]ounra/ of Clinical Pharmacy and Theraptics (1594) 19, 95-100

Stability and compatibility of cisplatin and carboplatin with PVC infusion bags B. Benaji,* T. Dine,* M. Luyckx,* C. Brunet,* F. Goudaliez,t M. L. Mallevais,t M. Cazin,* B. Gressier* and J. C. Cazin* 'Laboratoire de Pharmacologie, Pharmacocinltique et Pharmacie Clinique, Fanrltl des Sciences Pharmaceutiques et Biologiques de Lille, rue du Professeur Laguesse, BOP. 83 59006 Lille Cedex, and tkboratories Macophama, rue du Ponf-Rompu, B.P. 464, 59338 Tourcoing Cedex, France

SUMMARY

The availability and compatibility of drugs from solutions infused via PVC infusion bags through PVC administration sets have been examined. No significant drug loss was observed during simu- lated infusions using PVC infusion bags and administration sets over time periods used in hospitals (cisplatin, 2 h carboplatin, 1 h). The stability of carboplatin was studied in 5% dex- trose. In 0.9% NaCl, we observed that carboplatin could be converted to cisplatin in the presence of chloride ions. With cisplatin, no significant differ- ence was found between infusion solutions (5% dextrose or 0.9% NaCI). The stability of cisplatin (5% dextrose or 0.9% NaCl) and carboplatin (5% dextrose) was also studied in PVC bags after storage in the dark at room temperature. The results show that the drugs were stable over the 9-day storage period studied.

INTRODUCTION

The serendipitous discovery by Rosenberg et aI. (I) that cell replication could be inhibited by platinum complexes led to development and therapeutic use of cisplatin (Cis-diamine dichloroplatinum 11) as an anti- cancer agent. Cisplatin is routinely used in the treatment of testicular, ovarian and bladder cancer. Against a wide range of other malignanaes, such as carcinoma of the head, neck and stomach, or brain h o u r s , further research is needed to establish the drug's usefulness (2, 3). Side-effects of cisplatin

Correspondence Brahim Benaji, Facult6 des Sciences Pharma- ceutiques et Biologiques de M e , Laboratoire de Pharmacologie, PhannacocinCtique et Pharmacie Clinique, me du Proferseur Laguesse, B.P. 83 59006 Lille Cedex, France.

administration include myelosuppression, and neuro- logical and gastrointestinal toxicity, while neph- ropathy is regarded as the dose-limiting factor in humans (4, 5). Platinum, diamine [I,I cyclobutane dicarboxyIato(2-)-0,0'] carboplatin is a second- generation cisplatin analogue, with similar preclinical antiturnour activity (6, 7). Clinical phase I and I1 trials with carboplatin have demonstrated near-equal efficacy as well as reduced nephrokoxicity and nausea, indicating that carboplatin is one of the most promising cisplatin analogues developed to date (8, 9). The chemical structures of these cytostatic agents are shown in Fig. 1.

Pt

( b)

0

Fig. 1. Chemical structures of (a) cisplatin and (b) carbo- platin.

95

96 B. Benaji et al.

There is evidence from recent studies that the therapeutic index of anticancer agents may be improved if traditional bolus injection schedules are replaced with continuous infusion regimes (10, 11). Continuous infusion is also said to be advantageous over intermittent small-volume infusion or intravenous (i.v.) push (12). Therefore, with the increasing use of continuous i.v. infusion and intermittent small-volume i.v. infusion modes of administration, it is imperative that the stability and compatibility of antitumour agents in administration vehicles and PVC containers are investigated. Consequently, when drugs are administered by continuous i.v. infusion with PVC material, knowledge of the rate of drug delivery to the patient is essential (13).

Previous studies (14-16) have reported the loss of certain drugs (diazepam, nitroglycerin) from aqueous solutions stored in PVC infusion bags for various periods of time. Generally, these losses have been attributed to interaction (adsorption or absorption) between the drug and the plastic &sion bag, and, in some cases, may diminish the therapeutic response due to reduced drug delivery to the patient. Other studies have shown a loss of anthracyclines with glass bottles due to adsorption (17, 18).

