The aggressive peripheral T-cell lymphomas: 2012 update on diagnosis, risk stratification, and management


  • Conflict of interest: Consultant for Ziopharm, Seattle Genetics, Genentech, Allos, Roche.


Background: T-cell lymphomas make up approximately 10–15% of lymphoid malignancies. The frequency of these lymphomas varies geographically, with the highest incidence in parts of Asia. Diagnosis: The diagnosis of aggressive peripheral T-cell lymphoma (PTCL) is usually made using the WHO classification. The ability of hematopathologists to reproducibly diagnose aggressive PTCL is lower than for aggressive B-cell lymphomas, with a range of 72–97% for the aggressive PTCLs. Risk Stratification: Patients with aggressive PTCL are staged using the Ann Arbor Classification. Although somewhat controversial, positron emission tomography (PET) scans appear to be useful as they are in aggressive B-cell lymphomas. The most commonly used prognostic index is the International Prognostic Index. The specific subtype of aggressive PTCL is an important risk factor, with the best survival seen in anaplastic large-cell lymphoma—particularly young patients with the anaplastic lymphoma kinase positive subtype. Risk Adapted Therapy: Anaplastic large-cell lymphoma is the only subgroup to have a good response to a cyclophosphamide, doxorubicin, vincristine, prednisone (CHOP)-like regimen. Angioimmunoblastic T-cell lymphoma has a prolonged disease-free survival in only ∼20% of patients, but younger patients who have an autotransplant in remission seem to do better. PTCL-not otherwise specified (NOS) is not one disease. Anthracycline containing regimens have disappointing results and a new approach is needed. NK/T-cell lymphoma localized to the nose and nasal sinuses seems to be best treated with radiotherapy containing regimens. Enteropathy associated PTCL and hepatosplenic PTCL are rare disorders with a generally poor response to therapy, although selected patients with enteropathy associated PTCL seem to benefit from intensive therapy. Am. J. Hematol. 87:511–519, 2012. © 2012 Wiley Periodicals, Inc.


T-cell lymphomas make up approximately 10–15% of lymphoid malignancies [1]. The frequency of these lymphomas varies geographically, with the highest incidence in parts of Asia [2]. The current World Health Organization (WHO) classification of T-cell malignancies is presented in Table I. T-cell lymphomas can be divided into those of precursor T-cells (i.e., precursor-T lymphoblastic lymphoma) and those arising in more mature T-cells, with the later termed peripheral T-cell lymphomas (PTCLs). As is true for B-cell lymphomas, a subset of PTCLs such as mycosis fungoides and the CD30 positive cutaneous lymphoproliferative disorders (i.e., cutaneous anaplastic large-cell lymphoma and lymphomatoid papulosis) have a prolonged natural history. The aggressive PTCLs are associated with a short survival and are the topic of this update. These can be subdivided into those of primarily nodal origin and those that typically present in specific extra nodal sites and are often associated with characteristic clinical syndromes (Table II).

Table I. Peripheral T-Cell Lymphomas In The WHO Classification [8]a
  • a

    Italicized lymphomas are those discussed in this article.

Usually indolent with extra-nodal presentation
 Mycosis fungoides and variants
 Cutaneous CD30 positive cutaneous lymphoproliferative disorders
  Lymphomatoid papulosis
  Cutaneous anaplastic large cell lymphoma
 Subcutaneous panniculitis-like T-cell lymphoma
 Primary cutaneous CD4-postive small/medium T-cell lymphoma
Usually aggressive-typically nodal presentation
 Angioimmunoblastic T-cell lymphoma
 Anaplastic large cell lymphoma
  ALK positive
  ALK negative
 Peripheral T-cell lymphoma–not otherwise specified
 Adult T-cell lymphoma/leukemia
Usually aggressive-typically extra-nodal presentation
 Extra-nodal NK/T cell lymphoma, nasal type
 Enteropathy type intestinal T-cell lymphoma
 Hepatosplenic T-cell lymphoma
 EBV positive lymphoproliferative disorders of childhood
 Primary cutaneous gamma-delta T-cell lymphoma
 Primary cutaneous CD-8 positive aggressive epidermotropic cytotoxic T-cell lymphoma
Typically Leukemic Presentation
 T-cell prolymphocytic leukemia
 T-cell large granular lymphocytic leukemia
 Chronic lymhoproliferative disorders of NK cells
 Aggressive NK-cell leukemia
Table II. Characteristics of the Major Subgroups of Aggressive PTCL. [9]
Subtype% of PTCLs worldwide/no America% Reproducibility of histologic diagnosisMedian age% Stage I/II% Bone marrow positive“Standard” treatment% 5-year PFS OS“New” Treatments
  • a

