Liposome encapsulation of zoledronic acid results in major changes in tissue distribution and increase in toxicity

Contents lists available at SciVerse ScienceDirect

Journal of Controlled Release journal homepage: www.elsevier.com/locate/jconrel

Liposome encapsulation of zoledronic acid results in major changes in tissue distribution and increase in toxicity Hilary Shmeeda a, Yasmine Amitay a, b, Dina Tzemach a, Jenny Gorin a, Alberto Gabizon a, b,⁎ a b

Oncology Institute, Shaare Zedek Medical Center, Jerusalem, Israel Hebrew University-School of Medicine, Jerusalem, Israel

a r t i c l e

i n f o

Article history: Received 23 December 2012 Accepted 6 February 2013 Available online 16 February 2013 Keywords: Liposomes Folate-receptor Targeting Bisphosphonates Cancer Toxicity

a b s t r a c t Background: Zoledronic acid (Zol) is a potent inhibitor of farnesyl-pyrophosphate synthase with broad clinical use in the treatment of osteoporosis, and bone metastases. We have previously shown that encapsulation of Zol in liposomes targeted to the folate receptor (FR) greatly enhances its in vitro cytotoxicity. To examine whether targeted liposomal delivery of Zol could be a useful therapeutic approach, we investigated here the in vivo pharmacology of i.v. administered liposomal Zol (L-Zol) in murine models. Methods: Zol was passively entrapped in the water phase of liposomes containing a small fraction of either dipalmitoyl-phosphatidylglycerol (DPPG) or a polyethylene-glycol (PEG)-conjugated phospholipid with or without insertion of a folate lipophilic conjugate. Radiolabeled formulations were used for pharmacokinetic (PK) and biodistribution studies. Toxicity was evaluated by clinical, hematological, biochemical, and histopathological parameters. Therapeutic studies comparing free Zol, nontargeted and folate targeted L-Zol were performed in FR-expressing human tumor models. Results: Encapsulation of Zol in liposomes resulted in major PK changes including sustained high plasma levels and very slow clearance. DPPG-L-Zol was cleared faster than PEG-L-Zol. Grafting of folate lipophilic conjugates on liposomes further accelerated the clearance of Zol. L-Zol caused a major shift in drug tissue distribution when compared to free Zol, with a major increase (20 to 100-fold) in liver and spleen, a substantial increase (7 to 10-fold) in tumor, and a modest increase (2-fold) in bone. Liposomal formulations proved to be highly toxic, up to 50-fold more than free Zol. PEG-L-Zol was more toxic than DPPG-L-Zol. Toxicity was non-cumulative and appears to involve macrophage/monocyte activation and release of cytokines. Co-injection of L-Zol with a large dose of blank liposomes, or injection of a very low Zol-to-phospholipid ratio liposome formulation reduced toxicity by 2–4-fold suggesting that diluting macrophage exposure below a threshold Zol concentration is important to overcome toxicity. L-Zol failed to significantly enhance the therapeutic activity of Zol vis-à-vis free ZOL and doxorubicin. Folate-targeted L-Zol was marginally better than other treatment modalities in the KB tumor model but toxic deaths greatly affected the outcome. Conclusions: Liposome delivery of Zol causes a major change in tissue drug distribution and an increase in tumor Zol levels. However, the severe in vivo toxicity of L-Zol seriously limits its dose and its utility for in vivo tumor cell targeting. This strategy is under evaluation using liposomes carrying less toxic bisphosphonates. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Bisphosphonates are non-hydrolyzable pyrophosphate analogs, used to treat Paget's disease, osteoporosis, hypercalcemia of malignancy, and metastatic bone disease. Recent work has demonstrated anti-tumor and anti-angiogenic effects of a potent subclass known as amino- or nitrogen-containing bisphosphonates (NBP) which inhibit the mevalonate pathway and block post-translational prenylation of

⁎ Corresponding author at: Shaare Zedek MC, Oncology Institute, POB 3235, Jerusalem 91031, Israel. Tel.: +972 2 6555036; fax: +972 2 6555080. E-mail address: [email protected] (A. Gabizon). 0168-3659/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jconrel.2013.02.003

GTP-binding proteins such as Ras [1–3]. In addition, NBP appear to stimulate the expansion and anti-tumor activity of γδ (gammadelta) T lymphocytes, particularly a subset only found in primates and known as Vγ9Vδ2, via a cellular accumulation of phospho-antigens derived from mevalonate metabolites [4–6]. One of these NBP, zoledronic acid (Zol) is a potent inhibitor of farnesyl pyrophosphate synthase at nanomolar concentrations, interfering with the mevalonate pathway and related critical processes in cell signaling and growth [7,8]. Recent clinical studies have demonstrated effective prevention of disease recurrence in premenopausal breast cancer when zoledronic acid is given together with adjuvant chemo-hormonal therapy after resection of the primary tumor [9]. Because of its high water-solubility, and poor cell permeability [10,11], we hypothesized that the full antitumor potential

NANOMEDICINE

Journal of Controlled Release 167 (2013) 265–275

NANOMEDICINE

266

H. Shmeeda et al. / Journal of Controlled Release 167 (2013) 265–275

of Zol is not fully exploited and that an intracellular drug delivery mechanism should substantially facilitate the cytotoxic potential of Zol. Previously, we have demonstrated that liposome encapsulation of Zol (L-Zol) and targeting to the folate receptor (FR) results in a 200to 400-fold gain in cytotoxicity when compared to free ZOL and to the non-targeted formulation respectively in a variety of FR-expressing cell lines [11]. The cytotoxic potency of folate-targeted L-Zol (FTL-Zol) was equivalent on a molar basis to that of other powerful chemotherapeutic agents. Thus, targeted liposomes can transform a relatively inactive drug into a potent cytotoxic agent by facilitating cellular drug uptake via a receptor frequently over-expressed in tumor cells [12]. In addition, Zol encapsulation in long circulation liposomes should reduce renal clearance and, may increase accumulation of drug in the tumor [13] by a passive targeting process known as the EPR effect (enhanced permeability and retention) [14]. This would create an in situ depot of liposomal bisphosphonate in the tumor interstitial fluid, exposing tumor cells, endothelial cells, tumor-associated macrophages and other tumor-infiltrating cells to higher drug concentrations. Based on the folate receptor-mediated endocytosis pathway [15,16], folate-targeted liposomes could conceivably facilitate intracellular delivery of Zol to folate receptor-expressing tumor cells and further enhance its anti-tumor activity. This pathway has been shown to internalize folate-bound macromolecules and liposomes [17,18]. The use of liposomal bisphosphonates, specifically liposomal clodronate, a non-NBP, has been known for many years as a highly specific tool to suppress and deplete macrophages, as reviewed by Van Rooijen [19,20]. Depletion of blood monocytes has also been demonstrated with liposomal alendronate, an NBP agent [21]. Yet, Zol is a much more potent NBP and, except for a recent study [22], there is no published information on the in vivo activity of L-Zol and FTL-Zol. As a continuation of our previous in vitro work on L-Zol [11], we examined here the pharmacokinetics (PK), biodistribution, toxicity, and therapeutic efficacy of FTL-Zol and of non-targeted L-Zol. Two human epithelial tumor models expressing the FR were tested: KB derived from cervical cancer [23], and IGROV derived from ovarian cancer [24]. In contrast to the toxicity attenuation observed with liposomal encapsulation of most cytotoxic drugs, liposomal encapsulation of Zol resulted in an extra-ordinary increase in systemic toxicity in mice when compared to free Zol, thus severely restricting the therapeutic safety window. 2. Materials and methods 2.1. Animals Female Sabra (outbred), BALB/c, and CD1 athymic/nude mice, 8–10-week old, were used in these studies. Mice were purchased from Harlan Breeding Laboratories (Jerusalem, Israel) and housed either in the Animal Facility of Shaare Zedek Medical Center or at the SPF facility of Hadassah Medical Center with food and water ad libitum. All animal experiments were done under a protocol approved by the Hebrew University-Hadassah Institutional Review Board for use of animals in research. 2.2. Liposomal preparation and characterization The method of preparation, the sources of liposome components, and the formulations tested in this study were described in detail in a previous report [11]. The formulations tested in this study were: dipalmitoyl-phosphatidylglycerol (DPPG): partially hydrogenated soybean phosphatidylcholine (PHPC): cholesterol (Chol), molar ratio 5:55:40; and, methoxyPEG2000-distearoyl-phosphatidylethanolamine (mPEG-DSPE): fully hydrogenated soybean phosphatidylcholine (HSPC): Chol, molar ratio 5:55:40. They are abbreviated respectively as DPPG-L-Zol and PEG-L-Zol. The choice of PHPC over HSPC for the DPPG-containing formulation was based on our in vitro work [11] indicating that Zol encapsulated in folate-targeted DPPG–PHPC liposomes

