Clinical manifestations

Presentation and course

The spectrum of clinical manifestations is wide given that the peripheral nerve can be affected at any point from the level of the root to the most distal small, unmyelinated fibers. The etiology of the findings is usually due to a direct effect of HIV itself, opportunistic infections, neoplasms, or medication toxicity. The clinical presentation and etiology of the reported HIV complications of the peripheral nervous system will be considered individually, moving from the root level distally to the peripheral nerve.

Polyradiculopathy. HIV-related lumbosacral polyradiculopathy is characterized by asymmetric leg weakness involving proximal and distal muscles, flaccid paraparesis, hypo- or areflexia, sacral and lower extremity sensory loss, and sphincter disturbances such as urinary retention. The symptoms develop over a period of 1 to 6 weeks. The finding of “saddle” anesthesia, urinary retention, and a sensory level should raise suspicion for polyradiculopathy, rather than another process such as chronic inflammatory demyelinating polyneuropathy. In herpes zoster radiculitis, pain typically precedes a rash, which develops in a dermatomal distribution in the face or trunk. Motor involvement and weakness can also occur in more severe cases (146).

Cytomegalovirus infection has been identified as a cause of polyradiculopathy. It occurs in approximately 2% of patients with a CD4 cell count of 50 cells/µL or less (51). The patient typically presents with radiating, low back pain, progressive areflexic paraparesis, and distal sensory loss. Two thirds of patients complain of urinary difficulties, including hesitancy or retention (24). Weakness is typically more prominent than sensory loss (150). Cerebrospinal fluid evaluation may reveal polymorphonuclear pleocytosis, hypoglycorrhachia, and a raised protein concentration. Polymerase chain reaction of the cytomegalovirus DNA is positive in approximately 90% of affected patients (42). MRI of the lumbosacral spine may reveal thickened, enhanced nerve roots in approximately one third of patients (12; 217; 117). The classical electrophysiologic profile includes denervation of lower extremity and paraspinal muscles, with normal nerve conduction velocities and sensory potentials (15). However, a coexisting HIV neuropathy causing abnormal sensory potentials could cause confusion.

Anticytomegalovirus therapy with ganciclovir, foscarnet, or a combination of both drugs is imperative because of the rapid progression. In one study of patients with AIDS and acute lumbosacral polyradiculopathy, 15 of 23 patients had cytomegalovirus infection, and treatment with ganciclovir was followed by stabilization (213). Oral acyclovir is usually effective in herpes zoster radiculitis (191).

Radiculopathy has also been associated with syphilis (126), lymphoma (129; 74), vasculitis (163), toxoplasmosis (108), tuberculosis (18), primary CNS lymphoma with lymphomatous meningitis (120), and herpes zoster (53). Also, some data suggest that HIV itself may lead to lumbosacral radiculopathy (213).

Plexopathy. Bilateral brachial neuritis has been described at the time of seroconversion, occurring within 1 to 2 weeks of the acute febrile illness, with pain, weakness, scapular winging, and muscle atrophy (131; 29; 25).

Mononeuropathies. The most common cranial mononeuropathy is facial nerve palsy, with less frequent involvement of cranial nerves II, V, and VIII (131; 159; 86). Bilateral facial nerve palsies have also been reported with HIV infection (231; 198), most commonly occurring at the time of seroconversion, but also with advanced disease (13; 156), possibly secondary to opportunistic infection or lymphoma (190). Herpes zoster, neurosyphilis, and hepatitis C have been associated with facial nerve palsies and should be excluded.

Other mononeuropathies, such as neuropathy of the laryngeal recurrent nerve (134) and bilateral phrenic nerves (170) have been reported in HIV-infected patients. These neuropathies may occur as isolated events or may precede the development of multiple mononeuropathies, ie, mononeuropathy multiplex.

Mononeuritis multiplex. The clinical pattern of mononeuritis multiplex is that of the simultaneous or sequential involvement of individual major peripheral nerves in different limbs, resulting in the often abrupt onset of symptoms in variable distributions. This syndrome may develop early or late in the course of HIV infection (211). Several underlying etiologies of this syndrome should be recognized and include vasculitis, cryoglobulinemia, lymphoma, or cytomegalovirus infection.

Mononeuritis multiplex as a manifestation of polyarteritis nodosa, a vasculitis of medium-sized vessels, may occur at any stage of HIV infection. Though vasculitis classically presents as a mononeuritis multiplex, many patients present with an overlapping symmetrical sensorimotor polyneuropathy, or even a distal sensorimotor polyneuropathy, due to the cumulative involvement (20). The neuropathy symptoms may be accompanied by weight loss, myalgias, and leg tenderness (125; 189; 78). Systemic manifestations, though less frequent, include renal failure, skin rash, testicular pain, arthritis, or hypertension. The erythrocyte sedimentation rate is usually elevated, and patients may be positive for hepatitis B surface antigen (125; 74). Mononeuritis multiplex may also occur in association with an unspecified vasculitis (78), involving small vessels with neutrophilic inflammatory vascular disease. A milder form is associated with a pathological picture of perivasculitis in which there are perivascular inflammatory cells but no vasonecrosis or fibrosis (132), with the neuropathy following a more benign course. Cryoglobulinemia, frequently associated with hepatitis B and C, may cause vasculitis, resulting in mononeuritis multiplex in patients with HIV (123; 125; 10; 214; 130). Invasion of the nerves by lymphoma or Kaposi sarcoma can also result in the clinical picture of mononeuritis multiplex (131; 82; 74). In HIV-infected patients who present with mononeuritis multiplex, the possibility of leprosy should also be suspected, particularly in leprosy endemic areas (77).

