Current management of follicular lymphomas

Authors


Wolfgang Hiddemann, MD, PhD, Department of Internal Medicine III, University of Munich, Marchioninistr. 15, 81377 Munich, Germany. E-Mail wolfgang.hiddemann@med.uni-muenchen.de

Summary

After decades of stagnation, the prognosis of patients with follicular lymphomas (FL) has changed substantially within the last few years due to new and effective therapeutic modalities. These include myeloablative therapy followed by autologous stem cell transplantation (ASCT) in younger patients in first remission, which showed a significant prolongation of remission duration in three prospective randomised trials while the impact on overall survival still needs to be determined. Adding the anti-CD 20 antibody Rituximab to conventional chemotherapy resulted in a significant increase in remission rate, remission duration and, in two of four currently available prospective randomised studies, even in a longer overall survival. A prolongation of remission duration was also seen when Rituximab was given as maintenance after cytoreductive therapy including Rituximab. Radio-immunotherapy (RIT) with radioisotopes coupled to monoclonal antibodies produced encouraging data in several phase II studies. New therapeutic perspectives have also emerged from increasing insights into the biology of the disease that unravel molecular targets for novel agents, some of which have entered clinical evaluation already.

Incidence and characteristics

Follicular lymphoma (FL) is the second most frequent lymphoma subtype worldwide (The Non-Hodgkin's Lymphoma Classification Project, 1997; Anderson et al, 1998). Its incidence is rapidly increasing in Western countries and has nearly doubled within the last three decades (Groves et al, 2000; Chiu & Weisenburger, 2003). The reasons for this development are not completely clear but increasing evidence suggests that environmental factors most probably contribute (Grulich & Vajdic, 2005; Morton et al, 2005; Besson et al, 2006; Chiu et al, 2006; de Sanjose et al, 2006).

Frequently, the terms low-grade or indolent lymphoma, as introduced by the Kiel classification (Lennert et al, 1975) and the Working Formulation (The Non-Hodgkin's Lymphoma Pathologic Classification Project, 1982) respectively, are still used as synonyms for FL. These terms however, lumped several different entities together based on their cytological appearance or their indolent clinical course. Since the introduction and worldwide acceptance of the World Health Organaization (WHO) classification (Harris et al, 1999) both names should be avoided and the precise nomenclature of FL should be used only.

Follicular lymphoma is pathologically well characterised. Unlike other malignant lymphomas three different malignancy grades are defined according to the degree of centroblasts (Harris et al, 1999). Grade 3 FL is further subdivided into grade 3a and 3b (Ott et al, 2002). While it is broadly accepted that grade 3b FL behaves clinically more like a diffuse large B-cell lymphoma and should be treated accordingly, it remains questionable whether the discrimination of grades 1 to 3a is of clinical significance.

Prognosis and general treatment approaches

Prognosis and therapy are closely related to the extent of the disease at initial diagnosis, which is defined according to the Ann Arbor system. At most, 10–15% of FL are detected at the early stages I and II. In these patients radiotherapy is the treatment of choice and is applied as extended or involved field irradiation. This approach results in long-term disease-free survival (DFS) and probable cure in approximately 45–80% of cases (Table I). In the majority of patients the disease is diagnosed at an advanced stage III or IV and cannot be cured by conventional therapeutic approaches. Hence, anti-lymphoma therapy is usually withheld for a watch and wait period until the disease becomes symptomatic (Ardeshna et al, 2003). In this situation a broad spectrum of therapeutic options is available ranging from single agent to combination chemotherapy with or without additional irradiation. In spite of numerous efforts and the exploration of different regimens, chemotherapy had no major impact on survival and the prognosis of FL has literally remained unchanged over the last decades with a median survival time of 8–10 years (Horning & Rosenberg, 1984; Gallagher et al, 1986). Most recently, however, a significant improvement of survival was reported by several groups which was mainly due to the introduction of novel therapeutic strategies (Fisher et al, 2005; Liu et al, 2006).

