What is responsible for the recent improvements in outlook for patients with follicular lymphoma?

Since the earliest reports of the use of systemic chemotherapy for low grade or follicular lymphoma, the prevailing dogma has stated that the choice of initial therapy had no impact upon the overall survival (OS) of patients. This paradigm was given credence by the lack of change in outcome of patients with follicular lymphoma from series reported from the 1960s to the 1980s [1]. In this issue of the journal, Conconi et al. from the Oncology Institute of Southern Switzerland describe an improvement in the cause-specific survival of 281 consecutively treated patients with newly diagnosed follicular lymphoma at single institution over the period 1979–2007 [2]. This difference in outcome has been attributed to alterations in therapeutic agents chosen to treat the disease over this timeframe, namely alkylating agents, anthracycline based chemotherapy combinations, and use of the anti-CD20 monoclonal antibody rituximab in combination therapy. Unlike several recent previous reports from large co-operative groups and cancer institutions, an improvement in OS was not seen in the Swiss cohort, thought to be due to the shorter follow-up of the more recent cohort treated with immuno-chemotherapy. In this particular cohort, one-third of patients had early stage disease, where involved field radiotherapy remains the accepted standard of care. However, only 13% received this treatment, a figure similar to that seen in the US Surveillance, Epidemiology and End Results (SEER) database [3]. Excellent local control and high cure rates achieved with radiotherapy in this situation mean that alterations in outcome in this cohort with newer salvage strategies are very difficult to discern. Furthermore, the picture is sullied by the generally higher rates of bone marrow disease reported in more recent series of patients, which probably reflect more thorough and detailed examination of marrow trephine specimens [4]. Nevertheless, the apparent genuine improvement, albeit moderate, in the outcome of patients with follicular lymphoma over the last 25 years has provoked significant debate as to which of several potential factors is chiefly responsible.

Due to the use of radiation in many patients in even the earliest series reported, the natural history of untreated follicular lymphoma is surprisingly unclear. Despite this, follicular lymphoma is the archetypal example of an indolent neoplasm, albeit one with a remarkable variation in progression. The development of tools such as the Follicular Lymphoma International Prognostic Index (FLIPI) has aided clinicians in assessing individual prognosis. However, disease bulk at presentation remains the most important prognostic factor in progression free survival (PFS) and OS, unlike several other hematological diseases which utilize cytogenetic and molecular risk factors to stratify prognosis and treatment. Following responses to the alkylating agent chlorambucil in the 1950s, chemotherapy become the mainstay of treatment for advanced follicular lymphoma, and the disease became characterized by responses to treatment, although with inevitable relapses of shortening duration [5]. Early reported series of intensification using combination chemotherapy such as CVP (cyclophosphamide, vincristine, prednisone) or CHOP (cyclophosphamide, doxorubicin, vincristine, prednisone) failed to show a benefit in PFS or OS [6]. Nevertheless, a recent long term follow-up of large series suggested that initial treatment with CHOP reduced the risk of subsequent transformation to a high grade lymphoma by half [7], and another that there is a greater risk of transformation in patients treated expectantly, with the implication that occult transformation is present in a proportion of patients at presentation [8].

Meta-analyses of cooperative studies and tumor registries now suggest a true improvement in the OS of patients with follicular lymphoma. Using the SEER database, Swenson et al. demonstrated an improvement in the median OS from 84 to 93 months from 1978 to 1999 [3]. Most improvement was seen with younger patients with stage III/IV disease more likely to receive initial intensive combination chemotherapy. Effects of the chimeric anti-CD20 antibody rituximab on this cohort appeared unlikely to affect the analysis due to the short follow-up of treated patients at that time. Similar conclusions regarding the OS benefit to younger patients with advanced disease were drawn from an analysis of patients referred to Stanford University between 1986 and 2003. Although the contribution of rituximab was not assessed, improvements in OS pre-dated the availability of frontline rituximab [9]. The M.D. Anderson Cancer Center published treatment cohort data showing an improvement in 5-year OS from 64% to 95% over five successive clinical trials which latterly included rituximab, with improvements seen in all groups, but most significantly in those with a FLIPI score of 3 [10]. Cox regression analysis suggested the improvement in outcome was due to treatment rather than any differences in baseline clinical features, with progressive improvements seen in OS, failure free survival, and even survival after relapse in later cohorts. Similar results were reported from Southwest Oncology Group (SWOG) data, showing successive improvement in OS from patients treated on consecutive studies both before and after the introduction of monoclonal antibody therapy. The benefit of treatment appeared to occur across all subgroups when differences in IPI were adjusted for. An analysis by the Gruppo Italiano per lo Studio dei Linformi (GISL) of patients treated on consecutive studies suggested that rituximab combination therapies resulted in improvements in response rates, OS, and survival after first relapse [11].

