Chapter 99: Follicular Lymphoma

Follicular lymphoma (FL) is an indolent lymphoid neoplasm that is derived from mutated germinal center B cells and exhibits a nodular or follicular histologic pattern. It is typically composed of a mixture of small, cleaved follicle center cells (centrocytes) and large noncleaved follicle center cells (centroblasts). The disease has masqueraded under multiple monikers, including “nodular lymphoma” in the Rappaport classification, and “follicle center cell lymphoma” in the Working Formulation.¹ The current World Health Organization (WHO) classification proposes the terms follicular lymphoma, grades 1, 2, and 3, to differentiate cases based on the numbers of centroblasts per high-power microscopic field (see “Lymph Node Morphology and Lymphocyte Immunophenotype” below).²

FL accounts for approximately 20 to 25 percent of adult non-Hodgkin lymphomas (NHLs) in the United States, with an annual incidence of approximately 14,000 new cases per year.^3,4 FL is most common in North America and Western Europe, and much less frequent in Eastern Europe, Asia, Africa, and in Americans of African descent.² The median age at diagnosis is 59 years, and the male-to-female ratio is 1:1.7. The disease is rare in persons younger than age 20 years, and pediatric cases appear to represent a separate disease entity that is typically localized, lacks the t(14;18) translocation and BCL2 expression, and has a very good prognosis.^5,6

SYMPTOMS AND SIGNS

Patients with FL usually present with painless diffuse lymphadenopathy. Less frequently, patients may have vague abdominal complaints, including pain, early satiety, and increasing girth, which may be caused by a large abdominal mass or hepatosplenomegaly. Approximately 10 percent of patients present with B symptoms (fever, drenching night sweats, or loss of 10 percent of body weight). The disease usually is widespread at presentation, with involvement of multiple lymph node–bearing sites, liver, and spleen. The marrow is involved in 40 to 70 percent of patients at diagnosis. FL may occasionally present with primary involvement of extranodal sites, such as the skin, gastrointestinal tract, ocular adnexa, and breast, but CNS disease is rare, unless histologic transformation to diffuse large B-cell lymphoma has occurred.²

LYMPH NODE MORPHOLOGY AND LYMPHOCYTE IMMUNOPHENOTYPE

FL exhibits a predominantly nodular lymph node pattern, however, the neoplastic follicles are distorted and as the disease progresses, the malignant follicles efface the nodal architecture (see Chap. 96, Fig. 96–18), commonly resulting in the development of areas of diffuse involvement, which may predominate histologically. The WHO has developed a three-grade system for classifying FL according to the proportion of centroblasts detected microscopically: grade 1 lymphomas have 0 to 5 centroblasts, grade 2 lymphomas have 6 to 15 centroblasts, and grade 3 lymphomas have more than 15 centroblasts per high-power microscopic field (Fig. 99–1).² Grade 3 FL is further subdivided into grade 3A, in which some small centrocytes are present despite the predominance of centroblasts, and grade 3B, in which solid sheets of centroblasts are exclusively present and centrocytes are entirely absent.² Some, but not all, studies suggest that grades 1 and 2 lymphomas follow a more indolent course than grade 3 FL, and many authorities suggest that these lower grades should be treated more conservatively than grade 3 FL.⁷ Other studies indicate a similar natural history for grades 1, 2, and 3A.⁸ Nearly all authorities now agree, however, that grade 3B FL behaves aggressively and should be treated with anthracycline-containing regimens (e.g., rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone [R-CHOP]) similar to diffuse large B-cell lymphoma.² FL cells of all grades typically express monoclonal surface immunoglobulin, are positive for BCL-2, BCL6, and CD10, and express the pan–B-cell surface antigens CD19, CD20, CD22, and CD79a, but do not express CD5, CD23, CD11c, or CD43.

