The diagnostic value of subtraction radiography in the assessment of granular hydroxylapatite implants

oral and maxillofacial radiology Editor: ALLAN

G. FARMAN,

BDS,

PhD (Odont),

MBA

Department of Primary Patient Care University of Louisville School of Dentistry Louisville, Ky. 40292

The diagnostic value of subtraction radiography in the assessment of granular hydroxylapatite implants Werner Engelke, MD,DDS,a Serge de Valk, DDS,b and Urs Ruttimann, Bethesda, Md. NIDR/NIH

AND

UNIVERSITY

PhD,’

OF ZijRICH

Although histologic analysis of osseous changes around hydroxylapatite (HA) implants can be highly accurate, it is of limited use in human beings. Digital subtraction radiography may provide a noninvasive alternative. Ten patients with bony lesions were operated on and nine of the iatrogenic defects were filled with granulated HA. In one patient, the defect was left unfilled for reference. Customized film holders provided standardized radiography. Follow-up images after 4 to 6 months were subtracted from immediately obtained postoperative images, and changes around the implants were noted. From ten pairs of radiographs, eight could be successfully subtracted, whereas two pairs required corrective image transformation before subtraction. Although no bone loss was observed in any of the patients, the implants did not appear to enhance physiologic bone regeneration either. Hence, subtraction radiography holds the potential of clinical utility for the follow-up of HA implants. However, technical improvements are necessary to yield quantitative data. (ORAL SURC ORAL MED ORAL PATHOL 1990;69:636-41)

D

uring the last decade, hydroxylapatite implants have been used with increasing popularity. This remarkably biocompatible substance is noted for its ability to adhere to bone. Hydroxylapatite (HA) is available in a porous (natural) and a nonporous (synthetic) form, having a higher density.’ Combinations with other substances such as collagen and autogenous bone have been described. However, these combinations did not seemto improve the integration into natural bone compared to HA alone.2q3 Currently, HA is being used for several purposes in dental surgery, the most common of which are: @ augmentation of the atrophic edentulous ridge.4 aAssociate Professor, Department of Oral and Maxillofacial Sur&UJ, ~L~iveralty 01 Lurich; Vis~tmg Associate, Diagnostic Systems Branch, NIDR/NIH. bVisiting Fellow, Diagnostic Systems Branch, NIDR/NIH. CChief, Diagnostic Systems Branch, NIDR/NIH. T/16/16544 636

reconstruction of jawbones after pathologic bony defects like cysts, periodontal lesions, and osteomyelitis.5-7 l root implants as preprosthodontic treatment.s l alveolar ridge maintenance after extraction or trauma.9 l bone graft substitute in orthognathic surgery.lO,l4 An in vivo study showed bone regeneration at the periphery of HA implants in surgically created bony defects in rat mandibles within a period of 1 to 3 months. However, there was no bone growth into the central portion of the lesion, where only fibrous connective tissue containing macrophagesand giant cells was present.2 In a similar study with dogs, maxillary alveolar clefts were filled with HA. After 30 weeks, the fibrous tissue previously intervening between the host and the implant was replaced by bone encroaching onto the margins of the HA material.” In another study, HA was implanted in artificially produced periodontal bony defects in dogs. Histologic analysis af-

l

Assessment of HA implants with subtraction

Volume 69 Number 5

Table

radiography

637

I. Distribution of bony lesions and therapies Patient

Age

Sex

Diagnosis

Region

1 2 3 4 5 6 7 8 9 10*

42 40 40 48 68 45 68 38 38 40

M M M M M F M F F F

rad. cyst period. pock. trauma ridge resorpt. rad. cyst rad. cyst rad. cyst ridge resorpt. ridge resorpt. rad. cyst

22123 46 22 45146 31132 22123 34-36 16117 26127 22

= radicular cyst rad. cyst (n = 5) = periodontal pocket (n= I) period. pock. = ridge resorption (n = 3) ridge resorpt. (n= 1) trauma + lesion < 1 cm3 ++ 1 cm3 5 lesion < 2 cm3 +++ lesion > 2 cm3 enucl. = enucleation; reconstr. = reconstruction with HA; augment. = augmentation filling; M = male, F = female; *Patient no. 10 is the control patient.

