Amputation Pain

J. Leonard Goldner, M.D.

Persistent pain after amputation is related to either peripheral nerve involvement or to unusual phantom pain. This article will deal primarily with painful neuroma. An article concerned with recent advances in the management of a heretofore unsolvable condition-persistent phantom pain-appears elsewhere in this issue of the ICIB.

An "amputation neuroma" is a nonneoplastic tumor occurring at the central end of a cut nerve as the nerve fibers or axons grow from the open proximal end and become incorporated in fibroblastic tissue derived from supporting elements of the nerve or from adjacent traumatized muscle or fascia. This proximal neuroma begins immediately after the laceration and requires several weeks or months before becoming a mature, circumscribed, nonproliferating mass. The neuroma may become firmly attached to surrounding soft tissue or bone, resulting in a traction stimulus each time this segment of the extremity is moved. Traction, percussion, or emotional stimuli cause a painful, unpleasant sensation. When the neuroma is stimulated, paresthesias, usually described as "pins and needles," occur along the course of the normal dermatome or myotome, in addition to pain produced directly at the point of pressure.

Histopathology of Neuroma Formation

In the formation of a peripheral nerve, two cell types from ectodermal elements are involved. The neuroblasts are the embryonic nerve cells, and the sheath of Schwann cells are the supporting element of the nerve. The efferent fibers eventually end in motor end plates, while the afferent fibers attach to specialized sense organs such as free nerve endings for pain, tactile discs, Pacinian corpuscles, and neuromuscular spindles. As the embryonic nerve leaves the spinal cord, it requires a blood supply as well as a connective tissue covering of mesenchymal origin, closely investing the nerve as epineurium, perineurium, and endoneurium.

The mature nerve is made up of numerous fascicles of nerve fibers surrounded by layers of connective tissue. The fibers consist of axons and their associated sheaths of Schwann, the latter being composed of cells with flat oval nuclei which wrap themselves around the axon and connect with each other by cell processes at the nodes of Ranvier. The myelin developing in the Schwann cells occurs as a "highly ordered lamellar system of membranes." The outer layer of connective tissue, the epineurium, is composed of longitudinally arranged connective tissue cells with connecting fibers. The nerve fascicles are surrounded by perineurium from which strands of connective tissue pass through and around the individual nerve fibers, forming the endoneurium (sheath of Henle).

Nerve tumors may thus be derived from the Schwann cells, which are ectodermal in origin, or from specialized nerve endings. Mesodermal connective tissue at the periphery of the entire nerve unit may give rise to a tumor, and the blood vessels in and around the nerve contribute to specific lesions or serve as a source of metastasis from a distance. Axons do not form true neoplasms, but do contribute to the development of amputation neuromas. These must be differentiated from perineural fibrosis such as occurs in the plantar digital nerve. The latter is really a mesenchymal thickening with subsequent enlargement of the entire nerve but without hypertrophy of the axons.

Amputation Neuroma Pain

The amount of pain derived from an amputation neuroma depends on many factors:

  1. The circumstances surrounding the initial amputation, such as crushing stretching, incomplete laceration.

  2. Proximal stretching of the brachial or lumbosacral plexus and damage to the roots or posterior ganglia.

  3. Prolonged ischemia.

  4. Multiple nerve injuries with a summation factor.

  5. Environmental factors affecting the particular nerve involved, such as inadequate adjacent soft tissue, adherence of nerve to bone, or excessive length of amputation stump with excessive soft tissue.

The chronically painful neuroma is usually associated with an exaggeration of existing phantom pain or the recurrence of a phantom sensation which in itself is readily tolerated by the patient. In certain instances, however, the pain can be compared to causalgia or the burning sensation described by Weir Mitchell relative to sciatic and median nerve injuries. Causalgia, reflex sympathetic dystrophy, and Sudeck's atrophy and its associated pain are usually relieved by sympathetic nervous system ablation in the area involved. This treatment does not affect neuroma or phantom pain. The pain of neuroma is primarily local; the phantom pain is probably related to a central excitatory state as described by Head, with emphasis on the internuncial pool of the spinal cord or higher sensory centers, with resulting summation of abnormal stimuli and persistence of pain pattern due to higher-level involvement.