Only a minimal amount of information is available concerning the compatibility of cisplatin and carbo- platin administered by PVC infusion bags and i.v. administration sets. Such infomation is important when drug solutions are to be infused over long periods of time.

The present study was undertaken with the follow- ing objectives: (i) to survey a range of drugs (platinum complexes), including those presently being adminis- tered by i.v. infusion, for possible interactions with PVC infusion bags; (ii) to study the behaviour of these drugs in simulated infusions using PVC containers and administration sets under conditions of infusion routinely used in hospitals (infusion flow rate, dose, volume, temperature and exposure to light) and (iii) to identify differences in possible interactions between PVC containers and administration sets and to com- pare the stability of the drugs in 09% NaCl or 5% dextrose.

A high-performance liquid chromatography (HPLC) method developed in our laboratory was used in the present study. This method allowed the rapid deter- mination of drugs in infusion solutions (5% dextrose or 09% NaCl) using a sqitable chromatographic column and mobile phase.

MATERIALS AND METHODS

Chemicals

The drug substances studied were commercial prod- ucts suitable for clinical use. Cisplatin (A) was obtained from Roger Bellon Laboratories (Neuilly sur Seine, France) in vials containing 50mg sterile powder for injection. Carboplatin (B) was obtained from BristoI Laboratories in vials containing I50 mg sterile powder for injection. Tetrabutyl-ammonium hydroxide (ion- pairing agent) was obtained from Sigma (France). Water for injection (Htered and sterile) was obtained from Macopharma Laboratories (Tourcoing, France) and used for preparing the buffers, dilutions and standard solutions.

Chromatographic conditions and instrumentation

Chromatographic analysis was performed with an HP 1090 high-performance liquid chromatograph (Hewlett Packard, Orsay, France), equipped with a variable- volume injector, an automatic sampling system and an HP UV detector operating at suitable wavelengths. The output from the detector was connected to a Hewlett Packard 9000 model 300 integrator and the data recorded on an HP Thdqet terminal printer.

Analyses were performed on a 5-pm CIS Hypersil ODS column (100 x 4-6 mm i.d.1 (Interchim, Montlucon France) operating at room temperature.

Cisplatin separation was based on the method of Fleming &K Stewart (19), using a mobile phase (pH 6) consisting of 2% tetrabutyl-ammonium hydroxide (TBAH) (0.4 M) delivered at I d m i n . A reversed- phase ion-pair chromatographic method for determi- nation of carboplatin was developed.

Mobile phase preparation. A 2% (v/v) TBAH (0.9 M) solution was prepared in ultrapure water and adjusted to pH 6-35 with concentrated orthophosphoric acid at a flow rate of 1.6dmin. The mobile phase was filtered and degassed by helium for 15 min. Prior to analytical runs, the mobile phase was pumped through the column for 35 min for equilibration.

The detection wavelengths were 313 MI for cis- platii and 229 nm for carboplatin. In each case 10 pl of sample was injected. Under those conditions the retention times were 1.67 min for cisplatin and 2.06 min for carboplatin.

For simulated infusions, we used a volumetric lnfusion pump (ref. VIP 11) and PVC infusion sets (ref.

Cisplatinlcarboplatin with PVC infusion bags 97

Table 1. Test conditions

Infusion Infusidn bags Infusion bags Infusion pump Infusion . concentration volume concentration flow rate period Special

Cytostatic agents (mg) (ml) (mgh-4 ( d m i n ) (h) conditions

Cisplat in 150 250 0 6 2.08 2 RT+PL Carboplatin 600 250 2.4 4.16 1 RT+PL

RT, room temperature; PL, protection from light.

IS 05) obtained from Becton Dickinson Laboratories, Division Vial Medical (Saint-Etienne de Saint-Geoirs, France). For protection from light during simulated infusions of A and 8, we used opaque PVC infusion sets (ref. PERFECRAN 781547) obtained from Fandre Laboratories (Ludres,, France). Macoflex@ infusion bags of polyvinyl chloride (250 ml) were kindly pro- vided by Macrophama Laboratories (Tourcoing, France). The bags were filled with 0.9% NaCl or 5% dextrose in water to study the possible influence of solvent on the interaction of the drug.

