapping with angioimmunoblastic T-cell lymphoma (AITL), were reclassified under the umbrella of TFH lymphomas. This reclassification was in recognition of the biologic and clinical overlap between these diseases and AITL. The most recent World Health Organization classification (fifth edition) considers nodal TFH lymphomas to represent 3 entities: angioimmunoblastic-type, follicular-type, and NOS. Similarly, the new International Consensus Classification identifies TFH lymphoma as a single entity with 3 subtypes: angioimmunoblastic-type, follicular-type, and NOS.^1,2 

Among the rest of PTCL, NOS, 2 biologic subgroups have been identified, PTCL-GATA3 and PTCL-TBX21, which by gene expression profiling resemble T-helper 1 (Th1) and Th2 cells, respectively.⁴ PTCL-GATA3 was associated with poor prognosis, whereas PTCL-TBX21 was associated with favorable prognosis, except when characterized by a highly cytotoxic cell-like signature. These subgroups may predict drug sensitivity, as PTCL-GATA3 is associated with the upregulation of the phosphatidylinositol 3-kinase (PI3K) pathway, whereas PTCL-TBX21 is associated with the signal transducers and activators of transcription (STAT)3 pathway. However, whether the treatment should be altered based on the subtype is not yet known.⁴ 

It is likely that further molecular profiling will identify additional clinically meaningful subgroups within PTCL, NOS. For example, in a retrospective series of patients with PTCL treated with curative intent at Memorial Sloan Kettering Cancer Center, TP53 mutations were enriched among patients with PTCL, NOS (observed in 28%) and associated with inferior progression-free survival (PFS), suggesting that this subgroup is less sensitive to chemotherapy.⁵ Similarly, a retrospective study led by Corradini et al showed that cyclin-dependent kinase inhibitor 2A (CDKN2A) is frequently deleted in PTCL, NOS (46%) and is associated with refractory disease and poor prognosis, thus suggesting another group potentially requiring an alternative treatment approach.⁶ 

As mentioned above, a larger group of lymphomas is now classified under the umbrella of TFH lymphomas and includes angioimmunoblastic-type, follicular-type, and NOS. TFH lymphomas express at least 2 markers characterized by normal TFH cells (CD10, BCL6, CXCL13, PD1, and ICOS). These 3 subtypes have a common genetic landscape characterized by loss-of-function of genes that regulate DNA and histone methylation, such as TET2, DNA (cytosine-5)-methyltransferase 3A (DNMT3A), and isocitrate dehydrogenase 2 (IDH2). Mutations and fusions involving genes related to T-cell receptor signaling are also commonly seen, including RHOA, CD28, ICOS, and VAV1. In addition to the phenotypic and biologic overlap among these entities, the sensitivity of these diseases to histone deacetylase (HDAC) inhibitors and other epigenetic modifying agents justifies grouping them together.⁷ Furthermore, new insights into the interactions between tumor cells and the tumor microenvironment in TFH lymphomas are elucidating potential new therapeutic targets for this disease subset.^8,9 

ALCLs are characterized by pleomorphic tumor cells with strong CD30 expression. Nodal ALCLs include ALK-positive ALCL and ALK-negative ALCL based on the presence or absence of ALK translocation. Additional subtypes of ALK-negative ALCL have recently been identified, including DUSP22–rearrranged ALK-negative ALCL and TP63-R ALK-negative ALCLs. The TP63-R group is rare and is associated with poor prognosis, whereas DUSP22-rearrranged represents 19% to 30% of ALK-negative ALCL and may be associated with improved prognosis.¹ 

