Clinical utilization and complications of sural nerve biopsy

Clinical Utilization and Complications o f Sural Nerve Biopsy William D. Rappaport, MD, John Valente, Mr), Glenn C. Hunter, MO, Naomi E. Rance, Mr), PhD, Scott Lick, Mr), Tanya Lewis, BS, David Neal, MD, Tucson,Arizona

Surgeons frequently perform sural nerve biopsy as part of the work-up of patients with peripheral neuropathy. The indications for the procedure, therapeutic value, and complications associated with the procedure have received little attention in the surgical literature. A retrospective chart review of 6 0 patients with the suspected diagnosis of peripheral neuropathy undergoing sural nerve biopsy was performed. Vasculitis was suspected in 2 9 (48%) patients undergoing biopsy. This diagnosis was confirmed in 6 of the 2 9 patients and resulted in the alteration of therapy in 31% of patients with this suspected diagnosis. In 27 (45%) patients, the etiology of their peripheral neuropathy was unknown. Twelve (44%) patients in this group had sural nerve pathology; however, no change in therapy was required. Ten patients in our series had associated malignant tumors; some of these patients were diagnosed after referral for sural nerve biopsy. Twenty-five (42%) patients remained undiagnosed after biopsy. Nerve conduction studies were performed in 14 (22%) patients. Thirteen patients with abnormal lower extremity nerve conduction studies had 6 normal and 7 abnormal biopsy results. The one patient with a normal study had a normal nerve biopsy result. There were six (10%) patients with wound infectious, seven (12%) patients with delayed wound healing, and three (5%) patients with new onset of chronic pain in the distribution of the sural nerve, for an overall complication rate of 27%. There was no correlation between the preoperative use of antibiotics, type of local anesthetic used, or length of nerve excised and complication rate. We conclude that the complication rate after sural nerve biopsy is significant. Strict criteria should be employed in selecting patients for sural nerve biopsy including a careful neurologic history and physical examination, nerve conduction studies, appropriate work-up for vaseulitis if suspected, and implementation of a search for malignancy if this is not apparent. If the diagnosis is still in question, then sural nerve biopsy would seem appropriate, especially in patients with suspected vasculitis.

urgeons are often consulted to perform sural nerve biops~as part of the evaluation of patients with symptoms of peripheral neuropathy. The diagnostic information obtained from sural nerve biopsy, however, is frequently redundant or inconclusive [1]. Indeed, the reported 11% rate of significant lasting discomfort following the procedure has prompted the recommendation that sural nerve biopsy be limited to patients in whom important information could be expected [2]. The complication rate and therapeutic value of the procedure have received little attention in the surgical literature. The purpose of our study was to investigate the indications, value, and complication rate of sural nerve biopsy in patients with suspected peripheral neuropathy.

S

PATIENTS AND METHODS A retrospective chart review was conducted of patients with suspected peripheral neuropathy undergoing sural nerve biopsy at the Veterans Administration Medical Center, Tucson, AZ, between 1982 and 1989. Information collected included patient age, general medical condition, and presumed diagnosis at the time of referral. Preoperative studies including nerve conduction tests and electromyography were recorded if performed. Intraoperative information including the site and type of biopsy and the length of nerve removed was also reviewed. All wounds were re-examined at 2 to 3 weeks after surgery, at which time the sutures were removed. The diagnosis of wound infection was based on the clinical appearance of the wound and positive culture results. Wound separation, inability to remove sutures due to lack of healing, or persistent culture negative drainage was classified as delayed wound healing. Any patient complaining of new onset or significantly increased pain in the distribution of the biopsied sural nerve, present after 6 months of observation, was classified as having new onset of chronic pain. Statistical analysis was performed using X2 analysis,

RESULTS A total of 60 male patients with a mean age of 61 years (24 to 86 years) underwent sural nerve biopsy during the study period. All procedures were performed under local anesthesia, with 59 biopsies performed at the level of the ankle and 1 at the level of the gastrocnemius muscle in a patient with severe peripheral vascular disease. The mean length of nerve removed was 20 mm. Fromthe Departmentsof Surgery(WDR,JV, GCH, SL, TL, DN) and The final clinical diagnosis is listed in Table I. VasculiPathology(NER), Universityof ArizonaMedicalCenter,Tucson,Aritis was suspected in 29 (48%) patients with peripheral zona. Requests for reprints should be addressed to William D. Rappa- neuropathy, whereas an unknown etiology was the preopport, MD, Department of Surgery, Universityof Arizona Medical erative diagnosis in 27 (45%) patients. Multiple sclerosis, Center, 1501 North CampbellAvenue,Room 5409, Tucson,Arizona myasthenia gravis, acute myopathy, or amyloidosis was 85724. Manuscript submitted October 5, 1992, and accepted in revised suspected in each of the remaining patients. Confirmation of vasculitis was made in 6 (20%) of the 29 patients. formJanuary 19, 1993. 252

