Avascular necrosis of the mandibular ramus in the baboon (Papio ursinus) following orthognathic surgery

Int. J. Oral Surg. 1980: 9: 206-215 (Key words: osteotomy; necrosis; surgery, orthocnat/,ic; monkeY)

Avascular necrosis of the mandibular ramus in the baboon (Papio ursinus) following orthognathic surgery J. F. LOWNIE, P. E. CLEATON-JONES, J. C. AUSTIN, E. G. VIEIRA AND M. REITZIK MRCIUniversity oj the Witwatersrand Dental Research Institute, Johannesburg, South Africa

The maintenance of an adequate blood supply to the mandible following orthognatmc surgery is of extreme importance if satisfactory healing is to be achieved. With this object in mind, this study was undertaken to assess the effects of disrupting the blood supply of the baboon mandible by means of L- and C-osteotoIIlies. Ten adolescent baboons of the species Papio ursintls were used as the experimental animals for the study. On five of them a modified vertical subsigmoid osteotomy (L-osteotomy) lI."Ad on five a C-osteotomy of the mandibular ramus were performed. In each experimental animal the osteotomy was undertaken on one side of the mandible only, the other side serving as a control. After a postoperative period of 14 days the animals were killed, the entire mandible: removed, and the tissues prepared for examination with the light microscope. Histometric analysis revealed a significant degree of avascular necrosis in certain areas of the proximal or condylar fragments and to a lesser extent in the distal fragment. Avascular necrosis was also seen on the control side. It was also revealed that the periosteum and attached muscles of mastication play a role in the blood supply of the mandible and that the additional attachment of the temporalis muscle to the proximal fragment is of no advantage in enhancing the vascularity of that fragment.

.ABSTRACT -

(Received for publication 28 April, accepted 10 July 1979)

The vasculature of the mandible has been studied in man 6 ,7, the cat7,8, the dog7 and the monkey7 using perfusion, radiological and histological techniques6 , 7, 8. Our assessment of the blood supply of the human mandible based on these studies is shown in diagrammatic form in Fig. 1. The arterial supply to the mandible is threefold; firstly there is a supply from the

insertions of the muscles of mastication, secondly from the temporomandibular joint capsule and thirdly from the inferior dental artery-periosteal vessel anastomosis. During skeletal surgery to the ramus of the mandible, the masseter muscles, together with attached periosteum, are stripped from the lateral aspect of the underlying bone. This is accompanied by the partial de-

0300-9785/80/030206-10$02.50/0 @ 1980 Munksgaard, Copenhagen

..

_. .~

AVASCULAR NECROSIS OF RAMUS

.

c:::J .

Fig. 1. Arterial supply to the mandible. (1) Stippled area denotes supply from the insertion of the muscles of mastication. (2) Solid black line is the supply from the temporomandibular joint capsule. (3) Clear area is supplied by the inferior dental artery - periosteal vessel anastomosis.

tachment of the medial pterygoid muscle from the medial aspect of the mandibular ramus. Linking the surgical technique and the probable vasculature of the mandible, it can be postulated that in the L-osteotomy the proximal fragment would receive a blood supply only from the lateral pterygoid muscle and the capsule of the temporomandibular joint, the supply from the periosteum and inferior dental arteries being eliminated by the operative procedures. Stripping of the medial pterygoid and masseter muscles from the mandible removes this blood supply from the underlying proximal fragment and furthermore, the procedure would also result in disruption of the periosteal supply to the distal fragment. In the C-osteotomy, however, the temporalis muscle would be an additional source of vascular supply. The rate of bone healing and revaSClllarization of bone following vertical ramus osteotomies has been investigated by BELL & KENNEDYs in the rhesus monkey, using perfusion techniques. It was demonstrated that in the specimens where no soft tissue pedicle was left intact, vascular ischaemia,. intra-osseus necrosis and delayed healing resulted. Where the proximal fragment was left pedicled to soft tissues early