    Reference [40].

PTCL-NOS26/3475603122Clinical triala2032Pralatrexate, romadepsin, dasatinib
AITL19/1681651129Clinical triala1832Cyclosporine, bevacizumab, lenalidomide
 ALK+7/1697343512CHOP ± E6070Brentuximab, crizotinib (ALK+)
 ALK−6/87458427Clinical triala3649 
Nasal NK/T-Cell Lymphoma10/592527310Localized-XRT ± chemo disseminated SMILE2942L-Asparaginase
Enteropathy Type5/67961313Intensive chemo-therapy and autotrans-plantation420
Hepatosplenic1/37234574none07Pentostatin, alemtuzumab, cladribine

Diagnosis of Peripheral T-Cell Lymphoma

The most widely used system for classifying non-Hodgkin's lymphomas until the mid-1990s was the Working Formulation [3]. Although previous diagnostic systems recognized the distinction between T-cell versus B-cell lymphoma [4, 5], it was the development of the REAL classification that was the stimulus for wide spread recognition of the significance of diagnosing PTCLs as specific entities [6]. The REAL classification was shown to be clinically relevant [1] and the system was largely adopted as the WHO classification for lymphomas published in 2001 [7]. The system was updated in 2008 [8] (Table I). The new system has provided increasingly clear definitions for the diagnosis of the major subtypes of PTCL.

Hematopathologists have not been equally able to reproducibly diagnose B-cell and T-cell lymphomas. Although with the use of immunophenotyping they are able to recognize the presence of a PTCL as often as they can diagnosis the major subtypes of B-cell lymphoma [1], their ability to agree on the diagnosis of specific subtypes of PTCL ranged from 97% for anaplastic lymphoma kinase (ALK) positive anaplastic large-cell lymphoma to 72% for hepatosplenic PTCL and was only 75% for the most common subtype, PTCL-NOS, in a large international clinical study [9]. It is likely that the reproducibility of the diagnosis of specific subtypes of PTCL will improve as pathologists gain increasing experience in the use of the WHO classification.

Other than the use of immunophenotyping to prove that a neoplasm is a T-cell process, other molecular and genetic studies have not made the same inroads in diagnosing PTCLs as they have in B-cell lymphomas. The only frequent recurring chromosomal translocation seen in the PTCLs is the t(2;5)(p23;q35) that is characteristic of ALK positive anaplastic large-cell lymphoma. This translocation abuts the ALK gene on chromosome 2 to the NPM1 gene on chromosome 5 and leads to the overexpression of ALK. Several other variant translocations involving ALK have been identified. ALK is also expressed by a small subset of diffuse large B-cell lymphomas and by approximately 5–10% of non-small cell lung cancers. Other molecular associations with PTCLs include IRF4 rearrangements with ALK negative anaplastic large-cell lymphoma [10], deletions of chromosome 6q in nasal type NK/T-cell lymphoma [11], 9q abnormalities in enteropathy associated T-cell lymphoma [12], and isochromosome 7q in hepatosplenic PTCL [13]. Except for the t(2;5) in ALK positive anaplastic large-cell lymphoma, these genetic abnormalities have not made a major impact on the diagnosis of the other subtypes.

