Journal of Urthopaedic Re search 1 0 4 4 0 4 4 5 Kaven Press, Ltd., New York 0 1992 Orthopaedic Research Society

Slow Release of Anticancer Drugs from Porous Calcium Hydroxyapatite Ceramic A. Uchida, Y . Shinto, N. Araki, and K. Ono Department of Orthopaedic Surgery, Osaka University Medical School, Osaka, Japan

Summary:; We have developed a new delivery system for sustained release of an anticancer drug (cis-platinum) by enclosure into blocks of porous calcium hydroxyapatite ceramic. The slow release of this drug from this system was confirmed in in vitro experiments. When this system was implanted into normal back muscle, or the tibia, sustained release of cis-platinum was observed during a 12-week period after implantation. The diffusion rate of cis-platinum into blood and other organs (liver; kidney, brain) was less than 10% of that at the implanted site. This delivery system placed into experimental tumors of mice also showed a uniform release of anticancer drug for more than 3 months. Inhibition of tumor growth was more marked after local implantation of this system than after intraperitoneal administration of cis-platinum. These results indicate that this new approach to a drug delivery system may well have an important role in cancer chemotherapy. In bone tumors it is attractive because the mechanical strength of calcium hydroxyapatite ceramic permits partial surgical excision and replacement of the bone defect at the same time. Key Words: Drug delivery system-Calcium hydroxyapatite ceramic--Cisplatinum-Chemotherapy-Bone tumors.

Several materials and devices have been used as delivery systems. We considered that porous calcium hydroxyapatite ceramic (CHA), which has an excellent biocompatibility with bony tissue. could be used as a sustainer of anticancer drugs because of its interlinked pore structure (12,13). Ceramics have been used as vehicles for introducing chemical catalysts into a number of industrial processes. Furthermore, this ceramic delivery system might be useful for bone tumors, because the material shows chemical and physical properties similar to those of bone. Therefore, we developed a new anticancer drug delivery system using CHA as a retainer (14), and in lhi5 article, we report the effectiveness and feasibility of this system in vitro and in vivo.

In the treatment of cancer, the optimal effect of chemotherapy will be obtained by exposing the tumor to a high concentration of anticancer drugs for sufficiently long periods to kill all of the tumor cells. To achieve this objective, a variety of chemotherapeutic regimens have been implemented over many years. Chemotherapy has an important role for not only the prevention of metastases but also for local control of primary tumor. Selective administration of an anticancer drug having sustained release into the tumor appears to be a promising way to enhance its therapeutic efficacy for local control of the tumors, to make safer and easier resection possible. Received November 1 , 1990; accepted October I , 1991 . Address correspondence and reprint requests to Dr. A. Uchida at Department of Orthopaedic Surgery, Osaka University Medical School, 1-1-50 Fukushinia, Fukushima-ku, Osaka 553, Japan. This work was presented at the 36th Annual Meeting of Orthopaedic Research Society at New Orleans, Louisiana.

MATERIALS AND METHODS CHA blocks (4-mm cubic block) were sintered at 1,200"C for 2 h. The porosity of these blocks was 440

441

SLOW RELEASE OF ANTICANCER DRUG

approximately 50% and the micropore diameter was from 20 to 80 krn. The interconnecting pore structure was open on the surface of the ceramics (Fig. 1). One-half milligram or 1 mg of cis-diaminedichloroplatinum (CDDP) powder, an anticancer drug, was placed into a central cylindrical cavity in the CHA block. The hole was sealed with tricalcium phosphate glue (Fig. 2). This was called CDDPCHA composite.

In Vitro Experiments The CDDP-CHA composite was placed in plastic dishes containing 2 ml of phosphate-buffered saline. The concentration of CDDP released in pho3phatebuffered saline from the CDDP-CHA composite was measured at 11 different times up to 90 days. Four composites were used for each duration.

sintering ~1200°C temperature

CHA c3

F)

-

po ros ity 30 40O/o pore size 40-1 5 0 p m

L:

vT $.

