The clinical pharmacology of therapeutic monoclonal antibodies in the treatment of malignancy; have the magic bullets arrived?

Abstract

Monoclonal antibodies (Mabs) are proteins in the immunoglobulin family that bind to specific protein epitope targets on cancer and stromal cells, allowing them to be successfully exploited as therapeutic agents. The prototype Mabs were produced from fusion of mouse B lymphocytes and mouse myeloma cells and were entirely murine in sequence. Subsequent advances in technology have allowed for humanized Mabs, which have different pharmacokinetic properties than murine Mabs in humans. Mabs antitumour activity is mediated through direct interaction with specific target molecules, deployment of immune cytotoxic pathways, or through chaperoning cytotoxic agents to tumour. Mabs are typically administered intravenously, are generally well tolerated and can have powerful anticancer activity. Humanized Mabs have a t_1/2 in human sera of 2–3 weeks, which determines the frequency of administration. At present, nine clinically approved Mabs are used in the treatment of human cancer, and many others are in clinical trials. We discuss the pharmacology, clinical indications, and toxicity of the currently available anticancer Mabs in this review.

Introduction

Cancer therapeutic agents take advantage of differences between cancer and normal cells or tissues. Antibodies, through being highly specific for a particular molecule epitope, can effectively target cancer cells and thus be used for therapeutic purposes. The concept of modulating the body's own defence mechanisms to treat disease, including cancer, has been with us since the late 19th century, when the specificity of the immune response was recognized [1, 2]. Observations by Emil von Behring in the late 1800s that guinea pigs could be protected against diphtheria toxin when treated with sera from a different diphtheria toxin-immunized guinea pig provided the first evidence of a therapeutic molecule in the blood, and led the way to our understanding antibodies [2]. The recognition of target specificity of antibodies led Paul Ehrlich to propose the concept of the ‘magic bullet’ at the beginning of the 20th century [3]. Since that time, dramatic successes have helped propel receptor-specific monoclonal antibodies (Mabs) from the laboratory bench into the clinic.

In 1975 Kohler and Milstein introduced the hybridoma, a fusion of mouse myeloma and spleen cells, as a means of large-scale production of murine Mabs [4]. Initially, this method of murine Mab production provided antigen-specific reagents for laboratory studies and consequently clinical tools for assessment of tissue histopathology and serum markers of disease. Subsequently, Mabs have been developed for in vivo therapeutic and diagnostic imaging use.

Antibody structure and function

Endogenous antibodies are immunoglobulins (Ig) synthesized by B lymphocytes. Each B-lymphocyte clone produces a unique and specific immunoglobulin. Antibodies have two separate functions: (i) to bind specific antigen and (ii) to recruit mediators of the immune stem, including complement and effector cells.

Antibodies are proteins comprising four polypeptides with molecular weights between 150–900 kDa. The polypeptide chains contain two identical heavy chains (α, δ, ε, γ, μ) and two identical light chains (λ, κ) that join to form heterodimers linked by disulphide bonds to form a three-dimensional ‘Y’-shaped protein. The two outstretched arms of the ‘Y’, known as the ‘fragment antigen binding’ or Fab portion, are responsible for recognizing and binding specific antigen. The Fab is comprised of a constant region, a variable region and a hypervariable region that enable the antibody to bind to specific antigen epitope. The base of the ‘Y’ is known as the Fc portion, which mediates the physiological functions of the antibody such as triggering antibody-dependent cell-mediated cytotoxicity (ADCC) through Fc receptor on effector cells as well as providing the site for complement binding and complement-mediated killing [5] (Figure 2). There are five antibody classes: IgG, IgA, IgM, IgD and IgE. IgG (molecular weight 150 kDa) makes up approximately 70% of the antibody pool in humans and serves as the prototypical antibody. Therapeutic monoclonal antibodies are typically of the IgG type. IgG antibodies can then be divided into four subclasses, IgG1–IgG4. IgG1–IgG3 are the most active in antibody-dependent cellular toxicity [6].

Figure 2.

Antibody and target cell interaction

Monoclonal Abs

The first Mabs, derived from mice, have several short-comings when used in vivo in humans for therapeutic or diagnostic purposes. Patients treated with murine Mabs handle this construct as a foreign protein and develop a brisk human antimouse antibody (HAMA) response. HAMA will cause rapid clearance of the Mab, poor tumour penetration, as well as hypersensitivity reactions. In addition, Mabs with a murine Fc portion have limited ability to initiate antibody dependent cellular cytotoxicity in human subjects.

By integrating components of human immunoglobulin into murine antibodies, new molecules with improved ability to trigger in vivo immune pathways in humans and be administered on a repeating schedule have been developed. These recent humanized Mab constructs have different pharmacokinetic properties compared with murine Mabs in humans. Chimeric Mabs are 65–90% human protein and fuse the murine antibody variable region with a human IgG1 constant region, which allows for functional complement activation and ADCC in humans [7, 8]. Chimeric antibodies will still induce HAMA responses. Partially humanized and deimmunized Mabs, variations of chimeric Mabs, are 95% human protein and are composed of a few critical residues involved in the antigen binding site from the murine antibody, or modified murine variable domains containing non-immunogenic amino acid sequences, respectively. To prevent any HAMA response, fully humanized Mabs containing only human protein sequences have been developed from mice that have had human immunoglobulin genes placed in their genome.

To denote the different constructs of Mab, the suffixes umab (e.g. panitumumab), momab (e.g. tositumomab), ximab (e.g. cetuximab) and zumab (e.g. trastuzumab) are used (Figure 1).

Figure 1.