CYTOGENETICS

The classic cytogenetic finding detected in FL is the t(14;18)(q32;q21) translocation that juxtaposes the BCL-2 gene on band q21 of chromosome 18 with the immunoglobulin heavy-chain gene on band 32 of chromosome 14 (Fig. 99–2).⁹ The immunoglobulin enhancer element results in amplified expression of the translocated gene product and, thus, overexpression of BCL-2 protein leading to inhibition of apoptosis of affected B cells. Quantitative real-time polymerase chain reaction (PCR) assays on blood and marrow can determine the number of t(14;18)-expressing cells and may be useful in predicting the outcome of therapy. The t(14;18) translocation is found in approximately 85 percent of patients in the United States, but the translocation is present in a significantly lower percentage of Asian patients afflicted with FL. Detection of the t(14;18) translocation in lymphoid cells is neither necessary nor sufficient for the diagnosis of FL. Small numbers of B cells harboring the t(14;18) translocation can be detected in the blood of 25 to 75 percent of healthy individuals, as well as in reactive lymph nodes and tonsils if a very sensitive nested or reverse-transcription PCR assay is employed.² Additional cytogenetic abnormalities are found in the cells of 90 percent of patients with FL. The finding of multiple cytogenetic abnormalities is commonly associated with higher histologic grade and with the probability of transformation to aggressive lymphoma. A recent large, high-resolution, genome-wide copy-number analysis demonstrated that common recurrent chromosomal abnormalities include gains of chromosomes 2, 5, 6p, 7, 8, 12, 17q, 18, 21, and X and losses on 6q and 17p.¹⁰ Frequent small abnormalities are also commonly observed, including losses of 1p36.33–p36.31,6q23.3–q24.1, and 10q23.1–q25.1 and gains of 2p16.1–p15,8q24.13–q24.3, and 12q12–q13.13. Copy-number abnormalities more commonly observed in transformed FL include gains of 3q27.3–q28 and chromosome 11 and losses of 9p21.3 and 15q.¹⁰ Important candidate genes whose expression is affected by these copy-number abnormalities includeTNFRSF14, PRDM16, TP73, and ARIDIA on chromosome 1p36; BCL10 on chromosome 1p; REL and BCL11A on 2p16; BCL6 on chromosome 3q27; histocompatibility locus antigen (HLA)-B, HLA-C, CCND3, and PRDM1 on chromosome 6p21; TNFAIP3 or PERP on chromosome 6q23; CARD11 on chromosome 7p22; MYC on chromosome 8q24; CDKN2A or CDKN2B on chromosome 9p21; STAT6 on 12q13.3; and MDM2 on 12q15.¹⁰

STAGING THE DISEASE

Evaluation of FL involves performance of a medical history, physical examination (with attention to the lymph nodes in Waldeyer ring and size and involvement of liver and spleen), laboratory testing (including a complete blood count, examination of the blood film and a differential white cell count, lactic dehydrogenase [LDH], β₂-microglobulin, comprehensive metabolic panel, serum uric acid level); lymph node biopsy; marrow aspiration and biopsy; flow cytometric analysis of blood, marrow, and lymph node cells; and computed tomography (CT) of the chest, abdomen, and pelvis.⁷ Excisional lymph node biopsies are strongly preferred for the initial histologic diagnosis of FL, although in cases in which nodal masses are inaccessible, generous needle-core biopsies may suffice. The diagnosis should not be established merely on the basis of flow cytometry of the blood or marrow, or on cytologic examination of aspiration needle biopsies of lymph node or other tissue.¹¹ Hepatitis B serology should be assessed if rituximab therapy is contemplated, as hepatitis reactivation with rituximab may occasionally be life-threatening. In selected circumstances, additional CT scans of the neck, measurement of the cardiac ejection fraction, serum protein electrophoresis, quantitative immunoglobulins, and hepatitis C testing may be useful. Patients for whom chemotherapy is contemplated should receive counseling regarding contraception, fertility issues, and sperm or egg banking. The role of fluoro-2-deoxyglucose (FDG)-positron emission tomography (PET)/CT imaging in FL is rapidly evolving. Although PET/CT imaging was previously considered optional in FL, recent studies suggest that PET-negativity at the completion of induction chemotherapy is one of the most powerful predictors of both progression-free survival (PFS) and overall survival (OS) in this disease.¹²

CLINICAL AND LABORATORY VARIABLES

An international working group developed an international prognostic index (IPI) based on five independent variables (age, stage, LDH level, performance status, and number of extranodal sites) that affected OS of aggressive lymphoma patients treated with anthracycline-based combination chemotherapy.¹³ The IPI was subsequently applied retrospectively to FL and found to be predictive of both OS and PFS for FL (as well as diffuse large B-cell lymphoma). Nevertheless, the IPI was considered to be suboptimal for segregating indolent lymphoma patients into prognostic categories because only 10 to 15 percent of patients with FL fall into the poor risk category using this index. To redress this deficiency, a French cooperative group conducted a detailed prognostic factor analysis of 4167 patients with FL diagnosed between 1985 and 1992 for whom prolonged followup was available to assess OS.¹⁴ Five adverse prognostic factors were detected: age (>60 years vs. ≤60 years), Ann Arbor stage (III to IV vs. I to II), hemoglobin level (<120 g/L vs. ≥120 g/L), number of nodal areas (>4 vs. ≤4), and serum LDH level (high vs. normal). Three risk groups were defined: low risk (zero to one adverse factors, 36% of patients), intermediate risk (two adverse factors, 37% of patients, hazard ratio [HR] = 2.3), and poor risk (three or more adverse factors, 27% of patients, HR = 4.3). The Follicular Lymphoma International Prognostic Index (FLIPI) discriminated outcomes for FL better than the IPI, both in the original cohort and in later studies evaluating patients treated with modern combined rituximab-chemotherapy regimens (Fig. 99–3).¹⁵ A revised version of the FLIPI index (FLIPI2) was subsequently proposed to address perceived deficiencies in the original model.¹⁶ The FLIPI2 model is also based on assessment of five adverse risk factors, namely the presence or absence of an elevated β₂-microglobulin level, the longest diameter of the largest involved lymph node (>6 cm), presence of marrow involvement, hemoglobin level less than 12 g/dL, and age older than 60 years. Although several studies have demonstrated the superiority of the FLIPI2 model compared to the original FLIPI model, it has not been widely adopted in North America. A simpler prognostic model was developed based solely on the baseline serum LDH and β₂-microglobulin level that has been shown to be superior to the original FLIPI model and equivalent to the FLIPI2 model in prognostic power for predicting outcomes for FL patients.^17,18

