Recurrent bone regeneration in titanium implants. Experimental model for determining the healing capacity of bone using quantitative microradiography

Recurrentbone regenerationin titaniumimplants.Experimentalmodel for determiningthe healingcapacityof bone using quantitative microradiography P. K&leboand M. Jacobsson* Laboratory of Experimental Biology, Department ofAnaromy and the *E.N. % Clinic of Sahlgren Hospital, University of Gothenburg, 80x 3303 1. S-400 33 Gothenburg, Sweden (Received 27 April 198 7; accepted 15 August 1987)

An experimental technique for quantitative assessment of bone repair was tested. The recurrent regeneration capacity of cortical bone was analysed in consecutive 3 wk periods, using an osseointegrated titanium implant, the Bone Harvest Chamber (BHC), in the proximal tibia1 metaphysis of the rabbit. The BHC is a divisible implant penetrated by a canal into which newly formed bone tissue will grow during a 3 wk healing period. The newly formed bone tissue may easily be collected (harvested) without the animal being killed. After 3 wk. bone tissue can again be harvested, in principle, indefinitely. Intact harvested specimens were quantified by microradiography-videodensitometry, yielding a total specimen mass in mg aluminium equivalent. This unit correlated very well to a specimen dry weight (r = 0.996) and an average mineralized bone density (r = 0.937). The specimens were also examined by correlative histology. Three weeks after implant insertion, the chambers had become integrated in the bone tissue but the average bone mass varied widely, influenced by the surgical insertion trauma. Six weeks after insertion, the greatest average bone mass was found, indicating an intense ongoing osseointegration. The amount of bone regenerated at later havests was fairly equal, indicating a stabilization of the implant bed to the repeat stimulus, i.e. harvesting. Bone regeneration differed significantly between animals, but also intraindividual variations, i.e. different amounts of bone formed in the same chamber, were observed. Keywords: Rabbits, bone regeneration, osseointegration, titanium implants, quantitative microradiography, videodensitomatry

Quantitative assessment of bone regeneration is important in order to get objective information when investigating different factors that may influence the bone-healing response’. However, it would be desirable if a quantitative method could be combined with a complementary histological analysis*. One methodological problem in dealing with quantitative analysis of bone regeneration in situ is to identify and separate newly formed bone from previously existing tissue. This problem was overcome by the introduction of a refined titanium implant, the Bone Harvest Chamber (BHC)3. Following insertion and direct bone-to-implant anchorage of the implant into the cortex of the proximal tibia1 metaphysis of a rabbit, bone formation will occur in situ into a canal Correspondence to Dr Peter Kelebo. 0 1988

penetrating the implant. Ingrown tissue may easily be collected, i.e. harvested, and the chamber re-assembled enabling repeated harvesting of newly formed bone from the same canal, without killing the animal. Thus, an experimental model providing recurrent bone regeneration was presented, and later used in studies on bone formation following application of different factors, such as irradiation, ischaemia and the presence of osteoprogenitor cells4-6. Osseointegration is defined by living bone tissue in direct contact with the implant surface7-‘. Osseointegration will occur if a biocompatible implant material, such as commercially pure (cp) titanium and an atraumatic surgical technique are usedg-“. However, no strictly quantitative report exists on the regenerative response to a minor trauma in cortical bone adjacent to osseointegrated titanium implants. Using a recently developed nondestructive, quantitative

Butterworth & Co (Publishers) Ltd. 0142-9612/88/040295-07$03.00 Biomaterials

1988, Vol9 July

295

Bone regenerarion in Ti implants: P. KSlebo and M. Jacobsson

microradiographic-videodensitometric technique, accurate bone-mass determinations may be performed’2,13. Thus, the bone mass regenerated in a controlled environment may be assessed. The aim of this study was to evaluate bone-mass determination of tissue harvested in an implant. Furthermore, quantification of the bone regenerative capacity in the same tissue compartment was performed with this technique, in order to evaluate the healing response at different time intervals after implant insertion, and to assess inter- and intra-individual variations in bone regeneration. Any higher species may be used to study bone repair with the BHC, as there is no principle difference in the bone-healing process. In fact, a similar appearance of bone tissue regenerated in titanium implants was found in human beings, dogs and rabbits44. However, the fact that bone regeneration occurs rapidly in the rabbit contributed to the choice of this animal to allow for recurrent bone regeneration analysis.

