- Email: [email protected]
Ultrasonographic appearance of external callus in long-bone fractures
licrury Vol. 26, No. I, pp. 5- 12. 1995 Copyright 5, 1995 Elsevier Science Ltd Printed m Great Britain. All rights reserved 0020.1383/95
$lo.oo+o.oo
Papers
Ultrasonographic appearance long-bone fractures N. Maffulli
of external
callus
in
and A. Thornton’
Department of Orthopaedic Surgery, University of Aberdeen Medical School, Aberdeen, Scot-land, and *Department Radiology, Newham General Hospital, London, UK
High-resolution real-time ultrasound (US] scanning was used to monitor serially 24 patients with an acute long-bone fracture and fkree patients with humeral non-union, looking af the fracture site over a period of 12 months following the fracture. In both these groups, we found that US gave important information about the soft tissues surrounding the fracture site, and indicated callus formafion at an early stage. US was more sensitive than conventional radiography at showing the early phases of organization of the callus, and ifs progression to bridging new bone formation. US also clearly showed a disorganized eckopatfern at the fracture sife of the patients with non-union.
Injury, 1995, Vol. 26, 5-12, January
Introduction Fracture healing aims to restore the integrity of the injured bone. If no internal fixation is performed, a well-ordered series of events takes place’,‘, resulting in formation of callus that subsequently ossifies to reconstitute bone continuity’. The universally used method to follow the evolution of a fracture is radiography, which can show evidence of bone healing only when calcification, through seeding of hydroxyapatite crystals at the callus site, starts to take place’. Conventional radiography can therefore lag some weeks behind the physiological events of bone healing3. Magnetic resonance images (MRI) of tibia1 shaft fracture repair revealed a good correlation between images and histological events4. However, MRI is not widely available, is expensive, can be time consuming, and not enough dafa are available. High-resolution real-time ultrasonographic scanning (US) has recently been used in the diagnosis of fractures of the clavicle5-7, humerus? and femur’. US has also been used in the diagnosis of non-traumatic bone conditions9-“, and to monitor the process of new bone formation in limb lengthening’* 13. To our knowledge, only one study has been published on the longitudinal evaluation of callus and bone formation at the fracture site. Those patients were undergoing external fixation for a long-bone fracture14.
of
We report our experience in the follow-up of the fracture site in fractures of long bones in patients treated non-operatively.
Materials
and methods
Patients In the period January to December 1991, 24 patients (15 males, nine females, average age 42.9 i 10.3 years, range 4 to 71 years, six humeral, 11 femoral and seven tibia1 fractures) gave their informed consent to undergo serial US and radiographic examination of their fractured limb. Also, three patients with long-standing non-union of humeral shaft fractures gave their informed consent to undergo US scanning. Humeral fractures were treated by resting the forearm in a broad arm sling, taking care that it was lying horizontal. Femoral and tibia1 fractures were treated by bed rest and skin or skeletal traction until good alignment of the bony fragments had been achieved (generally, 10 days to 3 weeks after the injury). After that time, tibia1 and femoral fractures in adults were treated in a functional cast for 3 months. Children with femoral fractures were treated by traction only. Ultrasound scanning technique Sonograms were obtained in the longitudinal, oblique and coronal planes using a 5 MHz phase array or a linear array transducer on an Aloka SSD 650 or an ATL Mark 4 unit. A commercially-available coupling gel was used to ensure optimal contact between probe and skin. The US images were photographed, and all measurement were carried out on these. All radiographs and US scans were evaluated by a radiologist and an orthopaedic surgeon with particular attention to: (a) possible displacement or malalignment of the bony segments; (b) distance between the fracture ends and the bony fragments when these were present; (c) appearance and (d) maturity of the callus. US scans and radiographs were routinely taken on admission, after 3 weeks, and after 3 months from the time of injury. A final US scan and radiograph was taken I year after discharge. More frequent or longer follow-up occurred when the patient’s clinical condition merited it, or at the surgeon’s request.
6
Injury:
International
Journal
of the Care of the Injured
(1995)
Vol. 26/No.
