Denervation is a symptomatic treatment for patients experiencing pain originating from a joint. In the beginning of the last century, Camitz 1 was the first to propose a treatment for patients with hip–joint pain consisting of a neurotomy of the obturator nerve.
In 1931, Camitz 1 stated that the main advantages of the procedure were that it is not technically difficult to perform, not a big surgery for the patient, and it does not preclude any posterior surgery.
To avoid an undesired denervation of other structures, the knowledge of the specific innervation of the joint is essential. Therefore, different authors carried out anatomic studies to offer this possibility of treatment for patients.
In 1942, Tavernier and Trucher 2 tried to select only articular branches for hip denervation. Therefore, no major nerve trunks needed to be killed.
In 1958, at the wrist level, Wilhelm 3 published an exhaustive study on the innervation of the upper extremity and focused attention on the wrist joint. Posterior studies by Fukumoto et al. 4 confirmed, almost completely, the distribution and the source of articular branches. Based on these studies, Wilhelm described a technique for wrist denervation 5 that has been applied mainly by German surgeons. 6–9 In France, Alnot and other surgeons 10 also studied wrist innervation and noted some discrepancies with previous reports.
In spite of theses discrepancies, it has been possible to describe a technique for a wide wrist denervation. Based on other anatomic studies, in 1995, Ferreres et al. 11 concluded that it is not possible to achieve a complete wrist denervation, as there are some branches that would require an enormous exposure. Nonetheless, it is possible to perform an almost complete denervation through the technique described in the present article.
The following nerves innervate the wrist joint:
- Ulnar nerve (motor and dorsal sensory branch)
- Radial nerve (sensory branch and posterior interosseous nerve)
- Lateral antebrachial cutaneous nerve
- Medial antebrachial cutaneous nerve
- Median nerve (anterior interosseous nerve, palmar cutaneous branch, and thenar branch)
The dorsal sensory branch of the ulnar and the sensory branch of the radial nerve give tiny branches to the ulnar and radial sides of the wrist as they turn around the ulnar and radial border. Distally from the branch that is directed to the index finger, there emerges a small nerve that goes deeper to the first intermetacarpal space to participate in the innervation of the first and second carpometacarpal joints.
The motor branch of the ulnar nerve gives a branch from its concavity after it passes around the hook of the hamate. Other branches, described by Wilhelm, which pass through the intermetacarpal spaces, stay close to the intermetacarpal joints, and emerge dorsally where they innervate the carpometacarpal joints of the second–fifth fingers. 3
The lateral antebrachial cutaneous nerve takes part in the innervation of the wrist by means of two thin branches that at the level of the distal third of the forearm, pass near the radial artery and run distally with it. From there, some tiny branches emerge to innervate the first carpometacarpal joint as the artery passes across the base of the first intermetacarpal space.
The medial antebrachial cutaneous nerve gives a branch that goes deep along the radial border of the tendon of the extensor carpi ulnaris. More distally, it has an anastomosis with the most ulnar branch of the posterior interosseous nerve and (Fig. 1) innervates the dorsum of the carpometacarpal joints of the fourth and fifth fingers.
The median nerve participates in the innervation of the wrist joint by means of the anterior interosseous nerve, the palmar cutaneous branch, and the thenar branch (Fig. 2). The anterior interosseous nerve reaches the carpal joints running distally through the periosteum of the radius and also through the volar capsule of the distal radioulnar joint (Fig. 3).
Indications are very extensive. Any wrist pain may be managed by denervation and may be indicated in a wide variety of pathology, such as, scaphoid nonunion, painful carpal instability, occult ganglion, and arthritis, to name a few. Its effectiveness is predictable, and we can mimic denervation by local anesthetic injection (Nyakas test). 12
As a general rule, there are no contraindications to wrist denervation as a result of the technique involved, except for an active infectious process. Surgical technique, by itself, does not have any contraindications. Denervation is a surgical technique that does not preclude any subsequent posterior wrist surgery and may be performed after other previous surgical procedures.
EVALUATIONS AND TESTS BEFORE SURGERY
We believe that the origin of the pain in the wrist may be evaluated by selective anesthetic nerve block. This technique allows the surgeon to cut only the branches that conduct the pain. If after the nerve block, pain has been relieved, we can expect a significant relief of pain after surgery, although some fibers will always be preserved, and a complete denervation is not technically possible.
Tested Nerve Points
Posterior interosseous nerve is tested at a point 5 cm proximal to Lister's tubercle. The needle is introduced until it reaches the dorsal cortex of the radius. Then the ulnar border is localized. At that point, the patient may feel some paresthesiae directed to the dorsum of the wrist. One cubic centimeter of local anesthetic is enough to eliminate nerve conduction. The anterior interosseous nerve may be tested sequentially through the same needle insertion by passing through interosseous membrane.
The sensory branch of the radial nerve is then injected. It is done proximal to the radial styloid, at the radial border of the distal forearm. First, the intermetacarpal nerve has to be injected before the radial nerve and wait for its effectiveness.