A subset of HIV-infected patients develops a hyperimmune response to HIV infection by developing an oligoclonal expansion of CD8+ lymphocytes. Diffuse infiltrative lymphocytosis syndrome may develop due to diffuse visceral lymphocytic infiltration, particularly of the salivary glands and lungs (100; 101; 102; 73). A sensorimotor polyneuropathy may develop due to angiocentric infiltration by CD8 cells and vascular mural necrosis (79). In a series of 12 patients with diffuse infiltrative lymphocytosis syndrome and peripheral neuropathy, all had a sicca syndrome and multivisceral involvement. Four had an asymmetrical and eight a symmetrical neuropathy. Nerve biopsy showed marked angiocentric CD8 infiltrates without mural necrosis and with abundant HIV p24 protein in macrophages. Improvement with AZT was seen in six of six patients, and steroids were beneficial in four of five patients. Two patients developed a primary B cell lymphoma (155).

In later stages of HIV infection when patients have less than 50 CD4 cells, cytomegalovirus may result in a rapidly progressive multifocal neuropathy. This may be confirmed by demonstration of cytomegalovirus in the CSF by polymerase chain reaction accompanied by an elevated protein level and predominantly polymorphonuclear pleocytosis. Characteristic nerve biopsy findings include multiple foci of endoneurial necrosis, inflammatory infiltrates of mononuclear and polymorphonuclear cells, and cytomegalovirus inclusions in endothelial cells of endoneurial capillaries with surrounding inflammation (188; 184). Treatment includes antivirals, such as ganciclovir, foscarnet, and cidofovir, along with immune reconstitution with HAART (180).

Distal polyneuropathy. A predominantly sensory, distal symmetric polyneuropathy is the most common peripheral manifestation seen in HIV-1 infection and typically has been reported to occur in later stages of the disease. Advancing age and treatment with nucleoside analogue reverse transcriptase class of drugs are significant predictors of distal sensory neuropathy (154; 158; 105). These comorbidities include vitamin B12 deficiency, substance abuse (137), medications, and diabetes (71; 143). Hepatitis C co-infection is reportedly not associated with an increased risk of developing sensory neuropathy (35).

Older age, lower CD4 nadir, current combination antiretroviral therapy use, and prior “d-drug” therapy are significant risk factors for developing HIV sensory neuropathy (65; 158).

Patients present with paresthesias typically distal to the ankles. Deep tendon reflexes at the ankles are depressed or absent, and intrinsic foot muscle weakness may be seen in 37% of patients (43). Sensory loss and spontaneous pain are frequently reported (65). Physical examination shows decreased distal vibratory and temperature sensation with increased or decreased sensitivity to pinprick in the toes. Strength and proprioception are generally preserved (180).

Electrophysiologic findings are largely axonal, although demyelinating features have been described to a lesser degree (08; 212). Needle electromyography abnormalities are not typical, but chronic reinnervation changes may be seen (180). Pathologic features include axonal degeneration of long axons in distal regions in a “dying back” pattern, with reduction of both small and large myelinated fiber densities (43; 84; 06). Small fiber nerve damage may be assessed by skin biopsy with quantification of the epidermal nerve fiber density. A decreased epidermal nerve fiber density is associated with greater pain intensity and higher HIV plasma RNA levels (172).

Qualitative changes (such as focal swellings) may precede development of symptomatic distal symmetric polyneuropathy and decreased epidermal nerve fiber density (94; 61). Occasionally, HIV may be cultured from the CSF of affected individuals (98).

Antiretroviral toxic neuropathy has been a well-described complication of exposure to nucleoside analogues, particularly zalcitabine (ddC), but also didanosine (ddl) and stavudine (d4T) (60; 124; 36), which are collectively known as “d-drugs.” Symptoms typically occur in the first 6 weeks of treatment, and the risk of developing a toxic distal symmetric polyneuropathy from a d-drug is usually maximal in the first 3 months of therapy. Although symptoms may subside with termination of treatment, complete resolution is atypical (158). The toxic effect is dose-dependent and is estimated to occur in 15% to 38% of patients receiving these drugs (16; 46; 186). Although less commonly used in the Western nations, the “d drugs” remain a valuable therapy in the developing world (186). The combination of these drugs with hydroxyurea is associated with an increased risk of sensory neuropathy, suggesting a possible synergistic effect between the drugs (153). Now the incidence of peripheral neuropathy is expected to lower as stavudine and didanosine are no longer recommended (119). A study noted that in Sub-Saharan Africa the most common reason for antiretroviral drug regimen change was peripheral neuropathy (17).

