Update on peripheral neuropathy in HIV-infection

Paul Blanchard, HIV i-Base

A major plenary session at this year’s CROI was devoted to the important issue of distal symmetrical polyneuropathy (DSPN) in HIV-infection [1]. This formed part of an oral session devoted to neuropathogenic manifestations of HIV-1 infection, which together with additional poster sessions provided a useful forum for presenting advances in the field.

David Simpson, professor of neurology at Mount Sinai Medical Center, New York, gave an update on DSPN, which focused on issues of epidemiology, diagnosis and misdiagnosis, the toxicity of antiretrovirals to the peripheral nervous system and pain and the management of pain.

Since the widespread introduction of HAART the incidence of DSPN in the HIV-positive population appears to have increased – from 17.3 per 1000 person years in 1994 to 31.1 by the year 2000 (John Hopkins HIV Clinical Cohort).

New assessment tools

NARC, a sub-study in ALLRT (Adult Longitudinal Linked Randomised Trials) looked at the sensitivity and specificity of a brief peripheral neuropathy assessment (BPNS) carried out by a nurse coordinator and how these correlated with a total neuropathy screen (TNS) performed by neurologists [2]. Patient population used was a subset of subjects (n=301) enrolled in studies to assess the impact of HAART on neurological disease in HIV.

The rate of neuropathy diagnosed using the BPNS was 20% vs 32% using a TNS. This gives the BPNS a sensitivity of 46% and a specificity of 91%. The investigators conclude that this brief peripheral neuropathy assessment provides a quick, simple and low-cost screening tool with acceptable diagnostic efficiency. These characteristics should allow for easy incorporation into large scale clinical trials.

Why aren’t all of us diagnosing DSPN in a third of all our patients?

Many patients with DSPN can be asymptomatic at the time of presentation. Using an earlier form of BPNS 71% of patients with HIV-infection diagnosed with DSPN had no neuropathy symptoms [3]. In support of these figures a separate study (Manhattan HIV Brain Bank) found that 71% of cases meeting a neurologist evaluation for DSPN had not been diagnosed previously or mentioned in patients’ case notes [4]. Unsurprisingly, of the asymptomatic cases newly diagnosed, 86% were undiagnosed previously. This highlights the fact that many clinicians will not attempt to diagnose the disease when patients are not complaining of symptoms. Of more concern, however, is that 63% of symptomatic cases had not been diagnosed by clinicians prior to this study.

Risk factors for neuropathy

Over the years, attempts have been made to determine the risk factors associated with the development of DSPN. It is now well established that an increased risk of development occurs with increasing immunosuppression.

The DANA Consortium study also established risk factors for incident DSPN [5].

Patients entered the cohort with a CD4 cell count <200 cells/mL. Using neurological assessment it was found that by 12 months after entry 25%, and at two years over 50%, had evidence of symptomatic neuropathy.

Much of the data on incidence and risk factors comes from the pre-HAART era so an attempt has been made in the last couple of years to determine how this data may have been modified by HAART. It has been established that the higher the viral load at baseline prior to HAART, the higher the risk for the development of DSPN [6]. In a sub- study of the ACTG nerve growth factor trial the severity of neuropathy, both by intensity of pain and degree of abnormality of quantitative sensory testing (QST), also correlated with plasma viral load [7].

Does treatment of HIV make PN better?

Here there appear to be very few data so far. At least with QST, thermal threshold has been seen to improve with effective control of plasma viral load [8]. It has also been established that intraepidermal nerve fibre (IENF) loss at the distal leg is associated with increased neuropathic pain, lower CD4 counts, and higher plasma viral load in HIV-DSPN [9]. Early results presented at this meeting also suggest that reductions in IENF density may be reversed with effective HAART therapy and/or modification of didanosine and stavudine components of regimens [10].

A major study expected to improve on the available (mostly retrospective) data was introduced by Professor Simpson. ACTG 5117 is a DSPN study of 100 subjects with CD4 <300 cells/mm3 looking at the risk factors for the development and progression over 48 weeks using an intensive neurological evaluation:

  • ensory and motor signs and symptoms
  • QST (vibration, heat and cooling)
  • nerve conduction
  • intraepidermal skin biopsy (IENF)
  • lymphocyte MtDNA

Data are currently being analysed and should be available within the next few months.

