Indolent lymphomas: pushing the pace with novel agents

In 2014, at the age of 38, the patient was found to have asymptomatic abdominal adenopathy during a nonurgent cholecystectomy, which was biopsied and shown to be grade 1-2 follicular lymphoma (FL). His Follicular Lymphoma International Prognostic Index (FLIPI) at diagnosis was 4 for stage, number of involved lymph node regions, elevated lactate dehydrogenase, and anemia. He was enrolled in a clinical trial of rituximab, lenalidomide, and ibrutinib in September 2014, and he had a partial response. He remained on therapy for 16 months but eventually was taken off it in January 2016 because of disease progression. He was then enrolled in a randomized clinical trial of bendamustine/rituximab (BR) with or without venetoclax; he did not receive venetoclax. Scans and bone marrow biopsy at the end of 6 cycles of treatment showed a complete response (CR). This response lasted only 2 months, however, and in August 2017, he was noted to have new low-volume adenopathy. This was biopsied and showed grade 1-2 FL. He was enrolled in a clinical trial of an autologous anti-CD19 chimeric antigen receptor (CAR) T-cell product for the treatment of relapsed/refractory indolent B-cell non-Hodgkin lymphoma (B-NHL). He had grade 1 cytokine release syndrome (CRS) but no neurotoxicity after treatment in December 2017. At 1 month, scans showed a partial response to therapy, but by 3 months, he again had progressive disease. He was followed for several months until he had disease that was amenable to biopsy. This showed recurrent FL with retention of CD19. He was offered retreatment with CAR T cells but instead chose to enroll in a study of an anti-CD20 bispecific antibody. After mild CRS, a CR to therapy was achieved. He remains on this therapy, tolerating it well, to this day.

Indolent B-NHLs represent a group of lymphomas that historically have been very treatable but not generally curable with conventional therapies. Survival with these diseases tends to be long, however, with patients living 20 years or longer, on average, and many patients dying with their lymphoma rather than of their lymphoma. Chemotherapy has long been a standard treatment of these patients in the first-line setting. Advantages are that it offers time-limited therapy with a known and predictable side effect profile. Disadvantages are that it is chemotherapy, with a known and predictable side effect profile. Given the long natural history and expected relapse with these lymphomas, the goal of therapeutic advances should be either to achieve a cure or to maximize efficacy while minimizing toxicity.

The incorporation of the anti-CD20 monoclonal antibody rituximab (“R-”) into upfront chemotherapy regimens best exemplifies this treatment paradigm and was a major advance in the treatment of indolent B-NHL. A number of novel, nonchemotherapeutic agents have been developed for B-cell lymphomas in the past decade, including drugs that enhance host immunity against the cancer cell, such as lenalidomide; drugs that inhibit kinases that promote B-cell proliferation and survival, such as the Bruton tyrosine kinase (BTK) inhibitor ibrutinib or the phosphatidylinositol 3-kinase (PI3K) inhibitors idelalisib, copanlisib, and duvelisib; drugs that promote apoptosis, such as the B-cell lymphoma 2 (Bcl-2) homology 3 (BH3) mimetic venetoclax; and epigenetic modifier drugs, such as the enhancer of zeste homolog 2 (EZH2) inhibitor tazemetostat (Table 1). The side effect profile of these drugs is different from that of traditional chemotherapy, and most require indefinite therapy with chronic low-grade toxicity. For the majority of patients, though, they are generally better tolerated. In addition, many of them have activity both as single agents and in combination in relapsed disease. As such, efforts have been made to move these regimens into earlier lines of therapy in an attempt to offer patients a “chemotherapy-free” treatment of their lymphoma.

