EFFECT OF A NEWLY DEVELOPED BISPHOSPHONATE, YH529, ON OSTEOLYTICBONE METASTASES IN NUDE MICEAkira SASAKI1*, Kazuyuki KITAMURA 2, Rafael E. ALCALDE1, Toshimitsu TANAKA 2, Atushi SUZUKI1,Yohei ETOH1 and Tomohiro MATSUMURA1

1Department of Oral and Maxillofacial Surgery II, Okayama University Dental School, Okayama, Japan2Discovery Research Laboratories, Research and Development Division, Nippon Hoechst Marion Roussel Ltd.,Kawagoe/Saitama, Japan

YH529, [1-hydroxy-2-(imidazo [1,2-a] pyridin-3-yl) ethyl-idene]-bisphosphonic acid monohydrate, is a newly devel-oped third-generation bisphosphonate with a potent inhibi-tory activity toward osteoclastic bone resorption. The primarycellular mechanism of osteolysis associated with metastaticcancer is osteoclast-mediated. It is likely that bisphospho-nates would be efficacious in this situation. In the presentstudy, we examined the effect of YH529 in a nude mice bonemetastasis model, in which the intracardiac injection of ahuman breast cancer cell line, MDA-MB-231(MDA-231), leadsto osteolytic bone metastases. To examine whether YH529would prevent such bone metastasis, we administered YH529s.c. to nude mice simultaneously with cancer cell inoculationthrough the entire experimental period (protocol 1) orperformed short-term prophylactic administration beforeinoculation of the MDA-231 cells (protocol 2). In addition, toexamine the possible therapeutic effects of the drug onestablished bone metastases, we injected YH529 after radio-graphically small but distinct osteolytic bone metastases hadbeen detected (protocol 3). In all protocols, YH529 (2 mg/mouse/day) markedly inhibited bone metastases as well asthe progression of established metastatic foci that werequantified on the radiographs. Histological examination andhistomorphometrical analysis revealed that YH529 markedlyreduced the number of osteoclasts and the size of the tumorat the metastatic bone sites. Our results suggest that YH529may suppress metastasis formation and tumor growth inbone through inhibition of osteoclastic bone resorption. Int. J.Cancer 77:279–285, 1998.r 1998 Wiley-Liss, Inc.

Bone is the most common site of metastasis in human breastcancer. Autopsy studies demonstrated that 47–85% of breast cancerpatients had bone metastases (Averbuch, 1993; Body, 1992). Theoccurrence of osteolytic bone metastases causes serious morbiditydue to intractable bone pain, pathological fractures, hypercalcemiaand nerve compression syndromes, and decreases the quality of lifefor patients with cancer (Averbuch, 1993; Body, 1992; vanHolten-Verzantvoortet al., 1987).

Osteolytic bone metastases display a unique step of osteoclasticbone resorption that is not seen in metastasis to other organs. It iswidely accepted that osteolysis associated with cancer is essentiallymediated by osteoclasts, which appear to be activated, indirectlythrough osteoblasts or directly by tumor products (Galasko, 1976;Mundy, 1991). Therefore, it is expected that inhibition of osteoclas-tic bone resorption may become a useful adjunctive therapy for thetreatment of cancers that have colonized in bones.

Bisphosphonates, stable analogues of pyrophosphonate, arehighly potent inhibitors of osteoclastic bone resorption (Fleisch,1991). These bisphosphonates are emerging as important andwidely used drugs for the treatment of osteolytic lesions such asPaget’s disease of bone, hypercalcemia of malignancy, osteoporo-sis and metastatic bone disease (Averbuch, 1993; Fleisch, 1991).Several analogues of bisphosphonate have been developed andthese appear to operate via different functional mechanisms due tominor structural alterations (van Beeket al., 1994). Therefore,experimental investigation of the biological activity of newlysynthesized bisphosphonates is required before these compoundsmay be used for clinical purposes.

YH529, [1-hydroxy-2-(imidazo [1, 2-a] pyridin-3-yl) ethylidene]-bisphosphonic acid monohydrate, is a newly developed third-generation bisphosphonate that has a more potent inhibitoryactivity toward mouse osteoclastic bone resorptionin vitro and invivo than the previously developed bisphosphonates such asalendronate, pamidronate and risedronate. However, the efficacy ofYH529 in the treatment of osteolytic bone metastasisin vivo isuncertain.