was the most active formulation (lowest IC50) among those tested. The other formulation chosen for this study was a typical “Stealth” formulation, containing mPEG-DSPE and HSPC, which was found previously to be the most stable and longest circulating formulation [11]. Briefly, the lipid components were dissolved in tertiary-butanol, lyophilized overnight, and rehydrated in buffer containing 100 mM zoledronic acid in 15 mM Histidine at pH 7.0 to form liposomes. The liposomes were then down-sized by serial extrusion through polycarbonate filters with pore sizes from 1000 nm to 50 nm. They were then dialyzed against 500 ml, pH 7.0, 15 mM histidine in 9:1 volume ratio of 5% dextrose: 0.9% saline, and passed through a Dowex anion exchange resin 1 × 2-400 to remove residual free zoledronic acid. The liposomes were sterilized by filtration through 0.22 μM filters and stored in vacutainer tubes at 4 °C. Phospholipid content was determined by Bartlett phosphorus assay of the lower phase of Folch extracted samples (8:4:3 chloroform: methanol: DDW) [25]. ZOL-associated phosphorus (>99% as confirmed by testing liposomes spiked with C14-ZOL) remains in the upper (aqueous) phase and was determined by Bartlett assay of the upper phase after Folch extraction. The final preparation was tested for residual free material by fractionation on a BioGel A-15M (BioRad) chromatographic column to separate free zoldedronic acid and liposome-associated zoledronic acid. No residual free zoledronic acid was detectable after formulation preparation and the formulations were stable for ≥24 months with respect to pH, particle size, and retention of Zol. Biogel fractionation was also used to assess stability after liposomal exposure to plasma. Mean vesicle size determined on a HPPS Malvern Particle Sizer, was in the range of 80 to 100 nm (Gaussian mean) with polydispersity index ≤0.1. Surface charge (Zeta potential) was negative for DPPG-L-Zol (−35 mV) and neutral for PEG-L-Zol. DPPG-L-Zol was also tested in folate-targeted form, by micellar insertion of a lipophilic PEG2000-folate conjugate at 0.5% molar lipid ratio into preformed liposomes as described previously [11]. 2.3. PK and tissue distribution studies In some studies, a dose of 30 μg/mouse of either Free Zol or L-ZOL, spiked with ~40,000 cpm of C14-Zol, was injected through the tail vein of tumor-free Balb/C mice. In another study, we used a double-labeled liposome formulation to track both the drug and the liposomes in tumor-bearing mice: A dose of 30 μg/mouse of DPPG-L-Zol, spiked with 40,000 cpm C14-Zol and 60,000 cpm of H3-Cholesterol oleoyl ether (H3-Chol) was injected through the tail vein of Balb/C mice bearing M109-FR mouse lung carcinoma tumors [26] inoculated s.c. in both flanks. Groups consisted of 3 to 4 mice each. In all cases, Animals were anesthetized, terminally bled and tissues removed at 48 or 72 h (1 h for Free Zol). Plasma samples were extracted with Quicksafe (Zinsser Analytic, Maidenhead, United Kingdom) as described previously [26]. Radioactivity was measured by β scintillation counting. Tissue samples were processed as described below. Results were expressed as % of injected dose per ml plasma or gram tissue (%ID/ml, %ID/g). Since most of the liposomes were cleared from blood by 48–72 h, these time points were chosen to evaluate the tissue fate of L-Zol. Since the plasma level of free Zoledronic was very low already at 1 h after injection, this time point was chosen to evaluate the tissue distribution of free drug. 2.4. Solubilization protocol for tissue cpm counting Approximately 50–100 mg of fresh tissue samples of lung, spleen, kidney, liver, skin tissue, muscle and brain were digested in 1 ml Solvable™ (Perkin Elmer, Waltham, MA) and incubated in a shaking water bath at 60 °C, 100 strokes per minute, overnight. Samples were cooled to room temp and 100 μl of 0.1 M EDTA was added to each vial. To decolorize 300 μl of 30% H2O2 was slowly added to each vial and capped loosely. The peroxide was added in 100 μl aliquots with

2.5. Solubilization protocol for bone cpm counting A sample of bone was taken and weighed in a scintillation vial. 1.5 ml of Decalcifier-II® (SurgiPath, Richmond, IL) was added and left overnight. The supernatant was transferred to a new glass scintillation vial and 10 ml of Quicksafe was added to the supernatant and taken for counting. 1 ml Solvable was added to the precipitate in the dark. Samples were incubated in a shaking water bath at 60 °C, 100 strokes per minute, for 3 h. The decalcified bone is disintegrated easily by Solvable. The samples were cooled to RT. 100 μl of 0.1 M EDTA was added to each vial and the procedure continued as described above. This procedure allowed us to quantify the respective distribution of Zol and liposome in the mineral hydroxy-apatite compartment (supernatant), and in the cellular and extracellular matrix compartment (precipitate). 2.6. Toxicity studies These studies were carried out in female BALB/c and Sabra mice given intravenous injections of free Zol and L-Zol at various dose levels and followed for weight changes and clinical observation. Due to unexpected toxicity of the liposomal formulations, various attempts were made to protect the animals from toxicity by a variety of additional treatments (see Section 3). For clinical pathology studies, mice were anesthetized with isofluorane and bled into heparinized tubes. Hematology and biochemistry blood tests were done using the standard hospital automatized equipment. TNFα levels in plasma and in the supernatant of spleen cells cultured overnight were analyzed by ELISA using a PeproTech Murine TNFα Development Kit. Histopathology (hematoxilin–eosin staining) of various organs was performed in the veterinarian pathology service of the Hebrew University. 2.7. Therapeutic studies Therapeutic studies were performed in two human high FRexpressing tumor models in athymic CD1 nude mice: KB cervival (formerly thought to be of nasopharyngeal origin) and IGROV-1 ovarian carcinoma. Mice were fed a special low-folate diet (Harlan Tekled, Madison, WI) one week before tumor inoculation. Three days after the last treatment, mice were returned to a standard diet. Free doxorubicin was used as standard chemotherapy for comparison. Unpaired t test and log rank test (Graph-Pad Prism) were used for statistical analysis. For the KB model, mice were inoculated in the hind footpad with 1 million tumor cells. Approximately, 10 days after inoculation, mice developed measurable tumors as indicated by an increase in thickness of the footpad from 1.5 to 2 mm. Cages were inspected daily. Mice were monitored for body weight twice per week and for tumor size with precision calipers. 10 mice were allocated per group. All treatments were given i.v. ×2, 10 and 24 days after tumor inoculation. The doses were: DPPG-L-Zol, 0.2 mg/kg; DPPG-FTL-Zol 0.4 mg/kg, vehicle (plain liposomes), according to lipid dose given to DPPG-L-Zol group; Doxorubicin, 5 mg/kg; and Free Zol, 2 mg/kg. In the IGROV model, mice were inoculated with 5 million tumor cells in a volume of 200 μl in the peritoneal cavity. After ~3 weeks, abdominal swelling develops, indicating peritoneal tumor spread and ascites. 10 mice were allocated per group. Cages were inspected daily and monitored for survival. All treatments were applied i.v. ×2, 7 and 21 days after tumor inoculation. Except for a slightly lower dose of

