Immunomodulating antibodies and drugs for the treatment of hematological malignancies
- First Online:
- Cite this article as:
- Houot, R., Kohrt, H., Goldstein, M.J. et al. Cancer Metastasis Rev (2011) 30: 97. doi:10.1007/s10555-011-9274-3
- 246 Views
The aim of cancer immunotherapy is to induce immune cells to kill tumor and promote immunological memory that protects against tumor recurrence. Most current immunotherapies, such as monoclonal antibodies (mAb), target the tumor cells directly. Advances in our understanding of the immune system such as the role of co-stimulatory and co-inhibitory receptors, and the advent of new immunomodulatory agents provide new opportunities to target the immune system and enhance anti-tumor immune responses. These promising agents include immunomodulating mAbs, Toll-like receptor agonists, IMiDs, and cytokines. In this review, we discuss the current results of immunomodulating agents in the treatment of hematological malignancies and propose applications that include targeting of the innate and adaptive immune systems as well as combinations with tumor-specific mAbs.
KeywordsImmunomodulationImmunotherapyMonoclonal antibodiesCytokineCpGThalidomideLenalidomideCancerHematological malignanciesLymphomaLeukemiaMyeloma
1 Immunomodulating agents for cancer therapy
The aim of cancer immunotherapy is to induce the antitumor activity of immune cells to kill tumor and promote immunological memory that protects against tumor recurrence. Most of the current and successful approaches of immunotherapy such as monoclonal antibodies (mAbs) target tumor cells directly. However, several observations suggest that the tumor environment, particularly the immune environment, may play a crucial role in the tumor development . Therefore, manipulation of the immune system may result in tumor regression. A better understanding of the immune system, the development of new technologies, and the emergence of new immunomodulatory agents now open the door for new approaches aiming at targeting the immune system to enhance anti-tumor immunity. This goal might be achieved in one of several ways using immunomodulating mAbs or drugs such as Toll-like receptor (TLR) agonists (e.g., CpG), immunomodulatory drugs (IMiDs; e.g., thalidomide, lenalidomide), and cytokines (e.g., interleukin-2 (IL-2), interleukin-12 (IL-12), interleukin-21 (IL-21), interferon-α (IFN-α), granulocyte colony-stimulating factor (G-CSF), granulocyte-macrophage colony stimulating factor (GM-CSF)).
1.1 Modulating the innate and/or the adaptive immune system
Immunomodulating antibodies tested as single agents in preclinical models of cancer
Target expressing cells
Preclinical tumor models
Clinical grade Ab if available (trade name; sponsoring company)
T cells, NK cells, DCs, neutrophiles and monocytes
BMS-663513 (Bristol-Myers Squibb)
Induciblea (constitutive on Tregs)
Stimulatory (Abrogates Tregs suppression)
Induciblea (constitutive on Tregs)
Stimulatory (Abrogates Tregs suppression)
Sarcoma, colon carcinoma 
Induciblea (constitutive on Tregs)
Tremelimumab (CP-675,206; Pfizer); Ipilimumab (MDX-010; Bristol-Myers Squibb/Medarex)
CT-011 (Cure Tech); MDX-1106 (Bristol-Myers Squibb/Medarex)
B cells, DCs, macrophages
CP-870,893 (Pfizer); Dacetuzumab (SGN-40; Seattle Genetics); HCD122 (Novartis/XOMA)
1.2 Combination of immunomodulating agents with tumor-specific therapies
Because tumor-directed mAbs exert their antitumor effects through immune-dependant mechanisms, the integration of immunomodulatory agents with tumor-directed mAbs presents a rational approach in exploring new therapeutic combinations. Among the various mechanisms of action of tumor-directed mAbs, antibody-dependent cytotoxicity (ADCC) and antibody-dependent cellular phagocytosis play important roles through the recruitment of immune cells such as NK cells and/or monocytes/macrophages . In addition, some observations suggest that tumor-directed mAbs might also induce an adaptive immune response, often referred to as a “vaccinal effect” . Immunomodulating agents may therefore be used to stimulate innate immune cells to enhance ADCC and/or adaptive immune cells to enhance the vaccinal effect of tumor-specific mAbs. Rituximab, a mAb directed against CD20, is one of the most widely used tumor-directed mAb for the treatment of hematological malignancies, namely B-cell non-Hodgkin’s lymphoma (NHL). Here, we review clinical trials combining immunomodulating agents with rituximab for the treatment of NHL to illustrate how immunomodulating agents might be used to enhance a tumor-directed mAb’s efficacy (Table 4).