The clinical presentation often resembles distal sensory polyneuropathy, with predominantly axonal, sensory electrophysiologic findings. Antiretroviral toxic neuropathy may sometimes be distinguished from distal sensory polyneuropathy by its more painful character, abrupt onset with initiation, and eventual amelioration following cessation of the medication (14; 16; 153). There may also be a “coasting phenomenon” in which the symptoms intensify 2 to 4 weeks following withdrawal of the drug. HIV-infected diabetic patients have poorer blood sugar control and a higher incidence of neuropathy and nephropathy (when defined by overt proteinuria) (171).

Up to 90% of those with HIV-associated sensory neuropathy experience pain and have an adverse effect on quality of life (161). The intensity of distal neuropathic pain is not fully explained by the degree of peripheral nerve damage. Presence of early paresthesias significantly predicts disabling neuropathic pain later in life among older people with HIV having distal sensory polyneuropathy (57). A central mechanism has been suggested. The posterior cingulate is thought to be involved in inhibiting the perception of painful stimuli. It has been demonstrated that smaller posterior cingulate cortex structure may be related to reduced antinociception, leading to increased perception of distal neuropathic pain (109). In a study, Keltner and coworkers noted that smaller midbrain and thalamus were associated with paresthesia alone, but small posterior cingulate cortex was associated with both pain and paresthesia both (110). Abnormalities in calcium/calmodulin-dependent protein kinase (CAMKK2) gene predispose HIV-associated distal sensory peripheral neuropathy (187). A diffusion tensor imaging study using tractography and connectometry noted that severe HIV neuropathy was associated with lower white matter integrity extending from midbrain to somatosensory cortex. This phenomenon suggests an ascending deafferentation extending from damaged peripheral nerves up to third-order neurons (219). Ellis and co-workers observed that declining neurocognitive performance in persons with HIV infection was associated with worsened neuropathic pain. This phenomenon indicated that central pain processing mechanisms do contribute to pathogenesis of neuropathic pain in patients with HIV-associated distal polyneuropathy (66).

Neurotoxicity of antiretroviral drugs is associated with mitochondrial dysfunction via the inhibition of gamma-DNA polymerase, leading to depletion of the nerve’s mitochondrial DNA (46). Pathologic studies have revealed prominent mitochondrial disruption and abnormalities of the cristae (33; 165). Neuropathy and acquired lipodystrophy may be manifestations of these mitochondrial effects (27).

The protease inhibitor indinavir was found to be toxic to HIV-infected dorsal root ganglion cultures (167). However, analysis of 1159 HIV-infected individuals enrolled in a large, prospective, observational multicenter study revealed no statistically significant association when adjustments were made for well-known concomitant risk factors (eg, older age, more advanced disease, greater exposure to neurotoxic d-drugs) (64).

Other medications used in the treatment of AIDS-related complications may also lead to polyneuropathy. Vincristine, a vinca alkaloid used in the treatment of solid tumors, such as Kaposi sarcoma, lymphoma, and leukemia, binds with tubulin and is believed to cause a distal axonopathy by affecting axonal transport (220). Isoniazid for the treatment of tuberculosis and dapsone for the treatment of pneumocystis carinii pneumonia or toxoplasmosis may also cause neuropathy. The antifungal and antibacterial agent metronidazole may produce sensory neuropathy (83). Other potentially neurotoxic medications include chloramphenicol, ethambutol, etoposide, pyridoxine, and thalidomide (180).

Inflammatory demyelinating polyneuropathy. HIV is also associated with acute or chronic inflammatory demyelinating polyneuropathies. These are characterized by weakness, absent deep tendon reflexes, elevated CSF protein, and improvement either spontaneously or due to immunomodulation (44; 176).

Acute inflammatory demyelinating polyneuropathies or Guillain-Barré syndrome may have a bimodal pattern of occurrence, developing at the time of seroconversion and prior to the appearance of HIV antibodies (87; 169; 227; 21; 230) or during the chronic phase of HIV infection (09). Acute inflammatory demyelinating polyneuropathy has also been reported to occur as a complication of immune reconstitution (141; 218).

Electrodiagnostic findings consistent with a primary demyelinating neuropathy (such as severe conduction velocity slowing, prolonged distal latencies, and conduction block) are typical; however, other acute inflammatory demyelinating polyneuropathy variants, such as a motor axonal variant, have been reported (230). CSF pleocytosis in a patient with Guillain-Barre syndrome raises the possibility of HIV infection; however, it is frequently not seen (21). Cytomegalovirus inclusions in Schwann cells have also been found (62; 49).

First line therapy is plasmapheresis or intravenous immunoglobulin (IVIg) as it is in HIV-negative patients with AIDP (178).

There have been reports of patients presenting with a syndrome of rapidly progressive weakness, simulating Guillain-Barre syndrome, typically in the setting of lactic acidosis or symptomatic lactatemia. This is presumed to result from mitochondrial toxicity of nucleoside analogue therapy, particularly D4T, causing eventual disruption of the electron transport chain (233; 183; 204; 31). The electrodiagnostic features are primarily axonal, and most patients have not responded to intravenous immunoglobulin or plasmapheresis (144). The Miller Fisher variant (ophthalmoplegia, ataxia, and areflexia) of Guillain-Barre syndrome has also been described in a patient receiving D4T therapy (199).

