A. Demeyere, MD
Department of Radiology, Imelda Hospital, Imeldalaan 9, 2820 Bonheiden, Belgium
F. M. Vanhoenacker, MD, PhD
Department of Radiology, University Hospital Antwerp, Wilrijkstraat 10, 2650 Edegem, Belgium
C O N T E N T S 7.1 Introduction 86 7.2 Pathophysiology 86 7.2.1 Anatomy of Bone 86 7.2.2 Remodeling of Bone 87
7.2.3 Pathophysiology of Stress Fractures 88 7.3 Mechanisms of Injury 88
7.3.1 Muscle Fatigue Theory 89
7.3.2 Overload or Increased Muscle Strength
Theory
897.3.3 Remodeling Theory 89 7.4 Risk Factors 89 7.4.1 Training Factors 89 7.4.1.1 Alteration of Training 89
7.4.1.2 Footwear and Training Surfaces 89 7.4.2 Lower Extremity Biomechanics 90 7.4.2.1 Pronated Foot 90
7.4.2.2 Cavus Foot 90 7.4.2.3 Varus Alignment 90 7.4.2.4 Limb Length Discrepancy 90 7.4.3 Systemic Factors 90
7.4.3.1 Bone Mineral Density (BMD) 90 7.4.3.2 Female Sex 90
7.5 Diagnostic Imaging 91 7.5.1 Radiography 91 7.5.2 Bone Scintigraphy 93 7.5.3 CT Scan 95
7.5.4 Ultrasound 96 7.5.5 MR Imaging 97
7.6 Sites of Injury and Relationship to Specifi c Mechanism of Trauma 98 7.6.1 Location 98
7.6.2 Compressive vs Tensile Type of Stress Fracture
987.6.3 Subchondral Insuffi ciency Fracture, SONK and Rapid Destructive Joint Disease 99 7.6.4 Longitudinal Stress Fracture 100
Things to Remember 101 References 101
Overuse Bone Trauma and Stress Fractures 7
Annick Demeyere and Filip M. Vanhoenacker
Box 7.1. Plain fi lms
● Always the initial fi rst imaging modality
● First signs of stress fractures in cortical bone are gray cortex sign, cortical striations and periosteal new bone formation
● Early signs in cancellous bone are subtle and include blurring and faint sclerotic radiopaque areas
Box 7.4. MR imaging
● Comparable sensitivity and higher specifi city than scintigraphy distinguishing bone involve- ment from soft-tissue injuries
● More accurate in grading the stage of stress fractures and predicting the time of recovery
● No radiation exposure, non-invasive Box 7.2. Bone scintigraphy
● Very sensitive but lacks specifi city
● Particularly useful for demonstration of multi- ple stress fractures, and distinguishing bipar- tite bones from stress fractures
● Too sensitive for stress reactions and subclini- cal bone remodeling
● Not useful for follow-up the healing
Box 7.3. CT scan
● Less sensitive
● Specifi c indication are complex anatomical
sites, longitudinal stress fractures and differ-
ential diagnosis with tumoral lesions
7.1
Introduction
Stress related bone injuries are common in athletes and account for up to 10% of cases in sports medicine practice. Stress fractures have been classifi ed into two types: fatigue and insuffi ciency. The fatigue fracture is caused by an abnormal stress to a normally elastic bone. Fatigue fractures are thought to occur in differ- ent sites depending on the age, sex and activity of the athlete. Insuffi ciency fractures arise from the applica- tion of a normal stress on a bone that is mineral defi - cient or abnormally inelastic. Insuffi ciency fractures are most prevalent in nutrient-defi cient (osteomala- cia) and older populations in whom osteoporosis and rheumatoid arthritis are more common (Romani et al. 2002; Anderson and Greenspan 1996). The focus of this chapter is sports-related injuries and so this latter will not be further discussed.
Stress fractures are common injuries frequently seen in athletes and military recruits. Although the reported incidence of stress fractures in the general athletic population is less than 1%, the incidence in runners may be as high as 20%. But with the increas- ing emphasis on exercise for the elderly and the rec- reational athletic population, stress fractures should not be overlooked in this population.
Although stress fractures have been described in nearly every bone, they are most common in the weight-bearing bones of the lower extremities. Spe- cifi c anatomic sites for stress fractures may be asso- ciated with individual sports, such as the humerus in throwing sports, the ribs in golf and rowing, the spine in gymnastics, the lower extremities in running
activities, and the foot in gymnastics (Bennell and Brukner 1997). In a review of 370 athletes with stress fractures, the tibia was the most commonly involved bone (49.1% of cases), followed by the tarsals (25.3%), the metatarsals (8.8%) (Boden and Osbahr 2000) and pelvis. Bilateral stress fractures occurred in 17%
of cases.
Pain in the lower leg brought on by exercise but relieved by rest is a common complaint. Stress inju- ries involving the tibia account for up to 75% of exer- tional leg pain, and encompass several clinical syn- dromes such as medial tibial syndrome (shin splints), chronic compartment syndrome, soleus syndrome, and stress fracture (Bhatt et al. 2000).
