INTRODUCTION
Follicular lymphoma (FL) is the commonest indolent non‐Hodgkin lymphoma (NHL) comprising approximately 20% of all NHLs.1 The median age of diagnosis is 60–65 years.2, 3 Approximately 80% of patients are diagnosed with advanced‐stage disease and are stratified into high or low tumor burden (LTB) based on the Groupe d'Etude des Lymphomes Folliculaires (GELF) criteria.4 Observation with active surveillance, termed “watchful waiting” (WW), was developed from randomized trials that demonstrated that treatment at diagnosis did not improve overall survival (OS).5, 6, 7 Since then, the GELF criteria have been used as a framework to initiate treatment in high tumor burden (HTB) patients or select patients for WW with LTB. A recent Surveillance, Epidemiology, and End Results (SEER) analysis of 23,000 patients with newly diagnosed FL reported that 40% were managed with WW.8 In the future, more cases of LTB FL are expected from incidental diagnoses within imaging‐driven health care systems and known stage migration with the increasing availability of positron emission tomography (PET) imaging.9, 10, 11, 12 The emergence of cell‐free DNA‐based multicancer early detection platforms raises the possibility of diagnosing occult FL with a blood test, thereby redefining the earliest stage of disease recognition.13
LTB FL is understudied, common and heterogeneous. For decades, LTB FL has been a default diagnosis defined as the absence of criteria for treatment initiation. WW has continued as the standard of care for several reasons: similar OS to treatment initiation in randomized trials; limited access to rituximab in some regions; the absence of purpose‐built prognostic models or biology‐driven strategies for time to treatment (TTT); and familiarity.
Long‐term data comparing rituximab to WW have recently been published.14, 15 New prognostic models have been proposed to predict TTT to refine LTB patient selection for WW.16, 17, 18 Advances in molecular biology have led to improved understanding of FL lymphomagenesis, biomarker development and the use of circulating tumor DNA (ctDNA) for response assessment. As clinical heterogeneity is a product of biological heterogeneity, understanding the pathogenesis of FL will be crucial to consider cure. Trials with bispecific antibodies and ctDNA are currently underway examining if LTB FL can be cured.
In this review, we summarize pertinent literature underpinning current management strategies, provide a framework for individual management, and propose future directions for the field.
HISTORICAL PERSPECTIVE
Observations in the 1970–1990s demonstrated that advanced‐stage FL was incurable, had an indolent course yet most patients eventually deceased of disease with a median OS of 7–10 years.4, 5, 6, 7, 19, 20, 21 However, asymptomatic patients with LTB had a slower progression and more favorable prognosis. Data from Stanford demonstrated asymptomatic patients undergoing WW had a similar OS to patients receiving immediate treatment.20 The median TTT was ~2.5 years and a minority never required therapy. Another seminal observation was that a minority of patients undergoing WW had a partial or complete spontaneous regression.6 In contrast, a randomized trial compared intensive combined‐modality therapy (ProMACE‐MOPP plus radiation) against WW and demonstrated no difference in OS.22 These observations established the concept of WW and led to the development of randomized trials comparing WW to less intensive treatments.
PRE‐RITUXIMAB TRIALS AND THE DEVELOPMENT OF CRITERIA TO DEFINE LTB FL
In 1981, the British National Lymphoma Investigation (BNLI) compared immediate treatment with chlorambucil 10 mg/day continuously versus WW in asymptomatic advanced‐stage indolent NHL.21 They defined criteria for asymptomatic disease (Table 1). With a median follow‐up of 16 years, there was no difference in OS (median OS chlorambucil 5.9 years vs. WW 6.7 years, P = 0.84).
At 10 years, 19% of all WW patients, and 40% of those older than 70, had not received treatment. The authors concluded that an initial policy of WW was appropriate.
In 1986, the GELF criteria were developed to standardize the definition of tumor burden requiring treatment (Table 1).23 The OS of LTB patients undergoing WW was superior to HTB patients receiving chemotherapy, thereby validating the GELF criteria for patient selection.4 Patients with LTB were randomized to prednimustine, interferon alfa‐2b, or WW.23 There was no difference in freedom from treatment failure or OS among the three arms.