ter 2 months demonstrated regeneration of bone trabeculae in the apical area. However, signs indicative of direct bone contact with the particles were seldom seen;instead, the presenceeof fibrous connective tissue was evident.t2 Histologic and histometric findings, 5 to 16 months after imbedding of HA-blocks during human orthognathic surgery, indicated bone growth in all biopsy specimensdistributed throughout the porous HA matrix, except where the implant extended into the soft tissue. Inflammatory cells were very scarce.i3 Besides osteogenesis,HA implantation may have side effects such as volume loss caused by bone resorption, contraction of the wound, or loss of particles due to insufficient wound margin closure. After augmentation of the mandible, where volume decreasecan relatively easily be detected, the maximum height loss after 2 to 3 years is approximately 30%. Most of this loss occurs during the first 6 months.15 Although histologic analysis of the osteodynamics in and around HA implants is considered to be a highly accurate method, the need to obtain biopsies makesit of limited use in human beings. Radiography is a convenient, noninvasive alternative to detect changes in bone. Panoramic radiographs have been used to measure changes in the vertical height of augmented ridges. Is Individual film-holders registering the complete dental arch, or part of it, with or without marker or alignment aids, were constructed to assessperiodontal bone changes.16-I8Also, computerized tomography has been applied to evaluate augmented mandibles. l9 Recently, dual-photon absorptiometry has beenintroduced to estimate the bone mineral content.20

I

Size ++ + +++ +++ ++ ++ +++ ++ ++ +

~~-.-~--

with HA; rg = retrograde

I

Therapy enucl. rg/22 reconstr. reconstr. augment. muck og/3 1 enucl. rg/22 enucl. rg/34 augment. augment, enucl. rg/22

filling;

og = orthograde

As another alternative, digital subtraction radiography may be considered. This technique has already proved to be accurate in discriminating subtle changes in bony architecture. 21 With this method, geometrically standardized radiographs are produced longitudinally, and then digitized and subtracted by computer. 22 Because unchanged structures will cancel in the subtraction image, ideally only anatomic differences between one point in time to another will be displayed. Although this technique has already been successfully used for the detection of gain or loss in tooth-supporting bone,23the method has not been applied for the demonstration of osseous changes around implants. This article reports the follow-up of porous HA implants in patients, with the use of customized filmholder devices and the subtraction radiographic technique, to evaluate the osteodynamics at the periphery of the inserted material. MATERIAL

AND METHODS

Ten patients were evaluated over a period of 4 to 6 months after surgery. The study subjects were surgically treated for a variety of pathologic bony lesions. Detailed data are displayed in Table I. All operations were performed by the samesurgeon with the patients under local anesthesia (Ultracaine DS forte) at the Department of Oral and Maxillofacial Surgery of the University of Zurich, Switzerland. The type of HA used was Interpore 200 (Interpore International, Irving, California) porous granules of different diameters (20 to 40 mesh). In the case of a cyst, the surgical procedure consisted of a mucoperiosteal vestibular flap, enucle-

638

Engelke, de Valk, and Ruttimann

ORAL SURG ORAL MED ORAL PATHOL May 1990

Table II. Bone gain versus volume loss of the implant Patient

Time (months)

1 2 3 4 5 6 I 8 9 10*

4 4 4 4 4 6 4 4 6 6

Bone gain + 0 + 0 + ++ 0 0 0 +

LOC AL A,L A A,L,O

0 0 --

0,L 0 w0 w-

0 0 AL0

+ = bone gain. - = volume loss. ++ = gain >> loss.

-- = loss <
ation of the cyst, curettage of the bone surface, orthograde or retrograde filling of the root canal with gutta-percha points and Dycal or amalgam, and filling of the iatrogenic defect with Interpore 200 under low compression of the granules. The amount of implanted material varied from 0.5 to 1.5 gm. The flap was replaced with mattress sutures of nonabsorbable material. For localized ridge dehiscences, the samesurgical approach was used. After smoothening of the bone surface with a surgical drill, the alveolar crest was augmented with 1.0 to 1.5 gm HA. If stabilization of the granules was difficult, a thin slice of lyophilized human cartilage was overlayed. After splitting of periosteum and mucosa, the vestibular flap was replaced. In the case of the periodontal pocket, only a small flap was necessary, bone and root surfaces were carefully curetted, and 0.5 gm HA was inserted. For comparison to bone regeneration without HA implants, one patient’s defect was not filled after a radicular cyst was enucleated. This reference patient was evaluated in the samemanner as the other nine cases. Standardized radiography was achieved by using long-cone metal film-holder devices with bite blocks made of rapidly polymerizing acrylate, providing a customized occlusal imprint of the area of interest.24 The x-ray source was a Ritter Transdent 1502 unit (Ritter, Karlsruhe, FGR) set at 68 kVp and 12 mA, with an exposure time of 0.6 to 0.8 seconds.Depending on the available intraoral space, dental films of L X 3 or 3 X 4 cm (Kodak Ultraspeed) were used. By means of the paralleling technique (long-cone technique), the central ray of the x-ray beam was directed at right angles to the teeth and film to minimize geo-