Accurate diagnosis of a stress lesion is essential in the early phase after the onset of pain to apply spe- cifi c treatment and to insure an early return to sports activity.
7.2
Pathophysiology
7.2.1
Anatomy of Bone
Bone has both cortical and cancellous components (Fig. 7.1). Cortical bone is dense and highly orga- nized and withstands stress in compression better than in tension. Cancellous (trabecular) bone is an irregularly shaped meshwork and withstands stress according to the alignment of the fi ber matrix. The outer shafts of long bones are mainly cortical, with
Fig. 7.1. Anatomy of bone. Bone is made up of cortical and
trabecular bone surrounded by periosteum. The cortical bone
unit is the osteon
a large percentage of cancellous bone making up the ends of the bone and the central portion of the shaft.
Short and fl at bones such as the tarsals and pelvis have a higher amount of cancellous bone.
The fundamental unit of cortical bone is the osteon (Fig. 7.2). In the osteon, concentric layers of lamellar bone surround small channels called Haver- sian canals. These canals house nerves and blood ves- sels. On the outside of the lamellae are small cavities, known as lacunae. Each lacuna contains a single bone cell, or osteocyte. Canaliculi form a transport system between the lacunae and the Haversian canals which are responsible for the nutrition and metabolic trans- port system within the bone.
center. Therefore, most of the remodeling in long bones takes place in the outer cortex. Remodeling consists of both resorption of existing bone by osteo- clasts and formation of new bone cells by osteoblasts.
Participating in regular activity promotes bone strength through proper perfusion of nutrients to the osteocytes and normal bone remodeling. Con- versely, a sedentary lifestyle contributes to bony atro- phy (Mori and Burr 1993).
Activation of osteoclasts is required for the start of the remodeling process. The piezoelectric effect is one mechanism implicated in the activation of bone remodeling. Tension forces create a relative electro- positivity on the convex, or the tension side, of the bone. Thus, as bending produces repeated distrac- tion forces at a focal point of a bone, the electroposi- tive charge may stimulate osteoclastic resorption (Romani et al. 2002).
The streaming effect is the movement of extracel- lular fl uids in the Haversian canals and canaliculi during deformation. If the surface charge on the Haversian canal or canaliculi walls is positive, nega- tive ions in the fl uid are attracted to the outside of the fl uid stream, creating a positively charged current in the middle. As bone is bent, the positive stream is forced toward the open or distracted surface of bone.
This electropositive stream may, in turn, stimulate osteoclastic activity. Other possible activators are bone ‘sensors’ that recognize increased and decreased mechanical strains, hormones, decreased venous fl ow and decreased oxygen.
Upon activation, osteoclastic cells form a cone and begin to secrete proteolytic enzymes to cut longitu- dinal tunnels through the bone. This tunnels corre- spond to the striations, which are seen on all imaging modalities (see below). These new Haversian canals are aligned with the stresses placed on the bone.
Each osteoclast cone can resorb nearly three times its volume in burrowing a canal 3–10 mm deep. The new Haversian canals are fi lled with osteoblasts that create a mineralized matrix supporting the walls of the new channel. The remaining space of the channel is then fi lled with immature lamellar bone.
Haversian canal formation and osteoblast sup- port with lamellar bone begins 10–14 days after the onset of remodeling. The conversion of lamel- lar bone into mature osteocytes cells lags behind resorption by about a week and may continue for as long as 20–90 days. The result is a temporarily weakened bone due to the new, hollow Haversian canals. The infl ammation of periosteum is designed to bolster the weakened area of bone until it can
Fig. 7.2. Anatomy of an osteon. An osteon is made up of con- centric lamellae and lacunae (containing an osteocyte) and include a central Haversian canal. The canaliculi form the transport system between the lacunae and the Haversian canals
The outer surface of long bones is surrounded by a highly vascular outer coating called the periosteum.
The periosteum is responsible for providing nutri- tion to the outer portion of the cortex and enlarges during remodeling to provide support to the cortex.
On the inner portion of the cortex, medullary canals allow the vascular passage for nutrients and blood vessels to the inner two-thirds of the cortex (Romani et al. 2002).
7.2.2
Remodeling of Bone
Bone constantly remodels in order to endure exter-
nal forces (Mori and Burr 1993). According to the
column law, the magnitude of stress is greatest on
the surface of a column and decreases to zero at the
mature. However, the periosteum does not mature until about 20 days after the remodeling process begins. This six- to ten-day lag between the deposit of immature lamellar bone and periosteal maturity leaves the bone temporarily weakened at the point of stress during the third week of remodeling. Con- tinued stress applied to remodeling bone during the
‘weak third week’ may lead to an accelerated break- down of the cortex. It is at this time that a stress fracture is most likely to develop.
7.2.3
Pathophysiology of Stress Fractures
Until recently, the cause of stress fractures was thought to be due to the breakdown of bone after repetitive loading. It has been estimated that, at normal physiologic levels of strain, it would require 10
8cycles of loading to produce failure of a weight- bearing bone such as the tibia. This level of loading is not easily attained, and stress fractures commonly occur soon after the onset of a stressful activity.