Table 1 displays a summary of newer targeted drugs for nodal PTCLs. Brentuximab vedotin (BV) is the best example of a targeted agent that altered the standard of care for a subtype of PTCL. BV, an anti-CD30 antibody drug conjugate, targets CD30, which is universally expressed in ALCL and up to 30% of other PTCLs.²⁸ It induced responses in 86% of patients with relapsed or refractory ALCL, as well as in 54% and 33% of patients with CD30-positive AITL and PTCL, NOS, respectively.^13,14 The ECHELON-2 study was a phase 3 randomized study that compared frontline therapy with cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP) and BV-cyclophosphamide, doxorubicin, prednisone (BV-CHP) for CD30-positive PTCLs. The study was enriched for patients with ALCL by design; thus, ALCL represented 70% of patients enrolled. Treatment with BV-CHP improved both PFS and overall survival (OS) without increasing toxicity compared with CHOP.²⁹ The overall 5-year PFS rates in the BV-CHP and CHOP arms were 51.4% and 43%, respectively.³⁰ The benefit of BV-CHP was less clear for patients with AITL and PTCL, NOS as the study was not powered to independently assess the impact of these subtypes.³⁰ Based upon this study, BV-CHP is now the standard frontline therapy for ALCL and other CD30-positive T-cell lymphomas; however, it is likely that some of the “other” CD30-positive lymphomas, which includes TFH lymphomas, may benefit from different approaches such as upfront incorporation of epigenetic modifying agents.

The HDAC inhibitors (HDACis) most extensively studied in PTCL include romidepsin, belinostat, and chidamide. Romidepsin selectively inhibits class I HDACs and was granted accelerated approval by the FDA in 2011 primarily based upon a single-arm, international, phase 2 study.³¹ The study included 130 patients with relapsed or refractory PTCL, including PTCL, NOS (53%), AITL (21%), and ALK-neg ALCL (16%). The overall response rate (ORR) and complete response (CR) rates were 25% and 15%, respectively, and the median duration of response was 28 months.¹¹ Patients with AITL benefited the most from treatment, with an ORR of 33%. In addition, 19% of patients with AITL responded for ≥12 months.³² A phase 3 study of CHOP with or without romidepsin for frontline treatment of PTCL was meant to serve as a confirmatory study for romidepsin approval; however, it failed to meet its primary end point, leading to voluntary withdrawal of romidepsin for the indication of relapsed or refractory PTCL.³³ The 2-year PFS in the romidepsin-CHOP and CHOP arms were 43.2% and 36.3%, respectively. Interestingly, a post hoc analysis of patients with TFH lymphomas enrolled onto the study showed significantly improved PFS for the romidepsin-CHOP arm (median PFS 19.5 vs 10.6 months, P = .04), demonstrating the potential for improved outcomes for this subgroup with frontline integration of an epigenetic modifying agent.³⁴ 

Belinostat, a pan-HDAC inhibitor with efficacy similar to that of romidepsin, received accelerated approval in 2014 based on an international, phase 2, single-arm study of 129 patients with relapsed or refractory PTCL.¹⁰ The ORR, CR rate, and median duration of response were 25.8%, 10.8%, and 13.6 months, respectively. As seen with romidepsin, the highest efficacy was observed for patients with AITL, with an ORR of 45.5% compared with 23.3% for PTCL, NOS. The confirmatory study for belinostat opened in October 2023 and is a 3-arm phase 3 study comparing belinostat-CHOP, pralatrexate-CHOP and CHOP for patients with untreated PTCL.

Chidamide, an oral selective class I HDACi, acquired approval by the China Food and Drug Administration in 2014 following a single-arm phase 2 study of 79 patients with relapsed or refractory PTCL. The ORR and CR rates were 28% and 14%, respectively; once again, a higher activity was observed in AITL, with an ORR of 50%.²⁴ Currently, chidamide is only available in China and numerous studies evaluating chidamide-based combinations are ongoing (Table 2).