Table I.   Results of radiotherapy in follicular lymphomas (FL) at stages I to III.
ReferenceNo. patientsTime of analysis (years)Relapse-free stage I/II (%)Survival stage III (%)
Pendlebury et al (1995)581043 
Mac Manus and Hoppe (1996)1771544 
Stuschke et al (1997)117583/4456
Murtha et al (2001)6610 88
Wilder et al (2001)801566/26 
Ott et al (2003)581064 
Guadagnolo et al (2006)1061046 

New treatment modalities that improve the long-term outcome of patients with FL include, in particular, myeloablative therapy followed by autologous stem cell transplantation (ASCT) and, to a lesser degree, allogeneic stem cell transplantation (alloSCT) and monoclonal antibodies with an inherent antilymphoma activity, as carriers for radioisotopes or as immunotoxins. Future hope is based on increasing insights into the biology of the disease that are beginning to unravel molecular targets for novel agents that are currently entering clinical evaluation.

Myeloablative therapy with ASCT

Like most new treatment modalities ASCT was first applied for salvage therapy of relapsed or refractory FL. Encouraging data emerged from several phase II studies, which indicated that this approach may prolong DFS in patients with recurrent advanced stage FL (Bierman et al, 1997; Freedman et al, 1999; Apostolidis et al, 2000). However, the definite proof of a superiority of ASCT over conventional therapy in relapsed advanced stage FL still remains open as no valuable prospective randomised study is currently available. The only published study, by Schouten et al (2003), suffered from a slow recruitment and a small patient number as well as from a complex study design asking three questions simultaneously.

Autologous stem cell transplantation was also evaluated in first remission and revealed promising results. Currently, three prospective randomised cooperative group trials have been completed (Sebban et al, 2003; Lenz et al, 2004a; Deconinck et al, 2005) (Table II). All three studies showed a significant prolongation of response duration but inconsistent results on overall survival. In the study by the GLSG (German Low Grade Lymphoma Study Group) the observation time is still too short and the results on survival are therefore still blinded (Lenz et al, 2004a). The French GOELAMS (Groupe Ouest–Est des Leucémies et des Autres Maladies du Sang) study group reported a significant prolongation of response duration, which was predominantly seen in patients with a high-risk profile according to the FLIPI (>2). However, a high incidence of secondary fatal cancers was encountered in the ASCT arm and compromised the impact on overall survival (Deconinck et al, 2005). In contrast, the third randomised trial of the GELA (Groupe D'Etudes des Lymphomes De l'Adulte), which has as yet only been published in abstract form, resulted in a significant prolongation of response duration and also of overall survival (Sebban et al, 2003).

Table II.   Results of myeloablative therapy with subsequent autologous stem cell transplantation (ASCT) in first remission of follicular lymphoma.
Trial (Reference)InductionConsolidation (patients)Event-free survivalOverall survival
  1. CHOP, cyclophosphamide, doxorubicin, vincristine, prednisone; MCP, mitoxantrone, chlorambucil, prednisolone; VCAP, vindesine, cyclophosphamide, doxorubicin, prednisone; CHVP, cyclophosphamide, doxorubicin, teniposide, prednisone; IFN-α, interferon-α; ASCT, autologous stem cell transplantation; TBI, total body irradiation; Cyclo, cyclophosphamide; PFS, progression-free survival; EFS, event-free survivial; OS, overall survival.

  2. *5-year PFS; †5-year EFS; ‡7-year EFS.

Phase III Lenz et al (2004a) (GLSG)CHOP/MCP (4–6 cycles)TBI/Cyclo followed by ASCT (n = 153)64·7% vs. 33·3%* (P < 0·0001)Not yet available
CHOP/MCP (4–6 cycles)IFN-maintenance (n = 154)
Phase III Deconinck et al (2005) (GOELAMS)VCAP cycles)TBI/Cyclo followed by ASCT (n = 86)60% vs. 48%† (P < 0·050)Median not reached in both treatment arms
No significant difference
CHVP/IFN-α cycles)CHVP plus IFN-α (n = 80)
Phase III Sebban et al (2003) (GELA)CHOP (4 cycles)TBI/Cyclo followed by ASCT (n = 192)45% vs. 36%‡ (P = 0·5)86% vs. 74% (P = 0·05) (7-year OS)
CHVP/IFN-αcycles)CHVP plus IFN-α (n = 209)

