Ultrasound in the Management of Carpal Tunnel Syndrome
CTS is normally diagnosed with a thorough clinical history and examination and the addition of electrophysiology (EP) studies if necessary. More recently, the use of US in the diagnosis has been reported in a number of studies showing that sensitivity and specificity are approaching that of EP studies. US is highly acceptable to patients, with ease of use in the consultation room, and provides a capability to assess anatomical aspects of the carpal tunnel and guide treatment.
A thorough clinical history and examination are the most crucial elements in the diagnosis of CTS. Studies have been criticized for using EP criteria alone for inclusion, as they cannot fully exclude the diagnosis of CTS. Consensus opinion by experts in the field is that clinical diagnosis should be made, independent of EP studies, in both the research and the clinical setting. The American Academy of Neurology (AAN) has produced practice parameters for CTS with diagnostic criteria for history and physical examination and these are listed in Table 1.
Practice parameters for the performance of electrodiagnostic tests for CTS have been set out by a joint report of the American Association of Electrodiagnostic Medicine, American Academy of Neurology and American Academy of Physical Medicine and Rehabilitation. They also performed an extensive literature review and reported a sensitivity and specificity of >85% and >95%, respectively, for median sensory and motor NC studies when compared with clinical diagnosis. However, the same review found that EP studies missed the diagnosis of CTS in 16–34% of patients with clinically defined disease. Other more recent reviews put the sensitivity at 85–90% and the specificity at 82–85% and recognize that EP studies alone should not be used as the standard for diagnosis.
US is now well established as a diagnostic tool in CTS. There are many advantages to US, including that it is readily available, non-invasive, has a shorter examination time and can be used to assess a number of parameters of the median nerve such as size, vascularity (using power Doppler) and mobility (using dynamic imaging). In addition, US provides information on anatomical variations of the median nerve and surrounding structures that may be a causative factor in CTS. On US imaging at the wrist, the median nerve is easily visualized in transverse view as hypoechoic nerve fibres with hyperechoic rims immediately superficial to the flexor tendons, with the hyperechoic FR overlying it within the carpal tunnel. Fig. 1 demonstrates the probe position for examination of the median nerve and US images of the median nerve in CTS.
(Enlarge Image)
Figure 1.
Images of probe position and US in carpal tunnel syndrome
(A) Demonstration of probe position on the antral wrist for examination of the median nerve. (B) View of carpal tunnel with surrounding structures. MN: median nerve; FCR: tendon of the flexor carpi radialis; Sc: scaphoid bone; UA: ulnar artery; Pi: pisiform bone. (C) Tracing method to measure cross-sectional area of enlarged median nerve measuring 18 mm in a male with CTS. (D) Longitudinal view of the median nerve showing enlargement as it enters the carpal tunnel in a female patient. CTS: carpal tunnel syndrome.
A number of studies have examined the parameters of the median nerve that are most useful in diagnosing CTS. It has been shown that the cross-sectional area (CSA) of the median nerve is significantly greater in those with CTS compared with healthy controls. The CSA has also shown good concordance with NC studies for CTS severity when US cut-off points for CSA are used to discriminate between grades. In those with normal NC studies but a clinical picture of CTS, the CSA on US has been shown to be significantly larger than normal controls. Hypervascularity and hypoechogenicity of the nerve are also present in those with larger CSA, and the probability of having CTS has been stated to be 90% in those with normal EP studies and all three of these features present. Although it is recognized that the median nerve often becomes hypoechoic as it enlarges, the diagnostic accuracy of this remains uncertain. The thickness of the FR and the thenar muscles has also been shown to correlate with CTS diagnosed by EP studies, and these parameters can be used in combination to increase sensitivity and to assess patients with normal EP studies.
