New techniques for the diagnosis of carpal tunnel syndrome.

An overview is presented of the assessment of carpal tunnel syndrome (CTS) in manual workers, with emphasis on the distinction from other compression syndromes and from diffuse polyneuropathy. Clinical examinations, electrodiagnostic studies, the determination of vibration threshold, thermography, and, recently, computerized tomography have been used. This last-mentioned method has also been employed in a study indicating that a subgroup of the general population may possess a risk factor, small carpal canal, associated with the development of CTS.

A simple definition of carpal tunnel syndro me (CTS) is a disorder resulting from compression or irritation of the median nerve as it pas ses into the hand betw een the carpal bones and the transverse carpal ligament with subsequent discomfort and impaired use of th e hand . The carpal canal is formed by the concave arch of the carpal bon es and is roofed by the transverse carpal ligament. These structures fo rm a rigid compartment through which nine tendons and the median nerve must pa ss.

Clinical presentation
Unless associated with dire ct trauma , the on set of symptoms of CTS is usually no cturnal and insidious. The pain in th e hand may initially be described as episodic tingling and numbness with gradual progression to more severe symptoms referred to as burning , aching, pri cking, or as a pa inful numbness in the fin ger s and deep in the palm. With the pain and tingling there is a subjective feeling of uselessness in the fingers, which are sometimes describ ed as feeling swollen , even though , on inspect ion, little swelling is apparent. Many patients will have accompanying pain in the forearm , sometimes reaching the shoulder and describ ed as a dull aching pain felt deepl y in the limb .
Color changes in the fingers hav e been described , particularly with exposure to cold, but the y are not related to the att ack s of pain. In addition excessive sweating and mild degrees o f edema are related to the vasomotor imb alan ce known to occur in CTS (I As CTS progresses , the nocturnal pain and tingling may begin to wake the patient after a few hours of sleep. Relief ma y be obtained by hanging the arm out of bed or sha king or rubbing th e hand , but as symptoms increase, patients often get out of bed and walk about until the symptoms have ea sed. At thi s stage episod ic tinglin g may develop during the da y, but the associated pain in the arm occurs less often during the day than at night. In addition to sensory symptoms, there may be clumsiness and difficulty in performing cert ain ta sks, such as un screwing bottle tops, turning a key, or crocheting. Objective changes in sensation and strength may app ear in the hand, but some pati ents ma y suffer severe attacks of pain for man y yea rs without developing abnormal neurological signs. Sensory signs are best sought in the fingertips , where impairment is, as a rule, more pronounced. Occasionally, instead of decreased sensa tion, there is an overreaction to cutaneous stimuli in the median innervated digits. Isolated thenar wasting or sensory impairment in the distribution of one of the lateral three digital ner ves ma y be the pre senting feature of median nerve lesions at the wrist. Mild weakness of the abductor poll icis brevi s or of the opponens pollicis muscle is frequently present with no visually apparent atrophy. Manual pres su re over the flexor aspect of the wrist or prolonged hyper extension or hyperflexion of this joint ma y produce sen sor y symptom s (Phalen 's sign). Tinel 's sign, consistin g o f shocklike pa in and tinglin g elicited by per cussion of the median nerve at the wrist , is a less common finding th an often thought (10) .

