| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Extremities |
1 Department of Radiology, North Shore University Hospital, Great Neck, New York
 |
Commentary |
|---|
 Top Commentary ReferencesResponse |
|---|
The original US descriptions by Buchberger et al (2,3) describe the triad of swelling of the median nerve in the proximal carpal tunnel, flattening of the nerve in the distal tunnel, and increased palmar bowing of the flexor retinaculum. However, this triad was present in only 7% of 68 symptomatic wrists in a recent study by Duncan et al (4). Findings seen in carpal tunnel syndrome with MR imaging, such as swelling of the median nerve proximally, flattening of the nerve distally, increased signal intensity in the nerve on T2-weighted images, retinacular bowing, flexor tenosynovitis, and deep palmar bursitis, have been called into question (5). Radack et al (5) found that these parameters were not sensitive for detection of carpal tunnel syndrome and may have been the result of selection bias in previously published studies. One could make a similar argument with regard to US. Moreover, even if we accept that swelling of the median nerve at the level of the pisiform is an accurate criterion, there is no consensus in the literature as to what size is abnormal. Duncan et al (4) used an elliptical cross-sectional area of 9 mm2 or larger as indicative of carpal tunnel syndrome at US, whereas Lee et al (6) used an area of 15 mm2. Similarly, Middleton et al (7) used an elliptical area of 9.8 mm2 at MR imaging as their criterion for pathologic swelling, whereas Horch et al (8) reported an average elliptical area of 15.4 mm2 in their symptomatic population. Chronicity of the disease and severity of symptoms may have an influence on such criteria and need to be taken into account in future studies.
US and MR imaging of carpal tunnel syndrome have followed parallel paths with little comparison of the two modalities. In 1992, Buchberger et al (3) did compare the findings of carpal tunnel syndrome in 20 symptomatic wrists with both US and MR imaging and found that the measurements of cross-sectional area of the median nerve and flattening of the median nerve did not significantly differ between the two modalities; however, bowing of the flexor retinaculum, mild degrees of compression of the median nerve, and tendon sheath thickening and ganglion cysts were better demonstrated with MR imaging. As demonstrated by the high quality of the illustrations in the article by Martinoli et al (1), these differences may no longer be true.
Few comparisons of the two modalities with electromyography, the clinical standard, have been performed. Britz et al (9) found abnormalities of the median nerve with MR imaging in all 43 of the symptomatic wrists in their study and in none of the five control subjects. The criteria that they found most useful were abnormal signal intensity in the median nerve, abnormal signal intensity in the flexor tendon sheath, and bowing of the flexor retinaculum. Using US, Lee et al (6) found that median nerve swelling at the level of the pisiform correlated best with electromyographic results and that such swelling, with 15 mm2 as the criterion, had a sensitivity of 88%, specificity of 96%, positive predictive value of 97%, and negative predictive value of 86%.
On the basis of their findings, Lee et al (6) recommended a revised diagnostic algorithm in which US follows the physical examination: (a) The patient with clinically suspected carpal tunnel syndrome and a swollen median nerve at US is treated conservatively if the elliptical cross-sectional area of the nerve is less than 15 mm2 and is treated surgically if this area is greater than 15 mm2. (b) The patient with a normal median nerve at US but disease in the carpal tunnel itself, such as synovitis or a ganglion, is treated for that disease. (c) The patient with clinically suspected carpal tunnel syndrome but with both a normal median nerve at US and a normal carpal tunnel undergoes electromyography to check for intrinsic disease of the nerve or focal disease elsewhere along its course. This suggested algorithm is reasonable but has not been tested.
In terms of the six-tiered hierarchical model of efficacy of Thornbury (10), US and MR imaging of carpal tunnel syndrome have passed level I (technical efficacy) and are at level II (diagnostic-accuracy efficacy). Level III is the diagnostic-thinking efficacy tier, in which one determines if the imaging modality helps to strengthen or change the clinician's differential diagnosis, and level IV is the therapeutic efficacy tier, in which the actual effect of the modality on patient treatment is evaluated.