Efficacy of the hypomethylating agent azacytidine was initially reported in a retrospective series of 12 patients with AITL, demonstrating ORR and CR rates of 75% and 50%, respectively.⁴⁶ The Oracle study was a phase 3 study comparing azacytidine to the investigator’s choice (bendamustine, gemcitabine, or romidepsin) for patients with relapsed or refractory TFH lymphomas, which aimed to confirm the impressive efficacy observed in the small retrospective series. Treatment with azacytidine was associated with longer PFS (median PFS 5.6 vs 2.8 months; P = .042); however, the study failed to meet its primary end point which required improvement in PFS with significance of P < .025. In addition, the efficacy of azacytidine was lower than expected with 3-month ORR of 33.3%.⁴⁷ Given the modest activity of azacytidine observed in Oracle, it may be better suited in combination with other agents. A phase 1/2 study of azacitidine plus romidepsin showed ORR and CR rates of 61% and 43%, respectively, in 23 patients with PTCL, with a preference for activity in patients with TFH lymphomas.²⁵ The implication of these data is unclear, due to the inclusion of both treatment-naïve and relapsed/refractory patients in the study. A small phase 2 study evaluated frontline treatment with azacytidine plus CHOP in 20 patients with PTCL, including 17 patients with TFH lymphoma. The results appear promising, with a 2-year PFS of 65.8% for all patients and 69.2% for patients with TFH lymphomas.⁴⁸ The place for azacytidine-CHOP in frontline therapy of PTCL will be investigated in the ongoing Alliance study (A051902), in which azacytidine plus CHOP (for age >60 years) or cyclophosphamide, doxorubicin, vincristine, etoposide, and prednisone (CHOEP) (for age ≤60 years) is 1 of the 3 arms in this phase 2 randomized study (NCT04803201).

Enhancers of zeste homolog (EZH) 1 and EZH2 are histone methyltransferases that specifically methylate histone H3 lysine 27 (H3K27). They function as part of a multiprotein complex called polycomb repressive complex 2 (PRC2) and contribute to the silencing of tumor suppressor genes and cell differentiation genes.^49,50 Valemetostat is a dual inhibitor of EZH1 and EZH2 that is approved in Japan for HTLV-1 associated adult T-cell lymphoma/leukemia. This approval was based on a single-arm, 25-patient, phase 2 study showing an ORR of 48% for relapsed/refractory adult T-cell lymphoma/leukemia.⁵¹ The promising activity of valemetostat extends to the more common PTCLs as well.²² The VALENTINE-PTCL01 study was a phase 2, single-arm study that evaluated valemetostat 200 mg by mouth (po) daily for patients with relapsed or refractory PTCL after 1 or more lines of therapy. Among the 119 evaluable patients, 50 (42%) had TFH lymphomas, 41 (34%) had PTCL, NOS, and 9 (6.8%) had ALCL. The primary end point for this study was ORR on computed tomography; however, the PET-based response was included as an exploratory end point. The ORR and CR rates by computed tomography were 43.7% and 14.3%, respectively. The median duration of response was 11.9 months. The ORR and CR rates by PET were 52.1% and 26.9%, respectively. Responses were observed across all PTCL subtypes; however, there was a trend toward increased efficacy among TFH lymphomas, with an ORR of 54% for TFH lymphomas compared with 31.7% for PTCL, NOS, and 33.3% for ALCL. A phase 3 confirmatory study of PTCL is under development.

Approximately 20% to 45% of cases of AITL cases contain a mutation within the IDH2 gene, specifically IDH2^R172. This mutation leads to hypermethylation and repression of genes involved in T-cell signaling and differentiation, which likely contributes to lymphomagenesis.^52,53 A phase 1 study of enasidenib, a selective inhibitor of mutant IDH2, included 2 patients from Memorial Sloan Kettering Cancer Center with IDH2-mutant AITL (NCT01915498). Both patients achieved transient responses (personal communication, S. M. Horwitz, 2023), providing a rationale for further evaluation of IDH2 inhibition in IDH2-mutant AITL.

The clinical importance of the PI3K pathway in the T-cell lymphomas was uncovered in a phase 1 study with duvelisib, in which 8 (50%) of 16 patients with PTCL achieved objective responses.⁵⁴ This promising activity led to the initiation of the PRIMO study, a multicenter phase 2 study that evaluated single-agent duvelisib, a PI3K-δ,γ inhibitor, in patients with relapsed or refractory PTCL following at least 1 line of therapy. Among the 101 patients enrolled, the ORR and CR rates were 49% and 34%, respectively. Most patients had the most common PTCL subtypes, including 52 (51%) with PTCL, NOS, 30 (30%), with AITL, and 15 (15%) with ALCL. Responses were more commonly observed for patients with PTCL, NOS (ORR, 48.1%; CR, 26.9%) and AITL (ORR, 66.7%; CR, 53.3%), and less commonly for ALCL (ORR, 13.3%; CR, 13.3%).¹⁵ One challenge with PI3-kinase inhibitors is treatment-related toxicities, such as hepatitis, which can lead to treatment interruption or discontinuation. This barrier was potentially overcome in a phase 1 study of duvelisib plus romidepsin that not only showed high efficacy but also showed reduced rates of duvelisib-related hepatitis, potentially due to the anti-inflammatory effects of romidepsin.²⁶ Many other PI3-kinase combinations are under evaluation and are summarized in Table 2. The potential for frontline use of duvelisib is being evaluated in the ongoing Alliance study (A051902) in which duvelisib plus CHOP (age >60 years) or CHOEP (age ≤60 years) is 1 of the 3 arms in the phase 2 randomized study (NCT04803201).

