Therapeutic and diagnostic applications of dendrimers for cancer treatment☆

Available online at www.sciencedirect.com Advanced Drug Delivery Reviews 60 (2008) 1037 – 1055 www.elsevier.com/locate/addr Therapeutic and diagnostic applications of dendrimers for cancer treatment☆ Jesse B. Wolinsky, Mark W. Grinstaff ⁎ Departments of Biomedical Engineering and Chemistry, Boston University, Boston, Massachusetts 02215, USA Received 10 September 2007; accepted 14 February 2008 Available online 4 March 2008 Abstract Dendrimers are prepared with a level of control not attainable with most linear polymers, leading to nearly monodisperse, globular macromolecules with a large number of peripheral groups. As a consequence, dendrimers are an ideal delivery vehicle candidate for explicit study of the effects of polymer size, charge, composition, and architecture on biologically relevant properties such as lipid bilayer interactions, cytotoxicity, internalization, blood plasma retention time, biodistribution, and tumor uptake. Over the last several years, substantial progress has been made towards the use of dendrimers for therapeutic and diagnostic purposes for the treatment of cancer, including advances in the delivery of anti-neoplastic and contrast agents, neutron capture therapy, photodynamic therapy, and photothermal therapy. The focus of this review is on dendrimer developments from the last four years for oncological applications, with emphasis on distinct architectures and the biological responses these structures elicit. © 2008 Published by Elsevier B.V. Keywords: Dendrimer; Local Therapy; Nanoparticle; Cancer Treatment; Drug-conjugates; Drug Delivery Contents 1. 2. Introduction . . . . . . . . . . . . . . . . Architecture and composition. . . . . . . 2.1. Dendrimer-membrane interactions . 2.2. End groups and toxicity . . . . . . 2.3. Pharmacokinetics . . . . . . . . . 3. Drug delivery . . . . . . . . . . . . . . . 3.1. Drug-encapsulated dendrimers. . . 3.2. Dendrimer-drug conjugates . . . . 4. Targeted drug delivery . . . . . . . . . . 4.1. Folic acid . . . . . . . . . . . . . 4.2. Peptides . . . . . . . . . . . . . . 4.3. Monoclonal antibodies . . . . . . 4.4. Glycosylation . . . . . . . . . . . 5. Photodynamic therapy . . . . . . . . . . 6. Boron neutron capture therapy . . . . . . 7. Photothermal therapy . . . . . . . . . . . 8. Imaging . . . . . . . . . . . . . . . . . . 9. Conclusions. . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ☆ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1037 1038 1038 1038 1041 1042 1043 1043 1044 1046 1046 1046 1047 1047 1048 1049 1050 1051 1052 1053 1053 This review is part of the Advanced Drug Delivery Reviews theme issue on “Design and Development Strategies of Polymer Materials for Drug and Gene Delivery Applications”. ⁎ Corresponding author. Tel.: +1 617 358 3429; fax: +1 617 353 6466. E-mail address: mgrin@bu.edu (M.W. Grinstaff). 0169-409X/$ - see front matter © 2008 Published by Elsevier B.V. doi:10.1016/j.addr.2008.02.012 1038 J.B. Wolinsky, M.W. Grinstaff / Advanced Drug Delivery Reviews 60 (2008) 1037–1055 1. Introduction The emerging role of dendritic macromolecules for anticancer therapies and diagnostic imaging has highlighted the advantages of these well-defined materials as the newest class of macromolecular nano-scale delivery devices. As the relationships between dendrimer architecture, biocompatibility, retention, and delivery have become better elucidated, unique dendrimer derivatives have been synthesized for greater specificity and functionality, particularly with regards to pharmacokinetics and targeted delivery. Over the last several years, substantial progress has been made towards the use of dendrimers for therapeutic and diagnostic purposes for the treatment of cancer, including advances in the delivery of antineoplastic and contrast agents, neutron capture therapy, photodynamic therapy, and most-recently, photothermal therapy. The key properties of dendrimers that lend themselves as nano-carriers for biological applications are being identified and reviewed [1–3], and more recently, the increasing importance of these properties to dendrimer-based oncological approaches [4– 6]. As will be discussed herein, general principles are being established for designing dendrimer structures as delivery vehicles which include: 1) negatively-charged and neutral dendrimers are generally biocompatible, while positivelycharged species show varying degrees of toxicity; 2) dendrimer architecture can dramatically influence pharmacokinetics; 3) PEGylation increases water solubility and dendrimer size, and can lead to improved retention and biodistribution characteristics; 4) therapeutic agents can be internalized into the void space between the periphery and core, or covalently attached to functionalized surface groups; 5) targeting moieties bound to the dendrimer surface can be used to preferentially treat cancer cells with certain over-expressed receptor targets. As dendrimer structures have become more specialized, improved efficacy in in vitro and in vivo models is being realized. The focus of this review is on dendrimer developments from the last four years towards the treatment of cancer, with emphasis on distinct architectures and the biological responses these structures elicit. Mechanistic aspects including dendrimerlipid bilayer interactions, routes of cellular uptake, targeting, and biodistribution are discussed in order to relate composition to therapeutic effect. More broadly, a comprehensive survey of novel dendrimer structures and updates on existing technologies regarding cancer-specific applications is reported. 2. Architecture and composition Dendrimers can be prepared with a level of control not attainable with most linear polymers, leading to nearly monodisperse, globular macromolecules with a large number of peripheral groups. Dendrimers are usually synthesized by one of two strategies. The dendrimer can be grown outwards from a central core, a process known as the divergent method pioneered by Tomalia and Newkome [7–9], or it can be prepared by Fréchet's convergent method by which the dendrimer is synthesized from the periphery inwards, terminat- ing at the core [10]. The branching units are described by generation, starting with the central branched core molecule as generation 0 (G0) and increasing with each successive addition of branching points (i.e., G1, G2, etc.); dendrimers are often characterized by their terminal generation, such that a G5 dendrimer refers to a polymer with four generations of branch points emanating from a central branched core. With each successive generation, the number of end groups increases exponentially. Dendritic macromolecules tend to linearly increase in diameter and adopt a more globular shape with increasing dendrimer generation [3]. As a consequence, dendrimers have become an ideal delivery vehicle candidate for explicit study of the effects of polymer size, charge, and composition on biologically relevant properties such as lipid bilayer interactions, cytotoxicity, internalization, blood plasma retention time, biodistribution, and filtration. The majority of studies have been performed on modified polyamidoamine (PAMAM) dendrimer, in part because PAMAM generations 0 through 10 (G0–G10) are commercially available featuring a wide number of peripheral groups (4 to 4096), end-group functionality (e.g. amine, carboxylic acid, hydroxyl) and molecular weights (657 to 935,000 g/mol). Other dendritic molecules under active investigation include poly(propylene imine), poly(glycerol-co-succinic acid), poly(L-lysine), poly (glycerol), poly(2,2-bis(hydroxymethyl)propionic acid) and melamine dendrimers (Fig. 1). All together, these dendrimers represent a collection of macromolecules which possess varied chemical structures and properties (e.g., basicity, hydrogen bonding capability, charge, etc.) that can be manipulated by increasing dendrimer generation or modifying surface groups. Over the last few years, mechanistic and systematic studies have been taken to understand the relationships between the composition, architecture, and properties of dendrimers towards improved biocompatibility from cell to tissue and pharmacokinetic considerations including biodistribution and excretion. 2.1. Dendrimer-membrane interactions Previous studies have reported that large, cationic macromolecules can disrupt cell membranes to facilitate transport of biomolecules into cells [11,12]. These studies examined the interactions between positively-charged polymers with lipid vesicles and cultured cells, but more recently, a mechanistic understanding regarding the effects of size, charge, and functionality was reported by the groups of Banaszak Hall and Baker. For the first time, dendrimer-lipid bilayer interactions have been directly observed using characterization techniques including atomic force microscopy (AFM), dynamic light scattering (DLS), and 31P NMR. G7-PAMAM dendrimers (~ 8 nm in diameter) with amine or carboxylate peripheral functional groups formed 15–40 nm holes in supported 1,2dimyristoyl-sn-glycero-3-phosphocholine bilayers, a model phospholipid membrane [13]. The dendrimers showed an affinity for bilayer defect edges. Meanwhile, core–shell tectodendrimer clusters—multi-dendritic assemblies comprised of 10 to 12 G5 carboxylate-capped dendrimers covalently attached to a G7 amine-capped dendrimer (~ 28 nm diameter as J.B. Wolinsky, M.W. Grinstaff / Advanced Drug Delivery Reviews 60 (2008) 1037–1055 Fig. 1. Structures of dendrimers used for delivery of cancer therapies. (1) PAMAM, (2) melamine-based dendrimer, (3) dendrimer based on 2,2-bis(hydroxymethyl) propionic acid, (4) PPI, (5) dendrimer based on glycerol and succinic acid with a PEG core, and (6) dendrimer based on 5-aminolaevulinic acid. 1039 1040 J.B. Wolinsky, M.W. Grinstaff / Advanced Drug Delivery Reviews 60 (2008) 1037–1055 reported by the Tomalia group [14])—did not result in hole formation, indicating that charge is not the only factor resulting in membrane disruption. It was also proposed that dendrimer shape may play a part in forming dendrimer-lipid vesicles by removing individual lipid molecules from the membrane. This theory was further supported by 31P chemical shifts indicative of dendrimer–liposome interactions, and by DLS which showed complexes with a mean diameter larger than dendrimer alone but smaller than pure liposomes. The role of size and charge on lipid bilayer disruption was further analyzed with G3, G5, and G7 PAMAM dendrimers displaying positively-charged amine or neutrally-charged acetamide end groups (Fig. 2). Both the G7 amine and acetamide-capped dendrimers led to membrane hole formation. The G5 amine dendrimers showed a significantly diminished ability to remove lipid molecules from the bilayer, participating predominately in the growth of existing deformities, while the G3 amine, G3 acetamide, and G5 acetamide dendrimers showed no propensity to participate in hole formation. Several ideas were suggested relating dendrimer size and the ability of the end groups to associate with lipid molecules, based on dendrimer-lipid vesicle assembly. The ratio of lipid headgroups (L) in contact with the dendrimer surface to the number of dendrimer peripheral end groups (P) seems to be a determining factor in hole formation. G7 PAMAM dendrimers support low (L/P) ratios allowing for the formation of stable dendrimer-lipid vesicles [15]. In a separate study, it was shown that cationic PAMAM leads to hole formations only in the fluid phase of a membrane, whereas the existence of a gel phase in the plasma membrane is unaffected by the presence of these dendrimers, suggesting that the phase of the lipid bilayer may impact cellular uptake studies under certain experimental conditions [16]. The concept of dendrimer architecture and membrane bilayer hole creation was broadened to a range of linear and dendritic polycationic polymers commonly