1
pockets are seen as it liquefies (see Figure 3). The time scale of these changes is variable. As healing progresses, only an ill-defined 2 mm layer of more diffuse low-level echoes are seen next to the cortex (see Figures 5 and 6~). With complete healing, haematoma resorption is seen as wellorganized soft-tissue echotexture next to the cortex (see Figures 5 and 64. (6) In general, medullary canal texture is not well defined, as it is shielded by the overlying cortex. (7) Callus appears as very irregular bright echoes, especially when it mineralizes (see Figures 4 and 6a). Bright echoes are also present when the callus is disorganized, as it happens in non-unions, when the echotexture is bizarre and haphazard (see Figure 7A).
Results Soft tissues
Ring artifact
of probe
Acute peri-fracture haematomas were seen as hyperechoic areas within the muscle (Figure I). From an early stage (I week for younger patients, and up to 4 weeks in adults), they became progressively hypoechoic as haemolysis took place (Figuresh and 3). During the organization period, irregular echoes were seen in association with the progression of the echogenic linear bridge of callus and bone across the fracture (Figured). These disappeared by 3 months after the injury, and, at later review, no soft tissue abnormality was detected (Figure 5).
,
Bone b
Figure 1. a, Longitudinal US scan of a proximal femoral shaft
fracture 1 week after trauma in a 52-year-old male. The displacement at the fracture site is clearly shown with adjacent echogenic soft tissue haematoma. b, Diagram of US scan. C = cortex; H = haematoma; S = skin probe/skin interface; F = fracture site.
Interpretation
of ultrasound
scans
A simple guide to interpret US scans of bone fractures is as follows: (1) A 1 cm regular ring artifact is caused by the probe itself. This can be appreciated in all the scans presented. (2) Under this artifact, a thin regular bright line is formed at the skin-probe interface. Again, this can be appreciated in all the scans presented. (3) Inferior to the thin regular bright line at the skin-probe interface, an area of mixed echogenicity is found. This is formed by subcutaneous fat and/or muscle, which cannot always be separated from each other, especially when a recent irregular haematoma (with an US appearance of bright echoes) is present. Muscle has an organized striated echostructure (see Figure 5). (4) When intact, cortical bone margins (i.e. cortex) form a very bright smooth line I mm thick. This is interrupted and irregular when bone ends are malaligned, and not perpendicular to the probe. An acoustic shadow may be cast behind the cortex or area of calcifications depending on the probe-cortex-calcium angle (see Figures I and 3). (5) Haematoma forms irregular and initially fairly bright echoes (see Figure I). Later, well defined echolucent
The cortex of a normal bone was seen as a smooth, echogenic reflective surface I2 . It appeared as a curvilinear or a linear structure in the coronal and longitudinal planes, respectively. In the first scan, US showed a sharply demarcated vertical defect of the dense cortical bone, and an abnormality in its normal configuration (Figure I). Small separated bone fragments were seen as highly echogenic foci, and angulation, when present, was easily appreciated (Figrtre6). The gap between the fracture fragments was clearly and easily measured (Figure I ). This appearance progressed to one of organization of the peri-fracture haematoma, associated with gradual filling in of the cortical defect with increasingly more echogenic material. This progressed to an organized curving bridge across the fracture site (Figures2 and 4). Initially, the echogenic callus was irregularly aligned (Figure 6a). As new bone formed, the echogenic foci underwent progressive alignment along the long axis of the bone in an arching or ‘bowed’ shape, which gradually straightened (Figure 6b). The echogenic areas, representing early ossification foci, were often not detected radiographically (Figuresda and b). Eventuafly, the cortex was fully repaired, and regained its continuity (Figures 5 and 6b). Even 1 year after the injury, at full clinical resolution, it was possible to detect distorted bone modelling around the previously fractured area, and residual angulation (Figure 6~). The cortex had returned to its normal highly echogenic appearance, with sharp smooth definition when compared to the echostructure of soft tissues. Fractures
in which
non-union
had taken
place
When poor bony union occurred, the sonographic appearances were useful in confirming the lack of a bony bridge, and showed either several bone fragments and poor alignment, or a disorganized echopattem with loss of the linear echogenic cortical border (Figures 7a, b).