MATERIALS

AND METHODS

Six adult lop-eared rabbits 1 O-l 2 months of age of both sexes were used. They had all closed epiphyseal lines as evidenced by radiography. All animals received the same conventional laboratory diet and were allowed to move freely in the cages during the experiment.

The bone harvest chamber The BHC is made of commercially pure (cp) titanium and consists of a 10 mm high X 6 mm wide cylindrical screw ( Figure 7). In the centre of the cylinder, a removable centrepiece is fitted and attached by two screws. When the chamber is assembled, a transverse canal of 1 mm diameter X 5 mm length penetrates the implant. Insertion of an implant into

the cortex of the proximal tibia1 metaphysis of a rabbit results in osseointegration and invasion of newly formed-bone into the canal during a 3 wk healing period. The ingrown bone tissue is easily harvested by removal of the centre-piece, and the chamber is re-assembled to allow further bone ingrowth. The BHC may then be harvested at regular intervals,

Surgery and harvest procedure General anaesthesia was induced and maintained byfluanison and fentanyl, i.m., 0.7 ml/kg and hour (Hypnorm, Janssen) and diazepam, i.p., 0.5 ml/kg (Stesolid, Dumex). Under aseptic conditions, the skin and fascia were incised on the medial aspect of the proximal tibia1 metaphysis. A 6 mm circular periosteal incision was made. With a sharpened trephine of 5.5 mm diameter, a cortical hole was drilled -1 (< 2000 rev mtn ) under profuse saline solution cooling. Autologous bone marrow, 0.1 ml, was aspirated from the created defect, and injected into the canal of the assembled implant. The chamber was put into the bone defect and twisted 180”. which ensured a simple insertion, with the canal located at the cortical level and in a direction along the axis of the tibia. A radiograph of an anchored implant is seen in Figure 1. One implant was inserted per animal, thus leaving the other contralateral limb of each animal as a non-operated control. The animals were allowed full weight-bearing during the experiment. Each chamberwas harvested at consecutive 3 wk intervals, without intermission. With the animal under general anaesthesia and under aseptic conditions, the soft tissue over the chamber was incised and the centre-piece lifted out. Ingrown bone tissue was cut out with a sharpened scalpel and then immediately put into neutral buffered formalin solution. A visual postharvest examination using the naked eye provided a rough control of adequate harvesting. The centre-piece was then put back and the canal became filled with local haemorrhage from the cut. Following the first harvest, which served as a control of satisfactory implant incorporation, six further harvests were performed in each animal at 3 wk intervals. Thus, every time at specimen harvesting, a relatively standardized, non-traumatic stimulus for new bone formation was provoked by the scalpel cut.

Microradiography

Figure 1 Radiograph of the Bone Harvest Chamber integrated in the medial aspect of the proximal tibia1 meraphysis of rhe rabbit The canal penerraring rhe implant is located ar the lowest threaded patt. be. in the cortex.

296

Biomarerials

1988, Vol9 July

All specimens from each animal were put on filtre paper for 5 min to remove excess formalin. This procedure avoided damage to the photographic emulsion, on which the samples were directly put. Kodak High Resolution Plates, type 1A, with high contrast and high resolution properties were used’4V’5. The specimens were put in two rows between two parallel aluminium reference step-wedges. Each wedge consisted of 10 foils of 0.107 mm thick pure aluminium, adapted to a lo-step wedge of 25 mm length X 5 mm width. Four bone simulating aluminium phantoms of similar size and known weight were put at the ends of each specimen row. The wedges and phantoms enabled calibration of the simultaneously exposed samples from each animal. A Machlett OEG-50 X-ray tube, with a Cu-target and a focal spot of 1 mm, generated a prefiltred (0.5 mm aluminium) polyenergetic spectrum with the following exposure data: 27 kV. 14 mA and 20 min exposure. The film-focus distance measured 242 mm and theX-ray field of interest 25 X 15 mm.

Bone regeneration in Ti implants: P. Ktilebo and M. Jacobsson

The plates were processed during standardized agitation in fresh Kodak D-l 9 developer for 5.5 min at 20°C. Rinsing, fixing, washing and drying were performed as recommended by the manufacturer’5. Microradiography was performed twice in order to exclude uneven exposed-processed plates. A detailed description of the microradiographic procedure is given by Strid and Kalebo”.