The rapid onset of symptoms and bone remodeling (sometimes within the fi rst seven days) consistent with stress fracture suggest that mechanical stress cannot be the only cause.
Recently, several studies demonstrate that im- paired bone perfusion may be the most important factor in the pathogenesis of stress fractures.
Otter et al. (1999) proposed that the perfu- sion and reperfusion of bone after a repetitive load causes a temporary oxygen debt to the area of bone being stressed. This ischemia, in turn, facilitates bone remodeling and subsequent bone weakness and stress fracture. When a bone is loaded to normal physiologic levels, the small blood vessels that supply the cortex are squeezed. In most cases, this pressure is necessary for proper movement of the blood. When the load is higher, the blood fl ow may be temporar- ily cut off. The result is a brief period of ischemia in the cells that would normally be perfused by the compressed medullary vessels. Repeated loads over a prolonged period of an activity, such as a long run, cut off the oxygen during that period as well. This decrease in oxygen to the bone is believed to trigger the remodeling process.
Restriction of venous fl ow without any mechani- cal loading is another mechanism for stimulation of bone remodeling (Romani et al. 2002).
Repeated pressure to the capillaries is also believed to cause microdamage to the vessels. As neutrophils
respond to plug the damaged capillaries, the blood fl ow through the vessels is further restricted. In addition, small leaks in the vessels allow fl uid fl ow into the surrounding tissue during reperfusion, further restricting the perfusion of oxygen into the cells. This leaking increases with subsequent bouts of loading, worsening ischemia and triggering a fur- ther increase in remodeling. The repetition of this cycle causes an increase in remodeling, a breakdown in the cortex, a weakening of the bone, and poten- tially a stress fracture (Romani et al. 2002; Otter et al. 1999).
Direct damage to the extracellular matrix may also initiate angiogenesis (Globus et al. 1989). Cul- tured platelets have recently been shown to stimu- late osteogenic activity by releasing growth factors (Slater et al. 1995). Clearly, bone remodeling and angiogenesis are intimately related, and the major stimulus initiating each process may be altered bone perfusion, rather than an altered strain environment by itself.
7.3
Mechanisms of Injury
A stress fracture is a partial or incomplete fracture caused by the accumulation of stress forces to a local- ized area of bone. Stress fractures are not the result of a single insult. Instead, they arise as the result of repetitive applications of stresses that are lower than the stress required to fracture the bone in a single loading (Romani et al. 2002).
Bone endures a stress whenever a force is loaded upon it. Whether the stress comes from the pull of a muscle or the shock of a weight-bearing extrem- ity contacting the ground, it is defi ned as the force applied per unit area of the loadbearing bone. Low levels of these forces cause bone to deform, which is known as strain. The stress-strain response of bone depends on the direction of the load; the geometry, microarchitecture, and density of bone; and the infl u- ence of surrounding muscular contractions.
In most activities of daily living, when the force is
removed, the bone elastically rebounds to its original
position. The force that a bone can endure and still
rebound back to its original state without damage is
within the elastic range. Forces that exceed a criti-
cal level above the elastic range are in the plastic
range. Once forces reach the plastic range, a lower
load causes greater deformation; it is at this level that forces accumulate to cause permanent damage to bone.
Forces can be applied to bone through compres- sion, tension, bending, torsion or shear. Compres- sion forces are generally seen in cancellous bones, such as the calcaneus and femoral neck. Tension forces result in bone pulling away from bone, as is common in compact bones such as the tibia en femur. As the load is applied to the bony shaft through a bend; a tension strain is placed upon the convex surface of the shaft and compressive forces act on the concave side.
Several theories have been established to explain the development of stress fractures and it is likely a combination of the following factors that ultimately results in bone fatigue and fracture (Haverstock 2001; Spitz and Newberg 2002; Anderson and Greenspan 1996).
7.3.1
Muscle Fatigue Theory
One of the major roles of muscle is to minimize the tensile stress on bone (Spitz and Newberg 2002).
The muscles decrease bending or tensile forces in bone and increase compressive forces. Because bone is more resistant to force in compression than ten- sion, the supporting muscles help prevent fatigue fractures. When the supporting structures fatigue, the tensile forces increase, rendering bone failure more likely (Ross 1993). The muscles also have a shock absorption function by dissipating forces away from bone (Haverstock 2001; Anderson and Greenspan 1996). Accordingly, fatigue of muscles in the poorly conditioned athlete creates increased tensile stress on bone and decreased shock-absorb- ing capacity, resulting in stress fracture.
7.3.2
Overload or Increased Muscle Strength Theory Another explanation of stress fractures relates to increased muscle strength. Under normal conditions, when a new stress is applied, muscle tone is achieved more quickly than bone strengthening. This results in a mechanical imbalance, with muscle exerting an excess of load on bone, resulting in bone fatigue (Spits and Newberg 2002; Haverstock 2001;
Anderson and Greenspan 1996; Ross 1993).