Evidence of Janus kinase/STAT (JAK/STAT) pathway activation in numerous PTCL subtypes has sparked many studies that target this pathway. An investigator-initiated study of ruxolitinib, a JAK1/2 inhibitor, in patients with relapsed or refractory PTCL and CTCL provided clinical evidence of JAK/STAT pathway dependence in PTCL.¹⁶ This study was designed to enrich for patients with T-cell lymphomas characterized by JAK or STAT activating mutations or with evidence of upregulation of the pathway through overexpression of phospho-STAT3 by immunohistochemistry. Among the 53 patients enrolled, 46 (87%) had PTCL and 7 (13%) had CTCL. JAK/STAT pathway activation was observed in 68% of the patients and was associated with higher efficacy. Clinical benefit (ORR plus disease stability lasting >6 months) was observed in 48% and 36% of patients with JAK/STAT activation mutations and phospho-STAT3 overexpression, respectively, compared with only 18% of patients without JAK/STAT activation. Although this study provides evidence of JAK/STAT pathway dependence in PTCL, the efficacy was modest, potentially due to dual inhibition of JAK1 and JAK2, which contributes to dose-limiting cytopenias. Golidocitinib, a selective JAK1 inhibitor, potentially overcomes this barrier. Golidocitinib is being evaluated in the phase 2 JACKPOT8 study, which enrolled 104 patients with relapsed/refractory PTCL to date, including 46.2% with PTCL, NOS, 15.4% with AITL, and 9.6% with ALCL.⁵⁵ Among 80 evaluable patients, 35 (43.8%) achieved a response, including 25% with CR. Higher response rates were observed in AITL (ORR 56.3%), and PTCL, NOS (ORR 45.7%) and less commonly in ALCL (ORR 11.1%). The JAK/STAT pathway status was not reported in this study; therefore, the impact of JAK/STAT pathway activation on the efficacy of golidocitinib is not known, but it appears that golidocitinib may be associated with a higher efficacy than ruxolitinib, potentially because of an improved therapeutic index.

Another strategy to target the JAK/STAT pathway is to facilitate STAT3 degradation. KT-333 is a small molecule degrader that targets STAT3 for degradation through the ubiquitin-proteasome system, which is currently being evaluated in a phase 1a/1b study for relapsed/refractory T-cell lymphomas (NCT05225584).

As mentioned above, ALK-positive ALCL is characterized by the activation of the ALK receptor tyrosine kinase by chromosomal translocation. Crizotinib is a small-molecular inhibitor of several tyrosine kinases, including ALK, which demonstrated an ORR of 90% in pediatric patients with relapsed or refractory ALK-positive ALCL.²⁰ Crizotinib was evaluated in combination with a standard frontline pediatric regimen (ALCL99 protocol) for patients with untreated ALK-positive ALCL and demonstrated promising efficacy (2-year event-free survival [EFS] of 76.8%) but a concerning rate of thromboembolic events (27% before anticoagulation mandated; 8% after anticoagulation mandated).⁵⁶ A separate arm of this trial evaluated BV plus the ALCL99 regimen and showed similarly high efficacy (2-year EFS 79%) but no increase in toxicity. Therefore, based on this study, as well as on ECHELON-2, BV-based therapy is the current frontline standard for ALK-positive ALCL.⁵⁷ Small studies with next-generation ALK inhibitors, such as alectinib and ceritinib, have shown high efficacy in ALK-positive ALCL and are therefore options for patients who develop resistance to crizotinib.^21,58 Furthermore, next-generation ALK inhibitors are options for patients with CNS relapse because of their ability to cross the blood brain barrier.⁵⁹ 