Several themes emerged from these early studies.4, 6, 20, 21, 23 Treatment did not improve OS compared to WW. Patients managed with WW frequently required treatment with a median TTT of ~2–2.5 years with a small proportion not requiring therapy with available follow‐up, thereby justifying WW. Some patients experienced complete or partial spontaneous remission with WW.
There are limitations common among all criteria used for treatment initiation: they were developed based on OS not TTT; they provide a baseline assessment of symptoms, laboratory values and disease volume but were not designed to be used in a dynamic manner; they apply a binary cut‐off to variables that represent a biological continuum; there is subjectivity in assessing some criteria; they are agnostic to tumor biology. Recently, individual GELF criteria were evaluated and reported to have different prognostic weighting representing “hard” (constitutional symptoms, serous effusions, and cytopenias) and “soft” criteria with implications for management decisions.26 Patients with soft criteria may be safely observed with no detriment in survival. Despite these shortcomings, the GELF criteria have been widely used as a framework for clinical practice and trial entry for HTB FL studies.27, 28, 29 These early studies led to the establishment of GELF criteria to define LTB disease and the widespread practice of WW in the pre‐rituximab era.
RITUXIMAB TRIALS IN LOW TUMOR BURDEN FL
Rituximab is an anti‐CD20 monoclonal antibody that first improved OS when added to cyclophosphamide, vincristine, and prednisolone (CVP).30, 31 Registry data subsequently demonstrated substantial prolongation of median OS beyond 20 years in the rituximab era.32 This improvement in HTB FL led to the use of rituximab in LTB FL trials (Table 2).
In 2004, the UK National Cancer Research Institute (NCRI) commenced a randomized trial comparing rituximab induction (four weekly intravenous doses of 375 mg/m2), versus rituximab induction followed by maintenance (two monthly for 2 years) versus WW. The primary endpoint was time‐to‐new treatment (TTNT).25 With rituximab induction, the rate of complete response was 36%. With long‐term follow‐up, treatment was not initiated at 15 years in 65% of rituximab maintenance patients, 48% of rituximab induction patients, and 34% of WW patients with no difference in OS.14 Similarly, nearly a quarter of patients observed at diagnosis remained treatment free after 15 years in a recent retrospective study.37 For patients requiring first new systemic treatment, ~80% received immunochemotherapy (ICT) and responses were similar among the three groups.
An economic analysis demonstrated rituximab induction was cost effective compared to WW.38
In 2003, the Eastern Cooperative Oncology Group commenced a randomized trial examining rituximab induction followed by either rituximab maintenance or retreatment in LTB FL (RESORT).35 Patients responding to rituximab induction were randomized to re‐treatment with rituximab induction or maintenance rituximab (every 3 months), both until treatment failure. There was no difference in the median TTT failure (3.9 years retreatment vs. 4.3 years maintenance; P = 0.54).
The median progression‐free survival (PFS) for rituximab induction was 3.2 years. With long‐term follow‐up, there was no difference in 10‐year OS (83% vs. 84%).15
The LYSA FLIRT trial evaluated rituximab induction versus a subcutaneous rituximab schedule of induction followed by maintenance every 2 months for four doses.24 The primary endpoint of improved PFS was met (4 years PFS rituximab subcutaneous 58% vs. induction 41%; P = 0.0076). Higher rituximab exposure during the first 3 months was associated with complete remission (CR), PFS, and TTNT, suggesting that the short maintenance treatment had minimal impact. The LYSA BRIEF trial explored rituximab induction with two cycles of bendamustine followed by rituximab maintenance in patients aged >60 with FLIPI scores of ≥2.34 The trial was discontinued early due to treatment‐related mortality (5%).
Collectively, rituximab trials demonstrate several important lessons.14, 15, 24, 25, 35, 36 A proportion of WW patients did not require treatment after 15 years, and their OS was comparable to rituximab cohorts. With rituximab induction, the approximate overall and complete response rates are 75% and 35%, respectively, and the median PFS is approximately 3 years.24, 25, 35 Importantly, most trials used computed tomography (CT)‐based evaluation, which underestimates the response compared to rates of complete metabolic remission (CMR) with PET imaging. Most studies used rituximab induction consisting of four weekly doses.
