Doxo Stability

Journal of Clinical Pharmacy and Therapeutics (1 990) 15,279-289

STABILITY OF DOXORUBICIN, DAUNORUBICIN AND EPIRUBICIN IN PLASTIC SYRINGES AND MINIBAGS

M. J. Wood,* W. J. Irwin and D. K. Scott Drug Development Research Group, Department of Pharmaceutical Sciences, Aston University, Aston

Triangle, Birmingham B4 7ET, and *Pharmacy Department, Queen Elizabeth Hospital, Queen Elizabeth Medical Centre, Edgbaston, Birmingham BIS 2TH, U.K.

SUMMARY

The shelf lives of doxorubicin, daunorubicin and epirubicin in infusion fluids were studied using high-performance liquid chromatographic (HPLC) methods. Doxorubicin and epirubicin were stable (loss in potency of less than 10%) for 24 and 20 days respectively, when dissolved in sodium chloride solution (0.9%; pH, 6.47) at 25°C and stored in polyvinyl chloride (PVC) minibags, while daunorubicin was stable for at least 43 days. All three drugs were stable for at least 43 days in sodium chloride (0.9%; pH 6.47 and 5.20) and dextrose (5%; pH 4.36) at 4 and -20°C. Repeated thawing and re-freezing of these solutions at ambient temperature did not cause degradation. All three drugs were stable for at least 43 days when recon- stituted with Water-for-Injections BP and stored in polypropylene syringes at 4°C.

INTRODUCTION

Doxorubicin, daunorubicin and epirubicin have been used successfully to produce regression in a wide range of neoplastic diseases. They are often mixed with dextrose (5%) or sodium chloride (0.9%) for intravenous injection and are increasingly being administered as long-term infusions. For this reason, it is important to obtain information on the extended stability of these drugs. Information on the stability of doxorubicin is available in the literature (14) but data are limited and contradictory. Large differences in stability have been found by different groups for virtually identical experiments (2,5). In addition, stability data on the more recently developed analogues are lacking (6,7). The manufacturers recommend that reconstituted solutions of these drugs should be discarded 24 h after preparation as there is no preservative included in the formulation (8). Published data suggest, however, that these drugs may be stable for longer periods.

The purpose of these investigations was to determine the shelf lives of doxorubicin, daunorubicin and epirubicin when reconstituted with Water-for-Injections, dextrose

Correspondence: M. J. Wood, Pharmacy Department, Queen Elizabeth Hospital, Queen Elizabeth Medical Centre, Edgbaston, Birmingham B15 2TH, U.K.

279

280 M. J. Wood, W. J . Irwin and D. K . Scott

(5 yo) and sodium chloride (0.9%) using high-performance liquid chromatographic (HPLC) methods. Shelf lives were investigated in PVC minibags and polypropylene syringes. The studies were conducted at three temperatures [ambient temperature (25"C), 4°C and -2O"CI.

MATERIALS AND METHODS Chemicals and reagents

Doxorubicin, daunorubicin and epirubicin were purchased as the commercially available dosage forms Adriamycin, Pharmorubicin (both Farmitalia) and Cerubicin (Rh6ne-Poulenc). All reagents were of analytical grade or HPLC grade, as appropriate (BDH Ltd, Poole, U.K.). The admixtures were prepared with the following intra- venous fluids: dextrose (5%) injection BP (pH 4.36) and sodium chloride (0.9%) injection BP (pH 5.20 and 6.47) (both Viaflex, Baxter) and Water-for-Injections BP (Antigen).

Equipment The HPLC system consisted of an Altex lOOA pump which was used to deliver eluent

to a 10 cm x 4.6 mm stainless steel Shandon column packed with ODS Hypersil5 pm reversed-phase material. Injections were made with a Rheodyne model 7125 sample injector equipped with a 10 p1 injection loop. A Pye Unicam variable wavelength detec- tor was operated at 290 nm (0.16 a.u.f.s.) and connected to a J J instruments CR452 chart recorder.