A subset of patients with ALK-positive ALCL overexpress platelet-derived growth factor receptor-β, which was shown to be associated with less favorable outcomes.⁶⁰ Preclinical data have demonstrated the sensitivity of ALK-positive ALCL to PDGFR blockade, leading to a clinical trial assessing imatinib and BV in relapsed or refractory ALK-positive ALCL.^61,62 Unfortunately, the study was closed early due to poor accrual; however, imatinib was shown to be highly effective in 4 of the 6 patients treated, and PDGFR expression was predictive of response.⁶⁰ Based on these observations, additional studies targeting the PDGFR axis in ALK-positive ALCL are warranted.

Despite the success of chimeric antigen receptor (CAR)-T therapy, bispecific antibodies, and checkpoint inhibition for other lymphoma subtypes, progress with respect to immune therapy has been slower in PTCL. Checkpoint inhibition targeting programmed death-1 (PD-1) is associated with modest efficacy but also with the potential risk of hyperprogression.^63,64 Studies evaluating PD-1-based combinations with chemotherapy may overcome these challenges and are listed in Table 2. Early data with bispecific antibodies are encouraging, such as with AFM13, an anti-CD30/CD16a antibody that aims to activate antitumor immunity by redirecting natural killer cells to CD30-positive tumor cells. AFM13 was evaluated in a phase 2 study for patients with relapsed or refractory CD30-positive PTCL and demonstrated an ORR of 32.4% and CR rate of 10.2% in 108 evaluable patients. Compared with the other subgroups, efficacy was particularly high in AITL, with ORR and CR rates of 53.3% and 26.7%, respectively.⁶⁵ Further investigations of this agent in combination with cord–derived natural killer cells (AB-101) are planned (NCT05883449). Additional bispecific antibodies are under investigation, such as ONO-4685, which targets PD-1 and CD3 (NCT05079282). Finally, a novel monoclonal antibody targeting CD94, which functions via antibody-dependent cellular cytotoxicity by means of fratricide, is under investigation for cytotoxic T-cell lymphomas (NCT05475925).

Numerous challenges are faced with the development of CAR T-cell therapy for T-cell lymphomas, including the risk of T-cell aplasia, fratricide and bidirectional killing, and contamination of CAR T-cell products with malignant T cells.⁶⁶ Ongoing studies are tackling these obstacles through various strategies, such as targeting limited T-cell subsets (to avoid T-cell aplasia), use of gene editing to knock-out antigens on CAR T-cell surfaces (to avoid fratricide), and use of allogeneic T cells (to avoid product contamination). For example, CTX130 is an allogeneic CAR T-cell therapy that targets CD70, which is highly expressed in many T-cell lymphomas but is less frequently expressed in normal T cells. CTX130 is modified with CRISPR/Cas9 editing to eliminate the expression of CD70, the T-cell receptor, and major histocompatibility complex class I, which aid in mitigating fratricide, graft-versus-host disease, as well as distruction by the host's immune system.⁴¹ In an ongoing phase 1 study, among 15 patients evaluable for response (7 with PTCL, 8 with CTCL), efficacy is encouraging, with an ORR of 71% and a CR rate of 29% (NCT04502446).⁴¹ Additional CAR T-cell studies in PTCL are targeting CD5 (NCT04594135), TRBC1 (NCT03590574), CD37 (NCT04136275), and CD7 (NCT05290155).

Due to the lack of data to support the curative potential of alternative approaches, common nodal PTCLs have been treated similarly with CHOP, often with etoposide (CHOEP), followed by high-dose therapy and autologous stem cell transplantation (ASCT) in first remission.⁶⁷ CHOP has been the foundation of treatment for these diseases; however, with the exception of ALK-positive ALCL, <50% of the patients are cured with CHOP alone. Based on data from the retrospective International T-cell project, in which 85% of patients with common subtypes received CHOP, the failure-free survival rates at 5 years are as low as 20%, 18%, and 36% for PTCL NOS, AITL, and ALCL, respectively.⁶⁸ Similar outcomes were observed in a series from the British Colombia Cancer Center and Swedish National Registry.^69,70 Interestingly, despite the overall disappointing outcomes of CHOP in these retrospective registries, treatment with an anthracycline-containing regimen was associated with improved outcomes in the COMPLETE prospective PTCL registry.⁷¹ This observation from COMPLETE, as well as the better-than-expected outcomes showing the curative potential for patients enrolled in the CHOP control arms of ECHELON-2 (5-year PFS 43% and OS 61%) and the romidepsin-CHOP study (2-year PFS 43.2% and OS 63.4%), provide support to further build upon the CHOP backbone.^30,33 