The patient in the present clinical case, at the young age of 38 at diagnosis, opted for one such experimental strategy, combining rituximab, ibrutinib, and lenalidomide, for the treatment of his newly diagnosed FL. When he progressed through combination targeted therapy, the patient returned to a chemotherapy-based treatment, but he had a very transient response to this as well. Without confidence that further chemotherapy would lead to a durable remission, he turned to immunotherapeutic approaches in order to avoid allogeneic stem cell transplant. His FL represents an extreme example of resistant disease, but he is responding to bispecific antibody therapy after unsuccessful CAR T-cell therapy. This last class of therapies, those that harness the immune system for the treatment of these lymphomas, is being developed in parallel to some of these other new drugs in these diseases. This class includes immune checkpoint inhibitors, agonist antibodies against immunostimulatory molecules, antibody–drug conjugates, bispecific antibodies, and finally CAR T cells (Table 2). The hope with these therapies, largely stemming from the curative potential of allogeneic stem cell transplant for these lymphomas, is that they may offer a curative path for patients with these diseases without the risk of graft-versus-host disease and other complications of transplant. This update reviews the results of clinical studies exploring these targeted strategies in FL, marginal zone lymphoma (MZL), and lymphoplasmacytic lymphoma (LPL) and how they change and challenge long-standing treatment paradigms in these diseases.

The immunomodulating agents (IMiDs) were developed for the treatment of B-cell malignancies in part on the basis of their ability to enhance an antitumor immune response. The mechanism whereby they stimulate T cells and natural killer cells within the tumor microenvironment, among other antitumor effector actions, was not initially understood until the discovery that these drugs bind to the cereblon ubiquitin ligase, which in turn promotes selective proteasomal degradation of Ikaros proteins in both B cells and T cells.^1,2  Preclinical models in FL further demonstrated that lenalidomide can repair an impaired immune synapse between tumor-infiltrating lymphocytes and the tumor cell.³  As a single agent in multiply relapsed FL, with a median of 3 prior lines of therapy, lenalidomide yielded responses in 27% of patients, with a median duration of response (DOR) not reached at 16.5 months.⁴  On the basis of the mechanism of action of the drug, a randomized trial of lenalidomide alone vs lenalidomide combined with rituximab in relapsed FL was then conducted (CALGB 50401).⁵  The combination was particularly active, with an overall response rate (ORR) of 76% and a CR rate of 39% compared with 53% and 20%, respectively, in the lenalidomide-alone arm. Median time to progression was almost twice as long in the R-lenalidomide arm (2 years) as in the lenalidomide alone arm (1.1 years). This study initially had a third rituximab-alone arm, but this was discontinued because of slow accrual. However, the AUGMENT study subsequently successfully randomized patients with relapsed, rituximab-sensitive FL or MZL to R-lenalidomide or R-placebo and showed an improvement in progression-free survival (PFS), the studies’ primary end point, in the R-lenalidomide group (39.4 months vs 14.1 months).⁶  Response rates in the R-lenalidomide group were similar to those seen previously, with an ORR of 78% and a CR rate of 34%, compared with 53% and 18%, respectively, for R-placebo. Importantly, the combination of R-lenalidomide had an acceptable safety profile, with cytopenias, and neutropenia in particular, being the only grade 3 or 4 toxicities occurring in >5% of patients. In addition to cytopenias, infection and cutaneous reactions did occur more frequently in the R-lenalidomide arm. How to best combine or sequence rituximab with lenalidomide in order to minimize toxicity and maximize efficacy is not clear; induction therapy with single-agent rituximab followed by either lenalidomide alone or R-lenalidomide has not been tested.

It is precisely because this safety and efficacy profile aligned with the goal of maximizing responsiveness while minimizing risk for patients with these incurable but indolent lymphomas that R-lenalidomide was tested in previously untreated patients with FL in the phase 2 CALGB 50803 study.⁷  Sixty-six patients were treated for 12 cycles, with rituximab administered weekly in cycle 1 and then on day 1 of cycles 4, 6, 8, and 10, and 95% of patients responded, with a CR rate of 72%. At 5 years, the PFS was 70% and overall survival (OS) was 100%. Responses were seen across all grades and FLIPI risk groups. The most common side effects were fatigue, diarrhea, rash, and cytopenias, but most except for cytopenias were grade 1 or 2. This combination created such optimism that a chemotherapy-free approach would rival chemoimmunotherapy in the frontline setting that the RELEVANCE trial was designed to demonstrate superiority of R-lenalidomide over R-chemotherapy in previously untreated advanced FL.⁸  In this trial, over 1000 patients were randomized to R-lenalidomide or R-chemotherapy, followed by maintenance rituximab in each arm. Although the study failed to meet its primary end points of a superior CR rate at 120 weeks and PFS in the R-lenalidomide group, there was no statistically significant difference in ORR (65% vs 68%) or CR rate (55% vs 58%) at 120 weeks or in 3-year PFS (77% vs 78%) in the R-lenalidomide group compared with the R-chemotherapy group, respectively. There were trade-offs seen in each group with respect to toxicity, with anemia, fatigue, nausea/vomiting, neuropathy, leukopenia, febrile neutropenia, and alopecia being more common in the R-chemotherapy group, whereas rash, diarrhea, abdominal pain, myalgias, muscle spasms, and tumor flare reactions were more common in the R-lenalidomide group. The flaw in this study may have been in its design as a superiority study; the noninferior results are sufficiently compelling for many to support R-lenalidomide as an upfront option for patients with FL in need of treatment.