Composition of various types of monoclonal antibodies and associated suffix. Purple denotes human component orange murine component

In addition, through chemical and recombinant technologies, unique molecules have been developed from antibody components. Examples include bispecific antibodies, Fab fragments, Fsc (single chain) as well as others, which have potential pharmacodynamic advantages and disadvantages over Mabs. Few of these molecules are currently Food and Drug Administration (FDA)-approved for clinical use and are beyond the scope of this review.

Therapeutic Mabs may be divided into three main classes based upon their mechanism of action (Figure 2): (i) Mabs as directed targeted therapy: these Mabs either block or stimulate a particular cell membrane molecule (e.g. growth factor signal receptor) or ligand [vascular endothelial growth factor (VEGF)] and thereby inhibit tumour growth or activate effector cells; (ii) cytotoxicity by chaperoning cytotoxic molecules (immunoconjugates): these Mabs are conjugated to various cytotoxic molecules/atoms including chemotherapy or radio isotopes such as ⁹⁰yttrium, which is in clinical use, cellular toxins such as diphtheria toxin or biological agents such as interferon (IFN); (iii) modulating an immunological mechanism: in this case, Mabs exert their cytotoxic effects by ADCC or complement-dependent cytotoxicity. These Mabs may also have some non- immunological mechanisms of action, including induction of target cell apoptosis.

General clinical pharmacology of Mabs

The IgG framework, used for the majority of therapeutic antibodies, has a molecular weight of 150 kDa and has two identical antigen binding sites. Although there is a paucity of pharmacokinetic (PK) and pharmacodynamic (PD) studies in humans, some general observations have been reported [9].

Pharmacokinetics

Therapeutic antibodies can be administered via several different routes: subcutaneously, intramuscularly and intravenously. Antibody given by intramuscular (i.m.) or subcutaneous (s.c.) routes is primarily taken up from the interstitial fluid by the lymphatic channels via convection or passive movement with lymph fluid. Time to maximum plasma concentration ranges from 1–8 days and differs from antibody to antibody. Due to tissue degradation of antibody, between 50 and 100% of the Mab dose is available for absorption via these routes. The intravenous (i.v.) route, in general, is preferred because of the 100% bioavailability. The development of anti-Mab is dependent mostly on the construct for the Mab (e.g. humanized vs. chimeric), but may also be influenced by the route [10]. I.v. administration typically requires infusion in a clinic or hospital and is thus more cumbersome and less convenient than i.m. or s.c. routes, which may be administered easily in the home setting. The i.v. route can also incur toxicity related to the rate of infusion.

Once in the systemic circulation, Mabs enter the extravascular compartment (interstitial fluid, tissue) passively, primarily through convection, and via receptor-mediated pathways. Convection is driven by hydrostatic pressure, osmotic pressure, endothelial pore size and vessel tortuosity. Thus, the interstitial pressure of tumours can affect uptake. Interstitial pressure is typically elevated in epithelial cancers and less so in haematological malignancies. Tissue retention of Mab is influenced primarily by binding and affinity to its target, as the physical forces for tissue elimination are greater than tissue uptake.

The dominant route for elimination of antibodies is via uptake and catabolism by the reticuloendothelial system and target tissue. The binding of IgG antibody to the Fc-receptor of the neonate (FcRn) is a major component of IgG elimination. The FcRn, initially termed the Brambell receptor, was discovered following the observation that IgG gastrointestinal (GI) degradation does not occur in neonates, and is responsible for transport of IgG across the intestinal mucosa in newborns [11–13]. FcRn has been described in many tissues, including kidney, liver and lung, and is also important in tissue distribution of IgG. FcRn is species specific, and human FcRn has no affinity for murine Mabs, but does for humanized Mabs. For IgG, the plasma half-life is a function of increasing serum IgG, reflecting the role of FcRn as an IgG sink and a means of protection of antibody catabolism. The FcRn is a saturable system. Thus, with doses of 15 mg kg⁻¹ IgG, the mass of systemic IgG may increase by only 1–2% and thus be unlikely to alter t_1/2. Antibody doses of 1–2 gm kg⁻¹ would be required to alter half-life. By way of varying amounts of human and mouse components, different Mab constructs behave differently when given to humans. Chimeric antibodies typically have a half-life of 4–15 days, humanized from 3 to 24 days, and recombinant human 11–24 days [14]. HAMA response develops 7–10 days following exposure to murine antibody in humans and plays a more important role in elimination [9].

Safety overview of Mabs

In general, Mabs are a well-tolerated therapy. Mabs toxicity profile is related to the specificity of the target molecule and that molecule's function, as well as to the Mab construct and isotype. The majority of toxicities observed with therapeutic Mab are related to the target antigen. The most common nontargeted toxicity among nonconjugated therapeutic Mabs is a hypersensitivity reaction, which can be modified in part by the rate of infusion [15].

Specific Mabs in clinical use

A summary of the FDA-approved, clinically available Mabs in use to treat human malignancies is shown in Table 1.

Table 1.