2 Immunomodulatory antibodies
Monoclonal antibodies that recognize tumor-specific or tumor-associated antigens have been used to target cancer cells directly. This approach has lead to significant improvements in the treatment of cancer, particularly in hematological malignancies . Several of these tumordirected mAbs have been approved for the treatment of cancer . Recently, it was suggested that mAbs could also be used to target normal immune cells to enhance the anti-tumor immune response . Monoclonal Abs, depending on their agonistic or antagonistic properties, can activate stimulatory receptors or block inhibitory receptors expressed on the surface of immune cells and thereby enhance their function. Most of these mAbs initially targeted coreceptors known to be expressed on T cells such as CD137, OX40, CTLA-4, and programmed-death receptor 1 (PD-1). Similar approach has later been extended to other immune cell types such as dendritic cells (DC; through CD40 targeting) or NK cells (through KIR targeting). Therefore, immunostimulatory Abs can be used to enhance the innate immune response (NK cells), the antigen presentation (DCs), and/or the adaptive immune response (T cells).
Immunomodulating mAbs that have been tested with success as single agents in preclinical models of cancer are listed in Table 1. Some of these immunomodulatory receptors are constitutively expressed on immune cells and therefore carry the risk of non-specific activation. In 2006, a phase I clinical trial tested an agonistic mAb against CD28, a stimulatory receptor that is constitutively expressed on T cells, in six volunteer patients . Non-specific stimulation of T cells led to a severe systemic inflammatory response and multisystem organ failure in all patients. Conversely, some immunomodulatory receptors are only expressed following activation and therefore, may have a lower side effect profile. These distinctions might be important to take in consideration when developing immunostimulatory mAbs. Several of these immunomodulatory mAbs already have a clinical candidate directed against the human receptor available for clinical trials (Table 1).
Theoretical advantages of immunomodulatory mAbs over tumor-directed mAbs for the treatment of cancer
Pan-tumor activity (i.e., same Ab for different tumors)
Immunomodulating Abs do not depend on tumor Ag therefore may be used across different types of tumors
Induction of an adaptive immune response (i.e., prolonged and memory response)
Whereas tumor-directed Abs only trigger the innate immune system, immunomodulatory Abs can be used to stimulate an adaptive, long-lasting immune response
Polyclonal immune response (i.e., not restricted to a single TAA)
Whereas tumor-directed Abs only target one tumor Ag, immunomodulatory Abs can generate an immune response against several tumor Ags
Risk of target mutation (i.e., tumor escape by mutation)
Tumor-directed Abs target malignant cells which are genetically unstable and can thereby escape by mutation. Conversely, immunomodulatory Abs target normal cells which are less likely to mutate and escape therapy
Requires tumor Ags identification
Whereas tumor-directed Abs require prior identification of tumor-Ags to be directed against, immunomodulatory Abs can be used while ignoring expression of tumor Ags
Requires direct access to tumor Ags (i.e., surface expression, accessibility)
Wherease tumor-directed Abs are limited to surface tumor Ags and require direct access to them, immunomodulatory Abs can stimulate responses against intracellular Ags (“seen” by T-cells), even outside of the tumor site
Risk of non-specific adverse events
Tumor-directed Abs target TAA’s which can be present on normal cells. Therefore, they may cross-react with non tumor cells and damage normal host tissue. Immunomodulatory Abs target the immune system and may not directly damage host tissue through Ab binding. However, enhancing of the immune response with immunomodulatory Abs can result in immune-related adverse events
Results of clinical trials evaluating immunomodulating Abs in hematological malignancies
Generic Ab name (trade name; sponsoring company)
Ipilimumab (MDX-010; Bristol-Myers Squibb)
O’Mahony et al. 2007 