A new-generation nonphthalimide analog of the IMiDs, avadomide (CC-122), is currently in clinical development for the treatment of B-NHL, including indolent lymphomas.⁹  This drug is also a pleiotropic pathway modifier and has, in addition to enhanced immunomodulatory effects, enhanced antiproliferative and antiangiogenic activity in B-cell tumors. It also binds cereblon, leading to the ubiquitination and destruction of Ikaros proteins. It has been tested in combination with anti-CD20 antibodies in relapsed/refractory FL with good efficacy. In combination with obinutuzumab, responses were seen in 83% of patients (n = 18), and the CR rate was 50%.¹⁰  These responses were seen in both relapsed and refractory disease. Safety was similar to that of lenalidomide, with the most common treatment-related grade 3-4 adverse events being neutropenia, thrombocytopenia, febrile neutropenia, and tumor flare. Similarly, when combined with rituximab in a rituximab-refractory population (n = 37), the ORR and CR rate were 65% and 22%, respectively.¹¹  Similar responses were seen in patients who had previously received lenalidomide. One-year PFS was 66%, and toxicity was similar, although there was one grade 5 event due to infection and sepsis.

The activity of these agents is not limited to FL. Lenalidomide and rituximab were administered to 30 patients with previously untreated MZL, and the response rate was 93%, with 70% of patients achieving a CR.^12,13  Median PFS was 59.8 months, and 5-year OS was 96%; there were 2 deaths due to secondary malignancies. The same combination was administered to 50 patients with mucosa-associated lymphoid tissue lymphoma (gastric, 30%; nongastric, 70%). Approximately one-fourth of patients had received prior systemic therapy. The ORR was 80%, and CRs were seen in just over half of patients (54%).¹⁴  Long-term follow-up is not available for this study, but 2 patients had to discontinue treatment after just one cycle and maintained their response at 15- and 20-month follow-up.

B-cell receptor signaling is altered by a variety of mechanisms in a number of B-cell malignancies, including indolent B-NHL, making components of the pathway rational targets for the treatment of these diseases. Two such components, BTK and PI3K, have effective inhibitors, both in clinical development and with US Food and Drug Administration (FDA) approval. These include the BTK inhibitor ibrutinib and the PI3K inhibitors idelalisib, copanlisib, and duvelisib. These drugs have been evaluated for the treatment of indolent B-NHL with promising results.