List of FDA-approved therapeutic monoclonal antibodies

Generic/trade name	Target antigen	Antibody type	Approved use	Proposed mechanism of action
Alemtuzumab	CD52	Humanized IgG1κ	B-cell chronic lymphocytic leukaemia in patients who have been treated with alkylating agents and have failed fludarabine therapy	Antibody-dependent lysis of leukaemic cells
Bevacizumab	VEGF	Humanized IgG1	Metastatic carcinoma of the colon or rectum first-line treatment of patients with unresectable, locally advanced, recurrent or metastatic nonsquamous, nonsmall cell lung cancer, metastatic breast cancer	Binds to all active forms of VEGF and inhibits angiogenesis
Cetuximab	EGFR	Chimeric IgG1	EGFR expressing, metastatic, irinotecan-refractory colorectal carcinoma, locally advanced head and neck cancer	Competitively inhibits ligand binding leading apoptosis. Decreases autocrine production of growth factors
Gemtuzumab ozogamicin	CD33	Humanized IgG4κ	CD33+ AML in first relapse, >60, not candidates for cytotoxic chemotherapy	Calicheamicin released in lysosomes, binds DNA resulting in double strand breaks and subsequent cell death
90Y-ibritumomab tiuxetan	CD20	Murine IgG1κ	Relapsed or refractory low-grade, follicular, or transformed B-cell non-Hodgkin's lymphoma, including patients with rituximab-refractory follicular non-Hodgkin's lymphoma.	Beta emissions induce cellular damage through formation of free radicals. Naked antibody induces apoptosis
Panitumumab	EGFR	Humanized IgG2a	EGFR expressing metastatic colorectal cancer	Competitively inhibits ligand binding leading apoptosis
Rituximab	CD20	Chimeric IgG1κ	First treatment for follicular lymphoma Refractory indolent CD20+ B-cell non-Hodgkin's lymphoma First-line treatment for diffuse large B-cell lymphoma	ADCC, CDC, induction of apoptosis
Trastuzumab	HER2 (c-erbB2) receptor	Humanized IgG1κ	HER2 overexpressing metastatic breast cancer	Downregulation of HER2, inhibition of intercellular signalling, induction of apoptosis, ADCC
131I -Tositumomab	CD20	Murine IgG2aλ	CD20+ follicular, non-Hodgkin's lymphoma, with and without transformation, whose disease is refractory to rituximab and has relapsed following chemotherapy	Ionizing radiation from the I131 apoptosis, CDC, ADCC

ADCC, antibody-mediated cellular cytotoxicity; CDC, complement-dependent cytotoxicity; EGFR, epidermal growth factor receptor; VEGF, vascular endothelial growth factor; FDA, Food and Drug Administration; AML, acute myeloid leukaemia.

Mabs as target therapy

Trastuzumab is a humanized IgG1κ monoclonal antibody that binds the extracellular domain of the HER2 (c-erbB2) receptor, a member of the epidermal growth factor receptor (EGFR) family that plays a pivotal role in growth, differentiation and cell survival. HER2 is overexpressed in 25–30% of breast cancers and is associated with a poor prognosis [15]. Blocking HER2 with trastuzumab can arrest the breast cancer cell in G₁ phase of the cell cycle. Clinical trials in breast cancer patients have demonstrated the effectiveness of trastuzumab only in patients with tumours that overexpress HER2 protein [16–20]. Although the optimal method for assessing HER2 protein levels is still unknown, overexpression is typically determined by immunohistochemistry or fluorescence in situ hybridization (FISH). Significant reduction of mortality is seen in patients with overexpression of HER2 protein and treated with trastuzumab therapy [21, 22]. Recent data from a randomized clinical trial (CALGB 150002) found an improved response rate in metastatic breast cancer patients with low levels of HER2 expression as defined by FISH negative, polysomy 17 positive [23].

In patients with metastatic breast cancer, trastuzumab is given as a 4 mg kg⁻¹ loading dose followed by 2 mg kg⁻¹ weekly until disease progression [17]. Objective clinical response rate to single-agent trastuzumab therapy in chemotherapy failures is a modest 15% (4% complete response). However, when used in combination with chemotherapy (anthracycline plus cyclophosphamide or paclitaxel), trastuzumab significantly improves the objective response and survival of patients with metastatic breast cancer compared with chemotherapy alone [18]. The median time to progression was 7.4 months in the trastuzumab group compared with 4.6 months (P < 0.001) in the chemotherapy only group. The combination group also showed an overall response of 50% compared with 32% (P < 0.001) and a longer duration of response of 9.1 months compared with 6.1 months (P < 0.001). Perhaps most striking was an overall improvement in survival of 25.1 months for trastuzumab plus chemotherapy compared with 20.3 months (P = 0.046) for chemotherapy alone, corresponding to a 20% reduction in risk of death in patients treated with trastuzumab plus chemotherapy [18].

Subsequent studies have evaluated trastuzumab in addition to adjuvant multiagent chemotherapy in HER2+ early-stage breast cancer patients [18]. In the herceptin adjuvant trial (HERA), trastuzumab was given every 3 weeks for 1 or 2 years after adjuvant chemotherapy and was compared with surveillance following adjuvant chemotherapy [19]. The 2-year disease-free survival was 86% in the trastuzumab group compared with 77% in the observation group, with a corresponding hazard ratio (HR) of 0.54 (P < 0.0001) [19]. NSABP B-31 and NCCTG N9831 were two adjuvant studies that confirmed the benefit of trastuzumab in this setting [20]. The 3-year disease-free survival was 87% in the trastuzumab group compared with 75% in the control (P < 0.0001), with a corresponding HR of 0.48 [20].

Generally, trastuzumab is well tolerated; adverse effects include nontarget-related chills and/or fever, and rarely hypotension during the initial infusion. Slowing the rate helps to ameliorate these symptoms. In randomized clinical trails, infections were seen in a slightly higher frequency in the trastuzumab group, 47% vs. 29% in the chemotherapy only group [18]. Cardiac dysfunction presenting as congestive heart failure is seen more commonly when trastuzumab is administered with an anthracyclin-based chemotherapy regimen and less so with cyclophosphamide or paclitaxel [18]. Overall, trastuzumab is associated with a threefold increase in grade III or IV cardiac toxicity, and careful monitoring of ejection fraction using either multiple gated acquisition scan or echocardiogram is warranted [21].