In the present study, we examined the effects of YH529 on anexperimental bone metastasis model in nude mice in whichinoculation of human breast cancer cells into the arterial circulationthrough the left cardiac ventricle leads to osteolytic lesions withcharacteristics of those formed by human osteolytic metastasis.

MATERIAL AND METHODS

Cell culture

MDA-MB-231 (MDA-231), an estrogen-independent humanbreast cancer cell line, was cultured in Dulbecco modified Eagle’smedium (Life Technologies, Inc., Rockville, MD) containing 10%fetal bovine serum (Life Technologies, Inc.) and 1% penicillin–streptomycin solution (Life Technologies, Inc.) at 37°C, 5% CO2.

Intracardiac injection of MDA-231 cells into nude miceAll cultures used for intracardiac injections were harvested at

subconfluence after having been re-fed with fresh medium 24 hrbefore inoculation preparation. Cells (13 l05) were suspended in0.1 ml of phosphate-buffered saline and then injected into the leftheart ventricle of 4- to 5-week-old, female BALB/c–nu/nu mice(CLEA, Tokyo, Japan) with the use of a 27-gauge needle underpentobarbital anesthesia (0.05 mg/g body weight), according to atechnique described previously (Sasakiet al., 1995). The animalswere housed in a pathogen-free environment under controlledconditions of light and humidity in our animal facility for 4 or 5weeks.

Experimental protocolExperimental protocols to determine the effects of YH529

(Fig. 1) on osteolytic bone metastasis were established as follows(Fig. 2). All mice were radiographed and then sacrificed by cervicallocation for histological examination after 4 weeks of tumor cellinoculation.

Protocol 1: effects of continuous treatment with YH529 on thedevelopment of bone metastases.Protocol 1 was performed todetermine the effect of YH529 on the prevention of the develop-ment of new bone metastases. To this effect, we injected MDA-231cells into the left heart ventricle on day 1. From the same day on,YH529 (0.2, 2 and 20 µg/mouse/day) was injected s.c. daily for 4weeks. The untreated group received physiological saline solution.

*Correspondence to: Department of Oral and Maxillofacial Surgery II,Okayama University Dental School, 2–5-1 Shikata-Cho, Okayama 700–8525,Japan. Fax: (81)86235–6704. E-mail: aksasaki@dent.okayama-u.ac.jp

Received 21 October 1997; Revised 23 January 1998

Int. J. Cancer:77,279–285 (1998)

r 1998 Wiley-Liss, Inc.

Publication of the International Union Against CancerPublication de l’Union Internationale Contre le Cancer

Radiographs were taken 4 weeks after cell inoculation and prior tosacrifice, to evaluate osteolytic bone metastasis.

Protocol 2: effects of short-term prophylactic administration ofYH529. To examine the efficacy of YH529 as a prophylactictreatment, YH529 (2 µg/mouse/day) was administered s.c. everyday for 1 week prior to the inoculation of cancer cells.

Protocol 3: effects of short-term treatment with YH529 onestablished bone metastases.We confirmed the established osteo-lytic bone metastasis by radiographs taken after 17 days followinginoculation of MDA-231 cells. Animals with osteolytic metastaseswere divided into 2 groups. One group of mice received normalsaline solution daily, and the other, YH529 (2 µg/mouse/day) fromday 17 to day 28.

Assessment of the number and area of bone metastasesOsteolytic bone metastases were determined on radiographs as

previously described (Sasakiet al., 1995). The mice were deeplyanesthetized with a peritoneal injection of pentobarbital (0.05 mg/gbody weight), laid down in supine, prone and lateral positionsagainst the films (223 27 cm ; Fuji Industrial Film FR: Fuji,Tokyo, Japan), and exposed to soft X-rays at 35 kV for 20 sec withSofron (Tokyo, Japan). The number of osteolytic bone metastasesin whole bones was counted on radiographs. All of the radiographswere extensively and carefully analyzed by 3 different individuals,all of whom were without knowledge of the experiment. The boneradiolucent lesions in the lower limbs were recognized by micros-copy (320) and their area were quantitatively analyzed with aNikon Microphot FXA (Tokyo, Japan) and LUZEX 3U imageanalyzer (Nikon).