These patients may be at increased risk of developing CIDP (21). CIDP tends to develop during the early stages of HIV infection, although it may occur in the setting of AIDS (149; 173; 21). The CSF studies may reveal pleocytosis and the presence of viral antigens. Inflammatory infiltrates, sometimes seen in sural nerve biopsies, consist of CD8+ lymphocytes and macrophages, some of which harbor HIV antigens (47).

Autonomic neuropathy. Autonomic neuropathy becomes clinically relevant in advanced stages of HIV disease (185), generally occurring in conjunction with distal sensory polyneuropathy. Presenting symptoms typically include postural hypotension or gastroparesis (24). An increased incidence of papillary abnormalities (81) and denervation in the jejunal mucosa (11) has been reported. Increased abnormalities by cardiovascular tests, such as prolongation of the QT interval (229), may place the patients at risk during invasive procedures (45; 228). The HIV-associated adipose redistribution syndrome has also been attributed to the adverse effects of antiretroviral drugs on autonomic neurons in the CNS that project to visceral or subcutaneous white adipose tissue compartments (72). A study noted that the vagal nerve involvement in HIV-associated autonomic neuropathy is associated with immune and gastrointestinal function changes in treated HIV patients (179).

Amyotrophic lateral sclerosis (ALS)-like motor neuropathy. Amyotrophic lateral sclerosis (ALS) is a progressive disease resulting from upper and lower motor neuron degeneration. A small percentage of HIV-infected persons develop ALS-like syndrome. Neurologic manifestations in HIV-infected ALS patients resemble classical ALS. It occurs at a younger age and shows a dramatic improvement with antiretroviral therapy (04).

Prognosis and complications

The staging of the HIV infection (determined by the total CD4+ lymphocyte cell count, the total HIV viral load, and the patient’s clinical condition) is a determinant factor in the particular peripheral nerve complication that will occur. The staging also will affect the prognosis.

Clinical vignette

A 40-year-old man with HIV infection for 4 years complained of symmetrical pain in the soles of both feet. The pain was hard, sharp, and achy and had progressed over the past 3 and a half months. The pain was worse at night, and he reported hypersensitivity to touch with bedsheets or shoes. His symptoms began 1 month after starting the medication didanosine. Didanosine was replaced with stavudine; however, the symptoms persisted. He denied weakness or hand symptoms. He denied a history of alcohol or intravenous drug use, though he formerly used crack cocaine.

On neurologic examination, his mental status and cranial nerve functions were normal. His motor strength was normal without atrophy or fasciculations. Deep tendon reflexes were absent at the ankles. He had decreased sensation to vibration and pinprick in the legs distally. His gait was steady.

His CD4 count was 483, and his viral load was less than 200. The following lab studies were normal: hepatitis B and C antibody levels, BUN, creatinine, random glucose, vitamin B12 level, RPR, TSH, serum protein electrophoresis, and antisulfatide antibodies.

Amitriptyline, 75 mg at night, provided mild relief. Gabapentin, 600 mg three times daily, provided no benefit.

Five and a half months after the onset of his symptoms, stavudine was discontinued, and his antiretroviral regimen was changed to abacavir, efavirenz, and lamivudine. After 1 month, his pain improved briefly before worsening. He increased his gabapentin to 1200 mg three times daily, without any noticeable benefit. He had no improvement with zonisamide, 600 mg daily. Lamotrigine was begun at 25 mg per day, and a rash and gastrointestinal upset occurred. He was referred to a pain specialist.

Biological basis

Etiology and pathogenesis

The etiology of the various peripheral nerve disorders is usually due to opportunistic infections, neoplasms, medication toxicity, or possibly secondary to invasion and destruction by HIV itself. (See Clinical Manifestations section for further details.)

The mechanisms behind peripheral nerve complications of HIV have not been fully elucidated. Resulting disease is probably a multifactorial process with current data implicating: (1) toxicity due to macrophages, cytokines, antibodies, viral proteins, mitochondrial dysfunction, or oxidative damage; (2) direct infection; (3) nutritional deficiencies.

Toxicity due to macrophages and cytokines. HIV viral transcripts and antigens have been detected in perivascular and endoneurial macrophages (68; 234). Activated macrophages may result in the elevated production of destructive neural toxins (such as tumor necrosis factor, interleukin-1, and interleukin-6) and the depressed production of others (85; 200; 104). High numbers of macrophages expressing MHC class II markers, tumor necrosis factor alpha, interleukin-1, and interleukin-6 were noted in sural nerve specimens from patients with painful sensory neuropathy (85). These macrophages may play a critical role in causing demyelination, allowing activated lymphocytes and macrophages access across the blood-nerve barrier (47). In one study, the dorsal root ganglia of HIV-infected patients were exposed to supernatants from HIV-infected and activated macrophages, resulting in neuronal retraction in the dorsal root ganglion neurons and oxidative stress in the neuronal cell body (89).

