Treatment. The conventional chemotherapy regimens used to treat aggressive NHL (e.g. CHOP) have produced disappointing results in PTCL-NOS when compared to its B-cell counterpart or ALK-pos ALCL. This poor outcome for PTCL seems to be a combination of problems at all stages of the disease with lower initial response rates and a higher proportion of resistance and early death as well as a greater tendency to relapse after CR, mainly within the first 1–2 years. CHOP remains the most commonly used first line treatment despite the fact that it has never been established as the preferred or most effective treatment for non-ALK-pos PTCLs. Currently, however, there are insufficient data to recommend an alternative and trials are badly needed to explore new regimens.
First line therapy: CHOP has been evaluated in first-line treatment of PTCL-NOS in a number of studies. Allowing for the caveats in interpretation mentioned above, it achieves a CR rate of around 50% and 5-year OS of 30% (Melnyk et al, 1997; Gisselbrecht et al, 1998; Lopez-Guillermo et al, 1998; Sonnen et al, 2005). Higher relapse risks than for B-cell lymphomas are noted in these studies, contributing to a high rate of treatment failure in the first 1–2 years (Coiffier et al, 1990; Gallamini et al, 2004). These results have led to investigation of intensification of therapy.
There are examples of phase II and III studies addressing intensification, either with alternative chemotherapy, autografting or both. There is a tendency for single arm prospective data to show promising results with intensive approaches [e.g. CEOP-B (as for CHOP but with epirubicin instead of doxorubicin) + bleomycin, 5-year OS 49%] (Sung et al, 2006) but this has not been confirmed in a randomized setting (Gressin et al, 2006). A large retrospective comparison of CHOP and more intensive therapy from the M.D. Anderson Cancer Centre found no difference in outcome between the two (Escalon et al, 2005), with 3-year OS 62% vs. 56% respectively, and 43% vs. 49% after exclusion of ALCL. The Groupe Ouest Est d’Etude des Leucémies et Autres Maladies du Sang (GOELAMS) compared VIP/ABVD (etoposide, ifosfamide, cisplatin/doxorubicin, vinblastine, bleomycin dacarbazine) versus CHOP and showed no difference in event-free survival (EFS) or OS (Gressin et al, 2006). The Nordic group demonstrated some improvement with MACOP-B randomized against CHOP (Jerkeman et al, 1999). Etoposide added to CHOP has shown mixed results (Karakas et al, 1996). Seven high grade NHL studies by the German study group showed that young good risk patients had improved 3-year EFS (71% vs. 50%) if etoposide was added to CHOP (14 or 21) (Schmitz et al, 2010). But many patients in the series had ALCL, and if the ALK-positive ALCL cases are excluded the difference is no longer significant. The Groupe d’ Etude des Lymphomes de l’ Adulte (GELA) studies in all high grade lymphomas found ACVBP (doxorubicin, cyclophosphamide, vindesine, bleomycin, prednisone) to be superior to CHOP in patients aged 60–70 years but failed to show any difference in younger patients for this or other alternative regimens (Delmer et al, 2003; Tilly et al, 2003). Of particular interest is the observation from the ITLP (Vose et al, 2008) that the inclusion of an anthracycline in a chemotherapy regimen made no difference to outcome. This may be due, in part, to the high P glycoprotein (PGP) expression in many of the PTCLs that is associated with resistance to anthracyclines.
Most published data regarding alternative or intensified chemotherapy has also consolidated the patients with an autograft, which makes interpretation of the effects of chemotherapy schedules alone difficult (Mercadal et al, 2008). CHOP therefore remains essentially unchallenged outside clinical trials, if autografting is not considered an option for the patient at first line.