To improve upon CHOP, many studies evaluated CHOP plus X approaches; however, before ECHELON-2, which established the use of brentuximab vedotin (BV) plus cyclophosphamide, doxorubicin, and prednisone (BV-CHP) as frontline therapy for CD30-positive PTCLs, the only CHOP + X regimen that was widely used was CHOEP.²⁹ Data supporting the addition of etoposide to CHOP are limited and primarily based on a retrospective analysis of 7 prospective studies conducted by the German High-Grade Non-Hodgkin Lymphoma (DSHNHL) study group.⁷² This analysis demonstrated improved EFS in younger patients (<60 years old) with normal lactic dehydrogenase treated with CHOP plus etoposide compared with CHOP alone, whereas older patients had less benefit with etoposide, likely due to toxicity. Studies evaluating the addition of other agents to CHOP, such as alemtuzumab, denileukin difitox, bortezomib, and romidepsin, have typically shown increased toxicity without clear benefits in unselected patients with PTCL.^33,73-75 Current and future studies investigating novel CHOP-based combinations in selected patient populations are more likely to identify patients who will benefit from newer CHOP + X approaches and those requiring alternative approaches. For example, the single-arm azacytidine-CHOP study mentioned above showed promising efficacy in patients with TFH lymphomas, except for those whose disease harbored DNTM3A mutations, suggesting the need for a different treatment approach in this subgroup.⁴⁸ Likewise, the REVAIL study, which evaluated CHOP plus lenalidomide for patients with TFH lymphomas, showed significantly lower PFS for patients bearing certain DMNT3A mutations, again indicating that this subgroup may require an alternative strategy.⁷⁶ 

ASCT is frequently used in first remission to improve the outcomes observed with CHOP-based therapy alone; however, data supporting this approach are limited to single-arm phase 2 studies. The largest of these studies were from the Nordic group by d’Amore et al and included 160 patients treated with CHOEP or CHOP (for patients aged >60 years), followed by ASCT for responders.⁷⁷ This study enrolled 160 patients with PTCL and the 5-year PFS was 44%, which was considerably higher than that observed with CHOP alone in the International T-cell project and other retrospective series.^68-70 The patients with ALK-negative ALCL experienced the most favorable outcomes, with a 5-year PFS of 61%, compared with 49% and 38% for AITL and PTCL, NOS, respectively. A smaller prospective study by Reimer et al evaluating CHOP followed by ASCT mirrored these results, with a 3-year OS of 48% for all patients and 71% for patients proceeding to transplantation.⁷⁸ Notably, consolidation with ASCT for patients enrolled in ECHELON-2 who achieved CR after BV-CHP showed significantly improved PFS compared with those who did not pursue ASCT (5-year PFS, 65.3% vs 46.4%).⁷⁹ Although the plan for transplantation was based on physician discretion, this post hoc analysis provides further support for the use of ASCT in the first CR, which appears beneficial even after targeted induction therapy, such as BV-CHP.

Taken together, we use a CHOP-based approach for patients with nodal PTCL eligible for curative therapy (Figure 1). In the absence of a clinical trial, patients with ALCL and CD30-positive PTCLs receive induction therapy with BV-CHP. Patients with CD30-negative PTCL change to receive CHOEP (if eligible for intensive therapy) or CHOP. We offer consolidation with ASCT in the first CR for fit patients with higher-risk ALK-positive ALCL (International Prognostic Score ≥2) and those with other common nodal PTCL entities.