Maffulli
and Thornton:
External
callus in long-bone
fractures
artifact
Subcutaneous and muscle
tissue
Figure 2. II, Same patient as Figure I after 5 weeks of traction. There is good bone alignment with early callus formation bridging across the distracted bone. The adjacent soft tissues show a low echogenic pattern. b, Radiograph of the fracture site at the same healing stage for comparison. Hardly any callus is visible around the fracture site. c, Diagram of US scan. C = cortex; S = skin; H = haematoma; Ca = early callus; F = fracture site.
Injury:
International
Journal
of the Care of the Injured
(1995)
Vol. 26/No.
I
Probe artifact
S = skin C z Cortex
b
Figure 3. a, Longitudinal scan of a femoral fracture site in a J-year-old boy, 2 weeks after the trauma. There is a slight displacement of the fracture ends and the adjacent soft tissue haematoma is echolucent. b, Diagram of US scan. S= skin; C =cortex; H= haematoma - low echogenic focal areas of liquefying haematoma; F = fracture site.
Discussion Studies on fracture healing in humans are lacking, as it is not ethically feasible to biopsy patients, and provide serial correlation of radiographic images with histological samplesx5. Conventional radiography, whilst involving a relatively low radiation dose, is not sensitive enough to be used in the early stages of fracture healing, when nonossified callus is being formed16. Sensitive non-invasive methods are therefore necessary to detect early callus formation. Recently, serial MRI scanning of closed tibia1 shaft fractures showed a characteristic pattern, which correlated well with the histology of fracture repaiP. The soft tissues surrounding the fracture gradually developed signals corresponding to the calcified area. Radiographs were able to detect calcification only several weeks later. In this preliminary
study, US images
can be correlated
b
Figure 4. a, Fracture site of Figures 1 and 2 8 weeks after the trauma. The US scan now begins to show new bone ‘filling in’ across the fracture site with adjacent exuberant but irregular callus. b, Diagram of US scan. S = skin; C = cortex; Ca = early callus - irregular callus with some calcification; F = fracture site.
with radiographic views in the various stages of fracture healing. However, no selection of patients was performed, and therefore the fractures scanned are heterogeneous. Further investigations are required to ascertain whether US can become part of the routine armamentarium of an orthopaedic surgeon to assess callus formation and bone healing. US is cheap and can be taken to the bedside without any loss in the quality of the image, an important advantage with a patient in traction. In our hands, it has shown a staged sequence of events in the patients studied, reflecting the present knowledge on the histology of fracture healing l,z. In fact, the peri- fracture haematoma is gradually invaded by osteogenic repair tissue, leading to gradual mineralization
of the callus which
eventually
fracture ends’sz. At present, however, qualitative morphological information.
bridges
the
US only gives Further efforts
Maffulli
and Thornton:
External
callus in long-bone
9
fractures
tions in the early phases of the fracture healing process, and when it is necessary to date a fracture”. It also gives useful information in cases of non-union.
References
Probe
artifact \
Subcutaneous
tissues F = Fracture
site
b
Figure 5. u, Fracture site shown in Figure 3, 12 weeks after the original trauma. The fracture site is just visible. At this stage, no soft tissue abnormality is seen. b, Diagram of US scan. S = skin; C = cortex; F = fracture site.
1 McKibbin B. The biology of fracture healing in long bones. ] Bone Joint Surg. [Br] 1978; 60B: 150. 2 Brighton CT and Hunt RM. Early histological and ultrashuctural changes in medullary fracture callus. ] Bone joint Swg [Am] 1991; 73A: 832. 3 Tiedeman JJ, Lippiello L, Connolly JF and Strates BS. Quantitative roentgenographic densitometry for assessing fracture healing. Clin Orthop 1990; 253: 279. 4 Laarsonen EM, Kyro A, Korhola 0 and Bostman 0. Magnetic resonance imaging of tibia1 shaft fracture repair. Arch Orthop Trauma Surg 1989; 108: 40. 5 Graif M, Stahl-Kent V, Ben-Ami T, Strauss S, Amit Y and Itzchak, Y. Sonographic detection of occult bone fractures. Pediatr Radio1 1988; 18: 382. 6 Katz R, Landman J, Dulifzky F and Bar-Ziv J. Fracture of the clavicle in the newborn. An ultrasound diagnosis. ] Ultrasound Med 1988; 7: 21. 7 Bartoli E, Saporetti N, Marchetti S. Ruolo dell’ecografia nella diagnosi della fratture clavicolari de1 neonate. Radiologia Medica 1989; 77: 466. 8 Pat-fen RM, Mack LA, Wang KY and Lingel J. Nondisplaced fractures of the greater tuberosity of the humerus: sonographic detection. Radiology 1992; 182: 201. 9 Steiner GM and Sprigg A. The value of ultrasound in the assessment of bone. Br ] Radial 1992; 65: 689. 10 Fornage BD, Richli WR and Chuapetcharasopon C. Calcaneal bone cyst: sonographic findings and ultrasound guided aspiration biopsy. ] Clin Ultrasound 1991; 19, 360. 11 Nath AK and Sethu AU. Use of ultrasound in osteomyelitis. Br ] Radio1 1992; 65: 649. 12 Maffulli N, Hughes T and Fixsen JB. Ultrasonographic monitoring of limb lengthening. ] Bone joint Swg [Br] 1992; 74B:
305.