7.3.3
Remodeling Theory
The physiologic response of bone to stress is also important in the pathophysiology of stress injury. Bone is a dynamic tissue that requires stress for normal development, and it undergoes constant remodeling in response to changing environmental forces. Initially, osseous remodeling manifests as osteoclastic activity and resorption of lamellar bone. This is subsequently replaced by denser, stronger osteonal bone. In repeti- tive stress overload, however the accelerated remodel- ing results in an imbalance between bone resorption and bone replacement, leading to weakening of the bone. Continued stress results in further imbalance, leading to bone fatigue, injury, and fracture. Osseous stress injury is not an all-or-none phenomenon, but a physiologic continuum ranging from normal osseous remodeling, to accelerated remodeling with fatigue and early injury, to frank stress fracture (Spits and Newberg 2002; Haverstock 2001; Ross 1993).
7.4
Risk Factors
7.4.1
Training Factors
7.4.1.1
Alteration of Training
Stress fractures often result from participation in a new activity or from an alteration in a training pro- gram such as an increase in intensity, duration, or frequency of training in a short time span. Stress fractures related to a single training session have been described as well (Haverstock 2001).
7.4.1.2
Footwear and Training Surfaces
Inadequate footwear that has lost its shock absorp- tive properties does not provide suffi cient cushioning increasing the risk for fatigue fracture.
A hard running surface also can be a factor, par- ticularly if the runner’s footwear is inadequate or he/
she is running in a state of fatigue when the skeletal
structure is absorbing the impact of the running sur-
face (Haverstock 2001).
7.4.2
Lower Extremity Biomechanics
7.4.2.1 Pronated Foot
The pronated foot with a rearfoot valgus position produces increased stress on the lateral malleolus.
When the foot pronates through the midstance phase of gait into the toe-off phase, the fi rst metatarsal is hypermobile and dorsifl exes, resulting in an increased pressure load on the second metatarsal. If the second metatarsal is short, then increased load is transferred to the third metatarsal (Haverstock 2001).
7.4.2.2 Cavus Foot
The cavus foot is generally rigid in nature and the shock absorptive properties are poor. The plan- tarfl exed forefoot results in increased pressure on the metatarsals. A rigidly plantarfl exed fi rst metatarsal is placed under increased stress, which places the sesamoid complex at risk of stress fracture. A long or plantarfl exed lesser metatarsal is exposed to a greater load and is subjected to fatigue failure. Limited sub- talar joint motion produces increased pressure on the calcaneus and renders it susceptible to a stress fracture (Haverstock 2001).
7.4.2.3
Varus Alignment
Varus alignment such as genu varum, tibia vara, sub- talar varus and forefoot varus is associated with lower extremity stress fractures, 49% of which are located at the tibia. The resulting excessive pronation increases the eccentric work that must be done by the medial aspect of the soleus muscle and enforces the strain on the posteromedial side of the tibia ( Krivickas 1997).
7.4.2.4
Limb Length Discrepancy
Limb length discrepancy is a relatively common skeletal malalignment with predisposition to stress fractures. It has been reported that runners with a leg length discrepancy experienced stress fractures more often than runners without. During running, the short leg rotates externally to increase stability and overstriding may occur. The foot on the short
side is subjected to greater force over a short period of time (Haverstock 2001).
7.4.3
Systemic Factors
7.4.3.1
Bone Mineral Density (BMD)
The BMD increases with age, up to about the age of 35 years, and slowly decreases thereafter. Especially in the premenarcheal years, the increase in BMD and exercise-induced bone modeling is most signifi cant (Morris et al. 1997).
Lauder et al. (2000) found a strong positive asso- ciation between the body mass index and the BMD and the probability of stress fractures. Low body weight is a well-known risk factor associated with low BMD in women, especially female athletes.
Although exercise has been reported to increase BMD, it is well established that stress fractures occur more frequently in subjects who exercise intensely.
Consistent with these fi ndings, Lauder et al. (2000) described a gradual increase in BMD as the hours of exercise per week increased. At the same time, the per- cent of women with stress fractures increased from 12% among those performing 5 h or less of exercise per week to 50% among those exercising 10 h or more per week.
Other factors causing a low BMD are Caucasian race, any condition that results in altered gait or an ipsilateral joint prosthesis and systemic factors (Cohn et al. 1977; Anderson and Greenspan 1996;
Hester 2001; Romani et al. 2002). These are sum- marized in Table 7.1. As most of these factors are not seen in sports medicine, further discussion is beyond the scope of this chapter.
7.4.3.2 Female Sex
A high incidence of stress fractures has been reported
in women (Boden and Osbahr 2000; Anderson and
Greenspan 1996; Romani et al. 2002; Bennell and
Brukner 1997). They seem to occur more commonly
in the sacrum, pelvis and femoral neck. Therefore,
it is especially important to investigate intrinsic
factors in female athletes. The term ‘female athlete
triad’ refers to the combination of an eating disor-
der, amenorrhea, and osteoporosis. In an effort to
minimize body fat and maintain high athletic perfor-
mance, many young women develop eating disorders during puberty. Amenorrhea and oligomenorrhea are common fi ndings in competitive female distance runners. The resultant estrogen-defi cient state leads to decreased bone mineral density and an increased risk of stress fractures.