The only reliable curative treatment for PTCL in a relapsed/refractory setting is allogeneic stem cell transplantation (allo-SCT) and undergoing transplantation in complete remission ensures more favorable outcomes.⁸⁰ The choice of second-line therapy and beyond should take into account whether or not allo-SCT is planned as well as disease subtype. If allo-SCT is not considered, treatment that can be continued on an ongoing basis without significant cumulative toxicity should be prioritized. Generally, single-agent therapy, rather than combination chemotherapy, is preferred to reduce toxicity and optimize quality of life. For TFH lymphomas, epigenetic modifying agents and duvelisib should be considered early on, given their sensitivity to these agents. For ALCL, BV should be considered if the remission duration is >6 months from the frontline therapy with BV. For ALK-positive disease, an ALK inhibitor is used for patients already exposed to BV. When allo-SCT is planned, combination therapy can be considered to increase chance of achieving CR quickly. Again, for TFH lymphomas, therapy with HDACi and/or duvelisib should be prioritized.

Treatment of nodal PTCLs will continue to evolve in the upcoming years. Although we still have a long way to go, we are accumulating the necessary tools to test individualized therapies for PTCL. These tools include improved classification, identification of predictive biomarkers, and the development of targeted agents. Our growing knowledge of disease biology and drug sensitivity enables us to rationally design studies with a higher potential for success. We are also learning from previous negative studies, such as romidepsin-CHOP, to optimally design future studies. When the romidepsin-CHOP study was designed, the unique sensitivity of TFH lymphomas to HDACis was not yet recognized.^33,81 In the ongoing Alliance study (A051902), patients are stratified by diagnosis of TFH lymphoma, which will better position us to learn whether frontline incorporation of either azacytidine or duvelisib is appropriate for this group.

Continued revision of the PTCL classification will enable us to optimize therapy for patients in the future. Multi-institutional, international molecular profiling efforts in PTCL, such as T-cell Project 2.0 (NCT03964480), will aid in the improved characterization of the multiple entities that currently fall under the PTCL, NOS umbrella. These studies will potentially validate the clinical significance of PTCL-GATA3 and PTCL-TBX21 and expand our understanding of which patients benefit from which therapies, as well as identify other subgroups (such as TP53 mutated or CDKN2A deleted PTCL, NOS) that potentially warrant variations in current treatment approaches.

Promising new drugs, such as PI3K inhibitors, JAK inhibitors, EZH1/2 inhibitors, and other epigenetic modifying agents, are changing the landscape of PTCL. In addition, despite the barriers to the development of immune therapies for PTCL, novel strategies are under investigation and have the potential to significantly alter the treatment landscape as well.

Future treatment of PTCL will recognize numerous distinct entities and appropriately incorporate novel agents to optimize therapy for each individual (Figure 2). We anticipate that these efforts will lead to dramatic improvements in the prognosis for patients with PTCL.

The authors acknowledge and thank Nivetha Ganesan for her assistance in designing the visual abstract.

The authors are grateful for support from Nonna's Garden Fund and National Institutes of Health, National Cancer Institute Core Grant P30 CA008748. A.J.M is a Scholar in Clinical Research of The Leukemia & Lymphoma Society

Contribution: A.J.M., S.M.H., and R.N.S. designed and wrote the manuscript.

Conflict-of-interest disclosure: A.J.M. received research support from Seattle Genetics, Merck, AstraZeneca, Affimed, Bristol Myers Squibb, Incyte, and SecuraBio; and honorarium from Affimed, Merck, Seattle Genetics, and Takeda. S.M.H. received consulting fees from Affimed, Abcuro Inc, Corvus, Daiichi Sankyo, Kyowa Hakko Kirin, ONO Pharmaceuticals, SeaGen, SecuraBio, Takeda, and Yingli; and research support from ADC Therapeutics, Affimed, C4, Celgene, CRISPR Therapeutics, Daiichi Sankyo, Dren Kyowa Hakko Kirin, Millennium/Takeda, Seattle Genetics, and SecuraBio. R.N.S. declares no competing financial interests.

Correspondence: Steven M. Horwitz, David H. Koch Center for Cancer Care at Memorial Sloan Kettering Cancer Center, 530 E 74th St, New York, NY 10021; email: horwitzs@mskcc.org.