Videodensitometry The images were measured in an interactive image analysis system, IBAS I and II (Kontron Bildanalyse GmbH). The total bone content is determined by means of measurement of the light transmittance through the images. A stable, sensitive television camera, with a Newvicon tube of good linearity was used to translate the transilluminated images into a video signal. The signal was analog-to-digital converted into an array of 512 X 5 12 pixels, using a 256-level grey scale. A stored program in the analyser provided measurements of and adjustments to background density, step-wedge calibration with transformation of optical transmission into equivalent mass of aluminium, area calibration, density profile measurements and image assessments. Each image was scanned once and the bone mass obtained was expressed in mg aluminium equivalent. Aluminium is a valid and a good reference to wet, dense cortical bone of heterogenous composition, in measurements based on soft X-ray attenuation13. A total quantitative error within 2% is achieved with the present combined microradiographic-videodensitometric method, further described by Strid and Kalebo” and Kalebo and Strid13.

Validation

of assessment

In a previous study’ 3 the combined radiographic-densitometric method was validated in the case of cortical bone specimens. In order to test the possible relationships between quantitative measures of BHC specimens, the samples obtained from animal No. 1 were exempted from thorough histology. After radiography of the wet specimens, drying was performed to constant weight (90°C, 3 h), the samples weighed on an accurate analytical balance (M-3, Mettler), after cooling in a vacuum desiccator. The videodensitometric mass (mg-Al) was correlated to the dry weight (mg-Dw) of the six specimens from this animal. The correlation was assessed by linear regression analysis. To determine whether the videodensitometrically assessed bone mass is affected by the degree of mineralization, of which a relative measure is given bythe ratio of aluminium weight to dry weight, this ratio was correlated to the videodensitometric mass. Furthermore, a special program in the image analyser enabled calculation of the specimen maximal density (mg-Al/mm2), also reflecting the proportion of mineral.

Histology After microradiography, the specimens were decalcified in 15% formic acid and embedded in paraffin. The specimens were stained with Htx-eosin and cut into 3-6pm thick transverse and longitudinal sections.

Statistics The Wilcoxon signed-rank test was used to compare the average bone mass formed in the harvest groups. Testing of trend (i.e. bone regeneration with time) and inter-individual

Figure 2 Microradiographical appearance of the end of a harvested BHC specimen, adequately harvested without loss of bone

bone forming capacity (Kruskal-Wallis as described by Lehmann”.

test) was performed

RESULTS Animal No. 1 died during anaesthesia at the sixth harvest. All animals remained healthy during the experiment and the chambers became adequately incorporated, and the canals invaded by newly formed bone tissue in all harvests. All specimens were evenly harvested without fracturing or loss of tissue (Figure 2).

Specimen

mass evaluation

The comparison of videodensitometric mass (mg-Al) and specimen dry weight (mg-Dw) of samples from animal No. 1, yielded a very high agreement, r = 0.996 (Table 1). When the ratio mg-Al/mg-Dw was analysed, i.e. the uniformity of inorganic to organic composition, small variations were noticed. The mean ratio amounted to 1.074 with a coefficient of variation (CV) of 3.99% (Table 7). A low bone mass, 1.588/l .551 (first harvest) resulted in a low ratio, 1.024, while a greater bone mass 3.090/2.7 15 (second harvest) ied to an increased ratio, 1 .138. This indicates that the degree of mineralization increases slightly with increasing bone mass (mg-Al). In consequence, the degree of mineralization correlated well to bone mass (mg-Al), with a correlation coefficient of r = 0.937. Nevertheless, the variation in the ratio was so slight that for practical purposes it might be considered constant at 1.07 (Table 7). An

Table 1 Bone mass determination of six consecutively harvested BHC specimens at regular 3 wk intervals from the same animal. Quantification is performed by microradiography-videodensitometry (mg-AI) with a correlative dry weight (mg-LIw) assessment, enabling comparison of the two methods by means of a ratio (mg-Al/mg-Dw). The image maximal density in mg-Al/mm2 is also presented Harvest no.