THE AMERICAN JOURNAL OF SURGERY VOLUME 166 SEPTEMBER 1993

SURAL NERVE BIOPSY

Steroid therapy was started in five patients with suspected vasculitis, and empiric steroid therapy was discontinued in four patients, yielding a 31% alteration in therapy rate in this group of patients with suspected vasculitis. Twelve (44%) patients with a preoperative diagnosis of peripheral neuropathy with an unknown etiology had sural nerve pathology. No patient in this group required a change in therapy, and no case of unsuspected vasculitis was discovered. There were 10 patients with malignant tumors, some of whom were diagnosed after referral for sural nerve biopsy. All of these patients presented with weakness and symptoms of peripheral neuropathy, and two (20%) patients had abnormal findings on pathologic examination of the nerve specimen. Twenty-five (42%) patients remained undiagnosed after the biopsy. Nerve conduction studies were performed in only 14 (22%) patients. Thirteen patients with abnormal conduction studies had 6 normal and 7 abnormal biopsy results. The one patient with normal findings on the conduction study had a normal biopsy result. There were six (10%) patients with wound infections, seven (12%) patients with delayed wound healing, and three (5%) patients with chronic pain in the distribution of the biopsied sural nerve 6 months after surgery. Thus, the overall complication rate was 27%. There was no correlation between the preoperative administration of steroids, use of prophylactic antibiotics, type of local anesthetic used, and length of nerve removed and overall complication rate. COMMENTS The sural nerve is the biopsy site most commonly chosen to establish a diagnosis of peripheral neuropathy. It is easily accessible under local anesthesia, and this procedure has, for many years, been presumed to have a limited morbidity. The sural nerve supplies sensation to the lateral aspect of the foot and is made up of both myelinated and unmyelinated fibers. The unmyelinated fibers represent postganglionic, sympathetic innervation to the skin and blood vessels of the lateral aspect of the foot. The peripheral nerve trunk consists of nerve fascicles surrounded by the epineurium, which is composed of fibroblasts and collagen fibers. The epineurium is traversed by large arteries and veins running longitudinally between the fascicles. Each nerve fascicle is enclosed by the perineurium. Different forms of peripheral neuropathy may affect the ultrastructure of the nerve in characteristic ways [3]. The indications for performing sural nerve biopsy vary widely. In our series, the indication for biopsy included patients with clinical symptomatology of peripheral neuropathy. Peripheral neuropathy may result from a number of disease states including vasculitis, diabetes mellitus, inherited neuropathy, demyelinating neuropathy, toxins such as alcohol, and inflammatory and infectious conditions. Vasculitis was suspected in 48% of our patients. Sural nerve biopsy can be quite useful in confirming or ruling out the diagnosis of necrotizing vasculitis. The vasa nervorum frequently are involved with systemic vasculitis;

TABLE I Final Clinical Diagnosis of Patients With Peripheral Neuropathy Diagnosis

No. (%)

Vasculitis Diabetes mellitus Metabolic Paraneoplastic Multiple sclerosis Myasthenia gravis Acute rnyopathy Amyloidosis Unclassified

6 (10) 6 (10) 9 (t 5) 10 (17) t (2) 1 (2) 1 (2) 1 (2) 25 (42)

thus, it is not surprising that peripheral neuropathy may be present in these disease states (Figure 1). Peripheral neuropathy has been reported in 52% to 60% of patients with polyarteritis nodosa; however, it is less commonly seen in other collagen vascular diseases such as systemic lupus erythematosus and rheumatoid arthritis [4]. Prompt diagnosis of vasculitis is essential for appropriate management, and since early diagnosis may be difficult, sural nerve biopsy may contribute to diagnostic accuracy. As seen in our series, a normal biopsy result may lead to discontinuation of steroids in patients with a presumed diagnosis of vasculitis. In our series, six (10%) patients had a final clinical diagnosis of diabetic neuropathy. Patients with diabetes often complain of extremity weakness and abnormalities in sensation. The pathophysiology of diabetic neuropathy is poorly understood. Chronic hyperglycemia may directly affect Schwann cells and axon function, resulting in slowing of nerve conduction [5]. Functional and structural alteration of diabetic capillaries supplying the peripheral nerves may also contribute to the neuropathy seen in this group of patients [6]. Diabetics with polyneuropathy may have other disease entities present, such as vasculitis or a chronic demyelinating process [3]. In this group of patients, sural nerve biopsy may be helpful in differentiating the cause of neuropathy if confusion exists after thorough neurologic evaluation. Although hereditary neuropathy was not detected in our selected population (elderly men), this condition may result in peripheral neuropathy [I]. This can often be diagnosed without the need for sural nerve biopsy, based o n careful history, neurologic examination, and appropriate use of electrophysiologic testing. Biopsy, however, may be useful in detecting asymptomatic individuals [7]. Neuropathy resulting from exposure to neurotoxins such as alcohol and metabolic etiologies such as uremia was present in 15% of our patients. Toxins may cause a variety of abnormalities to the axons and myelin sheaths (Figure 2), resulting in alterations of peripheral nerve conduction [8]. This was strongly suggested based on history and neurologic examinations in our patients, and the need for biopsy in this group of patients is questionable.

