.I Oral Maxillofac 49:5a5-593.
Surg
1991
Infection
After Rigid Internal Fixation of Mandibular Fractures: A Clinical and Radiologic
Study
TATEYUKI IIZUKA, MD, DDS,* CHRISTIAN LINDQVIST, MD, DDS, PHD,t DORRIT HALLIKAINEN, MD, PHD,$ AND PERTTI PAUKKU, MD* Of 214 patients with mandibular fractures treated following the AO/ASIF principle of rigid fixation, 13 patients (6.1%) developed postoperative infection. The yearly percentage decreased remarkably during the study period. When rigid fixation became a routine method, the infection rate was as low as the corresponding figure for cases treated with nonstable techniques and maxillomandibular fixation (MMF). All but one of the infected fractures were in the angular region of the mandible. Teeth had been extracted from the fracture line in 9 of the 10 dentulous patients. It was concluded that erroneous techniques had been used in almost all infected cases, because compression could not be achieved when there was an irregular fracture line, an atrophic edentulous mandible, or inadequate stability due to removal of a tooth in the line of fracture. In five patients, successful reoperation was performed using a reconstruction plate. Detailed radiologic examination was useful in assessment of infection and in follow-up.
Operative treatment of mandibular fractures ideally should include reestablishing the correct occlusion and immediate, uncompromized mandibular function. Unstable fixation methods, however, often require maxillomandibular fixation (MMF), which may delay functional rehabilitation. Therefore, rigid internal fixation allowing early mobilization has increased in popularity during the last decade. I-9 Some authors, however, have claimed that the use of rigid plates increases the frequency of complications such as infection, malocclusion, delayed union, nonunion, osteomyelitis, and injury to the inferior alveolar and facial nerves.3*4,10-13Although these complications can also be caused by
the trauma itself, they are often associated with the method of treatment. Postoperative infection can be the most problematic complication in certain cases. Infection rates of 3% to 27% have been reported when using bone plates; the corresponding rates being lower when using transosseous wiring.3*‘4-‘5 It is difficult to compare such rates, however, because the treatment methods (immobilization or mobilization) are totally different from each other. Rigid fixation in mandibular bone surgery has been used in our department since 1983. The purpose of this study was to analyze cases with mandibular fractures in which infection developed after rigid plate osteosynthesis to evaluate the frequency, causes, and diagnostic measures.
Received from Helsinki University Central Hospital, Finland. * Resident, Department of Oral and Maxillofacial Surgery. t Head, Department of Oral and Maxillofacial Surgery. $ Senior Radiologist, Department of Radiology. Address correspondence and reprint requests to Dr Iizuka: DeDartment of Oral and Maxillofacial Suraerv. Helsinki Universit; Central Hospital, Kasarmikatu 1I-13,-Sl%O130 Helsinki 13. Finland.
Patients and Methods
0 1991 geons 0276-2391
American
Association
of Oral
and Maxillofacial
Between 1983 and 1989, 1,823 patients with mandibular fractures were treated in the Department of Oral and Maxillofacial Surgery, Helsinki University Central Hospital, Helsinki, Finland. The annual percentage of open reduction and osteosynthesis of the mandibular fractures varied between 18% and 29% (mean, 25%). The main indication for surgical
Sur-
I91 /4906-0006$3.00/O
585
586
INFECTIONANDMANDIBULARFRACTLJRES
treatment was displacement and/or instability of a fracture in the symphysis, body, or angular region. Nondisplaced, stable fractures and condylar fractures were usually treated conservatively with maxillomandibular immobilization. The use of rigid tixation increased and that of transosseous wiring decreased during the 6-year period. From 1987 on, rigid fixation was the routine method of osteosynthesis (Fig 1). TREATMENT
WITH RIGID FIXATION
During the study period, a total of 214 patients (175 male, 39 female) with 270 mandibular fractures were treated with rigid internal fixation. The mean age of the patients was 34 years (range, 8 to 83 years). Distribution of fractures according to the location, surgical approach, and type of plate is given in Table 1. These numbers do not include cases with large defects or comminution, often due to gunshot injuries, which were excluded from the present analysis. The mean delay from injury to treatment was 3.1 days. The operation was performed following the AO/ASIF principles described by Spiessl.16 Suction drains were placed routinely for the first 24 postoperative hours. Prophylactic antibiotics were used in each case where the frac340 1 320-
ture involved tooth-bearing areas of the mandible. This treatment was continued for 7 to 10 days postoperatively. INFECTIONS
In 13 of 214 patients (6.1%), infection occurred during the postoperative treatment period (Fig 2, Table 2). Ten were male and 3 were female, from 23 to 49 years old, with a mean age of 35.7 years. There were 11 dentulous and 3 totally edentulous mandibles. Criteria for a diagnosis of infection included the presence of pain, swelling, erythema, and/or purulent extra- or intraoral discharge, along with leukocytosis and elevated erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP). Most of the 13 patients developed an abscess at the fracture site, with a sinus tract to the skin or mucosa of the oral cavity. All infections were solitary. Cases of osteitis caused by a tooth in the fracture site requiring endodontic treatment only were not classified as infections, as was simple intraoral wound dehiscence not leading to major complications. Data for all patients with infection were analyzed to determine contributing factors. TREATMENTOF
INFECTIONS
In all 13 cases, the infection required removal of the osteosynthesis material. Secondary treatment consisted of sequestrectomy and, in 5 patients, application of an AO/ASIF reconstruction plate with or without a bone graft from the iliac crest. Three of these patients were operated (two with cancellous bone grafts and one without) in the acute stage of the infection. Of the other eight cases where no reconstruction procedure was undertaken, one fracture required external fixation, and another MMF. In one case, the fracture was stabilized with a Kirschner wire. In five patients, the fracture was left to stabilize by itself without any fixation or immobilization (Table 3).