DPPG-FTL-Zol (0.3 mg/kg), the same doses and test groups as for the KB model were applied. Experiment terminated on day 100. 3. Results 3.1. Pharmacokinetics and tissue distribution of C14-labeled L-Zol in normal mice In this study we compared the PK of free Zol, PEG-L-Zol, DPPG-L-Zol, and folate-targeted DPPG-L-Zol (FTL-Zol) in BALB/c mice. The results based on the plasma content of C14-Zol indicate long circulation times for all liposomal formulations with estimated half-lives of 9–18 h (Fig. 1). For comparison, free Zol is almost totally cleared from plasma within 1 hour with b 0.5% injected dose (ID)/ml left (Fig. 1). Note that DPPG-L-Zol is cleared faster than PEG-L-Zol, and that the folate-targeted formulation has a shorter circulation time than its non-targeted counterpart (Fig. 1). Due to the superior in vitro cytotoxic activity of the DPPG-based formulations [11], DPPG-L-Zol and DPPG-FTL-Zol, along with free Zol, were chosen for testing in the tissue distribution studies. As seen in Fig. 2, major differences between free Zol and DPPG-L-Zol were observed. Because of the differences in plasma clearance, the time points chosen for analysis of tissue drug concentration were 1 h for free Zol and 72 h for DPPG-L-Zol. Except for kidney, where the drug levels were similar for free Zol and DPPG-L-Zol, all other 4 tissues examined showed much greater drug levels after DPPG-L-Zol injection (~4-fold in liver, ~300-fold in spleen, ~6-fold in lung, ~8-fold in skin). When DPPG-L-Zol and DPPG-FTL-ZOL are compared, the latter shows an even greater uptake by spleen with somewhat lower uptake in skin. 3.2. Tissue distribution of double-labeled (C14-Zol, and H3-Chol) L-Zol in tumor-bearing mice A second biodistribution study was performed to determine whether the PK and biodistribution of the lipid component of the liposomes represented by H3-Chol correlates with that of C14-Zol, and to obtain drug uptake data on additional tissues, particularly on tumors. This study was done in BALB/c mice bearing subcutaneously implanted syngeneic M109 tumor. As in the previous study, DPPG-L-Zol, DPPG-FTL-Zol, and Free Zol were tested. Tissue distribution was examined 1 h after injection for free Zol-injected mice (Fig. 3), and 48 h after injection for mice injected with the DPPG-L-Zol formulations (Fig. 4). The H3-Chol label is non-degradable because it has an ether bond, and therefore remains for a prolonged time in tissues even if liposomes have broken down and released their contents giving a cumulative estimate of the liposomes reaching the tissue [27]. If the count ratio in circulation of H3-Chol/C14-Zol remains unchanged, this is

% Injected dose / ml plasma

swirling between additions to prevent excess foaming, incubated for 15 min at RT, followed by 30 min at 60 °C. The vials were closed tightly. Samples were cooled to RT and 15 ml of Ultima Gold (Perkin Elmer) was added in the dark with shaking. The samples were covered with aluminum foil, vortexed well until the solution was clear then counted.

267

100

t1/2 =18.6 h 10

t1/2 =13.8 h 1

Free Zol 0.1 0

6

PEG-L-Zol DPPG-L-Zol DPPG-FTL-Zol

t1/2 =9.2

12 18 24 30 36 42 48 54 60 66 72

Hours after Injection Fig. 1. Plasma clearance of C14-Zol labeled Free Zol and L-Zol in BALB/c mice. Dose: 30 μg Zol/mouse, i.v. Half-life calculated by one-phase decay, non-linear, least squares fit (GraphPad Prism). Free Zol not detectable after 1 h. N=3/group.

NANOMEDICINE

H. Shmeeda et al. / Journal of Controlled Release 167 (2013) 265–275

H. Shmeeda et al. / Journal of Controlled Release 167 (2013) 265–275

Skin

0 Li ve

% Injected dose / gram tissue

100 80

DPPG-L-Zol 30 μg/mouse i.v. 72h-post injection

B

or

Lung

Tu m

Kidney

on e

Spleen

r

Liver

1

B

0

2

n

2

3

Sk i

4

4

H ea rt

6

5

ng

8

6

Lu

40 10

7

id ne y

60

Free Zol 1h

8

K

Free Zol 30 μg/mouse i.v. 1h-post injection

A

Sp le en

80

% Injected dose / gram tissue

% Injected dose / gram tissue

100

Fig. 3. Biodistribution of C14-labeled free Zol, in BALB/c mice bearing M109 s.c. tumor implants. Dose, 30 μg Zol/mouse, i.v. N= 4/group.

60 40 10

markers are compared for the same liposome formulation, the plasma levels of the former were ~ 2-fold greater, suggesting some degree of drug leakage from circulating liposomes but, given the very low circulating levels of liposomes at this time point and the tissue distribution profile (see below), it is unlikely that drug leakage affects a major fraction of the injected dose. When the tissue drug levels are examined (Figs. 3 and 4), the most salient findings are:

8 6 4 2 0 Liver

100

% Injected dose / gram tissue

NANOMEDICINE

268

80

Spleen

Kidney

Lung

Skin

DPPG-FTL-Zol 30 μg/mouse i.v. 72h-post injection

C

60 40 10 8 6 4 2 0 Liver

Spleen

Kidney

Lung

Skin

Fig. 2. Biodistribution of C14-labeled Free Zol and L-Zol in BALB/c mice. Dose 30 μg Zol/mouse, i.v. N = 3/group.

indicative of liposome stability in circulation. If the ratio goes up in circulation, this is indicative of Zol leakage and clearance from plasma. If the ratio goes up in the tissue, this is indicative of breakdown of liposomes in tissues (as expected) and release of Zol. If the ratio goes down in a specific tissue (i.e., bone), this indicates influx of Zol released from other tissues and redistribution to tissues for which Zol has affinity. As seen in the previous experiment, free Zol was rapidly cleared from plasma within 1 h post injection (0.2% ID/ml); but unlike previous experiments, DPPG-L-Zol and DPPG-FTL-Zol were cleared from circulation faster (~0.3% and 0.1% ID/ml plasma in tumor-bearing mice, compared to ~ 7% and 3% ID/ml plasma in normal mice respectively at 48 h post-injection) suggesting that the tumor presence accelerates liposome clearance [28]. When the H3-Chol and C14-Zol

(1) Tissue levels for free Zol were highest in bone followed by kidney. (2) Spleen and liver were the dominant tissues of uptake in case of DPPG-L-Zol and DPPG-FTL-Zol. The spleen uptake of DPPGFTL-Zol is greater than that of DPPG-L-Zol and appears to be an important factor in the faster clearance of the folatetargeted formulations. Based on the H3-Chol liposome marker, liver and spleen clear ~ 15% more of the injected dose of DPPGFTL-Zol than that of DPPG-L-Zol. (3) The tissue distribution profile of the C14 and H3 labels of DPPG-L-Zol were largely similar, an observation consistent with the major part of the Zol dose being retained in liposomes during circulation, and delivered to tissues in liposomal form. However, in some tissues (kidney, heart, skin) and particularly in bone, there was a trend for lower H3-Chol levels compared to C14-Zol levels suggesting that part of the drug reaching these tissues may originate from the leakage of liposome contents in circulation and/or after uptake by liver and spleen, the major RES organs. (4) On a relative basis (%ID/g), the spleen uptake ranks highest for DPPG-L-Zol and DPPG-FTL-Zol, but on an absolute basis (%ID), the main deposit is the liver which is ~8-fold heavier than spleen and takes up ~35%ID of L-Zol. (5) Tumor uptake of C14-Zol was 7–10-fold greater for DPPG-L-Zol than for free Zol. Tumor uptake was greater for L-Zol (~10%ID/g) than for DPPG-FTL-Zol (~6.5%ID/g), probably due to the faster clearance and shorter circulation time of the latter. Based on H3-Chol label, liposome deposition in tumor of DPPG-L-Zol was higher reaching ~14% ID/g tumor, suggesting that a minor fraction of Zol leaks from liposomes and is not retained by the tumor. (6) The bone uptake of Zol is actually increased (~ 2-fold) by liposome delivery when compared to free drug. When the bone data are examined by sub-compartment (Fig. 5), after solubilization of the mineral fraction and precipitation of the protein fraction, the C14-Zol marker was found mainly in the mineral fraction in free Zol-treated mice as expected, and notably also