Bone mass

Bone mass

(mg-AU

(mg-Dw)

Ratlo (mg-Al/mg-Dw)

Maximal density (mg-Al/mm*)

1 2 3 4 5 6

1.588 3.090 2.132 2.046 2.364 2.665

1.551 2.715 2.025 1.969 2.145 2.455

1.024 1.138 1.053 1.039 1.102 1.086

0.587 0.748 0.711 0.703 0.718 0.748

Mean = 1.074 SD = 0.0429

Biomaterials

1988. Vol 9 July

297

Bone regeneration in Ti implants: P. Ktilebo and M. Jacobsson

except the third harvest. Following the first harvest, a significantly negative trend p < 0.05) in bone regenerative capacity occurred with time (Tab/e 2). The seventh harvest (2 1 wk after insertion) averaged 1.743 + 0.405 mg, but consisted only of five samples (animal No. 1 having died, see above) and is not further statistically analysed. However, no significant difference was found in average bone mass between harvests following the second, indicating a stabilization in bone regeneration (Figure 4). The inter-individual bone-mass distribution per harvest (Nos. 2-7) of the six rabbits resulted in CVs in the range of 18.8-3 1.5%, with a median average of 2 1.4% (Table 2). It was found that bone regeneration differed significantly (p < 0.01) between the animals, when the total amount of bone (i.e. harvests 2-7) from each animal was compared. The intra-individual bone regeneration, i.e. the average bone mass per animal (haNeStS 2-7). is presented in Table 2. A similar distribution was found with CVs in the range of 13.5-33.59/o, and a median of 17.8%; however, these results were also influenced by the time factor. The CV, when a fairly stable bone regeneration occurred (h6tWStS 3-7), was 2.3-37.79/o with a median of 1 1.8% Thus a clear variation of bone regeneration is evident

Microradiograph of 8 typical 3 wk sample, with the corresponding Figure 3 density profile. Note denser bone at the both ends of the specimen, and a trabeculated appearance in the remaining parts.

augmentation in sample image maximal density (mg-Al/ mm2) correlated positively with an increase in bone mass (mg-Al), r = 0.863, and hence in mineral content. A microradiograph with a corresponding density profile of an analysed sample is presented in Figure 3.

Quantitative

evaluation

3.0

2.5 2.0

I.5

The amount of bone formed in six BHC implants, one in each rabbit, in seven consecutive harvests is presented in Tab/e 2. The first harvest, i.e. 3 wk after insertion is not directly comparable to the other harvests, considering average bone mass, mainly because of the 2 X 0.5 mm longer way to grow (osseous invasion through both entrances to the chamber, each measuring 0.5 mm) thereby resulting in a lower average bone mass. However, the calculated bone mass of the first harvest amounted 0.939 + 0.686 mg (mean + SD) with a considerable scattering clearly differing from the following harvests. The greatest quantity of bone was found at the second harvest (6 wk after insertion) with a mass of 2.578 + 0.484 mg. This amount was significantly greater than (p < 0.05) the average mass assessed at the other harvests,

1.0

45 Tlme (weeks)

0

l

0

3

6

9

12

15

21

18

Figure 4 Graph of the bone regeneration response to time analysed in consecutive 3 wkperiods in six animals (numbered). After an initial low bone mass, bone regeneration increases apparently up ro the sixth week after implant insertion. Thereafter, bone regeneration seems to be fairly equal in each animal.

Table 2 Bone mass determination by microradiography and computer-assisted videodensitometry of BHC specimens harvested at consecutive 3 wk periods from six rabbits. Inter- and intra-individual variations in bone regeneration are presented (mean + SD), below and to the right in the table. respectively Harvest 1

2

3

4

5

6

7

Harvests 2-7

1 2 3 4 5 6

1.588 0.394 0.278 1.490 0.274 1.81 1

3.090 2.502 2.057 2.836 1.966 3.014

2.132 1.811 1.415 2.017 2.719 2.073

2.046 2.317 1.358 2.456 1.904 2.560

2.364 2.055 1.333 1.922 1.410 2.071

2.665 1.886 1.375 2.337 1.064 1.939

1.859 1.350 1.729 1.415 2.361

2.459 2.072 1.481 2.2 16 1.746 2.336

Mean SD

0.939 0.686

2.578 0.484

2.028 0.427

2.107 0.442

1.859 0.405

1.878 0.592

1.743 0.405

Animal

298

Biomaterieis

1988, Vol9 July

f f f f + +

0.426 0.280 0.283 0.405 0.584 0.402

Bone regeneration in Ti implants: P. Kalebo and M. Jacobsson

b

Histologic appearance of 3 wk BHC specimens: (a) longitudinal Figure 5 section; denser bone of lamellar character is present at the end of the specimen and thin trabeculae is seen in the middle: (b) transverse section through the end, with mostly lamellar bone. No fibrous tissue is surrounding the sample.

between animals, but also in the same chamber of each animal. However, the average amount of bone formed between each harvest in each of the six BHCs following the second harvest was similar in each animal.