Assay conditions The HPLC assays were developed from a method reported by Beijnen et al. (9). The

stability-indicating capacity of these methods was demonstrated by comparison of standard solutions, partially degraded and fully degraded solutions. Decomposition was accelerated by subjecting the solutions to extremes of temperature and pH.

The mobile phases for doxorubicin, daunorubicin and epirubicin contained acetonitri1e:water 40:60,55:45 and 5050 (v/v) respectively. Ten drops of diethylamine were added to each litre of mobile phase, which was then adjusted to pH 2.5 with 10% orthophosphoric acid (v/v). The flow rates were 1.4, 1.5 and 1.3 ml/min, respectively, and the chart speed was 5 mm/min.

Preparation of minibags Vials of each drug were reconstituted with Water-for-Injections and used immedi-

ately to prepare the admixtures for testing. Either 5ml of a 2mg/ml solution of doxorubicin or epirubicin, or 2 ml of a 5 mg/ml solution of daunorubicin, were added to 100 ml polyvinylchloride (PVC) minibags to yield initial theoretical concentrations of 95.2,95.2 and 98 pg/ml, respectively. Admixtures were stored at ambient temperature (25"C), in the refrigerator (4°C) or in the freezer (- 20°C).

The effect of repeated freezing and thawing at ambient temperature was investigated by comparison of the percentage parent drug remaining after two freeze-thaw cycles at 14 and 28 days, which was used as the control, and the percentage remaining in identical solutions which were frozen and thawed at ambient temperature 11 times over a period of 43 days.

Doxorubicin, daunorubicin and epirubicin 28 1

Preparation of syringes

Immediately after reconstitution with Water-for-Injections, the contents of two vials of either doxorubicin or epirubicin (both containing 50 mg in 25 ml) were drawn up into 60-ml polypropylene syringes (Becton Dickinson, Plastipak). For daunorubicin the contents of five vials (each containing 20 mg of drug in 4 ml) were drawn up into the syringes and made up to volume (50 ml) with Water-for-Injections to achieve a final concentration of 2 mg/ml. After all air had been excluded, the syringes were sealed with Luer-Lok hubs and stored at 4°C.

Two-millilitre samples were removed from both syringes and minibags for chromato- graphic analysis at appropriate time intervals over a period of 43 days. All samples were immediately frozen, stored at - 20"C, and thawed at ambient temperature just prior to assay. The colour and pH of each solution was noted on mixing and periodically during the course of the experiment. Experimental runs for each drug fluid admixture were duplicated. Duplicate samples were injected directly using the conditions described. All samples were protected from light throughout the study and during thawing prior to assay.

Treatment of data

Concentrations that remained were determined by interpolation of calculated peak areas on calibration curves which were constructed daily. All concentrations, which were expressed as a percentage of the initial concentration at time zero, were the mean of duplicate values. Significant degradation was defined as a loss of 2 10% of the original concentration.

Data were evaluated for a first-order kinetic model by construction of plots of the natural logarithm of the percentage parent drug remaining versus time. The rate constant, (k) for drug loss was calculated by linear regression analysis of the first-order plot.

Significant differences in the parent drug content of the analysed samples were calculated by means of a one-way, two-way or three-way analysis of variance (with replicates). Where it was necessary only to test the difference between two means, Student's t-test was employed. The pooled standard deviation for the two test samples and the standard error of the difference between the two sample means were calculated according to the usual equations (10).

RESULTS AND DISCUSSION

Stability in polyvinylchloride ( P V C ) minibags Data plotted as a first-order model exhibited a biphasic degradation pattern. For this

reason, rate constants were calculated from the terminal slope of the first-order plots. Figure 1 shows a typical first-order plot for doxorubicin. Doxorubicin has been reported to adsorb onto glass and certain plastics (1 1 , 12). This phenomenon may account for the biphasic degradation pattern observed for this and the other studied drugs. Results from studies which compared the rate of disappearance of doxorubicin (100 pg/ml) from glass containers, in the presence of an excess (8 g) of PVC squares (1 x 1 cm), with drug loss in identical control solutions, showed that drug loss, over the first 24 h, was much more

282 M . J . Wood, W . J . Irwin and D. K . Scott

100

- 9 8 - x

u Y)

._

._ z 0

c al u L X 2 92 -

f F g 6 : 94 \ c

X 90 -

L’ I I I I I 1 I I I I I I

Time ( h ) 0 200 400 600 800 I000 I200 0 200 400 600 800 I000 I200

Time ( h )

Fig. 1. First-order plot for doxorubicin (100pgglml) in PVC minibags in 0.9% sodium chloride, pH 6-47, at 4°C.

