`
`Compatibility of nitroglycerin, diazepam and chlorpromazine
`with a new multilayer material for infusion containers
`N.K. Kambiaa, T. Dinea,∗
`
`, T. Dupin-Sprieta, B. Gressiera, M. Luyckxa, F. Goudaliezb, C. Bruneta
`a Facult´e des Sciences Pharmaceutiques et Biologiques, Laboratoire de Pharmacologie, Pharmacocin´etique et Pharmacie Clinique,
`3 Rue du Professeur Laguesse, B.P. 83, 59006 Lille Cedex, France
`b Laboratoires MACOPHARMA, 96 Rue du Pont-Rompu, B.P. 464, 59338 Tourcoing Cedex, France
`
`Received 15 June 2004; accepted 18 October 2004
`Available online 28 November 2004
`
`Abstract
`
`The stability and compatibility of three drugs: nitroglycerin, diazepam and chlorpromazine, with a new multilayer infusion bag were studied.
`The study was carried out comparatively with PVC bags with which these drugs are incompatible. The drugs were diluted in 5% dextrose or
`in 0.9% sodium chloride isotonic solutions. Solutions were stored during 8 or 48 h with or without any protection against light. Remaining
`concentrations of drug were determined by high-performance liquid chromatography (HPLC) during the storage. The admixtures were also
`monitored for precipitation, color change and pH. Whatever the isotonic solution used, the loss of drugs is in discredit of the use of PVC bags
`for their storage. So, these three drugs would not be stored in PVC bags. In multilayer bags, no loss of drugs and no color change were detected
`throughout the storage period. pH values were stable during the same storage period. These three drugs were compatible with multilayer bags
`in all tested conditions for 8 or 48 h. The leaching of the plasticizer di-(2-ethylhexyl) phthalate (DEHP), that is incorporated into PVC to
`make the bags soft and pliable was not detected in the three drug solutions during storage period. Our study confirms that these three drugs
`are incompatible with PVC bags, on the contrary the new materiel tested was proved to be interesting for drug storage.
`© 2004 Elsevier B.V. All rights reserved.
`
`Keywords: Compatibility; Nitroglycerin; Diazepam; Chlorpromazine; Multilayer bags; PVC bags
`
`1. Introduction
`
`Polyvinyl chloride (PVC) materials for infusion bags are
`commonly used for the administration of infusion drug ad-
`mixtures because they offer several advantages over conven-
`tionnal glass containers, such as easier storage and shipping
`because of their resistance of breakage. However, there are
`still some drawbacks to resort to them because possible drug-
`package interactions can occur. Two major problems are often
`discussed in the literature. The first one is that some com-
`pounds of plastic as plasticizers can be leached into solutions.
`Several studies have shown the leaching of di-(2-ethylhexyl)
`
`∗
`
`Corresponding author. Tel.: +33 320 96 40 40; fax: +33 320 96 97 52.
`E-mail addresses: nicolas.kambia@libertysurf.fr
`(N.K. Kambia),
`tdine@pharma.univ-lille2.fr (T. Dine).
`
`0731-7085/$ – see front matter © 2004 Elsevier B.V. All rights reserved.
`doi:10.1016/j.jpba.2004.10.020
`
`phthalate (DEHP) from PVC bags [1,2]. DEHP is the pre-
`dominant plasticizer used to make the bags soft and pliable.
`This plasticizer is known to be responsible for change in struc-
`ture and function of liver in animals, reduction body weight
`and liver weight in adult male rats [3,4]. In rats, DEHP is
`both a male and female reproductive toxicant. Data from few
`studies in rodents reported that phthalates effects on repro-
`ductive cells are influenced by the stage of development at
`exposure [5,6]. Non-PVC bags such as multilayer containers
`do not contain plasticizers and are recommended by manu-
`facturers for admixture of total nutrient solution, containing
`lipophilic constituents, used for delivering total parenteral
`nutrition. The other main problem is the possible adsorption
`of drugs on the inner plastic surface of container [7,8] leading
`to the loss of drugs and a decrease of the injection concentra-
`tions. Therapeutic consequences can be observed because the
`
`Hospira, Exh. 2019, p. 1
`
`
`
`260
`
`N.K. Kambia et al. / Journal of Pharmaceutical and Biomedical Analysis 37 (2005) 259–264
`
`patient does not receive the dose, which has been prescribed
`him and the precise dose itself is unknown.