GENOMICS OF FOLLICULAR LYMPHOMA

New molecular approaches are revolutionizing our understanding of the pathogenesis of FL and providing insights into the pathways that might be targeted in the future by rationally designed therapies. Early gene-expression profiling studies of biopsy specimens from patients with untreated FL identified two gene-expression signatures that allowed construction of a survival predictor enabling segregation of patients into four quartiles with disparate median lengths of survival (13.6, 11.1, 10.8, and 3.9 years), independent of clinical prognostic variables.¹⁹ One signature (“immune response 1”) was associated with a good prognosis and included genes encoding T-cell markers (e.g., CD7, CD8B1, ITK, LEF1, and STAT4) as well as genes that are highly expressed in macrophages (e.g., ACTN1 and TNFSF13B). The “immune response 2” signature was associated with a poor prognosis and included genes known to be preferentially expressed in macrophages, dendritic cells, or both (e.g., TLR5, FCGR1A, SEPT10, LGMN, and C3AR1). Flow cytometry and cell sorting confirmed that these signatures reflected gene expression by nonmalignant tumor-infiltrating immune cells (CD19-negative cells) and not by the FL cells themselves (CD19-positive cells). The length of survival correlated with the molecular features of the nonmalignant immune cells present in the tumor at diagnosis and presumably reflected the robustness of the immune response mounted against the tumor.

“Next-generation sequencing” studies employing whole-genome or whole-exome sequencing and targeted mutational analyses have demonstrated that somatic mutations in epigenetic regulators are present in almost all cases of FL, including MLL2 (also known as KMT2D), a histone methyltransferase that is mutated in 89 percent of FL; CREBBP and EP300, two highly related histone and nonhistone acetyltransferases that act as transcriptional coactivators in multiple signaling pathways that are mutated in 30 percent and 11 percent of FL, respectively; EZH2, the catalytic subunit of PRC2 that catalyzes trimethylation of lysine 27 on histone H3, a repressive chromatin mark that is mutated in 27 percent of FL; and MEF2B, a calcium-regulated gene that cooperates with CREBBP and EP300 in acetylating histones that is mutated in 15 percent of FL.^{20,21,22,23,24} Other mutational targets include genes involved in immune modulation (β₂-microglobulin, CD58, and TNFRSF14), Janus kinase (JAK)-signal transducer and activator of transcription (STAT) signaling (SOCS1 and STAT6) and B-cell receptor–nuclear factor (NF)-κB signaling (BCL10, CARD11 and CD79B).²² Mutations in CREBBP and EP300 lesions are usually heterozygous, suggesting a haploinsufficient role in tumor suppression, apparently by impairing acetylation-mediated inactivation of the BCL6 oncoprotein and activation of the p53 tumor suppressor.²¹ A unifying model of the genetic evolution of FL postulates that the t(14;18) translocation serves as a “founder mutation” that is succeeded by development of additional “driver mutations” (e.g., CREBBP), irrelevant “passenger” mutations, and accelerator mutations (TNFRSF14).^22,23 Disagreement exists over whether MLL2/KMT2D mutations are “driver” or “accelerator” mutations.^22,23 Mutations in EBF1 and in regulators of NF-κB signaling (e.g., MYD88, TNFAIP3) appear to be acquired at the time of histologic transformation to diffuse large B-cell lymphoma.²²

LIMITED STAGE I OR II FOLLICULAR LYMPHOMA

Radiotherapy

Patients with stage I or II FL represent only 10 to 30 percent of cases in most series.^2,4 Standard management for stage I or limited contiguous stage II FL involves the administration of involved field radiotherapy (35 to 40 Gray [Gy]).⁷ Adjuvant chemotherapy does not appear to improve survival in this setting, although some studies suggest that combined chemoradiotherapy may improve PFS. A retrospective review of 177 patients with stage I or II and grade 1 or 2 FL reported a median survival of 14 years following radiation therapy as a single modality.²⁷ Approximately 50 percent of the patients were relapse-free after 5 to 10 years.