Nucleoside analogue related DSPN

Clinicians frequently encounter DSPN related to the use of dideoxynucleoside agents (stavudine, didanosine, zalcitabine). This form of DSPN is clinically indistinguishable from HIV-related DSPN in any given patient and in many patients they may often overlap. Pain when therapy is stopped should resolve within 8 – 16 weeks, although the signs of neuropathy may remain much longer. Additionally there are likely to be additive or synergistic effects between the impact of both HIV and dideoxynucleosides on nervous tissue. As Professor Simpson stated: “The more bad things you do to the peripheral nerve the less happy it gets”.

Up to a 28-fold increase in the incidence of DSPN over and above HIV alone has been seen with a variety of nucleoside or nucleoside plus hydroxyurea (HU) regimens. These side effects have been a major factor in the reduction in prescribing of certain regimens (eg. ddI+d4T, ddI+HU).

Pathogenesis of distal polyneuropathy

The pathogenesis of DSPN in AIDS is currently unknown even though the pathology is relatively well characterised.

Justin McArthur and colleagues have shown “deposition” of TNF-alpha in peripheral nerves correlating with neuropathy [11]. There are also neurotoxic byproducts of HIV replication such as gp120 [12]. At this meeting Keswani and colleagues presented data demonstrating that application of gp120 to dorsal root ganglia (DRG) culture leads to a reduction in the length of neuritic outgrowth with the nociceptive DRG neuronal population being particularly vulnerable [13]. Similarly when various nucleoside analogues (ZDV, ddI, d4T, ddC) are introduced to neuritic cell cultures the potency with which each NRTI inhibits neuritic outgrowth parallels the potency of each drug in causing peripheral neuropathy in HIV positive patients, ie ddC > ddI >d4T [14]. An intriguing question that remains is why different nucleoside analogues have different organ system toxicities?

The whole issue of mitochondrial toxicity and its potential pathogenetic role in numerous organ system toxicities has been a matter of great discussion and debate. Similarly, there are questions about the effects of dideoxynucleosides on mitochondria as a potential mechanism for the pathogenesis of both DSPN and another neurological disease – the ascending neuromuscular weakness syndrome.

What do we know about drugs and mitochondrial function?

Serum lactate, a presumed marker of mitochondrial dysfunction, has been measured in a study by Bruce Brew and colleagues in various groups of patients with and without neuropathy who were either receiving or not receiving d4T [15]. A significant elevation of serum lactate was seen in those patients with DSPN receiving d4T compared to those receiving d4T without DSPN and those with purely HIV-related PN. Additionally, it has been shown that ddC induces a mitochondrial neuropathy with depletion of the nerve’s mtDNA. These findings are consistent with the ability of ddC to selectively inhibit the gamma-DNA polymerase in neuronal cell lines [16]. However, Professor Simpson suggested that caution must be exercised in ascribing mitochondrial toxicity just to the presence of mitochondrial abnormalities in patients with HIV-infection receiving dideoxynucleosides. Similar, although less severe, mitochondrial abnormalities were seen in neuronal culture with the addition of gp120 alone [17], raising the possibility that there may be synergistic effects between HIV and the dideoxynucleosides on mitochondria.

HIV-associated neuromuscular weakness syndrome

This newly described syndrome also points to the role of mitochondrial dysfunction in DSPN. Sixty-nine patients have been identified in a retrospective series (Simpson et al. In Press, AIDS 2004). The syndrome is characterised by a rapidly progressing motor weakness that superficially mimics Guillain-Barre syndrome. In many of these patients, this syndrome is associated with lactic acidosis. It is hypothesised that there is a potential combined mitochondrial/immunological pathogenesis occurring through the interplay between HIV and dideoxynucleosides.

Management of DSPN

One of the most challenging aspects for physicians is the consideration of what to do with patients experiencing DSPN who are also receiving neurotoxic antiretrovirals.

Unfortunately, there is no black and white answer and Simpson suggested that consideration be made of the following factors;

  • identify and attempt correction of metabolic or nutritional causes
  • optimise HIV virological control
  • assess risk of progressive neuropathy
  • assess the risk to virological control of modifying antiretrovirals
  • availability of active non-neurotoxic antiretrovirals for a given patient.