DPPG-L-Zol 48h

A

100

C14-Zol (0.35 % ID/ml Plasma)

80

DPPG-L-Zol 48h H3-Chol

B

(0.6 % ID/ml Plasma)

80

m or

e

Tu

B on

in Sk

H ea rt

Lu

m or Tu

e B on

Sk in

H ea rt

Lu

ng

y

ve r Li

Tu

B on

Sk

ng Lu

ne id K

Li

(0.2 % ID/ml Plasma)

ne

20 16 14 12 10 8 6 4 2 0 m or

40

20 16 14 12 10 8 6 4 2 0 e

40

in

60

H ea rt

60

y

80

Sp le en

80

D

id

100

(0.1 % ID/ml Plasma)

K

C

DPPG-FTL-Zol 48h H3-Chol

Sp le en

100

ve r

% Injected dose / gram tissue

DPPG-FTL-Zol 48h C14-Zol

ng

y ne id K

m or

e

Tu

B on

in Sk

H ea rt

ng Lu

K

id

ne

y

20 16 14 12 10 8 6 4 2 0 Sp le en

40

20 16 14 12 10 8 6 4 2 0 Li ve r

40

Sp le en

60

60

Li ve r

% Injected dose / gram tissue

100

269

Fig. 4. Biodistribution of double-labeled L-Zol (C14-Zol, H3-Chol) in BALB/c mice bearing M109 s.c. tumor implants. Dose, 30 μg Zol/mouse, i.v. N = 4/group. A and B: DPPG-L-Zol, C14-Zol and H3-Chol data respectively; C and D: DPPG-FTL-Zol, C14-Zol and H3-Chol data respectively.

in DPPG-L-Zol-treated mice. In contrast, the liposome material was mainly deposited in the protein fraction as indicated by the H3-Chol marker. Release of Zol from this liposome material is likely to be the source of some of the mineral-associated Zol, but could not account for all of it (Fig. 5 indicates that a total of ~5% of the liposome dose as compared to ~14% of the Zol dose was deposited per gram bone), suggesting that there must be some redistribution to the mineralized bone of Zol released from liposomes residing in other tissues.

3.3. Toxicity of L-ZOL During the PK studies, we noticed that mice left on observation after injection of a flat dose per mouse of 30 μg PEG-L-Zol died without premonitory signs 5 to 7 days after injection. In contrast, mice tolerated 50–100 μg free Zol i.v. (~ 2.5 mg/kg) without apparent toxicity. Occasional deaths (10–20%), without significant weight loss, were observed at 100 μg/mouse which was considered the free Zol maximal tolerated dose (MTD) in BALB/c female mice.

C14-Zol L-Zol

Bone, % Injected Dose/gram

16 14

14

12

12

10

10

8

8

Free- Zol

6

6

4

4

2

H3-Chol

16

FTL-Zol

Free- Zol

L-Zol

FTL-Zol

0

2

L-Zol L-Zol

FTL-Zol

FTL-Zol

0 Mineral Fraction

Protein Fraction

Mineral Fraction

Protein Fraction

Fig. 5. Deposition of Free Zol, DPPG-L-Zol, and DPPG-FTL-Zol in the mineral-soluble compartment and in the protein compartment of bone. Experimental details as in Fig. 4. N=4/group.

NANOMEDICINE

H. Shmeeda et al. / Journal of Controlled Release 167 (2013) 265–275

H. Shmeeda et al. / Journal of Controlled Release 167 (2013) 265–275

24

A

PEG-L-Zol 5μg

Mouse Weight (g)

23

Mouse #1

22

Mouse #2

21

Mouse #3

20 19 18

Found dead d.7

17 16 0

1

2

3

4

5

6

7

8

Days after injection

Mouse Weight (g)

24

B

23

DPPG-L-Zol 25μg

22

Mouse #1 Mouse #2

21

Mouse #3

20 19

Found dead d.6

18 17 16 0

1

2

3

4

5

6

7

8

Days after injection 24

Mouse Weight (g)

DPPG-L-Zol 5μg

C

23

Mouse #1

22

Mouse #2

21

Mouse #3

20 19 18 17 16 0

1

2

3

4

5

6

7

8

Days after injection 24

Free Zol 100μg

D

23

Mouse Weight (g)

NANOMEDICINE

270

Mouse #1

22

Mouse #2

21

Mouse #3

20 19 18 17 16 0

1

2

3

4

5

Days after injection

6

7

8

ZO

10 5

L LZO

ZO Fr ee

U

Spleen Weights

ee

ZO

Fr

L-

ZO

L

ed re nt U

L

F

220 200 180 160 140 120 100 80

at

Spleen weight, mg/mouse

L

te d

0

L-

Zo

l

l

U

Fr

ee

Zo

ed at

Fr ee

15

nt re a

TNFα, pg/ml superatant

LZo

l

l Fr ee

Zo

te d nt re a

E

20

U

Platelets

C

re

L

te d nt re a U

TNFα in Spleen Cell Culture

Hgb

B

nt

L

0

Fig. 7. Effect of L-Zol on blood counts, spleen and plasma TNFα. Sabra mice injected i.v. with 50 μg/mouse free Zol or 20 μg/mouse DPPG-L-Zol. Five days later mice were bled and euthanized. Blood counts were done by automatic counting. The spleen was removed and weighed. TNFα levels were measured in plasma and in supernatant of overnight spleen cell cultures. N = 3–4/group.

Reduced liposomal Zol doses were tested to find a safe dose level. Doses of 20 μg and 10 μg/mouse of either DPPG-L-Zol or PEG-L-Zol were lethal when injected i.v. to BALB/c mice. At 5 μg/mouse, DPPG-L-Zol was well tolerated, but deaths were still observed with PEG-L-Zol. At 2.5 μg/mouse, PEG-Lip-ZOL formulation was also lethal in a major fraction (≥ 50%) of the animals. Finally, 1 μg/mouse was found to be a safe dose for PEG-L-Zol. In general, no significant weight loss was notable prior to death. Death occurred between 5 to 8 days after injection. Similar toxicity observations to those in BALB/c mice were made in outbred Sabra female mice which have ~ 50% greater body weight. These results are summarized in Fig. 6. Interestingly, DPPG-FTL-Zol appeared to be somewhat less toxic than since it was generally well tolerated at 10 μg/mouse, and its MTD was ~ 2-fold higher than that of DPPG-L-Zol (data not shown). A multiple injection toxicity study using doses of 5 μg/mouse of DPPG-L-Zol demonstrated good tolerability after 3 injections given once every two weeks, suggesting that toxicity was not cumulative.

3.4. Clinical pathology Sabra female mice were injected with 20 μg/mouse of DPPG-L-Zol or 50 μg/mouse of Free Zol. Five days later, blood counts and plasma levels of Na, Ca, BUN and creatinine were analyzed. Tissue samples were also collected and examined for histopathology. Significant leukocytosis, particularly neutrophilia, basophilia, and monocytosis, as well as thrombocytopenia were observed (Fig. 7A–C). In addition, a mild degree of hypocalcemia (7.5 ± 1.0 mg/dl vs. 8.6 ± 0.4 mg/dl in untreated mice), and elevated BUN (34 ± 7 mg/dl vs. 18 ± 1 mg/dl in untreated mice) was found in DPPG-L-Zol treated mice suggesting some level of functional kidney damage.