Eyres KS, Bell MJ and Kanis JA. Methods of assessing new bone formation during limb lengthening. Ultrasonography, dual energy X-ray absorptiomehy and radiography compared. ] Bone Joint Surg [Br] 1993; 75B: 358. 14 Ricciardi L, Perissinotto A and Visentin E. Ultrasonography in the evaluation of osteogenesis in fractures treated with Hoffmann external fixator. llal 1 Orthop Truuma 1986; 12: 13
197.
should be made to quantify this information, so as to be able to apply in clinical practice. In conclusion, serial US scanning allows the study of the fracture healing process in uivo, with apparent good matching between images and what is known of the histology of fracture healing. US scanning is not the method of choice for the detection of bone fractures, although some reports claim a role in special caseP.17. At present, we do not know whether the method can be used to detect early signs of post-traumatic pseudarthrosis, thus influencing treatment choices. However, as it is noninvasive, is available in all hospitals, and is a low-cost investigation, US may supplement radiographic examina-
15 Chapman S. The radiological dating of injuries. Arch Dis Child 1992; 67: 1063. 16 Nicholls PJ, Berg E, Bliven FE and Kling JM. X-ray diagnosis of healing fractures in rabbits. Ckn Orthop 1979; 142: 234. 17 Leitbeg N, Bodenteich A, Schweighofer F and Fellinger M. Sonographische Frackturdiagnostik. Ulfraschall in Medizin 1990;
11:205.
Paper accepted
4 August
1994.
Reqtlestsfor reprints should be addressed to: N. Maffulli, Department of Orthopaedic Surgery, University of Aberdeen Medical School. Polwarth Building, Foresterhill, Aberdeen AB9 2ZD, Scotland, UK.
Injury:
Subcutaneous and muscle
International
Journal
of the Care of the Injured
i199.5) Vol. 26/No.
tissue
Probe artifact
Almost fracture
C
d
complete site still
bon
1
Maffulli
and
Thornton:
Subcutaneous and muscle,
External
callus
in long-bone
fractures
tissue
Pro be artifact
Figure6. a, Longitudinal scan of a femoral fracture in a o-year-old girl, 5 weeks after the trauma. There is mild displacement and angulation at the fracture site with early bridging echogenic callus which is irregularly aligned. b, Same as in Figure 6a, 13 weeks after the trauma. There is some minimal cortical irregularity with good bone union at the fracture site. Minimal soft tissue abnormality is seen at this stage. c, Diagram of US scan. S = skin, C = cortex; F = fracture site; Ca = callus. d, Diagram of US scan. S= skin; C = cortex; F= fracture site; Ca=callus. c, At 1 year, there is complete bony union with normal cortical border but mild persistent angulation. d, Radiograph at 1 year taken in the same alignment as in Figure 6e. for comparison. g, Diagram of radiograph. S = skin
Injury:
12
International
Journal
of the Care of the Injured
(1994)
Vol. 26/No.
1
Subcutaneous tissue and muscle \
artifact
S = Skin C =Cortex Ca q Callus
Figure 7. Q, Longitudinal scan of a non-union of a humeral fracture at 1 year. US shows grossly disorganized exuberant callus and disruption of the cortical margin. b, Radiographs at the same healing stage for comparison. c, Diagram of US scan. S = skin; C = cortex; Ca = callus.