Gene expression profiling provides the opportunity to further improve our ability to recognize specific subtypes of PTCL, and to “sort out” the imprecise PTCL-NOS. Profiling studies to date have shown that several subtypes of PTCLs have reproducible and recognizable genetic signatures. These include angioimmunoblastic T-cell lymphoma, ALK positive anaplastic large-cell lymphoma, and adult T-cell lymphoma [14]. PTCL-NOS appear to have multiple subtypes, and approximately 20% have a genetic signature that suggests they are closely related to angioimmunoblastic T-cell lymphoma. Angioimmunoblastic PTCL appears to arise in follicular helper T-cells which might explain some of its histological and clinical characteristics [15].

Staging and Risk Stratification

PTCLs are usually staged by the Ann Arbor system [16] using the same techniques as are used for staging patients with aggressive B-cell lymphomas. However, in one large international study, patients with PTCLs, with the exception of anaplastic large-cell lymphoma, were much more likely to be Stage III or IV than patients with diffuse large B-cell lymphoma [1]. Although the value of utilizing PET scans in patients with PTCLs has been questioned, recent studies suggest that PET scans might be as useful in these patients as in those with the aggressive B-cell lymphomas [17–19]. However, it appears that the SUVMAX in patients with PTCLs is somewhat lower than in aggressive B-cell lymphomas, and that the SUVMAX in extra nodal PTCLs might be lower than in those presenting primarily in nodal sites [20]. At least one study has suggested that the value of post-treatment PET scans might be lower in PTCLs than in corresponding B-cell lymphomas, with a negative scan not necessarily associated with an improved treatment outcome [21]. Others have found that PET scans in T-cell lymphomas are predictive, similar to B-cell lymphoma [22]. Our group and others use PET scans in the management of patients with PTCLs in a manner similar to that used for diffuse large B-cell lymphoma.

The most commonly used prognostic system for patients with aggressive PTCL is the International Prognostic Index (IPI) [23]. The IPI score is inversely related to survival in patients with PTCL in a manner similar to that seen in diffuse large B-cell lymphoma [9], but the overall survival in each IPI category appears lower in patients with most PTCLs than seen in diffuse large B-cell lymphoma. Attempts to improve the predictive ability of the IPI have been carried out. The PTCL prognostic index (PIT) [24] does not appear to be clinically more useful than the original IPI.

Anaplastic large-cell lymphoma

Anaplastic large-cell lymphoma was the third most frequent type of aggressive PTCL found in the international review of more than 1,000 cases of PTCLs [9]. In that study 65% of the cases of anaplastic large-cell lymphomas were ALK positive and 45% were ALK negative. The patients with ALK positive anaplastic large-cell lymphoma were much younger (i.e., median age 34 years) than those with ALK negative anaplastic large-cell lymphoma (median age 58 years) [25]. With both ALK positive and ALK negative anaplastic large-cell lymphoma, more than 60% of the patients were male. Stage IV disease was seen in slightly more than one-third of patients with the most frequent extra-nodal sites of disease being bone, bone marrow, and lung—all present in approximately 10% of patients [25]. However, any extra-nodal site can be involved with anaplastic large-cell lymphoma.

The failure-free and overall survival for patients with ALK-positive and ALK-negative anaplastic large-cell lymphoma is presented in Fig. 1 and Fig. 2. The five-year overall survival (70% vs. 49%) and five-year failure-free survival (60% vs. 36%) both favor patients with ALK-positive anaplastic large-cell lymphoma [25]. However, age was a powerful predictor of survival in patients with ALK-positive anaplastic large-cell lymphoma and in patients over 40 years of age (i.e., few patients with ALK-negative anaplastic large-cell lymphoma are younger than 40 years of age) there was no difference in survival between the two subtypes.

Figure 1.