Proinflammatory cytokines and other mediators of inflammation have been demonstrated in the dorsal root ganglia of patients with HIV infection (234; 157). TNF-alpha can activate HIV viral production of latently infected neural cells (221; 216) and induce apoptosis in neuroectodermal cells (197). Arachidonic acid, another substance released by activated macrophages, can activate neuronal NMDA receptors, resulting in an influx of calcium ions and stimulation of nitric oxide synthase. Additionally, macrophage-derived factors such as prostaglandin E2 and leukotrienes may sensitize C polymodal nociceptors and contribute to the neuropathic pain and hyperpathia commonly seen in HIV-neuropathy (84).

Some studies have focused on understanding the high rates of HIV sensory neuropathy, even with successful antiretroviral therapy. These studies indicate that neuronal damage is caused by HIV itself and antiretroviral therapy neurotoxicity. This damage involves mitochondrial dysfunction and activation of the innate immune system in peripheral nerves, leading to increased secretion of proinflammatory cytokines. Additionally, the production profile of these cytokines, influenced by allelic polymorphisms in cytokine gene regulatory regions, affects various diseases, including HIV sensory neuropathy (127).

A study in Northern Greece involving 171 people living with HIV, categorized based on the presence or absence of peripheral neuropathy, highlighted the role of cytokine gene polymorphisms in the pathogenesis of HIV sensory neuropathy (162). For instance, carriers of certain genotypes (IL1a-889/rs1800587 TT and IL4-1098/rs2243250 GG) showed a higher risk for developing HIV sensory neuropathy, whereas others (IL2+166/rs2069763 TT) had a lower probability. These findings, though preliminary, suggest potential genetic markers for susceptibility to HIV sensory neuropathy.

Another study explored how HIV-associated autonomic neuropathy contributes to immune hyperresponsiveness. Forty-two adults with well-controlled HIV underwent autonomic testing, resulting in various Composite Autonomic Severity Scores. Analysis of blood-based immune markers and their correlations depicted increased network complexity with greater HIV-associated autonomic neuropathy severity, linking HIV-associated autonomic neuropathy to chronic immune activation in HIV.

A study on HIV sensory neuropathy in Indonesian and South African patients identified polymorphisms in CAMKK2, P2X7R, and P2X4R genes as risk factors (76). Skin biopsies from HIV+ Indonesians, with and without HIV sensory neuropathy, showed different expressions of these proteins, implicating their role in the condition. For example, CaMKK2 was found near nerve fibers in cases of nerve damage/repair, and varying expressions of P2X7R and P2X4R were observed in specific donor groups.

Toxicity due to viral proteins. The HIV-1 viral envelope protein gp120 has been discovered to exert a toxic effect on neurons through activation of the complement cascade (49; 05) or activation of macrophages to secrete neurotoxic factors (133; 80). The latter process may be enhanced by the loss of certain CD4 cells, which normally counteract macrophage activation by elaborating interleukin-4 and interleukin-10 (41; 232; 223). Gp120 has been shown to bind to CXCR4, a chemokine receptor found on neurons and Schwann cells (114) and can directly induce apoptosis or cellular dysfunction (97; 75). Direct epineurial application of gp120 has been implicated in inducing neuropathic pain (96).

Tat, a viral protein released from HIV-infected cells, is known to activate HIV viral expression, modifying several cellular functions with resulting neurotoxicity (139). Tat, through upregulation of inflammatory cytokines, may facilitate entry of the virus into the central nervous system and play a critical role in the progression of neurologic impairment in the course of HIV-associated dementia (175; 174).

Toxicity due to antibodies. There is some evidence that patients with HIV may have antibodies to components of the peripheral nerve, such as IgM antisulfatide (26; 52; 136). A higher incidence of IgG sulfatide antibodies was noted in HIV patients with distal sensory neuropathy versus HIV patients without distal sensory neuropathy (135). HIV-related autoimmunity has been related to the structural homologies between HIV-1 and immune-regulatory molecules and also to components of myelin, both in the CNS (164) and PNS (202).

Moodley and colleagues assessed the frequency of nodal-paranodal antibodies in HIV-infected patients with chronic immune-mediated radiculoneuropathies in 24 HIV-infected patients diagnosed with various forms of immune-mediated radiculoneuropathies, including chronic inflammatory demyelinating polyneuropathy, ventral root radiculopathies, and dorsal root ganglionopathies, with some patients exhibiting combined central and peripheral demyelination (151). The study evaluated IgG antibodies against nodal (neurofascin 186) and paranodal (neurofascin 155, contactin-1, and contactin-associated protein) cell adhesion molecules using a live cell-based assay. Further, the potential pathogenicity of these antibodies was explored by examining their binding to myelinated co-cultures and assessing complement activation as a possible mechanism for myelin injury. The authors observed that three patients (12.7%) tested positive for neurofascin IgG1 antibodies: one patient with ventral root radiculopathies was positive for NF186, and two others with dorsal root ganglionopathies and a mixed sensory-motor demyelinating neuropathy with optic neuritis were positive for NF155. The authors concluded that the frequency of nodal-paranodal antibodies in HIV-infected patients with immune-mediated radiculoneuropathies is similar to those without HIV. However, the pathogenicity of these antibodies, especially at low titers, remains uncertain in the context of HIV.