Observation

Excellent survival has also been observed in highly selected patients with early stage FL who received no initial therapy.²⁵ In a group of 43 patients, 56 percent were free from the requirement for treatment for at least 10 years and 86 percent were alive 10 years after diagnosis. Based on this study, many authorities have concluded that “watchful waiting” is an acceptable alternative to radiotherapy for stage I or II FL. A watchful waiting approach may be particularly appropriate for certain variants of FL, such as FL presenting in the small intestine, which pursues a remarkably indolent course, rarely exhibits progressive growth, very rarely disseminates (two of 63 patients) and does not transform to high-grade disease.²⁶

Chemoimmunotherapy

A large observational study, the National LymphoCare Study, assessed the outcomes of 471 patients with stage I FL according to treatment administered, including rituximab plus chemotherapy (28 percent of patients), radiotherapy alone (27 percent), observation (17 percent), systemic therapy + radiotherapy (13 percent), rituximab monotherapy (12 percent), and other treatments (3 percent).²⁸ This large, prospectively enrolled group of patients showed that national guidelines endorsing radiotherapy alone for stage I FL were not followed by practicing clinicians in the majority of cases. All treatment approaches resulted in excellent outcomes, though PFS was significantly better after a median followup of 57 months in patients treated with either rituximab plus chemotherapy (84 percent) or systemic therapy plus radiotherapy (96 percent) than in patients receiving radiotherapy alone (68 percent).²⁸ This study challenges the paradigm that radiotherapy alone should be the standard of care for patients with early stage indolent lymphoma, although the observational nature of this study, without randomization to treatment arm and the absence of differences in OS, attenuates the impact of the study.

ADVANCED STAGE FOLLICULAR LYMPHOMA

Observation Alone

Many patients with FL, particularly grades 1 or 2, will exhibit an indolent, asymptomatic course despite the absence of therapy. Because there is no conclusive evidence that survival of FL patients is improved by immediate institution of therapy, or that conventional management (other than allogeneic stem cell transplantation) can cure the disease, a “watch-and-wait” approach is often recommended for patients with extensive stage II or stage III or IV FL. In one study, survival was 82 percent at 5 years and 73 percent at 10 years after an initial strategy of observation alone, and the median time until therapy was required was 3 years.²⁹ Spontaneous regressions occurred in 23 percent of untreated patients. No differences in survival were observed in a trial of 309 patients randomized to initial watchful waiting or to chlorambucil.³⁰ In another trial, patients were randomized to either watchful waiting or to immediate aggressive combination chemotherapy with prednisone, methotrexate, doxorubicin, cyclophosphamide, etoposide, mechlorethamine, vincristine, procarbazine, prednisone (ProMACE/MOPP) chemotherapy followed by total nodal irradiation.³¹ The OS rates for the two groups were similar, although the disease-free survival rate was naturally higher in the patients treated with combined modality therapy. Criteria established by the Groupe d’Etudes des Lymphomes Folliculaires (GELF) are useful to identify patients who may benefit from intervention rather than “watchful waiting.” These criteria suggest that treatment is likely to be required for patients with a maximum diameter of any site of disease greater than 7 cm, more than three nodal sites greater than 3 cm in diameter, systemic B symptoms, a spleen size greater than 16 cm, pleural effusions, local compressive symptoms, circulating lymphoma cells, or cytopenias as a result of the lymphoma.³²

Single-Agent Chemotherapy

FL patients can be palliated effectively with a variety of single chemotherapy agents (Table 99–1). Responses to single-agent therapy, such as chlorambucil, a nucleoside analogue, or bendamustine, range from 70 to 90 percent and may last for several years.^30,33