In order to improve on the results with CHOP a number of recent studies have focussed on the addition of new agents to CHOP or other novel combination treatments (Table V) The current National Cancer Research Institute (NCRI) trial of CHOP chemotherapy with the addition of the MoAb alemtuzumab is open in some centres and is a recommendation for initial therapy. The Italian group have treated 18 evaluable patients who were given CHOP at a 4-weekly interval, together with alemtuzumab. Twelve of these patients were alive at 1 year, 11 in CR (Gallamini et al, 2007). An Asian study, also of CHOP and alemtuzumab, at 21 d intervals, was stopped early because of toxicity (Kim et al, 2007). The HOVON group have examined standard CHOP-14 with alemtuzumab and found a 90% ORR and median OS of 27 months (Kluin- Nelemans et al, 2008). There has also been a US National Cancer Institute study of DaEPOCH (dose-adjusted etoposide, vincristine, doxorubicin, cyclophosphamide, prednisone) + alemtuzumab showing a PFS of 45% and OS of 48% at 3 years, with a plateau emerging on the curves (Janik et al, 2005). The German study group have examined the combination of alemtuzumab with FCD chemotherapy (fludarabine, cyclophosphamide and doxorubicin), which gave a 58% CR rate in a small number of patients studied but with significant additional toxicity (Weidmann et al, 2010). These trials suggest that there may be an advantage in adding alemtuzumab to standard chemotherapy, albeit with increased toxicity, but this needs to be tested in prospective randomized trials and currently is not a strategy advised outside the trial setting. A current European study (ACT I/II) randomizes patients to 14-d CHOP with or without alemtuzumab. Patients under 60 years of age are autografted in first remission. A question remains regarding the CD52 expression in PTCL with some published data reporting around half of cases as CD52 negative (Rodig et al, 2006; Chang et al, 2007; Piccaluga et al, 2007b), whilst others suggest that the majority of PTCL-NOS are in fact positive (Jiang et al, 2009; Reimer et al, 2009). The discrepancy may be due to methodology because CD52 staining in paraffin embedded tissue is unreliable. In the future, CD52 staining on fresh tissue should be part of any prospective trial which includes alemtuzumab therapy.
The Eastern Cooperative Oncology Group (ECOG) are also looking at adding novel therapy to CHOP and currently have a Phase III trial comparing CHOP with or without bevazucimab.
Gemcitabine combinations are also being explored in the first-line setting e.g. CHOP, etoposide and gemcitabine (Kim JG et al, 2006), and the South Western Oncology Group (SWOG) are conducting a Phase II trial of gemcitabine, cisplatin, etoposide and methylprednisolone (PEGS).
A number of more novel agents have been investigated in PTCL but most data, as expected, is in relapsed/refractory disease (Foss, 2010).
Consolidation in first CR with auto-HSCT (Table VI): Several groups have examined the role of dose-escalated chemotherapy with auto-HSCT support as consolidation therapy for PTCL (Mounier et al, 2004; Corradini et al, 2006; Feyler et al, 2007; Rodriguez et al, 2007a) (Table VI). Mounier et al (2004) reported a series of carefully case matched patients drawn from the GELA LNH 87 and 93 trials comparing HDT with combination chemotherapy (ACBVP; doxorubicin, cyclophosphamide, bleomycin, vincristine, and prednisone) alone. They noted that there was no difference in DFS or OS in the 29 patients with non-anaplastic PTCL. Long-term follow-up of an Italian study of high dose sequential chemotherapy in PTCL reported a 12-year OS of only 21% in the non-ALK+ cohort compared to 62% in the ALK+ patients (Corradini et al, 2006). The intention-to-treat analysis in this prospective study showed that only 74% of patients underwent auto-HSCT because of a high incidence of disease progression during first-line treatment. In a multivariate analysis, achievement of CR at the time of transplant predicted for superior outcome, which has been corroborated in other studies (Corradini et al, 2006; Feyler et al, 2007).
Table VI. Prospective studies on first-line high-dose therapy and autotransplantation (auto-HSCT) in PTCL.