investigated for drug delivery applications, including poly-L-lysine (PLL), polyethylenimine (PEI), diethylaminoethyl-dextran (DEAE-DEX), and PAMAM, and compared against neutrally-charged polymers, including polyvinyl alcohol (PVA) and polyethylene glycol (PEG), in vitro [17]. Polymer solutions were added to KB and Rat2 cell lines, cytosolic enzymes lactate dehydrogenase (LDH) and luciferase (LUC) were measured to detect membrane permeability, and propidium iodide and FITC dye molecules were used to quantify transport in and out of the cells. Polymer charge density was found to significantly impact membrane permeability, with the most densely-charged polymer, PEI, releasing the largest amounts of cytosolic enzymes outside the cells, as well as facilitating the transport of dye molecules. PAMAM's increased ability to enhance membrane permeability was again attributed to its spherical architecture promoting interactions between the dendrimer and lipid molecules. Meanwhile, PVA and PEG polymers had no impact on membrane permeability. All cationic polymers in this study were capable of substantial hole formation at large concentrations, which resulted in cell death. Although membrane permeability may play a role in the cellular uptake of certain dendrimers, conventional modes of endocytotic internalization are attributed to the uptake of many dendrimers. A recent study by the group of Duncun examined the effect of structure on the rate and mechanism of cellular uptake of linear and branched PEIs, and PAMAM (G2, G3, and G4) dendrimers [18]. Binding, endocytic capture, and intracellular trafficking were evaluated with Oregon green-conjugated polymers, and all three polymers were internalized through an adsorptive endocytosis pathway by B16f10 melanoma cells. G4 PAMAM showed the highest rate of uptake, followed distantly by branched PEI, linear PEI, G3 PAMAM, and G2 PAMAM in decreasing magnitude of internalization. Non-specific binding to proteoglycans within the cell membrane was expected via ionic interactions. Branched PEI showed 5-fold greater binding to the plasma membrane than linear PEI, perhaps explaining increased endocytosis of the linear chains compared to the branched polymers. G4 PAMAM and branched PEI were internalized primarily by cholesterol-dependant pathways while Fig. 2. Interaction of PAMAM dendrimers with lipid bilayers. Large, cationic PAMAM dendrimers form nano-scale holes in lipid bilayers. These interactions are greatly reduced or eliminated by decreasing PAMAM generation and/or capping the end groups with neutral acetamide moieties (adapted from [15]). J.B. Wolinsky, M.W. Grinstaff / Advanced Drug Delivery Reviews 60 (2008) 1037–1055 linear PEI uptake was independent of cholesterol and clathrin pathways, suggesting that dendrimer architecture affects the mode of cellular internalization. 2.2. End groups and toxicity Several groups have shown that cell toxicity strongly correlates with dendrimer end group functionality. Positivelycharged groups such as amines generally demonstrate dosedependent toxicity; for this reason, positively-charged groups are often capped with neutral molecules such as acetyl and glycidol groups or poly(ethylene oxide) chains [19,20]. Recent studies have broadened the investigation of end groups on toxicity (Fig. 3). The Schluter group examined the impact of peripheral functionality on the cytotoxicity of MCF-7 breast cancer cells in vitro using low generation (G0, G1, and G2) polyamidoamine-like polymers [21]. The dendrimers were prepared featuring peripheral groups including tert-butoxycarbonyl or benzyloxycarbonyl-protected quaternized amines, tert-butoxycarbonyl-protected or unprotected L-phenylalanine, L-methionine, 1041 or L-aspartic acid amino acids, diaminopropionic acid (platinum-binding), or 5-dimethylaminonapthalene-1-sulphonyl (fluorescent label). The latter two end groups possess the capacity for the delivery of the anti-proliferative cisplatin or contrast imaging modalities, respectively. In general, most of the positively-charged materials led to cell toxicity, but interestingly not all, including diaminopropionic acid dendrimers, showed this effect. The dendrimer core structure did not seem to have an influence on toxicity for these low generation macromolecules. In another study by the Simanek group, the effect of surface groups on cytotoxicity, hemolysis, and acute in vivo toxicity was investigated using melamine polymers as drug delivery vehicles [22]. Unmodified melamine dendrimers have previously shown to be hemolytic. To improve biocompatibility of these polymers, amine, boc-protected amine, guanidine, carboxylate, sulfonate, phosphonate, and PEGylated G3 melamine dendrimers were synthesized and added separately to red blood cells, and acute toxicity and hemolysis was monitored. Positively-charged amine and guanidine groups Fig. 3. Survey of dendrimer end groups. Cationic and certain amino acid peripheral groups tend to invoke a cytotoxic response (⁎). 1042 J.B. Wolinsky, M.W. Grinstaff / Advanced Drug Delivery Reviews 60 (2008) 1037–1055 demonstrated dose and time-dependent hemolytic activity, negatively-charged sulfonate, phosphonate, and carboxylate dendrimers led to limited hemolysis only at high concentrations (~ 1 mg/mL compared to b0.01 mg/mL for amine-terminated at 24 h), and PEGylated melamine showed minimal activity. Cell viability studies on Clone 9 rat liver cells revealed similar trends. PEGylated melamine was chosen as a candidate delivery vehicle, injected into male C3H mice at acute loads to measure in vivo toxicity, and compared to unmodified G3 melamine dendrimers. The PEGylated dendrimers were non-cytotoxic, as evidenced by insignificant increases in urea nitrogen or liver enzyme levels. A previous study on acute and subchronic in vivo toxicity of unmodified G3 melamine dendrimers showed that high doses (160 mg/kg) resulted in 100% mortality [23]. Urea nitrogen blood levels remained normal for all dendrimer concentrations, while alanine transaminase levels indicated healthy liver function at or below 10 mg/kg for acute doses and 2.5 mg/kg for subchronic dosing schemes. These results are consistent with previous PAMAM toxicity investigations where it was shown that G3-PAMAM is well-tolerated in vivo at 2.6 mg/kg for 7 days, 30 days, and 6 months after exposure [24]. 2.3. Pharmacokinetics Circulation time, organ uptake, and tumor accumulation are all critical factors for efficacy in vivo with a polymeric drug delivery system. As such the pharmacokinetics need to be understood as a function of dendrimer composition, generation, and architecture. In a recent study, the groups of Fréchet and Szoka reported the pharmacokinetics of bow-tie dendrimer based on poly(ethylene glycol) (PEG) and 2,2-bis(hydroxymethyl)propionic acid with varying molecular weights and chain numbers to determine the effect of molecular weight and architecture [25]. The polymers were comprised of G3 hydroxyl-terminated branching on one side, and G1, G2, or G3 branching on the opposite side featuring PEG chains of 5000, 10,000, and 20,000 molecular weights. Examples of a [G1](PEG20k)2-[G3]-(OH)8, [G3]-(PEG5k)8-[G3]-(OH)8, and a [G3](PEG10k)8-[G3]-(OH)8 are shown in Fig. 4. All structures, regardless of molecular weight, demonstrated minimal cytotoxicity upon exposure to MDA-MB231 breast cancer cells. Incubation of the [G3]-(PEG5k)8-[G3]-(OH)8, polymer in buffer solutions at pH 5.0 or pH 7.4 led to significant degradation over 15 days, with a bimodal distribution of molecular weights appearing between days 5 and 10 as detected by size exclusion chromatography. The two distributions are indicative of carbamate chain hydrolysis and subsequent cleavage of PEG chains from the dendrimer branches. Ester hydrolysis is also expected to result in further degradation. Biodistribution studies with PEGylated 125I-labeled dendrimers injected into CD-1 mice showed several trends. G3 dendrimers with molecular weights of approximately 45,000, 85,000, and 165,000 (PEGylated with 8 chains per molecule; PEG5000, PEG10000, and PEG20000; [G3]-(PEG5k)8-[G3]-(OH)8, [G3]-(PEG10k)8[G3]-(OH)8, and [G3]-(PEG20k)8-[G3]-(OH)8,respectively) had long circulation times, with elimination half-lives of 31, 40, and 50 h, respectively. Less than 4% of these polymers were excreted in the urine over 48 h, with 7 to 16% excreted in the feces over the same time span. Circulation of the [G2]-(PEG)4[G3]-(OH)8 dendrimer system was significantly shorter, with the ≈ 45,000 and 85,000 molecular weight dendrimers (PEG10000 and PEG20000 respectively) showing half-lives of 26 and 25 h, while the 23,000 molecular weight [G2]-(PEG5k)4[G3]-(OH)8 had a decidedly smaller half life of 11 h. This result was not surprising as the molecular weight cutoff for renal filtration of linear PEG has been reported between 30,000 to 40,000 Da [26]. Importantly, an architecturally dependent response was observed, where the number of PEG macromolecules attached to the bow tie dramatically influenced circulation time. A [G1]-(PEG20k)2-[G3]-(OH)8 polymer of 44,000 Da has a half life of 1.5 h whereas a [G3]-(PEG5k)8[G3]-(OH)8 of similar molecular weight has a half life of 31 h. Next, the two largest [G3]-(PEG10k)8-[G3]-(OH)8 and [G3](PEG20k)8-[G3]-(OH)8 macromolecules were injected intravenously into tumored C57BL6 mice previously injected subcutaneously with B16F10 melanoma cells. Both polymers showed similar biodistribution characteristics, with high levels of material found in the tumors (10% to 15%) and blood (18% to 20%) at 48 h. The biodistribution characteristics of G4, G5, and G6 amino acid dendrimers based on poly(L-lysine) or poly(L-ornithine), and their PEGylated derivatives, were reported by Okuda et al. [27]. PEGylation of the dendrimer had remarkably different Fig. 4. Poly(ethylene oxide) Bow-tie hybrids. Pharmacokinetics of dendrons based on 2,2-bis(hydroxymethyl)propionic acid can be tuned by varying generation size (G1, G2, and G3) and PEG chain molecular weight (5000, 10,000, and 20,000). aMALDI-TOF MS (adapted from [25]). J.B. Wolinsky, M.W. Grinstaff / Advanced Drug Delivery Reviews 60 (2008) 1037–1055 biodistribution characteristics than its unmodified counterpart. Non-PEGylated G4, G5, and G6 amino acid dendrimers were all eliminated from circulation within minutes of injection, with accumulation primarily in the liver and kidneys. Hepatic accumulation increased with larger generation polymers, correlating to similar increases in dendrimer positive charge, while renal accumulation decreased. PEGylation of the G6 poly (L-lysine) resulted in a polymer of approximately 396,000 Da molecular weight with a particle size of about 17 nm compared to an original polymer of 16,000 Da molecular weight with a particle size of about of 6 nm. The retention time of this molecule was increased from minutes to over 24 h. Liver accumulation also substantially decreased by about half (to 25% over 60 min), and renal accumulation was not detected. The G6-PEGylated lysine dendrimers were also evaluated for their tumor-selective targeting as a consequence of the EPR effect through biodistribution studies in normal and tumorbearing mice [28]. The two PEGylated lysine dendrimers possessing low (10) or high (~ 76) number of PEG 5000 Da chains attached to the periphery groups (66,000 vs 396,000 Da) were investigated. The remaining unmodified end groups consisted of primary amines. Neither of the PEGylated derivatives accumulated in the kidney. The dendrimer with the highest PEG content accumulated effectively in tumor tissue with enhanced retention in the plasma, while the nonPEGylated lysine dendrimer showed negligible tumor accumulation and rapid clearance. The biodistribution of 3H-labeled G5 PAMAM positivelycharged and acetylated dendrimers was evaluated in B16 melanoma and DU145 prostate tumor models. Both dendrimers showed non-specific distributions with rapid clearance from the blood within 24 h post-injection [29]. Positivelycharged PAMAM showed greater tissue deposition. Accumulation was greatest in the lungs, liver, and kidneys, with approximately 3% of initial dendrimer-loading found in the tumor tissue after 1 h. Neutral-charged PAMAM was initially excreted through urine three-times more rapidly than positively-charged PAMAM (48% and 30% respectively over 7 days), and both dendrimers were excreted through feces to a lesser extent (5% and 3%). Longer biodistribution studies performed on non-tumor-bearing mice over 12 weeks time showed no deleterious effects. Small levels of dendrimers were still present throughout the major organs measured after 12 weeks, particularly in the kidney. 