Table 1. Rate constants (initial slope) for loss of doxorubicin in sodium chloride (0.9%), pH 5.2,

in PVC minibags at 4°C

Concentration (pglml) Rate constant (h-I)

200 100 50

1.37 x 10-4 3.37 x 10-4 4.72 x 10-4

rapid in the vials which contained the chopped plastic, than in the controls. However, the terminal slopeof both plots was similar. In another study the initial rates of disappearance of doxorubicin in PVC minibags increased with the concentration (Table 1).

Although these data suggest that doxorubicin was adsorbed onto PVC, this phenom- enon is usually complete within a few hours. In the present study, 8 days elapsed before equilibrium was reached. It is possible, therefore, that the initial slope of the first-order plot was due to rapid adsorption of the parent compound onto PVC followed by a slower dissolution and migration of the drugs into the plastic matrix. This sorption

Doxorubicin, daunorubicin and epirubicin 283

Table 2. Percentage doxorubicin, daunorubicin and epirubicin (100 pg/ml) sorbed to 100 ml PVC minibags after 192 h

Percentage loss due to sorption

Dextrose 5% NaC10.9yo NaClO.9% pH 4.36 pH 5.20 pH 6.47

25°C 4°C -20°C 25°C 4°C -20°C 25°C 4°C -20°C

DOX 5.2 3.0 2.3 7.3 6.8 3.2 7.4 7.7 3.9 EPI 4.4 2.1 3.4 4.9 2.8 2.2 8.6 4.3 4.4 DAU 4.7 3.3 2.7 4.1 4.0 3.3 3.8 3.5 3.1

model has been suggested for the interaction of glyceryl trinitrate, warfarin and the benzodiazepines with PVC (13-15).

Table 2 shows the percentage of each drug sorbed to 100ml PVC minibags after 192 h. The apparent partition coefficient (K) of these drugs between PVC and the solution was estimated using the following equation (13);

where I.'= the volume of solution, v, =the density of solution, Wp = the weight of plastic and a = the ratio of the final concentration to the total concentration drop according to the equation

F" 1 - F"

a=-

where F" =the equilibrium fraction of the drug remaining in solution. For solutions stored in sodium chloride (0.9%; pH = 6.47) at 25"C, the log K values

for doxorubicin, epirubicin and daunorubicin were 0.1, 0.2 and - 0.2, respectively. These values indicate that partition of all three drugs into PVC was limited as the partition coefficients were small compared to literature values for diazepam, warfarin and glyceryl trinitrate, which are strongly sorbed to PVC (log Kvalues 1.7,1.9 and 1.6, respectively) (13).

The pH of both the dextrose- and sodium chloride-containing admixtures increased slightly (up to 0.6 units) over the 43-day period of the study but the colour of the solutions was unchanged. When the minibags were emptied after completion of the study, the inside surface showed a dull-pink coating which confirmed that sorption to PVC had occurred. The constituents of this coating are, however, unidentified at present. Chromatograms obtained for solutions stored in these minibags showed no evidence of degradation products, even after prolonged storage, however degradation products were apparent on traces obtained from buffer solutions of similar pH values after relatively short time periods (Fig. 2). Results from previous studies (16) indicate that degradation in aqueous buffers follows first-order kinetics. In addition, the rate of degradation in buffers (pH 4.35,5.33 and 6.37) was much more rapid than in dextrose

284 M . J . Wood, W . 3. Irwin and D. K . Scott

’,1 0.0 I6 A U I

- 6 4 2 0

Time (rnin)

- 6 4 2 0

Time (rnin)

Fig. 2. Chromatogram of doxorubicin (100pg/ml) in (a) half-strength Sorenson buffer pH 4.0 ( t= 168 h), (b) (5%) dextrose, pH 4.36 (f= 1032 h). Doxorubicin=D, degradation product =DI.