3002802602400' 220213
RADIOLOGIC EXAMINATION
81
82
a3
84
85
86
a7
00
89
YEAR
FIGURE 1. Total annual number of mandibular fractures and open reduction according to different fixation methods in the Department of Oral and Maxillofacial Surgery, Helsinki University Central Hospital, Finland, in the period of 1981 to 1989.~a--, Total; @ rigid plate; !&I,transosseous wiring and MMF; n , miniplate and MMF.
Routine radiologic evaluation of the fractures preoperatively and postoperatively included panoramic images, and Townes and axial projections of the skull. For detailed evaluation, a Scanora x-ray unit (Orion Corp/Soredex, Helsinki, Finland) was used. These examinations included a panoramic image followed by detailed imaging with a narrowbeam technique and cross-sectional tomography. The images were evaluated for loosening of screws, bone resorption under the osteosynthesis plate, and
587
IIZUKA ET AL
Table 1. Fmctura Sltos and mthcds of Fixation In 214 Patlonts With Mandlbulrr Fractures Tmatod Following the AO/ASlF Rinclple of Rigld Flxatlon Approach Mandibular Fracture
Osteosyntheses
Intraoral
No.of Fractures Extraoral
Symphysis
20
14
6
Canine
39
23
16
Body
90
16
74
Angle
121
5
116
Total
270
58
212
Abbreviations:
DCP. dynamic compression
plate: EDCP. excentric dynamic compression
altered bone structure suggestive of infection. During follow-up, the same type of examination was used to evaluate healing after infection and osseous union after bone grafting. Since 1988, all patients with postoperative infection (9 to 13 in Table 2) have been followed radiologically by the same radiologist, from the diagnosis of infection through the whole treatment phase. At secondary operation, the local findings were all documented by the same surgeon. Results
The yearly percentage of postoperative infection changed remarkably during the study period (Fig 2). It decreased from 10.9% in the period 1983 to 1986 to 4.8% in 1987 to 1988 and to 3.6% in 1989. Detailed data on the type of fracture, primary osteoSl
_. u)
50 -
E: .s! iii n
40 -
5
30-
p
20-
53
55
10 ‘O
2 .
om 83
84
85
86
87
88
89
YEAR
FIGURE 2. Yearly distribution of infection cases in relation to total number of patients with mandibular fractures treated with rigid fixation. W. Rigid fixation; E, infection.
Fixation
Operated On 16 4 31 4 3
DCP RP DCP EDCP RP Lag screw DCP EDCP RP Lag screw DCP EDCP RP Lag screw
I 66 5 18
1 80 16 22 3 270
plate; RP, reconstruction
plate.
synthesis, and treatment of the infection are shown in Tables 2 and 3. There were eight fractures situated in the angular region and four in the proximal part of the body. In one edentulous patient, infection occurred in the parasymphysis region. Teeth (second or third molars) had been extracted from the fracture line before the osteosynthesis procedure in 9 of the 10 dentulous mandibles. In all 13 patients, plates causing compression were used primarily. Infection was diagnosed after a mean period of 7.3 weeks. No correlation was found between delay of surgical treatment and rate of infection. In 2 patients, clinical signs of infection were observed preoperatively (Fig 3A-D). Bacteriologic analysis of the pus from the infected area showed Staphylococcus epidermidis in eight cases, Staphylococcus aureus in four, and Streptococcus viridans in one. In all cases, loosening of screws was observed during removal of the osteosynthesis material, irrespective of the stage of fracture healing. Nonunion of the fracture and bone sequestration was found in six cases. In five cases, there was soft-tissue bridging in the fracture line without macroscopically pathologic bone. In two patients, an osseous union had occurred across the fracture site. In all three cases where reconstruction was performed in the stage of acute infection, the postoperative course was uneventful. There was no need of MMF, and the mandible could be mobilized immediately after surgery. Good bone formation was observed radiologically within an average of 7 months (Fig 4A-D). It was not possible to predict an infection from the postoperative radiographs taken on the first or
588
INFECTION
AND MANDIBULAR
FRACTURES
Table 2. Data on 13 Patients With Infection After Treatment of Mandibular Fracture With Rigid Internal Fixation Patient No.
Age (yr)
Sex F F M M M M F M M M M M M
1 2 3 4 5 6 7 8 9 10 11 12 13
Delay From Accident to Operation (d)
Location of Fracture
8 1 3 5 3 3 1 0 13 2 4 5 2
Body Body Angle Angle Angle Angle Symphy sis Angle Angle Angle Angle Body Body
43 44 23 38 23 31 42 32 21 36 49 46 30 35.7
Mean
Plate Type*
Surgical Approach
Extraction of Tooth (tooth no.)