The patient opted for entering a clinical trial of a chemotherapy-free regimen on the basis of the promising results of R-lenalidomide in previously untreated patients with FL and indolent B-NHL. He chose to participate in a clinical trial exploring a novel, triplet combination of R-lenalidomide plus ibrutinib. When used as a single agent in relapsed/refractory FL in the DAWN study, ibrutinib led to responses in 20.9% and CRs in 11% of patients.¹⁵  Although the response rate was disappointingly lower than the primary end point (ORR, 30%) and median PFS was only 4.6 months, median DOR was 19.4 months, which compares favorably with DOR seen in patients with multiply relapsed FL, half of whom were treated in the fourth line or beyond. Slightly more promising results were seen in 40 patients with relapsed/refractory FL treated in a second phase 2 study, in whom the ORR was 37.5% and the CR rate was 12.5%.¹⁶  Median PFS in this study was 14 months. Sensitivity to rituximab was associated with an increased likelihood of response (ORR, 52.6%), as was wild-type CARD11. When combined with rituximab in previously untreated patients, responses were seen in 82% of the 60 patients treated; CRs were seen in 27%.¹⁷  On the basis of these results, the Alliance for Clinical Trials in Oncology conducted a study of R-lenalidomide in combination with ibrutinib in previously untreated patients with FL, in which the patient was treated.¹⁸  This regimen will not move forward, owing to a high rate of dose-limiting rash seen in this study, with 82% of patients experienced a rash of any grade; 36% of patients experiencing a grade 3 rash; and 50% of patients requiring a dose reduction of the regimen for a variety of reasons, including rash. Although the ORR in the 22 patients was 95% and the 1-year PFS was 80%, these data were comparable to data for R-lenalidomide alone, and significantly more toxicity was seen. When combined with BR at a dose of 560 mg daily in a phase 1b study, 11 of 12 patients with FL treated in first line and beyond responded, and 6 of 12 patients had a CR.¹⁹  Importantly, the combination was safe and tolerable, with the most common high-grade toxicities being cytopenia and rash. There are ongoing combination studies of ibrutinib in FL, including the randomized study of R-ibrutinib vs R-placebo (NCT02947347) and the randomized SELENE study of ibrutinib in combination with either BR or R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone; NCT01974440), in previously untreated FL, for which we await the results.

Ibrutinib has greater activity in MZL and LPL, however, and is approved by the FDA for the treatment of both of these diseases. In relapsed/refractory MZL, FDA approval was based on the results of a phase 2 study of 63 patients.²⁰  In this study, ibrutinib 560 mg daily was continued until progression or toxicity, and the ORR was 48%; 7 patients achieved a CR. The median DOR had not been reached at a median 19.4-month follow-up, and median PFS was 14.2 months. Treatment was well tolerated, with the majority of grade 3 and 4 treatment-related adverse events being anemia, fatigue, and infection. There was one grade 5 bleeding event (cerebral hemorrhage), but all other bleeding events were grade 1 or 2. The rate of atrial fibrillation in this study was 6%, with all occurring in patients with known risk factors. Diarrhea led to treatment discontinuation in 3% of patients. The planned MALIBU trial will examine the combination of ibrutinib and rituximab in previously untreated MZL (NCT03697512).

When administered to 63 patients with relapsed/refractory LPL, ibrutinib 420 mg daily led to an ORR of 90.5% and a major response rate of 73%.²¹  Response was positively correlated with MYD88-mutant, CXCR4 wild-type tumors. Estimated 2-year PFS and OS were 69.1% and 95.2%, respectively. As in MZL, the most common toxicities were cytopenias; all bleeding events were low grade, and atrial fibrillation occurred in 5% of patients. Ibrutinib was examined in rituximab-refractory LPL, and the results were similarly encouraging, with 90% of patients responding and 71% of patients achieving a major response.²²  Responses were durable as well, with the 18-month PFS estimated to be 86%. Because of the safety and efficacy profile of ibrutinib in LPL, a randomized trial, iNNOVATE, comparing rituximab/ibrutinib with rituximab/placebo was conducted in both previously untreated and relapsed/refractory LPL.²³  One hundred fifty patients were randomized, including 68 treatment-naïve patients and 82 patients in relapse. ORR was 72% in the investigational group compared with 32% in the R-placebo group. Responses again correlated with MYD88L265P/CXCR4WT or MYD88L265P/CXCR4WHIM genotypes (WHIM: warts, hypogammaglobulinemia, infections, myelokathexis syndrome). PFS at 30 months was 82% in the R-ibrutinib group compared with 28% in the R-placebo cohort. This improvement was seen in both untreated and relapsed LPL. Specifically, in previously untreated patients, the 2-year PFS was 80%. R-ibrutinib differed significantly from R-placebo with respect to the incidence of hypertension (13% vs 4%) and atrial fibrillation (12% vs 1%). There were more serious adverse events with R-ibrutinib (43% vs 33%), and the most common of these were pneumonia/respiratory tract infection and atrial fibrillation. On the basis of these results, the FDA approved this combination for LPL as first-line therapy and beyond.