Advanced malignancies progressing after cancer vaccine, including 4 NHL (2 FL, 2 MCL)
2 tumor regressions in NHL patients including 1 mixed response (MCL) and 1 PR (FL)
Bashey et al. 2009 
Cancer patients relapsing after allogeneic hematopoietic stem transplantation including 27 hematological malignancies (14 HD, 6 MM, 2 AML, 2 CML, 2 CLL and 1 MCL)
2 CR (HD), 1 PR (MCL)
Ansell et al. 2009 
Refractory or recurrent B-cell NHL patients (14 FL, 3 DLBCL, 1 MCL)
1 CR (DLBCL), 1 PR (FL)
Dacetuzumab (SGN-40; Seattle Genetics)
Berger et al. 2008 
Advanced hematological malignancies (4 NHL, 1 HD, 3 CLL, 1 MM, 7 AML, 1MDS)
1 CR (FL), 1 minimal response (AML)
Advani et al. 2009 
Refractory or recurrent NHL patients (21 DLBCL, 12 FL, 10 MCL, 3 MZL, 1 SLL)
One third of patients experienced tumor regression including 6 OR (12%): 1 CR (DLBCL) and 5 PR (3 DLBCL, 1 MZL, 1 MCL)
Furman et al. 2010 
Hussein et al. 2010 
Recurrent or refractory MM
9 SD (20%)
Byrd et al. 2006 
Relapsed and refractory CLL
Transient decrease of peripheral CLL cells during Ab infusion in the majority of patients (not maintained week-to-week)
Bensinger et al. 2006 
Relapsed and refractory MM
Blinatumomab (MT103; Micromet)
Goebeler et al. 2010 
Relapsed indolent NHL (FL, MCL)
9 PR and 4 CR out of 13 evaluable patients (100% ORR) treated at the dose of 60 μg/m2/d
Topp et al. 2009 
B-precursor ALL patients in complete hematological remission with persistent or reappeared MRD after consolidation of frontline therapy
13/16 patients (81%) converted into a molecular CR
2.1 Anti-CTLA-4 mAb
Cytotoxic T-lymphocyte-associated protein 4 (CTLA-4, also known as CD152), a homolog of CD28, is a negative regulator of T cell activation that binds CD80 and CD86 with higher affinity than that of CD28. Its expression is induced on conventional T cells following activation . It is also constitutively expressed on regulatory T cells (Tregs) . CTLA-4 blockade can therefore sustain T cell activation and proliferation and/or prevent Treg suppression . AntiCTLA-4 mAbs have shown antitumor effects in various tumor models (Table 1). Based on these results, several clinical trials have been performed for patients with cancer. Impressively, a recent randomized phase III trial showed improved overall survival for patients with metastatic melanoma receiving a tumor vaccine with or without anti-CTLA-4 mAb (ipilimumab, MDX-010; Bristol-Myers Squibb) . Anti-CTLA-4 mAb has also been recently evaluated in phase I/II clinical trials for patients with hematological malignancies (Table 3). These studies showed encouraging results in heavily pretreated patients but need to be confirmed in larger cohorts of patients.
2.2 Anti-PD-1 mAb
PD-1 is another co-inhibitory receptor whose expression is induced after activation of T cells . PD-1 stimulation inhibits T cell activation and proliferation. There are two known ligands for PD-1: B7-H1/PD-L1, the predominant mediator of PD-1-dependent immunosuppression, and B7-DC/PD-L2. PD-L1 is expressed on human tumor cells, including leukemia, NHL, and Hodgkin’s lymphoma [14–18] and may participate in tumor evasion . Preclinical data indicate that interrupting the PD-1/PD-L1 interaction may lead to tumor regression [19–21]. Two antagonistic anti-PD-1 mAbs, CT-011 and MDX-1106, have entered clinical trials for cancer patients.
CT-011 (Curetech) is a humanized IgG1 mAb that binds to mouse and human PD-1 and inhibits its function. In mice, administration of CT-011 leads to tumor regression [22, 23]. A single trial using CT-011 has been reported in a small number of patients with heterogeneous hematological malignancies  (Table 3). Although only a few objective clinical responses were observed in this phase I study, the results indicate that CT-011 might induce durable antitumor immune responses. A phase II trial just opened which evaluates the combination of CT-011 with rituximab  (Clinicaltrials.gov-ID: NCT00904722).
MDX-1106 (Bristol-Myers Squibb) is a fully human IgG4 mAb specific for human PD-1. It has been recently evaluated in 39 patients with solid tumors resulting in one complete response (CR), two partial responses (PR), and two significant tumor regressions not meeting PR criteria . MDX-1106 has not yet been tested in patients with hematological malignancies.