Not every amputation neuroma is painful. Postoperative percussion of an amputated nerve end causes local pain and exaggerates painful phantom in the anatomic zone of the nerve. Gradually, the neuroma becomes more tolerable, provided that excessive stimulation from persistent traction or pressure does not occur. For example, a digital nerve lacerated in the palm and not repaired will form a neuroma, become less painful to tapping and digital motion, but will continue to be painful to direct firm pressure for an unlimited period of time. The nodule contains unmyelinated nerve fibers that respond to excessive compression in the same way that a normal nerve responds to pressure in the ulnar groove or around the fibular neck. The distal segment of the lacerated nerve is not painful because of absence of connection with the central nervous system and loss of its afferent pathways.

Variations in Neuroma Formations

Variability in the behavior of amputation neuromata depends upon the kind of injury, the level of amputation, and the condition of the soft tissue and bone adjacent to the neuroma. Following are illustrative examples :

  1. Traumatic amputation of the distal phalanx of the digit without proximal amputation of the digital nerve, resulting in adherence of the nerve to the flexor tendon, the phalanx, or the scar. If a painful amputation neuroma occurs, this will not be resolved until resection is done or proximal nerve section is completed.

  2. Digital nerve laceration within the digit or palm with spontaneous neuroma and fibrous formation and gradual subsidence of pain to percussion, but persistence of pain to firm compression. Once this neuroma is resected, the nerve resutured, and regeneration occurs, the neuroma pain gradually disappears. It is assumed that the involved nerve becomes insulated by the sheath of Schwann, with gradual protection of unmyelinated fibers.

  3. Median nerve adherence to flexor tendons, fascia, or distal radius after amputation of the hand at the wrist. Painful neuroma will persist until the median nerve is removed from the scar, shortened by nerve amputation, and placed in soft tissue. Many techniques have been proposed for amputating a nerve. These will be discussed later.

  4. Ulnar neuroma adherent in the midforearm or near the elbow, together with neuromata of the median and radial nerves, which interferes with prosthetic fitting. Adherence of three major nerves usually results from delayed primary closure of an amputation stump, accompanied by proliferating fibrous tissue reaction either from trauma or infection or inadequate resection of the involved nerve at the time of initial amputation or delayed closure.

Complete revision of the amputation stump can be avoided if the nerves are not adherent to bone or do not show excessive involvement in the terminal scar. Mobilization of the flaps can be avoided by isolating the nerves about 3 inches proximal to the end of the stump and resecting a 1 cm segment from each nerve. This does not interfere with cutaneous sensation because of adequate overlap from the higher cutaneous innervation. The proximal neuroma that forms is usually away from the pressure point of the socket, is not subjected to excessive traction or torsion, and usually remains comfortable.

This technique has been successful in eliminating the painful neuromata in both below-elbow and below-knee amputation stumps. It has been used in the upper arm by sectioning nerves in the axilla, rather than by revising an above-elbow amputation, and in the sciatic and femoral nerves at a level higher than the end point of the amputation. The same principle has been useful in the hand where all the digits have been amputated through the interphalangeal joint area, with resulting adhesions of the digital nerves to the ends of otherwise satisfactory amputation stumps. Rather than open each amputation stump, it has been expedient to isolate the common digital nerves 1 1/2 cm proximal to the end of the stump and resect a small amount of each nerve. This does not interfere with sensation of the terminal skin, as these areas are supplied by more proximal cutaneous branches.

5. The posterior tibial nerve is frequently the source of a painful neuroma in a long below-knee stump. Particular care must be taken when doing a Syme's or supramalleolar ablation to resect this nerve at least 3 cm proximal to the distal point of the tibia. The posterior tibial nerve will not tolerate excessive fibrosis, traction, or fixation, and pain resulting from any of these problems is managed either by resection when a necessary revision of the stump is done or by proximal section if the amputation end is in good condition except for the adherent nerve.

