Mantha, Venkat R. R. FFARCSI*; Nair, Harsha K.*; Venkataramanan, Raman PhD†; Gao, Yuan Yue MS†; Matyjaszewski, Krzysztof PhD‡; Dong, Hongchen PhD‡; Li, Wenwen PhD‡; Landsittel, Doug PhD§‖; Cohen, Elan MS§; Lariviere, William R. PhD*
Targeted delivery of chemotherapeutic agents combined with magnetic nanoparticles (MNPs) has been successfully demonstrated in animals1–4 and humans.5,6 MNPs are complexes that are attractable by a magnet and could potentially be used to direct anesthetic drugs to a specific tissue. In addition, MNPs such as those used in the current study are stable in aqueous solution below a certain temperature but shrink and release any drug loaded into them at higher temperatures. These 2 properties permit drug-associated MNPs to be injected IV and for the MNPs to be sequestered and concentrated in a superficial tissue with magnets where the drug is released at body temperature to have an effect in the tissue.
The objective of this pilot project was to test whether magnet-directed targeting of the commonly used local anesthetic drug ropivacaine associated with MNPs is able to produce anesthetic ankle block in the rat. For this purpose, ropivacaine-associated MNP complexes (MNP/Ropiv) were injected IV while attracting them to a magnet applied at the ankle. The rationale was that when the complexes in the circulation reach the ankle, they would be sequestered and concentrated there, the ropivacaine would be released locally, act on the nerves around the ankle, and produce an anesthetic block.
Approval from the University of Pittsburgh’s Institutional Animal Care and Use Committee (IACUC) was obtained for the animal experiments. Male Sprague Dawley rats were used for the ankle block experiments (n = 6–11/group; 300–350 g) and the pharmacokinetic studies (n = 2–3/group; 375–425 g). Rats were cared for in compliance with the IACUC policies, with free access to food and water and lights on from 6:00 AM to 6:00 PM. Each animal was used only once.
While animals were under 2% isoflurane anesthesia in oxygen with spontaneous ventilation, a 24-G cannula was placed in the tail vein and magnets placed around the ankle of the right hindpaw. The left hindpaw served as the untreated control without magnet application. The MNP/Ropiv complexes (or MNPs alone vehicle) were then manually injected via the cannula over 4 to 5 minutes by the same investigator. A volume of 2 mL MNP/Ropiv suspension (containing 14 mg ropivacaine or 40–46 mg/kg) was injected based on the assumption that approximately 10% or less of the injected dose would be sequestered and available at the ankle. Before injection, the MNP/Ropiv complexes were removed from the ice container used for transport and allowed to warm to 10°C to 15°C to avoid possible ventricular fibrillation. The magnets were removed after 15, 30, or 60 minutes (groups MNP/Ropiv/Mag15, MNP/Ropiv/Mag30, and MNP/Ropiv/Mag60, respectively). For the MNPs alone vehicle (group MNP/Mag30), the magnet was removed after 30 minutes. After the designated times, the isoflurane anesthesia was terminated; thermal nociceptive testing was started 5 to 10 minutes after this, at which time the animals had returned to baseline wakefulness. The 3 magnet application times were chosen based in part on the complexes’ in vitro drug release profile.
The effect of magnet-directed concentration of the MNP/Ropiv was compared with the effect of conventional ankle block with 0.1% or 0.2% ropivacaine (groups Ropiv 0.1% and Ropiv 0.2%), which served as benchmarks. Ankle block in humans is well described in standard anesthesia texts and is achieved with 5 injections around the ankle. The nerve supply of the rat’s hindpaw has close similarities to that of the human foot.7 Therefore, ropivacaine was given in 5 injections at the ankle in the isoflurane-anesthetized animal, using the malleoli as landmarks (Supplementary Online Methods, http://links.lww.com/AA/A869). A total of 0.8 mL was injected with a 1-mL syringe and 30-G hypodermic needle. The isoflurane anesthesia was then terminated, and testing for ankle block commenced after the animals returned to baseline wakefulness 5 to 10 minutes later.