DCP6 DCP4 DCP6 EDCP6 DCP2 + DCP4 DCP2 + DCP4 DCP4 DCP4 DCP6 DCP2 + DCP4 DCP6 DCP6 DCP4
E E E E E E E E E E E E I
Edentulous Edentulous Yes (48) Yes (38) Yes (37) Yes (38) Edentulous Yes (47) Yes (48) Yes (48) Yes (47) Yes (37) No
3.8
Abbreviations: DCP, dynamic compression plate; E, extraoral; EDCP, excentric dynamic compression * Numeral indicates number of holes in plate.
second postoperative day in any case. Radiologically, a screw in the fracture line was noted in three cases. Bone resorption around the plate preceded other radiologic signs of infection. The earliest it was visible was 2 weeks postoperatively. Screw loosening was noted in two patients before other signs of infection and in three when clear signs of infection were visible. No radiologic signs of infection were noted earlier than 4 weeks after surgery. The radiologic diagnosis of loosening of screws and bone resorption under the osteosynthesis plate was confirmed at surgery in each case. Bone infection was also confirmed in all cases, but the area of infected bone was radiologically estimated to be larger than the area of macroscopically pathologic Table 3.
plate; I, intraoral.
bone documented at surgery. The radiologic followup correlated well with the clinical findings. Discussion
Various factors may contribute to failure with the use of plates. The higher incidence of infection when using bone plates has raised a concern that plates and screws, with their increased quantity of metal, might predispose to infection. Infections were treated successfully, however, using larger quantities of metal than was used in the primary operation. The results of research by the Swiss Association for the Study of Internal Fixation show that suscep-
Method of Treatment of Infection and Resulting Late Complications
Patient No. 1 2 3 4 5 6 7 8 9 10 11 12 13 Mean
Abbreviations:
Infection Diagnosed Postoperatively (wk) 25 1 7 1 3 4 9 9 1 20 4 4 7
Fixation of Fragments
Reconstruction of Mandible
No No No
RP MMF No External Kirschner wire + MMF RP No RP RP RP
Secondary (bone)
Primary (bone) Secondary (bone) Primary (bone) Primary
Follow-up Time (mo)
Slight instability Slight instability None None None None None None None None None None None
4 6 6 49 8 4 12 18 19 15 15 9 9 13
7.3 MMF, maxillomandibular
Late Complication
fixation; RP, reconstruction
plate.
IIZUKA ET AL
589
FIGURE 3. A, Radiograph showing right preangular fracture. There was a suppurative infection on admission. E. Fracture treated with a sixhole dynamic compression plate (DCP) in spite of infection and comminution. Tooth was removed from the fracture line before reduction and osteosynthesis. C (following page), Infection was treated primarily by removal of osteosynthesis material and resection of infected bone. Nine months later, secondary reconstruction with a neutral reconstruction plate and bone graft was performed. D (following page), Radiograph after 14 months showing reconstruction plate removed and good bony union:
tibility to infection is related to mobility of the fragments. I6 Fracture site stability is dependent on the rigidity produced by the plate and screw system. In this study, we found some characteristics of fracture type and osteosynthesis techniques that may have contributed to an unstable osteosynthesis. In 12 of the 13 cases, infection was associated with fractures in the proximal part of the mandibular body or angular region, where bone is thinner than in the middle part of the mandible. The crosssectional surface in this region may not constitute a sufficiently large area of contact between segments for a technically sound compression osteosynthesis. In most cases, a tooth was also extracted from the fracture line before reduction and osteosynthe-
sis. This not only allows greater increase of oral contamination, but also it reduces the stability even more. Extraction of a molar tooth from the fracture line thus seems to be a contributing factor to infection. This has also been reported in the literature .4.8-‘7-19A tooth in the fracture line can give good support when the fracture is reduced and extraction of the tooth makes reduction difficult and decreases the cross-sectional surface of the fracture site. Thus, a tooth should probably not be extracted before the osteosynthesis is completed when rigid fixation is used. In our experience a tooth, even an impacted third molar either can safely be retained or else extracted immediately after the osteosynthesis.
590
FIGURE 3
INFECTION
AND MANDIBULAR
FRACTURES
(Continued).
Infection occurred only after osteosyntheses using compression plates. An explanation for this might be that the holes in the compression plate offer limited possibilities to position screws. Placing a screw in an unfavorable site, for example directly in the fracture line, may presumably be done more readily, especially by the inexperienced surgeon. The reconstruction plate affords greater latitude due to its length and greater opportunity to position the screws. Among our patients, no infections were observed when fractures of the kind described here were stabilized with reconstruction plates. If these two types of plates are compared, osteosynthesis using reconstruction plates is probably a more se-
cure method. This holds true especially for those cases where the potential to achieve interfragmentary compression is uncertain. Thinness of mandibular bone (angular region and proximal part of body, atrophic edentulous mandible), extraction of a tooth from the fracture line, irregularity of the fracture line, and even a slight degree of comminution and subclinical infection are factors that might interfere significantly with a compression technique . It appears that complications when using rigid fixation are closely related to technique and the operator’s experience and skill. The decrease of annual infection rate in our patient series supports this
IIZUKA ET AL
FIGURE 4. A, Fracture of mandibular body treated with a sixhole DCP. A second molar was removed from the fracture line before reduction and osteosynthesis. B, Detailed radiographic image showing bone resorption and a loose fragment 6 weeks postoperatively. C, Cross-sectional tomography shows loosening of one screw, which appears to be in the fracture line. D Cfdowing pup), Five months postoperative radiograph after treatment in the acute stage of infection with sequestrectomy. reconstruction of plate, and a bone graft.