Monoclonal Antibody Therapy

Rituximab is a human–mouse chimeric monoclonal antibody that binds to the CD20 antigen that is expressed on nearly all normal and malignant B cells but not on other human tissues. After binding to B cells, rituximab induces cell death via antibody-dependent cellular cytotoxicity (ADCC), complement-fixation (complement-dependent cytotoxicity [CDC]), induction of apoptosis, and by facilitating cross-presentation of lymphoma-associated antigens by dendritic cells. Rituximab was approved by the FDA for therapy of indolent lymphomas based on the results of a pivotal trial that evaluated treatment of 166 patients with relapsed or refractory indolent lymphoma with four weekly infusions of 375 mg/m². The response rate was 48 percent, including a 6 percent complete response rate and a median time to progression of approximately 1 year.³⁴ The response rate to first-line therapy with rituximab in newly diagnosed FL is approximately 70 to 75 percent with a complete remission rate of 18 to 27 percent.^35,36 A second response to rituximab may be achieved in 40 percent of patients who relapse after an initial remission to rituximab.³⁷ Extended courses of rituximab or “rituximab maintenance” therapy have become popular. Various schedules are employed, including administration of one dose of 375 mg/m² every 2 months for 2 years (usually as part of frontline therapy), one dose every 3 months for 2 years (for relapsed patients), four doses every 6 months for 2 years, or one dose every 2 months for four doses.^{38,39,40,41,42} No comparative studies of these disparate “maintenance” rituximab regimens have been performed. Several newer, humanized or fully human anti-CD20 monoclonal antibodies (ofatumumab, veltuzumab, obinutuzumab) have been engineered to exhibit superior ADCC, CDC, or improved induction of cell death. All are undergoing clinical trials to determine if they are superior to rituximab and one of them, obinutuzumab, has recently been approved for therapy of chronic lymphocytic leukemia (CLL) (but not FL). “Biosimilar” CD20 antibodies will also soon be available as an alternative to rituximab.

Rituximab Plus Chemotherapy

The introduction of rituximab into treatment protocols for FL has revolutionized the management of this disease. Multiple randomized, controlled clinical trials have documented the superiority of combining rituximab with chemotherapy compared to the use of chemotherapy alone in terms of overall response rates (ORRs), complete response (CR) rates, event-free survival (EFS), PFS, and OS (Table 99–2).^{43,44,45,46,47} In one study, induction therapy consisting of eight cycles of rituximab, cyclophosphamide, vincristine, and prednisone (R-CVP) was compared to eight cycles of CVP without rituximab in 321 patients with newly diagnosed FL (Fig. 99–4).⁴⁸ R-CVP was superior to CVP alone in terms of ORR (81 percent vs. 57 percent), CR rate (41 percent vs. 10 percent), time to progression (34 months vs. 15 months), time to treatment failure (27 months vs. 7 months), and OS (83 percent vs. 77 percent at 4 years, p = 0.029).⁴⁵ Similarly, R-CHOP was compared to CHOP for first-line treatment of 428 patients with advanced stage FL. R-CHOP exhibited a superior ORR (96 percent vs. 90 percent), time to treatment failure (p <0.001), duration of response (p = 0.001), and OS (p = 0.016) compared to CHOP alone.⁴⁶ Similar benefits have also been reported for the addition of rituximab to other regimens for both frontline therapy and relapsed FL.^41,47,49,50 The selection of the “optimal” chemotherapy regimen to combine with rituximab remains hotly contentious.⁵¹ Two recent large international phase III randomized studies compared the efficacy and toxicity of R-CVP, R-CHOP, and rituximab-fludarabine–based therapy, which until recently were the three most commonly employed regimens for frontline therapy of FL. Both trials demonstrated superior PFS for patients treated with R-CHOP or rituximab-fludarabine regimens than with R-CVP, although OS did not differ. Both studies also concluded that regimens with fludarabine and rituximab had significantly more hematologic toxicity and a higher risk of secondary malignancies than either R-CVP or R-CHOP, making fludarabine-based regimens less desirable. The investigators conducting these trials concluded that R-CHOP was the preferred regimen for FL, although some oncologists are hesitant to routinely employ R-CHOP in patients with indolent lymphomas because of its greater toxicity, including a 1 to 2 percent risk of cardiomyopathy, and the absence of a demonstrated OS advantage. These risks may be particularly justifiable, however, for patients with grade 3 FL, as many authorities believe anthracycline-based regimens may be curable for high-grade FL, which National Comprehensive Cancer Network (NCCN) guidelines advise treated identically to diffuse large B-cell lymphoma.⁷ To further amplify the controversy, two recent randomized trials comparing bendamustine plus rituximab (BR) with R-CHOP show similar efficacy for the two regimens, and less toxicity for BR.^52,53 Although these studies have been criticized for methodologic flaws, they have been highly influential and have catapulted BR to prominence as the most popular frontline induction regimen for FL, with approximately 65 to 70 percent of FL patients in the United States and Europe currently receiving this regimen for frontline therapy.

