Martin–gruber Anastomosis

The most commonly encountered anomaly in the upper extremity is a cross-over of median-to-ulnar fibers, the Martin–Gruber anastomosis (MGA). The anastomosis involves only motor fibers; sensory fibers are spared. The cross-over usually occurs in the mid-forearm originating from the branches of the median nerve supplying the superficial forearm flexor muscles, the anterior interosseous nerve, or directly from the main median nerve. The median fibers that have crossed over then run with the distal ulnar nerve to innervate any of the following ulnar muscles: (1) the hypothenar muscles (abductor digiti minimi), (2) the first dorsal interosseous muscle (FDI), (3) the thenar muscles (adductor pollicis, deep head of flexor pollicis brevis), or (4) a combination of these. The most common by far is for the anastomosis to innervate the FDI.

This particular anomaly is quite common and has been reported to occur in 15 to 30% of patients. When present, it may be unilateral or bilateral. During routine nerve conduction studies, the MGA may be recognized under the following circumstances.

Routine Ulnar Conduction Study: Pseudo-Conduction Block between the Wrist and Below-Elbow Sites

The MGA may be recognized during routine ulnar motor studies, recording the abductor digiti minimi, stimulating at the wrist and below-elbow sites (Figure 7–1). If the anastomotic fibers innervate the abductor digiti minimi, a characteristic pattern results: a drop in the ulnar compound muscle action potential (CMAP) amplitude is seen between the wrist and the below-elbow stimulation sites (Figure 7–2). With stimulation at the wrist, the CMAP reflects all motor fibers innervating the hypothenar muscles, including those that have crossed over more proximally from the median nerve. Stimulation at the below-elbow site activates fewer fibers, however, as this stimulation site is above the cross-over. Thus, the portion of fibers innervating the abductor digiti minimi that originate from the median nerve have already crossed over in the forearm and, therefore, do not contribute to the CMAP. The differential diagnosis of this pattern (i.e., higher amplitude distally than proximally) includes the following:

• Excessive stimulation of the ulnar nerve at the wrist resulting in co-stimulation of the median nerve

• Submaximal stimulation of the ulnar nerve at the below-elbow site

• Conduction block of the ulnar nerve between the wrist and below-elbow sites

• An MGA with crossing fibers innervating the hypothenar muscles

FIGURE 7–1 Martin–Gruber anastomosis (MGA).

Cross-over of median-to-ulnar fibers supplying the hypothenar muscles may occur in MGA. During routine ulnar motor studies, recording the abductor digiti minimi and stimulating the ulnar nerve (S_U) at the wrist (WR) and below-elbow (BE) sites, the ulnar compound muscle action potential (CMAP) amplitude with BE stimulation is lower than with WR stimulation. If an MGA is not recognized, a mistaken impression of a conduction block may occur. To demonstrate an MGA in this situation, the median nerve is stimulated (S_M) at the WR and antecubital fossa (AF) while recording the hypothenar muscles, looking for a CMAP stimulating at the AF that is not present stimulating at the WR.

FIGURE 7–2 Martin–Gruber anastomosis and pseudo-conduction block of the ulnar nerve in the forearm.

Recording hypothenar muscles (abductor digiti minimi), stimulating ulnar nerve at the wrist (WR) and below-elbow (BE) sites results in a drop in amplitude at the BE site. The anastomosis is demonstrated by stimulating the median nerve at the wrist and antecubital fossa (AF), recording hypothenar muscles. There is no potential with WR stimulation, whereas one is present with AF stimulation. The compound muscle action potential amplitude evoked with median nerve stimulation at the AF is approximately equal to the drop in amplitude on the ulnar studies.

If a reduced ulnar CMAP is found at the below-elbow stimulation site compared with the wrist, it is essential first to check that co-stimulation has not occurred at the wrist and that submaximal stimulation has not occurred at the below-elbow site. Note that up to a 10% drop in the ulnar CMAP amplitude at the below-elbow (compared with wrist) site is considered normal secondary to normal temporal dispersion. The major danger in not recognizing an MGA in this situation is that of mistakenly interpreting the findings as a conduction block in the forearm, an unequivocal sign of demyelination. This error is especially serious in that the presence of a conduction block at a non-entrapment site usually signifies an acquired demyelinating peripheral neuropathy, which often is treated with immunosuppressive or immunomodulating therapy.

Whenever there is a >10% drop in amplitude between the wrist and below-elbow sites on routine ulnar motor studies, median nerve stimulation should always be performed at the wrist and at the antecubital fossa while recording the hypothenar muscles to check for an MGA. If no MGA is present, a small positive deflection usually is recorded with both the wrist and antecubital fossa stimulation sites, reflecting a volume-conducted potential from median muscles (see Chapter 2). If an MGA is present, a small positive volume-conducted potential will be present with median nerve stimulation at the wrist; however, median stimulation at the antecubital fossa will evoke a small CMAP over the abductor digiti minimi. The amplitude of the CMAP evoked by stimulating the median nerve at the antecubital fossa (recording the hypothenar muscles) will approximately equal the difference between the CMAP amplitudes evoked with ulnar nerve stimulation at the wrist and below-elbow sites (recording the hypothenar muscles). However, it is important not to overstimulate the median nerve at the antecubital fossa, resulting in co-stimulation of the ulnar nerve at the elbow, and thereby giving the appearance of an MGA when none truly exists. This can be avoided by slowly moving the stimulator from the median nerve at the antecubital fossa to the ulnar nerve at the elbow, stimulating at several points. In a true MGA, the CMAP that is evoked with median nerve stimulation at the antecubital fossa will briefly disappear as the stimulator is moved toward, but not over, the ulnar nerve at the elbow. It will then reappear with ulnar nerve stimulation at the elbow.