592
FIGURE 4
INFECTION
AND MANDIBULAR
FRACTURES
(Continued).
premise. Initially, until the end of 1985, we reported an infection rate of 13% for our first 45 operations, including osteosyntheses using miniplates and reconstruction of large defects.4 The present study showed that the rate of infection was 4.4% in the period 1987 to 1989 when rigid internal fixation became a routine method of fracture treatment. This rate is not especially high compared with data from other studies8*9,‘4,‘5 and corresponds to that observed for patients treated with transosseous wiring and MMF in our unit (4%).4 Infection after plate osteosynthesis causes a difficult situation for both patient and the oral and maxillofacial surgeon. In general, initial treatment includes incision and drainage, removal of plates, and secondary reconstruction of the mandibular defect. However, the presence of infection has even been considered an indication for the use of bone plates by some surgeons.6S’6820 According to our very limited experience with three cases, it was possible to perform mandibular reconstruction in the acute stage of infection without any maxillomandibular immobilization or external fixation. Rigid fixation may thus possibly be an effective alternative for the treatment of infected mandibular fractures. However, this needs further investigation. Detailed radiologic examination, especially
cross-sectional tomography, was of great value in planning the secondary operation. Close cooperation between surgeon and radiologist resulted in a clear definition of the diagnostic problem and selection of an optimal imaging method in every case. This study demonstrates clearly the need of an early clinical and radiologic follow-up of patients with mandibular fractures treated surgically. The signs of infection are apparent within a few weeks after the primary operation, and postoperative appointments on a monthly basis only may result in delay of the diagnosis. The clinical signs of infection were preceded by radiological signs of bone resorption or loosening of screws in most cases. Further prospective studies may answer the question of the association of bone resorption and infection. References 1. Strelzow VV, Friedman WH: Dynamic compression plating in the treatment of mandibular fractures. Arch Otolaryngo1 108583, 1982 2. Klotch DW, Bilger JR: Plate fixation for open mandibular fractures. Laryngoscope 95:1374, 1985 3. Tu HK, Tenhulzen D: Compression osteosynthesis of mandibular fractures. A retrospective study. J Oral Maxillofac Surg 43:585, 1985 4. Lindqvist C, Kontio R, Pihakari A, et al: Rigid internal tixation of mandibular fractures. An analysis of 45 patients treated according to the ASIF method. Int J Oral Maxillofac Surg 15657, 1986
593
IIZUKA ET AL
5. Pogrel MA: Compression osteosynthesis in mandibular fractures. Int J Oral Maxillofac Surg 15:521. 1986 6. Schwimmer A, Greenberg A: Management of mandibular trauma with rigid internal fixation. Oral Surg 62:630, 1986 7. Jones JK, Van Sickels JE: Rigid fixation. A review of concepts and treatment of fractures. Oral Surg 65:13, 1988 8. Ardary WC: Prospective clinical evaluation of the use of compression plates and screws in the management of mandible fractures. J Oral Maxillofac Surg 47: 1150, 1989 9. Peled M, Laufer D, Helman J, et al: Treatment of mandibular fractures by means of compression osteosynthesis. J Oral Maxillofac Surg 47: 1989, 1989 IO. Becker R: Stable compression plate fixation of mandibular fractures. Br J Oral Surg 12:13, 1974 Il. Kahnberg KE, Ride11 A: Bone plate fixation of mandibular fractures. Int J Oral Surg 9:267, 1980 12. Bochlogyros PN: A retrospective study of 1,521 mandibular fractures. J Oral Maxillofac Surg 43:597. 1985 13. Oikarinen K, Altonen M, Kauppi H. et al: Treatment of
14.
15.
16. 17.
18. 19. 20.