`Furthermore, stability of the drug and the possible leach-
`ing of plasticizer can sometimes be related to the choice of
`the solution used for dilution [9,10] or to the volume of bag
`[11]. So, for every plastic material and each drug, it is nec-
`essary to carry out some tests of compatibility and stability.
`Indeed, the nature of the plastic material is the most impor-
`tant factor, since it determines the nature and the amount of
`drug binding. It is recognized that PVC causes the largest
`and the most numerous interactions with drugs, whereas
`polyethylene or polypropylene materials are considered more
`compatible. So, multilayer plastic materials would pro-
`vide useful solutions: maintaining physical properties, com-
`patibility with contents, protection against exterior agents
`[12].
`The aim of this study was to evaluate the compatibility
`between a new multilayer materials for infusion containers
`(named M312) and three drugs: nitroglycerin, diazepam and
`chlorpromazine. These drugs have been chosen according to
`their known uncompatibility with PVC bags. For instance,
`the loss of potency of nitroglycerin after 24 h was almost
`30% [8]. Remaining concentrations of diazepam after 24-
`h storage was only 60% of initial concentration [8]. It was
`also proved that chlorpromazine was able to interact with
`PVC container [13]. For these reasons, bags made of multi-
`layer materials have been manufactured for some years [12].
`Among them, M312 is a film made of five layers which might
`be assumed to combine a minor degree of sorption and a
`vapour barrier-effect as the inner layer in direct contact with
`the solution is made of a co-polymer of polyethylene and
`polypropylene, and a oxygen barrier-effect which does not
`permit the oxygen to enter and to oxidize the drugs because
`the nature of other layers in the middle and the outside of the
`film.
`This paper presents the compatibility results obtained for
`three drugs administered in infusion solution from PVC and
`multilayer containers.
`
`2. Materials and methods
`
`2.1. Chemicals and drugs
`
`HPLC-grade methanol was obtained from JT Baker (De-
`vender, Holland). The water used was de-ionized and puri-
`fied by Milli-Q Academic water purification (Millipore, Saint
`Quentin en Yvelines, France). Analytical grade ammonium
`acetate was purchased from Prolabo (Paris, France). Nitro-
`glycerin, chlorpromazine, diazepam, as specimen, and clon-
`azepam, used as internal standard, were obtained from Sigma
`(Saint-Quentin Fallavier, France).
`The drugs studied were the commercial products suitable
`for clinical use. Nitroglycerin applied to the treatment of
`angor and cardiac failure, chlorpromazine hydrochloride, a
`phenothiazine antipsychotic drug and diazepam, a benzodi-
`
`azepine anxiolytic drug were provided in injection forms.
`Nitroglycerin was the current clinical formulation Lenitral®
`intravenous injection. It was generously donated by Besins-
`Iscovesco Laboratories (Paris, France) in ampuls of 3 or
`15 mg diluted in, respectively, 2 or 10 ml of injectable sol-
`vent. Chlorpromazine hydrochloride was used as Largactil®
`intravenous injection. It was provided from Specia Rhone-
`Poulenc-Rorer Laboratories (Paris, France) in ampuls of
`25 mg diluted in 5 ml of injectable solvent. Diazepam was
`used as injections of Valium® and provided from Roche Lab-
`oratories (Neuilly sur seine, France) in ampuls of 10 mg di-
`luted in 2 ml of injectable solvent.
`
`2.2. Materials and chromatographic conditions
`
`During the experiments, the remaining concentrations
`of drugs in injection solutions were determined by high-
`performance liquid chromatography (HPLC). Chromato-
`graphic analysis was performed using a HP 1090 M HPLC
`system (Hewlett-Packard, Orsay, France) equipped with a
`variable volume injector, an automatic sampling system and
`a Hewlett-Packard Model 79994A linear photodiode array
`UV detector operating at suitable wavelengths. The output
`from the detector was connected to a Hewlett-Packard 9000
`model 300 integrator to control data acquisition and inte-
`gration. Retention times and peaks areas were determined
`by a computer connected to a Hewlett Packard Thinkjet ter-
`minal printer. Drugs analyses were performed on a 5 m
`Hypersil BDS C18 column (150 mm× 4.6 mm) (Life Sci-
`ences International, Epargny, France), operating at room
`temperature.
`• Nitroglycerin separation was based on an isocratic method
`using a mobile phase consisting of methanol and wa-
`ter (50/50, v/v). After degassing with helium stream for
`15 min, the mobile phase was pumped through the column
`−1. Samples (20 l) were in-
`at a flow rate of 1.5 ml min
`jected into the analytical column and the chromatographic
`separation was achieved with final detection at 215 nm.