3. Drug delivery Polymer-based drug delivery systems are designed to improve the pharmacokinetics and biodistribution of a drug and/or provide controlled release kinetics to the intended target [30]. The ideal dendrimer carrier should exhibit high aqueous solubility and drug-loading capacity, biodegradability, low toxicity, favorable retention and biodistribution characteristics, specificity, and appropriate bioavailability. In dendrimer-based drug delivery, a drug is either non-covalently encapsulated in the interior of the dendrimer or covalently conjugated to form macromolecular prodrugs. 1043 3.1. Drug-encapsulated dendrimers Poly(glycerol succinic acid) dendrimers, or PGLSA dendrimers, were investigated as delivery vehicles for camptothecins, a group of naturally-derived hydrophobic compounds with anti-cancer activity. In a preliminary study reported by the Grinstaff group, G4-PGLSA dendrimers with hydroxyl (G4PGLSA-OH) or carboxylate (G4-PGLSA-COONa) peripheral groups were used to encapsulate 10-hydroxycamptothecin (10HCPT) for delivery to cancer cells [31]. The G4-PGLSA-OH/ 10-HCPT solution precipitated upon standing after mixing; the more water-soluble G4-PGLSA-COONa dendrimer was used instead to improve overall solubility and 10-HCPT was successfully encapsulated. Upon exposure to MCF-7 human breast cancer cells, unloaded dendrimer showed no cytotoxic effects, while 10-HCPT-encapsulated dendrimers led to significant cytotoxicity with less than 5% of viable cells at higher concentrations (20 μM). An alternative triblock structure was explored by introducing a 3400 molecular weight PEG core to the G4-PGLSA dendrimer to afford (G4-PGLSA-OH)2PEG3400 [32]. A 20-fold increase in 10-HCPT water solubility was observed following encapsulation. The anti-cancer activity of the macromolecule/drug complex was examined using HT-29 human colon cancer cells and similar cytotoxicities were reported for encapsulated and free 10-HCPT. The conclusions drawn from these two studies led to the selection of G4PGLSA-COONa dendrimer as a delivery vehicle for 10-HCPT and 7-butyl-10-aminocamptothecin (BACPT), a highly potent lipophilic camptothecin derivative [33]. Anti-cancer activity was investigated for a human cancer cell line panel including HT-29 colon cancer, MCF-7 breast carcinoma, NCI-H460 large cell lung carcinoma, and SF-268 astrocytoma. Solubility, cellular uptake, and cellular retention studies were also performed for MCF-7 cells. The release profile of 10-HCPTencapsulated G4-PGLSA-COONa showed full release of the drug within approximately 6 h, suggesting that the delivery system may be best utilized by intratumoral injection. Dendritic delivery of 10-HCPT and BACPT resulted in lowered IC50s for all cell lines tested (Fig. 5); for 10-HCPT exposure to HT-29, MCF-7, NCI-H460, and SF-268 cells, IC50s were reduced by 3.5, 7.1, 1.9, and 2.8-fold, respectively, compared to free 10HCPT dissolved in DMSO. Exposure of BACPT led to IC50 reductions of 1.2, 3.2, 1.9, and 5.7-fold for the respective cell lines above. Uptake studies showed that dendrimer-encapsulated 10-HCPT was internalized much faster than free drug, with 16-fold intracellular concentrations at 2 h and 8-fold intracellular concentrations at 10 h. Drug delivered via the dendrimers also showed longer retention time in the cell, with 50% of delivered 10-HCPT present in the cell after 30 min, compared to 35% of free drug. Thus, increased toxicity of delivered camptothecins was attributed to enhanced uptake and retention. The cytotoxicity and encapsulation efficiency of star amphiphilic PAMAM block copolymers containing poly(γcaprolactone) and PEG arms has also been assessed [34]. The polymer forms micelles in solution which were non-cytotoxic. The anti-cancer drugs doxorubicin and etoposide were 1044 J.B. Wolinsky, M.W. Grinstaff / Advanced Drug Delivery Reviews 60 (2008) 1037–1055 Fig. 5. Drug delivery via a dendrimer based on glycerol and succinic acid. Left, chemical structure of G4.5-PGLSA-COONa dendrimer. Right, chemical structures of (1) 10-hydroxycamptothecin, and (2) 7-butyl-10-aminocamptothecin with IC50 values for HT-29 colorectal adenocarcinoma, MCF-7 breast carcinoma, NCI-H460 large cell lung carcinoma, and SF-268 astrocytoma human cancer cell lines (adapted from [32]). encapsulated within the micelles. Doxorubicin showed low encapsulation efficiency while the more lipophilic etoposide achieved a loading capacity of approximately 22% by weight. Unloaded dendrimer was non-cytotoxic to porcine kidney epithelial cells, while etoposide-encapsulated dendrimers showed comparable toxicity to free etoposide at similar drug concentrations. Enhanced aqueous solubility of paclitaxel was achieved with poly(glycerol) dendrimer formulations, showing that a hydrophobic dendrimer core is not necessary for encapsulation and solubilization of hydrophobic drugs [35]. Paclitaxel solubilities ranged from 80–128 μg/mL with increasing generations from G3–G5 of poly(glycerol), or three orders of magnitude higher than free paclitaxel. Nuclear magnetic resonance data suggests that the drug is not incorporated within the core of these dendrimers, but instead the methyne groups and aromatic rings of the paclitaxel are surrounded by the dendrimer structure leading to hydrotropic solubilization. Melamine-based dendrimers were used to solubilize the anticancer drugs methotrexate and 6-mercaptopurine, as well as to reduce drug toxicity [36]. C3H mice received subchronic doses of drug-encapsulated dendrimers and ALT levels were evaluated to determine hepatotoxicity. ALT levels were reduced by 27% for methotrexate-encapsulated dendrimers and by 36% for the 6-mercaptopurine dendrimers compared to animals treated with drug alone. Medium-generation dendrimers (i.e., G4–G6) have been shown to both enhance solubility and increase toxicity (lower IC50) of hydrophobic anti-cancer drugs through non-covalent encapsulation. Therapeutic agents are internalized within the interior core space or by micellar formation of the dendrimers. A major drawback to these delivery systems is a lack of controlled drug release kinetics, with most systems releasing their payload over the course of several hours. For this reason, drug-encapsulated dendrimer systems may best be utilized via direct intratumoral injection. 3.2. Dendrimer-drug conjugates Dendrimer-drug conjugates generally consist of an antineoplastic agent covalently attached to the peripheral groups of the dendrimer. This method offers distinct advantages over drug-encapsulated systems. Multiple drug molecules can be attached to each dendrimer molecule and the release of these therapeutic molecules is partially controlled by the nature of the linkages. The Kannan group reported the synthesis of PAMAM-methotrexate conjugates from carboxylic acid-terminated G2.5 PAMAM or amine-terminated G3 PAMAM in order to assess the activity of dendrimer-delivered methotrexate to sensitive and resistant CCRF-CEM human acute lymphoblastoid leukemia and CHO Chinese hamster ovary cell lines [37]. Although both polymers were conjugated to the drug by the formation of amide bonds, the carboxylic acid-conjugated G2.5 PAMAM system showed increased sensitivities of 8- and 24-fold towards the MTX-resistant cell lines CEM/MTX and RII, while amine-conjugated G3 PAMAM showed no such increases compared to free methotrexate. The differences in cytotoxicity were attributed to the charge of the dendrimer carrier after cleavage of methotrexate from the peripheral groups. It was proposed that the lysosomotropic effect, in which the displacement of small basic molecules from the lysosome by positively-charged dendrimers is accompanied by an increase in pH and eventual lysosomal disruption, was responsible for a decrease in lysosomal residence time for the cationic PAMAM. As a result, the conjugates experience reduced interactions with proteases and diminished drug release. The results indicate the potential of dendrimer-drug conjugates for the treatment of J.B. Wolinsky, M.W. Grinstaff / Advanced Drug Delivery Reviews 60 (2008) 1037–1055 cancer cells, particularly those that have demonstrated resistance to chemotherapeutics. Paclitaxel was conjugated to PEG or G4-PAMAM to compare the anti-cancer activity of the drug delivered by a linear or dendritic carrier [38]. Both PEG and PAMAM increased the aqueous solubility of paclitaxel (0.3 μg/mL) dramatically to 2.5 mg/mL and 3.2 mg/mL respectively. Upon exposure to human ovarian carcinoma A2780 cells, free paclitaxel accumulated in the cytoplasm near the plasma membrane. The polymer conjugates tended to distribute intracellularly in a more homogenous fashion compared to free drug. PEG-paclitaxel conjugates reduced the efficacy of the drug 25-fold, but the PAMAM-paclitaxel conjugates decreased the IC50 more than 10-fold when compared against free drug, leading to the conclusion that the availability of a drug is dramatically influenced by the architecture of its polymer conjugate. Doxorubicin-G4-PAMAM complexes have been encapsulated into liposomal formulations for potential local delivery to locations such as skin metastasis from breast cancer [39]. The dendrimer-drug complex was incorporated into one of two formulations to modulate release compared to doxorubicinliposomal systems. The first formulation was comprised of eggphosphatidylcholine, stearylamine, and the anti-tumor ether lipid hexadecylphosphocholine (HePC), while the second formulation was similar except did not include HePC. 1045 Incorporation efficiencies were above 90% and slow release was achieved with less than 20% released over 48 h for both systems. Cytotoxicity was assessed based on doxorubicin activity on several cancer cell lines including lung, colon, breast, prostate, and CNS. The doxorubicin-PAMAM liposome formulation with HePC showed the highest activity against most of the cell lines, with enhanced activity towards MDA-MB435 breast cells compared to the dendrimers conjugate alone, and high sensitivity towards DMS114 and NCI-H460 lung cancer cells. It should be noted that the dendrimer-liposomal complexes in this study increased in size from approximately 115 nm to 2000 nm after 18 weeks in storage at 4 °C, and this was attributed to the formation of liposomal aggregates facilitated by hydrophobic forces between dendrimers. A remarkable example of architecturally-optimized dendritic drug delivery was reported by Fréchet and Szoka where an asymmetric doxorubicin-functionalized bow-tie dendrimer was prepared by PEGylation of one side of a 2,2-bis(hydroxymethyl)propionic acid dendrimer (G3) and attachment of the drug via an acyl hydrazone linkage to the other side (G4) resulting in 8–10 wt.% doxorubicin content overall [40] (Fig. 6). They used a pH sensitive linkage between the drug and dendrimer to release the drug once in the cell. Following intravenous administration to BALB/c mice with s.c. C-26 colon carcinoma tumors, tumor uptake was approximately 9fold higher compared to free doxorubicin. A single injection of Fig. 6. Doxorubicin-functionalized bow-tie dendrimer. PEGylated polyester dendrimer with tunable molecular weight, drug-loading, water solubility, pharmacokinetics, and biodistribution cures C-26 colon carcinomas in mice with one dose (adapted from [40]). 