A major reason for the inconsistency of survival data between the three studies may be partly due to different patient risk groups. In addition the risk of inducing secondary malignancies, particularly secondary acute myeloid leukaemias (AML) or myelodysplastic syndromes (MDS), was substantially different. This risk was 0% in the GELA study, 8·5% in the GOELAMS study and 3·8% in the GLSG trial. A subgroup analysis of the GLSG study suggests that the risk of secondary MDS or AML may be associated with the type of initial chemotherapy rather than with the conditioning procedure (Lenz et al, 2004b).

Hence, myeloablative radiochemotherapy followed by ASCT is an effective treatment option for young patients (<65 years) with advanced stage FL as it significantly improves response duration and potentially also overall survival. Especially after a cyclophosphamide, doxorubicin, vincristine, prednisone (CHOP)-like induction, the increase of secondary haematological neoplasias is moderate and acceptable and should not preclude the offer of such treatment to young patients with a high or intermediate risk profile. As new approaches, such as the combination of chemotherapy with monoclonal antibodies, particularly Rituximab, are implemented in multimodal approaches, the role of ASCT may have to be redefined.

Myeloablative therapy with alloSCT

In patients with advanced FL allogeneic transplantation represents the only treatment modality with proven curative potential. Allogeneic transplantation can be performed after either conventional myeloablative or reduced-intensity conditioning. The latter approach may reduce treatment-related morbidity and mortality (TRM), consisting mainly of acute and chronic graft-versus-host disease (GVHD) and infectious complications.

AlloSCT after conventional conditioning

The largest report on conventional allogeneic transplantation (n = 176) with recurrent FL was published by the International Bone Marrow Transplant Registry (IBMTR), which retrospectively analysed the treatment outcome after mostly total body irradiation (TBI)-based conditioning and human leucocyte antigen (HLA)-identical sibling donor transplantation. The low recurrence rate of 19% was counterbalanced by a high non-relapse TRM of 30% from infections and GvHD resulting in an overall failure rate of 49% (van Besien et al, 2003). These data are in accordance with previous results from several smaller series which all demonstrate a low recurrence rate but a high TRM (Forrest et al, 2002; Peniket et al, 2003; Totze et al, 2004).

AlloSCT after reduced-intensity conditioning

The largest series of alloSCT after reduced-intensity conditioning (RIC) was reported by the Working Party of the European Group for Blood and Bone Marrow Transplantation (EBMTR) (Robinson et al, 2002). From a total of 188 patients, 52 cases had FL. When compared with other lymphoma entities FL had the lowest rate of progression and the highest proportion of progression-free (PFS) and overall survival, which exceeded 60% at 5 years. These data confirm prior results from smaller series of patients but also show that beneficial long-term effects are mainly seen in patients with chemosensitive disease (Khouri et al, 2001; Tanimoto et al, 2003).

These data suggest that alloSCT after RIC may be superior to conventional alloSCT. However, long-term follow-up is still relatively short and comparable groups of patients small. Therefore, both approaches must still be considered experimental and further data from prospective randomised comparisons are warranted.

Autologous versus allogeneic transplantation

In the largest retrospective survey of the IBMTR and the Autologous Blood and Marrow Transplant Registry (ABMTR) data from 904 patients were collected between 1990 and 1999 (van Besien et al, 2003); 19% had received alloSCT, 14% purged autologous blood cell transplantation (ASCT) and 67% ABCT. The three groups were not homogeneous in terms of risk factors and prior therapy. The 5-year TRM was higher in the alloSCT group with 30% compared with 14%, and 8% in the purged ASCT and ASCT cohorts, respectively. In contrast, the 5-year recurrence rates were considerably lower in the alloSCT group with 21% compared with 43% and 58% in the purged ASCT and ASCT groups respectively. This resulted in similar estimated 5-year survival rates of 51%, 62% and 55%. However, a continuous pattern of relapses was observed after ASCT whereas after alloSCT only a few relapses occurred after the first year after transplantation.