Failure-free survival for patients with aggressive peripheral T-cell lymphoma by subtype from the International Peripheral T-Cell and NK/T-Cell Lymphoma Study [9]. [Color figure can be viewed in the online issue, which is available at]

Angioimmunoblastic T-cell lymphoma

In the 1970s, Rappaport and coworkers first described an entity called angioimmunoblastic lymphadenopathy with dysproteinemia [26]. It was uncertain that this condition represented a malignancy, and some patients seemed to have long-term benefit from single agent prednisone. However, further observations made it clear that this entity represented a PTCL and it is now recognized in the WHO classification as angio-immunoblastic T-cell lymphoma [8]. The International Peripheral T-Cell and Natural Killer T-cell Lymphoma Study of more than 1,100 cases worldwide [9] found angioimmunoblastic T-cell lymphoma to make up 18.5% of all cases. The disease was not evenly distributed worldwide, with the highest incidence being 29% of all cases of PTCL found in Europe.

Patients with angioimmunoblastic T-cell lymphoma tend to be older and to present with systemic symptoms and advanced disease. In the International Peripheral T-Cell and Natural Killer T-Cell Lymphoma Study the median age was 65 years, 56% were male, and 89% had Stage III or IV disease. Only 14% presented with an IPI score of 0 or 1. In a French study of 157 patients, the median age was 62 years, 81% of the patients had Stage III or IV disease, 72% had B-symptoms, and 60% were male [27].

Peripheral T-cell lymphoma not otherwise specified

Although PTCL-NOS is the most common subtype found in series of mixed PTCLs, it does not represent a specific entity. As many as 20% of cases diagnosed as PTCL-NOS have been found to have a gene expression profile characteristic of angioimmunoblastic T-cell lymphoma [14]. In addition, another subgroup of patients with PTCL-NOS has features of cytotoxic T-cells and might represent a unique entity [14]. The latter group of patients had a worse prognosis than others with PTCL-NOS.

Patients with PTCL-NOS have a median age of about 60 years, a striking male predominance, and usually present with widespread disease [9]. While the International Prognostic Index has a significant association with treatment outcome, patients with IPI scores of 0/1 have only a 33% five-year failure-free survival and a 50% five-year overall survival [9]. Patients with high IPI scores of 4/5 have a five-year progression-free survival of only 6% and a five-year overall survival of only 11%.

NK/T-cell lymphoma

The tumor that is termed nasal NK/T-cell lymphoma in the WHO classification can be confusing to most North American oncologists. This lymphoma has a striking geographic variation with a high incidence in parts of Asia, Latin America, and those areas in North America with a large immigrant population from high incidence areas. Thus, most North America oncologists see this disease rarely. Terminology furthers the confusion. Most cases present involving the nose and nasal sinuses and represent modern terminology for the lesion that was once called lethal mid-line granuloma. Patients with this presentation often have local/regional disease, although metastases can occur. However, patients with what histologically appears to be the same lymphoma can present in other sites including the skin, soft tissue, gastrointestinal tract, testes, and others. The current terminology for these is “extra nodal NK/T-cell lymphoma, nasal type,” recognizing that some of the cases can be “non-nasal.”

These lymphomas morphologically show an angiocentric and angiodistructive pattern. They have an extremely high association with Epstein Barr Virus (EBV) and express viral DNA in the tumor cells. There is circulating EBV DNA and evidence that patients with higher titers have a poorer prognosis [28]. Patients with local/regional disease involving the nose and nasal sinuses have the best outlook, while patients who present in other sites and have wide spread disease havean extremely poor outlook with therapies applied to date [29].

Enteropathy associated T-cell lymphoma

Enteropathy associated T-cell lymphoma is an aggressive disorder frequently associated with celiac disease [30, 31]. Patients with celiac disease have an increased risk of various malignancies [32] including adenocarcinoma of the small bowel and diffuse large B-cell lymphoma. It is important to remember that a patient with known celiac disease who develops what appears to be a lymphoma in the small bowel might not have enteropathy associated T-cell lymphoma. The incidence of enteropathy associated T-cell lymphoma varies geographically and seems to follow the frequency of celiac disease [9].