Although inhibiting BTK is highly effective in MZL and LPL, its utility in FL is less compelling. Targeting PI3K, however, has been promising in FL in particular. The oral PI3K inhibitor idelalisib was examined in a phase 2 study in 125 patients with relapsed/refractory indolent B-NHL.²⁴  The majority of patients had FL (n = 72); 15 patients and 10 patients had MZL and LPL, respectively. The ORR was 57%, and although the CR rate was only 6%, the median DOR was 12.5 months. Median PFS was 11 months. Responses were similar across all histologies. The most frequent grade 3 or higher adverse events experienced in this study included diarrhea, pneumonia, and dyspnea; 25 patients discontinued treatment because of liver function abnormalities, pneumonia, pneumonitis, diarrhea, and/or neutropenia. Dose delays and/or reductions were successful in a majority of patients who experienced liver function abnormalities and/or diarrhea. There was one death resulting from pneumonitis. This efficacy and safety profile was maintained in subsequent subgroup analyses across multiple risk groups.²⁵  It is on the basis of the results of this study that idelalisib was approved by the FDA for the treatment of relapsed/refractory FL in the third line and beyond. After the approval of idelalisib in FL as well as in chronic lymphocytic leukemia, studies of idelalisib in combination were initiated (NCT01980888, NCT01732926). These studies were closed prematurely, however, on the basis of an unacceptable safety profile because patients experienced high rates of immune consequences, such as pneumonitis, colitis, and hepatitis, as well as immune dysfunction with an increased risk of opportunistic infections.

Idelalisib remains a treatment option for relapsed/refractory FL and MZL as a single agent, where it has been joined by two additional PI3K inhibitors: copanlisib and duvelisib. Although idelalisib is specifically an inhibitor of the δ-isoform of PI3K, copanlisib is an inhibitor of the α- and δ-isoforms, and duvelisib is an inhibitor of the δ- and γ-isoforms. The dual-isoform inhibition is believed to be important in terms of both efficacy and toxicity. In addition, copanlisib is dosed intravenously and intermittently, which may improve its safety profile in terms of its effects on immune alterations. Copanlisib was tested in 142 patients with relapsed or refractory indolent B-NHL (FL, 104; MZL, 23; LPL, 6).²⁶  The ORR was 59%, and 12% of patients had a CR to therapy. Patients with FL and MZL responded similarly well (ORRs, 59% and 70%, respectively), whereas only 1 patient with LPL had a partial response (ORR, 17%). The median DOR was 22.6 months, and the median PFS was 11.2 months. The side effect profile was different from that of idelalisib, with more transient hyperglycemia and hypertension seen because of inhibition of the α-isoform, but less pneumonitis, hepatitis, and colitis. There were 3 nonfatal opportunistic infections in this study. Copanlisib is currently being studied in randomized trials alone or in combination with rituximab, against placebo alone or in combination with rituximab (CHRONOS-2 and CHRONOS-3), in combination with standard immunochemotherapy in relapsed indolent B-NHL, and in combination with nivolumab in solid tumors and lymphoma. Similarly, duvelisib was investigated in an open-label phase 2 study in a total of 129 patients with lymphomas, including patients with refractory FL (n = 83), MZL (n = 18), and small lymphocytic leukemia (SLL; n = 28).²⁷  The additional inhibition of the γ-isoform is believed to inhibit T-cell and macrophage polarization, thus disrupting the tumor microenvironment believed to be important in the development and maintenance of these lymphomas. The ORR was 47.3%, but very few patients achieved a CR (1.6%). Responses were similar in FL and MZL (ORR, 42.2% and 38.9%, respectively) but were higher in SLL (67.9%). Median DOR and PFS were 10 months and 9.5 months, respectively. The most common toxicities were diarrhea, nausea, neutropenia, fatigue, and cough, each occurring in >27% of patients. Rates of diarrhea and colitis were similar to those seen with idelalisib, but rates of liver function abnormalities were lower. As with copanlisib, there were 3 opportunistic infections in this study.