2.3 Anti-CD40 mAb
CD40 is a member of the tumor necrosis factor receptor family expressed on B cells and DCs and is crucial for their function. CD40 is also expressed on the surface of monocytes/macrophages as well as non-hematopoietic cells such as epithelial and endothelial cells . CD40 is expressed by most B cell malignancies including NHL, multiple myeloma (MM), and chronic lymphocytic leukemia (CLL) [28–30]. Engagement of CD40 on antigen-presenting cells induces expression of costimulatory molecules and production of pro-inflammatory cytokines . On tumor cells, engagement of CD40 can have direct antitumor effects. Therefore, the antitumor activity of anti-CD40 mAb may be due to both direct cytotoxic effects and indirect immunomodulatory effects.
Three anti-CD40 mAbs are currently being evaluated in clinical trials: CP-870,893 (Pfizer), Dacetuzumab (SGN-40; Seattle Genetics), and HCD 122 (formerly known as CHIR-12.12; Novartis/XOMA).
CP-870,893 (Pfizer) is a fully human CD40 agonist mAb that has both immune-mediated and non immune-mediated effects on tumor cell death. CP-870,893 is an IgG2 immunoglobulin (in contrast, most approved mAbs are IgG1) and as such is unlikely to activate complement or bind Fc receptors efficiently. Accordingly, biological effects are primarily due to CD40 signaling. This mAb has been tested in 29 patients with solid tumors but has not yet been evaluated in patients with hematological malignancies .
The second anti-CD40 mAb, SGN-40 (Seattle Genetics), has been evaluated in three phase 1, dose-escalation studies in patients with relapsed or refractory NHL, MM, and CLL (Table 3). CD40 is almost uniformly expressed in all three tumors [33–35]. SGN-40 is a humanized IgG1 immunoglobulin and a weak agonist of CD40 signaling in blood mononuclear cells. Although the overall response rate (ORR) appeared to be modest in MM and CLL [34, 35], encouraging results were seen in NHL patients . It is not known whether the antitumor effect seen in these patients results from a direct anti-tumor effect or from an immunomodulating effect.
The third anti-CD40 mAb, HCD122 (Novartis/XOMA), is a fully human IgG1 mAb with antagonistic activity that mediates ADCC and blocks CD40L-induced survival and proliferation of normal and malignant B cells . It should be noted that, distinct from CP870,893 and SGN-40, HCD122 is an antagonistic Ab and therefore is not expected to have any direct immunostimulatory activity. This Ab has been evaluated in phase I trials for patients with CLL  and MM  (Table 3). No definitive conclusions can be drawn at that point regarding its clinical efficacy.
2.4 Other monospecific immunomodulatory mAbs
An anti-CD137 mAb directed against the human receptor is also available for clinical trials. Triggering CD137 signaling has been reported to elicit robust antitumor immune responses in various preclinical tumor models (Table 1). This effect is largely attributed to CD137-mediated signaling of tumor-specific T cells, enhancing their proliferation and cytotoxic activity, and preventing their activation-induced cell death and immune tolerance [5, 39, 40]. Additionally, it has been reported that CD137-mediated signaling induces the activation of NK cells and DC [41, 42]. An anti-CD137 mAb (BMS-663513, Bristol-Myers Squibb) has been recently tested in patients with solid tumors . Early results indicate only modest antitumor activity in patients with melanoma (6% PR) and renal cell carcinoma (14% SD). Patients with lymphoma may respond better to anti-CD137 therapy since we have shown that lymphomas overexpress CD137 mRNA compared with solid tumors, presumably due to the presence of CD137+ tumor-infiltrating T cells . Importantly, we have shown that tumor B cells do not express CD137 and therefore may not be stimulated by the agonistic anti-CD137 Ab. Despite these results and promising preclinical data in lymphoma and myeloma [44, 45] (Table 1), anti-CD137 mAb still remains to be tested in patients with hematological malignancies.
Finally, a phase I clinical trial has tested a mouse anti-OX40 mAb in patients with advanced cancer (W. Urba, personal communication). To our knowledge, no patient with hematological malignancy has been treated with this Ab yet.