Preparation of MNP/Ropiv Complexes
Polymer nanogels based on poly (diethylene glycol) methyl ether methacrylate (PM(EO)2MA) were synthesized by atom transfer radical polymerization (ATRP) in miniemulsion.8–14 Magnetic Fe3O4 nanoparticles modified with oleic acid (about 15 nm in diameter with narrow size distribution) were physically incorporated into the nanogels, forming MNPs complexes. MNP/Ropiv complexes were then prepared as published previously using the dye Rhodamine B as a model for hydrophilic drugs,15 the scheme for which is shown in Figure 1, upper panel. The Supplementary Online Methods (http://links.lww.com/AA/A869) provide a detailed description of the preparation and drug content and release study of the MNP/Ropiv complexes.
Two permanent neodymium magnets were used: a ring magnet (1-in OD, ½-in ID, ¼-in thick, product NR 014, Applied Magnets, Plano, TX) and a disk magnet (7/16-in OD × 3/16-in thick, product D73, K&J Magnetics, Inc., Philadelphia, PA). The surface fields of the 2 magnets are shown in Figure 2. We have no information about their tissue penetration. The disk magnet was placed immediately below the ring magnet, and both were placed over a ½-in long sleeve cut from a polyvinylchloride 9.0-mm OD endotracheal tube. The assembly was placed in such a way that the ring surrounded the ankle at the malleoli, and the disk overlay the proximal part of the dorsum of the paw. The ring magnet was used because ankle block in humans is essentially a ring block. The disk magnet was added because of the possibility of the ring not adequately covering the injection sites across the dorsum of the paw.
Assessment of Ankle Block
Anesthesia was assessed using the well-established method of Hargreaves et al.’s16 plantar thermal nociception assay and the IITC Plantar Test Analgesia Meter (IITC, Woodland Hills, CA). Before anesthetic injection, animals were placed in individual acrylic testing chambers on an elevated glass platform in a dimly lit, quiet room and tested for baseline sensitivity to radiant thermal stimuli focused on the plantar surface of the hindpaw from below. The active intensity of the apparatus was set to give a baseline withdrawal latency of 8 to 10 seconds. A cutoff time of the radiant heat source was set at 20 seconds to avoid tissue damage in the event that the animal did not respond. The latency to withdrawal was recorded to the nearest 0.1 second. Each hindpaw was tested 3 times in alternation and the mean withdrawal latency determined for each paw. Testing was done every 10 minutes for 120 minutes. All sensory testing was performed by a single investigator blind to the treatment condition of the animal.
The concentration of ropivacaine in plasma and ankle tissue was determined at various time points after IV injection of 2 mL of the following: 0.05% plain ropivacaine (group Ropiv 0.05%), and MNP/Ropiv with or without 30 minutes magnet application at the right ankle (groups MNP/Ropiv/Mag30 and MNP/Ropiv, respectively). In 2 animals not included in the analysis, IV injection of 1.2 mg plain ropivacaine (2.4 mL of 0.05%) resulted in immediate respiratory arrest; this did not occur with IV injection of 1 mg (2 mL of 0.05%) of the drug. All pharmacokinetics experiments were performed with the animals under 2% isoflurane anesthesia in oxygen with spontaneous ventilation. A jugular venous catheter was inserted for blood withdrawal and was flushed with 0.8 to 1.0 mL heparinized saline after each withdrawal.
To determine the concentration of ropivacaine in ankle tissue, the skin and subcutaneous tissue all around the right ankle were dissected 30 minutes after IV injection (n = 2/group). Blood samples (0.3 mL) were also drawn from these animals before and 15 and 30 minutes after completion of the injections. To determine the concentration of ropivacaine in plasma (separate experiments and rats, 3 animals per group), 0.3 mL blood samples were collected before and 15, 30, 60, 120, 180, 240, and 300 minutes after completion of the injections. The sample sizes above were chosen to collect pilot data for this study. Our Supplementary Online Methods (http://links.lww.com/AA/A869) provide details of methodology of the ropivacaine assay of plasma and ankle tissue.
Throughout the ankle block and pharmacokinetics experiments, the animals were continuously watched for gross clinical toxicity (seizures and/or cardiorespiratory arrest).
