Photodensitometric evaluation of osseous repair following Le Fort I osteotomy

J Oral Maxillofac Surg 44:977-986,1986

Photodensitometric Evaluation of Osseous Repair Following Le Fort I Osteotomy JEFFREY R. SCHANTZ, DDS, MDENTSC,* CHARLES N. BERTOLAMI, DDS, DMEDSC,t AND RAVINDRA NANDA, BDS, PHD* This study evaluated photodensitometry as a noninvasive method for quantitating bone mineral content (BMC) and osseous repair after Le Fort I osteotomy. Le Fort I osteotomies were performed on 6 Macaca fasicularis monkeys; maxillas were either advanced (Group I, n = 3) or impacted and advanced (Group II, n = 3). Postoperative, standardized lateral cephalometric films were taken at weekly intervals up to 25 weeks and osteotomy site repair was studied using photodensitometry. Segment stability was also evaluated and correlated with measured densities. In both experimental groups, clinical stability occurred at about the same time (45.7 and 48.7 days postoperatively) despite large differences in the size of the initial surgical defects. The net rate (slope) of osteotomy site remineralization was significantly different (Group II > Group I), but the relative difference in film absorbance between the osteotomy site and adjacent bone at the time of clinical stability was the same. This difference can be extrapolated from early postoperative films and may constitute a useful parameter for predicting when clinical stability will be achieved.

A need has been identified for a quantitative, noninvasive procedure for measuring the extent of osseous union following fractures and osteotornies.tr" Traditionally, bone repair has been monitored using radiographic and clinical exarninations." Although both techniques are routinely used to make treatment decisions, neither is quantitative or reproducible between different surgeons.v" As a result, accurate evaluation of the state of bone repair cannot be performed with confidence." Using photodensitometry, this study quantitatively and noninvasively evaluated changes in min-

eralization at maxillary osteotomy sites during the healing process and assessed the reproducibility of the photodensitometric method for determining radiographic film density along a specified scan path. This procedure indirectly measures osteotomy site mineralization relative to the mineral content of adjacent bone and allows the rate of osseous healing to be correlated with clinical stability. In addition, rates of repair can be compared readily between animals receiving different surgical procedures.

* Department of Orthodontics, School of Dental Medicine, University of Connecticut Health Center, Farmington, Connecticut. t Department of Oral and Maxillofacial Surgery, Massachusetts General Hospital and Harvard School of Dental Medicine, Boston, Massachusetts. Supported in part by USPHS Research Grants DE 07047-07 (I.R.S.), DE 05396-05 (R.N.), and DE 059241DE 06849(C.N.B.), from the National Institute of Dental Research. National Institutes of Heallh, Bethesda, Maryland. Address correspondence and reprint requests to Dr. Bertolami: Department of Oral and Maxillofacial Surgery, Massachusetts General Hospital, 32 Fruit Street, Boston, MA 02114. 0278·2391/86 SO.OO + .25

Six female Macaca fascicularis monkeys (aged 24-32 months) were studied over an 8-month period. All were housed under identical environmental conditions and were maintained on a Purina Monkey Chow diet (Ralston Purina, St. Louis, MO).

Materials and Methods

ANESTHETIC AND SURGICAL PROCEDURES

For inserting implants, performing osteotomies, and obtaining radiographs, the monkeys were se977

978

PHOTODENSlTIOMETRIC EVALUATION OF OSSEOUS REPAIR

Table 1. Surgical Movement in Experimental Groups Animal No. I 2 3 4 5 6

Surgical Group I I I 11 11 11

EVALUATION OF OSSEOUS REPAIR

Surgical Movement" (A-P)

(Vertical)

1.0 3.5 4.0 4.0 4.0 4.0

0.5 0.5 0.5 2.5 2.0 2.0

* Surgical movements expressed in mm. The amount of surgical movement planned and that actually obtained do not always coincide; therefore, these measurements were taken from postoperative cephalometric radiographs and are proportional to the amount of maxillary displacement actually achieved.