Table 3. Rate constants for drug loss and percentage potency of doxorubicin solutions (100 pg/ml) after 43 days storage in 5% dextrose and 0.9% sodium chloride in PVC

minibags

Infusion fluid

Percentage Temperature potency Rate constant

PH (“0 Z f SD (n =4) 0 - l )

Dextrose (5%) 4.36 25 4

- 20 Sodium chloride (0.9%) 5.20 25

4 - 20

Sodium chloride (0.9%) 6.47 25 4

- 20

92.8 f 0 3 96.3k2.2 96.3 k0.5 90.4 k 1.4 92.2 k 1 .O 95.4k0.6 89.1 k0.6 90.6 f 1.7 93.7 k2.8

2.33 x 10. 1.40 x 10-5 1.21 x 10-5 3.23 x 10-5 1.45 x 10-5 1.21 x 10-5 4.69 x 10.-5 2.38 x 10-5 2.64 x 10-5

Doxorubicin, daunorubicin and epirubicin 285 Table 4. Rate constants for drug loss and percentage potency of epirubicin solutions (100 pg/ml) after 43 days storage in 576 dextrose and 0.9% sodium chloride in PVC

minibags

Infusion fluid

Percentage Temperature potency Rate constant

PH ("C) SZk SD (n=4) 0-1)

Dextrose (5y0) 4.36 25 4

- 20 Sodium chloride (0,9°,0) 5.20 25

4 - 20

Sodium chloride (0.9°~0) 6-47 25 4

- 20

93.9k0.8 95.9 f 0.6 96.4 f 1.9 92.3 f 1.8 95.5 f 1.0 96.2k1.4 87.1 k2-0 93.3& 1-0 94.1 +0-5

2.46 x 10-5 1.73 x 10-5 1.33 x 10-5 3.73 x 10-5 2.10 x 10-5 1.28 x 10-5 5.25 x 10-5 3.07 x 10-5 2.18 x 10-5

Table 5. Rate constants for drug loss and percentage potency of daunorubicin solutions (100 pg/ml) after 43 days storage in 57; dextrose and 0.9% sodium chloride in PVC

minibags

Infusion fluid

Percentage Temperature potency Rate constant

PH ("C) 52kSD(n=4) 0-1)

Dextrose (5%) 4.36 25 4

- 20 Sodium chloride (0,9O/,) 5.20 25

4 - 20

Sodium chloride (0.9%) 6.47 25 4

- 20

93.3 k 1.6 95.6k 1.3 96.5 k0.6 93.9 k 1.2 945& 1.3 96.1 f 1.3 94-1 k 1.4 95.6 k 1.1 95.4 & 1-9

1.81 x 10-5 1.76 x 10-5 1.14 x 10-5 3.01 x 10-5 1.91 x 10-5 1.35 x 10-5 2.58 x 10-5 1.47 x 10-5 1.48 x 10-5

Table 6. Rate constants for degradation of doxorubicin, daunorubicin and epirubicin in half strength Britton-Robinson

buffer (pH 4.35,5-33 and 6.37)

Rate constant (h-I)

PH Doxorubicin Daunorubicin Epirubicin

4.35 30.5 x 10-4 3.06 x 10-4 2.47 x 10-4 5.33 5.88 x 10-4 4.15 x 10-4 4.65 x 10-4

- 5.08 x 10-4 6.37 -

-Rate of degradation not calculated due to precipitate formation.

286 M.J. Wood, W . J . Irwin and D. K. Scott

I P

PH

Fig. 3. Logarithm of the rate constant for degradation (k) versus the pH of the buffer for epirubicin.