When the patient returned to standard chemoimmunotherapy, he did so by participating in a clinical trial comparing standard BR with BR in combination with the BH3 mimetic venetoclax. Given the pathognomonic translocation involving the gene for Bcl-2 on chromosome 18 and the immunoglobulin heavy chain locus on chromosome 14 seen in FL, FL seemed like the prototypical disease in which to test the efficacy of this drug. However, 29 patients with FL were included in the phase 1 study of venetoclax in relapsed/refractory NHL, and responses were seen in only 29% of patients, and CRs were seen in only 4 patients (14%).²⁸  The median PFS in FL was 11 months, however, which compared favorably with that of the PI3K inhibitors, and responses were greater at the higher dose levels tested. There were only 4 patients with LPL and 3 patients with MZL in this study, but the ORR in LPL was 100%, and in MZL, it was 67%. Combining venetoclax with rituximab or chemoimmunotherapy, however, may have a greater impact on FL than its use as a single agent, because overexpression of Bcl-2 in this disease may contribute to resistance to conventional therapies. This hypothesis led to the study in which the patient was treated.²⁹  In this study, patients with relapsed/refractory FL either were treated with R-venetoclax (n = 53) or were randomized to BR-venetoclax (n = 51) or BR alone (n = 51). The ORR of R-venetoclax was 33%, with CRs seen in 14% of patients. The ORR was higher in the two BR groups, but it was not significantly different in the group that also received venetoclax in addition (ORR, 68% vs 64%, respectively; CR, 50% vs 41%, respectively). Data regarding response durability showed a nonsignificant improvement in median DOR with the addition of venetoclax to BR (24.87 months vs 15.64 months). The combination of BR-venetoclax was more toxic than BR alone, with more nausea, neutropenia, diarrhea, thrombocytopenia, and fatigue, as well as more early discontinuation of therapy. A variety of combination strategies with venetoclax are being explored in FL, including in combination with ibrutinib (NCT02956382), polatuzumab (NCT02611323), atezolizumab (NCT03276468), copanlisib (NCT03886649), and lenalidomide (NCT02992522).

Single-agent venetoclax has also been investigated in a phase 2 study in relapsed/refractory LPL.³⁰  Thirty patients were treated with 800 mg of venetoclax daily for up to 2 years, and the ORR was 80%, with a major response seen in 53% of patients. Response was not correlated with CXCR4 mutation status or prior BTK inhibitor exposure, and the therapy was well tolerated, with the most common toxicities being cytopenia, nausea, diarrhea, and rash. Venetoclax is a promising agent, then, for LPL that has relapsed after BTK inhibition.

Mutations in the histone methyltransferase EZH2 are among the most frequent seen in FL, where they are associated with a good prognosis.³¹  An oral inhibitor of EZH2, tazemetostat, has been evaluated in a phase 1 study in relapsed or refractory B-NHL and advanced solid tumors.³²  Among the 21 patients with B-NHL, the ORR was 38%, including some CRs. Seven of the 21 patients had FL, and 1 patient had MZL; 2 of the 7 patients with FL had a CR (29%). Responses were durable, with CRs ongoing at 2.3-year follow-up. The drug was well tolerated, with the most common toxicities including asthenia, anemia, anorexia, nausea, and vomiting; most were grade 1 or 2. In a phase 2 study in relapsed/refractory FL, 76 patients have been treated with tazemetostat.³³  Among patients with an EZH2 mutation (n = 22), the ORR was 82%, but only 5% of patients had a CR. Median DOR had not been reached at 32-week follow-up. Among the EZH2 wild-type tumors, responses were less frequent but still seen in 35% of patients, with similar CR rates of approximately 6%, and responses have been durable to date, with median DOR not reached at 56-week follow-up. The side effect profile was similar to that seen in the first-in-human study.

The role of allogeneic stem cell transplant in indolent B-NHL is discussed in detail by Gopal in his update³⁴ ; however, its success in inducing durable remissions in highly refractory lymphomas is evidence for the use of immunotherapy for the treatment of these diseases. These therapies, which include immune checkpoint inhibitors and immunostimulatory antibodies, antibody–drug conjugates, bispecific antibodies, and CAR T-cell therapy, have the advantage of generally being less toxic than transplant, but they have been shown to have variable efficacy to date in these diseases, and whether they have curative potential remains to be seen. There is certainly evidence that supports a role for immune evasion in the pathophysiology of FL, including the facts that FL cells affect gene expression patterns of tumor-infiltrating lymphocytes resulting in impaired T-cell function and alter polarization of macrophages within the tumor microenvironment, but FL cells themselves rarely express programmed death-ligand 1 (PD-L1), and results of using immune checkpoint agents in this disease have generally been disappointing.^35,36  Unraveling the complexity of the tumor cell–microenvironment interaction in FL, though, may highlight potential immunotherapy targets and mechanisms that can be effective, and potentially curative, in these diseases.