dated with intramuscular ketamine HCI (Bristol Laboratories, Syracuse, NY) (15 mg/kg) and acepromazine (Fort Dodge Laboratories, Fort Dodge, IA) (0.75 mg/kg). During the maxillary osteotomy, general anesthesia was achieved with 70% N 20 , 30% 02' and 0.5% fluothane. After the animals had been quarantined for 4 weeks, sterile, elemental tantalum implants (0.2 mm x 1.5 mm) (Bangsdal Dental, Ltd., Copenhagen, Denmark) were placed bilaterally in the area of the maxillary canine, first molar, and tuberosity, as well as in the midline of upper and lower jaws,7·8 Two implants were positioned at the cranial base midliner-'? and on each side of the zygomaticomaxillary, zygomaticotemporal, zygomaticofrontal, and frontomaxillary sutures ..These were placed to verify radiographically the accuracy of a given animal's repositioning within the cephalostat at the designated examination intervals and also to establish reproducible scanning paths between specified implants for each film. Six to eight weeks after implant insertion, the animals were divided into two groups (Table 1). Group I (animals 1, 2, and 3) underwent a Le Fort I advancement of 4 mm with minimal reduction of lower facial height and minimal autorotation of the mandible, as previously described (Fig. lA).8 Group II (animals 4,5, and 6) underwent a Le Fort I advancement of 4 mm with a 3-mm impaction of the maxilla to allow for autorotation of the mandible and a decrease in lower facial height (Fig. lB). In both groups, transosseous wires were placed in the piriform rim and buttress regions to stabilize the maxilla without the use of maxillomandibular fixation (MMF). The surgical procedure produced a Class II occlusion in both groups. Total body weight was measured immediately preoperatively and when maxillary stability was tested.

Stability. To test stability, the maxilla was manually manipulated in the sagittal, frontal, and transverse planes at postoperative intervals of 1, 2, 3,4, 5, 6, 7, 8, 9, 11, 13, 17,21, and 25 weeks. The animal's head and mandible were firmly held so that both condyles were seated in their most superior, posterior position. Changes in the overbite, overjet, and dental midlines induced by manipulation of the maxilla were measured. The maximum range of maxillary mobility was determined by subjecting the bone to reciprocal displacement forces in all three dimensions. Greater than 1.5 mm of movement was considered "severe" mobility, 1.5-1.0 mm was designated "moderate" mobility, and 1.0-0.5 mm was considered "slight" mobility. A maxilla showing less than 0.5 mm of movement in all planes was considered stable. However, because

FIGURE I. Design of the osteotomy used. A. Group I (maxillary advancement only); B, Group 11 (maxillary advancement with superior repositioning).

979

SCHANTZ ET AL.

mobility could only be measured to the nearest 0.25 mm, stability was defined as 0.25 mm mobility or less. Radiographic assessment. Lateral cephalometric films were taken immediately after implant placement, one day after maxillary osteotomy, and at weekly intervals for the first nine weeks postoperatively. Radiographs were then taken every two weeks for the next month (weeks 9-13) and then monthly until six months postoperatively. A modified Wehmer cephalostat (B. F. Wehmer, Franklin Park, IL),8,II,12 Kodak Ortho-H film (OH-I), 8 inch x 10 inch, (lot #4851) (Eastman Kodak Company, Rochester, NY), and a standardized film exposure and processing technique were used for cephalometric headplates. Each animal was radiographed each week using the same film cassette (Kodak XOmatic with Kodak Lanex regular screens). Films were exposed for 0.1 seconds at 15 mA and 80 kV(p) with a standard x-ray source to midsagittal plane distance of 60 inch. The film-midsagittal plane distance was 6 inch. Films were batch-developed each week (Kodak RP X-Omat processor) at 87° F for 48 seconds. PHOTODENSITOMETRY

A Corning model 760 Computing Fluorometer/ Densitometer (Corning Medical, Medfield, MA) was used. The detector output voltage of the photodensitometer is proportional to the amount of light detected and is a function of the density of the region of the film being studied (Fig. 2). If the film is dense, less light is transmitted and a lower detector output voltage results. The resulting output is electronically inverted causing dark areas of the film (high optical density) to produce positive deflections on the recorded plot. These cause light absorption (i.e., density) rather than light transmission to be plotted. Because of the property of computing densitometers to invert voltage output, and the desire to have the metallic implant images produce positive peak deflections, film reversals were used. Film reversals were produced in a standard film duplicating machine by exposing Kodak subtraction mask film (Kodak X-Omat Subtraction Mask Film, KP 63083E, 8 inch x 10 inch, Lot #2441321, Eastman Kodak Company, Rochester, NY) overlaid with the original lateral cephalometric film, for a period of 2 seconds (Fig. 3). This subtraction film was chosen because of a property of its emulsion which produces a reversal having the same contrast as that of the original film. All original films were reversed at the same time to avoid variation in processing chemistry. The computing densitometer used in this study automatically ad-