mandibular fractures. Need for rigid internal fixation. J Craniomaxillofac Surg 17:24, 1989 Theriot BA, Van Sickels JE, Triplett RG, et al: Interosseous wire fixation versus rigid osseous fixation of mandibular fractures. J Oral Maxillofac Surg 45:577, 1987 Dodson TB, Perrott DH, Kaban LB, et al: Fixation of mandibular fractures: A comparative analysis of rigid internal fixation and standard fixation techniques. J Oral Maxillofat Surg 48:362. 1990 Spiessl B: Internal Fixation of the Mandible. A Manual of AO/ASIF principles. Berlin, Springer-Verlag. 1989. p 5 Neal DC, Wagner WF, Alpert B: Morbidity associated with teeth in the line of mandibular fractures. J Oral Surg 36:859. 1978 James RB, Fredrickson C, Kent JN: Prospective study of mandibular fractures. J Oral Sure 39:275. 1981 Shetty V. Freymiller E: Teeth in the fracture line. J Oral Maxillofac Surg 47: 1303, 1989 Beckers HL: Treatment of initially infected mandibular fractures with bone plates. J Oral Surg 37:310. 1979
Surg
1991
Infection
After Rigid Internal Fixation of Mandibular Fractures: A Clinical and Radiologic
Study
TATEYUKI IIZUKA, MD, DDS,* CHRISTIAN LINDQVIST, MD, DDS, PHD,t DORRIT HALLIKAINEN, MD, PHD,$ AND PERTTI PAUKKU, MD* Of 214 patients with mandibular fractures treated following the AO/ASIF principle of rigid fixation, 13 patients (6.1%) developed postoperative infection. The yearly percentage decreased remarkably during the study period. When rigid fixation became a routine method, the infection rate was as low as the corresponding figure for cases treated with nonstable techniques and maxillomandibular fixation (MMF). All but one of the infected fractures were in the angular region of the mandible. Teeth had been extracted from the fracture line in 9 of the 10 dentulous patients. It was concluded that erroneous techniques had been used in almost all infected cases, because compression could not be achieved when there was an irregular fracture line, an atrophic edentulous mandible, or inadequate stability due to removal of a tooth in the line of fracture. In five patients, successful reoperation was performed using a reconstruction plate. Detailed radiologic examination was useful in assessment of infection and in follow-up.
Operative treatment of mandibular fractures ideally should include reestablishing the correct occlusion and immediate, uncompromized mandibular function. Unstable fixation methods, however, often require maxillomandibular fixation (MMF), which may delay functional rehabilitation. Therefore, rigid internal fixation allowing early mobilization has increased in popularity during the last decade. I-9 Some authors, however, have claimed that the use of rigid plates increases the frequency of complications such as infection, malocclusion, delayed union, nonunion, osteomyelitis, and injury to the inferior alveolar and facial nerves.3*4,10-13Although these complications can also be caused by
the trauma itself, they are often associated with the method of treatment. Postoperative infection can be the most problematic complication in certain cases. Infection rates of 3% to 27% have been reported when using bone plates; the corresponding rates being lower when using transosseous wiring.3*‘4-‘5 It is difficult to compare such rates, however, because the treatment methods (immobilization or mobilization) are totally different from each other. Rigid fixation in mandibular bone surgery has been used in our department since 1983. The purpose of this study was to analyze cases with mandibular fractures in which infection developed after rigid plate osteosynthesis to evaluate the frequency, causes, and diagnostic measures.
Received from Helsinki University Central Hospital, Finland. * Resident, Department of Oral and Maxillofacial Surgery. t Head, Department of Oral and Maxillofacial Surgery. $ Senior Radiologist, Department of Radiology. Address correspondence and reprint requests to Dr Iizuka: DeDartment of Oral and Maxillofacial Suraerv. Helsinki Universit; Central Hospital, Kasarmikatu 1I-13,-Sl%O130 Helsinki 13. Finland.
Patients and Methods
0 1991 geons 0276-2391
American
Association
of Oral
and Maxillofacial
Between 1983 and 1989, 1,823 patients with mandibular fractures were treated in the Department of Oral and Maxillofacial Surgery, Helsinki University Central Hospital, Helsinki, Finland. The annual percentage of open reduction and osteosynthesis of the mandibular fractures varied between 18% and 29% (mean, 25%). The main indication for surgical
Sur-
I91 /4906-0006$3.00/O
585
586
INFECTIONANDMANDIBULARFRACTLJRES
treatment was displacement and/or instability of a fracture in the symphysis, body, or angular region. Nondisplaced, stable fractures and condylar fractures were usually treated conservatively with maxillomandibular immobilization. The use of rigid tixation increased and that of transosseous wiring decreased during the 6-year period. From 1987 on, rigid fixation was the routine method of osteosynthesis (Fig 1). TREATMENT
WITH RIGID FIXATION
During the study period, a total of 214 patients (175 male, 39 female) with 270 mandibular fractures were treated with rigid internal fixation. The mean age of the patients was 34 years (range, 8 to 83 years). Distribution of fractures according to the location, surgical approach, and type of plate is given in Table 1. These numbers do not include cases with large defects or comminution, often due to gunshot injuries, which were excluded from the present analysis. The mean delay from injury to treatment was 3.1 days. The operation was performed following the AO/ASIF principles described by Spiessl.16 Suction drains were placed routinely for the first 24 postoperative hours. Prophylactic antibiotics were used in each case where the frac340 1 320-
ture involved tooth-bearing areas of the mandible. This treatment was continued for 7 to 10 days postoperatively. INFECTIONS
In 13 of 214 patients (6.1%), infection occurred during the postoperative treatment period (Fig 2, Table 2). Ten were male and 3 were female, from 23 to 49 years old, with a mean age of 35.7 years. There were 11 dentulous and 3 totally edentulous mandibles. Criteria for a diagnosis of infection included the presence of pain, swelling, erythema, and/or purulent extra- or intraoral discharge, along with leukocytosis and elevated erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP). Most of the 13 patients developed an abscess at the fracture site, with a sinus tract to the skin or mucosa of the oral cavity. All infections were solitary. Cases of osteitis caused by a tooth in the fracture site requiring endodontic treatment only were not classified as infections, as was simple intraoral wound dehiscence not leading to major complications. Data for all patients with infection were analyzed to determine contributing factors. TREATMENTOF
INFECTIONS
In all 13 cases, the infection required removal of the osteosynthesis material. Secondary treatment consisted of sequestrectomy and, in 5 patients, application of an AO/ASIF reconstruction plate with or without a bone graft from the iliac crest. Three of these patients were operated (two with cancellous bone grafts and one without) in the acute stage of the infection. Of the other eight cases where no reconstruction procedure was undertaken, one fracture required external fixation, and another MMF. In one case, the fracture was stabilized with a Kirschner wire. In five patients, the fracture was left to stabilize by itself without any fixation or immobilization (Table 3).