To move MR imaging and US of carpal tunnel syndrome from the technical efficacy and diagnostic-accuracy efficacy levels into the higher levels of diagnostic-thinking efficacy and therapeutic efficacy, the imaging criteria for these modalities may need to be standardized. In 1993, the Quality Assurance Committee of the American Association of Electrodiagnostic Medicine (11) surveyed the English-language literature regarding nerve conduction studies and electromyography to develop practice guidelines for the evaluation of carpal tunnel syndrome. Given the proliferation of published articles and presented abstracts regarding carpal tunnel imaging, we may be approaching the time when one of our professional societies should charge a subcommittee with the task of forming standardized imaging criteria for both MR imaging and US. Once these imaging modalities have reached the therapeutic efficacy level, the issue of which modality is preferred may need to be decided by another professional committee, such as exists for the Appropriateness Criteria Project of the American College of Radiology.
MR imaging of the ankle and foot has already entered the therapeutic efficacy stage, with Anzilotti et al (12) showing a change in treatment in 28 of 81 patients (35%) with various abnormalities of the foot and ankle diagnosed with MR imaging. For Morton neuroma specifically, Zanetti et al (13) showed that MR imaging resulted in a change in the treatment plan in 31 of 54 feet (57%). To my knowledge, no such work has yet been published for US of Morton neuroma. MR imaging has been shown to have a sensitivity of 87% and specificity of 100% (14), whereas US has a sensitivity of 85%–100% and specificity of 83%–100% (15–18). The general consensus in the US literature is that these neuromas are symptomatic when greater than 5 mm in diameter and may have varying echogenicity (15,17–20). Given that neuromas may be multiple in any particular foot, an advantage of US over MR imaging—in addition to the speed of the examination—may be the possibility of determining which neuromas are symptomatic by direct compression of the neuroma between the US probe and palpating finger.
Level V of the Thornbury classification is patient-outcome efficacy, in which the cost-benefit ratio of a test is analyzed. Although imaging of both Morton neuroma and carpal tunnel syndrome has not yet reached this higher level of assessment, Medicare reimbursement in 2000 for wrist MR imaging (Current Procedural Terminology code 73221) is $633.30, whereas for wrist US (Current Procedural Terminology code 76880) it is $105.26. The reimbursement for a single extremity electromyogram (Current Procedural Terminology code 95860) is $86.44. It remains to be seen how market forces will move the medical community. Moreover, the issue of cost-effectiveness of these modalities regarding the radiologist's time and effort also needs to be addressed. US is operator dependent and has a high learning curve but is a more rapid examination than MR imaging; in the study of Duncan et al (4), each wrist was evaluated in an average of 5 minutes. Given the potentially rapid patient throughput of US, I believe that it is currently the more efficient imaging examination for the focused evaluation of carpal tunnel syndrome.
An issue that should be a concern to all radiologists is control over US if this modality becomes widely accepted into the clinical diagnostic algorithm, whether it be for carpal tunnel syndrome or Morton neuroma. This issue is especially important in light of the new technology of handheld portable US units. The cost of these units is a fraction of that of a standard US machine, thus making them affordable for orthopedic surgeons, physiatrists, neurologists, podiatrists, and any other clinical specialists who deal with these syndromes.
Thus, although the article by Martinoli and colleagues is a beautiful review of the anatomy and US findings in nerve compression syndromes, it is merely the tip of the iceberg in regard to this topic.
|
References |
|---|
|
TopCommentaryReferencesResponse |
|---|
Department of Radiology R, University of Genoa, Italy; Division of Radiodiagnostics, Hôpital Cantonal Universitaire, Geneva, Switzerland
|
Response |
|---|
|
TopCommentaryReferencesResponse |
|---|
We fully agree with Dr Miller that some variability in diagnostic signs and cutoff values exists in US examination of carpal tunnel syndrome, although excellent accuracy has been demonstrated with this technique in different series. This variability can be explained, at least in part, by the fact that many studies in this field have been limited to small groups of subjects and performed without comparison with a standard of reference. However, it is interesting to reflect on certain aspects of US of peripheral nerves that still need to be discussed.