Histological and histomorphometrical examinationsLower hindlimbs taken for histological examination were fixed

with 10% neutral phosphate-buffered formalin. These specimenswere decalcified in 10% EDTA solution for 4 days and embedded inTechnovit-7100 (Heraeus Kulzer, Wehrheim, Germany). Histologi-cal sections were made according to standard conventional process-ing protocols and stained with hematoxylin and eosin or tartrate-resistant acid phosphatase (TRAP), a marker enzyme for osteoclasts.TRAP staining was performed by incubating the sections in 0.1 Macetate buffer, pH 5.6, containing 0.75% tartaric acid, 0.01%naphthol AS-BI phosphate and 0.044% pararosanilin, 0.04%sodium nitrite for 30 min at 37°C. Histomorphometrical determina-tion of the number of osteoclasts per millimeter of the tumor/boneinterface was assessed at sites of metastasis in femur specimensthat had been stained with TRAP. The metastatic tumors in boneand in the surrounding soft tissue adjacent to bone were recognizedusing a Nikon microscope (320) and their areas were measuredwith a Hamamatsu image processor (ARGUS-20; HamamatsuPhotonics, Hamamatsu, Japan) and National Institutes of Healthimage processing and analyzing software (NIH, Bethesda, MD).

Statistical analysisBartlett’s method was used to analyze the dispersion of the

numbers and area of bone metastasis. The tendency variations wereanalyzed using the cross-comparison method of SAS program(SAS Institute, Tokyo, Japan). The Williams method (1972) wasused for multiple group comparison, and the Student’st-test wasused when 2 groups were compared. Furthermore, for those inwhich the sample number was extremely small, the comparisonwas also performed with the Student’st-test. When Bartlett’smethod was used to analyze the number of metastasis observed onradiographs, the dispersion was uneven. However, even dispersionwas obtained when logarithmic conversion was done. Therefore,the logarithmic values of number and area of bone metastases wereplotted on the figures.

RESULTS

Osteolytic bone metastasis and cachexia

Nude mice injected with MDA-231 cells into the left cardiacventricle showed an increase in body weight for 3 weeks. Then,they developed a severe cachexia with a marked decrease in muscleand adipose tissue, leading to body weight loss (data not shown).Some mice had limb paralysis due to spinal cord compressioncaused by vertebral bone metastasis. Osteolytic bone metastaseswere commonly enumerated in limbs, vertebral bone, pelvis andscapulae as previously described (Sasakiet al., 1995).

Protocol 1: effects of continuous treatment with YH529 on theprevention of development of new bone metastases

Radiographs taken at day 28 in the untreated group showedmultiple osteolytic lesions at the distal end of femur and proximalend of tibia and fibula, as previously reported (Sasakiet al., 1995).In contrast, nude mice treated with continuous administration of alow dose of YH529 (0.2 µg/mouse/day) demonstrated not only asignificant decrease of osteolytic bone destruction, but also bonecondensation at the same sites (Fig. 3). Radiographically, YH529significantly reduced the number of osteolytic foci at concentra-tions of 2 and 20 µg/mouse/day and the area of osteolytic lesionswere decreased in a dose-dependent manner (0.2, 2 and 20µg/mouse/day; Fig. 4). Histological examination revealed thatmetastatic tumor had filled the bone marrow space at the proximalend of the tibia and numerous TRAP-positive osteoclasts werepresent between the metastatic tumor and trabecular bone, degrad-ing the bone matrix, in untreated nude mice (Fig. 5a, e). In theYH529-treated group (20 µg/mouse/day), however, metastaticbreast cancer cells were located only between the periosteum andcortical bone and trabecular bone volume was markedly increased(Fig. 5b), as previously described in the radiographs. Osteoclasts

FIGURE 1 – Chemical structure of YH529: [1-hydroxy-2-(imidazo [1,2-a] pyridin-3-yl) ethylidene]-bisphosphonic acid monohydrate.

FIGURE 2 – Summary of experimental protocols. MDA-231 cells wereinjected into the left heart ventricle of 4- or 5-week-old female nudemice. YH529 (protocol 1: 0.2, 2 and 20 µg/mouse/day, protocols 2 and3: 2 µg/mouse/day) was given daily s.c. All mice were sacrificed 28days after cancer cell inoculation.

280 SASAKI ET AL.

were not observed between tumor and bone surface (Fig. 5f). Thebone treated with other concentrations of YH529 (0.2 and 2µg/mouse/day) exhibited similar results (data not shown). In thehistomorphometrical analysis of TRAP-positive osteoclasts infemurs, YH529 significantly reduced the osteoclast number irrespec-tive of the concentration used (Table I). Although the number ofTRAP-positive osteoclasts did not decrease according to theYH529 concentration used, the number and area of osteolyticlesions were inhibited in a dose-dependent manner (Fig. 4). Theseresults thus indicate that YH529 may inhibit not only the formationof TRAP-positive osteoclasts, but also their osteoclastic boneresorption depending on the dose used.