Not all patients diagnosed with enteropathy associated T-cell lymphoma have known celiac disease [33]. In a large international study of T-cell lymphomas, only one-third of the patients with enteropathy associated T-cell lymphoma had previously been diagnosed with celiac disease [33]. Since celiac disease is frequently under diagnosed, it may be that many of the patients had unrecognized sprue. It has been proposed that enteropathy associated T-cell lymphoma associated with celiac disease typically has a chromosome 9q gain or a 16q deletion and the human leukocyte antigen (HLA)-DQ2 haplotype. Enteropathy associated T-cell lymphoma not associated with celiac disease is said to be more likely to have a chromosome 8q gain and less likely to display the HLA-DQ2 haplotype [34, 35]. The morphologic features of these two subtypes of enteropathy associated T-cell lymphoma are said to be different and correlate with the finding of clinical or histological evidence of sprue in the patient [33].

Patients with enteropathy associated T-cell lymphoma typically present in their 6th decade with abdominal pain, fatigue, and anorexia as the most frequent symptoms [9, 33]. It might be that increasing recognition of the risk of this disorder has reduced the likelihood that patients with refractory sprue and lymphoma will present with a bowel perforation. That presentation was previously associated with a very poor prognosis.

Hepatosplenic T-cell lymphoma

Hepatosplenic T-cell lymphoma is a rare disease seen most frequently in young men and often presents a diagnostic dilemma [36]. The tumor cells are most frequently found in the spleen, liver, and bone marrow. However, because they infiltrate the organs diffusely without growing as tumorous masses, the cells can be overlooked and the diagnosis missed unless slides are reviewed by an experienced hematopathologist who is aware of the disorder. Patients are often young males who present with a systemic illness with fever, enlarged liver and spleen, and cytopenias. This lymphoma has been reported in patients with an altered immune system including Crohn's disease [37] and after renal transplantation [38, 39]. In the absence of a high degree of suspicion, the diagnosis might be overlooked.

Risk-Adapted Therapy of Specific Subtypes (Table II and Figs. 1 and 2)

Figure 2.

Overall survival for patients with aggressive peripheral T-cell lymphoma by subtype from the International Peripheral T-Cell and NK/T-Cell Lymphoma Study [9]. [Color figure can be viewed in the online issue, which is available at]

Anaplastic Large-cell lymphoma

Anaplastic large-cell lymphoma is the only aggressive PTCL where a significant group of patients can be cured utilizing CHOP-like regimens. However, CHOP is not an optimal treatment for these patients with 40% of patients with ALK-positive disease and over 60% of patients with ALK-negative disease failing to be cured using this approach [25]. German investigators have reported an improved failure-free survival when patients receive CHOP plus etoposide [41]. The doxorubicin, cyclophosphamide, vindesine, bleomycin, prenisolone (ACVBP) regimen, which includes a high dose consolidation phase after the initial induction, is highly active in diffuse large B-cell lymphoma and has also been utilized in patients with anaplastic large-cell lymphoma [42]. The five-year survival rate was 63% utilizing ACVBP—results that did not appear greatly different from that achieved utilizing CHOP. In a pediatric series, vinblastine was shown to be a very active drug as a single agent in patients with relapsed anaplastic large-cell lymphoma [43]. However, when vinblastine was added to an intensive, leukemia-like chemotherapy regimen, it did not reduce the frequency of relapse [44].

The most exciting new drug in the treatment of patients with anaplastic large-cell lymphoma is the antibody toxin conjugate brentuximab vedotin. In a recently presented Phase II trial of 58 patients with recurrent anaplastic large-cell lymphoma the objective response rate was 86% and complete remission rate was 57%. The median duration of response was 13 months overall with a median of 17 months in patients who achieved a complete remission. The median overall survival in patients in this study has not been reached [45].