In some patients, a stop-and-switch strategy may be employed, while in others there may be no choice but to continue a regimen that includes dideoxynucleosides.

Pathophysiology of neuropathic pain

When one begins to address pain there is a complex set of mechanisms that we do not completely understand which may be going on within one particular patient. These pathophysiological mechanisms may have a huge bearing on the mechanism and efficacy of treatment. Among the processes contributing to the generation of pain the following are thought to be the most significant:

  • chemical excitation of nociceptors
  • recruitment of nerves outside of site of injury
  • excitotoxicity
  • sodium channels
  • ectopic discharge
  • deafferentation
  • central sensitisation (maintained by peripheral input)
  • sympathetic nervous system involvement
  • antidromic neurogenic inflammation
  • loss of neurotrophins (NGF)
  • neurotoxins (gp120, cytokines: TNF, IL-1, IL-6)

Pain interventions may be directed at one or many of these contributing processes. Similarly, interventions may be targeted at anatomical sites of pain generation and transmission (periphery to cortex) or some of the physiological pathways of pain perception (transduction, transmission, modulation, perception, interpretation, behaviour). Such guided therapy has become known as a mechanism or pathogenesis based approach [18].

Under-treatment of pain in AIDS and cancer

Pain is a huge and challenging problem in HIV-infection. Eighty-four percent of ambulatory patients with AIDS have been found to be receiving inadequate analgesia according to the WHO criteria compared to 42% of patients with cancer [19]. Why is this?

Barriers to the effective management of pain in HIV-infection have been explored by Breitbart [20, 21] and may be:

  • patient related
    • stoicism – a reluctance to report pain
    • fear of substance abuse and associated stigma
  • health care provider related
    • fear of addiction
    • fear of being conned (diversion of controlled drugs)
  • health care system related
    • restrictions and bureaucracy of prescribing controlled substances.

Neuropathy clinical trials

A wide range of clinical trials over many years has attempted to determine effective treatment for HIV related DSPN. These may be divided into those attempting to address the symptoms and those that attempt to modify the pathogenesis of DSPN, restoring nerve function and thereby reducing symptoms.

Simpson highlighted some of these in this meeting update.


Lamotrigine is an anticonvulsant that has generated interest as a potential treatment for DSPN. It has been explored most recently in a multicentre, double blind, placebo controlled trial of 11 weeks duration [22]. Patients with DSPN were stratified according to exposure to neurotoxic antiretrovirals versus no exposure. A significant benefit on pain of lamotrigine over placebo was seen in those patients receiving neurotoxic antiretrovirals. Interestingly, and somewhat paradoxically, no benefit was seen in patients not exposed to neurotoxic antiretrovirals.


Opioids can be an effective management strategy in neuropathic pain. Numerous controlled studies (in postherpetic neuralgia, phantom limb pain, diabetic neuropathy, but not yet in HIV) establish them as an important part of the treatment armamentarium.

Topical treatments for neuropathic pain

These have received increasing attention over the last couple of years with lidocaine and capsaicin being the prime candidates. A topical agent has the advantage of being delivered at the sensory level of the skin rather than as another systemic agent on top of the many systemic agents delivered by pill that the patient is already likely to be receiving.


This is the most active chemical constituent in chilli pepper (Capsicum sp.) – a volatile compound that is closely related to vanillin, a component of vanilla. It acts on a known receptor – the VR1 receptor and is mostly available as a low concentration cream (0.075% capsaicin), which is of questionable efficacy. However, high concentration patches are now being investigated in experimental studies.

McArthur, using skin biopsy, has shown a mechanism of action of capsaicin. Forty-eight hours after a single application of capsaicin, virtually complete depletion of IENF has taken place. Twenty-seven days later, there is regeneration and repletion of these fibres. This effect is thought to be mediated by capsaicin stimulating the depletion of substance P, neurokinin, somatostatin, and calcitonin from peripheral nerve fibres, particularly C-fibres [23].

It may be asked if this strategy is safe. You are taking away important fibres at the level of the skin, albeit temporarily. Are you removing important aspects of sensation? QST shows a change in a couple of degrees C for detection of thermal threshold – probably not clinically significant, but no change in mechanical sensitivity.