REVERSED FILM

I ~ --£--,I----H--'-~ FILTER

I

SLIT

COlLIMATION

ANALOG PlOT OF _ ABSORBANCE

FIGURE 2. Flow diagram depicting densitometer function. Since the output signal from the photo-diode detector is inverted. light absorbance rather than transmission is measured. 10 • intensity of the incident beam of light; I, beam intensity after passage through a reversed image lateral cephalometric film.

justed baseline and maximum deflections to maximize the range of the recorder plot (measurement error = ± 1.0%). To assure linearity of film density versus radiation exposure under designated conditions of kV(p), rnA, exposure time, film type, and development procedures, an aluminum step wedge was inset next to the animal's head and included in all radiographs. Static optical density values were measured for each of the 18 steps on the wedge for every scan in the series of a randomly chosen animal using a Tobias Associates Model TBX static densitometer. Comparisons were made between the optical densities of osteotomy sites and adjacent bone to ensure that these areas of interest were within the linear portion of the Hurter-Driffield curve (Fig. 4). The Hurter-Driffield curve is a sigmoidal plot of film density as a function of radiation exposure. At high and low exposures, radiographic film does not exhibit a linear response to changes in intensity of the incident beam. Regression analysis was performed to test the linearity of densitometer voltage response as a function of film density variation. An osteotomy site is less mineralized than adjacent bone prior to complete repair, thus it appears as a valley on the plot (Fig. 3). Because the densitometer integrates areas under peaks, and not valleys, osteotomy site areas were computed relative to each other and to the adjacent bone as shown in Figure 5. REPRODUCIBILITY

To ensure that film scanning was always performed across exactly the same site along the osteotomy at each postoperative interval, scan paths were chosen between specified metallic implants.

980

PHOTODENSITIOMETRIC EVALUATION OF OSSEOUS REPAIR

FIGURE 3. Top left , lateral cephalometric film of an experimental a nimal. Implants are radiopaque and osteotomy sites appear radiolucent in a con ventional image film. Top right, reversed image exposure of the film in top left. Implants appear radiolucent and the osteotomy site appe ars radiopaque. The arrows designate the linear scan path of the densitometer beginning at the superior border of the cranium. Bottom right, densitometric plot of the scan path identified in top right. The origin of the scan is on the left and represents the anterior cranium (AC) of the experimental animal followed by the maxilla and mandible. Implants for delineating the scan path appear as peaks, while the osteotomy site appears as a down ward deflection.

The densitometric peak configuration of each implant is qualitatively similar at each interval only if the animal's head position is constant relative to the incident beam and the film. The implant peaks of consecutive densitometric printouts taken at each interval were superimposed to establish animal positioning error (Fig. 6). Only scans with congruent implant peak configurations were used. Errors due to machine electronics, film positioning in the densitometer, variability in photon energies in the beam flux, scatter radiation, and film processing were determined. Reproducibility of osteotomy site area measurement on densitometric

A.C.

MAXIL

MA

D.

- - -+-- - - - - H plots was measured and error was expressed as the ratio of the 95% confidence interval over the mean area. Variations in radiation exposure and film processing were analyzed by comparing optical densities of films exposed through identical step wedge thicknesses and processed both at the same time and at different weekly intervals. Estimates of error, expressed as the ratio of the 95% confidence interval over the mean optical density, were determined. Comparisons were made between these estimates of error and those computed from the data used in determining the linearity of film density.

981

SCHANTZ ET AL.