We also examined the plasma level of TNFα in mice injected with DPPG-L-ZOL, and found a moderate rise (Fig. 7D). A similar finding was observed in spleen cell culture supernatants from DPPG-L-Zol treated mice (Fig. 7E). Despite the small differences and large variability, these findings are suggestive of cytokine overproduction. Except for increased mitotic activity in the spleen red pulp, no appreciable damage was detectable by standard H&E staining in liver, spleen, lung, heart, and kidney. However, spleen weight was consistently higher in the DPPG-L-Zol treated mice compared to free Zol treated and untreated mice (Fig. 7F), a finding possibly related to the observed leukocytosis. 3.5. Interventions to reduce toxicity of L-Zol A number of experiments to reduce the toxicity of L-Zol were attempted including daily physiologic saline injections, urine alkalinization with sodium bicarbonate, co-injection of corticosteroids, vitamin D3, calcium gluconate, or N-acetyl-cysteine, and others, all without success. Single administration of liposomal clodronate (0.5–1 mg/mouse) prior to L-Zol to mice delayed toxic deaths by about 3 to 4 days, but an attempt to re-challenge with liposomal clodronate on days 3 or 4 to improve protection failed and resulted in deaths. Most of these experiments were done using as control PEG-L-Zol 5 μg/mouse which is uniformly lethal at this dose. The only intervention that was able to prevent toxic deaths and reduce the toxicity significantly was the co-injection of a large amount of blank liposomes (10–20 μmoles phospholipid/mouse) of the same lipid composition as L-Zol which enabled us to increase the MTD of PEG-L-Zol to 10 μg/mouse (Table 1), a 4- to 10-fold increase in MTD. These observations clearly implicate the involvement of macrophage uptake in the mechanism of toxicity of L-Zol, and suggest that a dilution of the Zol dose into a larger

Fig. 6. A–D: Toxicity of PEG-L-Zol and DPPG-L-Zol in BALB/c mice. PEG-L-Zol is toxic at lower dose (5 μg/mouse) than DPPG-L-Zol. Free Zol is the least toxic formulation. Follow-up for 60 days did not show any more deaths beyond day 8 post-injection. Note in 6A and 6B a slight increase in weight before toxic deaths, apparently due to fluid retention.

NANOMEDICINE

5

LZO

TNFα, pg/ml plasma l Zo Fr ee

10

U

nt re a

15

l

550 500 450 400 350 300 250 200 150 100 50 0

D

20

LZo

Hgb g% (Mean±SEM)

17.5 15.0 12.5 10.0 7.5 5.0 2.5 0.0

271

TNFα in Plasma

WBC

A

te d

WBC x103/μl (Mean±SEM)

40 35 30 25 20 15 10 5 0

Plt x103/μl (Mean±SEM)

H. Shmeeda et al. / Journal of Controlled Release 167 (2013) 265–275

H. Shmeeda et al. / Journal of Controlled Release 167 (2013) 265–275

Table 1 Toxicity of PEG-L-Zol and its attenuation by co-injection of blank (drug-free) liposomes (PL = phospholipid; N20, N32, N33 are batch numbers). Formulation

Zol/PL mol/mol

Dose Zol μg/mouse

N mice dead/total

Time of death

PEG-L-Zol (N20)

0.10

5

0

3/3

PEG-L-Zol (N20) PEG-L-Zol (N20)

0.10 0.10

5 10

10 10

0/3 4/4

PEG-L-Zol (N20)

0.10

10

0/4

PEG-L-Zol (N32)

0.15

5×2 q2w 10

d6-1 d7-2 – d7-2 d10-1 d11-1 –

0

3/3

PEG-L-Zol (N32) PEG-L-Zol (N32)

0.15 0.15

10 20

10 10

0/4 2/4

PEG-L-Zol (N32)

0.15

10

0

3/4

PEG-L-Zol (N32) PEG-L-Zol (N33)

0.15 0.01

10 10

3.3 0

0/4 1/4

Blank PL μmoles/mouse

macrophage population reduces the toxicity signal. This hypothesis is supported by results obtained with a special PEG-L-Zol preparation with a 10-fold lower drug:lipid ratio than our standard PEG-L-Zol formulation. The low drug:lipid ratio L-Zol was significantly less toxic than the standard PEG- L-Zol formulation (Table 1). 3.6. Therapeutic studies In the KB model, DPPG-FTL-Zol at 0.4 mg/kg (~10 μg/mouse) appears to be effective in delaying tumor development and was the only therapeutic intervention demonstrating a significant advantage over control and free Dox-treated mice (Fig. 8). However, toxic deaths (5/10 mice) reduced the therapeutic value of this effect, and leave a very narrow therapeutic window in this tumor model. In the IGROV model, no significant advantage was detectable when analyzing the survival curves by the log rank test (Fig. 9). Nonetheless, 4/10 mice were alive and apparently tumor-free in the DPPG-FTL-Zol as opposed to 0/10 in the control group suggesting limited therapeutic activity but the study was underpowered to confirm it. Although DPPGFTL-Zol was injected at a lower dose (0.3 mg/kg or ~7.5 μg/mouse), toxic deaths were still observed hampering the therapeutic outcome. 4. Discussion

5

*

*

*

10

d6-2 d7-1 – d8-1 d13-1 d5-1 d7-2 – d7-1

7 47 47 43 43 43

Vehicle (plain liposome) Doxorubicin DPPG-L-Zol DPPG-FTL-Zol Free-Zol

70

*

*

* *

60 50 40 30 20 10

2 5

82

80

4 3

47 11

100

% Survival

6

7

90

Vehicle Doxorubicin DPPG-L-Zol DPPG-FTL-Zol Free Zol

7

Follow-up (days)

with hormonal therapy after resection of the primary tumor [9,29–31]. Zol treatment has been shown to reduce the level of serum VEGF in cancer patients implying an anti-angiogenic effect [32]. Preclinical studies also reveal non-skeletal anti-tumor effects of ZOL in various tumor models (breast cancer, myeloma, other) which suggest direct tumorinhibitory activity of Zol [33–35]. The anti-tumor effect appears to be mediated by a molecular mechanism similar to that observed in osteoclasts and related to the inhibition of the mevalonate pathway with blockade of FPP synthase [36]. The recommended dose of Zol in humans is 4 mg (~ 60 μg/kg for a 70 kg-patient) every 4 weeks. Due to potential nephrotoxicity [37], dose adjustments of Zol are required in patients with impaired renal function. After i.v. administration, Zol is rapidly cleared by the kidneys, plasma protein binding is negligible, and a significant fraction is retained in bone mineral matrix bound to hydroxy-apatite, where the osteoclast-inhibitory effect takes place [38–40]. Despite a well-established molecular mechanism for anti-tumor effects of NBP, new strategies are needed to overcome their in vivo pharmacologic limitations in anti-cancer therapy [41]. Our studies with radiolabeled free Zol are consistent with its known PK-biodistribution profile. In contrast, L-Zol causes a major prolongation of circulation time and a major shift in drug distribution when compared to free Zol. The differences between the various liposome formulations tested are minuscule when compared to the differences

Recent clinical studies have demonstrated effective prevention of metastases and recurrence of breast cancer when Zol is given together

Median Tumor Thickness (mm)

NANOMEDICINE

272

15

20

25

30

35

Days after tumor inoculation

0 0 20 20

30

40

50

60

70

80

Days after tumor injection Fig. 8. KB tumor model. All treatments i.v. × 2, 10 and 24 days after tumor inoculation. N=10/group. Vehicle (plain liposome); Doxorubicin 5 mg/kg; DPPG-L-Zol 0.2 mg/kg; DPPG-FTL-Zol 0.4 mg/kg; Free Zol 2 mg/kg. Toxic deaths: DPPG-L-Zol 2/10; DPPG-FTL-Zol 5/10. No apparent toxic deaths in al other groups. Symbol★ indicates a statistical significant difference between DPPG-FTL-Zol and Vehicle and Free Zol at p level between 0.02 and 0.002 (Unpaired T test, GraphPad Prism). Other comparisons were not significant.