Histologic

examination

The 3 wk samples consisted overall of trabeculated bone, with lining osteoblasts and interspersed connective tissue proper, indicating a not-completed bone growth or maturation. At both ends of the samples, a condensation of bone of a lamellar character was observed. The middle of the specimens consisted of part woven bone and part osteoid. Histologic sections are presented in Figure 5.

DISCUSSION Commercially pure titanium is a bone-tissue compatible material, with a high corrosion resistance. This is probably one factor contributing to its biocompatibility”, “. Titanium dental implants have been successfully integrated in the

edentulous jaw, with follow-up periods of 20 y’*. The BHC implant has advantages compared to some of the other titanium implants utilized in the experimental analysis of bone regeneration’g-25 . Owing to the construction, recurrent bone regeneration is made possible into a defined compartment of a known dimension, and the tissue easily harvested without killing the animal. It has been shown that the collected specimen will contain newly formed bone if careful harvesting is performed at regular intervals3. Furthermore, bone-healing analysis may be done without the potentially hazardous influence of a surgical implantation trauma, with its risks of thermal injury, emphasized by, among others, Linder and Lundskog” and Eriksson”. Rhinelande?, However, in spite of low-speed drilling and profuse saline cooling, a traumatic influencecannot be ruled out, because in the present study the first harvest, i.e. 3 wk after implant insertion, showed considerable scattering in bone mass between the animals, compared to the following harvests. Although osseointegration of titanium implants has been demonstrated 3 wk after implant insertion in the rabbitzoS2’, the first harvest was not further analysed, mainly because of the possible influence of the surgical trauma, as well as the longer distance of growth into the chamber. Nevertheless, as seen in Table 2, bone growth had occurred in the BHC canals at the first harvest, indicating a start to osseointegration. Hence, the possible negative influence of the implant insertion trauma was minimized at harvests 2-7. Only a standardized scalpel cut at the ends of the ingrown tissue, besides a soft tissue opening, was performed at the harvesting. These factors probably favour a relatively stable rate of bone regeneration. Microradiographic-videodensitometric analysis of cortical bone allows valid bone mass assessments, with a low total error (< 2%) and a very high correlation to dry weight, ash weight and calcium content”. 13. This technique was utilized in the present study to assess the mass of bone specimens after 3 wk growth. These consist of not fully mineralized bone, largely of trabeculae of lamellar structure, with lower density and ash content compared to intact cortical bone, which has relatively high values in these respects28-32. The videodensitometric bone mass to dry weight ratio of the present samples amounted to about 1.07, while samples emanating from the densest part of the midtibia of the rabbit show a ratio of 1 .39-l .40 (recalculated from Kalebo and Strid13). This means that 3 wk samples contain less mineral compared to intact bone, and a slight variation in the inorganic to organic composition is found, the mineral content increasing with bone mass. However, good correlations were found between bone mass, expressed in equivalent mass of aluminium, and specimen dry weight, with a correlation coefficient of r = 0.996, and bone mass to relative degree of mineralization, r = 0.937. An ash weight correlation would perhaps have yielded an even better linear regression value, but necessitating total specimen destruction. These findings are in agreement with those of Aro et a/.33 who, studying 4 wk old fracture callus in the rat tibia, found a significant correlation between organic and inorganic callus matrix and dry weight. Thus, the X-ray attenuation of pure aluminium and 3 wk BHC bone specimens of a similar thickness are of the same order of magnitude, indicating aluminium as a valid reference to the heterogenously composed bone tissue, the total content of which is quantified. Rapid mineralization occurs into the canal of the anchored titanium implant. Bone tissue is evident within 2 wk in a similar titanium implant inserted into the same