For patients with ALK-positive anaplastic large-cell lymphoma, the ALK inhibitor crizotinib is an oral agent that appears to be active in patients with recurrent ALK-positive anaplastic large-cell lymphoma. A recent report of two patients with recurrent disease reported a complete remission in both that was ongoing at the time of the report [46]. Crizotinib have been approved in the USA for use in ALK-positive non-small cell lung cancer.

The place of bone marrow transplantation in patients with anaplastic large-cell lymphoma is uncertain. It is clear that both autologous and allogeneic transplantation can cure some patients with recurrent lymphoma and a high proportion of the patients transplanted in initial remission [47]. However, the durability of complete remissions induced in patients with anaplastic large-cell lymphoma with CHOP-like regimens make it unclear that transplantation is likely to improve the results. However, some physicians do favor autotransplantation in first remission in patients with ALK-negative anaplastic large-cell lymphoma. Allogeneic hematopoietic stem cell transplantation should probably be reserved for patients who are young, have an HLA matched donor, and have recurrent lymphoma.

The therapy of a recently recognized clinical syndrome involving ALK-negative anaplastic large-cell lymphoma in women with breast implants deserves special mention. Although a variety of subtypes of lymphoma can arise in this situation [48], anaplastic large-cell lymphoma seems particularly frequent. Patients with this presentation of anaplastic large-cell lymphoma should have the implant removed. The relative merits of observation after having the implant removed as initial management for localized disease, or adjuvant CHOP-like chemotherapy or radiotherapy is uncertain.

Angioimmunoblastic T-cell lymphoma

The optimal treatment for patients with angioimmunoblastic T-cell lymphoma is not certain. Patients from the International Peripheral T-Cell and Natural Killer /T-Cell Lymphoma Study, largely treated with CHOP-like regimens, had a five-year overall survival of 32% and a five-year failure-free survival of 18% [9]. Patients with an IPI score of 0 or 1 had a five-year overall survival of 56% in contrast to 25% for those with an IPI score of 4–5. The five-year progression-free survival was only 34% for those with an IPI score of 0/1 and was 16% with an IPI score of 4/5 [9].

One-hundred forty-seven patients in a large French experience received some form of anthracycline-based chemotherapy, with the most common being ACVBP in younger patients [27]. A complete response was achieved in 46% of the patients, but 20% of the treated patients developed severe infections during therapy. The five-year overall survival rate was 33% and the five-year event-free survival rate was 29%. In the 1990s, a German group reported 67 patients with angioimmunoblastic T-cell lymphoma [49]. Twenty-eight patients were treated primarily with prednisone and received aggressive chemotherapy if they progressed. Another 11 received primary chemotherapy—usually utilizing an anthracycline containing regimen. The complete response rate with initial prednisone was 29% and with chemotherapy 64%. Chemotherapy-treated patients had a three-year event-free survival of 35%. The fact that three of eight patients who achieved a complete remission with prednisone alone continued in remission for as long two years at the time of the report is interesting. A review of other series of treatment of patients with angioimmunoblastic T-cell lymphoma revealed no obvious superior regimen [50].

A number of alternate therapies have been utilized in patients with angioimmunoblastic T-cell lymphoma. The presence of EBV infected B-cells in the tumor has led to clinical trials of rituximab. One report from France was initially positive [51], but a longer follow-up and more patients did not find a benefit of adding rituximab to a chemotherapy regimen [52]. Twelve patients were reported who received single agent cyclosporine as salvage therapy (i.e., 10 patients) or as initial treatment in two unwell/elderly patients [53]. Eight patients responded and three were continuing well at the time of the report. Bevacizumab has also been reported to induce complete remissions in patients with relapsed angioimmunoblastic T-cell lymphoma [54, 55]. A recent report has also suggested that lenalidomide has activity [56].