(5%) and sodium chloride (0.9%) (Tables 3-6). These data suggest that degradation products are likely to form after prolonged storage of these anthracyclines in infusion fluids. The absence of such degradation products on chromatograms suggests that they may have been sorbed to PVC in a similar way to the parent compounds. Therefore, initial results indicate that drug loss in PVC minibags appears to be due to a combination of sorption and degradation.

An evaluation of the total drug loss (sorptive losses and degradation), which was used to assess the overall stability of these drugs in PVC minibags, showed the following results. Doxorubicin and epirubicin became progressively more stable as the pH of the drug fluid admixture became more acidic (pH 6-47-4-36) and were optimally stable in dextrose (5%; pH = 4.36). Significant loss of both analogues (loss in potency 2 occurred in solutions of sodium chloride (0.9%; pH = 6.47) at 25°C (tgO., values 24 and 20 days, respectively) but not at pH 5.20 (Tables 3 and 4). Data from pH profiles (16) indicate that these analogues were expected to be stable optimally between pH 4.0 and 5.0, and as the pH of the admixture increased between pH 5.0 and 7.2, a rapid increase in the rate of drug loss was expected. Figure 3 shows the pH profile obtained for epirubicin. Conversely, daunorubicin was stable in PVC minibags (((10y0 loss in potency) in all the infusion fluids studied, for 43 days, at 25,4 and - 20°C (Table 5) , as predicted by the pH profile obtained for this analogue (Fig. 4) (16).

Doxorubicin, daunorubicin and epirubicin 287

0

-I

I - 'r

-t -2 - -

cn 0 J

- 3

-4 4

1 1 1 1 1 1 1 1 l 1 1 1

2 4 6 8 10 12

PH

Fig. 4. Logarithm of the rate constant for degradation (k) versus the pH of the buffer for daunorubicin.

Table 7. Percentage potency of solutions of doxorubicin, daunorubicin and epirubicin (100 pg/ml) after freeze-thaw treatment

Percentage remaining Percentage remaining after two re-thawings after 1 1 re-thawings

Drug SEkSD (n=3) Z k S D (n=3)

Doxorubicin 99.1 k2.0 96.7k 1.8 Epirubicin 97.5 k 0.7 95.6k 1.4 Daunorubicin 99.3 0.9 97.6& 1.5

In summary, intravenous infusions of doxorubicin, daunorubicin and epirubicin may be stored in PVC minibags for prolonged periods in the frozen state and thawed at room temperature when required for use. In addition, repeated freezing and thawing of these solutions (1 1 freezethaw cycles at ambient temperature) did not result in a significantly greater loss of drug compared to controls solutions which were frozen and thawed twice over a similar period (P>O.O5) (Table 7).

Stability in polypropylene syringes At a concentration of 2 mg/ml, doxorubicin, daunorubicin and epirubicin were stable

(loss in potency of less than 10%) for at least 43 days when reconstituted with Water-

288 M . 3. Wood, W. 3. Irwin and D. K . Scott

Table 8. Percentage potency of solutions of doxorubicin, daunorubicin and epirubicin (2 mg/ml) after storage for 43 days in poly-

propylene syringes at 4°C

Percentage remaining D w Z&SD (n=4)

Doxorubicin 98.7 1.0 Epirubicin 99.7 f 0.4 Daunorubicin 96.9f2.3

for-Injections and stored in polypropylene syringes at 4°C (Table 8). Plots of the natural logarithm of the percentage parent drug remaining versus time were linear ( r > 0.999), which indicated that drug loss followed first-order kinetics.

CONCLUSION The above data indicate that intravenous infusions of doxorubicin, daunorubicin and epirubicin in PVC minibags may be stored for prolonged periods in the frozen state and thawed at room temperature when required for use. Repeated freezing and thawing of solutions stored in minibags at ambient temperature does not appear to cause signifi- cant degradation. All three drugs are also stable, for prolonged periods, in Water-for- Injections when stored in polypropylene syringes at 4°C. Polypropylene also appears to be the best material in which to store these anthracyclines. The data indicate, therefore, that from a stability point of view, the manufacturers recommendations to discard solutions of these drugs (24 h after reconstitution) are very conservative. In conclusion, the shelf-life of solutions of doxorubicin, daunorubicin and epirubicin may be extended, at the discretion of the pharmacist in charge of the reconstitution service, provided that sterility of the product can be ensured at the time of use.