First, US examination of nerves is strongly technology dependent, and a substantial variation in the performances of US units is obvious when the results from the literature are compared. In the past 5 years, the interest in US examination of nerves has grown worldwide, especially due to the advent of broadband linear-array transducers with frequencies above 10 MHz and sophisticated focusing in the near field. The improved performance of these transducers has enabled US to demonstrate subtle anatomic details at least equal to or even smaller than those visible at surface-coil MR imaging and has made many of the results of early studies no longer applicable. As shown in our article, fine abnormalities of nerve fascicular structure and intraneural microvasculature can now be appreciated in nerve entrapment syndromes. When US research fully assesses the clinical significance of these new signs, it is conceivable that the value of quantitative studies based on calculation of nerve sizes will be reviewed and the criteria for diagnosing carpal tunnel syndrome will be further refined.
Second, as in other musculoskeletal applications, the use of US for imaging nerves is inherently dependent on the experience and interpretation of the operator. As a consequence, there is lack of standardization in this field, with special reference to the scanning technique and selection of the appropriate diagnostic criteria. For example, it is still unclear from the literature where measurement of the cross-sectional area of the median nerve should be performed, whether at the proximal tunnel (scaphoid-pisiform level), at the distal radius, before the entrance of the median nerve into the carpal tunnel, or at the point of maximum nerve swelling. Further experience based on controlled studies in large series of patients and use of high-end equipment will probably reduce the subjective nature of US studies with time. However, at present, we agree with Dr Miller that it is still premature to draw conclusions about level II (diagnostic-accuracy efficacy) of the Thornbury scale, whereas levels III–V could be correctly approached only in the future.
Third, at many institutions MR imaging is considered the most advanced imaging examination of the musculoskeletal system, and there are too few blinded research studies that assess the true effectiveness of US compared with this technique. Often, US is promoted as an alternative to MR imaging on the basis of its relatively low cost, portability, and, in most cases, improved accessibility. However, we should balance the indications for US and MR imaging on a scientific basis rather than purely according to economics. In nerve imaging, we believe that high-resolution US has specific advantages over MR imaging, including the ability to examine tissues in both the static and dynamic state with real-time scanning and the ability to explore long segments of nerve trunks in a single study to rule out other more proximal levels of compression. This latter capability is intrinsic to US and can facilitate diagnosis when the patient's symptoms and the electromyography report are not typical. To validate our clinical impression into scientific proof will be our goal in the near future.
Last but not least, performance by itself does not lead to knowledge. If continuing studies refine the diagnostic criteria for this imaging method, it is also important that radiologists not trained in peripheral nerve imaging be convinced to acquire this new skill. Not only the improved resolution of high-end technology but also a detailed knowledge of anatomy, time, energy, and resources are needed to create expertise. US imaging of peripheral nerves is rapidly evolving; thus, it is imperative for the radiologist involved in musculoskeletal US to gain the ability to combine clinical judgment and the necessary competence and performance capability to provide good patient care. Otherwise, US of peripheral nerves will become another turf battle between radiologists and other clinical colleagues.
In summary, US has great diagnostic potential in the assessment of peripheral nerve entrapments. We agree with Dr Miller that a more detailed analysis of the clinical efficacy of this technique will be needed. However, at present, the clinical significance of a US examination, which is now widely available with the appropriate high-resolution technology, still depends on the operator's experience and skill and the intrinsic performance of the apparatus being used. Nevertheless, we hope that our pictorial evaluation of nerve entrapment syndromes will encourage radiologists to enter this field and to use US for evaluating peripheral nerve disorders.
Related Article
RadioGraphics 2000 20: S199-S217.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOGRAPHICS | RADIOLOGY | RSNA JOURNALS ONLINE |