1046 J.B. Wolinsky, M.W. Grinstaff / Advanced Drug Delivery Reviews 60 (2008) 1037–1055 doxorubicin-conjugated dendrimer caused complete tumor regression and 100% survival of mice over two months, while no cures were observed with drug alone treatments. It is clear that dendrimer-drug conjugates are highly capable of delivering a payload with sufficient bioavailability to achieve a therapeutic goal. The release of covalently linked drug is dependent upon the chemical linkage binding the agent to the carrier. Novel dendrimer structures are being synthesized to further explore finer control of release kinetics. 4. Targeted drug delivery Macromolecular delivery of anti-cancer drugs using multifunctional dendritic architectures allows for the conjugation of both drugs and targeting moieties such as folic acid, monoclonal antibodies, and peptides to the dendrimer periphery for increasingly specific delivery. In the field of oncology, the targeted delivery of chemotherapeutics to tumor cells translates to significantly reduced side effects compared to systemic delivery where healthy tissue such as the liver, spleen, kidneys, and bone marrow can accumulate toxic levels of drug. The two general strategies of targeting include the passive targeting of bulk cancerous tissue and the active targeting of unique tumor cells. Non-specific or passive targeting of tumors is usually achieved by increasing the hydrodynamic radius of the dendrimer though PEGylation, leading to the accumulation of dendrimer in tumor tissue via the enhanced permeability retention (EPR) effect. The EPR effect is a result of tumorinduced angiogenesis leading to neovasculature that is irregular, leaky or defective with disorganized endothelial cells; tumor tissues also suffer from poor lymphatic drainage, all leading to the accumulation and retention of macromolecules in the tumor mass [41]. Specific or active targeting relies on the conjugation of one or more targeting moieties to the dendrimer to facilitate cell-receptor-mediated interactions. 4.1. Folic acid Studies have shown that folic acid-conjugated dendrimers preferentially target tumor cells that overexpress folic acid receptors [42–44]. A recent study by Hong et al. explicitly quantified the binding avidity of multi-valent targeted G5PAMAM containing different numbers of folic acid molecules [45]. Binding avidity to folic acid receptor-overexpressing cells increased with each additionally bound FA molecule conjugated to the dendrimer, saturating at 5–6 moieties per dendrimer, though the rate of intracellular internalization was not significantly affected with increased binding. The dendrimers demonstrated a dramatic enhancement of binding avidity of almost 5 orders of magnitude. It was suggested that aggregates of 5–6 FA receptors are preorganized on the membrane and that the key factor in reported tumor reduction is enhanced residence time on the cell and not the rate of endocytosis. In another example, DNA-assembled PAMAM dendrimer clusters were prepared by linking two dendrimer components with single but different functionalities for concurrent delivery of therapeutic, imaging, and targeting agents [46,47]. Complexes were formed between a folic acid-modified dendrimer and a FITC-modified dendrimer connected by a 34-base-pair long oligonucleotide. Clusters effectively targeted KB cells expressing folic acid receptors and were internalized by the cells (Fig. 7). The Baker group has investigated several variations of folic acid-conjugated dendrimers for targeted drug delivery. Surfaceconjugated folic acid G5-PAMAM dendrimers were prepared where the remaining free amine groups were capped with glycidol to neutralize the positive charges, and then further reacted with methotrexate (MTX) to form ester linkages [48]. A comparison between encapsulated MTX vs covalently bound drug release showed a rapid release for the free drug over 2.5 h (~ 75%), compared to a much slower release for the bound drug over the same period of time (~ 5%). Furthermore, encapsulated drug displayed diffusion characteristics similar to free drug. Folic acid-targeted MTX conjugates demonstrated high specificity for KB cells overexpressing folic acid receptors. When exposed to these cells, both free drug and dendrimer conjugates show similar cytotoxicity activity, but when the folic acid receptors are blocked or underexpressed, the conjugates lose their anti-proliferative effect, indicating receptor-mediated delivery. Improvements in the synthesis and scale-up of PAMAM-methotrexate conjugates have led to high synthetic reproducibility [49]. In a separate study, folic acid, fluorescein, and methotrexate were conjugated to PAMAM and examined in vitro against KB cells [50]. Anti-proliferative activity was slightly lower for the dendrimer-drug conjugates compared to free methotrexate. Dose-dependent binding to KB cells was demonstrated and compared to fluorescein-modified PAMAM not containing folic acid. Targeting was diminished yet still significant against KB cells underexpressing FA receptors. The drug-dendrimer conjugates became ineffective when the cells were pretreated with free folic acid. A comparable study was performed with folic acid, fluorescein, and paclitaxel conjugated to partially acetylated PAMAM dendrimers [51]. Again, folic acid-targeting occurred, preferentially delivering paclitaxel-conjugated dendrimers to KB cells. Internalization was not detected when dendrimers were exposed to down-regulated KB cells. In a related project, PAMAM dendrimer-based sensors have been targeted to tumor cells to detect the anti-cancer activity of therapeutics. Multi-functional folic acid-targeted PAMAM delivery vehicles were synthesized and covalently bound to the apoptotic sensor PhiPhiLux G1D2 in order to detect the extent of cell killing caused by a delivered anti-proliferative agent [52]. PhiPhiLux G1D2 is a caspase-specific FRET-based agent that responds to the release of the apoptosis-inducing agent, staurosporine. The dendrimers were internalized within the first 30 min of incubation with Jurkat cells and upon apoptosis, a 5-fold increase in intracellular fluorescence was detected, demonstrating the potential of chemotherapeutic delivery while monitoring cell-killing efficacy in vivo. 4.2. Peptides A doubly cyclized RGD (RGD-4C) peptide and Alexa Fluor 488 fluorescent label were conjugated to partially acetylated J.B. Wolinsky, M.W. Grinstaff / Advanced Drug Delivery Reviews 60 (2008) 1037–1055 G5-PAMAM for the targeting of tumor neovasculature via uniquely expressed αVβ3 integrins [53]. The RGD-4C peptide was attached by an acylhexanoic acid spacer to ensure adequate exposure of the conjugate to the target. Binding studies were performed on several cell lines with varying levels of integrin receptor expression. Free RGD-4C bound much more rapidly than the RGD-4C-dendrimer complexes, but the dendrimers disociated approximately 522 times slower, suggesting a synergistic effect of multiple peptide conjugation on binding avidity. Cyclic RGDs have also been attached to DOTA-conjugated mono, di-, and tetravalent dendrimeric alkynes for αVβ3 integrin targeting [54]. Binding characteristics were evaluated in vitro and in vivo in mice with human SK-RC-52 tumors and it was shown through biodistribution studies that the tetrameric RGD-dendrimer showed the highest level of tumor targeting. 4.3. Monoclonal antibodies Monoclonal antibody-conjugation to PAMAM has been explored for specific targeting of tumor cells that overexpress certain antigens. An anti-prostate specific membrane antigen (PSMA), J591, was conjugated to G5-PAMAM and evaluated in vitro for binding affinities and internalization [55]. PSMA is overexpressed in all prostate cancers, non-prostatic tumor neovasculature, and vascular endothelium in most solid sarcoma and carcinoma tumors [56]. The antibody–dendrimer conjugate was found to specifically bind to PMSA-positive (LNCaP.FGC) cells but not to PMSA-negative (PC-3) cells. Furthermore, the conjugate was internalized as determined by confocal microscopy, while the unconjugated dendrimer was not significantly taken up by cells. A similar study was performed using anti-HER2-G5-PAMAM for the targeting of human growth factor receptor-2, which is often overexpressed in breast and ovarian malignancies (Fig. 8) [57,58]. The conjugates showed binding and internalization into HER2expressing cells. Specific and increased binding to HER2expressing tumors was also demonstrated in vivo. A third study investigated two different antibody-G5-PAMAM conjugates, 60bca and J591, for the targeting of CD14 and 1047 prostate-specific membrane antigen (PMSA), respectively [59]. Targeting was achieved in vitro using two different antigen-expressing cell models including CD14-expressing HL-60 human myeloblastic leukemia cells and PMSAexpressing LNCaP cells. Methotrexate was covalently attached to G5-PAMAM bioconjugates containing cetuximab, a monoclonal antibody that acts as an epidermal growth factor receptor (EGFR) inhibitor and is currently used as a drug to treat colorectal, head, and neck cancers. The conjugate was designed for targeted delivery to EGFR-positive brain tumors, to build upon the demonstrated successful targeting and delivery of boronated PAMAM cetuximab conjugates to gliomas for neutron capture therapy (discussed in detail later) [60]. Approximately 13 methotrexate molecules were attached to each dendrimer as confirmed by UV/vis spectroscopy. The bioconjugate showed a modest 0.8 log unit reduction in its EC50, though the IC50 was 2.7 log units lower for the conjugate compared to free methotrexate. Unfortunately, tumor-bearing animals did not show a significant response from the bioconjugate compared to free methotrexate, possibly due to a lack of cleavage from the PAMAM scaffold. 4.4. Glycosylation One strategy to selectively deliver drug-conjugates to tumor cells used glycopeptide dendrimers conjugated to the anti-mitotic agent cholchicine. Glycodendrimers are a class of dendrimers that incorporate sugar moieties such as glucose, galactose, mannose [61], and/or disaccharides [62] into their structure. The dendrimer consists of 2,3-diaminopropionic acid branching featuring amino acids, a cysteine core, and four to eight glycoside groups on the periphery. The conjugates were prepared and evaluated against a cancer cell line (HeLa) and healthy cells (non-transformed mouse embryonic fibroblasts or MEFs) [63]. While the glycopeptide dendrimer conjugates were not as anti-proliferative as cholchicine alone, the dendrimers were 20–100 times more effective at inhibiting proliferation of HeLa cells than MEFs, Fig. 7. Advances in folic acid receptor (FAR) targeting. Left. Binding avidity of folic acid-conjugated PAMAM increases with increasing numbers of bound folic acid, and saturates at approximately six FA molecules per dendron (#FA). KD = dissociation constant. Right (1) DNA-assembled PAMAM dendrimer clusters, (2) trifunctional PAMAM covalently attached to fluorescein, folic acid, and a chemotherapeutic, and (3) folic acid-targeted PAMAM bound to apoptosis sensor PhiPhiLux G1D2 (adapted from [45–52]). 1048 J.B. Wolinsky, M.W. Grinstaff / Advanced Drug Delivery Reviews 60 (2008) 1037–1055 Fig. 8. Active targeting via a monoclonal antibody–dendrimer conjugate. A G5-PAMAM conjugated anti-HER2 mAb targets tumors that overexpress the human epidermal growth factor receptor-2 (HER2). The HER2 proteins is observed in breast and ovarian cancers in particular (adapted from [57,58]). whereas non-glycosylated dendrimers showed a selectivity of less than 10-fold for HeLa cells. 