Ulnar Conduction Study: Proximal Martin–Gruber Anastomosis and Pseudo-Conduction Block between the Below-Elbow and Above-Elbow Sites

In patients with ulnar neuropathy at the elbow, one of the classic electrophysiologic findings is conduction block across the elbow, whereby a drop in the CMAP amplitude is seen between the below-elbow and above-elbow sites during routine ulnar motor studies (see Chapter 19). This pattern is not typically confused with an MGA, because the drop in CMAP amplitude in a typical MGA occurs between the wrist and below-elbow sites, mimicking a conduction block in the forearm, not across the elbow. However, very rarely, the cross-over fibers of the MGA are very proximal, and if stimulated, will contribute to the CMAP amplitude at the below-elbow site. In contrast stimulation above the elbow will not excite these cross-over fibers, thereby giving the impression of a conduction block across the elbow. It is in these cases, where the below-elbow stimulation might occur below the MGA, that an MGA may result in the mistaken diagnosis of ulnar neuropathy with conduction block at the elbow (Figure 7–3). This is especially apt to occur when the below-elbow stimulation site is too distal, making it more likely that the below-elbow stimulation occurs below the MGA, thereby exciting the cross-over fibers.

FIGURE 7–3 Proximal Martin–Gruber anastomosis (MGA) mimicking ulnar neuropathy at the elbow.

An MGA may rarely result in the mistaken diagnosis of ulnar neuropathy at the elbow if the anastomosis is very proximal, the below-elbow (BE) stimulation site is too distal, or a combination of both. In this example, routine ulnar motor studies are performed, recording the abductor digiti minimi and stimulating the ulnar nerve (S_U) at the wrist (WR), below-elbow (BE), and above-elbow (AE) sites. The ulnar amplitude at the AE stimulation site is lower than at the WR and BE stimulation sites. If the MGA is not recognized, a mistaken impression of a conduction block across the elbow may occur. To demonstrate an MGA in this situation, the median nerve is stimulated (S_M) at the WR and antecubital fossa (AF) while recording the ulnar muscles, looking for a compound muscle action potential that is either present or higher in amplitude at the AF than when stimulating at the WR. This error is avoided if the BE stimulation site of the ulnar nerve is maintained at 3 cm distal to the medial epicondyle. In addition, one should always look for an MGA in a patient with an apparent ulnar neuropathy at the elbow that is diagnosed solely by a conduction block across the elbow without any other abnormalities or clinical symptoms to suggest an ulnar neuropathy.

In one anatomic study of cadavers found to have an MGA, the anastomosis joined the ulnar nerve an average of 8.4 cm (range 5–12) distal to the medial epicondyle, whereas electrophysiologic studies have suggested the possibility of an MGA even more proximal, as close as 3 cm distal to the medial epicondyle. Thus, if the below-elbow site is stimulated 3 cm or further distal to the medial epicondyle (especially >5 cm), there is a possible risk of a proximal MGA mimicking the pattern of an ulnar neuropathy with conduction block at the elbow. As ulnar neuropathies across the elbow typically occur either at the elbow or at the cubital tunnel (under the aponeurosis of the flexor carpi ulnaris), the below-elbow stimulation site needs to be at least 2 cm distal to the medial epicondyle, which is the most distal location of the cubital tunnel. This underscores the need to stimulate the below-elbow site of the ulnar nerve at the proper location, 3 cm distal to the medial epicondyle (see Chapter 10), and not more distally. In addition, one should always look for an MGA in any patient with a conduction block of the ulnar nerve at the elbow without any other supporting abnormalities to suggest an ulnar neuropathy at the elbow.

Ulnar Conduction Study Recording the First Dorsal Interosseous: Pseudo-Conduction Block between the Wrist and Below-Elbow Sites

The most common MGA (Figure 7–4) occurs with crossing over of median-to-ulnar fibers supplying the FDI. Nevertheless, this anastomosis is not often recognized during routine ulnar motor nerve conduction studies because the abductor digiti minimi is the muscle most often recorded for routine ulnar motor studies. However, it is not an uncommon finding when ulnar motor studies are performed recording the FDI. The FDI is commonly recorded in two situations: (1) looking for a lesion of the deep palmar motor branch of the ulnar nerve (i.e., ulnar neuropathy at the wrist), and (2) when evaluating a suspected ulnar neuropathy at the elbow (see Chapter 19).