B

J.5

10

2.5 __

J~2~L ______

Z: 2.0 'iii

c

D

E

F

I 1 1 I I 1

c: Q)

0

1.5

I

I I I I I 1 1 1(1.70)

1.0

J~~_

0.5

I

0.2 •

I Base+fog 0.5

'(Q701

I

1.0

1.5

2.0

2.5

10 A] ,-

3.0

Log Relative Exposure

......

f

I

~1: 1

FIGURE 4 (left). Hurter-Driffield curve for a representative cephalometric film. For film density to have a constant relationship with the subject's tissue density, measured optical densities must fall within the linear portion of this characteristic curve (delineated by the dashed lines.) If the average gradient (slope) of the curve is larger than 1.0, film contrast between different tissues will be greater than subject contrast; whereas, films with an average gradient equal to 1.0 will not exaggerate subject contrast. (From Curry TS, et al: Christensen's Introduction to the Physics of Diagnostic Radiology, 3rd ed. Philadelphia, Lea & Febiger, 1984. Reprinted with perrnission.) FIGURE 5 (right). A, graph of a portion of a densitometric plot showing the osteotomy site and adjacent implant. The osteotomy site appears as a downward deflection; the area of this deflection represents the relative difference in mineral mass between the osteotomy site and the adjacent bone. B, to measure the area of the osteotomy site deflection, points were chosen on each side of the osteotomy site by identifying the intersection of the densitometric plot and a line bisecting the vertical lines of the graph paper. C, a best-fit line was constructed through these points and represented the slope of the baseline used to delineate the area of the osteotomy site. D, the line constructed in C was horizontally translated until it coincided with the beginning of the osteotomy site. This line represents the hypothetical appearance of the plot if the osteotomy site were completely healed. E, magnification of the plot near the osteotomy site with the lines of the densitometer graph paper superimposed. F, further magnification with the lines of an ocular grid from a dissecting microscope superimposed over the smallest square on the graph paper. Thus, one square of the graph paper was divided into 100 additional squares. The osteotomy site area was computed by counting the number of squares it contained. A square was counted if any part of it was contained in the area of interest.

Results ANIMALS

Although no serious post-osteotomy complications occurred, all animals initially lost 10%of body weight (mean weight loss = 0.2 kg). By the fourth postsurgical week, all animals had returned to their presurgical weight and subsequently increased in weight. The magnitudes of the surgical repositioning of the maxillas in Groups I and II are shown in Table 1. Although a maxillary advancement of 4 mm was planned for all animals, animal No. 1 (Group I) received only a l-mm advancement. In

the vertical dimension the mean amount of impaction for Group I was negligible (~ 0.5 mm), as anticipated. Group II received a superior repositioning of2.2 ± 0.3 mm (mean ± 1.0 SD). This was slightly less than the 3-mm impaction that was planned. EVALUATION OF OSSEOUS REPAIR

BOlle Stability

The number of days between surgery and clinical stability for both groups is shown in Figure 7. The length of time required to achieve stability for

982

PHOTODENSITIOMETRIC EVALUATION OF OSSEOUS REPAIR

I----

AC--~-MAXILLA--+MAND-l

IMPLANT

l

DISTANCE FIGURE 6. Interval change between densitometric plots of absorbance versus scan path distance for an experimental animal scanned one week apart. The two scans represented by the solid and dashed lines are closely matched except for the segments of the plot denoted by the letters A and B. A corresponds to a plexiglass plug used to support the aluminum step wedge in one of the plots; B corresponds to the mandible and would not be expected to be reproducibly portrayed in the absence of MMF. Note perfect superimposition of the implant peaks; this ensures that the experimental animals were reproducibly positioned in the cephalostat at both points in time. AC. anterior cranium; MAND. mandible.

Group I was 45.7 ± 3.3 days (mean ± S.E.M.) and for Group II was 48.7 ± 0.3 days. Five out of 6 monkeys were stable at 7 weeks (48.8 ± 0.2 days); however, animal No. I was stable by 39 days. The difference between Group I and Group II in terms of the amount of time required for clinical stability to be achieved was evaluated by a t-test for independent variables and was found not to be significant. In addition to studying differences in the time needed for attaining clinical stability, comparisons were made regarding the time needed for each animal to proceed from "severe" to "moderate" mobility and from "moderate" to "slight" mobility as defined earlier (Fig. 7). Differences between the groups relative to the number of days required to pass each transition point also were not significant.