5. Photodynamic therapy Photodynamic therapy (PDT) relies on the activation of a photosensitizing agent with visible or near-infrared (NIR) light. Upon excitation, a highly energetic state is formed which upon reaction with oxygen affords a highly reactive singlet oxygen capable of inducing necrosis and apoptosis in tumor cells [64]. The tumor selectivity of porphyrin photosensitizers has been attributed to its characteristic leaky vasculature, compromised lymphatic drainage, and high degrees of newly synthesized collagen and lipid content, both for which porphyrins have an affinity for [65]. PDT has been shown to reduce tumors by direct cell killing, destruction of tumor neovasculature, and triggering of an acute inflammatory response that attracts leukocytes to the tumor [66]. Dendritic delivery of PDT agents has been investigated within the last few years in order to improve upon tumor selectivity, retention, and pharmacokinetics [67–69]. Several studies have investigated the use of dendrons and dendrimers composed in part of multiple 5-aminolaevulinic acid (ALA) for improved delivery and enhanced intracellular accumulation of porphyrins. ALA is formed during the first step of the heme biosynthetic pathway and leads to the conversion of the photosensitizer protoporphyrin IX (PpIX), which can selectively accumulate in tumors [70]. Di Venosa et al. reported the synthesis of a G0-ALA dendron with a free amine at the core and three ALA groups at the periphery [71]. Upon exposure to LM3 murine mammary adenocarcinoma cells, equimolar equivalents of dendron to free ALA showed similar efficacy inducing porphyrin generation. It was determined that only one of three ALA molecules was cleaved from the dendron within the cells. Compared to the widely investigated lipophilic hexyl ester derivative of ALA(He- ALA), G0-ALA dendrons led to high accumulations of porphyrin in vivo both through topical and systemic deliveries. Next, lipophilicity and esterase accessibility were examined through a set of G0-ALA dendrons with varying cores and linker lengths, including an amino core (3 m-ALA) with a methyl linker, a nitro core (3H-ALA) with a propyl linker, and an aminobenzyloxy carbonyl core (3Bz-ALA) with a methyl linker, each attached to three ALA molecules (Fig. 9) [72]. Partition coefficients were calculated for the dendrons and 3HALA was found to be the most lipophilic. While all dendrons induced higher porphyrin production compared to free ALA in vitro, 3H-ALA led to approximately 10-times porphyrin generation compared to free ALA at lower concentrations, while 3Bz-ALA was most effective at higher concentrations (3H-ALA precipitates at higher concentrations). The highest phototoxicity was achieved with 3H-ALA (10% survival at 0.05 mM), followed by 3Bz-ALA (40% survival at 0.05 mM), 3 m-ALA, and free ALA. When applied topically to explanted rat skin, both 3H-ALA and 3Bz-ALA generated higher porphyrin fluorescence compared to 3 m-ALA and free ALA. Analysis of the dendron structures indicate that both lipophilicity and accessibility to the ALA ester linkages may have led to the comparatively higher success of 3H-ALA. Noting the successes from the first two studies, a larger second generation ALA-based dendrimer comprised of 18 ALAs (18 m-ALA) attached to a tripodent aromatic core by polyamidoamine linkers was evaluated for its ability to deliver a high payload of ALA molecules [73]. At lower concentrations, 18-ALA was superior to free ALA regarding PpIX production; furthermore, PpIX generation was significantly higher for the dendrimer after 24 h of incubation compared to the free drug, indicating that ALA is gradually cleaved from 18 m-ALA over time. An earlier study performed with 18 m-ALA using acetamido linkers demonstrated slightly lower efficacy, suggesting that polyamidoamine linkers allow for greater esterase accessibility for cleaving ALA groups from the hyper-branched J.B. Wolinsky, M.W. Grinstaff / Advanced Drug Delivery Reviews 60 (2008) 1037–1055 1049 Fig. 9. Phototoxicity of dendritic derivatives comprised of 5-aminolaevulinic acid. The phototoxicity in cell culture of dendrons based on 5-aminolaevulinic acid (ALA) is dependent upon the dendron core chemical composition and likely related to lipophilicity and esterase accessibility (adapted from [72]). structure [69]. As a final note, 18 m-ALA dendrimers were found to be internalized by macropinocytosis, while the G0ALA dendrimers underwent active transport and passive diffusion. Another structure designed to enhance photodynamic efficacy is comprised of a negatively-charged G3-poly(benzyl ether) dendrimer with carboxylate periphery groups and a zinc porphyrin at the focal core, surrounded by positively-charged linear PEG-lysine block copolymers [74]. Remarkably, the encapsulated dendrimer PEG-lysine micelle system resulted in a 280-fold increase in phototoxicity against Lewis lung cells in vitro compared to free dendrimer. The carrier was then delivered to choroidal neovascularization (CNV) sites to determine accumulation in CNV lesions where it was found that the micelles selectively target the neovascular regions, attributed to the hyperpermeable nature of these lesions and amenable to the EPR effect in tumors. Free dendrimer exhibited decreased uptake attributed to its negatively-charged side groups. Accumulation of the micelles led to significant increases in phototoxicity which was explained by the prevention of porphyrin aggregation by steric hindrance resulting from the dendrimer-porphyrin complex. A closer investigation of the dendrimer micelle complex revealed that the micelles increase in size at lower pHs and precipitate under pH 5.6, indicating that the micelles may preferentially accumulate in the acid environment associated with tumors [75]. Isolation of the porphyrin at the dendrimer core may prevent fluorescence quenching and inhibit non-radiative decay, thus leading to increased fluorescence. Uptake of dendrimer micelles led to intracellular levels six to eight-times higher than dendrimer alone. It was conjectured that the high phototoxicity of dendrimer micelle complexes is due to localized aggregation in cell organelles, which are susceptible to photodamage. Photosensitizer carriers were also prepared from PEGylated G5-PAMAM and G4-PPI by the encapsulation of rose bengal or PpIX [76]. PEGylated G4-PPI formed stable complexes with PpIX. PEGylated G5-PAMAM demonstrated less stability as evidenced by the relatively low fluorescence intensity of the complexes after 3 h of release, suggesting that hydrophobic forces, as measured by shifts in the emission spectra of dendrimer-encapsulated 5-(dimethylamino)-1-naphthalenesulfonic acid (DNS), were an important factor for dictating dendrimer-drug complex stability. Free PpIX and dendrimerencapsulated PpIX showed similar photosensitivity in vitro, suggesting that PEGylated G4-PPI might be a promising carrier, particularly for passive targeting of tumor microvasculature. 6. Boron neutron capture therapy Boron neutron capture therapy (BNCT) is based on a lethal B(n,α)7Li capture reaction that occurs when 10B is irradiated with low-energy thermal neutrons to produce high energy αparticles and 7Li nuclei. These particles have limited path lengths in tissue (b10 μm) and thus their toxicity is limited to cells that have internalized 10B [77]. The emergence of BNCT as a significant clinical treatment modality has historically been limited by either a lack of sufficient tumor targeting or subtherapeutic 10B accumulation (~ 109 atoms/cell) in malignant tissues. To this end, macromolecular delivery vehicles have been prepared to enhance both the quantity of and targeting of 10 B to tumor cells by conjugating boron-containing complexes to monoclonal antibodies or receptor-targeting agents [78]. PAMAM has been the dendrimer system of choice for investigating intratumoral delivery of neutron capture therapy agents [79–82]. Human gliomas have been targeted with boronated G5-PAMAM conjugated to anti-EGF receptor monoclonal antibodies, which work against overexpressed tumor cell receptors. In one study by Fenstermaker et. al., a dendrimer was conjugated with cetuximab (Ctx), an EGF receptor-specific monoclonal antibody, and approximately 1100 boron atoms (Ctx-G5-B1100) and then evaluated in vitro and in vivo using F98 cells and Fischer rats with or without a mutant gene that causes overexpression of EGF [83]. The cell binding of cetuximab-conjugated dendrimer was comparable to that of free cetuximab in vitro. Intratumoral injections in mutated rats showed a 13.8-fold increase in tumor boron content for the targeted dendrimers over unmodified boronated G5PAMAM. 10 1050 J.B. Wolinsky, M.W. Grinstaff / Advanced Drug Delivery Reviews 60 (2008) 1037–1055 A more recent study evaluated the use of Ctx-G5-B1100 conjugates with or without boronophenylalanine (BPA) or sodium borocaptate (BSH), two drugs currently used for BNCT, for the treatment of F98EGFR glioma [84]. Boronated dendrimer was delivered via convection enhanced delivery (CED), a positive pressure method that facilitates transport across the blood–brain barrier, or intratumorally (i.t.) resulting in high retention of boron in the gliomas, with approximately 50% more accumulation resulting from the CED method. BNCT was performed on animals that received CED Ctx-G5-B1100, CED Ctx-G5-B1100 and i.v. BPA, i.t. Ctx-G5-B1100, or i.v. BPA, with mean survival times of 54.5, 70.9, 42.7, and 40.1 days respectively compared to irradiated and radiated controls (30.3 and 26.3 days) (Fig. 10). A second experiment evaluated CED Ctx-G5-B1100, i.v. BPA and BSH, CED Ctx-G5-B1100 BPA and i.v. BSH, and CED Ctx-G5-B1100 BPA and intracarotid (i.c.) BSH with mean survival times of 56.4, 50.9, 67.1, and 75.8 days respectively compared to irradiated and radiated controls (40.3 and 34.4 days), thus demonstrating the therapeutic value of Ctx-G5-B1100 with or without co-delivery of BPA for targeting EGF receptors towards the treatment of gliomas. A similar study was performed with L8A4-conjugated boronated G5-PAMAM [85]. L8A4 is a monoclonal antibody that specifically targets a mutant isoform of EGFR, EGFRvIII, that seems to be exclusively expressed in tumors, whereas EGFR, while found in malignancies, is also located in healthy liver and spleen. Biodistribution data confirmed undetectable amounts of boron in normal brain, liver, kidney, and spleen tissues, with high glioma accumulations. A similar battery of treatment regimes as the study above was performed with the delivery of CED BD-L8A4 and i.v. BPA resulting in a mean survival time of 85.5 days with 20% long-term survivors, compared to irradiated and radiated controls (30.3 and 26.3 days respectively). Besides brain malignancies, neutron capture therapy (NCT) has also been reported for tumor vasculature and micrometastatic lymphatics. Backer et al. derivatized boronated G5PAMAM (BD) with vascular endothelial growth factor (VEGF) Fig. 10. Treatment of EGFR-positive glioma with cetuximab-conjugated boronated PAMAM. Boronated G5-PAMAM (B-PAMAM) was targeted to epidermal growth factor receptors for localized boron neutron capture therapy (BNCT) via conjugation with the monoclonal antibody cetuximab. Treatment was compared with and without traditional over different delivery methods. BPA = boronophenylalanine, BSH = sodium borocaptate, i.v. = intravenous, i.t. = intratumoral, i.c. = intracarotid (adapted from [84]). and near-IR Cy5 dye, or VEGF-BD/Cy5, for targeting upregulated VEGF receptors overexpressed in tumor neovasculature [86]. Near-IR imaging confirmed accumulation of VEGF-BD/Cy5 and not the BD/Cy5 conjugate in 4T1 mouse breast carcinoma with increased concentrations at the tumor periphery. Gadolinium-based (Gd) neutron capture therapy is an alternative to BNCT that has been investigated due to Gd's high neutron absorbency properties, but has rarely been used as it is deemed too difficult to achieve therapeutic doses intravenously. Kobayashi et al. explored Gd-labeled PAMAM dendrimers for the delivery of MRI contrast agents that may also facilitate the use of neutron capture therapy to the sentinel lymph node, which is often imaged for breast cancer management [87]. Generation 2, 4, 6, and 8 PAMAM dendrimers, ranging in sizes from 3 to 12 nm, were evaluated to determine the optimal particle size for entering the lymphatic vessels while avoiding leaking, and it was shown that G6-PAMAM (9 nm) produced the earliest and most intense opacification of the sentinel lymph nodes with sufficient Gd concentrations for NCT. Conversely, G2- and G4-PAMAM do not retain in the lymphatic vessels, while G8-PAMAM is too large for rapid uptake. Based on these results, it was determined that gadolinium-labeled G6-PAMAM may simultaneous image and treat primary tumors or micro-metastasis in the sentinel lymph nodes. 