$lo.oo+o.oo
Papers
Ultrasonographic appearance long-bone fractures N. Maffulli
of external
callus
in
and A. Thornton’
Department of Orthopaedic Surgery, University of Aberdeen Medical School, Aberdeen, Scot-land, and *Department Radiology, Newham General Hospital, London, UK
High-resolution real-time ultrasound (US] scanning was used to monitor serially 24 patients with an acute long-bone fracture and fkree patients with humeral non-union, looking af the fracture site over a period of 12 months following the fracture. In both these groups, we found that US gave important information about the soft tissues surrounding the fracture site, and indicated callus formafion at an early stage. US was more sensitive than conventional radiography at showing the early phases of organization of the callus, and ifs progression to bridging new bone formation. US also clearly showed a disorganized eckopatfern at the fracture sife of the patients with non-union.
Injury, 1995, Vol. 26, 5-12, January
Introduction Fracture healing aims to restore the integrity of the injured bone. If no internal fixation is performed, a well-ordered series of events takes place’,‘, resulting in formation of callus that subsequently ossifies to reconstitute bone continuity’. The universally used method to follow the evolution of a fracture is radiography, which can show evidence of bone healing only when calcification, through seeding of hydroxyapatite crystals at the callus site, starts to take place’. Conventional radiography can therefore lag some weeks behind the physiological events of bone healing3. Magnetic resonance images (MRI) of tibia1 shaft fracture repair revealed a good correlation between images and histological events4. However, MRI is not widely available, is expensive, can be time consuming, and not enough dafa are available. High-resolution real-time ultrasonographic scanning (US) has recently been used in the diagnosis of fractures of the clavicle5-7, humerus? and femur’. US has also been used in the diagnosis of non-traumatic bone conditions9-“, and to monitor the process of new bone formation in limb lengthening’* 13. To our knowledge, only one study has been published on the longitudinal evaluation of callus and bone formation at the fracture site. Those patients were undergoing external fixation for a long-bone fracture14.
of
We report our experience in the follow-up of the fracture site in fractures of long bones in patients treated non-operatively.
Materials
and methods
Patients In the period January to December 1991, 24 patients (15 males, nine females, average age 42.9 i 10.3 years, range 4 to 71 years, six humeral, 11 femoral and seven tibia1 fractures) gave their informed consent to undergo serial US and radiographic examination of their fractured limb. Also, three patients with long-standing non-union of humeral shaft fractures gave their informed consent to undergo US scanning. Humeral fractures were treated by resting the forearm in a broad arm sling, taking care that it was lying horizontal. Femoral and tibia1 fractures were treated by bed rest and skin or skeletal traction until good alignment of the bony fragments had been achieved (generally, 10 days to 3 weeks after the injury). After that time, tibia1 and femoral fractures in adults were treated in a functional cast for 3 months. Children with femoral fractures were treated by traction only. Ultrasound scanning technique Sonograms were obtained in the longitudinal, oblique and coronal planes using a 5 MHz phase array or a linear array transducer on an Aloka SSD 650 or an ATL Mark 4 unit. A commercially-available coupling gel was used to ensure optimal contact between probe and skin. The US images were photographed, and all measurement were carried out on these. All radiographs and US scans were evaluated by a radiologist and an orthopaedic surgeon with particular attention to: (a) possible displacement or malalignment of the bony segments; (b) distance between the fracture ends and the bony fragments when these were present; (c) appearance and (d) maturity of the callus. US scans and radiographs were routinely taken on admission, after 3 weeks, and after 3 months from the time of injury. A final US scan and radiograph was taken I year after discharge. More frequent or longer follow-up occurred when the patient’s clinical condition merited it, or at the surgeon’s request.
6
Injury:
International
Journal
of the Care of the Injured
(1995)
Vol. 26/No.