Two other aspects of US assessment have attracted much debate: (i) which measure of the median nerve is best to establish the diagnosis and (ii) the best cut-off size for the CSA of the median nerve to diagnose CTS. There is a considerable body of evidence to indicate that the CSA at the level of the pisiform bone or the tunnel inlet is the most sensitive and specific US finding in patients with CTS. As can be seen in Table 2, the majority of studies use the tunnel inlet/pisiform bone as the site for measurement of the CSA. There is much less agreement on the best cut-off size for the CSA, with recommended cut-offs varying from 6.5 to 15 mm. Table 2 lists the sensitivity and specificity from 22 studies and includes the CSA cut-offs used and the location where the CSA was measured. These findings demonstrate the wide variation in sensitivity (62–97.9%) and specificity (57.1–100%) seen in studies comparing US with clinical assessment or EP tests. Most of the studies using clinical assessment as the reference standard do not document the sensitivity and specificity of EP studies. In the three studies that did, sensitivity was reported as 78%, 80% and 82% and specificity was reported as 83%, 84% and 97%. The reference standards used for comparing US and EP studies are included in Table 2.
Limited research has been carried out on inter- and intrareader reliability of US. The intrareader reliability for CSA of the median nerve is high in the studies identified that looked at this. There are two methods for measuring the CSA of the median nerve. The tracing method involves tracing a continuous line around the inner hyperechoic rim of the median nerve, then machine software is used to calculate the CSA. The second method involves measuring the anteroposterior and transverse distances of the median nerve, which are then input into the ellipse formula to calculate the CSA. Fig. 1 demonstrates the tracing method. There is strong correlation between these two methods. When measurements of the median nerve of an amputated limb obtained by US using both methods were compared with direct measurements made later on frozen section, correlation was found to be 0.992 for the anteroposterior and transverse distances and 0.982 for the direct measurement. Interreader reliability for measurement of the CSA at the tunnel inlet using both tracing and ellipse formula methods shows good reliability, with correlation coefficients of 0.81 and 0.97, respectively. However, the interreader reliability was poor at the tunnel outlet, which probably relates to the orientation of the median nerve at the tunnel outlet where it moves more dorsally, making good visualization and measurement difficult.
A few studies have reported data on the percentage of patients with a clinical diagnosis of CTS with normal CSA on US but positive NC studies. These studies used variable criteria for US diagnosis and are not directly comparable. Depending on the US criteria applied, NC studies were positive in 8.5%, 21% and 28.2% of patients with a clinical diagnosis but normal US in three studies identified. One of the main issues with studies looking at the diagnosis of CTS with US is the difference in parameters used, making comparison difficult. Recent evidence-based guidelines have been published by a panel of experts specializing in neurology, physical medicine and rehabilitation and radiology by the American Association of Neuromuscular and Electrodiagnostic Medicine. They concluded that based on class I and II evidence, median nerve CSA at the wrist is accurate for the diagnosis of CTS. In addition, they found that based on class II evidence, neuromuscular US probably adds value to electrodiagnostic studies in assessing CTS, as it can detect structural anomalies. However, further evidence-based guidelines would be useful to establish reference values and parameters for US in diagnosing CTS.
Assessment of the vascularity of the median nerve using colour and power Doppler as an aid in the diagnosis of CTS is gaining popularity, but evidence of relevance and sensitivity is limited to date. In one controlled study looking at power Doppler in CTS, 48% of patients with CTS showed increased power Doppler signal, with no signal seen in the control group. One group has suggested from their research that assessment of vascularity with colour Doppler in addition to CSA leads to improved sensitivity and specificity, in fact equalling that of EP studies. There is also a positive correlation between increased power Doppler score and CSA of the median nerve. Vascularity has been shown to decrease following local steroid injection. Despite these encouraging findings, no validated scoring system has been created for assessing the vascularity of the median nerve, thus subjective measures are often used.
Ophir et al. were the first to describe elastography, a phenomenon based on the principal that compression of tissue produces displacement (strain) within the tissue and that this strain is lower in harder tissues. US elastography is the method of generating a colour map of tissue movement in response to external compression. Initial research using US elastography indicates that the median nerve is stiffer in patients with CTS than in controls.Table 3 summarizes the comparison between US and EP studies in CTS.