Maintenance Rituximab

Despite the ability to induce long-lasting remissions in the majority of FL patients treated with chemoimmunotherapy regimens as outlined above, most patients will eventually relapse. Several strategies have emerged to prolong the remission durations and to delay lymphoma recurrence. The most popular approach is to administer extended courses of rituximab, also known as rituximab “maintenance” to forestall reappearance of FL. The PRIMA trial randomized 1217 patients with FL to either maintenance rituximab (375 mg/m² every 2 months for 2 years) or no maintenance therapy, following frontline induction therapy with R-CVP, R-CHOP, or rituximab, fludarabine, cyclophosphamide, and mitoxantrone (R-FCM).⁴² This study convincingly demonstrated the superiority of 2 years of maintenance rituximab (PFS 74.9 percent) compared to observation alone after induction (PFS, 57.6 percent, p <0.0001), regardless of which induction chemotherapy regimen is used. Despite the marked improvement in PFS afforded by maintenance rituximab in PRIMA, however, there was no significant difference in OS between the rituximab maintenance and observation arms. Similar improvements in PFS but not OS have also been demonstrated for maintenance rituximab given for 2 years following rituximab monotherapy in the frontline setting³⁰ and following CHOP or R-CHOP induction in the relapsed setting.⁵⁴

Radioimmunotherapy

Radiolabeled monoclonal antibodies targeting lymphoma-associated cell surface antigens, including idiotypic immunoglobulin, CD20, CD22, and HLA-DR, have emerged as effective and safe therapeutic agents for patients with FL.^55,56 Two radioimmunoconjugates targeting the CD20 antigen, ¹³¹iodine-tositumomab (Bexxar) and ⁹⁰yttrium-ibritumomab tiuxetan (Zevalin), have been extensively tested in indolent lymphomas.^57,58,59 Radioimmunotherapy (RIT) is an attractive therapeutic option for lymphomas because (1) many high-quality antibodies are available targeting pan–B-cell antigens expressed at high levels on lymphoma cells, (2) lymphomas are exquisitely sensitive to radiotherapy, and (3) cross-fire radiotherapy from β particles emitted by decaying radionuclides on targeted lymphoma cells can kill neighboring antigen-negative tumor cells (or inaccessible cells deep in tumor clumps), which would escape killing by nonradioactive antibodies. Several trials have demonstrated ORRs of 50 to 80 percent and CR rates of 15 to 40 percent in patients with relapsed or refractory indolent lymphoma treated with either ¹³¹iodine-tositumomab or ⁹⁰yttrium-ibritumomab tiuxetan.^57,59,60 In a randomized study comparing treatment of patients with relapsed FL with either⁹⁰ Y-ibritumomab tiuxetan or rituximab, the ORR (86 percent vs. 55 percent) and the CR rate (30 percent vs. 15 percent) were both statistically superior in the group treated with the radioimmunoconjugate.⁵⁹ Similarly, ¹³¹I-tositumomab was compared with unlabeled tositumomab in a randomized trial of relapsed indolent lymphoma and both the ORR (55 percent vs. 19 percent) and the CR rate (33 percent vs. 8 percent) were higher in patients receiving the radiolabeled antibody.⁶¹

Six phase II studies have studied frontline RIT for patients with newly diagnosed FL, either as a single agent or in combination with various chemotherapy regimens, including CVP, CHOP, and fludarabine. In all six studies, outstanding ORR rates (90 to 100 percent) and CR rates (50 to 96 percent) were observed with first-line RIT, with median PFSs in excess of 5 years in several of the studies.^62,63,64 A phase III randomized study evaluated the utility of consolidation therapy with ⁹⁰Y-ibritumomab tiuxetan for patients with FL in remission after frontline chemotherapy.⁶⁵ In this trial, 414 patients in either partial or complete remission after a variety of chemotherapy induction regimens (chlorambucil, CVP, CHOP, fludarabine or rituximab combinations) were randomized to either consolidation with RIT or to no consolidation. RIT dramatically improved the median PFS in the total patient population (36.5 months vs. 13.3 months, p <0.0001), and this advantage was observed regardless of whether patients were in partial remission (PR; 29.3 months vs. 6.2 months, p <0.0001) or CR (53.9 months vs. 29.5 months, p = 0.015) at the time of consolidation. Furthermore, RIT consolidation converted 77 percent of patients who were in PR after induction chemotherapy to CR following RIT. However, a recent study comparing six cycles of R-CHOP with six cycles of CHOP (without rituximab) followed by a single dose of ¹³¹I-tositumomab (CHOP-RIT) demonstrated similar outcomes in both treatment arms.¹⁷

The major toxicity of RIT is myelosuppression, with cytopenic nadirs occurring 4 to 7 weeks after treatment and requiring 2 to 4 weeks for recovery. Growth factor administration and transfusions are required in approximately 20 percent of patients. Human antimouse antibodies (HAMA) may develop in the serum of approximately 1 percent of patients treated with ⁹⁰Y-labeled ibritumomab and in 10 percent of patients treated with ¹³¹I-labeled tositumomab. A potential long-term concern with both radiolabeled antibody formulations is the potential development of myelodysplasia and acute leukemia as late complications. Hypothyroidism may also occur as a delayed toxicity of ¹³¹I-labeled tositumomab in approximately 10 percent of patients.