7. Photothermal therapy With the advent of metal nano-particles during the 1990's, photothermal ablation has burgeoned into a new niche of minimally invasive tumor therapies [88–91]. Gold-based nanoparticles have been developed that strongly absorb light in the near-infrared region, facilitating deep optical penetration into tissues, generating a localized lethal dose of heat at the site of a tumor [92]. The first methods for preparing metal-encapsulated dendrimers for use in biomedical applications were reported within the last decade with the goal of adding a finer degree of control for tuning the biological interactions elicited by the metal particles, including improved biocompatibility, retention, and ease of surface modification for potential use as biomarkers, contrast agents, and for photothermal therapy [92–94]. Only within the last year, dendrimer-encapsulated gold nanoparticles have been prepared and identified for their potential use towards the photothermal treatment of malignant tissue (Fig. 11). In one study, amine-terminated G5-PAMAM dendrimer-entrapped gold nano-particles were prepared and covalently conjugated to flourescein and folic acid for targeted delivery to tumor cells overexpressing folic acid receptors, as reported by the Baker group [95]. The dendrimers were shown to specifically bind to KB cells in vitro and were internalized into lysosomes within 2 h. The applicability of these particles for targeted hyperthermia treatment and as electron-dense contrast agents was recognized and in vivo performance studies are currently underway. The photothermal properties of gold-encapsulated PEGylated and non-PEGylated G4-PAMAM dendrimers as reported J.B. Wolinsky, M.W. Grinstaff / Advanced Drug Delivery Reviews 60 (2008) 1037–1055 1051 Fig. 11. Photothermal therapy. Conceptual representation of photothermal therapy using dendrimer-entrapped gold nano-particles. The nano-particles would be targeted to tumor cells via folic acid receptors, and upon exposure to near-infrared light, the gold particles would emit heat and kill the host cell (adapted from [96]). by Haba et. al., were evaluated and compared to conventionallyused gold nano-particles prepared with sodium citrate [96]. Gold was encapsulated by first introducing HAuCl4 and then chemically reducing the gold inside the dendrimers. PEGylated gold-encapsulated dendrimers demonstrated superior photostability compared to non-PEGylated, with the non-PEGylated absorbance decreasing to almost negligible levels by three days, whereas the PEGylated nano-particle absorbance was relatively unchanged over five days. It was noted that the non-PEGylated gold particles tended to aggregate. The photothermal properties of the PEGylated particles were only slightly lower compared to conventional gold particles. Future work for these studies will include a biological evaluation and attempts to extend the absorption spectra to the IR region. 8. Imaging Imaging modalities can be used in oncology to diagnose, locate, stage, plan treatment, and potentially find recurrence. Computed tomography (CT) and magnetic resonance imaging (MRI) are two standard methods of imaging associated with cancer diagnoses. Gadolinium (Gd) paramagnetic contrast agents for MRI have been complexed with dendrimer molecules over the last two decades for contrast enhancement, improved clearance characteristics, and potential targeting [42,97–97]. Gd-labeled PAMAM systems have been used for visualizing both tumor vasculature and lymphatic involvement. Changes in tumor permeability were visualized by magnetic resonance imaging using G8-Gd-PAMAM contrast agents after a single large dose of radiation treatment [100]. It was found that vessel permeability was temporarily enhanced in SCCVII tumors after radiation, possibly attributed to vessel leakiness resulting from decreased tumor interstitial pressure, or increases in vascular permeability factor or vascular endothelial growth factor [101,102]. These results suggest a new method of optimizing the use of concurrent therapies based on temporarily enhanced permeability of tumor vasculature to anti-cancer macromolecules. Micromagnetic resonance lymphangiography with G6-Gd-PAMAM was assessed in mice bearing hematomas to improve the contrast between intralymphatic and extralymphatic imaging [103]. A more accurate characterization of lymphoma could lead to increases in care as extralymphatic involvement may change the course of the chemotherapy regime. High spatial and temporal resolutions were obtained and the functional anatomy of the lymphatic system could be three-dimensionally imaged, defining both normal and abnormal lymphatics and distinguishing between intralymphatic and extralymphatic involvement. As mentioned earlier, Gd-labeled G6-PAMAM was also shown to accumulate in the sentinel lymph nodes, which are routinely imaged before surgery for breast cancer and melanoma (Fig. 12) [87]. Gadolinium contrast agents have been conjugated to PPI and evaluated for use as macromolecular MR contrast agents [104]. Higher generations of Gd-PPI (G3 and G5) had lower limits of detection compared to G0 and G1 agents but showed more gradual diffusion into tumors. Nonetheless, non-specific imaging of sub-millimeter-sized blood vessels was achieved regardless of dendrimer generation. The synthesis of polydiamidoproponoate-peptide nucleic acid assemblies with chelating dendrimer branches for enhanced magnetic resonance imaging of oncogene mRNAs in tumor cells by hybridization of complementary oligonucleotides was reported by Amirkhanov et al. [105]. In this manner, multiple contrast paramagnetic ions, such as gadolinium, may be complexed to the probe, thus enhancing the signal generated from cells with too few oncogene mRNAs. Similarly to Gd-dendrimer conjugates for MRI, iodinated contrast agents used for computer tomography (CT) could benefit from dendrimer conjugation with improved retention times and the potential for targeted delivery. Fu et al. reported the synthesis of a set of iodinated contrast agents based on iobitridol-conjugated G3–G5 poly(lysine) dendrimers with PEG cores of varying lengths for possible tumor microvasculature CT imaging [106]. A representative dendrimer, G4-poly (lysine) with PEG12000 core, showed strong visualization in vivo of normal rat vasculature with a monoexponential blood half life of about 35 min, compared to more typical exhaustion times of 5 min for small molecule CT contrast agents. Retention times were attributed to the large molecular weight of the 1052 J.B. Wolinsky, M.W. Grinstaff / Advanced Drug Delivery Reviews 60 (2008) 1037–1055 Fig. 12. Visualization of sentinel lymph node via gadolinium-labeled PAMAM. Rapid and enhanced visualization of the sentinel lymph node using magnetic resonance lymphangiography is achieved by optimizing the size of gadolinium-labeled PAMAM contrast agents (not to scale, adapted from [87]). dendrimer causing it to remain in the blood pool for longer durations. The use of fluorescent probes for tumor detection offers the advantage of improved biocompatibility compared to other types of contrast agents, but suffers from the poor penetration of light through tissue. To provide a minimally invasive solution to this obstacle, Thomas et al. prepared G5-PAMAM dendrimers conjugated to folic acid and a fluorescent probe (6-TAMRA, 6T) which were subsequently targeted to tumors in vivo and detected by a two-photon optical probe [107]. Human squamous KB cell tumors overexpressing folic acid receptors (FAR) were grown in vivo in mice and subsequently targeted by the dendrimer probes. An eight-fold increase in tumor accumulation was observed of targeted dendrimer (G5-6TPAMAM-FA) compared to non-targeted dendrimer (G5-6TPAMAM), and a three-fold increase in fluorescence was seen in FAR-positive KB cell tumors compared to FAR-negative MCA207 tumors. 9. Conclusions The culmination of these advances in dendrimer-based delivery systems along with fundamental work performed over the last couple decades has led to the founding of several start-up companies and a large number of patents focused, at least in part, in the development of dendrimer technologies for oncology applications [108–111]. For example, Starpharma is currently pursuing the development of drug-dendrimer conjugates as cancer therapeutics in order to ease administration and improve safety for patients (www.starpharma.com). The importance of the high architectural control characteristic of dendrimers has been increasingly supported by positive outcomes from in vitro and in vivo studies. Clinically-relevant carrier properties are being facilitated by controlling charge and functionality through choice of peripheral groups, and size through generation number and PEGylation. A recent example of such an architecturally- optimized dendritic drug delivery system, which highlights these principles, was reported using an asymmetric doxorubicinfunctionalized bow-tie dendrimer composed of PEG and 2,2-bis (hydroxymethyl)propionic acid [25,40]. The systematic exploration of these properties has been possible as a result of the growing number of variations in structure achieved by the collaborative efforts of many groups. A limitation that remains is the diversity of release mechanisms and the range of release kinetics for dendrimer-based drug delivery platforms. Dendrimer-encapsulated drugs tend to release rapidly, expelling their payload prematurely before the macromolecules can reach a target site, whereas the release of chemotherapeutics from drugdendrimer conjugates relies primarily on the chemical linkage connecting the drug to the dendrimer periphery. The continued development and exploration of functional or responsive dendrimers as well as environmentally sensitive linkages will offer solutions to optimized drug release. Another area that will need continued research is in the area of targeting to a specific cancer. This is realized and several groups are identifying unique peptides, antibodies, or aptamers using combinatorial screening techniques (e.g., phage display). The multi-valency of dendritic macromolecules may also be important in the design of personalized therapies. Patients respond differently to therapeutics, thus the selection of drug and reliance on receptor-mediated targeting must be considered for each individual. One can envision a customized multi-valent dendritic carrier that supports one or multiple drugs, a targeting moiety, a contrast agent to visualize delivery, and a diagnostic sensor to detect the extent of inflicted cell death. Studies have indeed demonstrated that a well-designed dendrimer structure can be potentially tuned simultaneously for desired biocompatibility, bioavailability, pharmacokinetics, and localized delivery of therapeutics to malignancies. Continued research in the area will bring compositions and architectures tailored for increasing specificity and efficacy towards the diagnoses and treatment of cancer in the clinic. J.B. Wolinsky, M.W. Grinstaff / Advanced Drug Delivery Reviews 60 (2008) 1037–1055 Acknowledgements The authors wish to thank Boston University for its support. [21] References [22] [1] C.C. Lee, J.A. MacKay, J.M. Fréchet, F.C. Szoka, Designing dendrimers for biological applications, Nat. Biotechnol. 