1
pockets are seen as it liquefies (see Figure 3). The time scale of these changes is variable. As healing progresses, only an ill-defined 2 mm layer of more diffuse low-level echoes are seen next to the cortex (see Figures 5 and 6~). With complete healing, haematoma resorption is seen as wellorganized soft-tissue echotexture next to the cortex (see Figures 5 and 64. (6) In general, medullary canal texture is not well defined, as it is shielded by the overlying cortex. (7) Callus appears as very irregular bright echoes, especially when it mineralizes (see Figures 4 and 6a). Bright echoes are also present when the callus is disorganized, as it happens in non-unions, when the echotexture is bizarre and haphazard (see Figure 7A).
Results Soft tissues
Ring artifact
of probe
Acute peri-fracture haematomas were seen as hyperechoic areas within the muscle (Figure I). From an early stage (I week for younger patients, and up to 4 weeks in adults), they became progressively hypoechoic as haemolysis took place (Figuresh and 3). During the organization period, irregular echoes were seen in association with the progression of the echogenic linear bridge of callus and bone across the fracture (Figured). These disappeared by 3 months after the injury, and, at later review, no soft tissue abnormality was detected (Figure 5).
,
Bone b
Figure 1. a, Longitudinal US scan of a proximal femoral shaft
fracture 1 week after trauma in a 52-year-old male. The displacement at the fracture site is clearly shown with adjacent echogenic soft tissue haematoma. b, Diagram of US scan. C = cortex; H = haematoma; S = skin probe/skin interface; F = fracture site.
Interpretation
of ultrasound
scans
A simple guide to interpret US scans of bone fractures is as follows: (1) A 1 cm regular ring artifact is caused by the probe itself. This can be appreciated in all the scans presented. (2) Under this artifact, a thin regular bright line is formed at the skin-probe interface. Again, this can be appreciated in all the scans presented. (3) Inferior to the thin regular bright line at the skin-probe interface, an area of mixed echogenicity is found. This is formed by subcutaneous fat and/or muscle, which cannot always be separated from each other, especially when a recent irregular haematoma (with an US appearance of bright echoes) is present. Muscle has an organized striated echostructure (see Figure 5). (4) When intact, cortical bone margins (i.e. cortex) form a very bright smooth line I mm thick. This is interrupted and irregular when bone ends are malaligned, and not perpendicular to the probe. An acoustic shadow may be cast behind the cortex or area of calcifications depending on the probe-cortex-calcium angle (see Figures I and 3). (5) Haematoma forms irregular and initially fairly bright echoes (see Figure I). Later, well defined echolucent
The cortex of a normal bone was seen as a smooth, echogenic reflective surface I2 . It appeared as a curvilinear or a linear structure in the coronal and longitudinal planes, respectively. In the first scan, US showed a sharply demarcated vertical defect of the dense cortical bone, and an abnormality in its normal configuration (Figure I). Small separated bone fragments were seen as highly echogenic foci, and angulation, when present, was easily appreciated (Figrtre6). The gap between the fracture fragments was clearly and easily measured (Figure I ). This appearance progressed to one of organization of the peri-fracture haematoma, associated with gradual filling in of the cortical defect with increasingly more echogenic material. This progressed to an organized curving bridge across the fracture site (Figures2 and 4). Initially, the echogenic callus was irregularly aligned (Figure 6a). As new bone formed, the echogenic foci underwent progressive alignment along the long axis of the bone in an arching or ‘bowed’ shape, which gradually straightened (Figure 6b). The echogenic areas, representing early ossification foci, were often not detected radiographically (Figuresda and b). Eventuafly, the cortex was fully repaired, and regained its continuity (Figures 5 and 6b). Even 1 year after the injury, at full clinical resolution, it was possible to detect distorted bone modelling around the previously fractured area, and residual angulation (Figure 6~). The cortex had returned to its normal highly echogenic appearance, with sharp smooth definition when compared to the echostructure of soft tissues. Fractures
in which
non-union
had taken
place
When poor bony union occurred, the sonographic appearances were useful in confirming the lack of a bony bridge, and showed either several bone fragments and poor alignment, or a disorganized echopattem with loss of the linear echogenic cortical border (Figures 7a, b).
Maffulli
and Thornton:
External
callus in long-bone
fractures
artifact
Subcutaneous and muscle
tissue
Figure 2. II, Same patient as Figure I after 5 weeks of traction. There is good bone alignment with early callus formation bridging across the distracted bone. The adjacent soft tissues show a low echogenic pattern. b, Radiograph of the fracture site at the same healing stage for comparison. Hardly any callus is visible around the fracture site. c, Diagram of US scan. C = cortex; S = skin; H = haematoma; Ca = early callus; F = fracture site.