In initial PK studies using a loading dose of 4 mg kg⁻¹ followed by a weekly maintenance dose of 2 mg kg⁻¹, the mean half-life suggested a terminal t_1/2 of 5.8 days, ranging from 1 to 32 days. A subsequent PK study using a 8 mg kg⁻¹ loading dose and a 6 mg kg⁻¹ dose every 3 weeks has found a terminal t_1/2 of 28.5, which justifies an every 3 weeks schedule [24]. Trastuzumab reached steady state at between 16 and 32 weeks, with a mean peak concentration of 79 µg ml⁻¹ and trough of 79 µg ml⁻¹. The disposition of trastuzumab is not altered by age or serum creatinine. The volume of distribution is approximately equal to serum volume, 44 ml kg⁻¹.

Cetuximab is a human/mouse chimeric IgG1κ Mab directed against the ligand binding site of the EGFR and competitively inhibits EGF binding, leading to cell growth inhibition and apoptosis. EGFR is a transmembrane receptor tyrosine kinase expressed on normal tissue and is upregulated by transforming growth factor-alpha (TGF-α) and EGF as well as radiation. Downregulation of TGF-α, amphiregulin, angiogenic factors such as VEGF, fibroblast growth factor and interleukin-8 occurs when endogenous ligand binds to EGFR [25–27]. EGFR is overexpressed in many tumour types, including 25–80% of colorectal carcinomas, where it is associated with advanced disease and poor prognosis [25, 28].

Cetuximab is approved for use in combination with irinotecan for patients with EGFR expressing colorectal carcinoma that is refractory to irinotecan and as a single agent in patients who are intolerant to irinotecan [25]. Cetuximab is given intravenously at 400 mg m⁻² loading dose followed by 250 mg m⁻² weekly until disease progression or unacceptable toxicity. Response rates in colorectal cancer to cetuximab and irinotecan in irinotecan-refractory metastatic colorectal cancer were doubled (22.9%) compared with cetuximab alone (10.8%) (P = 0.007) [29]. Time to progression improved to 4.1 months vs. 1.5 months (P < 0.001), whereas duration of response and median overall survival were not different. The best predictor of response was the development of acneiform skin rash [29]. The level of EGFR expression did not predict for response (positive staining for expression of EGFR was defined as >1% partial or complete membranous staining of any intensity) [30].

Cetuximab has also been approved in the treatment of head and neck cancer [31]. The justification for cetuximab use in head and neck cancer is both the overexpression of EGFR and the induction of EGFR by radiation. In a randomized study of stage III and IV nonmetastatic head and neck squamous cell cancer, patients received either radiation or radiation with weekly cetuximab. Locoregional control, progression-free survival (PFS) and overall survival were significantly improved in the combination arm [31]. The addition of cetuximab to radiation did not affect the radiation-related toxicity.

In general, cetuximab was well tolerated. Nontarget-specific toxicities include anaphylactic reactions, which occur in about 1% of patients and require drug discontinuation. Skin toxicity is related to interaction of cetuximab with its EGFR receptor and was the most common side-effect, including an acne-like rash that occurred in approximately 80% of patients. The hypomagnesaemia appears also to be related to the interaction with the target antigen, as it has been described in other EGFR Mab targeted therapy and is seen in approximately one-quarter of patients and may be severe and lead to symptoms. Measuring and replacing magnesium should be pursued if fatigue or hypocalcaemia are encountered during cetuximab therapy [32].

The plasma/serum t_1/2 of cetuximab was approximately 79–129 h, and its mean volume of distribution was 45–62 ml kg⁻¹, approximately equal to plasma volume. Clearance of cetuximab was thought to occur through binding and subsequent internalization of EGFR in various tissues, including liver and skin. Cetuximab clearance decreased with increasing dose, suggesting saturation of the metabolic pathways and nonlinear PK at higher doses [25].

Panitumumab is a fully human IgG2a Mab directed against the extracellular domain of EGFR. It competitively inhibits EGF and tissue growth factor-alpha binding to EGFR and leads to internalization of the receptor. Its mechanism of action is similar to that of cetuximab, although it has a higher affinity for the receptor (KD = 5 × 10⁻¹¹ M) [33].

Panitumumab is approved for monotherapy in colorectal cancer patients who have already failed prior chemotherapy and whose tumours show 1+ EGFR expression by immunohistochemistry in >1% of the tumour cells. Panitumumab is given intravenously at 6 mg kg⁻¹ over 1–1_1/2 h every 2 weeks [34].

A Phase III randomized study comparing panitumumab (6 mg kg⁻¹ every 2 weeks) with best supportive care (BSC) to BSC alone in metastatic colorectal cancer patients who had failed prior chemotherapy showed a 10% partial response (PR) rate for panitumumab compared with 0% in the BSC arm [34]. To determine the survival effect of panitumumab, PFS analysis was done with and without including responding patients. Even with panitumumab-responding patients removed, there was a significant difference in the treatment vs. BSC group (HR = 0.63, 95% confidence interval 0.52, 0.77; P < 0.0001). A Phase II study designed to evaluate the outcome of metastatic colorectal cancer patients with different EGFR staining characteristics by immunohistochemistry (≥2+ in ≥10% of cancer or ≥2+ in <10% of cancer and ≥10% of cancer with any staining) found no difference in partial and stable disease responses [35].

As with cetuximab, panitumumab causes skin toxicity as the most common side-effect, including conjunctivitis. For grade 1 or 2 skin toxicity occurring during the infusion, the infusion rate can be reduced, and for more severe immediate reactions the drug should be stopped. For persistent grade 3 or 4 skin toxicity that improves to grade 2 or less, the dose of panitumumab can be reduced by 50% and re-escalated if the skin toxicity does not recur. Hypomagnesaemia has also been described in nearly 40% of patients and may be severe and lead to symptoms. Other common side-effects include diarrhoea, cough and fatigue. There have been isolated reports of idiopathic pulmonary fibrosis in patients on panitumumab [36].