Although patients with FL treated in the early phase studies of nivolumab had responses to these agents, with an ORR of 40% and a CR rate of 10% with nivolumab, these results were not replicated in phase 2 studies.³⁷  Results of the CheckMate 140 trial of nivolumab in FL were disappointing, with only 4% of patients responding; only 1 patient of 92 achieved a CR. Among the few responding patients, though, the median DOR was 10.4 months (results available at www.clinicaltrials.gov, NCT02038946). Combination of PD1-blocking agents with anti-CD20 antibodies such as rituximab was more effective, with an ORR of 64%; however, patients treated with this combination had rituximab-sensitive disease, so response may be related to the anti-CD20 antibody.³⁸  These responses are durable, with median DOR for R-pembrolizumab not reached at 11 months. In addition, the CR rate with R-pembrolizumab was 48%, suggesting a possible synergistic effect of the combination. Rituximab in combination with immunotherapy targeting another T-cell receptor, 4-1BB, with the agonist antibody utomilumab has an ORR of 27%, including patients with R-refractory disease, in whom the ORR was 44%; CRs were seen in 17% of patients with R-refractory disease.³⁹  Numerous combination studies are underway combining PD-1– or PD-L1–blocking agents with histone deacetylase inhibitors, lenalidomide, ibrutinib, venetoclax, chemotherapy, or other immunotherapeutic agents. Correlative science involving tumor and blood samples in these trials will inform a rational immunotherapeutic approach to the treatment of this innately immunosensitive tumor.

Immune checkpoint inhibition of the innate immune system through antibodies against the “don’t eat me” molecule CD47 is a novel and promising strategy for the treatment of B-NHL, including FL. Hu5F9 has been tested in combination with rituximab in 7 patients with FL in a phase 1 study, with responses seen in 71% and CRs seen in 43% of patients.⁴⁰  Follow-up has been short, but at 8.1 months, 91% of patients remained in response. Therapy was remarkably well tolerated, with the most common toxicities being anemia and infusion-related reactions. The phase 2 portion of this study is ongoing, and this drug is now being combined with chemoimmunotherapy for the treatment of these diseases.

Polatuzumab vedotin is an antibody–drug conjugate directed against CD79b and with a monomethyl auristatin E payload. It has been tested as a single agent in relapsed/refractory B-NHL, with promising efficacy in both indolent and aggressive diseases. When combined with rituximab or obinutuzumab in a phase 2 study in relapsed/refractory FL, responses were seen in 70% and 78% of patients, with CRs seen in 45% and 30%, respectively.^41,42  Polatuzumab vedotin has also been combined with BR and compared with BR alone in a randomized study, but CR rates and PFS were similar between the two groups.⁴³  There are ongoing and completed studies of polatuzumab in combination with CD20 antibody therapy and lenalidomide (NCT02600897), venetoclax (NCT02611323), and atezolizumab (NCT02729896), for which results are awaited.

Loncastuximab tesirine is an anti-CD19 antibody conjugated to a pyrrolobenzodiazepine dimer toxin. In a first-in-human study, 15 patients with FL were treated, and 80% had a response to therapy, with over half of patients having a CR (53.3%).⁴⁴  Toxicity was manageable and included fatigue, elevated liver function test results, anemia, neutropenia, myalgia, dyspnea, nausea, constipation, anorexia, and headache. The expansion phase of this study is ongoing. Although not an antibody–drug conjugate, the novel Fc-engineered, humanized antibody against CD19, MOR208, has demonstrated early efficacy in relapsed/refractory FL. In a phase 2 study of MOR208 as a single agent, responses were seen in 9 of 34 patients with FL (29%) and 3 of 11 patients with other indolent B-NHL (27%).⁴⁵  Half of these responses were ongoing at 12-month follow-up, and the drug was very tolerable, with infusion reactions and neutropenia being the most common adverse events reported. Although these results are not as promising as results for some of the other targeted agents in FL, the favorable side effect profile of this drug lends itself to combination strategies that may prove more effective.