23 (2005) 1517–1526. [2] H. Yang, W.J. Kao, Dendrimers for pharmaceutical and biomedical applications, J. Biomater. Sci. Polym. Ed. 17 (2006) 3–19. [3] S. Svenson, D.A. Tomalia, Dendrimers in biomedical applications — reflections on the field, Adv. Drug Deliv. Rev. 57 (2005) 2106–2129. [4] M.K. Khan, S.S. Nigavekar, L.D. Minc, M.S. Kariapper, B.M. Nair, W.G. Lesniak, L.P. Balogh, In vivo biodistribution of dendrimers and dendrimer nanocomposites — implications for cancer imaging and therapy, Technol. Cancer Res. Treat 4 (2005) 603–613. [5] D.A. Tomalia, L.A. Reyna, S. Svenson, Dendrimers as multi-purpose nanodevices for oncology drug delivery and diagnostic imaging, Biochem. Soc. Trans. 35 (2007) 61–67. [6] E.R. Gillies, J.M. Fréchet, Dendrimers and dendritic polymers in drug delivery, Drug Discov. Today 10 (2005) 35–43. [7] D.A. Tomalia, H. Baker, J. Dewald, M. Hall, G. Kallos, S. Martin, J. Roeck, J. Ryder, P. Smith, Dendritic macromolecules: synthesis of starburst dendrimers, Macromolecules 19 (1986) 2466–2468. [8] D.A. Tomalia, H. Baker, J. Dewald, M. Hall, G. Kallos, S. Martin, J. Roeck, J. Ryder, P. Smith, A new class of polymers: starburst-dendritic macromolecules, Polym. J. 17 (1985) 117–132. [9] G.R. Newkome, Z. Yao, G.R. Baker, V.K. Gupta, Cascade molecules: a new approach to micelles. A [27]-Arborol. J. Org. Chem. 50 (1985) 2003. [10] C.J. Hawker, J.M. Fréchet, Preparation of polymers with controlled molecular architecture. A new convergent approach to dendritic macromolecules. J. Am. Chem. Soc. 112 (1990) 7638–7647. [11] Z.Y. Zhang, B.D. Smith, High-generation polycationic dendrimers are unusually effective at disrupting anionic vesicles: membrane bending model, Bioconjug. Chem. 11 (2000) 805–814. [12] N. Karoonuthasiri, K. Titiyevskiy, J.L. Thomas, Destabilization of fatty acid-containing liposomes by polyamidoamine dendrimers. Colloid Surf. B 27 (2003) 356–375. [13] A. Mecke, S. Uppuluri, T.M. Sassanella, D.K. Lee, A. Ramamoorthy, J.R. Baker Jr., B.G. Orr, M.M. Banaszak Holl, Direct observation of lipid bilayer disruption by poly(amidoamine) dendrimers, Chem. Phys. Lipids 132 (2004) 3–14. [14] S. Uppuluri, L. Piehler, J. Li, G.L. Hagnauer, D.A. Tomalia, Core–shell tecto(dendrimers): 1. Synthesis and characterization of saturated shell models, Adv. Mater. 12 (2000) 796–800. [15] A. Mecke, I.J. Majoros, A.K. Patri, J.R. Baker Jr., M.M. Holl, B.G. Orr, Lipid bilayer disruption by polycationic polymers: the roles of size and chemical functional group, Langmuir 21 (2005) 10348–10354. [16] A. Mecke, D.K. Lee, A. Ramamoorthy, B.G. Orr, M.M. Holl, Synthetic and natural polycationic polymer nanoparticles interact selectively with fluid-phase domains of DMPC lipid bilayers, Langmuir 21 (2005) 8588–8590. [17] S. Hong, P.R. Leroueil, E.K. Janus, J.L. Peters, M.M. Kober, M.T. Islam, B.G. Orr, J.R. Baker Jr., M.M. Banaszak Holl, Interaction of polycationic polymers with supported lipid bilayers and cells: nanoscale hole formation and enhanced membrane permeability, Bioconjug. Chem. 17 (2006) 728–734. [18] F.P. Seib, A.T. Jones, R. Duncan, Comparison of the endocytic properties of linear and branched PEIs, and cationic PAMAM dendrimers in B16f10 melanoma cells, J. Control. Release 117 (2007) 291–300. [19] S. Hong, A.U. Bielinska, A. Mecke, B. Keszler, J.L. Beals, X. Shi, L. Balogh, B.G. Orr, J.R. Baker Jr., M.M. Banaszak Holl, Interaction of poly (amidoamine) dendrimers with supported lipid bilayers and cells: hole formation and the relation to transport, Bioconjug. Chem. 15 (2004) 774–782. [20] M.S. Diallo, S. Christie, P. Swaminathan, L. Balogh, X. Shi, W. Um, C. Papelis, W.A. Goddard 3rd, J.H. Johnson Jr., Dendritic chelating agents. [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] 1053 1. Cu(II) binding to ethylene diamine core poly(amidoamine) dendrimers in aqueous solutions, Langmuir 20 (2004) 2640–2651. S. Fuchs, T. Kapp, H. Otto, T. Schoneberg, P. Franke, R. Gust, A.D. Schluter, A surface-modified dendrimer set for potential application as drug delivery vehicles: synthesis, in vitro toxicity, and intracellular localization, Chemistry 10 (2004) 1167–1192. H.T. Chen, M.F. Neerman, A.R. Parrish, E.E. Simanek, Cytotoxicity, hemolysis, and acute in vivo toxicity of dendrimers based on melamine, candidate vehicles for drug delivery, J. Am. Chem. Soc. 126 (2004) 10044–10048. M.F. Neerman, W. Zhang, A.R. Parrish, E.E. Simanek, In vitro and in vivo evaluation of a melamine dendrimer as a vehicle for drug delivery, Int. J. Pharm. 281 (2004) 129–132. J.C. Roberts, M.K. Bhalgat, R.T. Zera, Preliminary biological evaluation of polyamidoamine (PAMAM) Starburst dendrimers, J. Biomed. Mater. Res. 30 (1996) 53–65. E.R. Gillies, E. Dy, J.M. Fréchet, F.C. Szoka, Biological evaluation of polyester dendrimer: poly(ethylene oxide) “bow-tie” hybrids with tunable molecular weight and architecture, Mol. Pharm. 2 (2005) 129–138. T. Yamaoka, Y. Tabata, Y. Ikada, Distribution and tissue uptake of poly (ethylene glycol) with different molecular weights after intravenous administration to mice, J. Pharm. Sci. 83 (1994) 601–606. T. Okuda, S. Kawakami, T. Maeie, T. Niidome, F. Yamashita, M. Hashida, Biodistribution characteristics of amino acid dendrimers and their PEGylated derivatives after intravenous administration, J. Control. Release 114 (2006) 69–77. T. Okuda, S. Kawakami, N. Akimoto, T. Niidome, F. Yamashita, M. Hashida, PEGylated lysine dendrimers for tumor-selective targeting after intravenous injection in tumor-bearing mice, J. Control. Release 116 (2006) 330–336. S.S. Nigavekar, L.Y. Sung, M. Llanes, A. El-Jawahri, T.S. Lawrence, C.W. Becker, L. Balogh, M.K. Khan, 3H dendrimer nanoparticle organ/ tumor distribution, Pharm. Res. 21 (2004) 476–483. T.M. Allen, P.R. Cullis, Drug delivery systems: entering the mainstream, Science 303 (2004) 1818–1822. M.T. Morgan, M.A. Carnahan, C.E. Immoos, A.A. Ribeiro, S. Finkelstein, S.J. Lee, M.W. Grinstaff, Dendritic molecular capsules for hydrophobic compounds, J. Am. Chem. Soc. 125 (2003) 15485–15489. M.T. Morgan, M.A. Carnahan, S. Finkelstein, C.A. Prata, L. Degoricija, S.J. Lee, M.W. Grinstaff, Dendritic supramolecular assemblies for drug delivery, Chem. Commun. (Camb.) (2005) 4309–4311. M.T. Morgan, Y. Nakanishi, D.J. Kroll, A.P. Griset, M.A. Carnahan, M. Wathier, N.H. Oberlies, G. Manikumar, M.C. Wani, M.W. Grinstaff, Dendrimer-encapsulated camptothecins: increased solubility, cellular uptake, and cellular retention affords enhanced anticancer activity in vitro, Cancer Res. 66 (2006) 11913–11921. F. Wang, T.K. Bronich, A.V. Kabanov, R.D. Rauh, J. Roovers, Synthesis and evaluation of a star amphiphilic block copolymer from poly(epsiloncaprolactone) and poly(ethylene glycol) as a potential drug delivery carrier, Bioconjug. Chem. 16 (2005) 397–405. T. Ooya, J. Lee, K. Park, Hydrotropic dendrimers of generations 4 and 5: synthesis, characterization, and hydrotropic solubilization of paclitaxel, Bioconjug. Chem. 15 (2004) 1221–1229. M.F. Neerman, H.T. Chen, A.R. Parrish, E.E. Simanek, Reduction of drug toxicity using dendrimers based on melamine, Mol. Pharm. 1 (2004) 390–393. S. Gurdag, J. Khandare, S. Stapels, L.H. Matherly, R.M. Kannan, Activity of dendrimer-methotrexate conjugates on methotrexate-sensitive and -resistant cell lines, Bioconjug. Chem. 17 (2006) 275–283. J.J. Khandare, S. Jayant, A. Singh, P. Chandna, Y. Wang, N. Vorsa, T. Minko, Dendrimer versus linear conjugate: influence of polymeric architecture on the delivery and anticancer effect of paclitaxel, Bioconjug. Chem. 17 (2006) 1464–1472. A. Papagiannaros, K. Dimas, G.T. Papaioannou, C. Demetzos, Doxorubicin-PAMAM dendrimer complex attached to liposomes: cytotoxic studies against human cancer cell lines, Int. J. Pharm. 302 (2005) 29–38. C.C. Lee, E.R. Gillies, M.E. Fox, S.J. Guillaudeu, J.M. Fréchet, E.E. Dy, F.C. Szoka, A single dose of doxorubicin-functionalized bow-tie 1054 [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] J.B. Wolinsky, M.W. Grinstaff / Advanced Drug Delivery Reviews 60 (2008) 1037–1055 dendrimer cures mice bearing C-26 colon carcinomas, Proc. Natl. Acad. Sci. U. S. A. 103 (2006) 16649–16654. A.K. Iyer, G. Khaled, J. Fang, H. Maeda, Exploiting the enhanced permeability and retention effect for tumor targeting, Drug Discov. Today 11 (2006) 812–818. E.C. Wiener, S. Konda, A. Shadron, M. Brechbiel, O. Gansow, Targeting dendrimer-chelates to tumors and tumor cells expressing the high-affinity folate receptor, Invest. Radiol. 32 (1997) 748–754. A. Quintana, E. Raczka, L. Piehler, I. Lee, A. Myc, I. Majoros, A.K. Patri, T. Thomas, J. Mule, J.R. Baker Jr., Design and function of a dendrimerbased therapeutic nanodevice targeted to tumor cells through the folate receptor, Pharm. Res. 19 (2002) 1310–1316. K. Kono, M. Liu, J.M. Fréchet, Design of dendritic macromolecules containing folate or methotrexate residues, Bioconjug. Chem. 10 (1999) 1115–1121. S. Hong, P.R. Leroueil, I.J. Majoros, B.G. Orr, J.R. Baker Jr., M.M. Banaszak Holl, The binding avidity of a nanoparticle-based multivalent targeted drug delivery platform, Chem. Biol. 14 (2007) 107–115. Y. Choi, T. Thomas, A. Kotlyar, M.T. Islam, J.R. Baker Jr., Synthesis and functional evaluation of DNA-assembled polyamidoamine dendrimer clusters for cancer cell-specific targeting, Chem. Biol. 12 (2005) 35–43. Y. Choi, J.R. Baker Jr., Targeting cancer cells with DNA-assembled dendrimers: a mix and match strategy for cancer, Cell Cycle 4 (2005) 669–671. A.K. Patri, J.F. Kukowska-Latallo, J.R. Baker Jr., Targeted drug delivery with dendrimers: comparison of the release kinetics of covalently conjugated drug and non-covalent drug inclusion complex, Adv. Drug Deliv. Rev. 57 (2005) 2203–2214. I.J. Majoros, T.P. Thomas, C.B. Mehta, J.R. Baker Jr., Poly(amidoamine) dendrimer-based multifunctional engineered nanodevice for cancer therapy, J. Med. Chem. 48 (2005) 5892–5899. T.P. Thomas, I.J. Majoros, A. Kotlyar, J.F. Kukowska-Latallo, A. Bielinska, A. Myc, J.R. Baker Jr., Targeting and inhibition of cell growth by an engineered dendritic nanodevice, J. Med. Chem. 48 (2005) 3729–3735. I.J. Majoros, A. Myc, T. Thomas, C.B. Mehta, J.R. Baker Jr., PAMAM dendrimer-based multifunctional conjugate for cancer therapy: synthesis, characterization, and functionality, Biomacromolecules 7 (2006) 572–579. A. Myc, I.J. Majoros, T.P. Thomas, J.R. Baker Jr., Dendrimer-based targeted delivery of an apoptotic sensor in cancer cells, Biomacromolecules 8 (2007) 13–18. R. Shukla, T.P. Thomas, J. Peters, A. Kotlyar, A. Myc, J.R. Baker Jr, Tumor angiogenic vasculature targeting with PAMAM dendrimer-RGD conjugates, Chem. Commun. (Camb.) (2005) 5739–5741. I. Dijkgraaf, A.Y. Rijnders, A. Soede, A.C. Dechesne, G.W. van Esse, A. J. Brouwer, F.H. Corstens, O.C. Boerman, D.T. Rijkers, R.M. Liskamp, Synthesis of DOTA-conjugated multivalent cyclic-RGD peptide dendrimers via 1,3-dipolar cycloaddition and their biological evaluation: implications for tumor targeting and tumor imaging purposes, Org. Biomol. Chem. 5 (2007) 935–944. A.K. Patri, A. Myc, J. Beals, T.P. Thomas, N.H. Bander, J.R. Baker Jr., Synthesis and in vitro testing of J591 antibody–dendrimer conjugates for targeted prostate cancer therapy, Bioconjug. Chem. 15 (2004) 1174–1181. S.S. Chang, D.S. O'Keefe, D.J. Bacich, V.E. Reuter, W.D. Heston, P.B. Gaudin, Prostate-specific membrane antigen is produced in tumorassociated neovasculature, Clin. Cancer Res. 5 (1999) 2674–2681. R. Shukla, T.P. Thomas, J.L. Peters, A.M. Desai, J. Kukowska-Latallo, A. K. Patri, A. Kotlyar, J.R. Baker Jr., HER2 specific tumor targeting with dendrimer conjugated anti-HER2 mAb, Bioconjug. Chem. 17 (2006) 1109–1115. D.J. Slamon, W. Godolphin, L.A. Jones, J.A. Holt, S.G. Wong, D.E. Keith, W.J. Levin, S.G. Stuart, J. Udove, A. Ullrich, et al., Studies of the HER-2/neu proto-oncogene in human breast and ovarian cancer, Science 244 (1989) 707–712. T.P. Thomas, A.K. Patri, A. Myc, M.T. Myaing, J.Y. Ye, T.B. Norris, J.R. Baker Jr., In vitro targeting of synthesized antibody-conjugated dendrimer nanoparticles, Biomacromolecules 5 (2004) 2269–2274. [60] G. Wu, R.F. Barth, W. Yang, S. Kawabata, L. Zhang, K. Green-Church, Targeted delivery of methotrexate to epidermal growth factor receptorpositive brain tumors by means of cetuximab (IMC-C225) dendrimer bioconjugates, Mol. Cancer Ther. 5 (2006) 52–59. [61] E.K. Woller, M.J. Cloninger, Mannose functionalization of a sixth generation dendrimer, Biomacromolecules 2 (2001) 1052–1054. [62] R. Roy, M.G. Baek, Glycodendrimers: novel glycotope isosteres unmasking sugar coding. case study with T-antigen markers from breast cancer MUC1 glycoprotein, J. Biotechnol. 