Treatment of Nerves at Time of Amputation or Revision

The surgical technique of nerve amputation varies, both at the time of initial amputation or at later revision. One satisfactory method has been to apply a large hemostat to the nerve, exert moderate traction on the nerve trunk, and incise the nerve proximal to the hemostat about 4 cm above the compression point of the clamp. The epineurium will usually cover the open fascicles as they rebound. Prior to the incision, however, the nutrient artery is identified and ligated without placing a suture around the nerve end.

Experimental work and clinical investigation in the past have resulted in numerous materials being injected in and around nerves, including methyl 2-cyanoacrylate (Eastman 910 Monomer). Our initial use of Monomer was as a sutureless nerve glue, but our study showed that this was not practical or successful. However, very small neuromata occurred when this material was used, and subsequently we applied Monomer to open nerve ends in rabbits in order to abort neuroma formation. This procedure was successful, but the plastic material caused a foreign-body reaction with granuloma formation, and it was necessary to discontinue its use.

Absolute alcohol has been injected into nerves at the point of amputation for many years, causing degeneration of the myelin sheath. This has been helpful in some instances, but is usually a temporary measure. A 10% solution of formalin has been used to produce more permanent and vigorous proximal degeneration and inhibition of axon growth. Local cautery at the time of amputation and other forms of electrocoagulation, both at primary amputation and reamputation, have been used, but in our experience all of these can be avoided at the time of primary amputation. There appears to be a place for their use in secondary or reamputations when a painful pattern has already been established. We have depended on the use of electrocautery or 10% formalin if adequate soft tissue has not been available to protect the nerve at the time of reamputation, and this has yielded reasonable success.

Petropoulous and Stefanko, in an experimental study, have used cauterization with electrocoagulation and freezing, as well as ligation of nerve trunks above the site of resection. These methods produced only small neuromata. Similarly, implantation of the nerve into muscle, and nerve trunks injected with 5% formalin, hydrochloric acid, pepsin, alcohol, or phenol, resulted in the development of small neuromata or none. The authors indicated that their best results were obtained with local application or general administration of nitrogen mustard.

Comparison of various methods indicates that minimal crushing of the nerve in order to maintain the epineurium, cauterization of the terminal vessels to prevent bleeding, and application of a nylon or nonabsorbable suture about 1 cm proximal to the amputation site should prevent axon regeneration and diminish the possibility of axon attachment to surrounding tissue.

Congenital amputations are not associated with painful amputation neuromata and do not produce phantom sensations. Amputations in children prior to age 4 are not related to phantom sensation or phantom pain. The condition of the nerve after amputation has some effect on both phantom sensation and phantom pain, but the central image is important also.

Clinical Methods To Eliminate Neuroma Pain

Percussion. According to Russell and Spalding, the percussion technique includes tapping the neuroma with a rubber hammer a few minutes each hour each day for several weeks or months. Theoretically the unmyelinated nerve fibers degenerate or gradually fibrose as a result of external pressure, and pain diminishes accordingly. Frequently the nerve is so severely involved initially that the patient can not tolerate percussion directly. It is possible to diminish pain during the first few days by applying a blood pressure cuff to obstruct venous flow and then apply percussion.

Direct injections. A local anesthetic with or without cortisone injected into the involved area occasionally diminishes pain because the needle puncture causes enough hemorrhage and fibrosis to obliterate the unmyelinated nerve fibers. The anesthetic agent and the cortisone are probably of secondary influence.

Prolonged pressure. Application of a total-contact prosthetic socket applies enough direct pressure to the amputation stump to diminish peripheral pain while the closed socket is on the extremity. Excessive force applied to the arm or leg, firm traction, or suction will produce pain locally and possibly in the phantom area, and for these reasons such external stimuli should be avoided. Tranquilizers, antidepressants, and mood elevators have been tried and usually do not relieve the local discomfort. In our experience, the incidence of painful neuroma has been diminished noticeably in those patients who have had special attention paid to nerve placement and the method of resection, and in those persons who have had early use of the preparatory plastic socket pylon prosthesis.