3002802602400' 220213
RADIOLOGIC EXAMINATION
81
82
a3
84
85
86
a7
00
89
YEAR
FIGURE 1. Total annual number of mandibular fractures and open reduction according to different fixation methods in the Department of Oral and Maxillofacial Surgery, Helsinki University Central Hospital, Finland, in the period of 1981 to 1989.~a--, Total; @ rigid plate; !&I,transosseous wiring and MMF; n , miniplate and MMF.
Routine radiologic evaluation of the fractures preoperatively and postoperatively included panoramic images, and Townes and axial projections of the skull. For detailed evaluation, a Scanora x-ray unit (Orion Corp/Soredex, Helsinki, Finland) was used. These examinations included a panoramic image followed by detailed imaging with a narrowbeam technique and cross-sectional tomography. The images were evaluated for loosening of screws, bone resorption under the osteosynthesis plate, and
587
IIZUKA ET AL
Table 1. Fmctura Sltos and mthcds of Fixation In 214 Patlonts With Mandlbulrr Fractures Tmatod Following the AO/ASlF Rinclple of Rigld Flxatlon Approach Mandibular Fracture
Osteosyntheses
Intraoral
No.of Fractures Extraoral
Symphysis
20
14
6
Canine
39
23
16
Body
90
16
74
Angle
121
5
116
Total
270
58
212
Abbreviations:
DCP. dynamic compression
plate: EDCP. excentric dynamic compression
altered bone structure suggestive of infection. During follow-up, the same type of examination was used to evaluate healing after infection and osseous union after bone grafting. Since 1988, all patients with postoperative infection (9 to 13 in Table 2) have been followed radiologically by the same radiologist, from the diagnosis of infection through the whole treatment phase. At secondary operation, the local findings were all documented by the same surgeon. Results
The yearly percentage of postoperative infection changed remarkably during the study period (Fig 2). It decreased from 10.9% in the period 1983 to 1986 to 4.8% in 1987 to 1988 and to 3.6% in 1989. Detailed data on the type of fracture, primary osteoSl
_. u)
50 -
E: .s! iii n
40 -
5
30-
p
20-
53
55
10 ‘O
2 .
om 83
84
85
86
87
88
89
YEAR
FIGURE 2. Yearly distribution of infection cases in relation to total number of patients with mandibular fractures treated with rigid fixation. W. Rigid fixation; E, infection.
Fixation
Operated On 16 4 31 4 3
DCP RP DCP EDCP RP Lag screw DCP EDCP RP Lag screw DCP EDCP RP Lag screw
I 66 5 18
1 80 16 22 3 270
plate; RP, reconstruction
plate.
synthesis, and treatment of the infection are shown in Tables 2 and 3. There were eight fractures situated in the angular region and four in the proximal part of the body. In one edentulous patient, infection occurred in the parasymphysis region. Teeth (second or third molars) had been extracted from the fracture line before the osteosynthesis procedure in 9 of the 10 dentulous mandibles. In all 13 patients, plates causing compression were used primarily. Infection was diagnosed after a mean period of 7.3 weeks. No correlation was found between delay of surgical treatment and rate of infection. In 2 patients, clinical signs of infection were observed preoperatively (Fig 3A-D). Bacteriologic analysis of the pus from the infected area showed Staphylococcus epidermidis in eight cases, Staphylococcus aureus in four, and Streptococcus viridans in one. In all cases, loosening of screws was observed during removal of the osteosynthesis material, irrespective of the stage of fracture healing. Nonunion of the fracture and bone sequestration was found in six cases. In five cases, there was soft-tissue bridging in the fracture line without macroscopically pathologic bone. In two patients, an osseous union had occurred across the fracture site. In all three cases where reconstruction was performed in the stage of acute infection, the postoperative course was uneventful. There was no need of MMF, and the mandible could be mobilized immediately after surgery. Good bone formation was observed radiologically within an average of 7 months (Fig 4A-D). It was not possible to predict an infection from the postoperative radiographs taken on the first or
588
INFECTION
AND MANDIBULAR
FRACTURES
Table 2. Data on 13 Patients With Infection After Treatment of Mandibular Fracture With Rigid Internal Fixation Patient No.
Age (yr)
Sex F F M M M M F M M M M M M
1 2 3 4 5 6 7 8 9 10 11 12 13
Delay From Accident to Operation (d)
Location of Fracture
8 1 3 5 3 3 1 0 13 2 4 5 2
Body Body Angle Angle Angle Angle Symphy sis Angle Angle Angle Angle Body Body
43 44 23 38 23 31 42 32 21 36 49 46 30 35.7
Mean
Plate Type*
Surgical Approach
Extraction of Tooth (tooth no.)