Electrodiagnostic studies
Although CTS abnormalities are more likel y to be found in nerve cond uction studies in the presence of a defect found in a clinical examination, a significant number of patients with typical CTS symptoms have no abnormalities other than those detected in electrodiagnostic testing .
Compression of the median nerve in the carpal tunnel produces slow ner ve conduction , a block of nerve conduction, or axonal destruction. For man y years, tests for this syndrome have included measuring the time for a nerve impulse to travel through the median nerve from the wrist to the abductor poIIicis brevis muscle (distal motor latency) and from the index finger to th e wrist with orthodromic stimulation or from the wrist to the index finger with antidromic stimulation (distal sensory latency). The sensitivity of these measurements is limited because of the inclusion of the relatively normal segment of the nerve distal to the carpal ligament.
Several newer techniques are especially helpful in the early diagnosis of CTS . Recentl y the conduction time of sensory fibers has been measured in the segment across the carpal tunnel. With palmar stimulation the median nerve is stimulated distal to the carpal ligament in the midpalm , and the sensor y ner ve action potential is recorded 8 em proximally, just above th e wrist. The median palmar latency is compared with the ulna r palmar latency mea sured over the same distance . Prolonged median palmar latencies or latencies relatively greater than the ulnar value suggest CTS. Palmar stimulation identifies approximately 95 070 of patients with CTS (5). Prolongation of both the ulnar and med ian distal laten cies suggests low hand temperature or peripheral neuropathy.
Occasionally it is helpful to determine the site of maximal cond uction delay in the distal med ian nerve by stimulation at multiple sites along th e nerve . Latencies are measured from the inde x finger to the palm , from the palm to the wrist, and from the wrist to the forearm. These measurements allow one to distinguish between CTS and peripheral neuropathy affecting the median nerve in which the maximal slowing will be found in the segment from the palm to the finge r rather than across the carpal tunnel.
Stimulation of the median nerve from the palm to the wrist in l-cm increments in patients with CTS ha s shown that, in 53 % of affected nerves, there is a sharply locali zed latency increase acro ss a I-em segment , most commo nly 2-4 em distal to th e ori gin of the tranverse carpal ligament. This slowing appears to correspond to the region of the distal edge of a carpal ligament. In the remaining 47 % of nerves, the slowing is distributed more diffusely across the carpal tunnel (12). Median nerves are tested bilaterally in CTS patients because of the high incidence of conduction abnormalities in the asymptomatic extremities of CTS patients. Other nerves are tested because of the occa sional patient who presents a diffuse neuropathy with CTS.
Section of the transverse carpal ligament in CTS usually results in improvement in both clinical symptoms and the long median nerve distal latency. However, the latency abnormality often does not return to 386 normal (14). Needle-electrode examination of the thenar eminence is uncomfortable and less sensitive in CTS than are routine ner ve cond uctio n studies (16).
Calculation of residual latency, onl y recently reported (13), is important for eliminating the effect of variability in nerve conduction velocity upon distal latency measurements. Residual latencies hold constant into the ninth decade in contrast to conduction velocities which continue to slow with incre asing age. With a recent onset of CTS, prolonged residual latencies may be the only abnormality.
It has been recently demonstrated that median innervated lumbrical muscles are relatively spared in advanced CTS. Therefore, increased diagnostic sensitivity for CTS is obtained when comparison is made to the involved abductor pollicus brevis. In CTS , the distal latency to the abductor pollicis brevis is greater than the distal latency to the second lumbrical after stimulation of the median nerve at the wrist (15).

Vibration threshold
Th e large myelinated fiber s are decreased in number in the median palmar digital nerve s in CTS (17). Abnormalities in this population ofaxons should be reflected in vibration sensation, which is carried by this fiber population. Abnormal vibratory perception in the median inervated index finger in comparison to the perception in the ulnar innervated ipsilateral little finger has been reported as the earliest non provocative sign of CTS. Dellon (7) found an abnormal perception of vibratory stimuli in 72 % of pat ient s with a history compatible with CTS.
A practical tool for determining vibration perception threshold is a tactile stimulator. When the maximum sensitivity occurs at 200-300 Hz, the pacinian corpuscle, a rapidly adapting mechanoreceptor, is stimulated . Studies show that the difference in the vibratory thre shold between the invol ved digit II and the uninvolved digit V is signi ficantly elevated in CTS (3). Threshold tests of sensibility (pre ssure monofilament s and vibration) correlate accu rat ely with symptom s of nerve compression (9).

Thermography
Liquid crystal thermography has onl y recentl y been applied in the study of CTS. It appears to be sensitive in detecting asymmetries in heat distribution (8). Liquid crystal thermographs probably reflect vasomotor instability resulting from compromise of the sympathetic nerves , which are carried in the median nerve. As a noninvasive method for diagnosing and following CTS , this method ma y have potential, but further studies are still needed to determine whether thi s technique is identi fying onl y a subpopula tion of individuals with CTS or if these temperature changes are a universal finding in thi s disorder.