Our previous experiments have shown that tumor volume in thesurrounding soft tissue adjacent to bone tended to be larger inrisedronate-treated animals than in untreated animals (Sasakiet al.,1995). Therefore, to consider any possible side effects of YH529,we determined the tumor size in bone and in its surrounding softtissue by means of a histomorphometrical analysis and comparedthe results obtained with those of the untreated and YH529-treatedgroups. Histological analysis demonstrated that YH529 signifi-cantly decreased the metastatic tumors in bone in a dose-dependentmanner, yielding results similar to those obtained from theradiographic analysis. In contrast, tumors in soft tissue of micetreated with YH529 (0.2 and 2 µg/mouse/day) were larger thanthose in untreated mice, whereas high doses of YH529 (20µg/mouse/day) decreased tumor growth in the soft tissue surround-ing bone. YH529 tended to reduce tumors in soft tissue in adose-dependent manner (Fig. 6).

Protocol 2: effects of short-term prophylactic administrationof YH529

Next, we examined whether pretreatment with YH529 wouldexact a prophylactic action against the development of new bonemetastasis. Administration of YH529 (2 µg/mouse/day) for 7 daysprior to the inoculation of MDA-231 cells augmented the bonecondensation at the end of tibia and femur where metastases werefrequently observed on the radiographs (data not shown). Radio-graphic evaluation demonstrated that the number and area of bonemetastatic foci in YH529-treated bone had also decreased (Fig. 7).However, the radiographic impairment of bone destruction withshort-term prophylactic treatment of YH529 (2 µg/mouse/day) wasalmost the same as that attained by the continuous administration oflow-dose YH529 (0.2 µg/mouse/day) (Figs. 4, 7). According tohistological and histomorphometrical analysis, YH529 reduced thenumber of TRAP-positive osteoclasts and degenerated osteoclastswere observed histologically (Fig. 5c, g, Table I).

Protocol 3: effects of short-term treatment with YH529 onestablished bone metastases

Protocol 3 was performed to determine if established bonemetastatic tumor can be inhibited from further progression by theuse of YH529. In the radiographs taken on day 17 before treatmentwith YH529 (2 µg/mouse/day), there were a few small osteolytic

FIGURE 3 – Representative radiographs of osteolytic lesions in a hindlimb of nude mice untreated and treated with YH529 (0.2 µg/mouse/day)simultaneously with MDA-231 cell inoculation throughout the experimental period, according to protocol 1. Radiographs were taken at 4 weeksafter cell inoculation. Osteolytic bone metastases are shown by arrowheads.

FIGURE 4 – Effect of continuous treatment of YH529 on the develop-ment of new osteolytic bone metastases. YH529 (0.2, 2 and 20µg/mouse/day) was given continuously for 28 days according toprotocol 1. The number (left panel) and area (right panel) of osteolyticbone metastases in nude mice were radiographically assessed at day 28.The logarithm of counted number and area (mm2) was plotted. Point,mean; bars, SE. *Significantly different from the untreated group [p ,0.01, Williams (1972) multiple comparison].

281INHIBITION OF OSTEOLYTIC BONE METASTASIS

FIGURE 5 – Histology of bones of untreated nude mice (a ande) and of those treated with YH529 according to protocol 1 (b andf), protocol 2(c andg) and protocol 3 (d andh). Bones histologically examined were all distal femur. Hematoxylin and Eosin staining (a–d). TRAP staining(e–h). (a) Metastatic tumor cells (T) have replaced almost all the primary and secondary trabecular bone in this untreated nude mouse. (b)Metastatic cancer cells (T) are present only between the periosteum and the cortical bone in this nude mouse treated with YH529 (20µg/mouse/day). An increase of trabecular bone (b) is visible. (c) and (d) Although some metastatic cancer cells (T) are present between trabeculae,most of the trabecular bone has not been resorbed. (e) Higher magnification of osteolytic bone resorption in an untreated mouse. Numerousosteoclasts (arrows) are present between the cancer cells (T) and islands of bone (b). (f ) Higher magnification of inhibitory effect of YH529 (0.2µg/mouse/day; protocol 1) on osteoclastic bone resorption. Metastatic cancer cells (T) are located close to trabecular bone (b). There are noosteoclasts between the metastatic cancer cells and the bone. (g) Reduced TRAP activity in osteoclast (arrow) and degenerated osteoclasts(arrowheads) are observed. (h) Decreased TRAP activity and degeneration of osteoclasts are observed (arrowheads). Bars: 400 µm (a–d) and 40µm (e–h).