THE AMERICAN JOURNAL OF SURGERY VOLUME166 SEPTEMBER 1993 253

RAPPAPORTET AL

Figure 1. Photomicrograph of paraffin-embedded hematoxylin and eosinstained sural nerve biopsy specimen from a patient with vasculitis. Note transmural inflammation and fibrinoid necrosis of an epineural blood vessel (light microscope magnification: X 8oo).

Biopsy may also be required when a suspected neuropathy has a distinctive histologic picture, i.e., amyloidosis, sarcoidosis, leprosy, chronic inflammatory demyelinating neuropathy, and lipoidosis. Evaluation should be complete prior to referral for sural nerve biopsy. This should include a careful history and physical examination, a detailed neurologic examination, and appropriate use of electrophysiologic testing including nerve conduction studies. Argov and coworkers [1] evaluated 120 patients with peripheral neuropathy and noted that almost half could be definitively diagnosed without sural nerve biopsy. Although nerve biopsy was performed in 53 patients, it contributed to the final diagnosis in only 20 patients. These authors noted that the incidence of pathologic findings could be increased by choosing patients with marked slowing of motor conduction velocities of the lower extremity. Thus, they recommend that routine diagnostic sural nerve biopsy be res e r v e d for patients with unidentified neuropathies with abnormal motor conduction velocities. Wees et al [4], in reviewing patients with necrotizing vasculitis, noted that electrophysiologic studies were invaluable in choosing patients for nerve biopsy. They recommend sensory as well as motor nerve conduction studies of the lower extremity and consider abnormal sural nerve conduction a prerequisite for the demonstration of vasculitis by biopsy. In their series, all patients with vasculitis diagnosed by sural n e r v e biopsy had abnormal nerve conduction velocities. In our series, only 14 (22%) patients had electrophysiologic studies. The one patient with normal lower extremity nerve conduction velocities had a normal biopsy result, whereas slightly more than half of the 13 patients with abnormal nerve conduction studies had abnormal biopsy findings. If electrophysiologic studies are normal, the usefulness of sural nerve biopsy is questionable. Ten patients in our series had polyneuropathy associated with a neoplastic process. Some of these patients 254

diagnosed by clinical work-up after sural nerve biopsy was performed. Graus et al [9] noted peripheral neuropathy in 18% of patients with lung cancer and 10% of patients with lymphoma [9]. There appears to be a correlation between the degree of weight loss seen in cancer patients and the development of peripheral neuropathy [10]. Very little information is available in the literature regarding the complication rate following sural nerve biopsy. Wees and coworkers [4] reviewed 17 sural nerve biopsies performed in patients with vasculitis and noted delayed wound healing in 2 patients receiving steroids. Chronic pain in the distribution of the sural nerve was noted in 5% of our patients. Other authors have noted an incidence of such pain in 0% to 11% of patients [2,11]. The length of nerve removed, as well as whole versus fascicular biopsy, does not appear to correlate with the development of chronic pain [11]. Prior to biopsy, the peripheral pulses of the distal extremity should be carefully assessed, and noninvasive studies should be used to evaluate the peripheral blood supply if distal pulses are not palpable. Alternative nerves for biopsy are well described and should be used if there is a question of potential problems with distal wound healing [12]. Handling of the nerve should be kept to a minimum to avoid distortion of the anatomy, and at least 20 mm should be removed to allow for all necessary pathologic studies. Transection of the proximal end of the n e r v e as high as possible allows retraction and may minimize the formation of neuromas. The distal end of the nerve is transected in the same manner as the proximal end. The specimen should be placed on a saline-moistened sponge, and the sponge should be left in an airtight container. Nerves should be processed within 15 minutes after removal to avoid preparation artifacts. The processing of the sural nerve is highly specialized and requires a facility equipped to perform routine as well

were

THE AMERICAN JOURNAL OF SURGERY VOLUME166 SEPTEMBER1993

SURALNERVEBIOPSY

~

D~~

~~ e

IP,

Figure 2. Plastic-embedded sural nerve biopsy specimen from a patient with exposure to a neurotoxin (top). Note the decrease in large myelinated fibers and the thinning of myelin sheaths compared with the specimen from an age-matched control (bottom). This is a nonspecific finding and can be seen in patients with uremia or diabetes and in patients exposed to neurotoxins such as alcohol (Top and Bottom: paraphenylenediaminestained specimens; light microscope magnification: X 800).