Eiomaterials

1988. Vol9 July

299

Bone regeneration if? Ti imp&MS: f. @!ebo and M. Jacobsson

rabbit site, and the calcium content increases with time in a linear-like relationship up to 20 d34. The amount of bone regenerated in the present implants was fairly equal at the different harvests. However, a significantly better bone regeneration (p < 0.05) was observed at the second harvest, i.e. from the third to sixth week, compared to the other harvests (except harvest 3), and this harvest showed the greatest average maas of 2.578 + 0.484 mg. A possible explanation for this intense growth phase is a continuous implanl integration process occurring at this time with a general reparatory stimulus in the implant bed. This results in an increased mineralization, including enhanced bone formation into the implant. Lundskog” studied titanium implant incorporation in rabbit bone tissue using microradiography. It was found that 8 wk after implant insertion, a further reparative process had ensued, compared to 3 wk after insertion. This reparative process included cortical thickening and capsular arrangement of bone adjacent to the implant. A similar appearance was noticed at death in the present study. Following the second harvest, bone regeneration did not significantly differ between harvests. A reasonable explanation for the one-forming stabilization seen in this study is a combination of two factors. First, a completed osseointegration has occurred in terms of the amount of bone or mineral being deposited, although not in interfacial strength, a biomechanical parameter that also requires bone maturity, taking a longer time to reach35. Second, the implant bed adapts gradually to the same recurrent bone-regenerative stimuli, i.e. the consecutive harvesting. In fact, an indication of stabilization in bone formation was observed in a previous study3, where bone tissue was repetitively harvested for 1 y, A stabilization in the bone appearance was reported by Albrektsson36, who using vital microscopy, studied dynamic bone ingrowth into an optical titanium chamber, inserted in the rabbit tibia. Mature bone tissue was found 6-l 0 wk after insertion, and the appearance 1-2 months later resembled the same picture. The present regenerative response to time is in agreement with the chemical analysis repotted by Barth et CSI.~~, who quantified bone ingrowth into implants of porous fibre titanium-alloy, inserted into the tibia of the rat. A longer follow-up in the present study would perhaps have confirmed the stabilization discussed. A significant difference (P < 0.01) exists in bone regenerative capacity between animals. The bone mass from six rabbits, assessed at six consecutive harvests, i.e. interindividual bone regenerative capacity, varied with a CV about 2 1%. However, a clear variation was found in intra-individual bone formation capacity, i.e. different bone mass obtained from harvest to harvest in the same chamber, with a CV around 18%. One factor influencing the intra-individual CVis time, mainly the first weeks, discussed above. Hence, the CV was about 12% following the second harvest, when fairly stable bone regeneration occurred. Two major factors influence the bone mass determined in this study: methodological and biological variations. Methodological factors include too-traumatic insertion techniques and different chamber positioning in the implant bed, with subsequent differences in bone formation23’36’38. Uneven harvesting, inadvertent differences in sample handling and fixation, and small errors in the microradiogaphicvide~ensitometric procedures are other factors that cannot be ruled out. Biological variations seem to constitute a considerable factor in the present study. These variations consist of numerous factors that may influence bone regeneration of a

300

Biomaterials

t988,

Vol9

July

complex nature that is not fully understood. Reviews on this subject are given by Sevitt3’, Boskeyet z+/.~‘,Cruess4’ and Simmons4’ among others. Van Limborgh43 grouped factors controlling skeletal morphogenesis into intrinsic genetic factors, local and general epigenetic factors and local and general environmental factors. Influence on bone formation in the BHC implant is perhaps exerted by all these groups, to various degrees. This subject falls beyond the scope of this paper, and is therefore not further discussed.

CONCLUSIONS The present technique was developed as an instrument for quantifying bone regeneration where local as well as systemic influencing factors may be analysed. Both retarding and promoting factors on bone regeneration may be studied, and we believe that valuable information may be obtained with this methodology. The technique was found an adequate experimental model, yielding bone-mass values which correlate well with the weight of dried bone samples. Three weeks after insertion, the implant becomes incorporated in the cortical bone. The amount of bone formed in the canal of the implant during the first 3 wk appears to be influenced by the inse~ion surgical trauma. The osseointegrative process is, however, not completed at this time. The process continues at least up to 6 wk after implant insertion, as shown by the high bone mass obtained at this time. Later harvests yielded fairly equal amounts of bone, indicating a steady state had been reached in bone regenerative capacity into the osseointegrated titanium implant.

ACKNOWLEDGEMENTS The authors express their gratitude to Lotta Hallberg, Department of Mathematics, Chalmers Universi~ of Technol~y, for statistical calculations. This study was supported by grants from the Gothenburg Medical Society, the Greta and Einar Asker Foundation, the Trygg-Hansa Research Fund and the University of Gothenburg, Sweden.