`• Diazepam analyses were carried out at 254 nm with a mo-
`bile phase consisting of mixture methanol–water (65/35,
`−1 and degassing
`v/v) pumped at a flow rate of 1.7 ml min
`with helium stream. The volume of injection was 20 l.
`• For chlorpromazine, analyses were carried out at
`254 nm with a mobile phase consisting of the mixture
`methanol–acetate buffer pH 4 (65/35, v/v) pumped at a
`−1. For all drug analysed, clon-
`flow rate of 0.8 ml min
`azepam was used as the internal standard.
`
`The leaching of DEHP was also investigated using previ-
`ous conditions described by Faouzi et al. [1].
`A pH meter was used to measure the pH of injection solu-
`◦
`C.
`tions during storage in PVC or multilayer containers at 20
`It was a model HI 8520 N microprocessor equipped with a
`Micro pH electrode HI 1083 (Hanna Instruments, Lingol-
`stein, France).
`
`Hospira, Exh. 2019, p. 2
`
`
`
`N.K. Kambia et al. / Journal of Pharmaceutical and Biomedical Analysis 37 (2005) 259–264
`
`261
`
`2.3. Preparation of admixtures in multilayer and PVC
`bags
`
`PVC infusion bags (Macoflex) containing dextrose 5%
`injection or sodium chloride 0.9% injection were kindly
`provided by Macopharma Laboratories (Tourcoing, France).
`Multilayer bags were manufactured by Macopharma Lab-
`oratories using a new multilayer film, named M312 made
`of five layers, which is from the inner to the outer side
`of the bags, made of polyethylene, polypropylene and
`polyester. This multilayer material might be assumed to com-
`bine a minor degree of drug sorption as its inner layer is
`made of a polyethylene, and an oxygen barrier-effect and
`vapour barrier-effect because of the nature of some of the
`layers.
`A number of drug ampuls, chosen to obtain concentrations
`of the same magnitude as therapeutics, was added to 5% dex-
`trose or 0.9% sodium chloride isotonic injections, contained
`in PVC bags and in multilayer bags.
`
`• For nitroglycerin, one ampul (3 mg) was added to 50 ml
`of 5% dextrose or 0.9% sodium chloride isotonic so-
`lutions. The solutions were stored at room temperature
`◦
`(20–25
`C) during 48 h without any protection against
`light.
`• For chlorpromazine, 25 mg were added to 100 ml of 5%
`dextrose or 0.9% sodium chloride isotonic solutions. These
`admixtures had to be stored in the dark because the aque-
`ous solutions of phenothiazine-like compounds are very
`quickly oxidized in the daylight. So, the solutions were
`stored in the dark during 8 h because of the photosensitiv-
`ity of the molecule.
`• For diazepam, 20 mg were added to 500 ml of 5% dextrose
`or 0.9% sodium chloride isotonic solutions. The solutions
`◦
`were stored at room temperature (20–25
`C) during 48 h
`without any protection against light.
`
`Optimally, drug compatibility and stability trials should
`include both visual and chemical
`tests. The bags con-
`taining the drug solutions were agitated by bending,
`flexing, massaging and shaking for about 1 min after
`preparation to simulate the agitation that a bag may
`undergo during preparation, transportation and administra-
`tion.
`At specified time intervals, the bags were agitated and
`samples were directly taken from both PVC and multilayer
`bags up to the end of the storage period and placed in clear
`glass test tubes and were visually inspected for color and
`clarity by following European Pharmacopeia protocols V.6.1.
`(1983) and V.6.2. (1980). At the same time, the pH values of
`solutions were measured immediately. Then, samples were
`kept frozen in polypropylene tubes at −20
`◦
`C until analy-
`sis. The remaining concentrations of drug and DEHP con-
`tents were then determined by suitable methods mentioned
`above.
`
`3. Results and discussion
`
`3.1. Chromatography
`
`Drug concentrations were determined by using a stability-
`indicating HPLC assay. All assays were performed isocrati-
`cally at ambient temperature. The compounds were resolved
`with a satisfactory baseline separation under developed con-
`ditions. No, degradation product interfered or was eluted with
`the same retention time of the parent drug peak.
`• Nitroglycerin calibration curve was constructed at a con-
`centration range of 10–30 g/ml. A good linear response
`was found with a correlation coefficient better than 0.999.