207

osseus union, minimal osteonecrosis and minimum vascular ischaemia were found. It was concludeds that the viability of the proximal fragment was preserved when it was pedicled to the lateral pterygoid muscle and associated temporomandibular joint capsule. This study was a perfusion study without any histometric investigation into avascular necrosis. Vascular changes after orthognathic surgery in the mandible and maxilla have also been investigated by MEYER & CAVANAUGHI0. In their study subapical osteotomies were performed in the mandibles and maxillae of five dogs and six monkeys of unnamed species. By the use of an iscr tope fractionation technique using saRb the blood flow in these fragments was measured. There was an average decrease in blood flow to the alveolar bone ranging from 48 to 74%, as well as a decrease in the blood flow to the dental pulp from 54 to 82 % and 18 to 40 % in mucosal tissues following subapical osteotomies. The vascular response to the saggital split osteotomy of the mandibular ramus has been studied 9 • Sagittal split osteotomies were performed on one side of the mandible in 11 adult Macaca rhesus monkeys. By the use of radioactive Rubidium (SORb) the blood flow on the operated side was compared with that of the non-operated side immediately after operation. The results of this study showed that reduction in blood flow to the proximal fragment of the mandibular ramus below the horizontal bone cut was sufficient to cause avascular necrosis based on a decrease in the fractional uptake of BORb. The aims of this investigation were: (1) to determine whether or not avascular necrosis was present following L- and C-osteotomies in the baboon mandible (2) to quantitate the amount of avascular necrosis (3) to compare the amount of avascular

208

LOWNIE, CLEATON-JONES, AUSTIN, VIEIRA AND REITZIK

necrosis following the L- and C-osteotomies, to see whether the additional blood supply from the temporalis muscle (in the C-osteotomy) lessened the degree of avascular necrosis.

Material and methods Ten Chacma baboons (Papio ursinus) were used in the study. The animals were immobilized with intramuscular phencyclidine hydrochloride (1 mg/kg body weight), after which they were anaesthetized to effect by means of intravenous pentobarbitone sodium. A No. 4 Magill cuffed endotracheal tube was inserted orally and the baboons allowed to breathe room air. The submandibular area was shaved, swabbed and draped with a sterile towel. Using standard aseptic operative techniques, the mandible on each side was approached by means of a submandibular incision. The periosteum en,veloping the mandible was incised along the lower border of the bone and the lateral aspect of the ramus of the mandible exposed by reflecting the overlying masseter muscle. The L-osteotomy, a modification of the vertical subsigrnoid osteotomy, was performed on the first group of five animals in the left mandibular ramus area. In the second group of five animals the C-osteotomy was undertaken. The opposite ramus served as a control in both types of osteotomies. Here no bone cuts were made, but complete reflection of the masseter and medial pterygoid muscles from the surface of the mandibular ramus and angle was accomplished. In order to prevent ingrowth of fibrous tissue and to eliminate reattachment of overlying muscles, a 0.025 mm thick silastic sheeting envelope was constructed, and was placed over both the ascending mandibular rami from the sigmoid notch to lower border of the mandible to cover the surface of the ramus from anterior to posterior border. In both groups of animals after the immobilization of the segments had been accomplished, the soft tissues were re-apposed by interrupted subcutaneous sutures and final skin closure achieved by means of interrupted horizontal mattress sutures. Immediately after operation 1 ml of longacting antibiotic mixture containing benethamine and procaine penicillin was administered

Fig. 2. Five blocks excised for histology (a-e).

to each experimental animal by intramuscular injection. Each baboon was housed in a separate steel cage and, for the full 14-day postoperative period, was fed three times daily. Fresh fruit and drinking water were supplied as well as 2 ml vitamin syrup daily orally. On the 15th postoperative day, each experimental animal was immobilized by means of an intramuscular injection of phencyclidine hy~ drochloride. The baboons were then killed by administering an overdose of intravenous pentobarbitone, after which they were clinically examined for any signs of postoperative infection related to the operation site. Histological preoperation of the specimens After 2 weeks' fixation in several changes of fixative, the specimens were decalcified in 18 % E.D.T.A. (ethylene-diamine-tetra-acetic acid) at 60°C in a shaking water bath for a period of 3-4 weeks monitored by radiographic examination to determine when decalcification was complete, at the end of which the specimens were trimmed in the following manner: The tooth·bearing area O'f each hemi-mandible was separated from the ramus distal to the follicle of the developing third molar tooth. This tooth-bearing area was discarded. The entire mandibular ramus of both the experimental and control sides were retained for histological evaluation and were divided into five blocks labelled a-e in the horizontal plane as shown in Fig. 2. Each block was given a random code number, dehydrated, embedded in paraffin wax, and the blocks labelled a, C, e were cut on a rotary microtome at 6 .urn in a plane parallel to the lower border of the man-