Preparation of standard solutions

To obtain standard stock solutions, A and B were reconstituted with distilled water to give a suitable drug concentration, followed by protection from light. Working solutions were prepared from standard sol- utions of A and B by suitable dilutions with distilled water in polypropylene tubes. Calibration curves were constructed in the range 200-500~.@ml for A and 3O-lSOpg/d for B. Mobile phase was used for all dilutions.

Simulated infusions

The conditions used in the simulated infusions to mimic those routinely employed in hospitals are reported in Table 1.

Infusion solutions of cisplastin and carboplatin were prepared in PVC infusion bags containing 2 5 0 d of 5% dextrose or 0.9% NaCl immediately before the infusion. The bag containing the drug was then attached to an administration set connected to the infusion pump that allowed the solution to flow through at a constant rate. Samples (Id) were withdrawn at specified times from the PVC bags, and

a I-ml aliquot of effluent was simultaneously collected, in polypropylene tubes, from the administration set. Samples were analysed by HPLC immediately.

All simulated infusions were carried out at least in quadruplicate (two fusions in 0.9% NaCl and two infusions in 5% dextrose) for cisplatin and in duplicate (two infusions in 5% dextrose) for carboplatin at room temperature (20-24'C) protected from light.

Studies have shown that the nature of the parenteral solution influences the stability of carboplatin. Thus, the National Cancer Institute (NCI) currently does not recommend the use of 0.9% NaCl injection as diluent, because increased rates of decomposition have been observed (20). However, one report demonstrated that carboplatin 100 pg/ml was stable for 7 days in 0.9% sodium chloride injection (21). Although the loss of 5% of the initial carboplatin content over 24 h is small, it may have clinical importance. Other studies have indicated that carboplatin may be converted to cis- platin in the presence of chloride ions. The possible conversion of this 5% material loss to cisplatin could be serious because of the greater potency and the approximately five times higher toxicity of asplatin. Therefore, the use of chloride-containing infusion fluids for carboplatin is not recommended (22). For this reason, we only studied the stability of carboplatin in 5% dextrose solution.

Storage in infusion b a 5

To infusion bags containing 250 ml of 0.9% NaCl or 5% dextrose solution, a known amount of drugs was added to the following concentrations: cisplatin, 600 pg/ml; carboplatin, 2-4 mg/ml. After mixing the drug in the bag by rapid shaking, samples (I-ml) were withdrawn at regular intervals and stored in poly- propylene tubes at room temperature until HPLC analysis. Infusion bags containing the drug were

98 B. Benaji et al.

Table 2. Validation data of HPLC assay procedure ( n = 5 )

Linear regression Correlation Concentration Average concentration Interassay CV Accuracy equation coefficient (pg1ml) (pg/ml; mean f SD) (%I (%I (y=ax+ b) (4

~

Cisplatin 500 400 300 200

Carboplatin I 80 I20 60 30

~

498.62 f 4.59 400.99 f 2.02 303,44 f 2.37 198.99 f 1.14

180.07 f 2.34 119.86 f 2.10 59.43 f 1.95 32.11 f 1.66

0.92 0.55 0.78 0.5 7

1.30 1.75 2.29 3.29

99.72 y=0*131(~)+0-308 0.999 100.24 101-14 99.49

0.999 100.04 y=O*902(x) i- 0-596 9988 99.04

107.02

stored at room temperature for a period of 9 days with protection from light. Drug storage in these bags was carried out in Om% NaCl and 5% dextrose.

Sample preparation

Before HPLC analysis, each dmg sample was suitably diluted with mobile phase in order to obtain a drug concentration inside the calibration curve.

RESULTS AND DISCUSSION

Chroma topaphy

Cisplatin and carboplatin were rapidly separated, identified and quantified. Cisplatin had a retention time of 1-67 min and carboplatin of 2-06 min. Table 2 summarizes the data validating the assay procedure for each drug. We observed good linearity between peak area ratios and concentrations. The linear calibration curves fitted by the least-squares method gave corre- lation coefficients which were all above 0.999. No signhicant Werences were observed between the equation parameters.