Then the dorsal branch of the ulnar nerve is blocked at the level of the styloid process of the ulna.
The palmar cutaneous branch of the median nerve is blocked through its course parallel to the tendon of the flexor carpi radialis. Finally, the terminal branches of the musculocutaneous nerve (lateral antebrachial cutaneous nerve) are blocked at the dorsal and radial border of the distal forearm. This can be done at the same time that the sensory branch of the radial nerve is blocked.
Wilhelm proposes to block the recurrent branches of the ulnar nerve at the base of the intermetacarpal spaces.
A positive response to blocks is more of an indication than a formal requirement for denervation. Other indications may be related to the need of the patient for a painless wrist in a short period. These days, we are accustomed to fix every problem based on its origin, and we often perform partial arthrodesis, bone excisions, and ligament repair that have a long-recovery period. Patients often experience negative symptoms during long-recovery periods, and, thus, denervation may be a useful alternative to other forms of surgical treatment or even a way to postpone surgery.
The technique described here has mainly followed Wilhelm's description of surgical procedure, although Foucher has introduced several variations 13,14 that from our point of view have simplified the technique.
The following nerves will be severed: posterior interosseous nerve, branches of the anterior interosseous nerve, branches from the palmar cutaneous branch of the median nerve, sensory branch of the radial nerve, lateral antebrachial cutaneous nerve, medial antebrachial cutaneous nerve, and the dorsal branch of the ulnar nerve.
The operation is carried out, preferably, under axillary block. After exsanguination with an Esmarch bandage, the tourniquet is inflated at 250 mm of Hg pressure. Optical magnification is useful to identify some tiny branches.
There are five approaches, one anterior and four dorsal.
The anterior approach (Fig. 4) is made over the radial artery and is 5 cm long. It may be longitudinal or zigzag in shape. Periarterial stripping, over 2 cm long, of the radial artery is done to cut the branches away from the lateral antebrachial cutaneous nerve that run distally with it. The accompanying veins are cauterized.
The fascia over the tendon of the flexor carpi radialis is opened, and the palmar cutaneous branch of the median nerve is freed from the deep tissues. Deeper, between the radial artery and the tendon of the flexor carpi radialis, the distal border of the pronator quadratus is identified. The periosteum of the distal epiphysis of the radius is coagulated throughout its width and is elevated from distal to proximal for 0.5 cm long, and subsequently, the terminal branches of the anterior interosseous nerve, directed to the radiocarpal joint, are severed. Branches found in the distal radioulnar joint capsule 11 cannot be divided.
Subcutaneously, the dissection continues in the radial direction and below the radial artery to reach the radial styloid process over the first dorsal compartment where the sensory branch of the radial nerve is found and left attached to the subcutaneous tissue. In doing so, the articular branches of the radial nerve are detached from the main trunk (Fig. 5). The dissection is pursued towards the first intermetacarpal space. The approach is left open.
A second 5-cm–long incision, longitudinal or zigzag shaped, centered 1 cm ulnar to Lister's tubercle is performed (Fig. 6). After opening the extensor retinaculum, the posterior interosseous nerve is found in the floor of the 4th dorsal compartment, accompanied by the artery (Fig. 7). Two centimeters of the nerve are resected.
The plane of dissection is just superficial to the retinaculum or the antebrachial fascia, leaving all subcutaneous fatty tissue with the skin. Subcutaneous dissection is continued toward the radial styloid, and this approach is then connected with the volar one that has been previously described. In doing this, the articular branches of the radial nerve are disconnected from the joint capsule. Subcutaneous dissection is also extended to the ulnar styloid, and the dorsal branch of the ulnar nerve is separated from the deep plane in the same way the radial nerve was separated (Fig. 8). Thus, the articular branches to the ulnar side of the carpus, mainly piso-triquetrum and triquetrum-hamate joints are severed. Subcutaneous blunt dissection is also extended proximally over the antebrachial fascia, to disconnect the branch from the medial antebrachial cutaneous nerve that, distally, anastomoses with the posterior interosseous nerve.
The third incision is a short one over the base of the first intermetacarpal space. The radial dorsal collateral nerve of the index finger sends a recurrent branch. In the nineteenth century, Rauder described this branch in his dissections. This branch is accompanied by a vein that connects the superficial and the deep vascular planes that make its localization easier. Care must be taken to not catch the nerve with the venous net that may mimic the nerve of the index going deeper and be confused with its own branch.
Classically, incisions over the base of the second and third intermetacarpal spaces have been advised to cut branches from the deep branch of the ulnar nerve. Two short transverse incisions are used and the periosteum of the base of the metacarpals elevated by 0.5 cm. We do not agree with the anatomic basis for this approach but mention it for the sake of completeness.
After the procedure is completed, only closure of the skin is done. No drains are used. A cast can be used for 1 week after the removal of skin sutures, but patients are not restricted in any way.
Hematoma may be prevented by a good hemostasis with a bipolar coagulator.