Injury:
International
Journal
of the Care of the Injured
(1995)
Vol. 26/No.
I
Probe artifact
S = skin C z Cortex
b
Figure 3. a, Longitudinal scan of a femoral fracture site in a J-year-old boy, 2 weeks after the trauma. There is a slight displacement of the fracture ends and the adjacent soft tissue haematoma is echolucent. b, Diagram of US scan. S= skin; C =cortex; H= haematoma - low echogenic focal areas of liquefying haematoma; F = fracture site.
Discussion Studies on fracture healing in humans are lacking, as it is not ethically feasible to biopsy patients, and provide serial correlation of radiographic images with histological samplesx5. Conventional radiography, whilst involving a relatively low radiation dose, is not sensitive enough to be used in the early stages of fracture healing, when nonossified callus is being formed16. Sensitive non-invasive methods are therefore necessary to detect early callus formation. Recently, serial MRI scanning of closed tibia1 shaft fractures showed a characteristic pattern, which correlated well with the histology of fracture repaiP. The soft tissues surrounding the fracture gradually developed signals corresponding to the calcified area. Radiographs were able to detect calcification only several weeks later. In this preliminary
study, US images
can be correlated
b
Figure 4. a, Fracture site of Figures 1 and 2 8 weeks after the trauma. The US scan now begins to show new bone ‘filling in’ across the fracture site with adjacent exuberant but irregular callus. b, Diagram of US scan. S = skin; C = cortex; Ca = early callus - irregular callus with some calcification; F = fracture site.
with radiographic views in the various stages of fracture healing. However, no selection of patients was performed, and therefore the fractures scanned are heterogeneous. Further investigations are required to ascertain whether US can become part of the routine armamentarium of an orthopaedic surgeon to assess callus formation and bone healing. US is cheap and can be taken to the bedside without any loss in the quality of the image, an important advantage with a patient in traction. In our hands, it has shown a staged sequence of events in the patients studied, reflecting the present knowledge on the histology of fracture healing l,z. In fact, the peri- fracture haematoma is gradually invaded by osteogenic repair tissue, leading to gradual mineralization
of the callus which
eventually
fracture ends’sz. At present, however, qualitative morphological information.
bridges
the
US only gives Further efforts
Maffulli
and Thornton:
External
callus in long-bone
9
fractures
tions in the early phases of the fracture healing process, and when it is necessary to date a fracture”. It also gives useful information in cases of non-union.
References
Probe
artifact \
Subcutaneous
tissues F = Fracture
site
b
Figure 5. u, Fracture site shown in Figure 3, 12 weeks after the original trauma. The fracture site is just visible. At this stage, no soft tissue abnormality is seen. b, Diagram of US scan. S = skin; C = cortex; F = fracture site.
1 McKibbin B. The biology of fracture healing in long bones. ] Bone Joint Surg. [Br] 1978; 60B: 150. 2 Brighton CT and Hunt RM. Early histological and ultrashuctural changes in medullary fracture callus. ] Bone joint Swg [Am] 1991; 73A: 832. 3 Tiedeman JJ, Lippiello L, Connolly JF and Strates BS. Quantitative roentgenographic densitometry for assessing fracture healing. Clin Orthop 1990; 253: 279. 4 Laarsonen EM, Kyro A, Korhola 0 and Bostman 0. Magnetic resonance imaging of tibia1 shaft fracture repair. Arch Orthop Trauma Surg 1989; 108: 40. 5 Graif M, Stahl-Kent V, Ben-Ami T, Strauss S, Amit Y and Itzchak, Y. Sonographic detection of occult bone fractures. Pediatr Radio1 1988; 18: 382. 6 Katz R, Landman J, Dulifzky F and Bar-Ziv J. Fracture of the clavicle in the newborn. An ultrasound diagnosis. ] Ultrasound Med 1988; 7: 21. 7 Bartoli E, Saporetti N, Marchetti S. Ruolo dell’ecografia nella diagnosi della fratture clavicolari de1 neonate. Radiologia Medica 1989; 77: 466. 8 Pat-fen RM, Mack LA, Wang KY and Lingel J. Nondisplaced fractures of the greater tuberosity of the humerus: sonographic detection. Radiology 1992; 182: 201. 9 Steiner GM and Sprigg A. The value of ultrasound in the assessment of bone. Br ] Radial 1992; 65: 689. 10 Fornage BD, Richli WR and Chuapetcharasopon C. Calcaneal bone cyst: sonographic findings and ultrasound guided aspiration biopsy. ] Clin Ultrasound 1991; 19, 360. 11 Nath AK and Sethu AU. Use of ultrasound in osteomyelitis. Br ] Radio1 1992; 65: 649. 12 Maffulli N, Hughes T and Fixsen JB. Ultrasonographic monitoring of limb lengthening. ] Bone joint Swg [Br] 1992; 74B:
305.