A random-effects model was used to assess the significance of treatment group differences in withdrawal latency times using repeated measures (across time points from 10 through 120) within either the treated (right) paw or untreated (left) paw. More specifically, the measure of interest was the withdrawal latency for a given paw at that time point minus its mean baseline withdrawal latency. Sample size estimation for behavioral sensory testing experiments was based on finding paired differences in withdrawal latencies of at least 7.4 seconds between baseline and 30 to 60 minutes after treatment (duration of “dense block”). This assumption was based on an earlier study with sciatic nerve block with plain ropivacaine.17 We had no previous information on the variability of paired differences with peripheral nerve blocks with MNPs. We therefore assumed a conservative value for standard deviation (SD), of 50% of the mean difference, that is 3.7 seconds. In addition, an α error of 0.05 and power of 0.8 were used for the sample size estimation. In addition, paired differences between treated and untreated paws were also assessed within each treatment group. For these 2 sets of analyses, we first tested the significance of the overall group effect in the regression model. If that group’s differences as a whole were significant, we then tested differences between each pair of groups (e.g., Ropiv 0.1% vs Ropiv 0.2%) and controlled for multiple testing using the false discovery rate (FDR) criteria (allowing for a 5% expected FDR). The FDR was therefore applied across 15 different tests. Tabled results will indicate the original P values and will identify which of the P values were still significant after comparing to the FDR criteria. Based on the general shape of the raw data plots, a separate quadratic polynomial, which best accounted for continuous trends over time, was fit to each of the 6 treatment groups. Our next goal was then to use the confidence interval (CI) (for a given regression equation that corresponds to one of the 6 treatment groups) to identify when the predicted latency was significantly elevated above the baseline value. Using the CIs based on the regression model fit to all data points represents a more efficient way to identify the cutoff value for where latencies are significantly increased (as opposed to doing repeated testing).
The minimum time point at which the lower 95% CI crosses above 0 represents the time at which the treatment is statistically different from baseline. Stata version 11 (StataCorp LP, College Station, TX) was used for all polynomial modeling and testing. For pharmacokinetics, the WinNonlin® program (Pharsight Corporation, St. Louis, MO) was used to calculate various parameters as per standard procedures. Results were deemed statistically significant at α = 0.05.
The baseline withdrawal latencies for the untreated paw were: MNP/Mag30–9.7 seconds, Ropiv 0.1%–8.5 seconds, Ropiv 0.2%–9.1 seconds, MNP/Ropiv/Mag15–9.6 seconds, MNP/Ropiv/Mag30–11.0 seconds, and MNP/Ropiv/Mag60–10.1 seconds. The baseline withdrawal latencies for the treated paw were: MNP/Mag30–9.6 seconds, Ropiv 0.1%–8.1 seconds, Ropiv 0.2%–10.1 seconds, MNP/Ropiv/Mag15–9.2 seconds, MNP/Ropiv/Mag30–9.9 seconds, and MNP/Ropiv/Mag60–11.2 seconds. The raw data for withdrawal latencies for each of the treatment groups were plotted separately for the treated and untreated paws (Fig. 3). Figure 3 also quantifies the variability at each time point by plotting the lower halves of the 95% CIs. To model the withdrawal latencies over time more efficiently, the quadratic regression lines (using a mixed model with separate estimates and plots for each treatment), and lower 95% CIs, were also plotted separately for the treated and untreated paws (Fig. 4).
The mixed effects model was also used to test the outcome of paired differences between the treated and untreated paws. Results were highly significant for the group of animals with 30-minute magnet application after IV injection of MNP/Ropiv (group MNP/Ropiv/Mag30), and for the benchmark groups Ropiv 0.1%, and Ropiv 0.2% (at P < 0.0001 for all 3 groups, all of which remain significant after the FDR correction). They were not significant for MNP/Mag30, MNP/Ropiv/Mag15, or MNP/Ropiv/Mag60 (at P = 0.51, 0.64, and 0.24, respectively). Note that while each group consisted of 6 animals, MNP/Ropiv/Mag30 had 11, because it comprised 2 replication groups combined, each of which displayed similar results to the other subgroup and the entire group.
Overall treatment group effect was highly significant (P < 0.0001) within the treated paw, so each of the 15 pairwise comparisons between the groups was conducted to assess which groups were significantly different from each other. Table 1 gives the P values from the random effects model and the result with the FDR adjustment. In summary, 11 of the 15 comparisons were statistically significant; MNP/Ropiv/Mag30 was not significantly different from Ropiv 0.1% or Ropiv 0.2%, and MNP/Mag30, MNP/Ropiv/Mag15, and MNP/Ropiv/Mag60 were not different from one another. Overall treatment group effect was not significant for the untreated paw (P = 0.16), so no further statistical analysis was done strictly within the untreated paw.