The observation that the high TRM associated with conventional alloSCT compensates the low recurrence rate was confirmed in another retrospective analysis of the EBMT registry (Peniket et al, 2003).

Although alloSCT is still the only treatment option with curative potential in advanced stage FL, no clear survival benefit over ASCT has been demonstrated yet. Whether alloSCT after RIC will improve on these results also still needs to be demonstrated by prospective randomised studies.

From a practical viewpoint it must be stated, however, that individual patients, particularly younger patients in second relapse with matched sibling donors should be considered for alloSCT.

Anti-lymphoma antibodies with inherent anti-lymphoma activity

The only currently available anti-lymphoma antibody with a proven activity in FL is the monoclonal antibody (mAb) Rituximab. This mAb is directed against the CD 20 antigen that is expressed on most B-cell malignancies. The mechanisms of action are several and include the lysis of CD20+ cells by complement activation and antibody-dependent cell-mediated cytotoxicity, an induction of apoptosis, a block of the G1S-transition, an impairment of differentiation and an increased phosphorylation of cellular proteins (Reff et al, 1994; Shan et al, 2000). The efficacy of antibody-dependent cell-mediated cytotoxicity is strongly dependent on the activation of effector cells via the Fc gamma receptor. Accordingly, two polymorphisms of this receptor were shown to predict response rate and freedom from progression after Rituximab monotherapy (Weng & Levy, 2003).

In several phase II clinical trials Rituximab demonstrated a significant single agent activity in pretreated as well as in previously untreated patients with FL (Maloney et al, 1997; McLaughlin et al, 1998; Feuring-Buske et al, 2000; Colombat et al, 2001; Hainsworth et al, 2002). These results and in vitro data suggesting a potential synergistic effect of Rituximab with several chemotherapeutic agents lead to the clinical evaluation of Rituximab–chemotherapy combinations. In a first phase II trial the combination of Rituximab with CHOP induced responses in all evaluable patients with a complete remission (CR) rate of 63% and a median PFS of 82 months (Czuczman et al, 2004). The first prospective randomised comparison of the combination of Rituximab plus chemotherapy versus chemotherapy alone was carried out by the GLSG in patients with relapsed FL and mantle cell lymphomas (MCL). In this trial Rituximab was added to the FCM (Fludarabine, Cyclophosphamde, Mitoxantrone) combination (R-FCM), which was randomly compared with FCM alone. R-FCM showed a significantly higher remission rate, a significantly longer PFS and, in particular, a significantly longer overall survival for both lymphoma subtypes (Forstpointner et al, 2004).

For first-line therapy, the results of four prospective randomised phase III studies investigating Rituximab plus chemotherapy versus chemotherapy alone have become available (Herold et al, 2004; Salles et al, 2004; Hiddemann et al, 2005a; Marcus et al, 2005) (Table III). All four studies consistently showed a significant increase in initial response rates, a significant prolongation of response duration and, in two studies, a significantly longer overall survival.

Table III.   Rituximab plus chemotherapy in first line therapy of advanced stage follicular lymphoma.
ReferenceRegimen (No. patients)P-value
  1. CHOP, cyclophosphamide, doxorubicin, vincristine, prednisone; CVP, lomustine, vincristine, cisplatin; MCP, mitoxantrone, chlorambucil, prednisolone; CHVP; cyclophosphamide, doxorubicin, teniposide, prednisone; IFN-α, interferon-α; R, rituximab; CR, complete remission; CRu, complete remission unconfirmed.