Mitochondrial dysfunction. Mitochondrial abnormalities in axons and Schwann cells have been described in sensory nerves of patients with zalcitabine (ddC)-induced neuropathy. This dysfunction is attributed to inhibition of gamma-DNA polymerase, the enzyme responsible for mtDNA replication (46). Variations in the mitochondrial genome (138; 34; 30) and cytokine genotype (37) may influence susceptibility to nucleoside reverse transcriptase inhibitor toxicity. Roda and colleagues demonstrated that both the skin intraepidermal nerve fiber density and sural nerve amplitude correlated with the amount of mitochondrial mutation in people with HIV, specifically in those with HIV-associated sensory neuropathy (181).

Oxidative damage. Gp41 may be neurotoxic by a mechanism involving induction of iNOS that has been studied in HIV dementia (02). Similar mechanisms may play a role in HIV neuropathy.

Direct infection. HIV has been shown to infect ganglion neural cells in vitro (121) and in vivo. The blood-nerve barrier is relatively permeable at the site of the dorsal root ganglion, and HIV DNA and RNA sequences have been demonstrated in the dorsal root ganglion neurons and supporting cells of patients with a predominantly sensory peripheral neuropathy (20). Patients without neuropathy at a similar stage of HIV infection did not have evidence of the HIV genome in neurons. HIV viral RNA has also been amplified from sural nerve biopsies (54; 48), and HIV C2V3 envelope sequences have been amplified from peroneal nerve samples obtained from HIV/AIDS patients (104). As described previously, low levels of HIV replication have also been detected in the perivascular or supporting cells in the dorsal root ganglion (234). These findings, though suggestive of a correlation, do not prove that direct HIV infection of neurons results in neuropathy (47; 20). Because of the low level of neurons affected and the inability of techniques, such as immunohistochemistry, to demonstrate production of viral proteins, most investigators have focused on mechanisms other than direct HIV neuronal infection as the cause of neuropathy or HIV dementia.

Coinfection with HTLV-2 has been shown to increase the risk of HIV neuropathy (59; 236).

Nutritional deficiencies. Nutritional deficiencies resulting from anorexia, diarrhea, or malabsorption should be considered. Vitamin B12 deficiency was reported in some studies (92; 116) but not confirmed by others (177; 222; 226). Impaired production of nerve growth factor due to the depletion of CD4+ T-cells has been proposed as a possible mechanism. Acetyl-carnitine deficiency has also been reported in small studies of HIV patients taking dideoxynucleotide drugs (70), and its supplementation was beneficial in a small open-label study (93). (See the Management section for further details.) Plasma carnitine levels were not correlated with severity of neuropathy in larger studies (206).

Central pain processing mechanisms. Central pain processing mechanisms are held responsible in neuropathogenesis of HIV neuropathic pain. A functional magnetic resonance imaging-based study revealed a significant role of increased anterior insula activation in the pathogenesis of HIV neuropathic pain (215).

Gut microbial dysbiosis. Pathogenic changes in the spectrum of gut microbiota is also thought to play a role in the pathogenesis of HIV neuropathic pain. Ellis and co-workers noted that gut dysbiosis, characterized by reductions in diversity and relative increases in the ratios of Blautia and Clostridium to Lachnospira, contribute to pathogenesis of distal neuropathy in persons with HIV infection (63).

Prevention

As the incidence and severity of HIV-associated distal symmetric polyneuropathy have been shown to be positively correlated with the plasma HIV viral load (39; 205), aggressive control of the underlying disease with maintenance of the CD4 count above 500 cells/µL remains the key preventive measure against peripheral nerve dysfunction (223; 145). Administration of picomolar amounts of the hormone erythropoietin was demonstrated to prevent sensory axonal degeneration and in vitro dorsal root ganglion neuronal death by both gp120 and ddC (a neurotoxic dideoxynucleoside drug) (113). One study has also found that specific polymorphisms in the hemochromatosis gene may confer protection from the development of neuropathy with exposure to dideoxynucleoside drugs (106).

Differential diagnosis

Certain neurologic complications are known to occur at particular stages of HIV infection; therefore, an accurate determination of the patient’s viral load and CD4 cell count can lead to a more focused diagnostic evaluation. Acute demyelinating polyneuropathy (Guillain-Barre syndrome) and mononeuropathies are more commonly associated with the early phase of HIV disease, when the CD4 cell count is greater than 500 cells/µL. At the other end of the spectrum, distal sensory polyneuropathy, HIV-related lumbosacral radiculopathy, cytomegalovirus polyradiculopathy, cytomegalovirus-induced mononeuritis multiplex, medication-related neuropathy (antiretroviral drugs), and autonomic neuropathy occur more commonly in patients with AIDS or CD4 cell counts less than 200 cells/µL. Other conditions such as CIDP, mononeuritis multiplex, peripheral neuropathy in diffuse infiltrative lymphocytosis syndrome, syphilitic polyradiculopathy, hepatitis C neuropathy, HTLV type I neuropathy, and motor neuron disease syndrome have been more frequently described in moderately advanced disease, when the CD4 count is 200 to 500 cells/µL.