|Corradini et al (2006)||62||Mito/Mel or BEAM||66% CR,18% PR||73||30% (12-year EFS)|
55% (12-year DFS)
34% (12-year OS)
|2 phase II studies incl. ALK+ ALCL|
|d’Amore et al (2006)||121||CHOEP-16 × 4 + BEAM||71% CR/PR||73||63% (3-year OS)||No ALK+ ALCL|
|Rodriguez et al (2007b)||26||Mega CHOP ± BEAM|
IFE (ifosfamide and etoposide)
|65% CR, 16% PR||73||53% (3-year PFS)|
73% (3-year OS)
|No ALK+ ALCL|
|Mercadal et al (2008)||41||High CHOP/ESHAP||51% CR, 7% PR||41||30% (4-year PFS)|
39% (4-year OS)
|No ALK+ ALCL|
|Reimer et al (2009)||56/83||Cy/TBI||58% CR, 8% PR||66||36% (3-year PFS)|
48% (3-year OS)
|No ALK+ ALCL|
A study of 74 patients with PTCL transplanted in first remission mainly using high dose chemotherapy conditioning reported a 5-year OS and PFS of 68% and 63%, respectively (Rodriguez et al, 2007a). All patients entered into the study were however in remission at the time of transplant and the study included 23 cases of ALCL whose ALK status was not reported, which may have significantly biassed the outcome. On multivariate analysis the prognostic index for T-cell lymphoma (Gallamini et al, 2004) identified a poor risk subgroup with an OS of 21% at 5 years. A second study from the same group analysed outcome in poor risk cases, defined by exclusion of ALK+ disease and advanced stage (Rodriguez et al, 2007b). These patients received intensive induction with MegaCHOP prior to high dose therapy with BEAM (carmustine, etoposide, cytarabine, melphalan) conditioning in responders and salvage with ifosfamide and etoposide followed by BEAM in CHOP non-responders. Of 26 patients entered into this study 19 responded either to induction or salvage treatment and, after high dose therapy, 17 achieved CR. Thus, on an intention to treat basis, intensive induction followed by high dose therapy with autologous stem cell support resulted in a CR rate of 65% in poor prognosis PTCL. The 3-year OS and PFS was estimated at 73% and 53%, respectively. Reimer et al (2009) recently published results of a prospective multicentre centre trial of upfront HSCT in PTCL (PTCL-NOS n = 32; ALK− ALCL n = 13; AITL n = 27). Of 83 patients enrolled onto the study, only 55 patients (66%) proceeded to HSCT. Progressive disease was the predominant reason for not undergoing HSCT, as previously reported (Mercadal et al, 2008). The estimated 3-year OS and PFS were 48% and 36%, respectively (Reimer et al, 2009). The estimated 3-year OS was 71% for patients who underwent auto-HSCT compared to 11% for patients who did not. A Nordic group presented similar results in abstract form (d’Amore et al, 2006.)
Treatment of relapsed or refractory disease: Patients responding to further therapy and of acceptable fitness are usually considered for HSCT and whether patients should undergo allo-HSCT or auto-HSCT is contentious. Ideally this should be in the context of a trial, particularly if the stem cell source is allogeneic as this is experimental but, given the prognosis of relapsed PTCL, most clinicians would consider such approaches for any suitable patient as some evidence of efficacy does exist.
Salvage chemotherapy for relapsed or refractory disease: Re-induction or treatment of refractory disease is usually with combination chemotherapy. There are also a number of experimental agents that have shown promise and patients should be considered for inclusion in suitable clinical trials where available. There are no data on which to base the choice of re-induction and the conventional approach is to use a platinum-based schedule, particularly when intending to consolidate with a transplant.
There are emerging data of interest for other agents (Table V). Gemcitabine as a single agent in cutaneous and non-cutaneous T-cell lymphoma seems highly active in phase II studies (Zinzani et al, 2000; Sallah et al, 2001; Marchi et al, 2005). Studies of gemcitabine in combination with steroids and cisplatin (GEM-P) have yielded encouraging results in refractory patients (Emmanouilides et al, 2004; Arkenau et al, 2007; Spencer et al, 2007). Pentostatin has also been used in PTCL, but seems to be most effective in leukaemic and cutaneous subtypes (Mercieca et al, 1994; Tsimberidou et al, 2004).
The MoAb alemtuzumab is of interest. Combinations of this antibody with CHOP are being investigated in forthcoming trials in PTCL NOS as described above and this has followed on from experience in other diseases (notably CLL and T-PLL) and promising early clinical data in PTCL (Lundin et al, 1998; Dearden, 2006). A 36% ORR was seen with single agent alemtuzumab in a heavily pre-treated cohort of patients with PTCL (Enblad et al, 2004).