REFERENCES Brand, R.A., Fracture healing, in The Scientific Basis of Orthopaedics, (Eds J.A. Albright and R.A. Brand), Appleton Century Crofts, New York, USA, 1979, 289-311 Albright, J.A. and Skinner, H.C.W., Bone: remodeling dynamics, in The Scientific Basis a# Ofthopaedics, (Eds J.A. Albright and R.A. Brand), Appleton Century Crofts, New York, USA, 1979, 185-229 Albrektsson,T., Jacobsson, M. and Kalebo, P.. The Harvest Chamber a newly developed implant for analysis of bone remodeiiing in site, in Homaterials and Biomechanics 1983, (Eds P. Oucheyne, G. Van der Perre and A.E. Aubert), Elsevier Science Publishers, Amsterdam, 1984,283-288 Jacobsson, M., ICTlebo. P.. Albrektsson. T. and Turesson, I., Provoked repetitive healing of mature bone tissue following irradiation. A quantitative investigation, Acta. Radio/. Oncol. 1986, 25, 57-62. Kalebo, P., Jacobsson, M., Albrektsson, T. and Turesson, I., Bone healing following irradiation during tourniquet ischaemia.Acta. Oncol. 1907,26,63-68 Kalebo, P., Buch, F. and Albrektsson, T., Bone formation rate in osseointegrated titanium implants. Influence of locally applied haemostasis, peripheral blood, autofogous bone marrow and Fibrin Adhesive System (FAS), &and. J. Plast. Reconsrr. Surg. 1988, 22, 53-60 B&remark, P.I., Breine, U., Lindstrom, J., Adeli, R.. Hansson, B.O. and Ohlsson. A.. lntraosseous anchorage of dental prostheses. I. Ewperimental studies. Stand. J. Plast. Reconstr. Surg. 1969, 3, 8 1- 100 Branemark, P.I., Hansson, B.O., Adell, Ft., Breine, U.. Lindstrdm, J., Hallen, 0. and Ghman, A., Osseointegrated implants in the treatment

Bone regeneration in Ti implants: P. /Web0 and M. Jecobsson

9

10 11

12

13

14 15 16 17

18

19

20

21 22

23

24

25

of the edentulous jaw. Experience from a 1 O-year period, &and. J. P/as?. Reconstr. Surg. 1977. 11, 1-i 32. Albrektsson, T., Bn’memark, P.I. Hansson, H.A. and LindstrOm, J., Osseointegrated titanium implants. Requirements for ensuring alonglasting, direct bone-to-implant anchorage in man, Acta. Orthop. &and. 198 1, 52, 155-l 70 Albrektsson, T., The response of bone to titanium implants. CRC, Critical Reviews in Biocompatibility 1985, 1, 53-84 Albrektsson. T., Bone tissue response. In Tissue-integrated Prostheses. Osseointegration in CIinicei Dentistry, (Eds P. I. BrBnemark, G.A. Zarb and T. Albrektsson), Quintessence Publishing Co., Chicago, USA, 1985, 129-l 43 Strid, K.G. and Kaiebo, P., Bone mass determination from microradiographs by computer-assisted vid~ensitomet~. 1. Meth~ol~y. Acra Radio/ 1988, 29,465-472 Kalebo, P. and Strid. K.G., Bone mass determin~i~ from microradiographs by computer-assisted videodensitometry. 2. Aluminium as a reference. Acts Radial 1988, 29, in press Eastman Kodak Co. Kodak High Resolution Plates. Kodak, Rochester, New York, P-47. Eastman Kodak Co. Techniques in X-ray Microscopy: Microradiography. Kodak, Rochester, New York, 1983, EAM-12. Lehmann, E.L., Non-parametrics. Statisticalmethods based on rsnks. McGraw-Hill Book Co.. San Francisco, 1975 Kasemo, B. and Lausmaa, J., Metal selection and surface characteristics, in Tissue-integreted Prostheses. Osseointegration in Clinical Dentist/y. (Eds P.I. Branemark, G.A. Zarb and T. Aibrektsson), Quintessence Publishing Co.. Chicago. USA, 1985, 99-l 16 Adeli, R,, Eriksson, B., Lekholm, U.. Lindqvist. J. and Br&nemark, P.I., A 20-year study of osseointegrated implants in the treatment of fhe totally edenfutous jaw, in pre~r~i~ Galante. J., Rostoker, W., Lueck, R. and Ray, R., Sintered fiber metal composites as a basis for attachment of implants to bone, J. Bone Jt Surg. 1971,63-A, 101-l 14 Lundskog, J., Heat and bone tissue. An experimental investigation of the thermal properties of bone tissue and threshold levels for thermal injury, Thesis, University of Gothenburg, 1972 Linder, L. and Lundskog, J., Incorporation of stainless steel. titanium and vitallium in bone, injury 1975, 6, 277-285 Colella, S.M.. Miller, A.G., Stang, R.G.. Stoebe. T.G. and Spangler. D.M., Fixation of porous titanium implants in cortical bone enhanced by electrical stimulation, J. Biomed. Mater. Res. 1981, 15, 37-46 Albrektsson, T., Bach, A.. Edshage, S. and JC)nsson,A., Fibrin adhesive system (FAS) influence on bone healing rate. A microradi~raphicai evaluation using the Bone Growth Chamber, Acra. Orthop. &and. 1982.53.757-763 Albrektsson. T., Osseous penetration rate into implants pretreated with bone cement, Arch. Orthop. Trauma Surg. 1984, 102, 141147 Rsnningen, Ii., Solheim. L.F. and Langeland. N., Invasion of bone into porous fiber metal implants in cats, Act%. Orthop. Stand. 1984, 56. 352-358