2.5 T-cell engaging Abs
An alternative approach to harness the cytotoxic potential of T cells is the use of T-cell engaging Abs. Bispecific T-cell engager (BiTE) molecules are recombinant protein constructs consisting of two linked single-chain Abs. One is specific for CD3, a subunit of the T-cell receptor complex, and the other is specific for a selected tumor-associated antigen such as CD19. These molecules are capable of redirecting T cells toward target cells for elimination of tumor cells independent of peptide antigen presentation or T-cell specificity. Blinatumomab (MT103, Micromet) is a BiTE Ab with dual specificity for CD19 and CD3. In an ongoing phase I study in patients with relapsed B-cell malignancies (Table 3), blinatumomab induced tumor regression in all 13 patients treated at a dose level of 60 μg/m2/d (nine PR and four CR) [46, 47]. Tumor regression was observed in patients with follicular lymphoma (FL), mantle cell lymphoma (MCL), and CLL. For an unexplained reason, a significant number of patients experienced fully reversible neurological adverse events such as confusion, disorientation, speech disorder, tremors, and convulsions. Blinatumomab has also been evaluated in patients with B-precursor acute lymphoblastic leukemia (ALL). Patients in complete hematological remission with persistent minimal residual disease (MRD) after consolidation of front-line therapy received continuous i.v. infusion of blinatumomab. In these patients, treatment with blinatumomab converted MRD-positive ALL into molecular CR in 13 of 16 (81%) evaluable patients (Table 3) .
3 TLR-9 agonists
CpG oligodeoxyribonucleotides (ODN) are short sequences of DNA containing unmethylated CpG dimers mimicking bacterial DNA. These immunostimulatory ODNs are ligands for TLR-9 found in human B cells and plasmacytoid DCs. Stimulation with CpG ODNs involves both direct effects mediated by TLR-9 and indirect effects resulting in the activation of NK cells, T cells, B cells, monocytes, and DCs.
A phase I clinical trial evaluated intravenous administration of single-agent CpG ODN (PF3512676) in 23 patients with previously treated NHL . This study revealed significant signs of NK cell activity and late clinical responses were observed in two patients (FL and DLBCL).
Selected clinical trials combining immunomodulating agents with rituximab for patients with B-cell malignancies
Immunomodulating agent combined with rituximab
Friedberg et al. 2005 
Leonard et al. 2007 
Relapsed and refractory NHL
Friedberg et al. 2009 
Relapsed and refractory FL
Kaufmann et al. 2004 
Relapsed and refractory MCL
Treon et al. 2008 
Treon et al. 2009 
Wang et al. 2009 
Relapsed and refractory MCL
Dutia et al. 2010 
Relapsed and refractory indolent NHL
75 (85% in FL)
NA (38% in FL)
Fowler et al. 2010 
Untreated indolent NHL
79% (94% in FL)
Friedberg et al. 2002 
Rituximab-naïve, relapsed and refractory FL
Gluck et al. 2004 
Relapsed and refractory NHL
Khan et al. 2006 
Rituximab-refractory indolent NHL
Ansell et al. 2006 
Relapsed and refractory NHL
Timmerman et al. 2007 
Relapsed indolent NHL
Davis et al. 2000 
Relapsed and refractory indolent NHL
Sacchi et al. 2001 
Relapsed indolent NHL
Kimby et al. 2002 
Untreated and relapsed indolent NHL
Van der Kolk et al. 2003 
Relapsed indolent NHL
Ferrajoli et al. 2005 
Untreated and relapsed CLL
Mc Laughlin et al. 2005 
Cartron et al. 2008 
More recently, our group completed a phase I/II trial combining intratumoral injections of CpG ODN (PF-3512676) with low-dose, local radiotherapy in patients with low-grade B-cell lymphoma . Based on preclinical data , it was hypothesized that intratumoral injections of CpG might activate lymphoma B cells as well as nearby antigen-presenting cells to generate an anti-tumor response. Prior to local CpG administration, the same involved lymph node was irradiated to induce tumor B cell apoptosis and prevent their proliferation. A total of 15 patients were enrolled in this trial and overall objective (systemic) responses were seen in 27%, including one CR.
Thalidomide and its derivatives such as lenalidomide represent a new class of antineoplastic drugs (IMiDs) which have been especially effective in certain hematological malignancies. These agents have anti-inflammatory, antiangiogenic, and immunomodulatory properties. Antitumor responses observed with these agents may therefore not be entirely due to immunomodulation activity. Potent antitumor efficacy of thalidomide  and lenalidomide [59, 60] was first demonstrated in myeloma patients.
Thalidomide showed modest activity in NHL patients with only 12.5% OR (one CR and two PR) out of 24 patients with relapsed or refractory indolent lymphoma . However, combination of thalidomide with rituximab tested in two phase II trials suggested promising antitumor activity in patients with MCL  and Waldenström disease (WD)  (Table 4).