DP (05 m g or 1 Omg)

$.

FIG. 2. Illustration of a cis-diaminedichloroplatinum-calcium hydroxyapatite (CDDP-CHA) ceramic composite. A small cylindrical hole was made with ultrasound at the center of a ceramic block. Powdered cis-platinum (0.5 or 1 rng) was packed in the hole, which was sealed with a-tricalcium phosphate mixed in polylactic acid (TCP).

Implantation Experiments

FIG. 1. Calcium hydroxyapatite Ceramic block and its surface structure (top). Pores open on the surface of the ceramics were clearly seen in the scanning electron micrograph (bob tom).

For the in vivo experiment, a CDDP-CHA was implanted into the back muscles, or the proximal tibia, of C,/H mice weighing 80-100 g and of Sprague-Dawley rats weighing 500-800 g. All implant procedures were done under sterile conditions. The composites were sterilized by ethylene oxide gas. Implanted animals were anesthetized using halothane and nitrous oxide. Five implants were used in each species for each duration of implantation. Animals were killed by anesthetic overdose, and the portions of the back muscle or the tibia containing each implant were removed. Approximately 1 cm3 of tissue from muscle or bone surrounding the composite was used for measurement of CDDP concentration. Serum, liver, kidney, brain, and other organs were also removed for measurement of CDDP concentration. The distribution Of CDDP in the tibia was determined when the composite was implanted into the proximal tibia.

J Orrhop Kes, Voi. 10, No. 3 , 1992

A . UCHIDA ET AL.

442 Therapy Experiments

RESULTS

C,IH mice bearing subcutaneous Dunn osteosarcoma provided tumor fragments for use in the back region of mice of the same strain. Each CJH mouse, weighing 80-1 00 g, was subcutaneously inoculated with lo6 viable Dunn osteosarcoma cells suspended in a 0.2-1111 Ham F-12 + 10% fetal bovine serum. At 3 weeks after subcutaneous implantation, the solid tumor had developed to reach -3 cm3 size. A CDDP-CHA composite was implanted into solid tumor, and CDDP concentration of the tumors and other organs such as liver, spleen, lung, brain, kidney, and serum were determined at various intervals. CDDP distribution in the tumor was also determined after placing the composite in the central portion of the tumor. For the assessment of antitumor effect of this delivery system, the composite was implanted in the tumor, which grew - 1 cm3 in volume. As a control group, the CHA block without CDDP was implanted into the tumor. Solution of CDDP in saline was injected intraperitoneally at the concentration of 15 mgikg (median lethal dose) as another administration route. The tumor volume of each mouse was checked once daily, and any deaths were recorded. The percentage of change in weight from pretreatment values was calculated on each day, and maximal weight loss was determined for each mouse; this was then averaged for each treatment group.

CDDP Release In Vitro

The release of CDDP from the CDDP-CHA composite (0.5 mg or 1 mg of CDDP) increased rapidly during the first 4-5 days, followed by a decline for -1 week. After that, CDDP release into the solution remained uniform for 3 months, the maximum period for this experiment. More than 90% of 0.5 or 1 mg of CDDP originally contained in the composite was released during a period of 3 months or more (Fig. 3).

CDDP Release In Vivo When CDDP-CHA composites were implanted into the back muscles of C,/H mice, the CDDP release rate was lower than that in vitro, but the duration of CDDP release into the muscle was much longer. CDDP concentrations in liver, kidney, brain, and serum were less than 10% of that in the muscle implanted the composite (Table 1). The CDDP release rate from the composite implanted into rat tibia was slower than when implanted in muscle. The CDDP distribution pattern in the tibia is shown in Fig. 4. Even at the distal end of the tibia 3.5 cm or more away from the site of implantation of the CDDP-CHA composite, one-third of the maximum CDDP concentration in the proximal tibia was detected. However, the diffusion of the drug to other organs was very low.