Diagnosis of CTS
CTS is normally diagnosed with a thorough clinical history and examination and the addition of electrophysiology (EP) studies if necessary. More recently, the use of US in the diagnosis has been reported in a number of studies showing that sensitivity and specificity are approaching that of EP studies. US is highly acceptable to patients, with ease of use in the consultation room, and provides a capability to assess anatomical aspects of the carpal tunnel and guide treatment.
Clinical History and Examination
A thorough clinical history and examination are the most crucial elements in the diagnosis of CTS. Studies have been criticized for using EP criteria alone for inclusion, as they cannot fully exclude the diagnosis of CTS. Consensus opinion by experts in the field is that clinical diagnosis should be made, independent of EP studies, in both the research and the clinical setting. The American Academy of Neurology (AAN) has produced practice parameters for CTS with diagnostic criteria for history and physical examination and these are listed in Table 1.
EP Tests and US
Practice parameters for the performance of electrodiagnostic tests for CTS have been set out by a joint report of the American Association of Electrodiagnostic Medicine, American Academy of Neurology and American Academy of Physical Medicine and Rehabilitation. They also performed an extensive literature review and reported a sensitivity and specificity of >85% and >95%, respectively, for median sensory and motor NC studies when compared with clinical diagnosis. However, the same review found that EP studies missed the diagnosis of CTS in 16–34% of patients with clinically defined disease. Other more recent reviews put the sensitivity at 85–90% and the specificity at 82–85% and recognize that EP studies alone should not be used as the standard for diagnosis.
US is now well established as a diagnostic tool in CTS. There are many advantages to US, including that it is readily available, non-invasive, has a shorter examination time and can be used to assess a number of parameters of the median nerve such as size, vascularity (using power Doppler) and mobility (using dynamic imaging). In addition, US provides information on anatomical variations of the median nerve and surrounding structures that may be a causative factor in CTS. On US imaging at the wrist, the median nerve is easily visualized in transverse view as hypoechoic nerve fibres with hyperechoic rims immediately superficial to the flexor tendons, with the hyperechoic FR overlying it within the carpal tunnel. Fig. 1 demonstrates the probe position for examination of the median nerve and US images of the median nerve in CTS.
(Enlarge Image)
Figure 1.
Images of probe position and US in carpal tunnel syndrome
(A) Demonstration of probe position on the antral wrist for examination of the median nerve. (B) View of carpal tunnel with surrounding structures. MN: median nerve; FCR: tendon of the flexor carpi radialis; Sc: scaphoid bone; UA: ulnar artery; Pi: pisiform bone. (C) Tracing method to measure cross-sectional area of enlarged median nerve measuring 18 mm in a male with CTS. (D) Longitudinal view of the median nerve showing enlargement as it enters the carpal tunnel in a female patient. CTS: carpal tunnel syndrome.
A number of studies have examined the parameters of the median nerve that are most useful in diagnosing CTS. It has been shown that the cross-sectional area (CSA) of the median nerve is significantly greater in those with CTS compared with healthy controls. The CSA has also shown good concordance with NC studies for CTS severity when US cut-off points for CSA are used to discriminate between grades. In those with normal NC studies but a clinical picture of CTS, the CSA on US has been shown to be significantly larger than normal controls. Hypervascularity and hypoechogenicity of the nerve are also present in those with larger CSA, and the probability of having CTS has been stated to be 90% in those with normal EP studies and all three of these features present. Although it is recognized that the median nerve often becomes hypoechoic as it enlarges, the diagnostic accuracy of this remains uncertain. The thickness of the FR and the thenar muscles has also been shown to correlate with CTS diagnosed by EP studies, and these parameters can be used in combination to increase sensitivity and to assess patients with normal EP studies.