It is less clear whether 2 years of maintenance rituximab should be the standard after induction. Rituximab delays PFS and TTNT but does not appear to cure the disease and does not improve OS. TTNT is significantly influenced by response to rituximab.14, 15 Rituximab did not affect response or durability to subsequent ICT.14, 39 Rituximab will result in some patients avoiding ICT for life.
In LTB studies, rituximab is relatively safe with a low risk of infection (Grade ≥3 1%–2%), a low risk of neutropenia (Grade ≥3 1%–2%), and no Grade 5 events reported.24, 25, 35 However, the consequences of B‐cell depletion were unmasked by the immunologic challenge of the COVID‐19 pandemic, resulting in a markedly reduced ability to seroconvert following vaccination and a higher mortality from infection.40, 41
QUALITY OF LIFE: WW VERSUS RITUXIMAB
The goals of treatment in FL are for patients to live longer (OS) and better (quality of life [QOL]). Considering the favorable OS for most patients with FL, QOL is a vital endpoint.32 QOL is a true clinical endpoint for patients and clinicians as it assesses treatment efficacy and safety.42 With WW, QOL assesses the impact of the disease and well‐being in the absence of treatment. In the NCRI study, QOL scores improved in the induction and maintenance arms, indicating rituximab was not detrimental and resulted in improved QOL in some domains.43 Conversely, WW patients were more likely to avoid thinking about their illness, did not find learning about the illness helpful, and were more likely to attach unpleasant emotions to clinic visits.
Both groups reported their illness had less impact on QOL over time as patients adjusted to living with the disease. In the RESORT trial, surveillance with rituximab retreatment was not associated with increased anxiety compared to rituximab maintenance.44 A 12‐month longitudinal study demonstrated an improvement in social, physical, and functional QOL scores in patients receiving treatment compared to patients managed with WW.45 These data demonstrate that treatment with rituximab is typically associated with improved QOL. However, collective data may not apply to the individual.
These results should be considered when counseling patients regarding rituximab versus WW. Some patients will be content to avoid treatment, even with rituximab. Conversely, for some patients, WW may have a negative impact on QOL (“watch and worry” or “scanxiety”), despite the goal of delaying treatment.
WATCHFUL WAITING: BENEFITS, RISKS, AND PRACTICES
The obvious benefit of WW is the avoidance of ICT. However, clinical trials and retrospective analyses demonstrate a median TTT of 2.5–5 years with most of these patients receiving ICT.21, 25, 46 Although WW has been considered safe, there is limited data examining the risk of complications during WW. A Canadian study reported a one in four risk of organ dysfunction or transformation during WW and an inferior OS for patients experiencing organ dysfunction.47 A study from the Australasian Lymphoma Alliance (ALA) reported similar findings with approximately 10% of patients experiencing a complication from increasing tumor volume (e.g., organ compromise/obstruction) or active lymphoma (e.g., thrombosis).18
The optimal approach to disease surveillance during WW is uncertain with variation among guidelines or no specific guidance provided.48, 49, 50, 51 The National Comprehensive Cancer Network guidelines recommend history, physical examination, and blood tests every 3–6 months for the first 5 years, surveillance imaging no more than every 6 months for the first 2 years, and then surveillance imaging no more than annually after 2 years.50 These guidelines are based on expert opinion rather than data. Consequently, there are heterogeneous practices regarding follow‐up and the use of imaging during WW.
Asymptomatic progression in FL is common. Surveillance imaging has two purposes: (1) to monitor asymptomatic patients for disease progression to initiate treatment to avoid morbidity; (2) to identify patients without significant progression who can safely continue WW and avoid treatment. In the ALA study, imaging prior to treatment was defined as surveillance imaging in 20% or prompted by clinical findings in the remaining 80%.18 For patients undergoing prompted imaging, the indication in ~75% was symptoms and/or examination findings.
Consistent with prior reports, blood test results alone (e.g., raised lactate dehydrogenase) had limited value in detecting progression or relapse.18, 52 This highlights the importance of regular clinical appointments, and the role patients and clinicians play in monitoring disease progression.