Hiddemann et al (2005a)CHOP (205)R-CHOP (223) 
Response rate90%96%P = 0·011
Median time to treatment failure31 monthsNot reachedP < 0·0001
Marcus et al (2005)CVP (159)R-CVP (162) 
Response rate57%81%P < 0·0001
Median time to treatment failure7 months27 monthsP < 0·0001
Herold et al (2004)MCP (96)R-MCP (105) 
Response rate75%92%P < 0·001
Median event-free survival19 monthsNot reachedP < 0·0001
Salles et al (2004)CHVP/IFN-αR-CHVP/IFN-α (184) 
Response rate (CR/CRu)85% (49%)94% (76%)P < 0·0001
Median event-free survivalNot reachedNot reached 

A different strategy was applied in a study by the Eastern Cooperative Oncology Group (ECOG), in which remission induction by CVP (lomustine, vincristine, cisplatin) was followed by a randomised comparison of Rituximab maintenance over 2 years versus observation only. Rituximab maintenance achieved a 2·7 year longer response duration and a significant improvement in survival (Hochster et al, 2004, 2005.

Two recently completed studies by the GLSG and the European Organization for the Research and Treatment of Cancer/Hemato-Oncologie voor Volwassenen in Nederland (EORTC/HOVON) intergroup demonstrated that Rituximab maintenance is also effective in patients with relapsed FL who responded to Rituximab–chemotherapy combinations for salvage therapy (Hiddemann et al, 2005b; Van Oers et al, 2005). In both studies response duration was significantly prolonged with a tendency to a longer overall survival as well. In the two trials, however, Rituximab maintenance was applied on different schedules. While in the GLSG study Rituximab was administered at four weekly doses to be given at 3 and 9 months after entering remission, the EORTC/HOVON intergroup applied single infusions of Rituximab every 3 months over a period of 2 years. While Hainsworth et al (2005) and Hochster et al (2005) also used the four doses every 6 months schedule in their studies, Ghielmini et al (2004) repeated single infusions of Rituximab every 2 months for four times. To date it remains unanswered which of these regimens might be most effective and if they differ in their anti-lymphoma activity. In addition, it is currently unclear how long Rituximab should be given in remission and if the standard dose of 375 mg/m2/d is the most appropriate one. Based on pharmacokinetic analyses and assuming a serum target level of Rituximab of 25 μg/ml for a maintained anti-lymphoma activity, single applications of Rituxmab at a dose of 375 mg/m2/d, to be repeated every 2–3 months, appear most appropriate (Berinstein et al, 1998; Gordan et al, 2005).

In spite of these considerations, all currently available data demonstrate a significant activity of Rituximab in all phases of therapy for advanced stage FL. It is therefore no longer the question whether Rituximab should be applied for first and second line therapy of advanced stage FLs but rather when and how. Although further studies are needed to address this question in greater detail it may be speculated that the different ways of application might not be used competitively but rather complementary and might be appropriate for different patient populations as defined by age, performance status, International Prognostic Index (IPI) or the recently introduced IPI for FL (FLIPI) (Solal-Celigny et al, 2004) which also proved relevant under the conditions of Rituximab plus chemotherapy (Buske et al, 2006).

Anti-lymphoma antibodies as carriers for radio-isotopes (Radio-immunotherapy)

Because of the inherent high radiosensitivity of FL and the expression of potential target antigens on the cell surface, radio-immunotherapy (RIT) represents a promising concept in this disease. At present, the majority of radio-immunoconjugates target the CD 20 antigen, which is neither shed nor internalised thus representing an appealing target for RIT. Radiolysis is induced in both the targeted cells and adjacent lymphoma cells due to the radiation crossfire effect.

Currently, two different lymphocyte specific radio-immunoconjugates are approved for the clinical use in the European Union and the United States, the Yttrium-90 (90Y) labeled Ibritumomab tiuxetan (Zevalin®) and the Iodine-131 (131I) labeled Tositumomab (Bexxar®) (Table IV). Both compounds combine the tumour targeting attributes of a murine monoclonal anti-CD 20 antibody with therapeutic radioisotopes to be delivered to sites of disseminated disease. 90Y-Ibritumomab tiuxetan is a pure β-emitter of high energy with a short half-life of 64 h and therefore suitable for outpatient treatment. 131I-Tositumomab delivers both β and γ. Both constructs may be applied in a non-myeloablative as well as myeloablative activity.