`Within-run precision of the method was evaluated by repli-
`cate analysis (n = 5) of different concentrations (10, 15, 20
`and 30 g/ml), it was less than 2.83%.
`• For diazepam, calibration curve was constructed at a
`concentration range of 2.5–15 g/ml and the correla-
`tion coefficient was better than 0.999. The precision was
`validated by establishing the relative standard deviation
`(R.S.D. < 2.61%) with four concentrations (2.5, 5, 10 and
`15 g/ml).
`• In the same way, a good linear response was found for
`chlorpromazine assays. Correlation coefficient was 0.999
`for calibration curve (1.25–10 g/ml) and precision was
`established by determination of R.S.D. (n = 5), with four
`concentrations (1.25, 2.5, 5, 10 g/ml), which was less
`than 2.81%.
`
`Since DEHP is a persistent environmental polluant, rig-
`orous precautions were taken to avoid contamination during
`both sample handling and sample analysis. All the samples
`were prepared and diluted in glass or polypropylene tubes
`washed previously with a methanol/acetonitrile mixture, and
`rinsed with hexane.
`The intra-assay and inter-assay coefficients of variation
`(RDS values) were lower than 0.75 and 4.36%, respectively.
`A good linear response was found with a correlation coeffi-
`cient better than 0.999.
`
`3.2. Stability of nitroglycerin
`
`The analysis of each sample was performed by HPLC
`after a suitable dilution in the mobile phase in order to fit
`the calibration curve. At time zero, the initial concentration
`of nitroglycerin was designated as 100% and all subsequent
`measured concentrations were expressed as percentages of
`the initial concentration. Stability was defined as a concen-
`tration representing 90–105% of the initial one, in accordance
`with the Health Registration of France, the French Regulatory
`Agency for drug and drug-related products. So, drug instabil-
`ity and incompatibility with material in contact were defined
`as a > 10% decrease from the initial drug concentration.
`Fig. 1 shows the behaviour of nitroglycerin in both PVC
`and multilayer bags.
`
`Hospira, Exh. 2019, p. 3
`
`
`
`262
`
`N.K. Kambia et al. / Journal of Pharmaceutical and Biomedical Analysis 37 (2005) 259–264
`
`Fig. 1. Remaining concentrations of nitroglycerin after storage in multilayer or PVC bags.
`
`- In PVC bags, after 4 h, loss of nitroglycerin averaged 30%.
`During a 48-h period, remaining concentration were 20%
`of the initial concentration in dextrose solution and 30 %
`in sodium chloride solution.
`
`These findings indicated that, nitroglycerin was adsorbed
`by the plastic infusion bags that it comes in contact with dur-
`ing storage. Because this adsorption process is very fast with
`nitroglycerin, it is difficult to measure the exact initial con-
`centration. Consequence, the phenomenon could affect the
`response elicited in patient-treated with an i.v. nitroglycerin
`formulation. The results also confirmed that the loss is not
`due to degradation, since the assay method used is precise
`and specific and can detect degradation compounds. The rate
`of loss from sodium chloride injection appears to be slower
`than with dextrose injection. This indicates that adsorption is
`affected by the solution used to deliver nitroglycerin.
`
`- In multilayer bags, no loss was highlighted whatever the
`infusion solution. All concentrations remained above 90%
`of the initial value, and most were near 100%. There was
`no substantial difference between nitroglycerin concentra-
`tions at time zero and at any subsequent time point. Ni-
`troglycerin concentrations after various periods of storage
`showed no loss (>10%) of the drug. In contrast with PVC
`bags, the drug was not adsorbed by the plastic infusion
`bags. In conclusion, nitroglycerin remains stable for up to
`48 h in multilayer bags whatever the infusion solution used.
`
`3.3. Stability of chlorpromazine
`
`As shown Fig. 2, chlorpromazine slightly interacted with
`PVC bags and not with multilayer bags. After 8-h storage,
`about 30% of the drug were lost. Long-term stability tests
`confirmed these results [14]. No difference was noticed be-
`tween both sodium chloride or dextrose solutions. As previ-
`
`ously discussed, the present study is reported so as to alert
`clinicians of this phenomenon and to promote efforts to uti-
`lize non-PVC bags and the shortest possible infusion sets.
`Therefore, chlorpromazine dilutions may be prepared in mul-
`tilayer containers as an alternative to PVC bags and stored in
`the dark.