These data are consistent with reports demonstrating most relapses of FL were detected from signs or symptoms rather than imaging.53, 54 This suggests a reactive approach to imaging rather than planned surveillance may be appropriate in some cases. Conversely, surveillance imaging remains important in some cases based on disease location to monitor for progression to avoid complications. However, this must be balanced with the low rate of detection of progression, the risk of second cancer with repeat imaging, and cost to the health care system.53, 54, 55
TRANSFORMATION
Transformation is an event in the natural history of FL with disease evolution to DLBCL. Nonrandomized data demonstrate a reduced incidence of histological transformation in patients treated with rituximab‐chemotherapy versus chemotherapy alone (10‐year risk 5.2% vs. 8.7%).56 There is variability in retrospective studies as most but not all have identified an increased risk of transformation with WW versus immediate treatment.8, 39, 57, 58, 59, 60 Considering transformation is an uncommon event, large patient numbers are required for adequate power to examine the question.
The SEER analysis demonstrated upfront treatment reduced the risk of histological transformation compared with WW with only a minimal impact on OS; however, this analysis was in patients managed with WW not strictly LTB FL.8 There is potential bias in retrospective studies due to selection bias and the definition of transformation (clinical vs. histological). Importantly, in the randomized UK NCRI trial, despite an early increased risk with WW versus rituximab (hazard ratio [HR] 0.34, P = 0.052),25 15‐year follow‐up did not demonstrate an increased risk of transformation and OS was similar.14 In LTB FL, the available evidence does not support the use of rituximab to reduce the incidence of transformation. This is an important issue for patient counseling.
The development of a predictive biomarker for transformation remains a major unmet need.61
PROGNOSTIC MODELS FOR TIME TO TREATMENT
Variation in TTT arises from heterogeneity within LTB FL during WW ranging from early progression, stable disease, late progression, or, rarely, spontaneous regression.25 The GELF criteria define LTB but do not predict TTT. Clinicians do not appear to be purely using the GELF criteria to select patients for WW. Real‐world analyses of clinician behavior demonstrate that 20%–55% of GELF‐positive patients were observed26, 62 and 20%–30% of GELF‐negative patients were treated.26, 63 Consequently, there is a need to develop a prognostic model for TTT to better stratify patient selection for WW.
An ideal model would identify two groups: (1) a high‐risk group with a short TTT, where treatment at diagnosis should be considered, and (2) a low‐risk group with a longer TTT in whom WW would be appropriate.
Recently proposed prognostic models have been developed using TTT as the endpoint.16, 17, 64, 65, 66 Five clinical and laboratory factors (stage, elevated beta‐2‐microglobulin, lymphocyte‐monocyte ratio, bone marrow involvement, and splenomegaly) identified 35% of patients who progressed within 24 months of WW and had an inferior OS.17 The clinico‐genetic m3 PRIMA‐PI incorporates clinical parameters and gene mutations (KMT2D, EP300, and TP53) to stratify patients into three groups.16 A limitation across some of these studies has been the lack of a validation cohort. Only the m3 PRIMA‐PI was validated in an independent cohort.16 LTB‐TREAT uses elevated LDH and >4 nodal areas to identify three groups with distinct rates of TTT, where the low‐risk group is predicted to have a prolonged TTT with WW and a high‐risk group is predicted to experience early progression in whom treatment should be considered.18 These data have implications for treatment decisions and patient counseling regarding management strategies. Importantly, none of these models has been prospectively validated to guide treatment decisions or incorporated into international clinical guidelines.
Limited data suggest quantitative PET metrics may assist in predicting TTT. Standard uptake value max (SUV max)67 and total metabolic tumor value (TMTV) with a cut‐point of 14 cm3 have each been reported as independent predictors of TTT.64, 65 These potential biomarkers may be used at baseline and subsequently with monitoring to identify a subgroup less suitable for observation, but this requires prospective validation.
CIRCULATING TUMOR DNA
ctDNA is constantly shed into the blood from malignant B cells via apoptosis, proliferation, and secretion.68 Minimal residual disease (MRD) testing via a liquid biopsy represents a non‐invasive method of assessing ctDNA. Recent developments have markedly improved the sensitivity (~1/1,000,000) of next‐generation sequencing with the interrogation of multiple somatic mutations on the same DNA strand, called phased variant enrichment sequencing.69
Several themes have emerged across the limited data from FL studies incorporating ctDNA MRD testing.70, 71, 72, 73 ctDNA levels fall rapidly to MRD negativity after only several cycles of varying treatments in the front‐line and relapsed/refractory settings, and there is high concordance between CMR and MRD negativity (80%–90%). The combination of CMR and MRD negativity identifies a group with a superior PFS. Among PET‐positive patients, MRD‐positive patients had inferior outcomes compared to MRD‐negative patients.