Panitumumab has a nonlinear PK profile with increases in the AUC and decreases in clearance that are not proportional to dose increases [37]. The plasma/serum t_1/2 of panitumumab was approximately 180 h, with steady state occurring after three doses of 6 mg kg⁻¹ every 2 weeks.

Bevacizumab is a humanized monoclonal IgG1 antibody that binds and neutralizes all biologically active forms of VEGF-A, preventing it from interacting with its receptors on the surface of endothelial cells. Bevacizumab does not neutralize VEGF-B or VEFG-C. VEGF is a proangiogenic glycoprotein produced by normal and neoplastic cells and is involved with regulation of both normal and abnormal angiogenesis and tissue proliferation. Overexpression has been observed in several tumour types, including colorectal and renal cancers, and is associated with invasiveness, metastasis, recurrence and prognosis. In addition to its antiangiogenic properties, bevacizumab may potentiate both cytotoxic chemotherapy and radiation therapy by enhancing delivery through alterations of tumour vasculature and interstitial pressure [38–40].

Bevacizumab, in combination with 5-fluorouracil (5FU)-based chemotherapy, is used as first-line therapy for metastatic colorectal carcinoma. In this setting, 5 mg kg⁻¹ bevacizumab plus chemotherapy was better than 10 mg kg⁻¹[38, 39]. The objective response of bevacizumab plus 5FU was 40%, with a 70% improvement in PFS (9 months) and a benefit in median overall survival of 45% (10.6 months) over chemotherapy alone [38, 39]. Bevacizumab in combination with FOLFOX4 (5FU, leucovorin and oxaliplatin) has also been approved for the second-line treatment of metastatic carcinoma of the colon or rectum [41]. Median survival duration and PFS were significantly improved for the bevacizumab arm, which was associated with higher incidence of grade 3 hypertension.

At the end of 2007, the Eastern Cooperative Oncology Phase III breast cancer study of paclitaxel alone or with bevacizumab in 722 untreated metastatic breast cancer patients was published and led to the approval by the FDA of bevacizumab for breast cancer in February 2008 [42]. Response rate increased by 50% (from 21 to 37%) and PFS was doubled (11.8 vs. 5.9 months), although there was no overall survival difference.

Bevacizumab is also used in the treatment of patients with nonsmall cell lung or renal cell cancers [43, 44]. When bevacizumab is combined with carboplatin and paclitaxel for stage IIIB and IV nonsmall cell lung cancer, there is an improvement in objective response rate, PFS (7.4 months) and in median overall survival (17.7 months) [43]. Life-threatening haemoptysis or haematemesis occurred in a small number of patients with lung cancer, and this was noted in patients with centrally located tumours, cavitary or necrotic lesions and squamous cell histology [43]. In renal cell carcinoma (RCC), a large randomized placebo controlled Phase II study has reported a 92% improvement in median PFS benefit in the 10 mg kg⁻¹ bevacizumab group (4.8 months) [44]. The results of Phase III studies with bevacizumab and IFN-alfa have confirmed the activity of the combination over IFN-alfa alone in metastatic RCC patients [45].

Bevacizumab is well tolerated, with adverse effects related to its effect on the target molecule for the most part, including bleeding, clotting, GI perforation, hypertension and proteinuria. In August 2004 the US FDA warned that bevacizumab when given in combination with chemotherapy is associated with a twofold increased risk of arterial thromboembolic events (4.4 vs. 1.9%). These are mostly in the form of cerebrovascular and cardiovascular events. There is a marginal correlation with age >65 years and the increased risk of thromboembolic events. Bevacizumab-associated GI perforation typically presents with abdominal pain, nausea and vomiting, and has been reported in approximately 1.5% of patients. Bevacizumab-associated GI perforation is associated with tumours still in place in the GI tract and is not associated with history of peptic ulcer disease or diverticulitis [46]. GI perforation is an absolute contraindication for continuing bevacizumab therapy. More recently, the syndrome of reversible posterior leukoencephalopathy (headache, seizure, feeling tired, confusion, vision problems, and elevated blood pressure) has been seen in 0.1% of patients receiving bevacizumab.

The plasma/serum t_1/2 of bevacizumab is 20 days (range 11–50) [47]. Clearance varied by weight, gender and tumour burden. When corrected for body weight, men had a clearance of 0.26 l day⁻¹ and women a clearance of 0.21 l day⁻¹. Patients with higher tumour burden had a higher clearance compared with lower tumour burden with 0.25 l day⁻¹vs. 0.20 l day⁻¹, respectively. There was no evidence to suggest reduced antitumour efficacy in patients with higher clearance.

Immunological mediated cytotoxicity agents

Rituximab is a mouse/human chimeric anti-CD20 IgG1κ Mab that targets the CD20 antigen expressed on the surface of >90% of malignant and normal B lymphocytes. CD2O is expressed at lower density on chronic lymphocytic leukaemia cells. The CD20 antigen is an attractive target, in that it is not expressed on stem cells, does not circulate in the plasma, is not shed from the cell surface after binding and is not internalized or downregulated [48].

Although one of the most well-studied Mabs, its exact mechanism of action is still not completely understood. The proposed mechanisms include ADCC, complement-mediated cytotoxicity, as well as growth inhibition, cell cycle alteration and apoptosis via direct binding to CD20 [49, 50]. In addition, in vitro data suggest that it may sensitize lymphomas to the action of cyclophosphamide, doxorubicin, vincristine and prednisone chemotherapy [49].

Approximately 50% of patients with low-grade or follicular CD20+ lymphoma will have an objective response (6% complete response rate) [50]. The median time to progression for responders and duration of response were approximately 13 and 12 months, respectively.