DCP6 DCP4 DCP6 EDCP6 DCP2 + DCP4 DCP2 + DCP4 DCP4 DCP4 DCP6 DCP2 + DCP4 DCP6 DCP6 DCP4
E E E E E E E E E E E E I
Edentulous Edentulous Yes (48) Yes (38) Yes (37) Yes (38) Edentulous Yes (47) Yes (48) Yes (48) Yes (47) Yes (37) No
3.8
Abbreviations: DCP, dynamic compression plate; E, extraoral; EDCP, excentric dynamic compression * Numeral indicates number of holes in plate.
second postoperative day in any case. Radiologically, a screw in the fracture line was noted in three cases. Bone resorption around the plate preceded other radiologic signs of infection. The earliest it was visible was 2 weeks postoperatively. Screw loosening was noted in two patients before other signs of infection and in three when clear signs of infection were visible. No radiologic signs of infection were noted earlier than 4 weeks after surgery. The radiologic diagnosis of loosening of screws and bone resorption under the osteosynthesis plate was confirmed at surgery in each case. Bone infection was also confirmed in all cases, but the area of infected bone was radiologically estimated to be larger than the area of macroscopically pathologic Table 3.
plate; I, intraoral.
bone documented at surgery. The radiologic followup correlated well with the clinical findings. Discussion
Various factors may contribute to failure with the use of plates. The higher incidence of infection when using bone plates has raised a concern that plates and screws, with their increased quantity of metal, might predispose to infection. Infections were treated successfully, however, using larger quantities of metal than was used in the primary operation. The results of research by the Swiss Association for the Study of Internal Fixation show that suscep-
Method of Treatment of Infection and Resulting Late Complications
Patient No. 1 2 3 4 5 6 7 8 9 10 11 12 13 Mean
Abbreviations:
Infection Diagnosed Postoperatively (wk) 25 1 7 1 3 4 9 9 1 20 4 4 7
Fixation of Fragments
Reconstruction of Mandible
No No No
RP MMF No External Kirschner wire + MMF RP No RP RP RP
Secondary (bone)
Primary (bone) Secondary (bone) Primary (bone) Primary
Follow-up Time (mo)
Slight instability Slight instability None None None None None None None None None None None
4 6 6 49 8 4 12 18 19 15 15 9 9 13
7.3 MMF, maxillomandibular
Late Complication
fixation; RP, reconstruction
plate.
IIZUKA ET AL
589
FIGURE 3. A, Radiograph showing right preangular fracture. There was a suppurative infection on admission. E. Fracture treated with a sixhole dynamic compression plate (DCP) in spite of infection and comminution. Tooth was removed from the fracture line before reduction and osteosynthesis. C (following page), Infection was treated primarily by removal of osteosynthesis material and resection of infected bone. Nine months later, secondary reconstruction with a neutral reconstruction plate and bone graft was performed. D (following page), Radiograph after 14 months showing reconstruction plate removed and good bony union:
tibility to infection is related to mobility of the fragments. I6 Fracture site stability is dependent on the rigidity produced by the plate and screw system. In this study, we found some characteristics of fracture type and osteosynthesis techniques that may have contributed to an unstable osteosynthesis. In 12 of the 13 cases, infection was associated with fractures in the proximal part of the mandibular body or angular region, where bone is thinner than in the middle part of the mandible. The crosssectional surface in this region may not constitute a sufficiently large area of contact between segments for a technically sound compression osteosynthesis. In most cases, a tooth was also extracted from the fracture line before reduction and osteosynthe-
sis. This not only allows greater increase of oral contamination, but also it reduces the stability even more. Extraction of a molar tooth from the fracture line thus seems to be a contributing factor to infection. This has also been reported in the literature .4.8-‘7-19A tooth in the fracture line can give good support when the fracture is reduced and extraction of the tooth makes reduction difficult and decreases the cross-sectional surface of the fracture site. Thus, a tooth should probably not be extracted before the osteosynthesis is completed when rigid fixation is used. In our experience a tooth, even an impacted third molar either can safely be retained or else extracted immediately after the osteosynthesis.
590
FIGURE 3
INFECTION
AND MANDIBULAR
FRACTURES
(Continued).
Infection occurred only after osteosyntheses using compression plates. An explanation for this might be that the holes in the compression plate offer limited possibilities to position screws. Placing a screw in an unfavorable site, for example directly in the fracture line, may presumably be done more readily, especially by the inexperienced surgeon. The reconstruction plate affords greater latitude due to its length and greater opportunity to position the screws. Among our patients, no infections were observed when fractures of the kind described here were stabilized with reconstruction plates. If these two types of plates are compared, osteosynthesis using reconstruction plates is probably a more se-
cure method. This holds true especially for those cases where the potential to achieve interfragmentary compression is uncertain. Thinness of mandibular bone (angular region and proximal part of body, atrophic edentulous mandible), extraction of a tooth from the fracture line, irregularity of the fracture line, and even a slight degree of comminution and subclinical infection are factors that might interfere significantly with a compression technique . It appears that complications when using rigid fixation are closely related to technique and the operator’s experience and skill. The decrease of annual infection rate in our patient series supports this
IIZUKA ET AL
FIGURE 4. A, Fracture of mandibular body treated with a sixhole DCP. A second molar was removed from the fracture line before reduction and osteosynthesis. B, Detailed radiographic image showing bone resorption and a loose fragment 6 weeks postoperatively. C, Cross-sectional tomography shows loosening of one screw, which appears to be in the fracture line. D Cfdowing pup), Five months postoperative radiograph after treatment in the acute stage of infection with sequestrectomy. reconstruction of plate, and a bone graft.