Table IV.   Characteristics of the radioisotopes 90Yttrium and 131Iodine.
 90Yttrium131 Iodine
Source of radiationbetabeta and gamma
Energy (MeV)2·30·36/0·61
Depth of activity (mm)5–101–2
Half time (h)64193
ApplicationOutpatientInpatient (radioprotection)

In the non-myeloablative approach both conjugates demonstrated comparable efficacy and toxicity. In various clinical trials, RIT applied in relapsed or refractory FL yielded response rates of about 60–80% and complete response (CR)/unconfirmed CR (CRu)-rates in the range of 20–30% (Table V). In a randomised trial a single infusion of 90Y-Ibritumomab tiuxetan was superior to a standard 4 weekly infusion of Rituximab with regard to overall response (80% vs. 56%) and CR-rate (30% vs. 16%). Although the median PFS was not different in the total cohort, a long-term benefit was observed for patients achieving a CR after 90Y-Ibritumomab (Witzig et al, 2002a). As with many therapeutic approaches, higher response rates and longer duration of response were observed when RIT was used earlier in the treatment schedule (Emmanouilides et al, 2006). Kaminski et al (2005) reported on the experience with 131I-tositumomab in 76 previously untreated FL patients. Despite the favourable prognostic profile of this patient population, the high response rate [overall response rate (ORR) 95%, CR rate 75%] and the nature of the durable remissions, with 59% of the patients remaining in ongoing complete remission and a minimum follow-up of more than 4 years from treatment, are remarkable and support further testing of this approach in prospective randomised trials. The same holds true for 90Y Ibritumomab tiuxetan applied as first line treatment for FL with preliminary data suggesting a very high response rate with an ORR of 100% and a CR rate of 62% (Sweetenham et al, 2004).

Table V.   Efficacy of single-regimen RIT for relapsed/refractory FL.
ReferenceAgentPhasenPatient selectionResponse ratesDuration of response
  1. RIT, radio-immunotherapy; NHL, non-Hodgkin lymphoma; MCL, mantle cell lymphoma; FL, follicular lymphoma; ORR, overall response rate; CR, complete response; CRu, unconfirmed CR; DR, duration of response, TTP, time to progression; PFS, progression-free survival.

Kaminski et al (2000)Bexxar®I/II42Relapsed or refractory CD20+ B-cell low- and intermediate-grade NHLORR 71%
CR32%
Median PFS 12 months (for CR 20 months)
Vose et al (2000)Bexxar®II47Relapsed or refractory CD20+ B-cell low-grade and transformed NHLORR 57%
CR 32%
Median DR 9·9 months (for CR 19·9 months)
Kaminski et al (2001)Bexxar®II60Relapsed or refractory CD20+ B-cell low-grade and transformed NHLORR 65%
CR 20%
Median DR 6·5 months
Witzig et al (1999)Zevalin®I/II51Relapsed or refractory CD20+ B-cell low- and intermediate-grade NHL or MCLORR 67%
CR 26%
Estimated DR 11·7 + months
Witzig et al (2002a))Zevalin®III143Relapsed or refractory low-grade FL or transformed NHLORR 80%
CR 30%
CRu 4%
Estimated DR 14·2 months
Wiseman et al (2002)Zevalin®II30Relapsed or refractory low-grade FL or transformed NHL and mild thrombocytopeniaORR 83%
CR 37%
CRu 6%
Estimated TTP 9·4 months (for responders 12·6 months)
Witzig et al (2002b))Zevalin®II54Rituximab-refractory FLORR 74%
CR 15%
Estimated TTP 6·8 months (for responders 8·7 months)