26

Rhinelander, F.W., Tibia1 blood supply in relation to fracture healing,

27

C/in. Orthop. 1974, 105, 34-81 Eriksson. A.R., Heat-induced Bone Tissue /njury. An ln Vivo lnvesti-

28

29 30

31

32

33

34

35

36

37

38

39 40 41

42 43 44

gation of Heat Tolerance of Bone Tissue and Temperature Rise in the Drilling of Cortical Bone, Thesis, University of Gothenburg, 1984 Dallemagne, M.J. and Florkin. M. Etude comparative des &ments mineraux et de I’azote dans les cubitus, les radius et les tibias du lapin adulte, Acta. Biolog. Se/g. 194 1, 1, 406-409 Owen. M., Measurement of the variations in calcification in normal rabbit bone, J. Bone Jt. Surg. 1956, 36-B, 762-769 Balmain. N.. Legros, R. and Bonel. G., X-ray diffraction of calcined bone tissue. A reliable method for the determination of bone Ca/P molar ratio, C&/f. Tiss. Int. 1982, 34. S93-S98 Lane, J.M., Betts, F., Posner, A.S. and Yue, D.W., Mineral parameters in early fracture repair, J. Bone Jt. Surg. 1984, 66-A. 12891293 Laftman, P. and Stromberg, L., Body loading on atrophied bone and bone mineral content. An experimental study in the rabbi, Orthopaedics 1985,8, 1136-I 138 Aro, H., Eerola, E. and Aho. A.J., Determination of callus quantity in 4week-old fractures of the rat tibia, J. Orthop. Res. 1985, 3, lOl108 Listrom, R.D., Changes in Oxygen Tension, Blood Volume and Calcium Content of Tissue in Healing Bone Defects, Thesis, University of Toronto, 1984 Barth, E., Renningen. H., Solheim, L.E. and Saethren, B., Bone ingrowth into weight-bearing porous fiber titanium implants. Mechanical and biochemical correlations, J. Orthop. Res. 1986, 4, 356-361 Albrektsson, T., Repair of bone grafts. A vital microscopic and histological investigation in the rabbit, &and. J. Ptast. Reconsfr. Surg. 1980, 14, l-12 Barth, E., Ronningen, H., Solheim, L.E. and Saethren, B., Effect of CisPlatinum on bone ingrowth into porous fibertitanium. Mechanical and biochemical correlations, ht. J. Oral Maxillofacial implants 1986, 1, 123-127 Rubinacci. A. and Tessari, L.. A correlation analysis between bone formation rate and bioelectric potentials in rabbit tibia, C%/cif. Tiss. lnt. 1983,35. 728-731 Sevitt, S.. Bone Repair and Fracture Healing in Man, ChurchillLivingstone, Edinburgh, 1981 Boskey, A.L. and Posner. A.% Bone structure, composition and mineralization, Or?hop. C/in. North Am. 1984, 15, 597-612 Cruess, R.L.. Healing of bone. tendon and ligament, in Fractures in Adults, 2nd edn., (Eds CA. Rockwood Jr and D.P. Green), Vol. 1) J B Lippincott Co., Phjladelphi~, USA, 1984, 147-l 67 Simmons, D.J., Fracture healing perspectives, C/in. Orthop. 1985, 200. 100-l 13 van Llmborgh, J., Factors controlling skeletal morphogenesis, Progr. C/in. Biol. Res. 1982, 101, l-l 7 Tjellstram. A., Tympanoplasty with Preformed Autologous Ossicles, Thesis, University of Gothenburg, 1977

Biomaterials

1955, Vol9 J&y

301