very subtle differences were discerned among individual films and only when those films were viewed together. The osteotomy site was clearly visible one week postoperatively. At four weeks, the osteotomy line was no longer well demarcated and was difficult to distinguish from the adjacent bone. Additional radiographic changes up to the time of clinical stability were minimal. Further changes in the density of the osteotomy line could not be detected through visual inspection of radiographs from the eighth to the twelfth week postoperatively. Photodensitometry: As mentioned earlier, valid comparisons between the photometric density of the film at the osteotomy site and the adjacent bone require that density values fall within the linear range of the Hurter-Driffield curve. This criterion was verified by computing the mean optical density for each step of the aluminum wedge for an entire series of scans of one animal (Fig. 8). The 95% confidence interval of the mean optical density of the osteotomy site and adjacent bone (represented by the dashed lines in Fig. 8) is four-step wedge increments away from the end points used to determine the linear regression. All of the density values measured in this scan series were within the limits of the points used to determine the regression line. Therefore, the densities involved in studying osteotomy sites and adjacent bone were within the linear range of the Hurter-Driffield curve and were suitable for making comparisons between the osteotomy site and adjacent bone regardless of minor differences in film density caused by variation in film processing or radiation exposure or both under the conditions of the study. The difference in absorbance between the osteotomy site and adjacent bone on the negative image films was plotted as a function of time after surgery for each animal. Correlation coefficients, deter-

2.5

Surgical Group • I o II

Severe 1.5

- - - - - - - --- - - - - - - - - - - - - - - - - - - - - - - - - - - - - ---

Moderate

•

1.0

Radiographic Assessment Visual inspection. Visual examinations of serial lateral cephalometric films were not easily correlated with changes in density of the osteotomy site during repair nor with the transition times when segment mobility changed from "severe" to "moderate," "moderate to slight," or "slight" to "stable." By subjective clinical examination, only

Slight 05

Stable 0

0

\0

60

TIME (Days)

FIGURE 7. Maxillary segment mobility (mean values) versus time. Stability was defined as less than 0.5 mm of mobility in all dimensions.

983

SCHANTZ ET AL.

2.0

1.8 :>..

I..

Y>

1.4

C3 <::l -J

~ I:::

t.O

~

0.6

0.2

0.2

0.4

0.6

0.8

1.0

RELATIVE WEOGE THICKNESS (INCHES)

o

w

w

m

~

H

~

ro

~

ro

~

Time (Days)

FIGURE 8 (left). Reversed image film optical density as a function of step wedge thickness. The dashed lines represent the 95% confidence interval for mean values of optical density for the osteotomy sites and adjacent bone for all the films of a randomly selected animal. All optical density values for osteotomy sites and adjacent bone were within the limits of the points used to determine the regression line shown (r = 0.99). This criterion ensures that a constant relationship existed between film density and subject tissue mass. FIGURE 9 (right). Osteotomy site photodensity (area) versus time. The y-intercept represents the relative size of the osteotomy defect at the time of surgery. Group I = 339 :!: 78 units (rnean ee SEM); Group II = 692 :!: 102 units. The slope represents the relative rate of osteotomy site remineralization and is expressed as the mean z SEM. Arrows denote the point when clinical stability was reached.

mined for each regression line, varied from - 0.88 to -0.96. Osteotomy site area plotted as a function of time for surgical Group I and Group II is shown in Figure 9. The y-intercepts of the regression lines were the extrapolated values for osteotomy site area at the time of surgery and were representative of the relative size of the original bony defect. The slopes of the regression lines were estimates of the net rate of remineralization of the surgical defects. The mean y-intercept for Group I (maxillary advancement only) was 339 absorbance units; Group II (maxillary advancement plus superior repositioning with wedge of bone removed) had a mean y-intercept of 692 absorbance units. The difference between the two groups was statistically significant . (P < 0.05) and indicated that approximation of bony segments can be performed with greater precision when bone is not removed. The slope of the regression line for Group I (4.1 absorbance units/ day) was considerably less than the slope for Group II (10.2 absorbance units/day) and indicated a higher net rate of remineralization for osteotomies with a larger surgical defect (P < 0.05). When the percentage density change at the osteotomy site on the reversed films from day zero to the time of clinical stability of Group I was compared to that of Group II, no statistically significant difference was found between the surgical groups; nor was there a difference between the groups in the density at the time of clinical stability. The difference in absorbance between the osteotomy site

and the adjacent bone at the time of clinical stability was similar for surgical Group I and Group II (Fig. 9). REPRODUCIBILITY