Lenalidomide, whose immunomodulatory properties seem to be greater than those of thalidomide, showed interesting single-agent activity in hematological malignancies beside CLL [64, 65] myeloma including 5q-myelodysplastic syndrome (MDS) , and more recently NHL [67–69]. In 5q-MDS, 76% of patients treated with lenalidomide had a reduced need for transfusions, 67% became transfusion independent and 73% had a cytogenetic response . Additionally, as many as a quarter of MDS patients without 5q deletions can also respond to lenalidomide . In relapsed or refractory CLL, ORR to lenalidomide ranged between 32% and 47%, including 7–9% of CR [64, 65]. It should be noted that 50–60% of patients experienced tumor flare reactions [64, 65, 71]. In relapsed or refractory Hodgkin’s lymphoma, lenalidomide also gave encouraging results in a small cohort of 12 patients with 50% ORR including one long-lasting (>24 months) CR . Tumor flare reactions have also been reported in these patients . In B-cell NHL, ORR was 23% (7% CR) and 35% (12% CR) in relapsed or refractory indolent  and aggressive  NHL patients, respectively. In MCL, a phase II trial also showed efficacy of lenalidomide with 53% OR, including 20% CR, in 15 patients with relapsed or refractory diseases . In 24 patients with T-cell lymphoma, lenalidomide also demonstrated single-agent activity with 30% ORR .
Several trials have also studied the combination of lenalidomide with rituximab for the treatment of lymphoma (Table 4). Indeed, preclinical data indicate that lenalidomide can enhance NK cell-mediated ADCC of rituximab . Although a study suggested that lenalidomide could decrease CD20 expression on CLL cells , Hernandez-Ilizaliturri et al.  demonstrated that IMiD molecules enhanced the antitumor activity of rituximab in vivo using a xenograft mouse lymphoma model. These observations prompted several groups to test lenalidomide in combination with rituximab in patients with B-cell lymphoma. In patients with WD, the combination of rituximab with lenalidomide seemed somewhat less efficient than the combination with thalidomide showing lower response rates (50% vs. 72% CR and 25% vs. 64% MR, respectively) and shorter response duration (17.1 vs. 34.8 months median PFS, respectively) [63, 78]. Additionally, it should be noted that these two regimens had different toxicity profiles: development of acute (i.e., within 2 weeks) anemia with lenalidomide and peripheral neuropathy with thalidomide. Among 36 evaluable patients with relapsed or refractory MCL, 53% experienced an objective response and 31% achieved a CR with the combination of rituximab and lenalidomide . In patients with relapsed or refractory indolent NHL, 75% of 14 evaluable patients achieved an objective response, including 4/7 patients (57%) with rituximab-refractory disease . Finally, the ORR appeared to be particularly high (85%) in relapsed/refractory FL patients, including 38% CR. These promising results were further supported by the study by Folwer et al.  who administered the same combination of rituximab and lenalidomide to 30 untreated indolent B-cell NHL patients. The ORR was 86%, including 79% CR, and nearly all follicular lymphoma patients 16/17 (94%) achieved a CR. These impressive results should be tempered by the fact that patients with low tumor burden were enrolled in the study.
Most of these trials were not designed to evaluate the clinical benefit of the combination of IMiDs with rituximab over rituximab alone (low numbers of patients, no randomization, no comparison to rituximab alone arm). Therefore, efficacy can only be inferred relative to historical controls. As a landmark, one may keep in mind that studies evaluating rituximab 375 mg/m2 once weekly for 4 weeks in patients with indolent lymphoma produced ORR of 40–60% in the relapse/refractory setting and 50–70% in the first-line setting . For instance, in the pivotal study which also included rituximab-naïve patients, McLaughlin et al.  reported a 48% ORR for refractory indolent NHL patients (60% in the FL subgroup) with only 6% CR. The median PFS was 13 months. In light of these results, combinations of rituximab with IMiDs appear promising. However, these results will have to be confirmed in large randomized trials.
Cytokines regulate the innate and adaptive immune system and could be used to enhance anti-cancer immunity. Several studies have also used cytokines to stimulate effector cell function implicated in ADCC to augment the response to mAb therapy. These include IL-2, IL-12, IL-21, IFN-α, G-CSF, and GM-CSF.