Computerized tomography of the carpal canal
Characterization of overall hand size was tabulated by Arm strong & Chaffin (2) in a gro up of workers with CTS and a group of referents . Anthropometric data collected from surface measurement of the hand and wrist, including external measurement s of hand length , palm width, wrist width, and wrist thickness and the intern al dimension of the third metacarpal length, were selected for the characterization of overall hand size. No significant differences between the two groups were found. Even when several internal measurement s were mad e from radiographs to estimate the ca rpal angle, which is the cur vature of the carpal bone s, no association was found between hand shape and CTS .
Measurements of the cross-sectional area of the carpal tunnel reveal a narrow carpal tunnel in females with CTS (11). Rather than using verti cal roentgenograms of the bony boundaries of the carpal groove, Dekel et al (6) demonstrated carpal canal stenosis in idiop athic CTS with manual digiti zation of computerized tomograms through the carpal canal. Both of these studies examined only females with CTS since idiopathic CTS is generally associated with females between 40 and 60 yea rs of age.
Carpal canal stenosis may be congenital or acquired by trauma or arthritis, but , as an etiologic facto r in occupationally related CTS, it ha s been given little a ttent ion . Ca rpal canal size as a pred ictor for the development of this entrapment syndrome was studied in a group workers (4). Seven of 14 electricians had symptomatic carpal tunnel syndro me . Workers classified as ha ving clinicall y apparent CTS had appropriate responses on the symptom questionn aire and electrophysiological finding s of CTS. Subclinical cases included asymptomatic workers with physiological evidence of the synd rome. "Normal" subjects had no clinical or physiological evidence of nerve entrapment.
Computerized tomography was performed on a GE 8800 CT I T scanner. A scout film was obtained to insure proper alignment. The first tomographic slice was taken at the palmar crease, followed by 5-mm contiguous slices through the carpal canal to the metacarpal level. At each level, a cursor in the scanner outlined the circumference of the carpal canal. The computer calculated canal area for each level. The depth at the median nerve and width of the canal were measured, as well as the th icknes s of the transverse carpal ligament. Affected wor kers had a mean crosssectio nal area of 1.75 + 0.21 cm-; the reference values were 2.23 + 0.15 ern? (P < 0.05). Indi viduals with a subclinical for m of CTS had an area of 1.93 + 0.22 crn-, similar to that of the symptomatic group . Wrist circumference was not a reliable predictor of the small est carpal canal area. Unusual bon y and soft tissue structures within the carpal canal were easily ident ified with the computerized tomography.
In a seco nd study, carpal canal size was examined in eight men with an age distribution (20-60 yea rs) comparable to that of the aforementioned study, a normal physical examination of the upper extremities, normal nerve conduction studies, and no exposure to occupations or hobbies which involved repeated and forceful movement of flexor tendons through the carpal tunnel. The range of the smallest carpal canal size per wrist was identical to that found among the electricians with a median of 1.75 (range 1.34-2.01) crn-. These results suggest that a subgro up of the general population may possess a risk factor , small carpal canal, associated with the development of Cl'S, Howe ver, with no exposure to known ergonomic stresse s or other risk factors associated with the development of crs, these individuals may remain free of symptoms secondary to median nerve compression in the carpal canal.
Computerized tomography may reveal spaceoccupying structures within the canal that are not apparent by external examination. Accumulations of fat, muscle insertions within the canal, bony abnormalities (such as a protruding head of the capitate), fractures, tenosynovitis (causing adhesions between the long flexor tendons), and a thickened transverse carpal ligament are all features that compromise the area in the canal needed for the long flexor tendons and the median nerve. Wrist circumference has been a weak predictor for CTS because it does not reflect these canal anomalies. Body weight affects wrist circumference , but does not alter the size of the canal.
An association between bod y weight and the quantit y of fat in the canal was not found . Further evaluation is needed to establish the significance of the relationship between long bones , as mea sured by height, and the small carpal bones composing the canal.
Computerized tomography has been used on our clinic pat ients with idiopathic crs in the nondominant hand. It revealed carpal canal stenosis. This etiology should be considered when symptoms are relieved by splinting and recur whenever the splint is removed. Establishing carpal canal stenosis as the etiology may eliminate months of fruitless conservative therapy before release surgery is performed.