Analysis of Cis-Platinum Aliquots of solution or blood, 0.2-0.5 ml, were lyophilized until platinum determination. The tissues werc minced with scissors, and a weighed quantity (1-2 g) of the mince was homogenized with deionized water (2 ml). An aliquot of the homogenate was lyophilized in the same manner as was the solution or blood. These samples were soaked in nitric acidlwater, and rinsed three times with deionized water before use. For platinum determination, a flameless atomic absorption spectrometer (Hitachi 2-8000, Tokyo) was used to monitor the platinum line at 265.9 nm, with a slit band width of 0.4 nm (8). All determinations were performed five times. Statistical Analysis The mean values and the SEM for the resulting data were calculated and compared using Student’s t test.

J Orthop Res, Vol. 10, No. 3, 1992

Therapy Experiments When this delivery system was implanted into a solid tumor transplanted subcutaneously, the phar-

80

70

. ?

E 60

50

v

a 40 30 20 10

n 1

3

5

7

12

17

21

25

30

60

OO(davs)

FIG. 3. Cis-platinum (CDDP) concentration in the surrounding solution at different times after insertion of cis-platinumceramic composite [CDDP-calcium hydroxyapatite (CHA)]. A CDDP-CHA block was set in phosphate-buffered saline in a plastic dish and incubated in an atmosphere of 5% CO, in air at 37°C. Samples were obtained every 2 or 5 days. Values were mean 2 SD for five samples.

443

SLOW RELEASE OF ANTICANCER DRUG TABLE 1. Tissue and plasma concentrations of cis-diaminedichloroplatinum (CDDP) after implantation of CDDP-calcium hydroxyupatite slow release material into muscle CDDP 0.5 mg

Days 1 3 5

I 10 14 28 56 120

Muscle (implanted site)

Plasma

Liver

Kidney

Lung

Brain

10.8 2 1.3 39.6 t 4.3 45.1 f 5.3 86.1 -t 5.9 123.0 2 10.9 146.0 9.6 60.1 2 11.2 17.8 i 3.4 3.1 f 2.5

NU 0.5 f 0.2 0.5 t 0.1 0.5 2 0.1 0.3 f 0.2 ND ND ND ND

3.3 2 1.1 27.7 i 4.1 42.1 2 3.6 31.1 1.9 33.7 2 1.8 35.4 2 5.0 14.3 t 3.4 0.2 f 0.1 ND

8.0 2 1.1 17.9 2 3.9 42.1 2 9.1 23.5 2 5.6 32.1 I 3.5 36.4 = 8.0 16.4 = 2.4 0.2 = 0.1 ND

0.8 f 0.4 1.9 f 0.2 1.3 2 2.8 5.1 f 1.8 3.3 1.2 1.3 2 0.6 0.2 f 0.1 ND

ND

ND 0.6 2 0.3 0.5 f 0.2 0.4 f 0.1 ND ND ND ND ND

28.3 f 4.6 36.1 2.3 42.3 f 4.4 79.4 t 5.9 226.0 i 24.0 195.0 ? 14.4 66.5 2 15.1 23.4 i 4.4 5.3 2 4.6

0.3 0.1 0.4 2 0.1 0.6 f 0.2 1.3 -+ 1.0 1.4 -C 0.1 0.9 f 0.1 ND ND ND

24.2 +- 4.8 42.9 2 7.8 46.0 2 10.4 38.5 8.1 38.4 f 10.6 23.6 2 3.3 11.3 2 2.5 0.8 2 0.2 ND

1.9 ? 0.8 15.9 f 3.1 8.6 2 2.4 4.2 f 1.1 3.3 f 1.1 1.8 f 0.8 0.3 0.1 ND ND

2.4 i 1.2 4.3 -+ 1.1 1.4 f 0.8 0.5 2 0.2 ND ND ND ND ND

*

*

CDDP 1.0 mg 1 3 5

I 10

14 28 56 120

Values are means

*

-t

*

11.4 2 3.7 35.7 f 1.4 47.5 2 5.7 40.1 f 1.5 41.2 t 3.3 37.6 -t 8.0 1.5.8 f 3.1 0.9 i 0.1 ND