AVASCULAR NECROSIS OF RAMUS

dible. Step-serial sections every 60 pm were stained with haematoxyline and eosin. The blocks labelled band d were retained for fuM ture reference. Histometric analysis Histometric analysis was carried out on blocks a, c, e on both the experimental and control sides in the following way. Step-serial sections every 60 ,urn were exam· ined with the aid of a calibrated graticule of 1.2 mm2 containing 100 equally sized squares of 0.012 mm2 each. Each histological section was examined by placing the graticule over the areas 1-8, as indicated in the orientation diagram (Fig. 3). Five sections per specimen were examined wherever possible. The edge of the graticule was aligned with either the cut surface of the bone in the areas of the osteotomy cut (areas 1-4) or the outer and inner surfaces· of the cortical place of the mandible in areas 5-8. As no bone cuts were made on the control side of each animal, the placing of the graticule in areas 1-4 was done as accurately as possible to reproduce the site examined on the osteotomy side. The cell count was carried out at random in one of the 0.012 mm2 squares in all eight positions of the histological section. Empty lacunae were taken to indicate cell necrosis. The total number of lacunae, as well as the number of empty lacunae in each numbered area of each slide were recorded, transferred to punch cards and analysed in a computer using the Statistical Package for the Social Sciencesl l . The difference in cellular necrosis between the control and operation sides for each section examined in each area (1-8) or in combination of areas was calculated and a paired

Fig. 3. Orientation of the sections examined. Areas 1-8 are those in which avascular necrosis was assessed.

209

Student's t-test carried out using the following formulae: empty lacunae 1. % empty lacunae (X) = -t:-7-;'-171--ot a acunae X 100 2. Xc-Xop = Xdill 3. t= X _ diflVN S.d·dill Where Xc = X dead cells in control site

X ov = X dead cells in operative site xdiff = mean difference in % dead cells between control and operative sites N = number of sections examined S.d'dill = standard deviation of differences in % dead cells between control and operative sites. The level of statistical significance chosen was P < 0.01. It proved impossible to carry out a check on intraexaminer variation in quantitating cell deaths. This was because it was not possible to place a grid over exactly the same area as was previously measured.

Results In none of the animals was there any clinical evidence of enlargement of either the submandibular or jugulodigastric lymph nodes. During the subsequent preparation of the specimens, which was undertaken with great care so as not to disturb any tissues, no further evidence of infection as judged by the presence of pus was detected. In all cases the siIastic barrier had not been displaced, but still enveloped the operation site as had been intended on both the osteotomy and control sites. A clinical examination of the mandibular rami showed no evidence of bony union. A prominent feature in all the rami on which either the L-osteotomy or the C-osteotomy had been performed, was the presence of a diffuse, white appearance of an area of cortical bone approximately 5 mm in size, which involved the distal-most tip of the proximal osteotomy segment. This was

210

LOWNIE, CLEATON-JONES, AUSTIN, VIEIRA AND REITZIK

Fig. 4. Specimen from the distal-most tip of the proximal fragment showing the appearance seen after both types of osteotomy. Note that almost all the lacunae are empty. Hand E X 480.