A common approach to treating patients with angioimmunoblastic T-cell lymphoma, and the one most often utilized by our team, involves induction chemotherapy utilizing an anthracycline containing regimen and autologous hematopoietic stem cell transplantation in remission. A recently reported German trial found that patients with angioimmunoblastic T-cell lymphoma benefited more from this approach than those with other subtypes of T-cell lymphoma [57]. A large European series of 146 patients with angioimmunoblastic T-cell lymphoma who underwent autologous transplantation found that patients transplanted in first initial remission did much better than those who were transplanted as part of salvage therapy [58]. Patients transplanted in first complete remission had a four-year progression-free survival of 56% in contrast to 30% for patients with chemotherapy sensitive relapse disease and 23% for those with chemotherapy refractory disease. Allogeneic hematopoietic stem cell transplantation has been reported from the same group in 45 patients with angioimmunoblastic T-cell lymphoma [59]. The one-year treatment-related mortality was 25% and the three-year progression-free survival was 53%. Eleven of these patients had previously failed an auto transplant.

The most common treatment currently applied for young, fit patients with angioimmunoblastic T-cell lymphoma is remission induction with an anthracycline containing chemotherapy regimen and an autologous transplant in first response. For older or unfit patients, less intensive chemotherapy, cyclosporine as a single agent, or prednisone might all be considered. Autologous transplantation can salvage a few patients who relapse after standard chemotherapy regimens, and allogeneic transplantation can lead to long-term survival in patients who fail other treatment approaches—albeit with a significant mortality risk.

Peripheral T-cell lymphoma not otherwise specified

The best treatment for patients with PTCL-NOS remains uncertain. In an International study of almost 300 patients, there was no hint that patients who received an anthracycline as part of their initial treatment did any better than patients who did not receive therapy with an anthracycline [9]. The National Comprehensive Cancer Network Guidelines for the treatment of patients recommend a clinical trial as the best initial treatment option [40]. There is some evidence that the combination of CHOP plus etoposide is more effective than CHOP alone [41], and this is the regimen that I favor. Other regimens that have been utilized include CHOP plus alentuzumab [60], CHOP plus denileukin diftitox [61] CHOP plusbortezomib [62], and other intensive regimens such as ACVBP [63] and hyper-CVAD [64]. A variety of gemcitabine containing regimens have also been studied [65]. However, none of these regimens has been shown to be clearly superior.

A number of groups, including ours, routinely do an autotransplant in first remission in high-risk patients with PTCL-NOS. There is at least some evidence that adjuvant autologous transplantation might have a superior outcome to what has been seen with standard regimens [66, 67]. For patients failing primary therapy, a wide variety of new drugs are being studied. Pralatrexate [68] and romidepsin [69] have now been approved for use in the United States for relapsed/refractory PTCL. Other new agents including lenalidomide [70], desatinib [71], and the aurora kinase inhibitor alisertib [72] all show some promise. Autologous transplantation at the time of treatment failure can cure some patients with chemotherapy sensitive disease [47]. Allogeneic hematopoietic stem cell transplantation, although more risky than autologous transplantation, has also been shown to cure some patients with relapsed/recurrent PTCL-NOS, and should be considered in younger patients with HLA-matched sibling donors [47].

NK/T-cell lymphoma

Patients with local/regional nasal NK/T-cell lymphoma involving the nose and nasal sinuses seem to uniquely benefit from the use of radiotherapy [73]. This is the only situation with aggressive lymphomas where initial treatment with radiotherapy rather than chemotherapy seems to be most effective. Whether adjuvant chemotherapy is necessary to have the highest cure rate is not certain, but it is almost always administered—sometimes in combination with radiotherapy and sometimes after completion of radiotherapy. In the International Peripheral T-Cell/NK Cell Lymphoma Study, patients with nasal NK/T-cell lymphoma who presented with a low IPI score of 0/1 had a five-year progression-free survival of 53% and a five-year overall survival of 57%. However, patients who presented with a high IPI score of 4/5 had 0% survival at five years [9]. Patients who had primary sites other than the nose and nasal sinuses rarely had localized disease and had a five-year failure-free and overall survival of 6% and 9%. The best treatment for patients who present with wide spread disease is uncertain. Obviously, involved field radiotherapy would have no chance to be curative in these patients. Typical anthrocycline containing chemotherapy regimens have yielded poor results with few patients have extended survival [29]. This may be in part because these tumors express P-glycoprotein and have the multi-drug resistance phenotype [74]. Recognition that nasal NK/T-cell lymphomas are sensitive to L-asparaginase [75, 76] led to incorporation of L-asparaginase into primary chemotherapy regimens for this disorder. A recent report of 38 patients with widespread, relapsed, or refractory extra nodal nasal NK/T-cell lymphoma utilizing the SMILE (dexamethasone, methotrexate, ifosphamide, L-asparaginase, and etoposide) regimen showed a 45% complete response rate and a one-year survival of 55% [77]. At the present time, the SMILE regimen probably represents the best choice for reasonably fit patients with wide spread disease, and could be considered as adjuvant therapy in patients with local regional disease.