*

SD for 10 samples (micrograms per gram); the units in plasma are micrograms per milliliter. ND, not detected.

macokinetics of CDDP was similar to that in normal muscle or bone (compare Table 2 and Table 1). Local tissue levels in the tumors were only half those measured in muscle, and liver and kidney levels were less than half in animals in which the implantation was into tumor compared with animals with implantation into muscle. This CDDP delivery system produced a high concentration and a prolonged retention of the drug compared with other delivery systems such as intraperitoneal administration. CDDP distribution in the tumor was determined by placing the composite in the central portion of the tumor. More than 20% of the CDDP concentration adjacent to the composite in the tumor could be detected at the edge of the tumor 3 cm or more away from the composite (Fig. 5 ) . In contrast, CDDP concentrations in principal organs such as liver, kidney, bone marrow, brain, and serum were significantly lower than in the tumor. Tumor growth was markedly inhibited, with an inhibition rate of 90% at 30 days after implantation of the CDDP-CHA composite. In contrast, the inhibition rate after intraperitoneal administration of the drug was lower (Fig. 6 ) . In histological evaluation, more than 50% necrosis of the tumor was found after this composite implantation into the tumor, but no tissue necrosis was found in other organs.

Animal toxicity for the CDDP-CHA delivery system was lower than that for intraperitoneal administration of CDDP, as is clear from the weight-loss data (Table 3). DISCUSSION To obtain a high concentration of anticancer drugs for long enough periods in tumors, various types of drug delivery systems have been consid( w/g)

300 -

-

S

0 ._ +-’ d

2 2000

-

8 a

100-

(u

5

0

-

n 0

O

I

I

1

I

J Orthop Res, Vol. 10, N o . 3, 1992

444

A . UCHIDA ET AL.

TABLE 2. Tissue and plasma concentrations of cis-diaminedichloroplutinum (CDDP) after implantation of CDDP-calcium hydroxyapatite slow-release material into tumor CDDP 0.5 mg Days 1 3 5 7 10 14 28 56 I20

Tumor (implanted site) 4.1 5.9 27.0 46.0 77.9 80.5 19.1 10.5 1.3

f 0.9

f 1.3 2 3.7

5.0 4.8 f 4.2 t 3.7 k 1.0 0.8 f

5

*

Plasma 0.2 f 0.1 0.9 2 0.2 0.9 f 0.3 0.4 t 0.1

ND ND ND ND ND

Liver 1.3 2.4 5.5 3.8 20.1 5.6 2.3

0.2 ? 1.0 t 1.4 f 1.0 f 1.1 ? 0.8 t 0.1 2

Kidney 1.4 5.4 7.6 9.9 28.4 12.1 9.1

2

0.7

t 1.8 f 2.0 -c 0.1 2 4.5 2 1.1

+ 2.0

ND ND

ND

ND

CDDP 1.0 mg 1 3 5 7 10 14 28 56 120

1.6 + 0.5 9.0 + 2.2 37.3 f 3.9 61 0 2 15.6 111.4 2 8.9 17.9 2 10.7 59.6 f 6.5 26.5 + 6.0 2.0 2 1.7

1.3 t 0.1 0.1 2 0.1

ND 0.6 2 0.1 1.2 f 0.2 0.2 0.1

*

ND

1.1 7.2 7.4 8.4 7.4 12.1 3.5

f 0.1 2 0.4 2

2.5

t 1.4 ?