in contrast to the shiny appearance of the cortical bone in the equivalent area on the control side. Histological examination of the specimens from the mandibular rami on which osteotomies had been performed revealed two pieces of bone, namely the proximal and distal fragments, separated by granulation tissue composed of immature fibroblasts, indicated by the absence of both fully formed processes and collagen fibres. Further examination of both bony fragments revealed cortical and medullary bone arranged in typical Haversian systems. However, high power microscopic examination of sections of the proximal fragment starting from the condylar region and progressing down the ramus towards the distal-most tip of the proximal fragment revealed pyknosis of the nuclei of the osteocytes. This was minimal in the condylar region, but increased gradually with increasing distance from the condylar blood supply. In the distal-most tip of the proximal fragment most of the lacunae were empty, which was taken to indicate cellular death (Fig. 4). The adjacent area of the distal fragment in each section revealed cellular destruction to

Fig. 5. The corresponding area of the distal

fragment to that in Fig. 4 is shown. Note that there are few empty lacunae. Hand E X 480.

a lesser extent than that in the proximal fragment (Fig. 5). The histological examination of the control sides where no bone cuts had been made revealed pyknosis of few of the osteocyte nuclei. The histometric analysis results are presented in Tables 1-5, arranged according to the orientation diagram in Fig. 3. The mean difference of dead cells denoted by x was evaluated in all eight areas in each slide. A -ve sign indicates a greater degree of necrosis on the operative side. A + ve sign shows more necrosis on the control side. In all cases N varies because some sections had to be discarded because of folding in the area to be examined. (1) Cell death in all areas examined In the L-osteotomy specimens, the amount of cellular death in all eight areas examined was significantly greater on the operative side than on the control side (P < 0.001). This is indicated by the minus values in Table 1. Cell death was present in both proximal and distal fragments but was approximately two times greater in the four areas of the proximal fragment (3, 4, 5, 6) than those in the distal fragment (1, 2, 7, 8) (Table 1).

AVASCULAR NECROSIS OF RAMUS

211

Table 1. Cell death following L- and C':osteotomies of the mandibular ramus. The number of specimens examined and the results obtained in each area are shown C-Osteotomy

L-Osteotomy 52 specimens (± s.d.)

II

II ""

Area Proximal fragment anteromedial (No.3) anterolateral (No.4) posteromedial (No.5) posterolateral (No.6) Distal fragment anteromedial (No.7) anterolateral (No.8) posteromedial (No.1) posterolateral (No.2)

mean

mean

= 65 specimens (± s.d.)

-28.2 -29.1 -25.3 -26.9

35.2 34.9 32.0 33.7

-5.8"" -6.0"" -6.3** -5.8**

-36.8 -27.7 -26.5 -25.3

36.3 34.2 33.6 35.3

-5.9** -6.5** -6.3** -5.8**

-7.4 -12.0 -8.3 -12.0

18.9 24.6 18.1 23.9

-2.8** -3.5"" -3.3"" -3.6**

-3.0 -4.0 -3.4 -3.3

21.9 21.8 23.9 15.7

-1.1 -1.6 -1.1

-1.7

A -ve sign indicates a greater degree of necrosis on the operative side. * p < 0.01 ** P < 0.001

In the C-osteotomy specimens, no significant differences in amount of cell death were seen between the control and operative sides in the distal fragment. In contrast, significantly more cell death was seen on the operative side of the proximal fragment than on the control side. There was approximately five times more cell death in the proximal fragment than in the distal fragment (Table 1).

(2) Cell death adjacent to osteotomy cuts Table 2 indicates the amount of cell death seen following the L- and C-osteotomies in the areas adjacent to the osteotomy cuts. Within the combined areas studied, ie. 1 + 2 and 3 + 4, there was significantly more cell death on the operative side compared to the control side (P < 0.001) except for the area adjacent to the osteotomy cut in the distal fragment of the C-osteotomy.

