Two processes account for most instances of ulnar neuropathy at the elbow: compression in the retroepicondylar groove, and compression by the humeroulnar aponeurotic arcade joining the two heads of the flexor carpi ulnaris. While conventional electrodiagnostic criteria may localize an ulnar neuropathy to the elbow, separating retroepicondylar compression from humeroulnar arcade compression is more difficult. In 130 cadaver elbows, we examined the relationships between the medial epicondyle, flexor carpi ulnaris, and ulnar nerve. The humeroulnar arcade lay from 3 to 20 mm distal to the medial epicondyle, the intramuscular course of the nerve through the flexor carpi ulnaris ranged from 18 to 70 mm, and the nerve exited the flexor carpi ulnaris 28 to 69 mm distal to the medial epicondyle. In 6 specimens, dense fibrous bands bridged directly between the medial epicondyle and the olecranon proximal to the cubital tunnel proper; accessory epitrochleoanconeus muscles were present in 14 specimens: both may cause ulnar neuropathy at the elbow. Anatomical variations may contribute to the difficulty in separating causes of ulnar neuropathy at the elbow.

At the elbow the ulnar nerve may be compressed either in the retrocondylar groove or at the cubital tunnel. Optimal surgical therapy should be directed at the specific site of involvement. Intraoperative electroneurography performed in conjunction with 19 ulnar nerve explorations helped localize the precise site of compression. Of the primary procedures, abnormality was at the retrocondylar groove in 9, cubital tunnel in 4, both locations in 3, and at an unusual distal point in 1; 12 anterior subcutaneous transpositions, 4 cubital tunnel releases, and 1 distal decompression resulted. Intraoperative studies helped identify residual compression in two patients undergoing reexploration. Although routine electrodiagnosis may localize an ulnar neuropathy to the elbow, reliably separating retrocondylar from cubital tunnel compression is more difficult. Preoperatively, percutaneous serial short increment studies were more accurate than simple "inching" in predicting the site of compression.

The ulnar nerve normally enters the flexor carpi ulnaris (FCU) proximally and anteriorly between the humeral and ulnar heads of the muscle. After an intramuscular course of several centimeters, the nerve exits the FCU distally to lie in a tissue plane between the FCU and the flexor digitorum profundus (FDP). A patient with ulnar neuropathy studied by intraoperative electroneurography demonstrated major focal conduction block at the point where the nerve exited the FCU. A fivefold increase in amplitude and reversal of a dispersed, irregular compound muscle action potential to a more normal configuration occurred with stimulation just distal to the point of exit. There was no evidence by inspection, probing, or electroneurography of compression in the retrocondylar groove or at the cubital tunnel. In a series of 100 cadaver dissections, the intermuscular septum between FCU and FDP was thick and tough in several specimens. This septum may represent a site of ulnar nerve entrapment.

A common misconception attributes sparing of the flexor carpi ulnaris (FCU) in ulnar neuropathy at the elbow (UNE) to its innervating branch arising "at or above the elbow." We examined the relationship of FCU branches to the medial epicondyle (ME) and humeroulnar aponeurotic arcade (HUA) in 30 cadaver elbows. In only three did the first FCU branch arise at or proximal to the ME. In 36 UNE cases with fibrillations in the first dorsal interosseous, the FCU was normal in 10, mildly abnormal in 11, and severely abnormal in 15. FCU involvement correlated with the severity of the neuropathy and with whether compression was retroepicondylar or at the HUA. We conclude that sparing of the FCU in UNE is unrelated to the level of origin of its innervating branch, but rather is related to the internal neural topography and to the severity and level of compression.