Bennell et al. (1996) showed in his study that age of menarche and calf girth were the best independent predictors of stress fractures in women. Their bivari- ate model correctly assigned 80% of the female ath- letes into their respective stress fracture or nonstress fracture groups. Their results suggest that it may be possible to identify female athletes most at risk for this overuse bone injury.
7.5
Diagnostic Imaging
Bone response to stress is evaluated on a dynamic continuum between early remodeling and perios- titis to a cortical stress fracture. It is important to note that these bony changes refl ect a wide spec- trum of physical fi ndings and radiographic pre- sentations.
7.5.1
Radiography
Plain radiography plays an important role in the initial work-up of a suspected stress fracture. They can be used to confirm the diagnosis at a rela- tively low cost. Unfortunately, initial radiographs are often normal, which is not surprising given the degree of microscopic remodeling that occurs in the early stages of stress injury. The sensitiv- ity of early radiographs can be as low as 15%, and follow up radiographs will demonstrate diag- nostic findings in only 50% of cases (Nielsen et al. 1991). The lag time between manifestation of initial symptoms and detection of radiographic findings ranges from one week to several months, and cessation of physical activity may impede the development of typical plain radiographic findings (Fig. 7.3) (Anderson and Greenspan 1996; Spitz and Newberg 2002).
In addition to the time course, radiographic changes are dependent on the type of bone involved (Table 7.2).
Initial changes in the cortical bone are the so called “gray cortex” sign seen as a subtle ill defini- tion of the cortex (Mulligan 1995) or faint intra- cortical radiolucent striations, which are presum- ably related to the osteoclastic tunneling found early in the remodeling process ( Muthukumar et al. 2005; Daffner and Pavlov 1992). These changes may be easily overlooked until periosteal new bone formation and/or endosteal thickening develops in an apparent attempt to buttress the temporar- ily weakened cortex. As damage increases, a true fracture line may appear. These injuries typically involve the shaft of a long bone and are common in the posterior portion of the tibia in runners (Fig. 7.4).
Stress injuries in cancellous bone are notori- ously diffi cult to detect. Subtle blurring of trabec-
Table 7.1. Systemic factors that infl uence the development of (insuffi ciency) stress fractures
Metabolic disorders - Osteoporosis - Osteomalacia
- Hyperparathyroidism, Cushing’s syndrome, hypothyroid- ism, renal insuffi ciency
- Rickets - Scurvy
Infl ammatory conditions - Rheumatoid arthritis - Osteomyelitis Bone dysplasias
- Osteogenesis imperfecta
Neurological disorders
- Neurotrophic diseases
- Poliomyelitis
Paget’s disease
Pharmacological
- Corticosteroids
- Diuretics
- Anticonvulsants
Nutritional defi ciencies
Sleep deprivation
Smoking
ular margins and faint sclerotic radiopaque areas may be seen secondary to peritrabecular callus, but a 50% change in bone capacity is required for these changes to be radiographically detectable.
With progression, a readily apparent sclerotic band will be seen. Common sites for cancellous lesions
Table 7.2. Radiologic grading system for stress injuries: the correlation between histology, radiography, bone scintigraphy and MR imaging
Stress reaction Grade 1 Grade 2 Grade 3 Grade 4
Histology
- Periosteal bone and
cortical tunneling
Cortical resorption more than periosteal reaction
Extensive cortical tun- neling and periosteal reaction
Trabecular microfrac- tures, granulation tissue and necrotic areas
Radiography:Cortical bone
- - Cortical striations,
gray cortex sign
Periosteal and endos- teal new bone forma- tion
True fracture line
Radiography:
Cancellous bone
- - Blurring of trabecu-
lar margins, faint sclerotic densities
Sclerotic band True fracture line
Scintigraphy
Amorphous lesion in the bone marrow
Small, ill-defi ned lesion with mildly increased activity in the cortical region
Larger well-defi ned, elongated lesion with moderately increased cortical activity
Wide fusiform lesion with highly increased activity in the cortico- medullary region
Wide extensive lesion with intensely increased activity in the transcor- ticomedullary region
MR imaging
Ill-defi ned zone of bone marrow edema (only on T2W)
Mild periosteal edema, without bone marrow edema (only on T2W)
Moderate to severe periosteal edema and bone marrow edema (only on T2W)
Moderate to severe periosteal edema and bone marrow edema (on both T1W and T2W)
Moderate to severe periosteal edema and bone marrow edema,
‘low signal’ fracture line (on both T1W an T2W) Fig. 7.3a–c. Metatarsal stress fractures on radiography. a Subtle faint periosteal reaction on the lateral aspect of the diaphysis of the second metatarsal in a 42-year-old patient (stress fracture grade 2). b The plain fi lm of the same patient one month later shows obvious more periosteal new bone formation (stress fracture grade 3). c Stress fracture of the second metatarsal from another patient (30-year-old) with a true fracture line and periosteal new bone formation (grade 4)
a b c
include the calcaneus (Fig. 7.5), proximal tibia, distal tibia and fi bula, pelvis and femoral neck (Fig. 7.6) (Anderson and Greenspan 1996; Spitz and Newberg 2002).
Sometimes stress fractures may mimic tumor and
tumorlike conditions (see Chap. 8).