Interferon-α₂

Interferon (IFN)-α₂ has been studied in 10 large phase III studies evaluating its utility in both the induction phase of treatment and for maintenance therapy. A meta-analysis of 1922 newly diagnosed patients with FL treated in these trials concluded that the addition of IFN-α₂ to induction chemotherapy did not significantly influence response rates, but did show a significant difference in favor of IFN-α₂ with regard to survival.⁶⁶ Results differed greatly from trial to trial and further analyses were carried out to define the circumstances in which IFN-α₂ prolonged OS. The survival advantage was seen when IFN-α₂ was given (1) in conjunction with relatively intensive initial chemotherapy (p = 0.00005), (2) at a dose of 5 million units or greater (p = 0.000002), (3) at a cumulative dose of 36 million units or greater per month (p = 0.000008), and (4) when given with induction chemotherapy rather than as maintenance therapy (p = 0.004).⁶⁶ With regard to remission duration, there was also a significant difference in favor of IFN-α₂, irrespective of the intensity of chemotherapy used, IFN dose, or whether IFN was given as a maintenance strategy or with chemotherapy. Despite these salutary findings, IFN is rarely employed to treat FL in the United States because of its unfavorable toxicity profile (asthenia, fatigue, flu-like symptoms, cytopenias) and because of the perception that rituximab confers similar or superior advantages with much less toxicity.

Idiotype Vaccines

The idiotypic immunoglobulin protein expressed on the surface of B lymphoma cells represents a true tumor-specific antigen and is an ideal target for immunotherapeutic strategies. Several groups have reported favorable phase II trials using idiotypic vaccines produced either by rescue hybridoma fusions or recombinant DNA approaches. The idiotypic immunoglobulin in most of the vaccines is coupled to keyhole limpet hemocyanin (KLH) and administered with sargramostim (granulocyte-macrophage colony-stimulating factor) to enhance immunogenicity. Specific immune responses are generated in approximately 50 percent of immunized FL patients who are in complete remission at the time of vaccination. Patients exhibiting immune responses to the vaccine experience longer remission durations and superior survival compared to patients failing to mount immune responses to the vaccine. Three phase III randomized trials have been conducted using idiotypic vaccination following induction therapy. Two of these studies found no statistical differences in PFS between an idiotypic KLH vaccine and a control vaccine.^67,68 The third study reported an improved disease-free survival for specifically vaccinated patients,⁶⁹ although the study has been criticized for a variety of methodologic imperfections.⁶⁸

Hematopoietic Stem Cell Transplantation

The role of high-dose chemoradiotherapy and hematopoietic stem cell transplantation in the management of patients with FL remains highly controversial. Proponents of autologous stem cell transplantation for indolent NHL note the favorable outcome of a collaborative study of 121 adult patients conducted by St. Bartholomew’s Hospital and the Dana-Farber Cancer Institute, where an apparent plateau in the remission duration curve was observed in 48 percent of patients with a median followup of 13.5 years.⁷⁰ Survival was longer in patients transplanted in second remission compared to those transplanted later in their disease course. The value of autologous stem cell transplantation was also tested in a randomized trial of 89 patients with relapsed FL (Fig. 99–5). Transplanted patients experienced a marked advantage in PFS and a marginal OS advantage compared to patients randomized to continued conventional salvage chemotherapy without transplantation.⁷¹ When used as part of initial therapy for high-risk patients, randomized studies demonstrate a prolongation of PFS, but no improvement in OS.^72,73 Adverse outcomes associated with autologous stem cell transplantation include treatment-related mortality (3 to 5 percent) and a substantial increase in the incidence of secondary myelodysplasia and acute myelogenous leukemia, occurring in 7 to 19 percent of patients, particularly if total-body irradiation is employed in the conditioning regimen.

Allogeneic transplantation affords long-term PFS for approximately 40 to 50 percent of patients with relapsed FL, but widespread adoption of this approach has been hampered by transplant-related mortality ranging from 20 to 40 percent. Consequently, careful patient selection and informed consent are essential. When allogeneic and autologous stem cell transplantation are compared for patients with relapsed FL, the long-term survival rates are comparable.^74,75 Autologous stem cell transplantation is associated with a greater likelihood of dying from recurrent disease, and allogeneic stem cell transplantation results in a higher frequency of death from graft-versus-host disease, infection, and venoocclusive disease. Nonmyeloablative and reduced-intensity allogeneic transplantation conditioning regimens have been developed to exploit the benefit of a graft-versus-lymphoma effect while minimizing transplant-related morbidity and mortality. Preliminary results of this approach are encouraging, with 52 to 85 percent OS and 43 to 83 percent PFS after 3 to 11 years with 8 to 43 percent nonrelapse mortality.^76,77