CURRENT CLINICAL TRIALS IN LTB FL
The last decade has seen an unprecedented expansion of novel agents in FL. For a novel therapy to change the natural history of LTB FL, several criteria would have to be met: (1) manageable acute toxicity; (2) minimal long‐term toxicity; (3) the absence of treatment‐related mortality; (4) the ability to avoid ICT for life; and (5) curative potential.
Bispecific antibodies are a promising class for LTB FL. CD20xCD3 and CD19xCD3 bispecific antibodies enable a T cell to form a functional immune synapse with a B cell and effect cell death. They have demonstrated remarkable activity in relapsed/refractory HTB FL.70, 74, 75, 76 Several themes emerged across these trials, which were conducted during the COVID‐19 pandemic.70, 74, 75, 76 These agents have a high rate of CMR (60%–70%) with a rapid time to response (median time of 1.5–3 months).
High grades of cytokine release syndrome (Grade ≥3 1%–2%) and immune effector cell‐associated neurotoxicity (ICANS) (Grade ≥3 0%) are uncommon in FL. The major adverse event is infection (Grade ≥3 infection 20%–40%). Many infections were viral respiratory tract infections (e.g., COVID).
The incidence of infection appears proportional to exposure with higher rates seen with continuous therapy. Opportunistic infections remain a rare but significant complication.
This activity has led to the development of trials in LTB FL77, 78, 79, 80 (Table 3). Despite different designs, each trial is exploring the hypothesis of cure. The MorningSun trial examines mosunetuzumab (up to 8–17 cycles depending on response) in LTB FL.77 The rate of CMR was 87%, the incidence of grade ≥3 infection was 8%, and there was no Grade 5 adverse event with a median follow‐up of only 9 months.
These early data demonstrate the potential of bispecific antibodies to induce remission.
FUNCTIONAL CURE VERSUS BIOLOGICAL CURE
OS in FL remains inferior to the age‐ and gender‐matched control population (relative survival ~85%).82, 83, 84 As the commonest cause of death for patients with LTB FL is lymphoma, exploring cure is an important concept, especially for young patients.14 What does cure mean? Functional cure refers to a survival estimate that is comparable to the age‐ and gender‐matched population and where patients die with, rather than from, FL or its treatment.85 If patients receive treatment and relapse after long remissions, avoid ICT for life and die with a normal life expectancy, then this approach has still modified the course with disease‐free survival (functional cure). Biological cure refers to the absence of a single neoplastic cell in a patient's body but is an elusive concept to both achieve and prove.
Achieving biological cure requires a highly active therapy. Proving biological cure requires a highly sensitive response assessment tool.
FUTURE DIRECTIONS
There are several areas for future research to advance the field: refine LTB by its biology (risk stratification), endpoint selection, clinical trials beyond rituximab, incorporating ctDNA and prioritizing QOL and minimizing treatment burden.
A biologically informed approach to LTB FL
There is a need to move beyond the GELF criteria (biology agnostic) to a biologically informed approach to FL. There has been limited work performed in understanding the pathogenesis of LTB FL. Across studies, differences between LTB and HTB are subtle, and comprehensive integrated analyses of tumor genomics together with the tumor microenvironment remain limited.86, 87, 88 These observations suggest LTB FL is best viewed not as a distinct biological entity, but as a clinical state of a genomically established lymphoma.
The key question is why it remains controlled in some patients despite an advanced‐stage presentation. It is plausible that host‐tumor interactions play a dominant role, with preserved immune surveillance maintaining a dynamic equilibrium that limits tumor expansion.
Future translational work will provide a biologically informed approach to enable risk‐stratification to guide front‐line management decisions (WW vs. rituximab vs. clinical trial). The development of a pre‐therapy biomarker to predict early progression, transformation, or an indolent course would facilitate this. This will require international coordination to pool patient material as prior LTB trial cohorts and real‐world studies have smaller numbers of patients than HTB FL trials.
To be implemented, it must be developed and then independently validated, linked to a clinically relevant turn‐around‐time to facilitate management and remain inexpensive. Finally, the development of predictive biomarkers would identify therapeutic vulnerabilities in biologically defined subsets of FL patients, thereby tailoring treatment based on biology.