2.4

k 1.1 1.0

+

ND

ND

ND

ND

3.3 f 1.0 11.0 2 3.4 9.5 t 1.7 12.8 ? 2.5 13.6 2 3.9 15.6 f 2.3 0.3 + 0.1

ND ND

Values are means f SD for 10 samples (pglg). the units In plasma are pg/ml. ND. not detected.

ered. Local administration of anticancer drugs is feasible and effective for local control in bone and soft-tissue tumors that occur mostly in the extremities, for reasons of safer administration of the drug, easier assessment of chemotherapeutic ef’fect, and low frequency of systemic side effects. However, single administration of a high concentration of anticancer drugs is impractical and risky because of local side effects such as skin necrosis or infection of the tumor. Sustained release of appropriate drugs may make the frequency of systemic or local side effects lower. The safety of sustained release is supported by the present evidence that no experimental animals had local side effects. It is still unclear why the diffusion rate of CDDP from this delivery system to the muscle and the tumor is different. Vascular differences in tumors and normal tissues are possible to explain. The diffusion rate to other organs was very low, and weight loss in the animal-implanted CDDP-CHA was significantly lower than that in another administration route. These results suggest that this delivery system makes the frequency of systemic side effects lower, and makes application of a higher concentration of the drug in a local site to obtain total kill of the tumor cells possible. Several slow-release drug delivery systems have been developed for parenteral use and include im-

J Orthop Res, Vol. 10, N o . 3, 1992

plantable pumps, biodegradable implants, and controlled-release polymeric systems (1-7,9-11). Because of the simplicity and reproducibility of a ceramic slow-release system, we chose porous ceramics as a retainer of drugs. Another advantage of this method is the possibility of making various controlled pore sizes of apatite ceramics to control the release rate of drugs. A 20-80-pm pore size of the ceramics used in this study showed continuous and uniform release of the drug during a period of more than 3 months. We also suggest that this delivery system is suitable for local control of malignant bone tumors for the reasons of excellent biocompatibility with bone and similar chemical and physical properties. Adequate local control of bone tumors should make subsequent wide excision of the tumor easier and safer. In cases of metastatic tumors, moreover, this type of drug delivery retainer is helpful as a mechanical support of fragile bone caused by metastatic bone disease. Bone ingrowth into pore structure of this ceramic system was observed in preliminary experiments even after implantation into a bone tumor produced experimentally. Because the broad spectrum of the tumors and severe systemic side effects such as nephrotoxicity and neurotoxicity , we introduced cis-platinum into a slow-release delivery system. Our results demonstrated that CDDP impregnated into porous ceramics was more effective than intraperitoneal administration was for local control of tumor. Several anticancer drugs such as doxorubicin hydrochloride (adriamycin), methotrexate, and other antibiotic

I

I 2 (cm) Distance away from the implanted site FIG. 5. Cis-diaminedichloroplatinum (CDDP) distribution pattern in Dunn tumor after a CDDP-calcium hydroxyapatite composite was implanted into the central portion of the tumor of 3 cm or more diameter. After 2 weeks of implantation, the tumor was removed and samples were obtained at a distance of 5 mm. 0

1

445

SLOW RELEASE OF ANTICANCER DRUG

FIG. 6. Inhibition of tumor growth by local administration of cis-diam i nedic hloroplatinum-calcium hydroxyapatite (CDDP-CHA) composite. When the tumor reached the size of 1-cm diameter, the composite implantation (1 mg) into the tumor or intraperitoneal administration of CDDP (1 mg) was treated. Tumor size was measured every 2 days. Growth of the tumor-implanted CDDP-CHA composite was markedly inhibited compared with control or intraperitoneal administration groups. Values are mean 5 SD for three

0

animals.

drugs may be suitable for use by slow release from porous ceramics. Moreover, the potential immunotherapeutic agents introduced by this delivery system may become a new modality for cancer therapy. It is important to know how a drug distributes itself in a tumor after local administration. We have shown that cis-platinum diffuses totally throughout a Dunn osteosarcoma, although drug concentration decrease in inverse proportion to the distance from the composite delivery system. This system may have little effect on circulating tumor cells or cells in metastatic sites. Therefore, for practical use of this drug delivery system, it will be necessary to combine systemic administration of anticancer drugs for preventing metastasis. The major contribution of this study has been to demonstrate that CDDP is effective when presented in a calcium hydroxyapatite ceramic carrier. This may lead to more effective therapy in diseases that are inadequately controlled by current management procedures. TABLE 3. Comparison of animul toxicity by different administrations of cis-diaminedichloroplutinum (CDDP) ~~~