metric distortion. Radiographs were made immediately after surgery, and 4 to 6 months postoperatively. The films were developed with a Dilrr Dental AC 245L automatic developer (D&r, Bietigheim, FGR) under standardized conditions (28” C/5 minutes) . All radiographs were converted into 5 12 X 5 12 X 8 bit digital images, such that as many of the 255 available gray levels as possible were used, to obtain maximal contrast resolution. This was accomplished by using a lightbox with an adjustable light level, a charge-coupled-device video camera (Applied Intelligent Systems, Inc., Ann Arbor, Michigan), and an analog-to-digital converter, interfaced with an image processing system (DeAnza IP6400, Gould Inc., Fremont, California) running on a VAX/750 (Digital Equipment Corp., Maynard, Massachusetts)host computer. The images of the baseline radiographs were stored and intensity inverted. Thereafter, the followup images were superimposed over their negative counterparts, by placing the secondary radiographs under the video camera and mixing the corresponding real-time video signal with that of the stored negative. The images were registered by careful rotations and translations with the use of a mechanical manipulator attached to the lightbox carrying the radiograph. When the region of diagnostic interest seen in realtime on a video monitor attained best cancellation of the structures, the second image was digitized in the same way as its baseline image. Approximate film contrast matching was achieved by adjusting the light level of the lightbox, and final matching was obtained by a digital contrast correction method.25 The changes in the region of the implants, as demonstrated in the subtraction image, were analyzed by measuring the relative areas over which bone gain or volume loss occurred. A gray-value threshold was selected at mean background plus estimated noise level above which all gray-values were interpreted as bone gain. Similarly, all gray-values below a threshold set at the mean background, minus estimated noise level, were considered as volume loss. The number of pixels representing either gain or loss was counted at each implant site for quantitation of the results. RESULTS

Ten pairs of radiographs were acquired over a time period of four to six months. Eight of thesepairs could be successfully subtracted and analyzed, whereastwo pairs showed poor geometric standardization. This was caused by migration and tilting of the teeth toward the diastema, preventing an optimal fit of the customized film holder in the area of interest. After additional scaling (affine transformation) of these images by computer, the tentative assessmentmade

Assessment of HA implants with subtraction radiography

Volume69 Number 5

639

Fig. 1. Digital subtraction of standardized radiographs evaluating reconstruction of alveolar crest with HA after traumatic loss of maxillary left lateral incisor. A and B are the digitized and gamma corrected radiographs, immediately and 4 months after surgery. C shows the subtraction image. In D the regions of bone gain (white) and volume loss (black), as determined by gray level thresholds, are superimposed over the subtraction image. Only peripheral changes were scored, in this case as + and -.

with accordingly reduced sensitivity was that there was no bone gain or volume loss. Fig. 1 shows as a representative example the standardized radiographs from patient 3, together with the corresponding subtraction image and the relative areas of bone

gain (white) and volume loss (black). Relative gain and loss, as well as their localization, were evaluated for each patient separately (Table II). In patients in whom both bone gain and volume loss

occurred simultaneously, gain exceeding loss by more than factor 2 is indicated by ++, and vice versa by --. None of the subtraction images indicated bone loss in the area of interest. In five patients (SO%),including the control patient, we detected significant bone regeneration, in three of these patients combined with volume loss of the implant

material.

In two cases

(20%) only volume loss occurred, and the three

remaining cases(30%) did not show any changes over time. If there was new bone formation, this occurred at all times at the apical area of the implant. Volume loss, on the other hand, was most apparent at the oral cavity site of the implant. Furthermore, it was noticed that volume loss occurred around the larger implants and that bone gain was only seen in intrabony defects (enucleated cysts). Alveolar ridge augmentation with HA did not seemto induce bone regeneration, nor did the implant in the periodontal pocket. However, the granules remained in situ, in most caseswithout visible volume loss. DISCUSSION

A feasibility study was performed to investigate the value of subtraction radiography in the assessmentof implanted HA granules in several types of bony