TRANSFORMED FOLLICULAR LYMPHOMA

Approximately 30 to 40 percent of patients with FL undergo documented transformation to a more aggressive histology, usually diffuse large B-cell lymphoma, with an annual rate of transformation of approximately 3 percent (Fig. 99–6A). Clinically, histologic transformation is characterized by the sudden explosive growth of a single lymph node site (or extranodal mass). Anthracycline-based chemotherapy (e.g., R-CHOP) is the most appropriate therapy for patients experiencing transformation, however, most studies indicate that the prognosis is poor despite aggressive management. Whereas 50 to 65 percent of patients presenting with de novo diffuse large B-cell lymphoma are cured with R-CHOP chemotherapy, less than 10 percent of patients with diffuse large B-cell lymphoma arising by transformation from FL will be cured by this regimen (Fig. 99–6B). Most series report median survivals of 6 to 20 months for patients undergoing transformation,^29,78,79,80 although one recent study reports a 50-month median survival, with particularly good survival in patients experiencing transformation more than 18 months after initial diagnosis of FL.⁸¹ Because of the historically poor outcome of chemotherapy alone, some authorities advise either autologous or allogeneic stem cell transplantation following induction of remission with R-CHOP. A recent multicenter cohort study of 172 patients with transformed FL concluded that patients undergoing autologous stem cell transplantation had better outcomes than those treated with rituximab-containing chemotherapy alone. However, allogeneic transplantation did not improve outcomes compared with rituximab-containing chemotherapy because of high rates of transplant-related mortality.⁸²

A PRAGMATIC APPROACH TO THERAPY OF FOLLICULAR LYMPHOMA

There is currently little consensus among lymphoma experts with regard to the optimal management of patients with either frontline or relapsed FL.⁵¹ Therefore, at this time all patients with FL should be considered for entry into clinical trials to define the best regimens and to allow evaluation of the multitude of promising new drugs and antibodies that are now available. Patients who are ineligible for trials or who decline enrollment should receive individualized treatment. Patients with localized stage I or II disease may be offered local radiotherapy, observation or chemoimmunotherapy. Elderly patients with asymptomatic, stage III or IV disease are best monitored with observation alone, particularly if their disease is of low volume and if they have multiple coexistent medical illnesses. Patients who are symptomatic, have cytopenias, massive splenomegaly, effusions, or bulky adenopathy should be treated with rituximab plus chemotherapy. Several regimens are acceptable including BR, R-CVP, and R-CHOP, with the latter regimen being most appropriate for young patients with aggressive presentations and rapidly growing bulky adenopathy, B symptoms or for patients with grade 3 FL. The roles of maintenance rituximab and consolidative RIT following initial induction chemoimmunotherapy of newly diagnosed patients are contentious. It appears that either rituximab maintenance for 2 years or a single dose of consolidative RIT with ⁹⁰Y-ibritumomab tiuxetan can prolong initial remission duration and PFS, but neither improves OS. Management of FL following relapse depends on the patient’s initial treatment and the resultant remission duration. If the first remission lasts many years, the initial treatment regimen may again be used (except for anthracycline-containing regimens). If the initial remission is short, an alternative second-line regimen should be selected from among the many available options, including BR, R-CVP, R-CHOP, R-FND (rituximab, fludarabine, mitoxantrone [Novantrone], and dexamethasone), and RIT. In the near future, many additional attractive options will be available, including ibrutinib,^83,84 idelalisib,^85,86 ABT-199,^87,88 antibody–drug conjugates,⁸⁹ and adoptive immunotherapy with chimeric antigen receptor modified T lymphocytes.^90,91 Patients with a good performance status, who experience very short response durations, should be considered for autologous or allogeneic stem cell transplantation. Patients who undergo histologic transformation should receive R-CHOP and be offered the option of stem cell transplantation.

FL has been considered an indolent but incurable disease for which the median survival of approximately 10 years is minimally affected by medical interventions. This attitude is no longer valid, and survival of FL patients has been progressively increasing over the past 20 years (Fig. 99–7).^92,93,94 Much of the improvement in survival appears attributable to the introduction of rituximab, better salvage therapies (bendamustine, RIT), improved supportive care measures, and the wider implementation of stem cell transplantation. Controversy persists over whether any of the grades of FL are curable with standard chemoimmunotherapy regimens, although it is clear that OS has improved significantly in the last decade. Further clinical research will settle these disputes and permit formulation of a consensus standard of care for the management of both newly diagnosed and relapsed FL.