Fig. 7.4. Radiograph of a stress fracture of the tibia shows a sclerotic band perpendicular to the trabeculae
Fig. 7.6. Example of a tensile stress fracture of the femoral neck on a coronal T2-weighted MR image with fat suppression
Fig. 7.5. a Radiograph of a stress fracture of the calcaneus shows nicely a sclerotic band in de posterior calcaneus per- pendicular to the trabeculae. b Another calcaneus stress frac- ture from another patient on sagittal T1-weighted image a
b
7.5.2
Bone Scintigraphy
Bone scintigraphy is an effective modality in the evaluation of athletes with clinically suspected osse- ous stress injuries. Before the advent of MR imaging
it had been the gold standard for evaluating stress fractures (Zwas et al. 1987).
Bone scintigraphy demonstrates abnormal fi ndings
early in the continuum of the stress response in bone,
by detecting the increased bone metabolism and osteo-
blastic activity associated with osseous remodeling.
Scintigraphy is typically abnormal one to two weeks or more before the radiographic changes appear.
Bone scintigraphy should optimally be performed using a three-phase technique. The most widely used radiopharmaceuticals for skeletal imaging are the technetium-99m phosphate analogues; these are taken up at sites of bone turnover. The degree of uptake depends primarily on the rate of bone turn- over and local blood fl ow, and abnormal uptake may be seen within 6–72 h of injury (Anderson and Greenspan 1996).
The three-phase technique can help differentiate between soft tissue injury and osseous injury (Spitz and Newberg 2002). In the fi rst phase, the blood fl ow phase, imaging is performed by acquiring dynamic 2- to 5-s images over the area of clinical concern for 60 s after the bolus intravenous injection. In the second phase, the blood pool or soft tissue phase, imaging is acquired after 5 min. The fi nal phase of imaging is the delayed skeletal phase. These images should be acquired approximately 2–4 h after injection to max- imize clearance of the radiopharmaceutical from the overlying soft tissues.
Zwas et al. (1987) described a scintigraphic classifi - cation of stress fractures ranging from grade 1 (mild) to grade 4 (severe) according to size, extent, and tracer concentration in the lesions (see Table 7.2.) (Fig. 7.7).
Despite its high sensitivity, the specifi city of scin- tigraphy is slightly lower than that of radiography
because other conditions such as tumors, infec- tion, bone infarction, and shin splints or periostitis can produce similar fi ndings (Fig. 7.8). This can be improved by using the three-phase technique. The localisation and distribution of the tracer accumu- lation and the timing can help in the differentiation (Nielsen et al. 1991; Zwas et al. 1987). For example, stress fractures are hot within all phases and have a more focal or fusiform morphology, whereas tracer uptake in shin splints is limited to the delayed phase and has a linear aspect.
On the other hand, bone scintigraphy is more sen- sitive than magnetic resonance, especially in evalu- ating suspected lesions in the spine or pelvis, iden- tifying multiple stress fractures, and distinguishing bipartite bones from stress fractures.
However, scintigraphy may be too sensitive; as many as 50% of scintigraphically positive fi ndings can occur at asymptomatic sites. Areas of increased uptake may represent subclinical sites of bone remod- eling or stress reactions.
Although bone scanning is good for detection of stress fractures, it is not useful for follow-up of heal- ing, because the intensity of activity decreases over 3–18 months as the bone remodels and often lag- ging behind clinical resolution of symptoms (Boden and Osbahr 2000). Since these entities are usually treated differently, CT or MR imaging can be helpful for better delineation.
Fig. 7.7. Longitudinal stress fracture grade 4 with a wide trans- corticomedullary increased activity on scintigraphy
Fig. 7.8. Bone scintigraphy of shin splints dem-
onstrates faint linear hot spots on the postero-
medial aspect of both tibiae
the injury. A fracture that courses longitudinally may be particularly well demonstrated with CT in com- parison to one coursing transversely to the long axis of the bone. This drawback of axial CT scanning may be overcome by multidetector CT with appropriate reformatting.
Gaeta et al. (2005) demonstrated that CT had the best performance in the identifi cation of corti- cal abnormalities. Early CT changes consist of osteo- penia, resorption cavities and striations within the cortical bone. They believe that these changes may precede formation of a cortical fracture.
CT fi ndings in stress fracture are sometimes non- specifi c, but fi ndings of endosteal and periosteal reaction surrounding a thin cortical translucency are diagnostic of stress fracture. In longitudinally stress 7.5.3
CT Scan
CT is limited in its ability to detect early osseous stress injuries, and is less sensitive than bone scin- tigraphy and MR imaging. It does, however, have a role in more advanced injuries and injuries in spe- cifi c anatomic locations where radiography fails to detect early changes due to superposition of adjacent structures. CT is particularly well suited for stress fractures of the tarsal bones (Fig. 7.9), longitudinal stress fracture of the tibia (Fig. 7.10), pars interar- ticularis stress fractures (spondylolysis), and stress fractures in the sacrum.