Dideoxynucleoside analogue drugs--eg, zalcitabine (ddC), didanosine (ddI), and stavudine (d4T)--(38) have been associated with distal sensory polyneuropathy. Antiretroviral medications have been implicated in autonomic dysfunction. In addition, medications used to treat opportunistic infections and tumors, such as dapsone, isoniazid, metronidazole, and vincristine, may also cause neuropathy.

Diagnostic workup

A careful history and examination are essential to determine the disease stage and to localize the area of involvement. Electromyography or nerve conduction studies are useful in determining the distribution of nerve involvement and the degree of axonal or demyelinating damage. A lumbar puncture may detect elevated protein levels in AIDP or CIDP (08; 43) and pleocytosis in cytomegalovirus polyradiculopathy or mononeuritis multiplex (184). CSF studies may reveal evidence of HIV itself (98), cytomegalovirus (by polymerase chain reaction) (184), or elevated lactate levels, which suggest lymphoma (82). MRI of the lumbosacral spine is rarely useful; however, enhancement of the cauda equina in cytomegalovirus polyradiculopathy has been described (12; 217).

Skin biopsy of the calf and thigh may reveal a reduced intraepidermal nerve fiber density in patients with polyneuropathy (172) and has been used to follow the response to potentially therapeutic agents such as nerve growth factor (147; 195). Epidermal nerve fiber density assessment may have a role as a predictor of the development of symptomatic HIV neuropathy (94; 95) and may correlate with the clinical and electrophysiologic severity of the neuropathy (237). One study has shown that HIV+ individuals (including those without neuropathy symptoms) have an impaired capacity to regenerate these small nerve fibers (90).

Sural nerve biopsy is typically performed if the diagnosis remains unknown or if vasculitis, cytomegalovirus or lymphoma is suspected. The array of pathologic findings includes axonal loss, segmental demyelination and perivascular mononuclear cell infiltrates with complement or immunoglobulin deposition. The inflammatory cells usually are macrophages or T-lymphocytes, predominantly of the CD8+ cytotoxic or suppressor cell type.

Management

Supportive measures are discussed in Peripheral neuropathies: supportive measures and rehabilitation.

The primary management goal is control of the underlying process with HAART.

Nerve growth factor, an important neurotrophin that modulates the activity of small sensory nerve fibers, was investigated as a potential target for intervention in distal sensory polyneuropathy. Nerve growth factor was safe and well-tolerated and significantly improved pain symptoms; however, there was no improvement of neuropathy severity as assessed by neurologic examination, quantitative sensory testing, and epidermal nerve fiber density (195).

The management of demyelinating neuropathies, including AIDP and CIDP, is the same as for non-HIV-1 infected patients. Both conditions respond to plasmapheresis (43; 118) or IVIg (40; 142). There is one case report of a patient who developed acute inflammatory demyelinating polyneuropathies during the chronic phase of HIV infection, who had dramatic resolution of the neuropathy following HAART therapy alone (09). Mononeuropathies in HIV-infected patients, particularly recurrent laryngeal and bilateral phrenic neuropathies, have been reported to respond to high-dose IVIg (134; 170). Corticosteroids may be beneficial in patients with a painful distal sensory polyneuropathy due to vasculitis (19). It should be mandatory that all patients receiving isoniazid receive pyridoxine to prevent isoniazid-associated neuropathy (119).

Treatment of neuropathic pain symptoms remains a challenge (21), with many medications showing little or no superiority to placebo. Typical agents include anticonvulsants, antidepressants, topical agents, and opioids.

Anticonvulsants. Gabapentin has been reported to be helpful (160; 88); however, a randomized clinical trial did not demonstrate superiority to placebo (88; 168). Nonetheless, it is a desirable treatment given the decreased incidence of drug-to-drug interactions (182). A typical starting dose is 100 mg three times daily, with subsequent escalation up to as much as 3600 mg/day as tolerated.

Pregabalin is a GABA analogue that acts on the alpha-2-delta ligand of voltage-activated calcium channels. It is an FDA-approved treatment for diabetic neuropathy, postherpetic neuralgia, fibromyalgia, and partial seizures. A placebo-controlled trial, however, showed that pregabalin was well tolerated but no better than placebo for treatment of the painful symptoms of HIV neuropathy (209). A Cochrane review failed to demonstrate efficacy of pregabalin in HIV neuropathy (55).

Lamotrigine has been proven in placebo-controlled trials to be effective for HIV-associated painful sensory neuropathies (208; 207). Patients not receiving drugs known to induce the metabolism of lamotrigine started at a dose of 25 mg every other day for the first 2 weeks, with a target maintenance dose of 600 mg/day (300 mg twice daily). Patients who were receiving drugs known to induce the metabolism of lamotrigine started at a dose of 25 mg daily, with a target maintenance dose of 400 mg/day (200 mg twice daily).