Rituximab has been studied in several other haematological disorders. In diffuse large B-cell lymphoma, rituximab combined with chemotherapy significantly improved the complete response rate (CR = 76%), as well as event-free and overall survival [51]. In addition, rituximab and chemotherapy improved the CR rate by 3.9-fold (CR = 34%) and the objective response rate by 25% (OR = 94%) in mantle cell lymphoma, although no overall survival advantage was observed [52]. Additional studies have suggested a role of rituximab in the treatment of chronic lymphocytic leukaemia [53, 54] as well as in refractory idiopathic thrombocytopenic purpura, thrombotic thrombocytopenic purpura and autoimmune haemolytic anaemia [55, 56].

Maintenance rituximab has been studied in several randomized trials, both when used as a single agent and in combination with chemotherapy [57–60]. In a recent study of advanced-stage relapsed or refractory follicular or mantle cell lymphoma, patients received rituximab plus chemotherapy and were then randomized to receive maintenance rituximab 375 mg m⁻² weekly for 4 weeks at 3 and 9 months after achieving a PR or CR [58]. The median PFS was 17 months for the observation arm, whereas the median PFS for the maintenance arm has not been reached at 3 years of follow-up. Overall survival was improved with a trend toward significance at 3 years of 77% in maintenance arm vs. 57% in observation (P = 0.1) [58]. The optimal schedule of maintenance rituximab and which subsets of non-Hodgkin's lymphoma (NHL) patients benefit most from this treatment remain to be clearly defined.

Rituximab is well tolerated and can be administered to most patients regardless of age or performance status. Adverse events thought to mediated by the effector function of the Mab are infusion-related and commonly include fever, chills, flushing, and, less commonly, bronchospasm and hypotension. They typically occur during the first hour of infusion and last only 1–2 h. Anaemia, thrombocytopenia and neutropenia are seen in approximately 10% of patients and are mostly grade 1–2 in severity [61]. The US FDA issued an alert regarding the development of progressive multifocal leukoencephalopathy, a rare demyelinating disease caused by human polyomavirus (JC) of oligodendrocytes, which caused death in two patients on rituximab.

Rituximab is given as an i.v. infusion of 375 mg m⁻² weekly for 4–8 weeks [62]. In the initial clinical trial the mean serum t_1/2 after the first infusion was 76.3 h (range 31.5–152.6 h) and after the fourth infusion was 205.8 h (range 83.9–407 h). This wide range of half-lives may be explained by variable tumour burden as well as changes in CD20+ B cells. Metabolism and excretion are not well defined, but antibody-coated cells are thought to undergo elimination via Fc receptor binding and phagocytosis by the reticuloendothelial system. Very low concentrations of rituximab can be detected in serum 3–6 months following completion of therapy, suggesting drug accumulation and a complex PK.

Alemtuzumab is a recombinant humanized IgG1κ Mab against CD52. Human CD52 is a glycosylphosphatidylinositol-anchored antigen expressed on normal and malignant B and T lymphocytes as well as on some male genital tract epithelial cells. CD52 is not expressed on haematopoietic stem cells. Whereas the CD52 protein core is the same for haematopoietic and male genital tract epithelial cells, the N-linked glycosylated carbohydrate side-chain differs [63].

Alemtuzumab is used for the treatment of B-cell chronic lymphocytic leukaemia (B-cell CLL) after treatment failure with alkylating agents and fludarabine [64]. Alemtuzumab was titrated up to 30 mg m⁻² three times a week for up to 12 weeks. The overall response was 33%, with 2% achieving complete remission [64]. The median time to response was 1.5 months, median time to progression was 4.7 months overall, 9.5 months for responders. Median survival was 16 months overall and 32 months for responders. Alemtuzumab is being explored as first-line therapy in patients with B-cell CLL [65]. A preliminary study has shown a high overall response rate of 87%; 19% complete response was observed [65]. To confirm this observation, a Phase III study is currently underway comparing single-agent alemtuzumab with chlorambucil for first-line therapy in B-cell CLL.

The common adverse events to alemtuzumab are thought to be mediated by the effector function during infusion and include rigors, fever, nausea, vomiting, rash, pruritus, urticaria, dyspnoea, diarrhoea and hypotension. Premedication with antihistamines, paracetamol, anti-emetics, meperidine and corticosteroids is used to reduce infusion-related toxicity. Starting at a lower (10 mg) dose with escalation on subsequent days has also been useful in reducing infusion-related toxicity. Patients are at risk for opportunistic infections both during treatment and 2 months post treatment, with reactivation of cytomegalovirus being most frequent. All patients should receive antibiotic and antiviral prophylaxis [64]. Postmarketing reports have also attributed myocardial infarction, cardiac arrhythmias, adult respiratory distress syndrome, respiratory arrest and syncope associated with alemtuzumab therapy [66].

The mean t_1/2 is 11 h (2–32 h) after the first 30-mg dose and is 6 days (1–14 days) after the last 30-mg dose. At steady state the mean volume of distribution is 0.18 l kg⁻¹ (0.1–0.4 l kg⁻¹) [67]. Systemic clearance of alemtuzumab decreases with repeated administration.

Conjugated monoclonal antibodies

Rituximab therapy has been successful, but tumour cell resistance still plays a significant role in treatment failure. Mechanisms of resistance include inadequate serum/tissue antibody concentrations, limited access to tumour cells with bulky disease, defective tumour effector mechanisms, and genetic polymorphisms in the FcγRIII gene, which may lower the binding affinity of the antibody for the Fc receptor. One approach to overcome tumour resistance has been the development of conjugated antibodies. Given lymphoma's sensitivity to radiation, attaching a radionuclide to a tumour-specific antibody is a rational combination-targeting approach. This technology is termed radioimmunotherapy (RIT). The β emissions from radionuclides that are commonly used in RIT have a cytotoxic effect of approximately 100–200 cell diameters and therefore are lethal to neighbouring tumour cells that may be inaccessible to antibody or may not express the target antigen. This is termed the ‘cross-fire’ or ‘by-stander’ effect [68].