282 SASAKI ET AL.

lesions (data not shown). The animals were then treated withYH529 for 10 days. On day 28, the same animals were radiographi-cally examined again for changes in the number and size ofosteolytic lesions (data not shown). The osteolytic lesions devel-oped aggressively and the number of metastases increased in theuntreated group. On the other hand, there were few changes in theosteolytic lesions in the YH529-treated group. A quantitativeassessment of the radiographs revealed that YH529 considerablyinhibited the increase of the number and area of establishedosteolytic bone metastases from day 17 to day 28 (Fig. 8). Lightmicroscopic observation showed that there were many osteoclastsfacing the endosteal bone surface along with an aggressiveinfiltration of cancer cells at bone metastatic sites in untreated bone(Fig. 5e), and that there were degeneration and decrease ofosteoclast TRAP activity in the YH529-treated bone ( Fig. 5d, h).Histomorphometric analysis of femur bones from mice subjected tothis protocol also showed that YH529 suppressed the number ofosteoclasts, despite the presence of an established tumor (Table I).

DISCUSSION

The pathophysiological mechanism of bone destruction associ-ated with metastatic tumors is related essentially to osteoclastic

bone resorption and not to direct bone resorption produced bytumor cells (Galasko, 1976; Mundy, 1991). This peculiar processmakes bone metastasis different from the formation and develop-ment of metastatic tumors in other organs. From the specific eventof osteolysis mediated by malignant tumors, we expected thatinhibitors of osteoclastic bone resorption would be a useful strategyfor cancer therapy of bone metastasis, besides conventional antican-cer therapies such as radiotherapy or chemotherapy (Body, 1992).Among bone resorption inhibitors, bisphosphonates are the most

TABLE I – HISTOMORPHOMETRICAL ANALYSIS OF OSTEOCLAST NUMBERIN FEMUR BONES

Dose of YH529 Number ofosteoclasts/mm2

Protocol 1Untreated Vehicle 14.36 2.0 (n 5 5)YH529 0.2 µg/mouse/day 2.76 1.8* (n 5 5)

2 µg/mouse/day 2.36 0.6* (n 5 2)20 µg/mouse/day 2.26 0.5* (n 5 3)

Protocol 2Untreated Vehicle 22.76 1.8 (n 5 5)YH529 2 µg/mouse/day 4.66 3.1* (n 5 12)

Protocol 3Untreated Vehicle 13.86 4.4 (n 5 12)YH529 2 µg/mouse/day 5.06 1.7* (n 5 9)

Values are mean6 SD. *Significantly different from untreated group( p , 0.01, Student’st-test).

FIGURE 6 – Comparison of the effect of continuous treatment ofYH529 on tumor size in femur and surrounding soft tissue. YH529(0.2, 2 and 20 µg/mouse/day) was given continuously for 28 daysaccording to protocol 1. The tumor size in bone (left) and soft tissuesurrounding bone (right) was assessed by histomorphometrical analysison day 28. Columns, mean; bars, SE. *Significantly different from theuntreated group (p , 0.01, Student’st-test).

FIGURE 7 – Effect of short-term prophylactic treatment with YH529 (2µg/mouse/day) on the development of new osteolytic bone metastases.YH529 (2 µg/mouse/day) was given for 7 days before MDA-231 cellinoculation according to protocol 2. The number (left panel) and area(right panel) of osteolytic bone metastases in nude mice was assessedon radiographs on day 28. The logarithm of counted number and area(mm2) are plotted. Point, mean; bars, SE. *Significantly different fromthe untreated group (p , 0.01, Student’st-test).

FIGURE 8 – Effect of YH529 on established osteolytic bone metastases.YH529 (2 µg/mouse/day) was given between day 17 and day 28according to protocol 3. The number (left panel) and area (right panel)of osteolytic bone metastases were evaluated on radiographs on days17 and 28. The logarithm of counted number and area (mm2) wereplotted. Untreated group (●, n 5 15), YH529 treated group (C, n 5 14).Point, mean; bars, SE. *Significantly different from the untreated group( p , 0.01, Student’st-test).