Unsurprisingly, debate centers around the impact of rituximab on response rates and duration remissions for frontline therapy as a single agent and in combination in salvage and maintenance therapy for both previously untreated and relapsed follicular lymphoma. The benefit of frontline immuno-chemotherapy and maintenance rituximab in previously untreated patients has been the focus of two separate Cochrane meta-analyses. With regard to frontline immuno-chemotherapy, a total of 1943 patients from seven randomized clinical trials in indolent lymphoma confirmed marked improvement in response rates, but also a reduction in the hazard ratio for death to 0.63 (95% confidence interval [CI] 0.51–0.79; p < 0.001) [12]. Data from five randomized controlled trials of 985 patients showed a reduction in the hazard ratio for death to 0.58 (95% CI 0.42–0.79; p = 0.0003) with the inclusion of a variety of administration schedules of maintenance rituximab [13]. The European Organisation for Research and Treatment of Cancer (EORTC) randomly allocated patients with relapsed/refractory disease to CHOP or rituximab and CHOP. Patients in complete remission or partial remission after induction were further randomized to maintenance rituximab or observation. Rituximab maintenance significantly improved PFS to a median of 3.7 years, compared to 1.3 years with observation. Five-year OS was 74% in the rituximab maintenance arm, compared to 64% in the observation arm, which may have resulted in statistical significance but for the far greater number of patients who received CHOP receiving a salvage regimen containing rituximab at progression [14]. The Groupe d'Étude des Lymphomes de l'Adulte (GELA) have reported benefits in OS for patients relapsing after chemotherapy with rituximab and autologous transplant at first relapse [15], while improvements in responses to salvage therapies appear enhanced by the addition of rituximab. Crucially, there remains a lack of prospective data of the benefit of induction immuno-chemotherapy and maintenance rituximab, and the eagerly awaited results of the randomized Primary Rituximab and Maintenance (PRIMA) study should provide much needed clarification. Questions will however remain regarding the optimum duration and schedule, as well as the merits of rituximab maintenance versus retreatment at relapse.

Despite the understandable excitement regarding the effects of rituximab more than a decade after its introduction into treatment algorithms, at least some of the aforementioned studies suggest that improvements, at least in some areas, pre-dated frontline antibody therapy. Key to the debate is the question as to whether the intensity of firstline treatment of follicular lymphoma truly affects long term outcome, or results in a selection of a group of patients deemed high risk by their short remission duration. High-risk patients may go on to receive salvage immuno-chemotherapy followed by autologous stem cell transplant, fludarabine based regimens, or transplant with adjuvant radio-immunotherapy, options only lately routinely available. The potential impact of improvements in treatment options at relapse are particularly relevant, given that single agent immuno- and radio-immunotherapy at least appear able to induce remissions of similar durability to those with initial use. Successive effective treatments in all patient groups may confer survival advantages in a disease characterized by a relapsing–remitting course. The development of further effective drugs such as humanized anti-CD20 antibodies, immunomodulatory agents such as lenalidomide, BH-3 mimetics, or the rediscovery of older drugs such as bendamustine may mean the improvement seen thus far may be the tip of the clinical iceberg. It does, however, appear highly unlikely that we will ever know the correct permutation and sequencing of drugs required for the inevitable later relapses, and whether a ‘one size fit all’ strategy based on patient health or a specific patient-tailored approach based on clinical and biological parameters will be superior. However, regardless of the relative contribution of rituximab, the combined weight of these recent reports demonstrates that the paradigm of initial therapy having no impact on OS has clearly has been broken, paving the way for clinical trials in previously untreated patients which include OS as a primary measure of efficacy.

In the ideal world, improved understanding of the use of immuno-chemotherapy and the tailored succession of salvage therapies will be supplemented by an increased understanding of the role of host stroma and immune response in determining disease biology. The mechanisms of rituximab-induced tumor cell death, and the nullification of the previously adverse effect of high tumor-infiltrating macrophage content, are prime examples of areas requiring ongoing investigation. Furthermore, the merits of achieving a greater depth of remission as judged by positron emission tomography (PET) or molecular negativity and later impact on clinical course will hopefully become clearer over the next decade.