⁹⁰Y-ibritumomab tiuxetan is an immunoconjugate composed of ibritumomab, a murine IgG1κ anti-CD20 Mab, covalently bonded to the linker-chelator tiuxetan, which provides a high-affinity chelation site for the ¹¹¹indium used for imaging or ⁹⁰yttrium used for therapy. The Mab was left in its murine form in order to accelerate clearance and limit the effects of prolonged irradiation exposure. In follicular B-cell NHL refractory to rituximab treatment, ⁹⁰Y-ibritumomab tiuxetan therapy demonstrated an overall response rate of 74%, with 15% of patients achieving a complete response [69]. Median time to progression was 6.8 months and the median duration of response was 6.4 months [69].

⁹⁰Y-ibritumomab tiuxetan appears to provide an improved response rate to rituximab in patients with relapsed or refractory low-grade lymphoma, follicular lymphoma or transformed B-cell NHL, heavily pretreated low-grade follicular NHL with or without transformation [overall response rate was 80% (30% CR) in the ⁹⁰Y-ibritumomab tiuxetan arm] compared with 56% (P = 0.002; 16% CR) for the rituximab-alone arm [70]. There was substantially greater haematological toxicity for ⁹⁰Y-ibritumomab tiuxetan-treated patients in both studies.

The dose (0.4 mCi kg⁻¹ actual body weight) of ⁹⁰Y-ibritumomab tiuxetan is given only after the biodistribution of ¹¹¹In-ibritumomab tiuxetan is determined by whole-body γ-camera images to be adequate [71]. The t_1/2 of ⁹⁰Y-ibritumomab tiuxetan is 30 h. Over 7 days, a median of 7.2% of the injected activity was excreted in the urine [72].

Tositumomab and ¹³¹I-tositumomab: ¹³¹I-tositumomab is comprised of tositumomab, a murine IgG2aλ murine Mab directed against CD20 covalently linked to ¹³¹iodine. Tositumomab is used for therapy of CD20+ follicular NHL, with and without transformation, that is refractory to rituximab and has relapsed following chemotherapy. Response rates to ¹³¹I-tositumomab in low-grade or transformed low-grade chemotherapy-refractory CD20+ B-cell lymphoma were 65% and were better than additional chemotherapy [73].

Tositumomab has a 65% overall response rate in low-grade or transformed lymphoma failing rituximab therapy, with a median duration of response of 14.7 months [74]. A complete response was seen in 38% of patients [74].

The common toxicities relate to severe bone suppression. More than 70% of patients experience National Cancer Institute (NCI) Common Toxicity Criteria grade 3–4 cytopenias. Less than 10% of patients experience hypersensitivity, including life-threatening anaphylactoid reactions. Infusion-related pyrexia, rigors, hypotension dyspnoea and bronchospasm are seen in the first 48 h and the fever and rigors can be seen after as long as 14 days. Hypothyroidism related to the ¹³¹I can develop, and thyroid blocking agents are suggested.

The median clearance following administration of 485 mg of ¹³¹I-tositumomab in 110 patients with NHL was 68.2 mg h⁻¹ (range 30.2–260.8 mg h⁻¹). Those patients with high tumour burden, splenomegaly or bone marrow involvement are noted to have a faster clearance, shorter half-life and larger volume of distribution.

The overwhelming majority of the elimination of iodine ¹³¹I occurs through renal excretion. ¹³¹I decay accounts for a minority of elimination. ¹³¹I has a half-life of 8 days. Five days following the dosimetric dose, the whole body clearance was 67% of the injected dose [75, 76].

Gemtuzumab ozogamicin is a humanized IgG4κ anti-CD33 Mab covalently linked to a semisynthetic derivative of the potent cytotoxic antibiotic calicheamicin. Gemtuzumab ozogamicin binds to the CD33 receptor, resulting in complex formation and internalization. Once internalized, the calicheamicin derivative is thought to be released in the lysosomes and then binds to DNA, resulting in double strand breaks and subsequent cell death [77]. In addition, calicheamicin may cause cell death through a caspase-mediated pathway [78]. CD33 is expressed on approximately 90% of acute myeloid leukaemia myeloblasts as well as normal myeloid precursor cells. CD33 expression is downregulated with maturation of myeloid cells, which results in a low level of expression on circulating granulocytes and tissue macrophages. CD33 is not expressed on CD34+ pluripotent haematopoietic stem cells. Approximately 50% of the antibody is linked to calicheamicin with an average of 4–6 moles of calicheamicin per mole of antibody. The remaining 50% is unconjugated antibody.

Gemtuzumab ozogamicin is used in the treatment of CD33+ acute myelogenous leukaemia (AML) in first relapse, and those ≥60 years old and not considered candidates for cytotoxic chemotherapy. Gemtuzumab ozogamicin is given intravenously at a dose of 9 mg m⁻² over 2 h for up to three doses, with a minimum of 14 days and a maximum of 28 days between doses [79]. Complete response as defined by US NCI consensus criteria was seen in 16% of patients [79]. CR plus CR with incomplete platelet recovery were reported in 30% of patients. Ongoing studies are evaluating gemtuzumab ozogamicin in combination with standard induction chemotherapy regimens in both young and older adult patients with newly diagnosed AML [80–82].

Common transient and reversible adverse events related to the infusion include chills, fever and hypotension. These events occurred despite use of prophylactic paracetamol and antihistamines. Infusion-related adverse events diminished in subsequent infusions. Treatment-related adverse events include grade 3–4 neutropenia, thrombocytopenia, neutropenic fever, infection and sepsis, as well as haemorrhage including intracranial bleed and haemorrhagic death in a small percentage. Additional serious grade 3–4 toxicity includes hyperbilirubinaemia and transaminitis [79, 83, 84]. Gemtuzumab ozogamicin has been associated with increased risk for veno-occlusive disease in patients who went onto stem cell transplantation [85].