In the meantime, the application of current prognostic models for predicting TTT in patients with LTB FL will aid patient counseling and management decisions. These models may be used as tools for risk stratification, closer surveillance, and the identification of populations for prospective early‐intervention trials. The selection of one model for widespread application to guide TTT will harmonize future management strategies.
Future studies incorporating biological variables (TMTV, ctDNA, and genomic data) with clinical factors are needed to improve risk stratification. Incorporating tumor biology is a vital focus for future models.
Endpoint selection
Endpoints are useful to define goals in clinical practice and trials. In LTB FL, the most appropriate endpoints differ based on management strategy.39 For patients managed with WW, TTT is an important endpoint reflecting how long a patient can be safely observed before disease progression necessitates treatment; however, this requires long‐term follow‐up and is influenced by clinician judgment and patient wishes. Conversely, for actively treated patients, PFS represents an assessment of response and durability.
For all patients, OS is a vital endpoint, but due to the indolent nature of LTB FL and the presence of effective therapies, long follow‐up is required to compare strategies. All these endpoints require considerable follow‐up. For future trials, MRD will be explored as a potential early surrogate endpoint as it may provide an assessment of long‐term disease control sooner than traditional time‐to‐event outcomes; however, it is currently unclear what MRD means for long‐term outcomes.
Clinical trials beyond rituximab
Rituximab prolongs TTNT but does not improve OS versus observation or appear to cure the disease. Early intervention trials are currently exploring the hypotheses that LTB FL can be cured, altered, or transformation can be prevented. Bispecific antibody therapy will presumably result in high rates of CMR and ctDNA MRD negativity.
The question is whether these responses will be durable with serial response assessment. If so, brief therapy would be acceptable for some patients as it would be expected to minimize or shorten T‐cell exhaustion, B‐cell aplasia, hypogammaglobulinemia, and reduce the incidence of infection. Importantly, these concepts represent hypotheses that are being tested in ongoing clinical trials.
They are not presently supported by mature, randomized clinical trial data.
The potential benefits of bispecific antibodies in LTB FL must be weighed against risks. The incidence and severity of adverse events, most notably infection, and the monitoring required may represent overtreatment for some patients. In an asymptomatic cohort where WW would be acceptable, this is an important consideration.
Antigen loss (i.e., CD20 or CD19) affecting subsequent therapies may represent another potential challenge. The results of ongoing trials will determine if improvements in disease control, MRD clearance and prolonging TTT will outweigh the risks associated with bispecific antibodies. In the future, other novel agents, personalized neoantigen vaccines or CAR‐T‐cell therapy may represent therapeutic options.81, 89, 90
Incorporating ctDNA in the management of LTB FL
Currently, ctDNA represents an investigational biomarker in LTB FL. It remains in the research phase and has not been incorporated into clinical practice or management guidelines. Most evidence has been generated in HTB FL, where it also remains in the research phase.
There are several potential applications in LTB FL. Baseline quantitative ctDNA is prognostic in HTB FL.91 In LTB FL, quantitative ctDNA has predicted TTT.92 ctDNA may be expanded to complement current or future prognostic models for TTT. Baseline tumor genotyping with a qualitative assessment of reporter variants could predict early progression, transformation, or an indolent course. ctDNA is being explored as a potential highly sensitive response assessment tool. ctDNA combined with PET imaging could determine the deepest possible remission state to explore biological cure.
Although undetectable MRD has not been used as a primary endpoint in a trial for regulatory approval, it represents an important exploratory endpoint for future development. However, at present there is no data with serial ctDNA assessment to determine if FL can be cured with any therapy. Large data sets with long‐term follow‐up are needed to define the value of MRD negativity in FL. ctDNA may become a tool for disease surveillance to guide treatment initiation.
This would be an important advance to minimize reliance on imaging to reduce the long‐term risk of second cancer from radiation exposure.55 There are important limitations and future considerations for ctDNA. There is a need for international harmonization of pre‐analytical, analytical, and post‐analytical issues. Currently, there is insufficient clinical evidence to support the routine use of ctDNA to guide decisions regarding WW versus active treatment.