Radio-immunotherapy has been successfully used as consolidation therapy. The Southwest Oncology Group (SWOG) conducted a phase II trial with 90 previously untreated, advanced stage FL patients treated with six cycles of CHOP followed by 131I-tositumomab. The ORR rate for the entire regimen was 90%, including 67% CR/CRu. Of note, 57% of the evaluable patients who achieved less than a CR with CHOP improved their remission status after RIT. After a median follow-up of 2·3 years, the estimated 2-year PFS was 81% with a 2-year overall survival of 97% (Press et al, 2003). Similarly, promising results have been reported from another phase II trial using an abbreviated fludarabine regimen followed by 131I-tositumomab as front-line therapy of FL patients (Leonard et al, 2005). Consolidation therapy with 90Y-Ibritumomab tiuxetan has been tested in a phase II multicenter trial of the Minnie Pearl Cancer Research Network in 42 previously untreated patients with stage II - IV FL after a standard course of rituximab followed by three cycles of R-CHOP (CHOP + Rituximab). An ORR of 100% with a CR rate of 28% was observed after R-CHOP; importantly, after 90Y-ibritumomab tiuxetan the CR rate rose to 67% and 2-year PFS was 77% at a median follow up of 20 months (Shipley et al, 2005). Based on the experience that long-term benefit is primarily observed in lymphoma patients with high-quality responses, consolidation strategies are currently widely tested in ongoing clinical trials.

Principally, if murine antibodies are administered, a human anti-mouse antibody (HAMA) may be seen, but is diminished in patients with prior chemotherapy. HAMA response was 17% after 131I-tositumomab treatment in patients with relapsed or refractory B-cell lymphoma, however, in chemotherapy-naïve patients, HAMA responses of up to 63% have been described. By interfering with biodistribution and tumour targeting, presence of high titres of HAMAs seem to influence outcome of RIT and also precludes re-administration of the foreign protein. (Kaminski et al, 2005).

Of note, both radio-immunoconjugates induce considerable myelosuppression with delayed nadirs occurring 6–10 weeks post-treatment. Thus, the pretreatment bone marrow infiltration by lymphoma cells should not exceed 25% and haematopoeitic recovery should be intact to avoid severe and prolonged myelosuppression.

Myeloablative radio-immunotherapy with subsequent re-infusion of autologous peripheral stem cells has been successfully applied in patients with refractory or relapsed B-cell lymphoma (Press et al, 1993). A single myeloablative dose of 131I–Tositumomab followed by ASCT was tested in 27 patients with FL and achieved an ORR of 93% with a CR/CRu rate of 85% and an estimated 5-year PFS and OS of 48% and 67%, respectively, in comparison with 29% and 53% after conventional high-dose chemotherapy (Gopal et al, 2003). Importantly, the risk for secondary haematological neoplasias did not differ between the two treatment strategies and was estimated to be 7·6% at 8 years for RIT and 8·6% at 7 years for chemotherapy alone. A phase I/II trial evaluated high-dose 90Y-ibritumomab tiuxetan (median dose 2·65 GBq, range 1·35–3·88 GBq) followed by etoposide and cyclophosphamide prior to ASCT in 31 patients with CD20+ B-cell lymphoma, including 12 cases of FL. Treatment was generally well tolerated and addition of RIT did not increase transplant related toxicity or delay engraftment. At a median follow-up of 22 months estimated overall survival and relapse-free survival were 92% and 78% respectively (Nademanee et al, 2005). Thus, RIT may be a safe and important component of a pretransplant conditioning regimen.

In summary, RIT represents a highly attractive and promising approach for the treatment of FL. However, comparison with other treatment modalities in prospective randomised studies is warranted to define the definite role and the mode and timing of RIT in the overall strategy of FL therapy.