Table 2. A comparison of cell death in the proximal and distal fragments of the L- and C-osteotomies in areas adjacent to the osteotomy cuts Area of mandible Proximal fragment Areas 3 + 4 (adjacent to osteotomy cut) Distal fragment Areas 1 + 2 (adjacent to osteotomy cut) .... P

< 0.001

L-Osteotomy n = 52 specimens

x

s.d. t

diff

x diff s.d. diff t

= -28.89 = = = =

33.79 -6.16**

":'10.3 19.8 -3.7··

n

C-Osteotomy specimens

= 65

x diff s.d. diff t

= =

-27.3 34.3 -6.42**

x diff = -2.4 s.d. diff = 21.1 t -0.9

L-Osteotomy vs C-Osteotomy L>C t = 0.25 N.8. L>C t = 2.05 P < 0.02

LOWNIE, CLEATON-JONES, AUSTIN, VIEIRA AND REITZIK

212

Table 3. A comparison of cell death in areas of the proximal and distal fragments of the L- and C-osteotomies in the region of anterior and posterior borders of the mandibular ramus L-Osteotomy vs C-Osteotomy

Area of mandible

L·Osteotomy n = 52 specimens

C-Osteotomy n = 65 specimens

Proximal fragment Areas 5 + 6 (posterior border of ramus)

x diff = -25.8 s.d. diff = 31.6 t = -5.87**

x diff = -25.8 s.d. diff = 33.8 t = -6.2*'"

L=C

Distal fragment Areas 7 + 8 (anterior border of ramus)

X diff = -9.8 s.d. diff = 19.9 t -3.6*'"

.i: diff = -3.6 s.d. diff = 20.2 t -1.4

L>C t = 1.65 0.05 P

**

P

<

< 0.001

In the L-osteotomy the amount of cell death in the proximal fragment was double that of the distal fragment. In the C-osteotomy the cell death in the proximal fragment was seven times greater than that in the distal segment. There was no significant difference in cell death between the two types of osteotomies in the proximal and distal fragments (i.e. P 0.01).

<

(3) Cell death at anterior and posterior borders of the ramus Table 3 indicates the amount of cell death following both L- and C-osteotomies at the

anterior and posterior borders of the mandibular ramus. Within the combined areas studied, i.e. 5 + 6 and 7 8, there was significantly more cell death on the operative side compared to the control side (P 0.001) except for the anterior border of the ramus in the distal fragment of the C-osteotomy. In the L-osteotomy the amount of cell death in the posterior border region, viz. proximal segment is almost double that seen at the anterior border, viz. distal fragment. In the C-osteotomy the amount of cell death in the posterior border region is

+

<

Table 4. A comparison of cell death on the medial aspect of the mandibular ramus in the proximal and distal fragments of both the L- and C-osteotomies Area of mandible

L-Osteotomy n = 52 specimens

C-Osteotomy n = 65 specimens

Proximal fragment Areas 4 + 6 (medial aspect of proximal fragment)

x diff = -28.0 s.d. diff = 33.1 t = -6.0**

.t diff s.d. diff t

Distal fragment Areas 2 + 8 (medial aspect of distal fragment)

.i diff = -12.3 s.d. diff = 20.9 t = -4.2**

x diff = s.d. diff =

**

p

< 0.001

* P < 0.01

t

= -26.5 = 33.4

L-Osteotomy vs C-Osteotomy

-6.4**

L>C t = 0.24 N.S.

-2.8 19.7 -1.1

L>C t = 2.54 P < 0.01*

AVASCULAR NECROSIS OF RAMUS

213

Table 5. A comparison of cell death on the lateral aspect of the mandibular ramus in the proxi· mal and distal fragments of both the L- and C-osteotomies Area of mandible Proximal fragment Areas 3 + 5 (lateral aspect of proximal fragment) Distal fragment Areas 1 + 7 (lateral aspect of distal fragment) ** P

L-Osteotomy n = 52 specimens

C-Osteotomy n = 65 specimens

x diff = -26.7 s.d. diff = 32.3 -6.0** t

x diff = -26.7 s.d. diff = 33.1

x diff = -7.7 s.d. diff = 17.1 -3.2** t

oX diff = -3.3 s.d. diff = 20.7 -1.3 t

t

-6.0 H

L-Osteotomy vs C-Osteotomy

L=C N.S. L>C t = 0.81 N.S.