Eyres KS, Bell MJ and Kanis JA. Methods of assessing new bone formation during limb lengthening. Ultrasonography, dual energy X-ray absorptiomehy and radiography compared. ] Bone Joint Surg [Br] 1993; 75B: 358. 14 Ricciardi L, Perissinotto A and Visentin E. Ultrasonography in the evaluation of osteogenesis in fractures treated with Hoffmann external fixator. llal 1 Orthop Truuma 1986; 12: 13
197.
should be made to quantify this information, so as to be able to apply in clinical practice. In conclusion, serial US scanning allows the study of the fracture healing process in uivo, with apparent good matching between images and what is known of the histology of fracture healing. US scanning is not the method of choice for the detection of bone fractures, although some reports claim a role in special caseP.17. At present, we do not know whether the method can be used to detect early signs of post-traumatic pseudarthrosis, thus influencing treatment choices. However, as it is noninvasive, is available in all hospitals, and is a low-cost investigation, US may supplement radiographic examina-
15 Chapman S. The radiological dating of injuries. Arch Dis Child 1992; 67: 1063. 16 Nicholls PJ, Berg E, Bliven FE and Kling JM. X-ray diagnosis of healing fractures in rabbits. Ckn Orthop 1979; 142: 234. 17 Leitbeg N, Bodenteich A, Schweighofer F and Fellinger M. Sonographische Frackturdiagnostik. Ulfraschall in Medizin 1990;
11:205.
Paper accepted
4 August
1994.
Reqtlestsfor reprints should be addressed to: N. Maffulli, Department of Orthopaedic Surgery, University of Aberdeen Medical School. Polwarth Building, Foresterhill, Aberdeen AB9 2ZD, Scotland, UK.
Injury:
Subcutaneous and muscle
International
Journal
of the Care of the Injured
i199.5) Vol. 26/No.
tissue
Probe artifact
Almost fracture
C
d
complete site still
bon
1
Maffulli
and
Thornton:
Subcutaneous and muscle,
External
callus
in long-bone
fractures
tissue
Pro be artifact
Figure6. a, Longitudinal scan of a femoral fracture in a o-year-old girl, 5 weeks after the trauma. There is mild displacement and angulation at the fracture site with early bridging echogenic callus which is irregularly aligned. b, Same as in Figure 6a, 13 weeks after the trauma. There is some minimal cortical irregularity with good bone union at the fracture site. Minimal soft tissue abnormality is seen at this stage. c, Diagram of US scan. S = skin, C = cortex; F = fracture site; Ca = callus. d, Diagram of US scan. S= skin; C = cortex; F= fracture site; Ca=callus. c, At 1 year, there is complete bony union with normal cortical border but mild persistent angulation. d, Radiograph at 1 year taken in the same alignment as in Figure 6e. for comparison. g, Diagram of radiograph. S = skin
Injury:
12
International
Journal
of the Care of the Injured
(1994)
Vol. 26/No.
1
Subcutaneous tissue and muscle \
artifact
S = Skin C =Cortex Ca q Callus
Figure 7. Q, Longitudinal scan of a non-union of a humeral fracture at 1 year. US shows grossly disorganized exuberant callus and disruption of the cortical margin. b, Radiographs at the same healing stage for comparison. c, Diagram of US scan. S = skin; C = cortex; Ca = callus.