Fig. 9. IGROV-1 model. All treatments i.v. ×2, 7 and 21 days after tumor inoculation. Experiment terminated on day 100 without further change in survival. N=10/group. Vehicle (plain liposome); Doxorubicin 5 mg/kg; DPPG-L-Zol 0.2 mg/kg; DPPG-FTL-Zol 0.3 mg/kg; Free Zol 2 mg/kg. Toxic deaths: DPPG-L-Zol=3/10; DPPG-FTL-Zol=4/10. No apparent toxic deaths in all other groups. All comparisons between the various groups (log rank test, GraphPad Prism) were not significant.

between L-Zol and free Zol. As noticed in a previous report [11], the clearance of FTL-Zol is faster (~2-fold) than that of non-targeted L-Zol. The reason for the faster clearance of folate-targeted liposomes has been discussed before [26,42] and may be related to folate-receptor mediated uptake by activated macrophages or to a liposome-opsonizing effect by serum folate-binding protein. Co-injection of FTL-Zol with a large dose of folic acid (5 mg/mouse) maintains plasma levels of FTL-Zol equal to those seen with L-Zol (data not shown). There appears to be slight drug leakage from circulating DPPG–PHPC liposomes in view of the lower plasma drug concentration (C14-Zol) in comparison to the plasma liposome concentration (H3-Chol) 48 h after injection (Fig. 4). This observation should not be extrapolated to PEG-HSPC liposomes since in vitro plasma stability studies have shown a slight (~5%) leakage of C14-Zol from DPPG–PHPC liposomes, but none from the former [11]. Despite the high and expected hepato-splenic uptake of L-Zol, the bone uptake of Zol is maintained and even increased ~ 2-fold by liposome delivery when compared to free drug. Thus, even at the level of bone and bone mineral matrix, liposomal delivery of Zol rates higher than free drug. In terms of potential for antitumor efficacy of Zol, the most valuable result of liposome delivery is a substantial increase in the tumor levels of Zol, between 7 to 10-fold, as compared to free Zol. As seen in a past study with folate-targeted liposomal doxorubicin [26], we did not find here any enhancement of uptake in the FR-expressing M109 tumor when DPPG-FTL-Zol is compared to DPPG-L-Zol. In fact, the tumor level of DPPG-FTL-Zol was slightly lower than that of DPPG-L-Zol, probably resulting from the higher hepato-splenic uptake and shorter circulation time of the former. This is consistent with our conclusion from previous studies that liposome circulation time and extravasation efficiency are the rate-limiting factors for tumor liposome deposition, for non-targeted liposomes and for liposomes targeted to extravascular target receptors [43–45]. Despite the toxicity seen at doses of of DPPG-L-Zol ≥ 10 μg/mouse, we could not use less than 30 μg/mouse in tissue biodistribution studies because this was the minimal amount needed to have a reasonable amount of C14 and H3 counts for detection in vivo. The lipid dose is in any case relatively low for both 30 and 10 μg-dose levels, suggesting that this is not a critical issue in interpreting the results. In contrast to past experience of toxicity reduction with liposomeencapsulated cytotoxic drugs [46–48], the liposome formulations of Zol demonstrated much higher toxicity than free Zol. DPPG-L-Zol was safe at a maximal dose of only 5 μg/mouse (~ 200 μg/kg), while PEG-L-Zol was significantly more toxic with a safe dose as low as 1 μg/mouse (~40 μg/kg). In fact, L-Zol is more toxic on a molar basis than powerful cytotoxic drugs such as mitomycin-C, and doxorubicin. The toxicity of L-Zol may well be explained by macrophage activation leading to a cytokine storm [49]. The macrophage is the major in vivo destination of any liposomal formulation, and excess of cytokine production (high levels of TNFα and IL-6) has been reported after free ZOL treatment [50]. The lag time from injection till death (~ 5–7 days) is consistent with the time required for liposomes to be cleared from circulation, break down, and release Zol inside macrophages. Zol released from liposomes in the extracellular fluid compartment is unlikely to account for the increased toxicity since it should be redistributed and cleared rapidly in the same way as free Zol. A direct nephrotoxic effect as cause of death is unlikely since kidney histopathology, as well as that of other organs, indicated no overt signs of toxicity. The pathology observed includes splenomegaly, leukocytosis, and thrombocytopenia, which may be an expression of the activated state of macrophages and are known to occur in the cytokine storm syndrome [49]; and, a trend to a weight increase just prior to death is consistent with capillary leak syndrome and edema. Finally, the most effective manipulation to decrease toxicity was the co-injection of blank liposomes or the injection of a formulation with very low Zol:lipid ratio. Increasing the liposome lipid dose is likely to saturate

273

hepato-splenic macrophage-mediated clearance and dilute the dose of Zol into an extended pool of macrophages and other liposomeengulfing cells, thus reducing Zol concentration in macrophages below the threshold required to switch on a cytokine storm. Of note, the toxicity potency of L-Zol correlates with the circulation half-life of the various formulations (PEG-L-Zol > DPPG-L-Zol > DPPGFTL-Zol). This may have to do with the higher stability of PEG-L-Zol, resulting from the protective effect of PEG [51] and the high phase transition temperature (~50 °C) of the bulk component HSPC [52,53], replaced by PHPC in DPPG-L-Zol. Partial leakage of Zol from DPPG-L-Zol in extracellular fluids prior to liposome internalization by macrophages will reduce macrophage exposure to Zol and thereby the ensuing toxic response. While liposomal clodronate has well-known macrophagesuppressing and depleting effects which are not seen with free clodronate [19,20], it is important to point out that clodronate is a non-NBP with a mechanism of action completely different from nitrogen-containing bisphosphonates [7]. Liposomal clodronate is generally non-lethal and is injected at doses as high as 1 mg/mouse. Therefore, clodronate may not be a relevant comparator in the context of liposomal Zol which has lethal effects at a dose of a few micrograms per mouse. A more appropriate comparator would be liposomal alendronate, another NBP, albeit much less powerful than zoledronic acid. Most of the liposomal alendronate work was done in the context of prevention of coronary re-stenosis and inflammatory disease [54], with a recent paper dealing with cancer therapy [55]. Our results suggesting cytokine release from macrophages as mechanism of toxicity of liposomal Zol are consistent with the results of the liposomal alendronate work of the Golomb laboratory [21]. However, no lethal toxicity was reported with liposomal alendronate. In our view, the severe lethal toxicity of liposomal zoledronic acid cannot be equated with any prior observations in this field. The fact that Zol is by far the leading bisphosphonate in use in cancer patients is also of special relevance. A recent publication of Marra et al. on a different liposomal formulation of Zol points to a higher tolerated dose (≥20 μg Zol/mouse) than the MTD found in this his study [22]. No PK-biodistribution information was reported. However, a close look reveals significant differences between the formulations that may account for the discrepancy in the toxic effects. The main component of the formulation of Marra et al. is egg phosphatidyl–choline, which is less stable than HSPC and PHPC, and may lead to more drug leakage in circulation [52]. In addition, the liposomes were freeze-dried resulting in a loss of 25–80% of the encapsulated Zol upon reconstitution, thus decreasing the ratio of encapsulated Zol per lipid [22] which, as shown here, reduces toxicity. Given the dose limitations imposed by the restricted safety window of L-Zol found in our studies, an in vivo direct antitumor effect with our formulations is unlikely. Indeed, the therapeutic results in the KB and IGROV tumor models were rather disappointing. However, the current mouse models are inadequate to detect indirect antitumor effects of nitrogen-containing bisphosphonates, such as increasing the levels of phospho-antigens and boosting the antitumor activity of Vγ9Vδ2 T lymphocytes [56–58]. In summary, liposome encapsulation of Zol results in a major change in PK and biodistribution, including a rise in tumor drug levels, but also renders the drug extremely potent and toxic in vivo to rodents. The use of a stable and macrophage-tropic nanoparticulate carrier transforms this relatively safe drug into a highly potent and toxic compound, unleashing the great potential of zoledronic acid to impact on the host biological response. Exploiting the biological effects and the major rise in tumor drug levels of liposome-encapsulated zoledronic acid for anticancer applications while restraining toxicity is challenging. Replacing Zol with less toxic nitrogen-containing bisphosphonates such as pamidronate or alendronate, and combining with chemotherapy are alternative ways to exploit the potential contribution to cancer therapy of liposomal amino-bisphosphonates that deserve further exploration.