Two other aspects of US assessment have attracted much debate: (i) which measure of the median nerve is best to establish the diagnosis and (ii) the best cut-off size for the CSA of the median nerve to diagnose CTS. There is a considerable body of evidence to indicate that the CSA at the level of the pisiform bone or the tunnel inlet is the most sensitive and specific US finding in patients with CTS. As can be seen in Table 2, the majority of studies use the tunnel inlet/pisiform bone as the site for measurement of the CSA. There is much less agreement on the best cut-off size for the CSA, with recommended cut-offs varying from 6.5 to 15 mm. Table 2 lists the sensitivity and specificity from 22 studies and includes the CSA cut-offs used and the location where the CSA was measured. These findings demonstrate the wide variation in sensitivity (62–97.9%) and specificity (57.1–100%) seen in studies comparing US with clinical assessment or EP tests. Most of the studies using clinical assessment as the reference standard do not document the sensitivity and specificity of EP studies. In the three studies that did, sensitivity was reported as 78%, 80% and 82% and specificity was reported as 83%, 84% and 97%. The reference standards used for comparing US and EP studies are included in Table 2.
Limited research has been carried out on inter- and intrareader reliability of US. The intrareader reliability for CSA of the median nerve is high in the studies identified that looked at this. There are two methods for measuring the CSA of the median nerve. The tracing method involves tracing a continuous line around the inner hyperechoic rim of the median nerve, then machine software is used to calculate the CSA. The second method involves measuring the anteroposterior and transverse distances of the median nerve, which are then input into the ellipse formula to calculate the CSA. Fig. 1 demonstrates the tracing method. There is strong correlation between these two methods. When measurements of the median nerve of an amputated limb obtained by US using both methods were compared with direct measurements made later on frozen section, correlation was found to be 0.992 for the anteroposterior and transverse distances and 0.982 for the direct measurement. Interreader reliability for measurement of the CSA at the tunnel inlet using both tracing and ellipse formula methods shows good reliability, with correlation coefficients of 0.81 and 0.97, respectively. However, the interreader reliability was poor at the tunnel outlet, which probably relates to the orientation of the median nerve at the tunnel outlet where it moves more dorsally, making good visualization and measurement difficult.
A few studies have reported data on the percentage of patients with a clinical diagnosis of CTS with normal CSA on US but positive NC studies. These studies used variable criteria for US diagnosis and are not directly comparable. Depending on the US criteria applied, NC studies were positive in 8.5%, 21% and 28.2% of patients with a clinical diagnosis but normal US in three studies identified. One of the main issues with studies looking at the diagnosis of CTS with US is the difference in parameters used, making comparison difficult. Recent evidence-based guidelines have been published by a panel of experts specializing in neurology, physical medicine and rehabilitation and radiology by the American Association of Neuromuscular and Electrodiagnostic Medicine. They concluded that based on class I and II evidence, median nerve CSA at the wrist is accurate for the diagnosis of CTS. In addition, they found that based on class II evidence, neuromuscular US probably adds value to electrodiagnostic studies in assessing CTS, as it can detect structural anomalies. However, further evidence-based guidelines would be useful to establish reference values and parameters for US in diagnosing CTS.
Assessment of the vascularity of the median nerve using colour and power Doppler as an aid in the diagnosis of CTS is gaining popularity, but evidence of relevance and sensitivity is limited to date. In one controlled study looking at power Doppler in CTS, 48% of patients with CTS showed increased power Doppler signal, with no signal seen in the control group. One group has suggested from their research that assessment of vascularity with colour Doppler in addition to CSA leads to improved sensitivity and specificity, in fact equalling that of EP studies. There is also a positive correlation between increased power Doppler score and CSA of the median nerve. Vascularity has been shown to decrease following local steroid injection. Despite these encouraging findings, no validated scoring system has been created for assessing the vascularity of the median nerve, thus subjective measures are often used.
Ophir et al. were the first to describe elastography, a phenomenon based on the principal that compression of tissue produces displacement (strain) within the tissue and that this strain is lower in harder tissues. US elastography is the method of generating a colour map of tissue movement in response to external compression. Initial research using US elastography indicates that the median nerve is stiffer in patients with CTS than in controls.Table 3 summarizes the comparison between US and EP studies in CTS.
SHARE