Enteropathy associated T-cell lymphoma

Patients with enteropathy associated T-cell lymphoma still have a poor outlook and one of the poorest for any subtype of PTCL. In a large international study, the five-year overall survival was 20% and the five-year progression-free survival only 4% [9]. Even patients with a favorable IPI score had only a 29% five-year overall survival and 7% five-year failure-free survival. The best treatment for patients with enteropathy associated T-cell lymphoma remains uncertain. In a series reported more than a decade ago from the United Kingdom, many were felt to be unable to tolerate chemotherapy, and a majority of the remaining patients were unable to complete a course of chemotherapy. In that series, the five-year survival rate was 20% and the five-year failure-free survival rate was 3% [78]. A recent report from the United Kingdom suggests that a better treatment outcome is possible in patients able to tolerate an aggressive chemotherapy regimen followed by an autologous transplant [79]. Of 26 patients treated with a regimen including ifosphamide, vincristine, etoposide, methotrexate, and an autologous transplant, the five-year overall survival was 60% and the five-year progression-free survival 52%. At the present time, this intensive therapeutic approach should be considered in patients with enteropathy associated T-cell lymphoma who are fit enough to tolerate the regimen.

Hepatosplenic T-cell lymphoma

The treatment of patients with hepatosplenic T-cell lymphoma has been unsatisfactory. In the International Peripheral T-Cell/NK Cell Lymphoma Study Lymphoma Study, the five-year survival for patients with hepatosplenic PTCL was 7% and the five-year progression-free survival was 0% [9]. In a series of 21 patients from France treated mostly with CHOP-like regimens, 43% achieved a complete remission to induction therapy and 2 patients (10%) were alive and in remission at the time of the report—both having undergone adjuvant autologous transplantation in first remission [80]. The median survival in that series was 16 months. Occasional patients have been reported to have prolonged survival after treatment with alemtuzumab and cladribine [81], pentostatin [82], and allogeneic hematopoietic stem cell transplantation [83, 84]. The only patient that I have personally treated who achieved long-term disease-free survival received single agent pentostatin.


PTCLs have been the “poor step-child” of the diseases we refer to as lymphomas. The aggressive PTCLs were lumped in clinical trials with the aggressive B-cell lymphomas. Since this meant that 80–90% of patients included in the trials had diffuse large B-cell lymphoma, the treatments determined to be best were those that worked for diffuse large B-cell lymphoma. It is increasingly clear that the aggressive PTCLs are distinct diseases—quite different from the aggressive B-cell lymphomas. They have different biological features and, presumably, this underlies the dramatic difference in response to various therapies. The WHO classification has made it possible to distinguish the major subgroups of aggressive PTCL, and gene expression studies support most of these subgroups as unique entities with PTCL-NOS as the major exception. Therapeutic trials focusing on these specific subgroups are starting to identify more effective treatment options, but today only patients with ALK positive anaplastic large-cell lymphoma have a greater than 50% chance to be cured. This field is beginning to advance rapidly, and with good fortune an update of this review next year will reflect improved prospects for our patients.