Storage at 4 and - 20'C

In 0.9% sodium chloride solution cisplatin precipitated from solutions containing 0-6-1 mglml and stored at 4 or - 2O'C. At a concentration of 0.1-0-5 mglml, no

precipitation was observed. Therefore, infusion sol- utions should not be at concentrations higher than 500 pg/ml to allow storage under refngeration with- out precipitation.

Stability of drugs during simulated infusions using PVC infusion bags and sets

When solutions of cisplatin and carboplatin were infused through PVC infusion sets from PVC infusion bags over periods normally used in hospitals (2 h and 1 h, respectively), the variation in drug concentration in both the PVC bags and effluent never exceeded 10%. No significant Uerence was observed with respect to drug stability during simulated infusions using 5% dextrose or 0.9% NaCI. It is, however, necessary to use opaque PVC administration sets to avoid the formation of a precipitate in infusion sol- ution. In all cases, no degradation product was detected by HPLC in infusion solution.

StabiZity of drugs in infusion bags during storage at room temperature with protection from light

The concentrations of cisplatin and carboplatin present in solution after various periods of storage in plastic infusion bags at room temperature with protection from light are given in Table 3. No significant dis- appearance of cisplatin (exceeding 10%) was observed in PVC infusion bags for drugs over 9 days of storage,

Cisphfinlcarboplafin wi fh PVC infusion bags 99

Table 3. Concentrations of cisplatin apd carboplatin present in solution after storage in PVC bags at room temperature (n=2)

Cisplatin Carboplatin (150 mg/250 d) (600 mg/25 0 ml)

Infusion solution 0.9% NaCl 5% dextrose 5% dextrose

To 1 day 2 days 3 days 4 days 5 days 6 days

9 days 7 days

150.00 150.19 153.41 14913 ND 152.69 15 1.47 147.98 146.05

150-00 148.50 148.60 149.89 ND 147.09 148.69 147.53 146.43

600.00 608.39 601.74 604.20 604.27 605.80 ND 604.19 602.22

~ ~~ ~ ~-~

ND, none determined.

irrespective of the infusion solution (5% dextrose or 0.9% NaCI). .

During storage at room temperature and with protection from light, no degradation product was detected by HPLC and no suspicious coloration nor precipitation of drugs was observed over 9 days of storage. The pH of the infusion solution was stable during the storage period.

CONCLUSION

In conclusion, the HPLC procedures described in this paper are rapid and reproducible for the determination of cisplati and carboplatin in parenteral solutions. With the increasing use of continuous i.v. infusion of cytostatic agents, the present study has examined the kinetics of cisplati and carboplatin concentrations during simulated infusion using PVC infusion bags and administration sets. The results demonstrate satisfac- tory compatibility of these drugs with PVC infusion material during infusion periods commonly used in hospitals. It is likely that other drugs interact with PVC infusion bags and administration sets, leading to a reduction in the clinical effectiveness of the drug. This type of study is of importance to the packaging of pharmaceuticals in plastic containers in general, and might be carried out for all drugs administered in PVC infusion bags.

ACKNOWLEDGEMENTS

This work was supported by grants of the ComitC du Nord de la Ligue Nationale Francaise Contre le Cancer and the Conseil Regional du Nord-Pas de Calais. The authors also wish to thank Macopharma, Roger Bellon and Bristol Laboratories for their co-operation.

REFERENCES

1. Rosenberg B, Van Camp L, Krigas T. (1965) Inhibition of cell division of l3chrrichia coli by electrolysis products from platinum electrode. Nature, 205, 69& 699.

2. Rozencweig M, Von Hoff DD, Slavik M, Muggia FM. (1977) Cisdiamminedichloroplatinurn (11). A new anticancer drug. Annals of Internal Medicine, 86, 803- 812.

3. Al-Smaf M. (1988) Head and neck cancer: chemo- therapy concepts. Seminars in Oncology, IS, 70-85.

4. Carter SK. (1984) Cisplatin-past, present and future. In: Platinum Coordinafion Complexes in Cancer C h a o - therapy, eds Hacker MP, Douple EB, Krakog IH, pp. 359-376. Martinus Nijhoff Publishing, Boston.