< 0.001

about five times greater than that in the region of the anterior border. Once again the amount of cell death in the distal fragment of the L-osteotomy is twice that of the Costeotomy, the proximal fragments being similar. No statistically significant differences were found between the L- and Costeotomies. (4) Cell death at the medial aspect of the ramus The amount of cell death following both Land C-osteotomies on the medial aspect of the ramus in both proximal and distal fragments is shown in Table 4. Within the combined area studied, i.e. 2 + 8 and 4 + 6, there was significantly more cell death on the operative side compared to the control side (P< 0.001) except for the medial aspect of the distal fragment in the C-osteotomy. In the L-osteotomy the amount of cell death in the medial aspect of the proximal fragment is slightly higher than that in the distal fragment. In the C-osteotomy however, cell death in the medial aspect of the proximal fragment is six times greater than that in the distal fragment, although the amount of cell death in the proximal frag-

ments of both osteotomies is similar. The amount of cell death in comparing the distal fragments of the two types is four times greater in the L-osteotomy than the C-osteotomy. This difference was significant at the 0.1 % level. (5) Cell death at the lateral aspect of the ramus In Table 5 is listed the amount of cell death following both the L- and Costeotomies on the lateral aspect of the mandible in both the proximal and distal fragments. Within the combined areas studied, i.e. 1 + 7 and 3 + 5, there was significantly more cell death on the operative side compared to the control side (P < 0.001) except for the lateral aspect of the distal fragment in the C-osteotomy. In the L-osteotomy the amount of cell death in the proximal fragment is almost double that of the distal fragment. In the C-osteotomy, however, the amount of cell death in the proximal fragment is four to five times greater than that of the distal fragment, although the amount of cell death in the proximal fragments of both types of osteotomies is equal. In the distal fragments the cell death in the L-osteotomy

214

LOWNIE, CLEATON-JONES, AUSTIN, VIEIRA AND REITZIK

is almost three times greater than the Costeotomy. This was not statistically significant, however.

Discussion In previous studies in which the vascular response to orthognathic surgery has been investigated in both the mandible and the maxilla, the rhesus monkey (Macaca mulatta) was selected as the experimental nonhuman primate l , 2, 4, 5, 9, 10. This species was chosen because of musculoskeletal architecture similar to that of man. The equivalent non-human primate to the rhesus monkey in South Africa is the vervet monkey (Cercopithecus aethiops). This is, however, a small laboratory animal. For the present study, in order to obtain good immobilization of the bony fragments postoperatively, and also to obtain adequately sized histological sections, it was thought better to select a non-human primate with a more robust mandible, still exhibiting musculoskeletal features similar to those of the human mandible. To fulfil both these criteria, the Chacma baboon (Papio ursinus) was selected. It is clear from the previous studies that the arterial blood supply to the mandible is not simply via two inferior dental arteries, but via complex arterial and capillary anastomoses derived from the inferior dental arteries and vessels supplying muscles of mastication, temporomandibular joints and overlying periosteum. Any interference with one of the above structures, either surgically or pathologically, could adversely affect the vascular supply of a part of the mandible. It has been demonstrated 5 that the revascularisation of devitalized bone from the overlying tissues occurs during a period of days and not hours postoperatively. Thus it was postulated that an inadequate blood supply im-

mediately postoperatively would remain in· adequate for a long enough period to produce avascular necrosis of part of an osteotomy fragment. From the present investigation it has been proved that there is evidence of necrosis of osteocytes in the proximal and distal osteotomy fragments of both the L-osteotomy and the C-osteotomy. Furthermore, when the results of the hig,. tometric analysis following both types of osteotomies are compared, there is no evidence to suggest that the additional blood supply to the proximal fragment in the Costeotomy due to the temporalis attachment is advantageous. The additional fact that a lesser degree of osteocyte necrosis was present in the distal fragment of both the osteotomy and control sides is sufficient evidence to prove that there is a small but significant vascular supply to the mandible from the attached muscles of mastication and overlYing periosteum, and that disruption of this supply by the operative technique results in cellular death within the mandible. An interesting observation is that the amount of cell death in the distal fragment of the L-osteotomy is significantly higher than that in the C-osteotomy. The only explanation that can account for this discrepancy is that there must be a small but significant blood supply to that area of the mandibular ramus from the capsule of the temporomandibular joint and lateral pterygoid muscle. This supply would be eliminated in the L-osteotomy and not in the Costeotomy. It may be argued that the cellular necrosis adjacent to the bone cuts is a direct result of operative trauma even in the presence of adequate cooling of the rotatory cutting instrument used. This is disproved, however, by the fact that the amount of cellular damage on the posterior border of the proximal fragment, an area far removed from any bone cut, is similar