Whether the fracture is actually seen at CT may depend on orientation and the stage in evolution of
Fig. 7.9a–d. Navicular stress fracture with a true fracture and diastasis on CT and MR-imaging in a 31-year-old women. Note the parallel ori- entation of the fracture line with the bone trabeculae. a Axial CT image.
b Axial T1- weighted MR image. c Axial T2-weighted MR image with fat saturation. d Coronal CT image
a b c
d
fractures CT has a particular interest in demonstrat- ing a fracture line that extends along the axis of the bone (Allen 1988).
Additionally, CT may be problem solving when fi ndings are equivocal on radiographs, MR imaging or scintigraphy. The value of CT in this regard lies in the detection of a discrete lucent or sclerotic frac- ture line or periosteal reaction. CT is also extremely helpful in differentiating between stress fracture and osteoid osteoma, because both entities may be hot on bone scan, show bone marrow edema at MR imaging, and demonstrate sclerosis on radiographs. The pres- ence of a radiolucent nidus – however – is diagnostic for osteoid osteoma.
CT has also proven its value in the diagnosis of pediatric stress fractures. Initial radiographs may demonstrate marked periosteal bone formation, which may mimic tumor. Demonstration of endos- teal bone formation on CT often leads to the correct diagnosis (Anderson and Greenspan 1996).
7.5.4 Ultrasound
Ultrasound is not reliable in the diagnosis of stress fractures and therefore is not recommended as the imaging technique of choice. Literature available
Fig. 7.10a–e. Longitudinal stress fracture of the tibia in a 41-year-old man. a Radiograph demonstrates periosteal new bone formation on the posterior aspect of the tibia diaphysis without a clearly lucent fracture line. b Scintigraphy shows an extensive long linear increased activity with a transcorticomedullary extension. c Axial CT-image shows nicely the vertical oriented true fracture line and the periosteal new bone formation. d Axial T2-weighted MR image with fat suppression shows a hyperintense transcortical line on the posterior aspect of the tibia with periosteal edema and hypointense new bone formation. e Hyperintense transcortical fracture line and hypointense new bone formation on an axial T1-weighted MR image
a
c d e
b
about the use of ultrasound is scarce. Ultrasound fi ndings consist of a thickened hypoechogenic peri- osteum (Fig. 7.11), hyperechogenic cortical irregular- ities compatible with new bone formation (Fig. 7.12) or cortical discontinuity compatible with a fracture (Van Holsbeeck and Introcaso 1991). According to Boam et al. (1996), a sensitivity of 43% and a specifi city of 49% was reported.
MR imaging is helpful in grading the stage of certain stress fractures (Table 7.2) and, therefore, predicting the time to recovery. In addition, MR imaging avoids radiation exposure, is non-invasive and requires less time than three-phase bone scintigraphy (Boden and Osbahr 2000).
The MR imaging sequences necessary for evaluat- ing stress injuries are summarized in Table 7.3. MR- imaging should also be performed in multiple orthog- onal planes (Anderson and Greenspan 1996).
Fig. 7.11. Ultrasound in shin splints demonstrates a hypoechoic thickened periosteum on the anterior aspect of the tibia
Fig. 7.12. The hyperechoic cortical irregularity on ultrasound corresponds to the new bone formation in a healing fracture
7.5.5 MR Imaging
Magnetic resonance imaging is a valuable tool in identifying stress fractures in case of high index of suspicion. MR has a comparable sensitivity and higher specifi city than scintigraphy in distinguishing bone involvement from soft-tissue injuries. Moreover,
Table 7.3. Recommended MR-sequences for stress injury
Sequence Characteristic detectionT1-weighted sequence Anatomy and more advanced
stress-related fi ndings Edema-sensitive sequence:
- Short tau inversion recov- ery (STIR)
- T2-weighted with fat sup- pression
Detection of the earliest changes of stress reaction, such as periosteal, muscle or bone marrow edema
MR imaging fi ndings will differ along with the stage of the lesion and the sequences used. The MR grading system, developed by Fredericson et al.
(1995) corresponds very well with the scintigraphic grading system of Zwas et al. (1987).
As discussed earlier, both resorption and replace- ment of bone constitute the most early histological changes of stress injury to bone. This stress reaction is accompanied by local hyperemia and edema, which is easily depicted by signal changes on MRI, especially on fl uid sensitive sequences. This is why MR can be regarded as an excellent modality for the detection of early stress injury (Table 7.2) (Zanetti et al. 2002).
Indeed, MR can detect subtle cortical abnormali- ties like on CT on both T2-weighted and fast STIR images. The resorption cavity appears as a round or oval area of high signal intensity, and striation or cor- tical tunneling appears as a slightly hyperintense line extending through only a part of the cortical thick- ness. Usually, multiple parallel striations can be seen within the cortex. Osteopenia appears as an area of intermediate signal intensity. Irregularity of subperi- osteal bone can be seen as well.
The appearance of resorption cavities and stria-
tions within the cortex may correlate with osteoclas-
tic proliferation. High signal intensity depicted on
T2-weighted MR images within striations and resorp-
tion cavities may be well explained by cell accumula-
tion (osteoclasts, osteoblasts, and fi broblastic cells),
increased amount of blood vessels, and production of
connective tissue and osteoid matrix, which all occur in osteoporotic bone (Gaeta et al. 2005).