1.

2.

3.

4.

5.

6.

7.

REFERENCES

Benvenuto, J.A., Anderson, R., Kerkof, K., Smith, R.G. & Loo, T.L. (1981) Stability and com- patibility of antitumour agents in glass and plastic containers. AmericanJournal of Hospital Pharmacy, 38, 1914-1918. Poochikian, G.K., Cradock, J.C. & Flora, K.P. (1981) Stability of anthracycline antitumour agents in four infusion fluids. American3ournal of Hospital Pharmacy, 38,483-486. Karlsen, J., Hjort Thonnensen, H., Resberg Olsen, I., Horne Sollien, A. & Skobba, T.J. (1983) Stability of cytotoxic intravenous infusions subjected to freezethaw treatment. Norvegica Pharmaceutica Acta, 45,61-67. Hoffman, D.M., Grossano, D.D., Damin, L.-A. & Woodcock, T.M. (1979) Stability of refrigerated and frozen solutions of doxorubicin hydrochloride. American Journal of Hospital Pharmacy, 36,

Tavoloni, N., Guarino, A.M. & Berk, P.D. (1980) Photolytic degradation of Adriamycin. Communications: Journal of Pharmacy and Pharmacology, 32,860-862. Keusters, L., Stolk, L.M.L., Umans, R. & Van Asten, P. (1986) Stability of solutions of doxorubicin and epirubicin in plastic minibags for intravesical use after storage at -20°C and thawing by microwave radiation. Pharmaceutisch Weekblad. [Scientific Edition], 8,194-197. Beijnen, J.H., Rosing, H., De Vries, P.A. & Underberg, J.M. (1985) Stability of anthracycline antitumour agents in infusion fluids. Journal of Parenteral Science and Technology, 39,220-222.

1536-1538.

Doxorubicin, daunorubicin and epirubicin 289

8. Doxorubicin and Epirubicin Data Sheets, (1989) Farmitalia Car10 Erba Ltd., St. Albans, Herts. Package insertfor daunorubicin. May & Baker Pharmaceuticals, RhBne-Poulenc Ltd, Dagenham, Essex.

9. Beijnen, J.H., Wiese, G. & Underberg, W.J.M. (1985) Aspects of the chemical stability of doxorubicin and seven other anthracyclines in acidic solution. Pharmaceutisch Weekblad [Scientific Edition], 7,

10. Neville, A.M. & Kennedy, J.B. (1964) Basic Staristical Methodsfor Engineers and Scientists. Intertext Books, London.

11. Bosanquet, A.G. (1986) Stability of antineoplastic agents during preparation and storage for in vztro assays. 11. Assay methods, adriamycin and the other antitumour antibiotics. Cancer Chemotherapy and Pharmacology, 17,l-10.

12. Tomlinson, E. & Malpeis, L. (1982) Concomitant adsorption and stability of some anthracycline antibiotics. Journal of Pharmaceutical Sciences, 71,1121-1 125.

13. Illum, L. & Bundgaard, H. (1982) Sorption of drugs by plastic infusion bags. InternationalJournal of Pharmaceutics, 10,339-351.

14. Illum, L., Bundgaard, H. &Davis, S.S. (1983) A constant partition model for examining the sorption of drugs by plastic infusion bags. Internariaal Journal of Pharmaceurics, 17,183-192.

15. Allwood, M.C. (1987) Sorption of parenteral nitrates during administration with a syringe pump and extension set. InternationalJournal of Pharmaceutics, 39,183-188.

16. Wood, M.J. (1988) Stability of anthracycline cytotoxic agents in solution and infusionfluids. MPhil thesis, Aston University, Aston.

109-1 16.