A VASCULAR NECROSIS OF RAMUS to that in the area adjacent to the bone cuts. Finally, in all 10 animals examined, there was no evidence whatsoever of secondary infection as judged macroscopically by the absence of enlarged regional lymph nodes or presence of pus and microscopically by the complete absence of any chronic inflammatory cellular response. This fact is of paramount importance as the resulting osteocyte necrosis can only be a direct result of a vascular insufficiency resulting from the operative technique and can thus be described as avascular necrosis. CONCLUSIONS It can be concluded from this investigation that in the Chacma baboon (Papio ursinus) following orthognathic surgery to the ramus of the mandible that: (a) A significant degree of avascular necrosis in the proximal fragments of both the L- and C-osteotomies was present. (b) There was avascular necrosis present in the distal fragments of both types of osteotomies. (c) There was some avascular necrosis on the control sides where the masseter and medial pterygoid muscles were reflected and separated from the underlYing bone by a silastic envelope. (d) From the above facts, it was evident that the periosteum and attached muscles of mastication playa small but significant role in the blood supply of the mandible. (e) The additional attachment of the ternporalis muscle to the proximal fragment in the C-osteotomy was of no advantage in enhancing the vascularity to that fragment. Acknowledgments - This study was supported in part by an Elida-Gibbs grant. The authors would like to thank Mrs. H. BALL for typing the manuscript.

215

References 1. BELL, W. H.: Revascularization and bone healing after anterior maxillary osteotomy. A study using adult rhesus monkeys. J. Oral Surg. 1969: 27: 249-255. 2. BELL, W. H., FONESCA, R. J., KENNEDY, J. W. & LEVY, B. M.: Bone healing and revascularization after total maxillary osteotomy. J. Oral Surg. 1975: 33: 253-260. 3. BELL, W. H. & KENNEDY, J. W.: Biological basis for vertical ramus osteotomies - a study of bone healing and revascularization in adult rhesus monkeys. J. Oral Surg. 1976: 34: 215-224. 4. BELL, W. H. & LEVY, B. M.: Revascularization and bone healing after posterior maxillary osteotomy. J. Oral Surg. 1971: 29: 313-330. 5. BELL, W. H. & LEVY, B. M.: Revascularizadon and bone healing after maxillary corticotomies. J. Oral Surg. 1972: 30: 640648. 6. CASTEllI, W.: Vascular architecture of the human adult mandible. J. Dent. Res. 1962: 42: 786-792. 7. COHEN, L.: Methods of investigating the vascular architecture of the mandible. J. Dent. Res. 1959: 38: 920-931. 8. COHEN, L.: Further studies into the vascular architecture of the mandible. J. Dent. Res. 1960: 39: 936-946. 9. GRAMMER, F. C., MEYER, M. W. & RICHTER, K. J.: A radioisotope study of the vascular response to sagittal split osteotomy of the mandibular ramus. J. Oral Surg. 1974: 32: 578-582. 10. MEYER, M. W. & CAVANAUGH, G. D.: Blood flow changes after orthognathic surgery: maxillary and mandibular subapical osteotomy. J. Oral Surg. 1976: 34: 495-501. 11. NID, N. H., HULL, C. H., 8TErNBR'ENNER, K. & BENT, D. H.: Statistical Package for the Social Sciences. 2nd ed. McGraw-Hili, New York 1975.

Address: J. F. Lownie MRC/University of the Witwatersrand Dental Research Institute 1 Jan Smuts Avenue Johannesburg

2001

South Africa