Quantitative assessment of error associated with animal positioning in the cephalometer is beyond the scope of this study; however, a subjective determination was possible by inspection of implant peak configurations (Fig. 6). Implant peaks from different scans of the same animal that were not di-rectly superimposable were discarded. In this project, more than 75% of the implant peaks were congruent. The inherent error of the Corning densitometer in scanning the same film repeatedly, without removing the film between scans, was less than ± 0.5% (P < 0.05) for implant peak area, peak height, and implant-osteotomy distance, and less than ±0.2% for implant-implant distance. An estimate of film positioning error, with removal of the film between each scan, again demonstrated high reproducibility (~± 1.0% for implant peak area, peak height, and implant-osteotomy distance, and less than ±0.3% for implant-implant distance.) The reproducibility of measuring osteotomy site area on the densitometric plot was high (1.5% for very small areas to 0.44% for large areas.) Measurement of the optical density of the middle step of the wedge using a static densitometer for all films processed at the same time provided an estimate of the variability in photon energy of the primary

984

PHOTODENSITIOMETRIC EVALUATION OF OSSEOUS REPAIR

beam, scatter radiation, and film quality (± 6.0%). For films processed at different times the error increased to ±9.0%. Discussion

of the osteotomy site occur after stability is achieved-"; nonetheless, the present work demonstrates that subtle differences in density (not apparent visually) are detectible photometrically and provide a useful index of the rate at which osseous repair is occurring throughout the healing period.

EVALUATION OF OSSEOUS REPAIR RADIOGRAPHIC ASSESSMENT

The time required for bony stability to occur was approximately 46 days for Group I and 49 days for Group II. In previous studies utilizing the same technique, Nanda and Sugawara'? reported stability at six to eight weeks postoperatively while Bell'! reported stability at four weeks. Because Bell does not give specific criteria for stability, comparison with this study is difficult; however, if 1.0 mm or less of segment mobility was defined as a stable maxilla, then the time required for clinical stability to occur in the present study is in accord with Bell's data. Differences in the time required for clinical stability to occur between Group I and Group II were small and statistically not significant. However, the maxillary mobility of Group II (Le Fort I advancement plus impaction) was consistently greater than that of Group I during the two-month postoperative interval (Fig. 7). This finding supports the idea that different surgical movements may require different periods of time before achieving stability" and that different durations of MMF may be appropriate depending on the specific movements involved. The lack of a statistically significant difference in clinical mobility between the surgical groups may reflect both the small sample size and the semiquantitative nature of clinical mobility estimates. Bone formation occurs on the endosteal and periosteal surfaces of the osteotomy site two weeks after Le Fort I advancement surgery in Macaca mulattaP At four weeks, young bone and osteoid is present throughout the osteotomy area. At six weeks mature bone and osteoid are present with new bone bridging the defect in some areas. Depending on how it is defined, clinical stability may result either from dense fibrous connective tissue and foci of new bone (four weeks postoperatively) or from bridging of bone across the surgical defect (six weeks postoperatively). In the present study the greatest change in mean mobility for Groups I and II occurred from three to seven weeks after surgery (Fig. 7). This almost certainly represents a continuum of biologic events beginning with the formation of dense fibrous connective tissue across the osteotomy gap, formation of foci of immature bone, and, finally, bridging of the osteotomy site with new bone. Regardless of how clinical stability is defined, complete mineralization and repair