5. Von Hoff DD, Schilsky R, Reichert CM, t t al. (1979) Toxic effects of cisdichlorodiammineplatinum (11) in man. Cancer Treatment Reports, 63, 1527-1531.

6. Micetich KC, Barnes D, Erickson LC. (1985) A compara- tive study of the cytotoxicity and DNA-damaging ef- fects of cis-(diammino)( 1,l-cyclobutanedicarboxy1ato)- platinum (11) and as-Diammine dichloroplatinum (11) on L 1210 cells. Cancer Research, 45,40434047.

7. Lelieveld P, Van Der Vijgh WJF, Veldhuizen RW, et al. (1984) Predinical studies on toxiaty antitumour activ- ity and pharmacokinetics of cisplati and three recently developed derivatives. European ]oumal of Cancer Clinical Oncology, 8, 1087-1104.

8. De Andres Basauri L, Pousa AL, Alba E, Sampedro F. (1984) Carboplatin, an active drug in advanced head and neck cancer. Cancer Treatment Reports, 70, 1173- 1176.

9. Canetta R, Rosencweig M, Carter SK. (1985) Carbo- platin: the clinical spectrum to date. Cancer Treatment Reviews, 12, (Suppl. A) 125-136.

10. V o g e h g NJ (1984) Continuous infusion chemo- therapy: a aitical review. ]ournal of Clinical Oncology, 2, 1289-1304.

11. Lokich JJ. (1987) Introduction to the concept and prac- tice of idusion chemotherapy. In: Cancer Chemotherapy by Infwion, ed. Lokich JJ, pp. 3-11. Precepts Press Inc., Chicago.

12. Carlson RW, Sikic BI. (1983) Continuous infusions or bolus injection in cancer chemotherapy. Annals of Internal Medicine, 99, 823-833.

100 B. Bemji et al.

13. DArcy PF. (1983) Drug interactions and reaction update. D r q s Intelligence and Clinical Pharmacy, 17, 72673 1.

14. Moorhatch P, Chi0 WL. (1974) Interactions between drugs and plastic intravenous fluid bags. I. Sorption studies on 17 drugs. Amnican Journal of Hospital

15. Kowaluk EA, Roberts MS, Blackbum HD, ct al. (1981) Interactions between drugs and polyvinylchloride infu- sion bags. American loumal of Hospital Phnnnacy, 38, 1308-1313.

16. Uum L, Bundgaard H. (1982) Sorption of drugs by plastic infusion bags. In tmt iona l ]ournu1 of P h a m e u - tics, 10, 339-351.

17. Benvenuto JA, Anderson RW, Kerkoff K, et al. (1981) Stability and compatibility of antitumor agents in glass and plastic containers. American ]ouml of Hospital Pharmacy, 38, 1914-1919.

18. Magnam J, Martin ]P. (1988) Les anticanckew sont ils compatibles avec les matikres plastiques? Lethe QS Canchologie, 72, 1-6.

P h a m ~ y . 31, 72-78.

19. Fleming RA, Stewart C. (1990) Stability-indicating high-performance liquid chromatographic method for the simultaneous determination of cisplatin and 5-fluorouracil in 0.9% sodium chloride for injection. ]oumal of Chromatography, 528, 517-525.

20. Trissel LA, Davignon JP, Kleinman LK, et al. (1985) NCI Inocstigationa~ Drugs: Phannaceutical Data 198.5. Bethesda, MD: National Cancer Institute; 135, (NCH publication no. 85-2141).

21. Krd IS, Ding XD, Browerman S, et al. (1983) Trace analysis for cisplatinum anticancer drugs via LCEC. Journal of Chromatography Sciences, 21, 166173.

22. Ardiet C, Bastien G, Favre R, ct al. (1989) Le cisplatin et son analogue. In: Lc guide pratiqw de phamcocinitique cliniqw en oncologic. ed. Frison-Roche, France, pp. 124- 143.