In terms of identifying the time points at which a given treatment group is significant for a given paw, refer to the lower 95% CIs in Figure 4. For the treated paw, compared with their baseline latencies, MNP/Ropiv/Mag30, Ropiv 0.2%, Ropiv 0.1%, and MNP/Ropiv/Mag60 had significantly increased withdrawal times for (up to, but not including), 100, 100, 60, and 50 minutes, respectively. Neither MNP/Mag30 nor MNP/Ropiv/Mag15 had elevated withdrawal times. For the untreated paw, only MNP/Ropiv/Mag30 and MNP/Ropiv/Mag60 had withdrawal times that were elevated, up to 100 and 20 minutes, respectively.
The final aqueous suspension was composed of 0.7% w/v ropivacaine and 0.8% w/v MNPs containing 12% w/w Fe3O4 and 0.069% w/v elemental iron. Ropivacaine release kinetics showed that at 37°C about 43% of the drug was released in the first 30 minutes, 50% in the first hour, and 71% in 2 hours. In contrast, at 0°C and 18°C, only about 18% of the ropivacaine was released in the first 30 minutes and <40% in 6 hours (Fig. 1, lower panel).
Thirty minutes after IV MNP/Ropiv injection, the absolute concentration of ropivacaine in the ankle tissue and the ankle tissue-to-plasma concentration ratio were higher in the group with magnet application (MNP/Ropiv/Mag30) compared with the group without magnet application (MNP/Ropiv) (Table 2). The ropivacaine concentration in ankle tissue was higher in the group injected with 1 mg plain ropivacine IV (group Ropiv 0.05%); however, the ankle tissue-to-plasma concentration ratio was lower than in the MNP/Ropiv/Mag30 group (Table 2).
After IV administration of plain ropivacaine or of MNP/Ropiv with or without magnet application, the plasma concentration of ropivacaine peaked at the first time point assessed 15 minutes after injection and declined in an exponential manner in all 3 groups (Fig. 5).
Note that the plasma concentrations of ropivacaine are higher in the group that received 1 mg plain ropivacaine even though the groups injected with MNP/Ropiv received 14 mg ropivacaine associated with the MNPs. The concentration of ropivacaine in plasma was similar between the 2 MNP/Ropiv groups (Fig. 5 and Table 3), and there were no significant differences in any pharmacokinetic parameters. This lack of difference in plasma ropivacaine concentrations indicates that the systemic exposure was not significantly different in the 2 groups. However, the higher ankle tissue-to-plasma ropivacaine ratio in the MNP/Ropiv group with magnet application suggests sequestration of the drug locally by the magnet.
In this study, we produced ankle block in the rat by IV injection of MNP/Ropiv complexes with magnet application at the ankle. Possible disposition of the complexes at the ankle could be that with each cardiac cycle, they are sequestered by the liver, spleen, other organs, and by the magnet. In the next several minutes, the complexes permeate the capillaries and accumulate in the ankle tissue. Irreversible ropivacaine release occurs; only the free drug is expected to be active. The released drug is no longer under the magnet’s influence; some of it acts on the nerves with which it comes in contact, and some gets reabsorbed into the circulation. The MNPs, with or without the drug, and the polymer nanogels also eventually get reabsorbed.
Thirty minutes of magnet application to the ankle of animals injected with MNP/Ropiv was more effective than 15 or 60 minutes of magnet application (Table 1, Figs. 3 and 4). This may be explained by the drug release profile of the complexes. The maximum ropivacaine release occurred in the first 30 minutes (43%), after which it slowed to 7% in the next 30 minutes and 21% in the subsequent 60 minutes. In the MNP/Ropiv-injected group of animals with 60 minutes of magnet application (MNP/Ropiv/Mag60), the thermal sensitivity in the right, magnet-treated paw was significantly reduced from baseline sensitivity, but there were no right-to-left paw differences. Although speculative, one possible reason for this could be that some of the ropivacaine that accumulated at the ankle in the first 30 minutes was reabsorbed in the subsequent 30 minutes, without adequate replacement by further release from intact complexes. Indeed, the half-life of ropivacaine in plasma was approximately 1.5 hours. The sensitivity of untreated paws (without magnet application) of MNP/Ropiv-injected animals with 30 and 60 minutes of magnet application was significantly affected compared with their baseline. This is likely due to systemic exposure to the circulating ropivacaine. Finally, 15 minutes of magnet application did not produce effective anesthesia in animals injected with MNP/Ropiv, likely because either the ropivacaine release or the time of exposure of the nerves to the released ropivacaine, or both, was not adequate.