Interleukin 2 is normally produced by T cells and enhances NK cell proliferation and cytotoxicity. Early studies of high dose IL-2 resulted in tumor responses, including in patients with NHL . In preclinical models, IL-2 was capable of increasing rituximab-induced ADCC and synergized with rituximab anti-lymphoma activity . Several clinical trials have tested the combination of IL-2 with rituximab [86–88] (Table 4). Although the ORR did not appear particularly higher than response rates expected with rituximab alone, prolonged response durations were observed in some patients.
Interleukin 12 receptor is predominantly expressed on T cells, NK cells, and DCs. IL-12 is produced by DCs, monocytes, and macrophages. This cytokine is known to play a pivotal role in enhancing the cytotoxicity of T and NK cells and in promoting ADCC . In a phase II clinical trial for patients with relapsed and refractory NHL and Hodgkin’s disease (HD), IL-12 monotherapy induced objective responses in six of 29 NHL patients (21%), including two CR (7%) . None of the ten HD patients responded. Treatment was well tolerated and the most common toxicity was flu-like symptoms. In a phase II trial, Ansell et al.  evaluated the combination of IL-12 with rituximab in 58 patients with relapsed B-cell NHL. Administration of IL-12 was either concomitant or sequential relative to rituximab infusion. Overall, only modest activity was seen in both arms with an ORR of 44%, including 16% CR.
Interleukin 21, a recently described common γ-chain cytokine, can induce maturation and enhance cytotoxicity of NK and CD8+ T cells. In preclinical models, exogenous IL-21 has antitumor effects via immunological mechanisms and can also have direct antitumor effects against lymphoid malignancies (CLL, NHL) . Roda et al.  also demonstrated in a murine model that IL-21 could enhance ADCC. A phase I/II trial tested IL-21 with rituximab in patients with relapsed indolent B-cell NHL . Objective responses were seen in five of the 15 patients (33%), including two CR.
Interferons possess antitumor, antiviral, and immunomodulating activity. Recombinant IFN-α has shown single-agent clinical activity in relapsed patients with indolent-lymphoma . Additionally, administration of IFN-α may enhance rituximab’s efficacy by increasing CD20 expression on tumor B cells  and by stimulating effector cells implicated in ADCC such as NK cells . Several studies evaluated the benefit of combining IFN-α with rituximab [95–97] (Table 4). Compared to historical studies, the combination seemed to give equal or higher response rates. But above all, the combination was accompanied by a significantly longer PFS . Finally, one study compared in a randomized manner, rituximab alone versus rituximab plus a short course of IFN-α in patients achieving either a PR or a minor response after rituximab alone . This study showed that the addition of IFN-α increased the proportion of CR (48% vs. 22%) and prolonged the duration of response (72% vs. 50% PFS at 2 years).
5.5 G-CSF and GM-CSF
G-CSF is another cytokine that enhances the activity of cytotoxic effector cells—specifically neutrophils—and potentiates neutrophil killing of tumor cells by ADCC. The combination of rituximab with G-CSF was evaluated in a single phase II study of 26 relapsed indolent lymphoma patients  (Table 4). Overall response rates were similar to that reported for rituximab monotherapy. However, the combination resulted in a longer duration of response than normally seen with rituximab monotherapy.
GM-CSF is a hematopoietic cytokine that can increase granulocyte proliferation and phagocytosis, promotes macrophage/monocyte ADCC, and enhances monocyte differentiation into DCs . Given the expression of FcRIIIa and FcRIIa by granulocytes and monocytes, their stimulation with GM-CSF may enhance rituximab-mediated ADCC and phagocytosis. Therefore, several studies investigated the benefit of combining rituximab with GM-CSF (Table 4). In CLL patients, a study by Ferrajoli et al.  showed an ORR of 65%, including 9% CR, which compared favorably to historical controls for both untreated (ORR 47%, CR 4%)  and previously treated (ORR 25–45%) [101, 102] patients. In NHL, two studies also suggested that GM-CSF enhances the efficacy of rituximab with an increase in CR rate over rituximab monotherapy in patients with relapsed FL patients [99, 103].
Recent therapeutic developments offer new opportunities to enhance anti-cancer immunity. Immunomodulating agents such as immunostimulatory mAbs, IMiDs, TLR agonists, and cytokines are starting to be evaluated in clinical trials. Beside single-agent activity, some of these immunomodulatory agents also showed promising results in combination with tumor-directed mAbs such as rituximab. Future progress may therefore be achieved by targeting simultaneously both the tumor and its immune environment. Randomized clinical trials are now needed to prospectively evaluate the benefit of these immunomodulating agents in the treatment of hematological malignancies.