New agents and therapeutic approaches

Biological agents

Follicular lymphoma is characterised by a translocation between the chromosomes 14 and 18 that brings the BCL2 gene under the transcriptional control of the immunoglobulin heavy chain promoter. Resulting overexpression of the Bcl-2 protein can be detected in the vast majority of FL and has been associated with cellular resistance to apoptosis and chemotherapy (Gascoyne et al, 1997). Specific inhibition of gene expression can be achieved by chemically modified single-strand DNA molecules with a nucleotide sequence complementary to that of their target mRNA. Oblimersen (formerly known as G3139) is an 18-mer phosphorothioate oligonucleotide targeting the first six codons of the BCL2 mRNA open reading frame. A dose escalating study of oblimersen administered as a 14-day subcutaneous infusion in 21 patients with advanced, relapsed Bcl-2 positive non-Hodgkin lymphoma (NHL), including nine cases of FL, demonstrated feasibility and suggested clinical activity (Waters et al, 2000). Based on promising in vitro data, ongoing studies are currently exploring the combination of oblimersen with Rituximab (Ramanarayanan et al, 2004).

The ubiquitin-proteasome pathway is essential for maintaining intracellular protein homeostasis, thus representing a valid target in the treatment of malignant disease. At the centre of this degradation pathway is the 26S proteasome, an ATP-dependent multicatalytic protease. Various oncogenes and regulatory proteins for cell cycle progression and apoptosis are processed by this pathway including p53, p21, p27, nuclear factor kappa B (NF.B) and Bcl-2 (Adams, 2004). Bortezomib is a potent, selective, reversible inhibitor of the 26S proteasome and demonstrated clinical activity in relapsed multiple myeloma. Encouraging data were also observed in patients with other lymphoproliferative malignancies, including FL and mantle cell lymphoma (Orlowski et al, 2002; O'Connor et al, 2005). Aiming for enhanced antitumour activity of anthracyclins, a phase I trial has been successfully conducted defining the maximum tolerated dose (MTD) of bortezomib combined with pegylated liposomal doxorubicin (Orlowski et al, 2005). As preclinical data indicated additive cytotoxic activity of bortezomib and rituximab, a phase II trial is currently evaluating this combination therapy for patients with indolent lymphoma (de Vos et al, 2005). Another interesting agent is the orally available dipeptidyl peptidase inhibitor (PT-100, Talabostat), which acts via novel immune mechanisms by upregulated gene expression of certain cytokines within the haematopoetic tissue (Adams et al, 2004).

Increasing insights into the biology of rituximab mechanism and resistance form the rationale for developing clinical strategies to optimise anti-CD 20-directed therapy, including the combination of rituximab with granulocyte colony-stimulating factor (Niitsu et al, 2004) or with immunostimulatory oligonucleotides containing non-methylated CpG motifs (Friedberg et al, 2005). Furthermore fully humanised monoclonal antibodies with potential advantages over murine or chimeric compounds are now entering clinical evaluation, (Morschhauser et al, 2005).

Adoptive immunotherapy

Active immunotherapy in FL aims to generate a humoral and/or cellular immune response of the host against highly-specific lymphoma antigens. The hypervariable region (idiotype, Id) is unique to each lymphoproliferative clone, representing an attractive target for vaccination strategies. A recent retrospective analysis of 136 patients with FL who received Id vaccination following chemotherapy demonstrated anti-Id responses in 47% and significantly prolonged PFS (8·21 vs. 3·38 years, P = 0·018) in those with detectable idiotype-specific humoral immune response (Weng et al, 2004). Efficacy of this approach might be further improved by adding granulocyte-macrophage colony-stimulating factor for enhanced induction of T-cell response (Bendandi et al, 1999). Probably the optimal setting for vaccination strategies is the minimal residual disease (MRD) state with adequate recovery of immune effector mechanisms after chemotherapy. Improved manufacturing methods are being developed to make these individualised vaccines feasible for more widespread use.

In the light of constantly increasing clinical experience with novel therapeutic agents there is justified optimism to shift treatment towards pathogenesis-oriented and lymphoma-specific approaches. Whether this will translate into an impact on the natural history of FL remains to be awaited. Obviously, further comparative clinical trials are desperately needed, as clinicians will have to face the challenge of how to best integrate these exciting new options into existing established treatment algorithms.

Acknowledgements

This paper is dedicated to Prof. Dr. Günter Brittinger, a pioneer of lymphoma therapy in Germany, on the occasion of his 75th birthday.

Ancillary