Anticonvulsants, such as phenytoin and carbamazepine, which are highly protein-bound or induce cytochrome P450, should be avoided as they might affect the efficacy of protease inhibitors (182). Valproate should also be used cautiously as it has been shown in vitro to stimulate HIV (152) and cytomegalovirus replication (122).

Antidepressants. Tricyclic antidepressants and selective serotonin-norepinephrine reuptake inhibitors (SNRIs) are the two major classes used in the treatment of neuropathic pain. Controlled studies of amitriptyline, mexiletine, memantine, and acupuncture have been negative (111; 115; 201; 196). Duloxetine is an SNRI that is an FDA-approved treatment for diabetic neuropathy. A typical starting dose is 30 mg daily, with slow titration up to a maximum of 120 mg daily. It is also being studied for HIV-related neuropathy. In one randomized trial, amitriptyline was found ineffective in painful HIV-associated sensory neuropathy (58).

Topical treatments. Capsaicin, the active component of chili peppers, is known to bind to the excitatory vanilloid receptor (TRPVR1), which is expressed on small-diameter afferent neurons specialized for the detection of noxious sensations. By activating these receptors, the capsaicin initially produces a burning sensation, and at higher concentrations, desensitizes these nociceptors. Low concentrations are not effective in painful HIV neuropathy; however, a single application of a high-concentration (8%) dermal patch (NGX-4010) was found to provide some benefit (203; 107). This treatment is currently approved by the United States FDA for postherpetic neuralgia only. Topical lidocaine, in gel or patch form, has not been found to be effective (69). High-concentration topical capsaicin can be considered when other forms of drug treatment have failed (56).

Other treatments. Oral acetyl-L-carnitine (1500 mg twice daily) has been demonstrated to result in a reduction in neuropathic pain and a significant increase in cutaneous innervation in one open cohort of 21 HIV-positive patients followed for up to 33 months (93). In another study, pain improvement was noted with 3000 mg of acetyl-L-carnitine daily; however, no change in epidermal fiber density was noted after 24 weeks of therapy (224). In a double-blind, placebo-controlled trial involving patients with ATN, acetyl-L-carnitine (500 mg twice a day intramuscularly for 14 days, followed by 1000 mg twice a day orally for 42 days) was found to result in improved pain ratings versus placebo (235).

Smoked cannabis in a monitored setting was shown to have efficacy in relieving pain in a controlled trial (01). In a randomized, crossover trial, 46% of patients experienced at least 30% pain relief, compared to18% for placebo (67). In this study, cannabis of potency between 1% and 8% tetrahydrocannabinol was smoked four times daily for 5 consecutive days during each of two treatment weeks, separated by a 2-week washout period. Medical marijuana is a potentially effective therapy for HIV-associated sensory neuropathy, and its controlled use may be recommended in future (128).

Randomized controlled trials of subcutaneous recombinant human nerve growth factor (rhNGF) showed that it was superior to placebo for pain relief (147; 168). Studies of newer agents, such as subcutaneous prosaptide and intranasal peptide T, were negative (168).

Opioids have not been formally studied in painful HIV-related sensory neuropathy (168); however, they may be useful for moderate to severe neuropathic pain. Careful patient selection and screening for substance abuse is imperative.

Hypnosis has shown efficacy in clinical trials but due to limited access is not typically used in the clinical setting. Acupuncture has not been shown to be effective (180).

Management of autonomic symptoms focuses on removing exacerbating factors, such as drugs, which might cause orthostatic hypotension or anticholinergic side effects. If salt supplementation and compressive stockings are ineffective for treating orthostatic hypotension, fludrocortisone may be tried. Promotility drugs such as cisapride may be used to treat gastroparesis. Sildenafil can help those with erectile dysfunction (24).

The immunophilin ligand, FK506, was found to have a neuroprotective effect in an in vitro model of toxic neuropathy due to ddC. Some studies have indicated that neuroprotection is via inhibition of calcineurin, a calcium-dependent protein phosphatase (112). The hematopoietic growth factor, erythropoietin, has also been shown to prevent sensory axonal degeneration and in vitro dorsal root ganglion neuronal death by both the HIV envelope protein gp120 and zalcitabine (ddC) (113).

Outcomes

With the advent of HAART, the incidence of some complications (eg, HIV-related lumbosacral radiculopathy, cytomegalovirus polyradiculopathy, cytomegalovirus-induced mononeuritis multiplex, medication-related neuropathy [antiretroviral drugs], and autonomic neuropathy) has diminished, whereas there has been an increase in the incidence of peripheral neuropathy, both distal sensory polyneuropathy and antiretroviral toxic neuropathy.

A multinational study demonstrated that there was substantial reduction in the prevalence of distal sensory peripheral neuropathy following combination antiretroviral treatment in treatment-naive people living with HIV. Before initiating combination antiretroviral treatment, 21.3% of people living with HIV had distal sensory peripheral neuropathy compared with 8.5% of people not living with HIV. This study also noted that combination antiretroviral treatment-naive patients with HIV with distal sensory peripheral neuropathy were more likely to experience depression and loss of productivity. Overall prevalence of distal sensory peripheral neuropathy among those virally suppressed on combination antiretroviral treatment significantly decreased (225).