283INHIBITION OF OSTEOLYTIC BONE METASTASIS

effective and widely used drugs for the treatment of osteolyticlesions, because of their effectiveness in suppressing osteoclasticbone resorption (Averbuch, 1993; Body, 1992).

YH529 is a newly developed third-generation bisphosphonate,one that has a more powerful inhibitory activity than that demon-strated by alendronate, pamidronate and risedronate toward osteo-clastic bone resorptionin vitro and hypercalcemia induced byparathyroid hormone-related protein or interleukin-1b in vivo (datanot shown). In the present study, we tested thein vivo efficacy ofYH529 on osteolytic bone metastasis according to 3 protocols thatwe had used previously to investigate the effect of a third-generation bisphosphonate, risedronate, on bone metastasis (Sasakiet al., 1995). These experimental protocols were performed basedon the following clinical assumptions : protocol 1, continuouslong-term administration with YH529 for breast cancer patientswith bone micrometastases that could not be detected clinically;protocol 2, short-period prophylactic treatment with YH529 forbreast cancer patients without bone micrometastases; protocol 3,therapeutic administration of YH529 against the progression ofalready-established metastatic lesions in bone, which is morerelevant to the clinical situation. We used an experimental bonemetastasis model in nude mice, in which injection of breast cancercells into left cardiac ventricle preferentially causes establishmentof metastasis in the bones. This model results in osteolytic lesionssimilar to those observed in patients with bone metastasis. There-fore, it is certainly useful not only for screening potential therapeu-tic agents for osteolytic bone metastasis, but also for investigatingthe functional mechanism of bone metastasis (Sasakiet al., 1995;Yonedaet al., 1997).

In all protocols, YH529 inhibited not only the formation anddevelopment of new bone metastasis, but the progression ofalready-established metastatic tumors. Histological and histomor-phometrical assessments demonstrated that this drug significantlydiminished tumor-mediated bone destruction by reducing osteo-clast formation and decreasing their number as well as their TRAPactivity. Thus, YH529, the molecular mechanism of which remainsunknown, probably prevents bone metastasis by inhibiting osteo-clastic bone resorption instead of having antitumor activity. It hasbeen shown that the action of bisphosphonates may be due to thereduction of ATP-dependent proton accumulation (Caranoet al.,1990), to the disruption of the actin rings in osteoclasts (Murakamiet al., 1995), or to the promotion of osteoclast apoptosis (Hughesetal., 1995). Satoet al.(1991) demonstrated that alendronate binds toresorption surfaces and then is released locally during acidification,thus leading to a rise in its concentration. The elevated concentra-tion of alendronate inhibits bone resorption through destruction ofthe osteoclastic membrane ruffling. It has been reported thattyrosine phosphatase activity, which plays an important role inosteoclast formation and function, is a putative molecular target ofalendronate action (Schmidtet al., 1996). Although the mechanismremains unknown, it has been suggested that the osteoblast mayalso be a target for bisphosphonate (Sahniet al., 1993). Bisphospho-nate may affect cancer-cell attachment to the bone extracellularmatrix (Boissieret al., 1997). However, the precise mechanism ofbisphosphonate action remains uncertain and further studies arerequired.

Breast cancer patients, who have bone-dominant or bone-onlymetastasis, frequently present a prolonged survival, but they areusually afflicted by clinical morbidities such as severe pain,pathological fractures and hypercalcemia. Therefore, the preven-tion of bone metastasis is a very important issue to be solved. Toexamine the possibility of prophylactic treatment for patients withor without bone micrometastases, we investigated the effect ofcontinuous long-term administration or short-term pretreatmentwith YH529 in protocols 1 and 2, respectively. Continuouslong-term administration not only blocked formation of bonemetastases and further development of already established metasta-

ses in a dose-dependent manner, but also increased the trabecularbone mass at the end of long bones.

An increase in osteoclastic bone resorption promotes the releaseof certain growth factors and cytokines from the bone matrix intothe bone microenvironment, thus increasing the incidence of bonemetastasis, according to previous experiments (Sasakiet al., 1994).Also, patients with enhanced bone turnover as in active Paget’sdisease have been reported to preferentially develop the firsthematogenous metastases at the sites of active bone remodelling(Powellet al.1983). These clinical and experimental observationssuggest that the bone microenvironment involved in skeletalmetabolic enhancement could facilitate the formation and develop-ment of bone metastasis. Therefore, in protocol 2, we examinedwhether pretreatment with YH529 would prevent the inoculatedcancer cells from becoming arrested in bone tissue. However, thispretreatment could not completely prevent the initial attachment ofcancer cells circulating in arteries on the bone, despite theinhibition of bone turnover.