640

Engelke, de Valk, and Ruttimann

lesions. From the ten pairs of radiographs in our investigation, eight pairs could be successfully subtracted and analyzed by a computer digitized system. In two cases,the radiographs presented poor geometric standardization because of misfitting of the occlusal imprints after the follow-up interval. Both cases involved ridge augmentation of a large dorsal diastema in the maxilla and, in addition, an unstable occlusion of the posterior antagonists. This enabled the superior molars and premolars to rotate and migrate toward the defect. Since the occlusal tooth surfaces are the only available hard structures in the oral cavity to be used for customized film-holder devices, slight movements of the teeth will hinder optimal reproducibility. If the projection angle deviation is within a reasonable range, proper affine transformation of the radiographic images can improve the quality of the subtraction images. However, some reduction in the sensitivity for detecting changes remains. When movements of teeth adjacent to large defects are likely to occur, it is suggested to expand the individual bite blocks. This permits the utilization of a larger number of tooth surfaces, which are farther away from the lesion site and are in a more stable area of the jaw. An alternative technique in such cases is to use cephalostat head-stabilization.26 A calibration step-wedge or ramp is indispensable for converting the density changes in the two-dimensional subtraction images into massor equivalent volume changes. No such wedge was used in this preliminary investigation.27 Since the gray-level threshold settings in the subtraction images were somewhat subjective, only a categorical quantitation was performed, distinguishing bone regeneration and volume loss of the implanted HA and their relative magnitudes of gain and loss at each site. Among the different kinds of bony lesions evaluated in this investigation, the subtraction images of the radicular cysts were qualitatively the best. This is probably due to the relatively stable position of the teeth next to the defect, and the localization of the cysts in the anterior part of the jaw, providing an easy access for the operator to place the films. Slight movements of the HA particles causing large changes in the gray-level patterns did not allow us to evaluate changes in density in the central portion of the implants other than on an average basis. In the peripheral area, on the other hand, density changes could be detected easily. In none of our ten patients did bone loss occur. Volume loss due to compaction of the implanted material was mainly seen after filling of larger defects and always occurred at the oral cavity site. In five

ORAL SURG ORAL

MED ORAL PATHOL May 1990

cases, including the control patient, bone gain was observed. These casesinvolved four enucleated cysts and the traumatic loss of an incisor tooth. Bone regeneration was mainly seen apically, in the area of the thickest bone mass with the best vascularization. The HA implants did not appear to causebone loss in any of our patients, but compared to our reference patient, the implants did not seem to improve physiologic bone regeneration either. In conclusion, subtraction radiography with the use of customized film holders can be very useful in the evaluation of HA implants. It is a convenient, noninvasive procedure for the patient, but in the present stage of development, it remains a rather complicated diagnostic method for the clinician. In this preliminary study, the computerized subtraction analysis allowed us to identify changes in the periphery of the implants. However, the method needsto be improved by incorporating a calibration step-wedge,permitting determination of mass gain or loss. The use of customized bite blocks for standardization of the radiographs is not advisable if tooth movements are likely to occur and extraoral head-stabilization should be considered. REFERENCES

OgilvieA, Frank RM, Benqu’e EP, Gineste M, Heughebaert M, Hemmerle J. The biocompatibility of hydroxyapatite implanted in the human periodontium. J Periodont Res 1987;4:270-83. 2. Bell R, Beirne OR. Effect of hydroxylapatite, tricalcium phosphate, and collagen on the healing of defects in the rat mandible. J Oral Maxillofac Surg 1988;7:589-94. 3. Frame JW, Rout PG, Browne RM. Ridge augmentation using solid and porous hydroxylapatite particles with and without autogenous bone or plaster. J Oral Maxillofac Surg 1987;9: 771-8. 4. Desjardins RP. Hydroxyapatite for alveolar ridge augmentation: indications and problems. J Prosthet Dent 1985;3:374-83. 5. Shepard WK, Bohat 0, Joseph CE, Lopiccolo P, Bernick S. Human clinical and histological responses to a Calcitite implant in intraosseous lesions. Int J Periodontics Restorative Dent 1986;3:46-63. 6. Meffert RM, Thomas JR, Hamilton KM, Brownstein CN. Hydroxylapatite as an alloplastic graft in the treatment of human periodontal osseousdefects. J Periodontol 1985;2:63-73. I. Block MS, Zide MF, Kent JN. Excision of sclerosing osteomyelitis and reconstruction with particulate hydroxylapatite. J Oral Maxillofac Surg 1986;3:244-6. 8. Brook IM, Sattayasanskul W, Lamb DJ. Dense hydroxyapatite root replica implantation: tooth site and successrate. Br Dent J 1988;7:212-5. 9. Kentros GA, Filler SJ, Rothstein SS. Six month evaluation of particulate Durapatite in extraction sockets for the preservation of the alveolar ridge. Implantologist 1985;2:53-62. 10 Wolford LM, Wardrop RW, Hartog JM. Corallin porous hydroxylapatite as a bone graft substitute in orthognathic surgery: J &al Maxillofac Surg 1987;12:1034-42. 11. Cullum PE. Frost DE. Newland TB. Keane TM. Ehler WJ. Evaluation of hydroxylapatite in repair of alveolar clefts in dogs. J Oral Maxillofac Surg 1988;4:290-6. 12. Minabe M, Sugaya A, Satou H, et al. Histological study of the 1.