90 (2002) 291–309. [63] D. Lagnoux, T. Darbre, M.L. Schmitz, J.L. Reymond, Inhibition of mitosis by glycopeptide dendrimer conjugates of colchicine, Chemistry 11 (2005) 3941–3950. [64] M. Triesscheijn, P. Baas, J.H. Schellens, F.A. Stewart, Photodynamic therapy in oncology, Oncologist 11 (2006) 1034–1044. [65] T.J. Dougherty, C.J. Gomer, B.W. Henderson, G. Jori, D. Kessel, M. Korbelik, J. Moan, Q. Peng, Photodynamic therapy, J. Natl. Cancer Inst. 90 (1998) 889–905. [66] A.P. Castano, P. Mroz, M.R. Hamblin, Photodynamic therapy and antitumour immunity, Nat. Rev. Cancer 6 (2006) 535–545. [67] N. Nishiyama, H.R. Stapert, G.D. Zhang, D. Takasu, D.L. Jiang, T. Nagano, T. Aida, K. Kataoka, Light-harvesting ionic dendrimer porphyrins as new photosensitizers for photodynamic therapy, Bioconjug. Chem. 14 (2003) 58–66. [68] G.D. Zhang, A. Harada, N. Nishiyama, D.L. Jiang, H. Koyama, T. Aida, K. Kataoka, Polyion complex micelles entrapping cationic dendrimer porphyrin: effective photosensitizer for photodynamic therapy of cancer, J. Control. Release 93 (2003) 141–150. [69] S.H. Battah, C.E. Chee, H. Nakanishi, S. Gerscher, A.J. MacRobert, C. Edwards, Synthesis and biological studies of 5-aminolevulinic acidcontaining dendrimers for photodynamic therapy, Bioconjug. Chem. 12 (2001) 980–988. [70] Q. Peng, K. Berg, J. Moan, M. Kongshaug, J.M. Nesland, 5Aminolevulinic acid-based photodynamic therapy: principles and experimental research, Photochem. Photobiol. 65 (1997) 235–251. [71] G.M. Di Venosa, A.G. Casas, S. Battah, P. Dobbin, H. Fukuda, A.J. Macrobert, A. Batlle, Investigation of a novel dendritic derivative of 5aminolaevulinic acid for photodynamic therapy, Int. J. Biochem. Cell Biol. 38 (2006) 82–91. [72] S. Battah, S. O'Neill, C. Edwards, S. Balaratnam, P. Dobbin, A.J. MacRobert, Enhanced porphyrin accumulation using dendritic derivatives of 5-aminolaevulinic acid for photodynamic therapy: an in vitro study, Int. J. Biochem. Cell Biol. 38 (2006) 1382–1392. [73] S. Battah, S. Balaratnam, A. Casas, S. O'Neill, C. Edwards, A. Batlle, P. Dobbin, A.J. MacRobert, Macromolecular delivery of 5-aminolaevulinic acid for photodynamic therapy using dendrimer conjugates, Mol. Cancer Ther. 6 (2007) 876–885. [74] R. Ideta, F. Tasaka, W.D. Jang, N. Nishiyama, G.D. Zhang, A. Harada, Y. Yanagi, Y. Tamaki, T. Aida, K. Kataoka, Nanotechnology-based photodynamic therapy for neovascular disease using a supramolecular nanocarrier loaded with a dendritic photosensitizer, Nano Lett. 5 (2005) 2426–2431. [75] W.D. Jang, N. Nishiyama, G.D. Zhang, A. Harada, D.L. Jiang, S. Kawauchi, Y. Morimoto, M. Kikuchi, H. Koyama, T. Aida, K. Kataoka, Supramolecular nanocarrier of anionic dendrimer porphyrins with cationic block copolymers modified with polyethylene glycol to enhance intracellular photodynamic efficacy, Angew Chem. Int. Ed. Engl. 44 (2005) 419–423. [76] C. Kojima, Y. Toi, A. Harada, K. Kono, Preparation of poly(ethylene glycol)-attached dendrimers encapsulating photosensitizers for application to photodynamic therapy, Bioconjug. Chem. 18 (2007) 663–670. [77] R.F. Barth, J.A. Coderre, M.G. Vicente, T.E. Blue, Boron neutron capture therapy of cancer: current status and future prospects, Clin. Cancer Res. 11 (2005) 3987–4002. [78] G. Wu, R.F. Barth, W. Yang, R.J. Lee, W. Tjarks, M.V. Backer, J.M. Backer, Boron containing macromolecules and nanovehicles as delivery agents for neutron capture therapy, Anticancer Agents Med. Chem. 6 (2006) 167–184. [79] W. Yang, R.F. Barth, D.M. Adams, A.H. Soloway, Intratumoral delivery of boronated epidermal growth factor for neutron capture therapy of brain tumors, Cancer Res. 57 (1997) 4333–4339. J.B. Wolinsky, M.W. Grinstaff / Advanced Drug Delivery Reviews 60 (2008) 1037–1055 [80] W. Yang, R.F. Barth, D.M. Adams, M.J. Ciesielski, R.A. Fenstermaker, S. Shukla, W. Tjarks, M.A. Caligiuri, Convection-enhanced delivery of boronated epidermal growth factor for molecular targeting of EGF receptor-positive gliomas, Cancer Res. 62 (2002) 6552–6558. [81] H. Kobayashi, S. Kawamoto, T. Saga, N. Sato, T. Ishimori, J. Konishi, K. Ono, K. Togashi, M.W. Brechbiel, Avidin-dendrimer-(1B4M-Gd)(254): a tumor-targeting therapeutic agent for gadolinium neutron capture therapy of intraperitoneal disseminated tumor which can be monitored by MRI, Bioconjug. Chem. 12 (2001) 587–593. [82] R.F. Barth, D.M. Adams, A.H. Soloway, F. Alam, M.V. Darby, Boronated starburst dendrimer-monoclonal antibody immunoconjugates: evaluation as a potential delivery system for neutron capture therapy, Bioconjug. Chem. 5 (1994) 58–66. [83] G. Wu, R.F. Barth, W. Yang, M. Chatterjee, W. Tjarks, M.J. Ciesielski, R. A. Fenstermaker, Site-specific conjugation of boron-containing dendrimers to anti-EGF receptor monoclonal antibody cetuximab (IMC-C225) and its evaluation as a potential delivery agent for neutron capture therapy, Bioconjug. Chem. 15 (2004) 185–194. [84] G. Wu, W. Yang, R.F. Barth, S. Kawabata, M. Swindall, A.K. Bandyopadhyaya, W. Tjarks, B. Khorsandi, T.E. Blue, A.K. Ferketich, M. Yang, G.A. Christoforidis, T.J. Sferra, P.J. Binns, K.J. Riley, M.J. Ciesielski, R.A. Fenstermaker, Molecular targeting and treatment of an epidermal growth factor receptor-positive glioma using boronated cetuximab, Clin. Cancer Res. 13 (2007) 1260–1268. [85] W. Yang, R.F. Barth, G. Wu, S. Kawabata, T.J. Sferra, A.K. Bandyopadhyaya, W. Tjarks, A.K. Ferketich, M.L. Moeschberger, P.J. Binns, K.J. Riley, J.A. Coderre, M.J. Ciesielski, R.A. Fenstermaker, C.J. Wikstrand, Molecular targeting and treatment of EGFRvIII-positive gliomas using boronated monoclonal antibody L8A4, Clin. Cancer Res. 12 (2006) 3792–3802. [86] M.V. Backer, T.I. Gaynutdinov, V. Patel, A.K. Bandyopadhyaya, B.T. Thirumamagal, W. Tjarks, R.F. Barth, K. Claffey, J.M. Backer, Vascular endothelial growth factor selectively targets boronated dendrimers to tumor vasculature, Mol. Cancer Ther. 4 (2005) 1423–1429. [87] H. Kobayashi, S. Kawamoto, M. Bernardo, M.W. Brechbiel, M.V. Knopp, P.L. Choyke, Delivery of gadolinium-labeled nanoparticles to the sentinel lymph node: comparison of the sentinel node visualization and estimations of intra-nodal gadolinium concentration by the magnetic resonance imaging, J. Control. Release 111 (2006) 343–351. [88] C. Loo, A. Lowery, N. Halas, J. West, R. Drezek, Immunotargeted nanoshells for integrated cancer imaging and therapy, Nano Lett. 5 (2005) 709–711. [89] M. Everts, V. Saini, J.L. Leddon, R.J. Kok, M. Stoff-Khalili, M.A. Preuss, C.L. Millican, G. Perkins, J.M. Brown, H. Bagaria, D.E. Nikles, D.T. Johnson, V.P. Zharov, D.T. Curiel, Covalently linked Au nanoparticles to a viral vector: potential for combined photothermal and gene cancer therapy, Nano Lett. 6 (2006) 587–591. [90] I.H. El-Sayed, X. Huang, M.A. El-Sayed, Selective laser photo-thermal therapy of epithelial carcinoma using anti-EGFR antibody conjugated gold nanoparticles, Cancer Lett. 239 (2006) 129–135. [91] X. Huang, P.K. Jain, I.H. El-Sayed, M.A. El-Sayed, Determination of the minimum temperature required for selective photothermal destruction of cancer cells with the use of immunotargeted gold nanoparticles, Photochem. Photobiol. 82 (2006) 412–417. [92] L.R. Hirsch, A.M. Gobin, A.R. Lowery, F. Tam, R.A. Drezek, N.J. Halas, J.L. West, Metal nanoshells, Ann. Biomed. Eng. 34 (2006) 15–22. [93] W. Lesniak, A.U. Bielinska, K. Sun, K.W. Janczak, X. Shi, J.R. Baker Jr., L.P. Balogh, Silver/dendrimer nanocomposites as biomarkers: fabrication, characterization, in vitro toxicity, and intracellular detection, Nano Lett. 5 (2005) 2123–2130. [94] X. Shi, S. Wang, H. Sun, J.R. Baker Jr., Improved biocompatibility of surface functionalized dendrimer-entrapped gold nanoparticles, Soft Matter 3 (2007) 71–74. 1055 [95] X. Shi, S. Wang, S. Meshinchi, M.E. Van Antwerp, X. Bi, I. Lee, J.R. Baker Jr, Dendrimer-entrapped gold nanoparticles as a platform for cancer-cell targeting and imaging, Small 3 (2007) 1245–1252. [96] Y. Haba, C. Kojima, A. Harada, T. Ura, H. Horinaka, K. Kono, Preparation of poly(ethylene glycol)-modified poly(amido amine) dendrimers encapsulating gold nanoparticles and their heat-generating ability, Langmuir 23 (2007) 5243–5246. [97] E.C. Wiener, M.W. Brechbiel, H. Brothers, R.L. Magin, O.A. Gansow, D.A. Tomalia, P.C. Lauterbur, Dendrimer-based metal chelates: a new class of magnetic resonance imaging contrast agents, Magn. Reson. Med. 31 (1994) 1–8. [98] M.W. Bourne, L. Margerun, N. Hylton, B. Campion, J.J. Lai, N. Derugin, C.B. Higgins, Evaluation of the effects of intravascular MR contrast media (gadolinium dendrimer) on 3D time of flight magnetic resonance angiography of the body, J. Magn. Reson. Imaging 6 (1996) 305–310. [99] H. Kobayashi, N. Sato, S. Kawamoto, T. Saga, A. Hiraga, T.L. Haque, T. Ishimori, J. Konishi, K. Togashi, M.W. Brechbiel, Comparison of the macromolecular MR contrast agents with ethylenediamine-core versus ammonia-core generation-6 polyamidoamine dendrimer, Bioconjug. Chem. 12 (2001) 100–107. [100] H. Kobayashi, K. Reijnders, S. English, A.T. Yordanov, D.E. Milenic, A. L. Sowers, D. Citrin, M.C. Krishna, T.A. Waldmann, J.B. Mitchell, M.W. Brechbiel, Application of a macromolecular contrast agent for detection of alterations of tumor vessel permeability induced by radiation, Clin. Cancer Res. 10 (2004) 7712–7720. [101] H.F. Dvorak, Vascular permeability factor/vascular endothelial growth factor: a critical cytokine in tumor angiogenesis and a potential target for diagnosis and therapy, J. Clin. Oncol. 20 (2002) 4368–4380. [102] C.A. Znati, M. Rosenstein, Y. Boucher, M.W. Epperly, W.D. Bloomer, R.K. Jain, Effect of radiation on interstitial fluid pressure and oxygenation in a human tumor xenograft, Cancer Res. 56 (1996) 964–968. [103] H. Kobayashi, S. Kawamoto, M.W. Brechbiel, M. Bernardo, N. Sato, T. A. Waldmann, Y. Tagaya, P.L. Choyke, Detection of lymph node involvement in hematologic malignancies using micromagnetic resonance lymphangiography with a gadolinum-labeled dendrimer nanoparticle, Neoplasia 7 (2005) 984–991. [104] S. Langereis, Q.G. de Lussanet, M.H. van Genderen, E.W. Meijer, R.G. Beets-Tan, A.W. Griffioen, J.M. van Engelshoven, W.H. Backes, Evaluation of Gd(III)DTPA-terminated poly(propylene imine) dendrimers as contrast agents for MR imaging, NMR Biomed. 19 (2006) 133–141. [105] N.V. Amirkhanov, E. Wickstrom, Synthesis of novel polydiamidopropanoate dendrimer PNA-peptide chimeras for non-invasive magnetic resonance imaging of cancer, Nucleosides Nucleotides Nucleic Acids 24 (2005) 423–426. [106] Y. Fu, D.E. Nitecki, D. Maltby, G.H. Simon, K. Berejnoi, H.J. Raatschen, B.M. Yeh, D.M. Shames, R.C. Brasch, Dendritic iodinated contrast agents with PEG-cores for CT imaging: synthesis and preliminary characterization, Bioconjug. Chem. 17 (2006) 1043–1056. [107] T.P. Thomas, M.T. Myaing, J.Y. Ye, K. Candido, A. Kotlyar, J. Beals, P. Cao, B. Keszler, A.K. Patri, T.B. Norris, J.R. Baker Jr., Detection and analysis of tumor fluorescence using a two-photon optical fiber probe, Biophys. J. 86 (2004) 3959–3965. [108] N. Malik, R. Duncan, D.A. Tomalia, R. Esfand Dendritic-antineoplastic drug delivery system (2001) Pat. No. 7,005,124. [109] J.R. Baker Jr., D.A. Tomalia Multifunctional nanodevice platform (2000) Pat. No. 6,471,968. [110] J. Fréchet, H.R. Ihre Dendrimeric support or carrier macromolecule (2006) Pat. No. 7,097,856. [111] E.S. Handy, R. Ivkov, D. Ellis-Busby, A. Foreman Thermotherapy via targeted delivery of nanoscale magnetic particles (2006) Pat. No. 7,074,175.