`
`3.4. Stability of diazepam
`
`The loss of diazepam after 24-h period reached 50% of
`the initial concentration in PVC bags whereas no interac-
`tion was noticed in multilayer bags. So, extremely rapid loss
`of diazepam occurred in PVC containers. The loss was so
`rapid that accurate time-zero determinations were not pos-
`sible. Losses of 50% or more occurred in both dextrose 5%
`injection and in sodium chloride 0.9% injection. In multilayer
`bags, diazepam remained stable for at least 48 h (Fig. 3).
`The leachability of DEHP into intravenous solutions was
`not highlighted with the various studied drugs. Moreover, no
`drugs precipitation or crystallization and no solution color
`change were observed in any solutions stored in PVC or mul-
`tilayer bags. In the same way, no modification of pH was ob-
`served with the various solutions neither with the PVC bags
`nor with the multilayer ones.
`The new materiel tested (multilayer bags) was proved to
`be interesting for drug package. Multilayer bags were man-
`ufactured using a new multilayer film, named M312, which
`provides superior performance for packaging pharmaceuti-
`cal solutions. Several polymers, each contributing a specific
`property, are coextruded into a multiply film. The result is an
`exceptionally clear benefit: exceptional product protection
`with optimum leach resistance, no presence of plasticizers,
`adhesives or chlorine and ultra-low particulate levels. M312
`is a highly inert material exhibiting extremely low extracta-
`bles with a wide variety of solutions. Moreover, this material
`
`Hospira, Exh. 2019, p. 4
`
`
`
`N.K. Kambia et al. / Journal of Pharmaceutical and Biomedical Analysis 37 (2005) 259–264
`
`263
`
`Fig. 2. Remaining concentrations of chlorpromazine after storage in multilayer or PVC bags.
`
`minimises oxygen transfer, improves long-term stability and
`can facilitate 48 h or more therapy for stable patients. Other
`types of multilayer films for their anti-adhesive property are
`used not to manufactured bags, but to manufactured a mate-
`rial making it possible to reduce the bacterial adherence of
`surface of implants [15].
`PVC bags containing injection solutions offer several ad-
`vantages over conventional glass containers, such as easier
`storage and shipping because of their resistance to breakage.
`However, several problems are reported with their use such
`as the loss of substantial amounts of drug from the solution
`by adsorption or absorption onto the plastic bags, and the
`leaching of potentially harmful substances into the solution,
`particularly a plasticizer, DEHP [1,7,8,11,16]. This plasti-
`
`cizer is known to be responsible for change in structure and
`function of liver in animals, reduction body weight and liver
`weight in adult male rats [3,4]. In rats, DEHP is both a male
`and female reproductive toxicant. Data from few studies in
`rodents reported that phthalates effects on reproductive cells
`are influenced by the stage of development at exposure [5,6].
`So, patients on regular exposure of this plasticizer are there-
`fore at risk for toxic consequences.
`The container-content (PVC-drug) interactions with loss
`of drugs were not limited to the three studied drugs. Interac-
`tions of the same type exist with other drugs. Fentanyl cit-
`rate was rapidly lost when admixed with fluorouracil in PVC
`containers, losing about 25% in the first 15 min and about
`50% in the first hour [17]. Clomipramine hydrochloride and
`
`Fig. 3. Remaining concentrations of diazepam after storage in multilayer or PVC bags.
`
`Hospira, Exh. 2019, p. 5
`
`
`
`264
`
`N.K. Kambia et al. / Journal of Pharmaceutical and Biomedical Analysis 37 (2005) 259–264
`
`clorazepate dipotassium salt slightly interact with PVC bags
`[13]. In certain cases, one can note that there is not loss of
`drugs, but a strong leaching of the plasticizer DEHP occurred.
`Thus, strong amounts of DEHP were extracted from PVC
`i.v. infusion bags of taxol, teniposide, 5-FU, quinine and cy-
`closporine [1,11,16,18,19]. The magnitude of leaching of the
`plasticizer appeared to be directely dependent on the area of
`PVC, the duration of storage and the concentrations of Cre-
`mophor. Sometimes, one can note a physical drugs incompat-
`ibility with the PVC bags such as pH change, color change or
`the formation of precipitates [11,16]. All these interactions
`with PVC involving the loss of drugs or the leaching of the
`plasticizer DEHP were not observed with the multilayer bags.
`So, the multilayer bags represent a better alternative in the
`preparation and the storage of the medicamentous formula-
`tions incompatible with PVC bags.
`So, that is why, specific content/container studies are nec-
`essary to prove the drug compatibility with plastic packag-
`ings.