Simple visual inspection of radiographs was not helpful in evaluating the size of the surgical defect, the rate of osseous repair, or the time when clinical stability would be achieved; nor could it be used for confirming when bone mass at the osteotomy site was the same as that of adjacent bone. This result is consistent with other studies. 16 - 18 In contrast, photodensitometric methods were useful in quantitating the relative size of the surgical defect, the net rate of remineralization, and the difference in photodensity between the osteotomy site and the adjacent bone at the time of clinical stability and throughout the postoperative period (Fig. 9). To quantitate osteotomy site repair from lateral cephalometric films, a constant relationship between subject tissue mass and film density over the range of measured optical densities is necessary regardless of minor fluctuations in exposure and film processing. This was verified by plotting optical density versus step wedge thickness (Fig. 8), and it established the relative difference in photodensity between the osteotomy site and the adjacent bone as being indicative of differences in BMC. A plot of the difference in photodensity (expressed as osteotomy site area) versus time (Fig. 9) provided information regarding the relative size of the surgical defect (y-intercept), and the net rate of rernineralization of the osteotomy site (slope). If the net rate of remineralization of an osteotomy or fracture is related only to the size of the surgical defect, one might expect a difference in the time required to achieve bony stability and complete repair. However, this does not seem to be the case. The time at which clinical stability occurred in both groups was similar though a large difference existed in the size of the defect at the time of surgery. For this to occur the net rate of rernineralization of the osteotomy site for Group II had to be significantly higher than for Group I (Fig. 9). Also, the relative difference in bone mass between the osteotomy site and adjacent bone (y-coordinate) at the time of clinical stability was similar despite differences in the size of surgical defect and rates of remineralization. This may be an important parameter in predicting when clinical stability will occur and could be useful in objectively determining when MMF should be discontinued.

985

SCHANTZ ET AL.

The net rate of remineralization of the osteotomy sites (slope of osteotomy photodensity versus time) followed a first-order (linear) relationship from postoperative weeks I to 10 for Group II and weeks I to 13 for Group I. Subsequently, the rate gradually diminished as the difference in photodensity between the osteotomy site and adjacent bone approached zero. This relationship is similar to that reported by Hellewell et al." for rooster metatarsus. Correlation coefficients for photodensity versus time for individual animals generally varied from - 0.88 to - 0.96. When the monkeys were grouped according to surgical procedure , the correlation coefficients decreased slightly (r = - 0.70 to - 0.84) because of variability in the healing response between the animals. REPRODUCIBILITY

In studies designed to simulate clinical situations, the error of experimental measurement is increased by variations in biologic response. This error is compounded when experimental conditions (for example, magnitude of surgical movement) vary or when subject positioning error prevents measurement of the same area of interest before and after the experimental treatment. In this study, the same surgeon performed all the osteotomies in an attempt to limit inadvertent treatment variation among animals of the same group. Subject positioning error was minimized during radiographic procedures by having the same operator position the monkeys in the same cephalostat throughout the study. Moreover, because photodensitometric methods have the capability of detecting changes in subject position (due to changes in peak configuration of stable anatomic structures), a high level of confidence existed regarding the precision of animal positioning during radiographic procedures when configurations of densitometric plots were congruent. A decrease in positioning error would be anticipated for human subjects, who are both conscious and cooperative, in contrast to the sedated monkeys used in this study. Therefore, the use of standardized cephalometric radiography plus verification from photo densitometric methods should be adequate for monitoring healing of craniofacial osteotomies and fractures in a human population. Because optical density values were within the linear range of the Hurter-Driffield curve for the films used in this study, differences in film density reflected real differences in BMC. Although variation in film density from radiation exposure and film processing were not critical in this project, irradiation of the aluminum step wedge allowed com-

pari sons to be made between films processed at the same time (radiation exposure error) and those processed at different times (film exposure and development error). For films processed at the same time and under the same conditions, the variation of photodensity of the aluminum wedge was 6%. This increased to 9% for films processed at different weekly intervals and is consistent with other studies. 19 - 22 Thus, at least an additional 6% error due to variability in film exposure and processing would be expected if actual value s for BMC were needed. Conclusions 1) Progressive mineralization of osteotomy sites was readily quantified by means of photodensitometric methods. Plots of osteotomy site radiodensity versus time yielded regression lines (r ~ 0.9) whose negative slope reflected the net rate of osteotomy site remineralization . 2) Optical densities of film used in evaluating the relative BMC of osteotomy sites and adjacent bone were within the linear range of the Hurter-Driffield curve and thus are suitable for quantitation by photodensitometry. 3) Densitometer-based instrumentation and measurement error were less than 3% for evaluating bone mineral changes at osteotomy sites. 4) Visual appearance of radiographs could not be correlated with the onset of clinical stability or with the progressive changes in bone mineral mass at osteotomy sites. 5) The amount of time necessary for clinical stability to occur was not statistically different between the surgical groups despite a large difference in the size of the surgical defect at day zero (Group II > Group I). However, this finding may be a reflection of the small sample size. 6) The net rate of remineralization of osteotomy sites was significantly different between the groups. Group II (impaction + advancement) had a higher rate of remineralization than Group I (advancement only). 7) The relative difference in film absorbance between the osteotomy site and adjacent bone at the time of clinical stability was essentially the same between the two groups and may be indicative of the time when stability will be achieved.