In animal studies, gemtuzumab ozogamicin undergoes hepatobiliary elimination. The pathway of elimination has not been defined in humans. The half-lives of total and unconjugated calicheamicin were approximately 45 and 100 h, respectively, after the administration of the first 9 mg m⁻² of gemtuzumab ozogamicin. Clearance of the antibody and unconjugated calicheamicin were consistently decreased with subsequent doses, leading to longer half-lives [83].

Future directions

By virtue of their antigen specificity, Mabs can affect multiple pathways that mediate immune function, tumour growth and tumour vascularity. Thus, using Mab cocktails may provide improvement in clinical response in cancer patients. One such attempt has been studied in a randomized Phase II design of cetuximab and bevacizumab alone or with irinotecan in irinotecan-refractory colorectal patients, the BOND-2 Study [86]. Both agents alone have shown benefit with chemotherapy in colorectal cancers. In this approach the Mabs are being used to block both EGF and VEGF pathways. The time to progression for the combination of cetuximab, bevacizumab and irinotecan was 7.3 months compared with 4.9 months in the cetuximab + bevacizumab arm, and better than the 4.1 months seen in the original cetuximab + irinotecan arm observed in irinotecan-resistant colorectal cancer observed previously [29, 86]. The toxicities seen in the BOND-2 study appeared acceptable, although the three patients (n = 83) had serious GI complications including perforation, gastric ulcer and rectal fissure. Other serious complications included two deaths, one from enterococcal endocarditis, one from myocardial infarction, and an additional two patients with arterial thrombotic events. These toxicities underscore the need to proceed with Mab cocktails cautiously and under controlled clinical trial designs.

With the advent of small molecules that disrupt the growth receptors’ signal transduction pathways it is now possible to use Mabs in conjunction with these agents to provide either vertical (within the same signal pathway) or horizontal (across different signal pathways) blockade. One such approach in RCC employed bevacizumab to block the VEGF pathway and erlotinib, a signal transduction inhibitor of the EGFR [87]. Blockade of EGFR will suppress the expression of VEGF. Median PFS was 11 months and the toxicity was acceptable, with one grade 4 GI bleeding noted (n = 59). In a subsequent randomized Phase II study of bevacizumab + erlotinib vs. bevacizumab alone, there was no additional benefit observed in the combination arm (n = 104) [88]. Thus, it is unclear if this approach will ultimately prove successful.

Similar approaches have been taken in other cancers (colorectal, lung, head and neck) with cetuximab and the EGFR signal transduction pathway inhibitor gefitinib [89]. This preliminary report suggests a benefit in colorectal cancer, with five of nine patients responding. Further exploration of these approaches is warranted.

Mab constructs (ipilimumab and ticilimumab) have been used to block T-lymphocyte receptor (CTLA4) responsible for negative regulation of cellular immune responses [90, 91]. Although not yet approved by the FDA, the preliminary data to date for both of these Mabs are encouraging in the treatment of metastatic melanoma patients. VEGF is another immune inhibitory molecule, and bevacizumab has been explored with other biological response modifiers such as IFN and high-dose aldesleukin [45, 92]. As new Mabs that target immune pathways are developed, combination immunotherapy strategies will be explored as cancer therapy.

Conclusions

Paul Ehrlich's first proposed the concept of a ‘magic bullet’ that would specifically seek out, target and destroy diseased cells with minimal damage to normal tissue >100 years ago. Since that time tremendous advances in science and technology, most notably the discovery of methods for production of Mabs in the 1970s and the advent of genetic engineering in 1980, have led to dramatic advances in the development and approval of several Mabs. Currently, there are Mabs approved for use in lymphoma, acute and chronic leukaemia, colorectal cancer or breast cancer. With >150 therapeutic Mabs in early-phase clinical trials and >700 clinical trials involving Mabs, they will clearly continue to have an increasing presence in the antineoplastic therapeutic armamentarium.

Specific clinical indications for these agents continue to develop. For example, bevacizumab approval now includes first-line treatment of patients with unresectable, locally advanced, recurrent or metastatic nonsquamous, nonsmall cell lung cancer. Rituximab has been approved for use as first-line therapy for follicular lymphoma as well.

Most importantly, all antibodies designed for clinical use have powerful biological effects and need to be tested initially in well-designed Phase I studies safely conducted. In a recent Phase I clinical trial of TGN1412 performed in London, six of eight healthy volunteers receiving the experimental Mab suffered cytokine storm syndrome and multisystem organ failure [93]. None of the patients died and at this time all have been discharged from the hospital; the two subjects who received placebo sustained no ill effects. TGN1412 is a humanized monoclonal superagonist of the CD28 T-cell receptor, which leads to pronounced T-cell activation and expansion. One hypothesis is that activation of T cells in vivo may have led to a cytokine storm, causing the catastrophic multisystem organ failure [93]. The agent underwent preclinical testing in vitro and in vivo in animals, including nonhuman primates, which were reportedly safe and revealed biological effect. This has raised considerable concern and awareness in first in human Phase I clinical trials of so-called ‘high-risk agents’ and the general safety and performance of new medications not yet evaluated in humans.

In addition to the study of new Mabs in other malignancies, future studies are needed to continue to optimize the schedule in maintenance therapy as well as in combination with chemotherapy, use in adjuvant setting, and the role in other relapse/progressive disease.

Have the magic bullets arrived? We can now answer this question with a resounding yes. Mabs have made and are making major contributions to the therapy of human malignancies, and as new targets and pathways of human disease are identified, the role of Mabs for treatment of human disease will only continue to grow and expand.