MR also allows depiction of periosteal and endos- teal marrow edema, which are believed to be useful ancillary markers for stress injury (Schweitzer and White 1996; Fredericson et al. 1995).
Especially in cancellous bone, stress injuries result in non-specifi c bone marrow alterations simi- lar to those described in patients with bone marrow edema syndrome of bone bruise (Anderson and Greenspan 1996) (see Chap. 6).
The most common pattern of a frank fatigue-type fracture is a fracture line that has a low signal on all pulse sequences, surrounded by a larger, ill-defi ned zone of edema. The fracture line is continuous with the cortex and extends into the intramedullary space oriented perpendicular to the cortex and the major weight-bearing trabeculae.
Fredericson et al. (1995) also concluded that MR imaging is more accurate than bone scan in correlat- ing the degree of bone involvement with clinical symp- toms, allowing for more accurate recommendations for rehabilitation and return to activity (Fig. 7.13).
The term shin splints has been used to describe the clinical entity of activity related lower leg pain, typically associated with diffuse tenderness along the posteromedial tibia. A similar entity in the upper leg has been described as thigh splints or adductor inser- tion avulsion syndrome (Anderson et al. 2001; Van de Perre et al. 2003). Recent MR imaging studies have suggested that shin splints are a part of the continuum of fatigue damage in bone (Anderson et al. 1997;
Fig. 7.13. Example of a grade 4 stress fracture of the second metatarsal on a sagittal T1-weighted MR image
Fredericson et al. 1995). On T2-weighted images, a linear abnormal high signal is seen along the postero- medial surface of the tibia representing subperiosteal edema (Fig. 7.14). This implies that one cause of shin splints may be traction periostitis along the insertion of the soleus fascia and tibialis posterior. It can also appear as a linear high signal along the subcortical aspect of the bone marrow. This is considered to be secondary to edema or hemorrhage related to micro- damage and the associated reparative response (Aoki et al. 2004). In thigh splints the periosteal reaction is located at the proximal third of the medial femoral shaft near the insertions of the adductor brevis and longus muscles (Anderson et al. 2001).
7.6
Sites of Injury and Relationship to Specifi c Mechanism of Trauma
7.6.1 Location
Although most common in the lower extremity, stress injury to bone and stress fractures have been reported in nearly every bone in the body. This pos- sible sites of stress injury are summarized in Table 7.4 (Resnick 1996).
7.6.2
Compressive vs Tensile Type of Stress Fracture There are different types of stress fractures. The vast majority of stress fractures are perpendicular to the trabeculae, in cortical bone as well as in cancellous bone (Fig. 7.5). This type can be at the compressive side or at the tensile side, as seen in the femoral neck or tibial shaft (Fig. 7.15).
In the more common compression stress fractures of the femoral neck, that is more frequent in younger patients; the injury begins at the inferior cortex of the femoral neck. The lateral or tensile stress fracture, more common in older patients, starts in the superior cortex of the femoral neck and may advance across the femo- ral neck as a fracture line perpendicular to the axis of the femoral neck (Fig. 7.6) (Spitz and Newberg 2002;
Boden and Osbahr 2000; Resnick 1996).
The tibial shaft is the most common location of
stress fractures in athletes. Depending on the patient
population, the incidence may range from 20% to 75% of all stress fractures. Tibial stress fractures may occur at any site along the shaft of the bone, but are most frequently encountered in the posteromedial cortex (compression side), in either the proximal of the distal third of the shaft (Fig. 7.16). A less common tibial stress fracture occurs on the tension side of the bone or anterior cortex of the midshaft of the tibia (Krivickas 1997).
7.6.3
Subchondral Insuffi ciency Fracture, SONK and Rapid Destructive Joint Disease
A more special form to be mentioned is the spontane- ous osteonecrosis of the knee (SONK), mostly occur- ring in the area of maximum weight-bearing of the
Table 7.4. Possible sites of stress injuries Upper extremity:
– Phalangeal tufts
– Carpal bones: hook of hamate
– Ulna: coronoid, olecranon process, diaphysis – Humerus: distal diaphysis
– Scapula: coracoid, inferior edge of glenoid fossa Lower extremity:
– Sesamoids – Metatarsals – Navicular – Talus – Calcaneus
– Tibia: mid and distal diaphysis, proximal shaft, medial mal- leolus
– Fibula: distal and proximal diaphysis – Patella
– Femur: diaphysis, neck
– Pelvis: obturator ring, pubic rami, sacrum Spine:
– Lumbar vertebra: pars interarticularis – Lower cervical, upper thoracic spinous process Thorax:
– Ribs – Clavicle – Sternum
aFig. 7.14a–c. Shin splints. a Axial T2-weighted MR image with fat saturation demonstrates a discrete hyperintense signal on the posteromedial aspect of the tibia represent- ing subperiosteal edema in shin splints in a 28-year-old man. b Axial T2-weighted MR image with fat suppression from the tibia in another patient shows minimal sub- periosteal edema with subcortical bone marrow edema.
c Radiograph of the same women demonstrates no “gray cortex” sign or periosteal reaction
b
c