If local pain persists after several months' trial of percussion, injections, systemic medication, or a total-contact socket, and if percussion continues to result in paresthesias and stimulation of phantom pain, surgical revision is indicated.

Prognosis of Variable Pain Syndromes

The pain syndromes already mentioned are managed in different ways:

  1. The painful neuroma is usually cured or improved by appropriate resection and neurolysis.

  2. Causalgia is improved or eliminated by nerve decompression alone or by neurolysis and sympathectomy.

  3. Severe painful phantom has not responded to any of the usual treatments, including nerve block, sympathectomy, local resection, proximal resection, injection of formalin, tranquilizers, electroshock therapy, hypnotism, or other methods. The lower-extremity phantom has been diminished by spinal thalamic tractotomy, but this type of treatment has not affected upper-extremity problems. Stereotactic neurosurgery has eliminated pain that had previously been impossible to eradicate.

It has been stated that the painful phantom is a depressive equivalent, but the evidence is not in favor of this opinion. The persistent, severe, painful phantom is usually associated with brachial plexus stretch injuries, severe crushing trauma, twisting or avulsion episodes, or similar severe injuries. Other conditions, such as a painful extremity resulting from prolonged vascular insufficiency or chronic fibrosis and associated nerve entrapment, will accentuate phantom sensation and particularly phantom pain.

Dr. Nashold's Presentation

On page 13 Dr. Blaine S. Nashold, of the Division of Neurosurgery, Duke University Medical Center, describes a method of management of persistent painful phantom which had seriously affected the emotional and physical condition of the involved patients.

Each year we see several such individuals with unrelenting painful phantom despite exposure to many forms of treatment. Fortunately, this condition occurs infrequently and involves only .5% to 1% of the total number of amputee patients seen each year. The upper extremity is involved more frequently than the lower.

Recently two patients in our clinic have obtained complete relief from severe phantom pain, which had been interfering with sleep and causing severe depression, by stereotactic surgery. This procedure should not be recommended lightly, as certain complications have occurred, including mild spasticity in the lower extremity on the same side as the phantom, loss of upward visual gaze, and at least temporary interference with balance. Dr. Nashold has assured us that these complications will subside or diminish in intensity and, in fact, both patients already show noticeable improvement. Each has indicated that, in spite of the physical problems involved, the relief of pain has been worth the early postoperative sequelae.

Additional clinical and laboratory work is essential if we are to thorough understand the pain problems of the amputee and develop methods of effective treatment.

J. Leonard Goldner, M.D. is the Professor of Orthopaedic Surgery and Chief of Orthopaedic-Amputee Clinics Duke University Medical Center Durham, North Carolina

Ferlic, Donald C, and Goldner, J. Leonard. "Evaluation of the Effect of Methyl 2-Cyanoacrylate (Eastman 910 Monomer) on Peripheral Nerves," South. Med. J. 58, June 1965, pp 679-685. 
Haymaker, W., and Woodhall, B. Peripheral Nerve Injuries, ed. 2, Philadelphia and London: W.B. Saunders Co., 1953. 
Mitchell, S.W.. "Phantom Limbs," Lippincott's Mag. Pop. Lit. Sci, 8: 563-569, 1871. 
Petropoulous, P.C., and Stefanko, S. "Experimental Studies of Post-Traumatic Neuromas Under Various Physiological Conditions," J. Surg. Resi. 1:235-241, 1961. 
Russell, W.P.R., and Spalding, J.M.K. "Treatment of Painful Amputation Stumps," Brit. Med. J., 2:68, 1950. 
White, James C, and Sweet, William H. Pain. Its Mechanisms and Neurosurgical Control. Springfield, Ill.: Charles C. Thomas.