References I. Siegal 1M. Anast GT. Fields T: The determination of frac-

ture healing by measurement of sound velocity across the fracture site. Surg Gynecol Obstet, September 1958. pp

327-332 2. Gerlanc M. Haddad D. Hyatt GW. et al: Ultrasonic study of

986

3. 4. 5.

6.

7. 8.

9.

10. II.

PHOTODENSITIOMETRIC EVALUATION OF OSSEOUS REPAIR

normal and fractured bone. Clin Orthop Rei Res 111 :175, 1975 Sonstegard DA, Matthews LS: Sonic diagnosis of bone fracture healing; a preliminary study. J Biochem 9:689, 1976 Nicholl s PJ, Berg E, Bliven FE, ct al: X-ray diagnosis of healing fractures in rabbits. Clin Orthop ReI Res 142:234, 1979 Matthews LS, Kaufer H, Sonstegard DA: Manual sensing of fracture stability: a biomechanical study. Acta Orthop Scand 45:373, 1974 Hellewell AB, Beljan JR: A new in vivo method of quantitating the rate and magnitude of mineralization of an experimental osseous defect. Invest Rad 12:467, 1977 Bjork A: Facial growth in man studied with the aid of metallic implants. Acta Odont 13:9, 1955 . Nanda R, Topazian RG: Craniofacial growth following Le Fort I osteotomy in adolescent monkeys, ill McNamara JA Jr, Carlson DS, Ribbens KA (cds): The Effect of Surgical Intervention on Craniofacial Growth. Monograph No. 12, Craniofacial Growth Series. Ann Arbor, Center for Human Growth and Development, 1982, pp 99-129 McNamara JA Jr: Neuromuscular and skeletal adaptations to altered orofacial function. Monograph No. I, Craniofacial Growth Series. Ann Arbor, Center for Human Growth and Development, University of Michigan, 1972 McNamara JA Jr, Riolo ML, Enlow DH: Growth of the maxillary complex in the rhesus monkey (Macaca 11I1/latta.) Am J Phys Anthrop 44:15, 1976 Nanda R, Legan H: Craniofacial adaptations after total

12. 13. 14. IS. 16. 17. 18. 19. 20. 21. 22.

maxillary osteotomy in Macaca irus: a cephalometric and histologic study. Am J Orthod 73:410, 1978 Nanda R: Protraction of the maxilla in rhesus monkeys by controlled extraoral forces. Am J Orthod 74:121, 1978 Nanda R, Sugawara J: Mandibular adaptations following total maxillary osteotomy in adolescent monkeys. Am J Orthod 83:391, 1983 Bell WH, Fonseca RJ, Kennedy JW, et al: Bone healing and revascularization after total maxillary osteotomy. J Oral Surg 33:253, 1975 Sevitt S: Bone Repair and Fracture Healing in Man (Current Problems in Orthopaedics). New York, Churchill Livingstone, 1981, p 300 Mooney V, Nichel VL , Harvey JP Jr, et al: Cast brace treatment for fracture of the distal femur. J Bone Joint Surg 52A:1563,1970 Worth HM: Principles and Practice of Oral Radiologic Interpretation. Chicago, Yearbook, 1972, pp 383-419 Kappel DA, Craft PO, Robinson DW, et al: The significance of persistent radiolucency of mandibular fractures. Plast Reconstr Surg 53:38, 1974 Mack PB, Brown WN, Trapp HD: The quantitative evaluation of bone density. Am J Roentgen Rad Ther 61:808, 1949 Griffith ER, Stonebridge 18, Lehman JF : Current methods of in-vivo measurement of osteoporosis. Am J Phys Med 52:75, 1973 Meema HE: Non-invasive measurements of bone loss in peripheral bones. Clin Invest Med 5:157, 1982 Tumland J, Margen S: Use of a spectrophotometer for radiographic photodensitometry of bone. Calcif Tissue Int 28:1, 1979