NANOMEDICINE

H. Shmeeda et al. / Journal of Controlled Release 167 (2013) 265–275

NANOMEDICINE

274

H. Shmeeda et al. / Journal of Controlled Release 167 (2013) 265–275

Acknowledgments Supported by the Israel Cancer Research Fund and Novartis Pharmaceuticals Corporation. References [1] S.P. Luckman, D.E. Hughes, F.P. Coxon, R. Graham, G. Russell, M.J. Rogers, Nitrogen-containing bisphosphonates inhibit the mevalonate pathway and prevent post-translational prenylation of GTP-binding proteins, including Ras, J. Bone Miner. Res. 13 (1998) 581–589. [2] M.J. Rogers, S. Gordon, H.L. Benford, F.P. Coxon, S.P. Luckman, J. Monkkonen, J.C. Frith, Cellular and molecular mechanisms of action of bisphosphonates, Cancer 88 (2000) 2961–2978. [3] M.S. Greenberg, Intravenous bisphosphonates and osteonecrosis, Oral Surg. Oral Med. Oral Pathol. Oral Radiol. Endod. 98 (2004) 259–260. [4] I. Benzaid, H. Monkkonen, V. Stresing, E. Bonnelye, J. Green, J. Monkkonen, J.L. Touraine, P. Clezardin, High phosphoantigen levels in bisphosphonate-treated human breast tumors promote Vgamma9Vdelta2 T-cell chemotaxis and cytotoxicity in vivo, Cancer Res. 71 (2011) 4562–4572. [5] P. Clezardin, Bisphosphonates' antitumor activity: an unravelled side of a multifaceted drug class, Bone 48 (2011) 71–79. [6] P. Clezardin, M. Massaia, Nitrogen-containing bisphosphonates and cancer immunotherapy, Curr. Pharm. Des. 16 (2010) 3007–3014. [7] J.R. Green, A. Guenther, The backbone of progress–preclinical studies and innovations with zoledronic acid, Crit. Rev. Oncol. Hematol. 77 (Suppl. 1) (2011) S3–S12. [8] J.R. Green, Bisphosphonates: preclinical review, Oncologist 9 (Suppl. 4) (2004) 3–13. [9] M. Gnant, B. Mlineritsch, W. Schippinger, G. Luschin-Ebengreuth, S. Postlberger, C. Menzel, R. Jakesz, M. Seifert, M. Hubalek, V. Bjelic-Radisic, H. Samonigg, C. Tausch, H. Eidtmann, G. Steger, W. Kwasny, P. Dubsky, M. Fridrik, F. Fitzal, M. Stierer, E. Rucklinger, R. Greil, C. Marth, Endocrine therapy plus zoledronic acid in premenopausal breast cancer, N. Engl. J. Med. 360 (2009) 679–691. [10] F.P. Coxon, K. Thompson, A.J. Roelofs, F.H. Ebetino, M.J. Rogers, Visualizing mineral binding and uptake of bisphosphonate by osteoclasts and non-resorbing cells, Bone 42 (2008) 848–860. [11] H. Shmeeda, Y. Amitay, J. Gorin, D. Tzemach, L. Mak, J. Ogorka, S. Kumar, J.A. Zhang, A. Gabizon, Delivery of zoledronic acid encapsulated in folate-targeted liposome results in potent in vitro cytotoxic activity on tumor cells, J. Control. Release 146 (2010) 76–83. [12] N. Parker, M.J. Turk, E. Westrick, J.D. Lewis, P.S. Low, C.P. Leamon, Folate receptor expression in carcinomas and normal tissues determined by a quantitative radioligand binding assay, Anal. Biochem. 338 (2005) 284–293. [13] A. Gabizon, Liposome circulation time and tumor targeting: implications for cancer chemotherapy, Adv. Drug Deliv. Rev. 16 (1995) 285–294. [14] H. Maeda, The enhanced permeability and retention (EPR) effect in tumor vasculature: the key role of tumor-selective macromolecular drug targeting, Adv. Enzyme Regul. 41 (2001) 189–207. [15] S. Wang, P.S. Low, Folate-mediated targeting of antineoplastic drugs, imaging agents, and nucleic acids to cancer cells, J. Control. Release 53 (1998) 39–48. [16] J. Sudimack, R.J. Lee, Targeted drug delivery via the folate receptor, Adv. Drug Deliv. Rev. 41 (2000) 147–162. [17] C.P. Leamon, P.S. Low, Delivery of macromolecules into living cells: a method that exploits folate receptor endocytosis, Proc. Natl. Acad. Sci. U. S. A. 88 (1991) 5572–5576. [18] R.J. Lee, P.S. Low, Delivery of liposomes into cultured KB cells via folate receptormediated endocytosis, J. Biol. Chem. 269 (1994) 3198–3204. [19] N. van Rooijen, Liposomes for targeting of antigens and drugs: immunoadjuvant activity and liposome-mediated depletion of macrophages, J. Drug Target. 16 (2008) 529–534. [20] N. van Rooijen, E. Hendrikx, Liposomes for specific depletion of macrophages from organs and tissues, Methods Mol. Biol. 605 (2010) 189–203. [21] H. Epstein-Barash, D. Gutman, E. Markovsky, G. Mishan-Eisenberg, N. Koroukhov, J. Szebeni, G. Golomb, Physicochemical parameters affecting liposomal bisphosphonates bioactivity for restenosis therapy: internalization, cell inhibition, activation of cytokines and complement, and mechanism of cell death, J. Control. Release 146 (2010) 182–195. [22] M. Marra, G. Salzano, C. Leonetti, P. Tassone, M. Scarsella, S. Zappavigna, T. Calimeri, R. Franco, G. Liguori, G. Cigliana, R. Ascani, M.I. La Rotonda, A. Abbruzzese, P. Tagliaferri, M. Caraglia, G. De Rosa, Nanotechnologies to use bisphosphonates as potent anticancer agents: the effects of zoledronic acid encapsulated into liposomes, Nanomedicine 7 (2011) 955–964. [23] M. McHugh, Y.C. Cheng, Demonstration of a high affinity folate binder in human cell membranes and its characterization in cultured human KB cells, J. Biol. Chem. 254 (1979) 11312–11318. [24] M.J. Turk, D.J. Waters, P.S. Low, Folate-conjugated liposomes preferentially target macrophages associated with ovarian carcinoma, Cancer Lett. 213 (2004) 165–172. [25] H. Shmeeda, S. Even-Chen, R. Honen, R. Cohen, C. Weintraub, Y. Barenholz, Enzymatic assays for quality control and pharmacokinetics of liposome formulations: comparison with nonenzymatic conventional methodologies, Methods Enzymol. 367 (2003) 272–292. [26] A. Gabizon, A.T. Horowitz, D. Goren, D. Tzemach, H. Shmeeda, S. Zalipsky, In vivo fate of folate-targeted polyethylene-glycol liposomes in tumor-bearing mice, Clin. Cancer Res. 9 (2003) 6551–6559.