Volume 69 Number 5

13. 14.

15. 16. 17. 18. 19. 20.

21.

Assessment of HA implants with subtraction

hydroxylapatite-collagen complex implants in periodontal osseous defects in dogs. J Periodontol 1988;10:671-8. Holmes RE, Wardrop RW, Wolford LM. Hydroxylapatite as a bone graft substitute in orthognathic surgery: histologic and histometric findings. J Oral Maxillofac Surg 1988;8:661-71. Vanassche BJ, Stoelinga PJ, de Koomen HA, Blijdorp PA, Schoenaers JH. Reconstruction of the severely resorbed mandible with interposed bone grafts and hydroxylapatite. A 2-3 year follow-up. Int J Oral Maxillofac Surg 1988;3:157-60. Block MS, Kent JN. Long-term radiographic evaluation of hydroxylapatite-augmented mandibular alveolar ridges. J Oral Maxillofac Surg 1984;12:793-6. Renggli HH, Steiner E, Curilovic Z. Reproducible radiographs and photographs in periodontal diagnosis. Paradontologie 197 1;3:66-74. Rosling B, Hollender L, Nyman S, Olson G. A radiographic method for assessing changes in alveolar bone height following periodontal therapy. J Clin Periodontol 1975;4:21 l-7. Nery EB, Olson WJ, Henkin JM, Kalbfleisch JH. Film-holder device for radiographic assessment of periodontal tissues. J Periodont Res 1985;1:97-105. Osborn JF, Brecht G, Kapovits M, Stuckenholz CA. Computerized tomographic analysis of the mandible augmented with hydroxylapatite ceramic granules. Zahnarzt 1986;3:165-76. Von Wovern N, Worsaae N. Bone mineral content of the maxilla estimated by dual-photon absorptiometry after augmentation with bone or hydroxyapatite. J Dent Res 1988; 11:1405-8. Rethmann MP, Ruttimann UE, O’Neil R, et al. Diagnosis of bone lesions by subtraction radiography. J Periodontol 1985; 6:324-9.

radiography

641

22. Grsndahl HG, Griindahl K, Webber RL. A digital subtraction technique for dental radiography. ORAL SURC ORAL MED ORAL PATHOL 1983;1:96-102. 23. Schmidt E, Webber RL, Ruttimann UE, Loesche WJ. Effect of periodontal therapy on alveolar bone as measured by subtraction radiography. J Periodontol 1988;10:633-8. 24. Engelke W, Velvart P. Eine praxisnahe Methode zur Herstellung von reproduzierbaren ZahnrBntgenaufnahmen mit der Rechtwinkeltechnik. Schweiz Monatsschr Zahnmed 1987; 7: 864-8. 25. Ruttimann UE, Webber RL, Schmidt EF. A robust digital method for film contrast correction in subtraction radiography. J Periodont Res 1986;5:486-95. 26. Jeffcoat MK, Reddy MS, Webber RL, Williams RC, Ruttimann UE. Extraoral control of geometry for digital subtraction radiography. J Periodont Res 1987;5:396-402. 27. Ruttimann UE, Webber RL, Saffer A. Calibrated volume determination of localized bone lesions by subtraction radiography. In: Salamon R, Blum B, Jd rgensen M, eds. MEDINFO 86. Amsterdam: Elsevier Science Publishers B.V., 1986;634-7. Reprint requests to: Dr. Werner Engelke Kieferchirurgische Klinik und Poliklinik Direktor: Prof. Dr. Dr. H.F. Sailer Frauenklinikstrasse 10 8092 Ziirich/Switzerland