It is highly improbable that the mechanism of the paw anesthesia in the magnet-treated paw is similar to that of IV regional anesthesia (Bier block), with the magnet ring acting as a tourniquet; there are fundamental differences between the 2. In Bier block, the injection is made distal to the tourniquet, and the anesthesia disappears as soon as the tourniquet is released. In our experiments, the injection was made proximal to the magnet ring “tourniquet,” indeed into the systemic circulation, and the block persisted long after the magnet was removed.
The half-life, apparent volume of distribution, and apparent clearance of ropivacaine after IV administration of plain ropivacaine are comparable to values published in the literature.18 The dose of plain drug (1 mg) tolerated was much smaller than that found by other investigators.19–21 This may be because the animals in our experiments were under isoflurane anesthesia (with its associated respiratory depression) and spontaneous ventilation, while in the other studies, they were either under anesthesia with controlled ventilation20,21 or awake.19,20
Even though the groups injected with MNP/Ropiv received doses of ropivacaine that were 14 times higher than the 1 mg dose received in the group injected IV with plain ropivacaine, the maximum plasma concentration (Cmax) in the former groups was lower by a factor of 5 to 7.5. This is consistent with studies in which bupivacaine22 and ropivacaine23 formulated in liposomes also showed lower plasma drug concentrations compared with plain drug administration. The absolute concentration of ropivacaine in ankle tissue and the ankle tissue-to-plasma concentration ratio after MNP/Ropiv injection were higher in the group with, than in the group without, magnet application, as would be expected. However, the absolute ankle tissue drug concentration in animals injected with only 1 mg of plain ropivacaine was higher by a factor of 7 to 10 compared with MNP/Ropiv-injected animals. This is surprising, but one possible explanation is that the dissected tissue in all groups was contaminated with capillary blood. In future research, alternate methods of isolating the ankle tissue to avoid blood contamination may need to be considered.
An interesting and exciting observation was that the gross clinical safety of ropivacaine when combined with MNPs may be increased at least 14 times compared with the plain drug. There were no seizures during the experiments and no mortality 24 hours or even up to 2 weeks after MNP/Ropiv injection containing 14 mg of ropivacaine, based on none of 23 animals dying during that period in normal housing conditions. The reason for the higher gross clinical safety with MNP/Ropiv could be that the drug is active only in its free form; the rats therefore had much less systemic drug exposure. However, the drug should eventually (beyond the 5 hours of blood sampling done in the study) get released into the circulation and could produce cardiorespiratory arrest. This did not happen; the reason could be that the ropivacaine metabolism was greater than its release from MNPs, resulting in subtoxic plasma drug levels. In fact, we used a much smaller dose of ropivacaine than that used in some toxicity studies in rats.19,21
We have not yet studied possible toxic effects of degradation products of the complexes. However, excellent biocompatibility of ferrofluids has been established by Weissleder et al.24 at an elemental iron dose of 168 mg/kg in rats and dogs, and by Lübbe et al.2 in a dose of 30 mg Fe/kg in rats and 120 mg Fe/kg in mice. In contrast, we used a much smaller dose of 1.38 mg, or 4 to 4.5 mg Fe/kg. Clinical tests in humans have also showed excellent biocompatibility of ferrofluids.5
We chose the ankle to test the hypothesis because the nerves here are superficial, with a high likelihood of the drug coming in contact with them with magnet application on the skin surface. In other areas of the body, muscle, adipose tissue, or a “sheath” covering the nerves might intervene between the magnet and the nerves and make it difficult to achieve the block. Although any local anesthetic drug can be combined with MNPs, we chose ropivacaine because it is between lidocaine and bupivacaine in safety and potency.
Future research will address limitations of the current study, which include lack of a formal toxicity study of MNP/Ropiv, a separate experimental group with ankle injections of saline to serve as negative controls, another group with MNP/Ropiv IV injections without magnet application (although the left paws served this purpose), determination of ropivacaine concentration in the control ankle and in major organs, and histological examination of the ankles and major organs for iron deposition.