[27] J.T. Derksen, H.W. Morselt, G.L. Scherphof, Uptake and processing of immunoglobulincoated liposomes by subpopulations of rat liver macrophages, Biochim. Biophys. Acta 971 (1988) 127–136. [28] S.M. Moghimi, H.M. Patel, Altered tissue-specific opsonic activities and opsonorecognition of liposomes in tumour-bearing rats, Biochim. Biophys. Acta 1285 (1996) 56–64. [29] R.E. Coleman, H. Marshall, D. Cameron, D. Dodwell, R. Burkinshaw, M. Keane, M. Gil, S.J. Houston, R.J. Grieve, P.J. Barrett-Lee, D. Ritchie, J. Pugh, C. Gaunt, U. Rea, J. Peterson, C. Davies, V. Hiley, W. Gregory, R. Bell, Breast-cancer adjuvant therapy with zoledronic acid, N. Engl. J. Med. 365 (2011) 1396–1405. [30] H. Eidtmann, R. de Boer, N. Bundred, A. Llombart-Cussac, N. Davidson, P. Neven, G. von Minckwitz, J. Miller, N. Schenk, R. Coleman, Efficacy of zoledronic acid in postmenopausal women with early breast cancer receiving adjuvant letrozole: 36-month results of the ZO-FAST Study, Ann. Oncol. 21 (2010) 2188–2194. [31] M. Gnant, B. Mlineritsch, H. Stoeger, G. Luschin-Ebengreuth, D. Heck, C. Menzel, R. Jakesz, M. Seifert, M. Hubalek, G. Pristauz, T. Bauernhofer, H. Eidtmann, W. Eiermann, G. Steger, W. Kwasny, P. Dubsky, G. Hochreiner, E.P. Forsthuber, C. Fesl, R. Greil, Adjuvant endocrine therapy plus zoledronic acid in premenopausal women with early-stage breast cancer: 62-month follow-up from the ABCSG-12 randomised trial, Lancet Oncol. 12 (2011) 631–641. [32] D. Santini, B. Vincenzi, S. Galluzzo, F. Battistoni, L. Rocci, O. Venditti, G. Schiavon, S. Angeletti, F. Uzzalli, M. Caraglia, G. Dicuonzo, G. Tonini, Repeated intermittent low-dose therapy with zoledronic acid induces an early, sustained, and longlasting decrease of peripheral vascular endothelial growth factor levels in cancer patients, Clin. Cancer Res. 13 (2007) 4482–4486. [33] A. Guenther, S. Gordon, M. Tiemann, R. Burger, F. Bakker, J.R. Green, W. Baum, A.J. Roelofs, M.J. Rogers, M. Gramatzki, The bisphosphonate zoledronic acid has anti-myeloma activity in vivo by inhibition of protein prenylation, Int. J. Cancer 126 (2010) 239–246. [34] P.D. Ottewell, D.V. Lefley, S.S. Cross, C.A. Evans, R.E. Coleman, I. Holen, Sustained inhibition of tumour growth and prolonged survival following sequential administration of doxorubicin and zoledronic acid in a breast cancer model, Int. J. Cancer 126 (2010) 522–532. [35] A.C. Hirbe, A.J. Roelofs, D.H. Floyd, H. Deng, S.N. Becker, L.G. Lanigan, A.J. Apicelli, Z. Xu, J.L. Prior, M.C. Eagleton, D. Piwnica-Worms, M.J. Rogers, K. Weilbaecher, The bisphosphonate zoledronic acid decreases tumor growth in bone in mice with defective osteoclasts, Bone 44 (2009) 908–916. [36] L.M. Mitrofan, J. Pelkonen, J. Monkkonen, The level of ATP analog and isopentenyl pyrophosphate correlates with zoledronic acid-induced apoptosis in cancer cells in vitro, Bone 45 (2009) 1153–1160. [37] M.A. Perazella, G.S. Markowitz, Bisphosphonate nephrotoxicity, Kidney Int. 74 (2008) 1385–1393. [38] T. Chen, J. Berenson, R. Vescio, R. Swift, A. Gilchick, S. Goodin, P. LoRusso, P. Ma, C. Ravera, F. Deckert, H. Schran, J. Seaman, A. Skerjanec, Pharmacokinetics and pharmacodynamics of zoledronic acid in cancer patients with bone metastases, J. Clin. Pharmacol. 42 (2002) 1228–1236. [39] A. Skerjanec, J. Berenson, C. Hsu, P. Major, W.H. Miller Jr., C. Ravera, H. Schran, J. Seaman, F. Waldmeier, The pharmacokinetics and pharmacodynamics of zoledronic acid in cancer patients with varying degrees of renal function, J. Clin. Pharmacol. 43 (2003) 154–162. [40] H.M. Weiss, U. Pfaar, A. Schweitzer, H. Wiegand, A. Skerjanec, H. Schran, Biodistribution and plasma protein binding of zoledronic acid, Drug Metab. Dispos. 36 (2008) 2043–2049. [41] M. Marra, A. Abbruzzese, R. Addeo, S. Del Prete, P. Tassone, G. Tonini, P. Tagliaferri, D. Santini, M. Caraglia, Cutting the limits of aminobisphosphonates: new strategies for the potentiation of their anti-tumour effects, Curr. Cancer Drug Targets 9 (2009) 791–800. [42] A. Gabizon, H. Shmeeda, A.T. Horowitz, S. Zalipsky, Tumor cell targeting of liposomeentrapped drugs with phospholipid-anchored folic acid-PEG conjugates, Adv. Drug Deliv. Rev. 56 (2004) 1177–1192. [43] D. Goren, A.T. Horowitz, S. Zalipsky, M.C. Woodle, Y. Yarden, A. Gabizon, Targeting of stealth liposomes to erbB-2 (Her/2) receptor: in vitro and in vivo studies, Br. J. Cancer 74 (1996) 1749–1756. [44] A.A. Gabizon, H. Shmeeda, S. Zalipsky, Pros and cons of the liposome platform in cancer drug targeting, J. Liposome Res. 16 (2006) 175–183. [45] A. Gabizon, H. Shmeeda, H. Baabur, R. Satchi-Fainaro, Liposomes and polymers in folate-targeted cancer therapeutics, in: A.L. Jackman, C.P. Leamon (Eds.), Targeted Drug Strategies for Cancer and Inflammation, Springer, New York, 2011, pp. 217–247. [46] A. Gabizon, F. Martin, Polyethylene glycol-coated (pegylated) liposomal doxorubicin. Rationale for use in solid tumours, Drugs 54 (Suppl. 4) (1997) 15–21. [47] T. Minko, R.I. Pakunlu, Y. Wang, J.J. Khandare, M. Saad, New generation of liposomal drugs for cancer, Anticancer Agents Med Chem. 6 (2006) 537–552. [48] T.M. Allen, W.W. Cheng, J.I. Hare, K.M. Laginha, Pharmacokinetics and pharmacodynamics of lipidic nano-particles in cancer, Anticancer Agents Med Chem. 6 (2006) 513–523. [49] S.W. Canna, E.M. Behrens, Making sense of the cytokine storm: a conceptual framework for understanding, diagnosing, and treating hemophagocytic syndromes, Pediatr. Clin. North Am. 59 (2012) 329–344. [50] G. Dicuonzo, B. Vincenzi, D. Santini, G. Avvisati, L. Rocci, F. Battistoni, M. Gavasci, D. Borzomati, R. Coppola, G. Tonini, Fever after zoledronic acid administration is due to increase in TNF-alpha and IL-6, J. Interferon Cytokine Res. 23 (2003) 649–654. [51] M.C. Woodle, D.D. Lasic, Sterically stabilized liposomes, Biochim. Biophys. Acta 1113 (1992) 171–199. [52] J. Senior, G. Gregoriadis, Stability of small unilamellar liposomes in serum and clearance from the circulation: the effect of the phospholipid and cholesterol components, Life Sci. 30 (1982) 2123–2136.

[53] A. Gabizon, M. Chemla, D. Tzemach, A.T. Horowitz, D. Goren, Liposome longevity and stability in circulation: effects on the in vivo delivery to tumors and therapeutic efficacy of encapsulated anthracyclines, J. Drug Target. 3 (1996) 391–398. [54] D. Gutman, G. Golomb, Liposomal alendronate for the treatment of restenosis, J. Control. Release 161 (2012) 619–627. [55] D. Gutman, H. Epstein-Barash, M. Tsuriel, G. Golomb, Alendronate liposomes for antitumor therapy: activation of gammadelta T cells and inhibition of tumor growth, Adv. Exp. Med. Biol. 733 (2012) 165–179. [56] N. Caccamo, S. Meraviglia, F. Scarpa, C. La Mendola, D. Santini, C.T. Bonanno, G. Misiano, F. Dieli, A. Salerno, Aminobisphosphonate-activated gammadelta T cells

275

in immunotherapy of cancer: doubts no more, Expert Opin. Biol. Ther. 8 (2008) 875–883. [57] V. Stresing, F. Daubine, I. Benzaid, H. Monkkonen, P. Clezardin, Bisphosphonates in cancer therapy, Cancer Lett. 257 (2007) 16–35. [58] K. Sato, S. Kimura, H. Segawa, A. Yokota, S. Matsumoto, J. Kuroda, M. Nogawa, T. Yuasa, Y. Kiyono, H. Wada, T. Maekawa, Cytotoxic effects of gammadelta T cells expanded ex vivo by a third generation bisphosphonate for cancer immunotherapy, Int. J. Cancer 116 (2005) 94–99.

NANOMEDICINE

H. Shmeeda et al. / Journal of Controlled Release 167 (2013) 265–275