Neuromas will form from a section of the nerves, but have not been clinically significant. Transient paresthesiae are sometimes reported by the patients 9,15,16 and have always been reported in the radial nerve region.
Wilhelm described branches from the motor branch of the ulnar nerve that follow a volar-dorsal direction and emerge at the level of the base of the metacarpals. So, he advised doing two supplementary incisions on the second and third intermetacarpal spaces to cut these recurrent branches. He advised that elevating the periosteum of the bases of the metacarpals over a 0.5-cm length was sufficient to sever these branches.
Although some authors disagree with this anatomic finding of Wilhelm, 10,11 the first author of the present work noted that in two cases, while operating under peripheral blocks for a dorsal ganglion and a carpal boss, he had to block the ulnar nerve to continue the procedure without patient discomfort.
In 1983, another author, Zeman 17, proposed cutting the posterior and anterior interosseous nerves through a dorsal incision made in the distal forearm. After localizing and cutting the posterior interosseous nerve underneath the tendons of the extensor digitorum communis, he proposes opening the interosseous membrane and cutting the anterior interosseous nerve. Although a useful concept, the anterior interosseous nerve is thus cut before innervating the pronator quadratus muscle. In our opinion, this is not justified, because the pronator quadratus is a potent stabilizer of the distal radioulnar joint. We prefer the method described by Wilhelm to cut branches from the anterior interosseous nerve. Although incomplete, it preserves the innervation of the pronator quadratus.
Different authors give an overall expectancy of more than 66% for a good result. 6,7,9,16–19 This is definitely a good percentage, keeping in mind the kind of procedure we are dealing with, the quick recovery, and the fact that it does not preclude any subsequent surgery.
Because the denervation is never anatomically complete, Charcot-joint changes never develop, and this constitutes more of a theoretical complication rather than a real complication. If arthrosis progresses, we have to assume that it is caused by the evolution of the underlying disease or traumatic process.
Although it cannot be called a complication, pain recurrence may occur in one third of the cases after several years.
1. Camitz H. Die deformierende Hüftgelenksathritis und speziell ihre behandlung. Acta Orthop Scand 1933; 4:193–213.
2. Tavernier L, Truchet P. La section des branches articulaires du nerve obturateur dans le traitement de l'arthrite chronique de la hanche. Rev Orthop 1942; 28:62–8.
3. Wilhelm A. Zur Innervation der Gelenke der oberen Extremität. Z Anat. Entwicklungsgesch 1958; 120:331–71.
4. Fukumoto K, Kojima T, Kinoshita Y, et al. An anatomic study of the innervation of the wrist joint and Wilhelm's technique for denervation. J Hand Surg [Am] 1993; 18:484–9.
5. Wilhelm A. Die Gelenkdenervation und hire anatomischen Grundlagen. Hefte zur Unfallheilk 1966; 86:1–109.
6. Geldmacher J, Legal HR, Brug E. Results of denervation of the wrist and wrist joint by Wilhelm's method. Hand 1972; 4:57–9.
7. Buck-Gramcko D. Denervation of the wrist joint. J Hand Surg [Am] 1977; 2:54–61.
8. Reill P. Denervierung des Handgelenkes. Aktuelle Probleme in der Chirurgie und. Orthopädie 1977; 2:67–71.
9. Ekerot L, Holmberg J, Eiken O. Denervation of the wrist. Scand J Plast Reconstr Surg 1983; 17:155–7.
10. Dubert T, Oberlin C, Alnot JY. Anatomie des nerfs articulaires du poignet. Ann Chir Main 1990; 9:15–21.
11. Ferreres A, Suso S, Ordi J, et al. Wrist denervation, anatomical considerations. J Hand Surg [Br] 1995; 20:761–8.
12. Nyakas A. Unsere Erfahrungen mit der Denervation des Knöchel-und tarsalen Gelenkes. Zbl Chir 1958; 83:2243–9.
13. Foucher G. Technique de dénervation du poignet. Ann Chir Main 1989; 8:84–7.
14. Ferreres A, Suso S, Foucher G, et al. Wrist-denervation, surgical considerations. J Hand Surg [Br] 1995; 20:769–72.
15. Abbiati G, Petrolati M, Delaria G, et al. La denervazione articolare del polso. G Ital Ortop Trauma 1993; 19:543–52.
16. Foucher G, Da Silva JB, Ferreres A. La dénervation totale du poignet: a propos de 50 cas. Rev Chir Orthop 1992; 78:186–90.
17. Zeman M. Denervace radiokarpálního kloubu. Sborníc lékařský 1985; 87:344–9.
18. Martini AK, Frank G, Küster HH. Klinische Erfahrungen mit des Handgelenks-denervation nach Wilhelm. Z Orthop 1983; 121:767–9.
19. Röstlund T, Somnieer F, Axelsson R. Denervation of the wrist joint: an alternative in conditions of chronic pain. Acta Orthop Scand 1980; 51:609–16.