In conclusion, we have established proof of principle that it is possible to produce ankle block in the rat by IV injection of MNP/Ropiv complexes and magnet application at the ankle. Another exciting observation was that the safe dose of ropivacaine when combined with MNPs (as judged by gross clinical effects) was higher by a factor of at least 14 when injected IV. If it is shown to be safe in humans, there is a potential for translation of this technique to clinical practice.
Name: Venkat R. R. Mantha, FFARCSI.
Contribution: This author helped conceive and design and conduct the study, review the original study data, and write, review, and edit the manuscript.
Attestation: Venkat R.R. Mantha has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Harsha K. Nair (Undergraduate Neuroscience Student)
Contribution: This author helped conduct the study and collect the data.
Attestation: Harsha K. Nair has approved the final manuscript.
Name: Raman Venkataramanan, PhD
Contribution: This author helped design the pharmacology part of the study, assay and analyze the blood and tissue samples, and helped write, review, and edit the manuscript.
Attestation: Raman Venkataramanan has approved the final manuscript.
Name: Yuan Yue Gao, MS.
Contribution: This author helped analyze the samples, conduct pharmacokinetic calculations and write the manuscript.
Attestation: Yuan Yue Gao approved the final manuscript.
Name: Krzysztof Matyjaszewski, PhD.
Contribution: This author helped design the MNP/Ropiv complexes and write, review, and edit the manuscript.
Attestation: Krzysztof Matyjaszewski has approved the final manuscript.
Name: Hongchen Dong, PhD.
Contribution: This author helped prepare and analyze the MNP/Ropiv complexes and write the manuscript.
Attestation: Hongchen Dong has approved the final manuscript.
Name: Wenwen Li, PhD.
Contribution: This author helped prepare and analyze the MNP/Ropiv complexes and write the manuscript.
Attestation: Wenwen Li has approved the final manuscript.
Name: Doug Landsittel, PhD.
Contribution: This author helped in the statistical analysis of the ankle block experiments and helped write the manuscript.
Attestation: Doug Landsittel has approved the final manuscript.
Name: Elan Cohen, MS.
Contribution: This author helped in the statistical analysis of the ankle block experiments and helped write the manuscript.
Attestation: Elan Cohen has approved the final manuscript.
Name: William R. Lariviere, PhD.
Contribution: This author helped design the anesthesia assessment portion of the study, conduct the study, review the original study data, and write, review, and edit the manuscript.
Attestation: This author has seen the original study data, reviewed the analysis of the data, approved the final manuscript, and is the author responsible for archiving the study files.
This manuscript was handled by: Terese T. Horlocker, MD.
The authors gratefully acknowledge the following people: Dr. Gerald F. Gebhart, PhD., Professor and Director, Center for Pain Research, Department of Anesthesiology, University of Pittsburgh, for acting as an advisor and guide throughout the study; Stacy Lee Cashman and Lacey M Geibel, veterinary surgical technicians, University of Pittsburgh, for their help with anesthesia and central line placement.
1. Alexiou C, Arnold W, Klein RJ, Parak FG, Hulin P, Bergemann C, Erhardt W, Wagenpfeil S, Lübbe AS. Locoregional cancer treatment with magnetic drug targeting. Cancer Res. 2000;60:6641–8
2. Lübbe AS, Bergemann C, Huhnt W, Fricke T, Riess H, Brock JW, Huhn D. Preclinical experiences with magnetic drug targeting: tolerance and efficacy. Cancer Res. 1996;56:4694–701
3. Pulfer SK, Gallo JM. Enhanced brain tumor selectivity of cationic magnetic polysaccharide microspheres. J Drug Target. 1998;6:215–27
4. Widder KJ, Morris RM, Poore G, Howard DP Jr, Senyei AE. Tumor remission in Yoshida sarcoma-bearing rts by selective targeting of magnetic albumin microspheres containing doxorubicin. Proc Natl Acad Sci U S A. 1981;78:579–81