as sophisticated studies of the nerve. A well-trained neuropathologist is invaluable to ensure proper handling and fixation of the specimen. Whereas paraffin-embedded histologic preparations are useful in demonstrating vasculitis, plastic embedding is required to demonstrate axonal pathology. Light microscopy can discern the density and distribution of myelinated nerve fibers; however, electron microscopy is useful in assessing unmyelinated nerve fibers. A teased nerve preparation may demonstrate demyelination or active degeneration when this finding is not demonstrated by other studies (Figure 3). Additional nerve specimens may be frozen for biochemical, histochemical, and immunofluorescent studies if indicated [12]. Due to the highly specialized nature of the processing and interpretation of the study results, sural

nerve biopsy specimens are often sent to centers with these capabilities. Based on our study, we recommend the following indications for sural nerve biopsy in patients with symptoms of peripheral neuropathy: (1) when other causes of neuropathy are not apparent after a thorough neurologic evaluation; (2) if vasculitis is suspected; (3) if hereditary neuropathy is suspected, even if the patient is asymptomatic, to aid in genetic counseling; (4) if neurotoxin exposure is suspected, and cause and effect need to be established; and (5) if the suspected neuropathy is undiagnosed and has a distinctive histologic appearance, i.e., amyloidosis. Prior to biopsy, all patients should undergo a thorough neurologic evaluation as well as lower extremity nerve conduction studies. Normal nerve conduction stud-

THE AMERICAN JOURNAL OF SURGERY VOLUME166 SEPTEMBER1993 255

RAPPAPORTET AL

Figure 3. Teased nerve preparation demonstrating two axons undergoing wallerian degeneration (osmiumstained; electron-microscope magnification: X 1,600)(top). Bottom shows normal teased nerve fiber with node of Ranvier. ies, a l t h o u g h not a contraindication to sural nerve biopsy, reduce the value of the procedure. REFERENCES 1. Argov Z, Steiner I, Soffe D. The yield of sural nerve biopsy in the evaluation of peripheral neuropathies. Acta Neurol Scand 1989; 79: 243-5. 2. Solders G. Discomfort after fascicular sural nerve biopsy. Acta Neurol Scand 1988; 77: 503-4. 3. Johnson PC. Diagnostic peripheral nerve biopsy. Barrow's Neurologic Institute Quarterly 1985; 1: 2-7. 4. Wees S J, Sunwoo IN, Oh SJ. Sural nerve biopsy in systemic necrotizing vasculitis. Am J Med 1981; 71: 525-32. 5. Archer AG, Watkins PH, Thomas PK, et al. The natural history of acute painful neuropathy in diabetes mellitus. J Neurol Neurosurg Psychiatry 1983; 46: 491-9. 6. Dyck P J, Hansen S, Karnes J, et al. Capillary number and 256

percentage closed in human diabetic sural nerve. Proc Natl Acad Sci U S A 1985; 82: 2513-7. 7. Dyck P J, Oviatt KF, Lambert EH. Intensive evaluation of referred unclassified neuropathies yields improved diagnosis. Ann Neurol 1981; 10: 222-6. 8. Ahonen RE. Peripheral neuropathy in uremic patients and in renal transplant recipients. Acta Neuropathol (Berl) 1981; 54: 43-53. 9. Graus F, Ferrer I, Lamarca J. Mixed carcinomatous neuropathy in patients with lung cancer and lymphoma. Acta Neurol Scand 1983; 68: 40-8. 10. Hawley RJ, Cohen MH, Saini N , et al. The carcinomatous neuromyopathy of oat cell lung cancer. Ann Neurol 1980; 7: 65-72. 11. Pollock M, Nukada H, Taylor P, et al. Comparison between fascicular and whole sural nerve biopsy. Ann Neurol 1983; 13: 65-8. 12. Johnson PC. Peripheral nerve. In: Davis RL, Robertson, DM, editors. Textbook of neuropathology. Baltimore: Williams and Wilkins, 1991: 1004-87.

THE AMERICAN JOURNAL OF SURGERY VOLUME 166 SEPTEMBER 1993