In FL, clonal haematopoiesis of indeterminate potential or early founder mutations may be present but not representative of disease.93
A focus on QOL
A key focus on management in clinical trials and real‐world practice in LTB FL must be QOL to ensure patient‐centered care. QOL should be considered an essential endpoint in LTB trials. Clinical trials must be designed with a consideration of toxicity (e.g., drug combination, treatment length, depth, and duration of immunosuppression).
An assessment of short‐ and long‐term toxicity and QOL data from trials will be vital in asymptomatic patients. If treatment does not cure disease, it must have a favorable safety profile and preserve or improve QOL. Some patients may be unnecessarily exposed to toxicity where instead therapeutic advancements through improving QOL and de‐escalation strategies may be more appropriate goals.
In clinical practice, WW will remain an important strategy for many patients. The burden of WW is underrecognized in some patients. There is a need for a greater awareness and assessment of QOL to continually engage the patient in the decision‐making process.
This will support all patients with LTB FL to live better.
CURRENT MANAGEMENT RECOMMENDATIONS
Current management decisions require an individualized approach to consider the priorities for each patient (Figure 1) and the goals of each management strategy (Table 4). This must consider clinical variables, age, life expectancy, comorbidities, disease location, and the patient's wishes. With ~25%–35% of patients managed with WW remaining untreated at 15 years, WW remains a standard approach for LTB FL.14, 37 This is regardless of age or fitness as randomized data demonstrate similar OS and risk of transformation compared to rituximab.14 For older, less fit or comorbid patients, WW is an appropriate approach, particularly when avoiding treatment‐related toxicity and preserving QOL are dominant considerations.
Similarly, young patients may wish to defer treatment with the hope of avoiding or prolonging time to ICT. As WW is a management strategy, not a treatment, it requires a dynamic assessment (i.e., active surveillance) regarding its appropriateness to balance the risk of complications against the benefits of avoiding treatment. Long‐term follow‐up from two randomized trials demonstrates rituximab effectively prolongs TTNT with minimal acute‐ and long‐term toxicity and may result in a lifetime avoidance of ICT; however, it does not appear to cure LTB FL.14, 15 These data establish rituximab as a reasonable alternative to WW and provide a benchmark against which future clinical trials will be compared.
Rituximab can be considered for all patients (young and old) with: a preference to treat to delay or avoid ICT; anxiety related to WW; documented disease progression during observation not meeting criteria for HTB; or anatomical location. Current clinical trials with bispecific antibodies are exploring the hypothesis of cure. Patients most likely to consider these trials will be younger and fitter who are willing to accept the possibility of treatment‐related adverse events with the presumption of disease control that is more durable than rituximab.
CONCLUSION
LTB FL is common and heterogeneous. Evidence now supports an individualized, nuanced approach to management. The historical practice of WW remains supported by mature randomized evidence demonstrating no impact upon survival compared to immediate treatment with rituximab.
Conversely, rituximab provides a meaningful prolongation of TTT and improvements in QOL, establishing this as a reasonable alternative. The future challenge is refining patient selection as current approaches rely on clinical criteria to describe tumor burden without consideration of underlying biology. Future research exploring biological risk will aim to provide a personalized approach to management decisions to determine which patients can be safely observed versus benefit from which early treatment.
Current early intervention trials are examining if novel agents can induce deep and durable remissions or possibly cure. The ultimate goal is to match treatment intensity with biological risk, preserve QOL, and minimize toxicity for all patients. In practice, an informed discussion between clinician and patient will make the right decision for each patient to live longer and better.
AUTHOR CONTRIBUTIONS
Greg Hapgood: Conceptualization; writing—original draft; writing—review and editing. Joshua Tobin: Writing—review and editing. Robert Kridel: Writing—original draft; writing—review and editing.
Brad Kahl: Writing—review and editing.
CONFLICT OF INTEREST STATEMENT
G.H.: none. J.T.: none. R.K.: research funding (to institution): AbbVie, AstraZeneca, Bristol Myers Squibb, Merck, and Roche.
Honoraria: Roche and Kite/Gilead. B.K.: Consulting: AstraZeneca, Genmab, ADCT, BMS, Roche, Nurix, Octapharma, BeOne, Incyte, AbbVie, Lilly, Genentech, and Foresight. Research funding: Roche and BeOne.
FUNDING
This research received no funding.