`
`4. Conclusion
`
`This study confirms that these three drugs are incompati-
`ble with PVC container; in contrast, the new materiel tested
`was proved to be interesting for drug package. Moreover,
`it clearly indicates that diazepam and nitroglycerin undergo
`marked adsorption on the PVC bags surface. No significant
`difference was observed between the infusion solutions used
`as vehicles of the drugs. So, nitroglycerin (3 mg/50 ml) and
`diazepam (20 mg/500 ml) in dextrose 5% or sodium chloride
`0.9% solutions could be stored 48 h at room temperature,
`without precaution with light in this multilayer bag. Chlor-
`promazine (25 mg/100 ml) in dextrose or sodium chloride so-
`lution is stable for 8 h in the dark in the multilayer bags.
`This study highlights the interest of multilayer infusion
`bags. Further studies should be conducted with additional
`drugs in order to explore the behaviour of these new materials
`extensively.
`
`Acknowledgements
`
`The technical assistance of F. Khalfi and X. Meersseman
`is acknowledged.
`
`References
`
`[1] M.A. Faouzi, T. Dine, M. Luyckx, B. Gressier, F. Goudaliez, M.L.
`Mallevais, C. Brunet, M. Cazin, J.C. Cazin, Int. J. Pharm. 105 (1994)
`89–93.
`[2] K. Kambia, T. Dine, B. Gressier, S. Bah, A.F. Germe, M. Luyckx, C.
`Brunet, L. Michaud, F. Gottrand, Int. J. Pharm. 262 (2003) 83–91.
`[3] A.E. Ganning, U. Brunk, G. Dallner, Hepatology 4 (1984) 541–547.
`[4] K. Rathinam, S.P. Srivastava, P.K. Seth, J. Appl. Toxicol. 10 (1990)
`39–41.
`[5] P.M.D. Foster, E. Mylchreest, K.W. Gaido, M. Sar, Hum. Reprod.
`Update 7 (2001) 231–235.
`[6] T. Tanaka, Food Chem. Toxicol. 40 (2002) 1499–1506.
`[7] T.D. Sokoloski, C.C. Wu, A.M. Burkman, Int. J. Pharm. 6 (1980)
`63–76.
`[8] J.K. Yliruusi, A.G. Sothmann, R.H. Laine, R.A. Rajasilta, E.R.
`Kristoffersson, Am. J. Hosp. Pharm. 39 (1982) 1018–1021.
`[9] S.D. Pearson, L.A. Trissel, Am. J. Hosp. Pharm. 50 (1993)
`1405–1409.
`[10] L.A. Trissel, Pharmacotherapy 17 (1997) 133s–139s.
`[11] W.N. Waugh, L.A. Triseel, V.J. Stella, Am. J. Hosp. Pharm. 48
`(1991) 1520–1524.
`[12] L. Tchiakp´e, C.B. Airaudo, O.M. Abdelmalik, A. Gayte-Sorbier, M.
`Verdier, J. Guerri, J. Biomater. Sci. Polym. Ed. 7 (1995) 199–206.
`[13] C.B. Airaudo, A. Gayte-Sorbier, C. Bianchi, M. Verdier, Int. J. Clin.
`Pharmacol. Ther. Toxicol. 31 (1993) 261–266.
`[14] B.A. Sarsfield, J.T. Maloy, J. Pharm. Sci. 87 (1998) 1130–1137.
`[15] F. Boulmedais, B. Frisch, O. Etienne, P. Lavalle, C. Picart, J. Ogier,
`J.C. Voegel, P. Schaaf, C. Egles, Biomaterials 25 (2004) 2003–2011.
`[16] T. Dine, F. Khalfi, M. Duban, B. Gressier, M. Luyckx, C. Brunet,
`X. Meersseman, F. Goudaliez, M. Cazin, J.C. Cazin, Biomaterials
`20 (1999) 655–661.
`[17] Q.A. Xu, L.A. Trissel, J.F. Martinez, Ann. Pharmacother. 31 (1997)
`297–302.
`[18] R. Venkataramanan, G.J. Burckart, R.J. Ptachcinski, Am. J. Hosp.
`Pharm. 43 (1986) 2800–2802.
`[19] M.A. Faouzi, F. Khalfi, T. Dine, M. Luyckx, C. Brunet, B. Gressier,
`F. Goudaliez, M. Cazin, J. Kablan, A. Belabed, J.C. Cazin, J. Pharm.
`Biomed. Anal. 21 (1999) 923–930.
`
`Hospira, Exh. 2019, p. 6