The inhibitory activity against the initial development of metas-tases exacted by short-duration pretreatment with the usual dose ofYH529 (2 µg/mouse/day) in protocol 2 was almost the same as thatof the continuous low dose ( 0.2 µg/mouse/day) used in protocol 1,in terms of the number and area of osteolytic bone metastatic foci.Also, the number of TRAP-positive osteoclast after continuouslow-dose of YH529 was almost 2 times lower than that aftershort-duration pretreatment (Table I). Although bisphosphonate hasa high degree of bone affinity, long half-life in the skeleton and thesurface-bound bisphosphonate is biologically active for inhibitingbone resorption, its activity is reduced by a cover of newly formedbone matrix (Krempien and Manegold, 1993). Therefore, if thetherapy-free interval is lengthened, bisphosphonate might becomeinactivated. It is possible to provide cancer patients who have bonemetastasis with a long-term administration schedule using thelower total dose of YH529 in lower individual dosages. Althoughsuch administration schedule may need further trials, our resultssuggest that bisphosphonate should be administrated continuouslyon a long-term basis to cancer patients when attempting to improvetheir quality of life.

Once the patients have developed significant bone destruction,the usage of bisphosphonate might not be as effective as standardtreatments such as chemotherapy or radiotherapy. In a clinicalstudy, the usefulness of bisphosphonates for the survival of patientswith bone metastasis remains controversial, despite bisphospho-nates’ clinical effectiveness to protect morbidities related to bonemetastasis in breast cancer patients. It is difficult to estimate theeffect of bisphosphonates on survival, because most of thesepatients may die of metastasis to other organs and not of bonemetastasis. Therefore, the use of an experimental model is ex-tremely important to clarify this point (Sasakiet al., 1995; Yonedaet al., 1997). In protocol 3, therapeutic treatment with YH529essentially suppressed the increase of bone destruction, suggestingthat therapy might improve the quality of life of breast cancerpatients with bone metastases.

Our results with our bone metastasis model suggest that the newpowerful bisphosphonate, YH529, may be used as an effectivetherapy not only for established osteolytic bone metastases, butalso for the prevention of new osteolytic bone metastases, particu-larly in patients at high risk, such as breast cancer patients.Moreover, low-dose, long-term, prophylactic administration maybe effective to prevent bone metastasis. However, the clinicalutility of this compound depends on its potential side effects.Kostenuiket al. (1993) have demonstrated that a second-generationbisphosphonate, pamidronate, conversely increased tumor burdenin the skeleton of bisphosphonate-treated animals. In the present

284 SASAKI ET AL.

study, high doses of YH529 inhibited tumor growth in soft tissues.Although not statistically significantly, other dosages of YH529 inprotocols 1, 2 and 3 (data not shown) conversely showed a trendtowards an increase of tumors in soft tissue, as previously

described in a trial of risedronate (Sasakiet al., 1995). To minimizeside effects, further studies to determine the adequate dose,therapeutic regimen and the clinical utility of this new drug areneeded.