Bispecific antibodies represent an alternative way to target CD19 in B-NHL. Bispecific antibodies involve antibody molecules with dual specificity so that a single molecule can bind a tumor antigen and CD3 on the surface of T cells simultaneously, thus redirecting the adaptive immune system against the tumor cell to which the antibody is also bound. Blinatumomab, the anti-CD19 bispecific T-cell engager, was tested in a phase 1 study in B-NHL, including 28 patients with FL.⁴⁶  Responses were seen in 80% of patients treated with the target dose of the drug, with 40% of patients achieving a CR. Six of 12 responding patients had responses that lasted 2 years or longer. CRS (including fever) and neurologic toxicity (not including headache) were seen in 75% and 18% of patients, respectively, but these were transient and resolved with time or discontinuation of therapy. A combination study of blinatumomab and lenalidomide in B-NHL is ongoing (NCT02568553). Bispecific antibodies targeting CD20, such as the one the patient is currently being treated with, are also in clinical development and include drugs such as mosunetuzumab and REGN1979, both of which have significant activity in FL. These drugs are discussed by Sharman in his update on CD20 targeting agents.⁴⁷ 

Anti-CD19 directed CAR T cells have had a remarkable impact on the treatment and prognosis of relapsed and refractory aggressive B-NHL, with up to 40% of patients achieving durable remissions.^48,49  This efficacy is associated with a significant risk of CRS and neurologic toxicity that is more pronounced than that seen with the bispecific antibodies, however, and it is this toxicity profile that has the potential to limit its applicability in more indolent diseases with a number of other effective therapeutic options. However, if anti-CD19 CAR T cells are potentially curative for these largely incurable malignancies, they could spare a patient multiple relapses and several lines of treatment. Furthermore, for a high-risk population such as that represented by the young patient in the present clinical case, with a high FLIPI, short remission to chemoimmunotherapy, and multiply relapsed disease, CAR T-cell therapy may represent the best chance for long-term survival. In the first report of the anti–CD19-28z–CAR T-cell construct that would become axicabtagene ciloleucel, 2 of 2 patients with indolent B-NHL (1 with splenic MZL and 1 with FL) had responses to therapy for >23 and >11 months, respectively, that were ongoing at the time of reporting.⁵⁰  The anti–CD19-41BBz–CAR T-cell construct that would become tisagenlecleucel was tested in 14 patients with FL.⁵¹  The CR rate in this population was 71%, with 89% of patients remaining in response at a median follow-up of 28.6 months. Severe CRS was seen in 18% of patients in this study, which also included 17 patients with diffuse large B-cell lymphoma, and high-grade neurotoxicity was seen in 11%, with 1 death due to complications related to encephalopathy in a patient with FL. A study of anti-CD19 CAR T cells with axicabtagene ciloleucel in relapsed/refractory indolent B-NHL, including FL and MZL (NCT03105336), is ongoing, and the efficacy and safety results of this therapy in a larger cohort of patients with longer follow-up are awaited to understand how this therapy may impact the treatment landscape and natural history of these diseases.

Indolent B-NHLs are a group of slow-growing, mature B-cell malignancies that have a prolonged natural history but are largely incurable outside an allogeneic stem cell transplant. Patients with these diseases survive 20+ years, though, on average, and as such, they often experience the benefit of clinical and therapeutic advances in their diseases made subsequent to their diagnosis. The goals of such advances are twofold: to minimize toxicity while not sacrificing efficacy and to turn these diseases into curable malignancies. The two goals are not necessarily mutually exclusive. As new combinations are explored that can achieve minimal residual disease–negative responses, as new insights are made into the genomic and molecular underpinnings of these neoplasms, and as the relationships between these tumor cells and their microenvironment are unraveled, effective and potentially curative strategies appear to be moving within reach. Many tout a goal of achieving “chemo-free” therapies, but perhaps more important than disposing of chemotherapy is the ability to predict which patients and which tumors would benefit most from any specific type of therapy and to spare patients not chemotherapy, but ineffective therapy.

Caron Jacobson, Dana-Farber Cancer Institute, 450 Brookline Ave, Boston, MA 02215; e-mail: caron_jacobson@dfci.harvard.edu.