Early data on a number of other molecules exists including histone deacetylase inhibitors (depsipeptide) (Piekarz et al, 2004), antibodies to CD25 and the IL2–toxin conjugate dinileukin-difitox (Dang et al, 2007; Foss et al, 2007; Waldmann, 2007), a novel anti-folate, praletrexate (O’Connor et al, 2007, 2008, 2009), lenalidomide, mammalian target of rapamycin (mTOR) inhibitors, anti- CD4 (zanolimumab) (d’Amore et al, 2010) and bortezomib (Zinzani et al, 2007). Most promising of these are the histone deacetylase inhibitors (Piekarz et al, 2011), bortezomib and pralatrexate, with a response rate of 39% in heavily pre-treated refractory patients (O’Connor et al, 2007, 2008, 2009). Praletrexate has recently been approved in the US for treatment of relapsed PTCL and will be reviewed for licencing in the European Union this year. In vitro praletrexate has been shown to have synergy with gemcitabine and bortezomib and clinical trials evaluating combination therapy are ongoing.The place of these newer agents in therapy is not yet established although, given the poor response in PTCL-NOS to conventional chemotherapeutic agents, they are likely to be critical for progress in the future.
Auto-HSCT for relapsed/refractory PTCL: A number of groups have reported their experience with high dose therapy and auto-HSCT as salvage for relapsed PTCL (Blystad et al, 2001; Song et al, 2003; Kewalramani et al, 2006; Smith et al, 2007). In the main these are retrospective uncontrolled studies and many include cases of ALK+ ALCL which, as previously noted, have a better prognosis than other histological categories. Overall the efficacy of this approach, in patients with disease that was not ALK+, was disappointing, with 5-year OS of <35% in most studies (Song et al, 2003; Jantunen et al, 2004; Zamkoff et al, 2004; Kewalramani et al, 2006; Smith et al, 2007). Zamkoff et al (2004) specifically reported 15 ALK-negative ALCL cases that were followed up after being autografted for relapse. Thirteen of these relapsed once more and the median survival was only 72 weeks.
Allo-HSCT for relapsed/refractory PTCL: Most retrospective studies of allo-HSCT in lymphoma have not analysed results for patients with T-cell disease separately; however, a small number of reports have been published. The TRM for standard intensity conditioning regimens in patients with PTCL has been very high (30–50%), presumably because of more advanced age and the effects of prior therapy (Dhedin et al, 1999; Le Gouill et al, 2008). This unacceptably high toxicity stimulated the development of RIC regimens, which seek to maintain the graft-versus-tumour effect whilst minimizing regimen-related problems. So far there are few studies of RIC -allo HSCT in T-cell lymphoma. A pilot study of 17 patients which included nine PTCL-NOS, four AITL and four ALK- ALCL reported a non-relapse mortality at 2 years of only 6% following a conditioning regimen that incorporated thiotepa, fludarabine and cyclophosphamide (Corradini et al, 2004). Severe acute or chronic extensive graft-versus-host disease (GvHD) was reported in three patients overall and with a median follow up of 28 months, 12 were in CR. Two patients achieved disease control after DLI, providing further evidence for a graft-versus-T-NHL effect. Short follow up of 10 patients who received fludarabine and cyclophosphamide conditioning after an alemtuzumab containing induction regimen showed a similarly encouraging outcome with six continuing remissions but chronic extensive GvHD in 5 (Wulf et al, 2005). Further prospective trials addressing the role of RIC-allo-HSCT in PTCL NOS are warranted.
CNS prophylaxis: This remains contentious in all the aggressive lymphomas. There is no consensus as to the optimal strategy or indeed which lymphomas should receive prophylaxis. The data on PTCL does not allow specific recommendations distinct from B-NHL. Guidelines on prophylaxis are being drawn up by the BCSH and have been the subject of recent reviews (McMillan, 2005; Hill & Owen, 2006). There is a 5% incidence of CNS relapse in most large studies of aggressive NHL and the factors of importance include: IPI score, LDH, involvement of extranodal sites and specific sites, such as bone marrow, testis and sinuses. It seems logical to apply the same approach to prophylaxis in PTCL as for the more common diffuse large B-cell lymphoma (DLBCL). The nature of PTCL is that it will tend in more cases to have the high risk features listed above and so a larger proportion of patients may receive CNS prophylaxis for that reason. T-cell phenotype alone is not an indication to use prophylaxis.