REFERENCES

AVERBUCH, S.D., New bisphosphonates in the treatment of bone metastases.Cancer (suppl.), 72, 3443–3452 (1993).BODY, J.J., Metastatic bone disease: clinical and therapeutic aspects.Bone,13, S57–S62 (1992).BOISSIER, S., MAGNETTO, S., FRAPPART, L., CUZIN, B., EBETINO, F.H.,DELMAS, P.D. and CLEZARDIN, P., Bisphosphonates inhibit prostate andbreast carcinoma cell adhesion to unmineralized and mineralized boneextracellular matrices.Cancer Res.,57,3890–3894 (1997).CARANO, A., TEITELBAUM, S.L., KONSEK, J.D., SCHLESINGER, P.H. andBLAIR, H.C., Bisphosphonates directly inhibit the bone resorption activityof isolated avian osteoclastsin vitro. J. clin. Invest., 85, 456–461 (1990).FLEISCH, H., Bisphosphonates: pharmacology and use in the treatment oftumor-induced hypercalcemic and metastatic bone disease.Drugs, 42,919–944 (1991).GALASKO, C.S.B., Mechanisms of bone destruction in the development ofskeletal metastases.Nature (Lond.), 263, 507–508 (1976).HUGHES, D.E., WRIGHT, K.R., UY, H.L., SASAKI, A., YONEDA, T., ROODMAN,G.D., MUNDY, G.R. and BOYCE, B.F., Bisphosphonates promote apoptosisin murine osteoclastsin vitro and in vivo. J. Bone Miner. Res., 10,1478–1487 (1995).KOSTENUIK, P.J., ORR, F.W., SUYAMA , K. and SINGH, G., Increased growthrate and tumor burden of spontaneously metastatic Walker 256 cancer cellsin the skeleton of bisphosphonate-treated rats.Cancer Res., 53, 5452–5457(1993).KREMPIEN, B. and MANEGOLD, C., Prophylactic treatment of skeletalmetastases, tumor-induced osteolysis, and hypercalcemia in rats with thebisphosphonate Cl2MBP. Cancer, 72, 91–98 (1993).MUNDY, G.R., Mechanisms of osteolytic bone destruction.Bone, 12, S1–S6(1991).MURAKAMI , H., TAKAHASHI , N., SASAKI, T., UDAGAWA, N., TANAKA , S.,NAKAMURA , I., ZHANG, D., BARBIER, A. and SUDA, T., A possible mechanismof the specific action of bisphosphonates on osteoclasts: tiludronatepreferentially affects polarized osteoclasts having ruffled borders.Bone, 17,137–144 (1995).POWELL, N., Metastatic carcinoma in association with Paget’s disease ofbone.Brit. J. Radiol., 56, 582–585 (1983).

SAHNI, M., GUENTHER, H.L., FLEISCH, H., COLLIN, P. and MARTIN, T.J.,Bisphosphonates act on rat bone resorption through the mediation ofosteoblasts.J. clin. Invest., 91, 2004–2011 (1993).

SASAKI, A., BOYCE, B.F., STORY, B., WRIGHT, K.R., CHAPMAN, M., BOYCE,R., MUNDY, G.R. and YONEDA, T., Bisphosphonate risedronate reducesmetastatic human breast cancer burden in bone in nude mice.Cancer Res.,55, 3551–3557 (1995).

SASAKI, A., WILLIAMS , P.J., MUNDY, G.R. and YONEDA, T., Osteolysis andtumor growth are enhanced in sites of increased bone turnoverin vivo. J.Bone Miner. Res., 9, S294 (1994).

SATO, M., GRASSER, W., ENDO, N., AKINS, R., SIMMONS, H., THOMPSON,D.D., GOLUB, E. and RODAN, G.A., Bisphosphonate action. Alendronatelocalization in rat bone and effects on osteoclast ultrastructure.J. clin.Invest., 88, 2095–2105 (1991).

SCHMIDT, A., RUTLEDGE, S.J., ENDO, N., OPAS, E.E., TANAKA , H., WE-SOLOWSKI, G., LEU, C.T., HUANG, Z., RAMACHANDARAN , C., RODAN, S.B.and RODAN, G.A., Protein–tyrosine phosphatase activity regulates osteo-clast formation and function: inhibition by alendronate.Proc. natl. Acad.Sci. (Wash.), 93, 3068–3073 (1996).

VAN BEEK, E., HOEKSTRA, M., VAN DE RUIT, M., LOWIK, C. and PAPAPOULOS,S., Structural requirements for bisphosphonate actionsin vitro. J. BoneMiner. Res., 9, 1875–1882 (1994).

VAN HOLTEN-VERZANTVOORT, A.T., BIJVOET, O.L.M., CLETON, F.J., HER-MANS, J., KROON, H.M., HARINCK, H.I.J., VERMEY, P., ELTE, J.W.F., NEIJT,J.P., BEEX, L.V.A.M. and BLIJHAM, G., Reduced morbidity from skeletalmetastases in breast cancer patients during long-term bisphosphonate(APD) treatment.Lancet, 31, 983–985 (1987).

WILLIAMS , D.A., The comparison of several dose levels with a zero dosecontrol.Biometrics, 28, 519–531 (1972).

YONEDA, T., SASAKI, A., DUNSTAN, C., WILLIAMS , P.J., BAUSS, F., DECLERCK, Y.A. and MUNDY, G.R., Inhibition of osteolytic bone metastasis ofbreast cancer by combined treatment with the bisphosphonate ibandronateand tissue inhibitor of the matrix metalloproteinase—2.J. clin. Invest., 99,2509–2517 (1997).

285INHIBITION OF OSTEOLYTIC BONE METASTASIS