592
FIGURE 4
INFECTION
AND MANDIBULAR
FRACTURES
(Continued).
premise. Initially, until the end of 1985, we reported an infection rate of 13% for our first 45 operations, including osteosyntheses using miniplates and reconstruction of large defects.4 The present study showed that the rate of infection was 4.4% in the period 1987 to 1989 when rigid internal fixation became a routine method of fracture treatment. This rate is not especially high compared with data from other studies8*9,‘4,‘5 and corresponds to that observed for patients treated with transosseous wiring and MMF in our unit (4%).4 Infection after plate osteosynthesis causes a difficult situation for both patient and the oral and maxillofacial surgeon. In general, initial treatment includes incision and drainage, removal of plates, and secondary reconstruction of the mandibular defect. However, the presence of infection has even been considered an indication for the use of bone plates by some surgeons.6S’6820 According to our very limited experience with three cases, it was possible to perform mandibular reconstruction in the acute stage of infection without any maxillomandibular immobilization or external fixation. Rigid fixation may thus possibly be an effective alternative for the treatment of infected mandibular fractures. However, this needs further investigation. Detailed radiologic examination, especially
cross-sectional tomography, was of great value in planning the secondary operation. Close cooperation between surgeon and radiologist resulted in a clear definition of the diagnostic problem and selection of an optimal imaging method in every case. This study demonstrates clearly the need of an early clinical and radiologic follow-up of patients with mandibular fractures treated surgically. The signs of infection are apparent within a few weeks after the primary operation, and postoperative appointments on a monthly basis only may result in delay of the diagnosis. The clinical signs of infection were preceded by radiological signs of bone resorption or loosening of screws in most cases. Further prospective studies may answer the question of the association of bone resorption and infection. References 1. Strelzow VV, Friedman WH: Dynamic compression plating in the treatment of mandibular fractures. Arch Otolaryngo1 108583, 1982 2. Klotch DW, Bilger JR: Plate fixation for open mandibular fractures. Laryngoscope 95:1374, 1985 3. Tu HK, Tenhulzen D: Compression osteosynthesis of mandibular fractures. A retrospective study. J Oral Maxillofac Surg 43:585, 1985 4. Lindqvist C, Kontio R, Pihakari A, et al: Rigid internal tixation of mandibular fractures. An analysis of 45 patients treated according to the ASIF method. Int J Oral Maxillofac Surg 15657, 1986
593
IIZUKA ET AL
5. Pogrel MA: Compression osteosynthesis in mandibular fractures. Int J Oral Maxillofac Surg 15:521. 1986 6. Schwimmer A, Greenberg A: Management of mandibular trauma with rigid internal fixation. Oral Surg 62:630, 1986 7. Jones JK, Van Sickels JE: Rigid fixation. A review of concepts and treatment of fractures. Oral Surg 65:13, 1988 8. Ardary WC: Prospective clinical evaluation of the use of compression plates and screws in the management of mandible fractures. J Oral Maxillofac Surg 47: 1150, 1989 9. Peled M, Laufer D, Helman J, et al: Treatment of mandibular fractures by means of compression osteosynthesis. J Oral Maxillofac Surg 47: 1989, 1989 IO. Becker R: Stable compression plate fixation of mandibular fractures. Br J Oral Surg 12:13, 1974 Il. Kahnberg KE, Ride11 A: Bone plate fixation of mandibular fractures. Int J Oral Surg 9:267, 1980 12. Bochlogyros PN: A retrospective study of 1,521 mandibular fractures. J Oral Maxillofac Surg 43:597. 1985 13. Oikarinen K, Altonen M, Kauppi H. et al: Treatment of
14.
15.
16. 17.
18. 19. 20.
mandibular fractures. Need for rigid internal fixation. J Craniomaxillofac Surg 17:24, 1989 Theriot BA, Van Sickels JE, Triplett RG, et al: Interosseous wire fixation versus rigid osseous fixation of mandibular fractures. J Oral Maxillofac Surg 45:577, 1987 Dodson TB, Perrott DH, Kaban LB, et al: Fixation of mandibular fractures: A comparative analysis of rigid internal fixation and standard fixation techniques. J Oral Maxillofat Surg 48:362. 1990 Spiessl B: Internal Fixation of the Mandible. A Manual of AO/ASIF principles. Berlin, Springer-Verlag. 1989. p 5 Neal DC, Wagner WF, Alpert B: Morbidity associated with teeth in the line of mandibular fractures. J Oral Surg 36:859. 1978 James RB, Fredrickson C, Kent JN: Prospective study of mandibular fractures. J Oral Sure 39:275. 1981 Shetty V. Freymiller E: Teeth in the fracture line. J Oral Maxillofac Surg 47: 1303, 1989 Beckers HL: Treatment of initially infected mandibular fractures with bone plates. J Oral Surg 37:310. 1979