Cis-platinum do5e (mgbody) Control lntraperitoneal injection Slow-release system of CDDP-calcium hydrox yapatite Data represent the mean

Peak weight loss (%)"

Mortality

(%I

0

0.55

* 3.9

0

0.5

24.3

2

4.3

40

2.2 2 1.6 3.3 2 2.7

0 0

0.5 1.0

* SD of 10 animals, but the mean i

SD of 6 animals in intraperitoneal injection group. a Peak weight loss (76) means the rate of the lowest body

weight during the experiment compared with the weight before treatment in each animal.

Acknowledgment: We thank Professor Sydney Nade, Westmead Hospital, New South Wales, Australia for valuable comments and suggestions.

REFERENCES 1. Beacock CJ, Buck AC, Zwinck R, Peeling WB, Rees RWM, Tnrkes A , Walker D, Grifiths K: The treatment of metastatic prostatic cancer with the slow release LH-RH analogue zoladex ICI 118630. Br J UrolS9:432-442, 1987 2. De Flora A, Benatti U, Guida L, Zocchi E: Encapsulation of adriamycin in human erythrocytes. Yroc Natl Acad Sci U S A 83:7029-7033. 1986 3 . Kato T, Nemoto R , Mori H , Kumagai I: Sustained-release properties of microencapsulated mitomycin C with ethylcellulose infused into thc renal artery of dog. Ctrncer 46:14-21,1980 4. Kim S, Kim DJ, Geyer MA, Howell SB: Multivesicular liposomes containing 1-p-D-arabinofuranosykytosine for slowrelease intrathecal therapy. Cancer Rev 47:3935-3937, 1987 5 . Langer RS, Peppas NA: Present and future applications of biomaterials in controlled drug delivery- systems. Biomuierials 2:201-205, 1981 6 . Mattie DR, h i p a i PK: Analysis of the biocompatibility of ALCAP ceramics in rat femurs. J Biomed Muter Res 22:1101-1126, 1988 7. Ormrod DJ, Cawley S , Miller TE: Extended immunosuppression with cyclophosphamide using controlled-release polymeric implants. Int J lmmunopharmacol7:443448, 1985 8. Pera MF Jr, HC Harder: Analysis for platinum in biological material by flameless atomic absorption spectrometry.Clin Chem 23:1245-1247, 1977 9. Raham YE, Patel KR, Cerny EA, Maccoss M: The treatment of intravenously implanted Lewi\ lung carcinoma with two sustained release forms of 1-P-D-arabinofuranosyicyto\ine. Eur J Cancer CIin Ontol 20:1105-1112, 1984 10. Schroder U: Crystallized carbohydrate spheres as slow release matrix biologically active substances. Biornaterials 5: 10CL104, 1984 11. Sugitachi A, Takasuka Y, Kido T, Sat0 E: A sustained release anticancer device. Trans A m Soc Artif Intern Organs 33:62&630, 1987 12. Uchida A, Nade S , McCartney E, Ching W: Bone ingrowth into three different porous ceramics implanted into the tibia of rats and rabbits. J Orthop Res 3:65-77, 1985 13. Uchida A, Nade S, McCartney E, Ching W: The use of ceramics for bone replacement. A comparative study of three different porous ceramics. J Bone Joint Surg [Br] 66:269-275, 1984 14. Uchida A, Shinto Y, Araki N , Ono K: Development of a slow release system of anticancer drug retained in calcium hydroxyapatite ceramic. Jpn J Cancer Chemother 16:32313235, 1989

J Orthop Res, Val. 10, No. 3 , 1992