5. Lübbe AS, Bergemann C, Riess H, Schriever F, Reichardt P, Possinger K, Matthias M, Dörken B, Herrmann F, Gürtler R, Hohenberger P, Haas N, Sohr R, Sander B, Lemke AJ, Ohlendorf D, Huhnt W, Huhn D. Clinical experiences with magnetic drug targeting: a phase I study with 4’-epidoxorubicin in 14 patients with advanced solid tumors. Cancer Res. 1996;56:4686–93
6. Wilson MW, Kerlan RK Jr, Fidelman NA, Venook AP, LaBerge JM, Koda J, Gordon RL. Hepatocellular carcinoma: regional therapy with a magnetic targeted carrier bound to doxorubicin in a dual MR imaging/conventional angiography suite–initial experience with four patients. Radiology. 2004;230:287–93
7. Greene EC Anatomy of the Rat (American Philosophical Society Transactions, New Series; No.27). 1955 New York Hafner
8. Averick SE, Magenau AJD, Simakova A, Woodman BF, Seong A, Mehl RA, Matyjaszewski K. Covalently incorporated protein-nanogels using AGET ATRP in an inverse miniemulsion. Polymer Chemistry. 2011;2:1476–78
9. Matyjaszewski K. Atom transfer radical polymerization (ATRP): current status and future perspectives. Macromolecules. 2012;45:4015–39
10. Matyjaszewski K. Atom Transfer Radical Polymerization: From Mechanisms to Applications. Isr J Chem. 2012;52:206–20
11. Matyjaszewski K, Spanswick JMatyjaszewski K, Möller M. Copper-Mediated Atom Transfer Radical Polymerization. Polymer Science: A Comprehensive Reference. 2012 Amsterdam Elsevier BV:377–428
12. Matyjaszewski K, Tsarevsky NV. Nanostructured functional materials prepared by atom transfer radical polymerization. Nat Chem. 2009;1:276–88
13. Matyjaszewski K, Xia J. Atom transfer radical polymerization. Chem Rev. 2001;101:2921–90
14. Siegwart DJ, Oh JK, Matyjaszewski K. ATRP in the design of functional materials for biomedical applications. Prog Polym Sci. 2012;37:18–37
15. Dong H, Mantha V, Matyjaszewski K. Thermally Responsive PM(EO)2MA Magnetic Microgels via Activators Generated by Electron Transfer Atom Transfer Radical Polymerization in Miniemulsion. Chem Mater. 2009;21:3965–72
16. Hargreaves K, Dubner R, Brown F, Flores C, Joris J. A new and sensitive method for measuring thermal nociception in cutaneous hyperalgesia. Pain. 1988;32:77–88
17. Brummett CM, Padda AK, Amodeo FS, Welch KB, Lydic R. Perineural dexmedetomidine added to ropivacaine causes a dose-dependent increase in the duration of thermal antinociception in sciatic nerve block in rat. Anesthesiology. 2009;111:1111–9
18. Guilhaumou R, Boulamery A, Deluca B, Deturmeny E, Bruguerolle B. Effects of induced hyperthermia on pharmacokinetics of ropivacaine in rats. Fundam Clin Pharmacol. 2010;24:463–8
19. Danielsson BR, Danielson MK, Böö EL, Arvidsson T, Halldin MM. Toxicity of bupivacaine and ropivacaine in relation to free plasma concentrations in pregnant rats: a comparative study. Pharmacol Toxicol. 1997;81:90–6
20. Dony P, Dewinde V, Vanderick B, Cuignet O, Gautier P, Legrand E, Lavand’homme P, De Kock M. The comparative toxicity of ropivacaine and bupivacaine at equipotent doses in rats. Anesth Analg. 2000;91:1489–92
21. Ohmura S, Kawada M, Ohta T, Yamamoto K, Kobayashi T. Systemic toxicity and resuscitation in bupivacaine-, levobupivacaine-, or ropivacaine-infused rats. Anesth Analg. 2001;93:743–8
22. Yu HY, Li SD, Sun P. Kinetic and dynamic studies of liposomal bupivacaine and bupivacaine solution after subcutaneous injection in rats. J Pharm Pharmacol. 2002;54:1221–7
23. Shen Y, Ji Y, Xu S, Chen da Q, Tu J. Multivesicular liposome formulations for the sustained delivery of ropivacaine hydrochloride: preparation, characterization, and pharmacokinetics. Drug Deliv. 2011;18:361–6
24. Weissleder R, Stark DD, Engelstad BL, Bacon BR, Compton CC, White DL, Jacobs P, Lewis J. Superparamagnetic iron oxide: pharmacokinetics and toxicity. AJR Am J Roentgenol. 1989;152:167–73