A study of the efficacy of high concentration capsaicin patch versus low concentration patch has been performed in post herpetic neuralgia (Backonja et al. AAN 2003). Significant pain reduction was achieved in 33% of high concentration patch recipients versus 4% of low concentration patch recipients over one month.

High concentration capsaicin patch in HIV-associated DSPN

An open label study performed in painful HIV-associated DSPN was presented by Simpson et al. at this meeting [24]. Twelve patients received a single application and were then followed for three months. Topical anaesthesia was used before the application of capsaicin and over 40% pain reduction was achieved lasting for three months after this single application. Eight of 12 patients were responders based on thresholds of greater than or equal to 30% pain reduction from baseline. Overall, good tolerability was observed although exposure to high concentration capsaicin did lead to a significant increase in pain, but of short duration. Local dermal changes were limited to a transient erythema. Simpson concluded that further research in controlled studies is now needed as such local treatments are attractive as an alternative or complement to centrally acting agents.

Standard concentration capsaicin patches normally contain 10–40 µg/cm2 whereas those used in this study contained 640 µg/cm2. The pain and discomfort induced by application of such a high concentration should not be underestimated and although 100% of subjects completed the one application it is unknown if any would subject themselves to the procedure a second time. Previous studies have been performed with high concentration capsaicin for patients with intractable pain [25]. Here, capsaicin concentrations of 5–10% were used after the administration of regional anaesthesia by epidural or peripheral nerve block and sedation with midazolam. After nerve blocks wore off, intravenous fentanyl was required for pain relief and oral morphine after discharge. Only four of the original 10 patients in this study went on to receive additional applications.

Future directions for the research of DSPN highlighted by Professor Simpson at the end of his update included diagnostic criteria, risk factors, antiretroviral toxicity and both symptomatic and pathogenesis based treatments.


  1. Simpson DM and McArthur JC. Peripheral neuropathy associated with HIV and ARV: update on diagnosis and management. 11th CROI 2004, Abstract 34.
  2. Ellis R, et al. NARC007: Clinical validation of the AACTG NeuroScreen. 11th CROI 2004, Abstract 493.
  3. Marra CM, Boutin P and Collier AC. Screening for distal sensory peripheral neuropathy in HIV-infected persons in research and clinical settings. Neurology, 1998. 51(6): p. 1678-81.
  4. Estanislao L. in 5th International Symposium on Neurovirology. 2003. Maryland, Baltimore.
  5. Schifitto G et al. Incidence of and risk factors for HIV-associated distal sensory polyneuropathy. Neurology, 2002. 58(12): p. 1764-8.
  6. Childs EA et al. Plasma viral load and CD4 lymphocytes predict HIV-associated dementia and sensory neuropathy. Neurology, 1999. 52(3): p. 607-13.
  7. Simpson DM et al. Severity of HIV-associated neuropathy is associated with plasma HIV-1 RNA levels. Aids, 2002. 16(3): p. 407-12.
  8. Martin C et al. Antiretroviral therapy may improve sensory function in HIV-infected patients: a pilot study. Neurology, 2000. 54(11): p. 2120-7.
  9. Polydefkis M et al. Reduced intraepidermal nerve fibre density in HIV-associated sensory neuropathy. Neurology, 2002. 58(1): p. 115-9.
  10. Cherry C et al. Response of HIV-associated sensory neuropathies to early detection and treatment manipulation. 11th CROI 2004, Abstract 497.
  11. Wesselingh SL et al. Cytokine dysregulation in HIV-associated neurological disease. Adv Neuroimmunol, 1994. 4(3): p. 199-206.
  12. Keswani SC et al. Schwann cell chemokine receptors mediate HIV-1 gp120 toxicity to sensory neurons. Ann Neurol, 2003. 54(3): p. 287-96.
  13. Keswani SC, Polley M and Hoke A. Nociceptive neruons are particularly vulnerable to gp120 neurotoxicity. 11th CROI 2004, Abstract 463.
  14. Keswani SC et al. An in vitro model of antiretroviral toxic neuropathy in dorsal root ganglion sensory neurons. Programme and abstracts of the 9th Conference on Retroviruses and Opportunistic Infections; 24-28 February, 2002; Seattle, Washington. Abstract 70.
  15. Brew BJ, Tisch S and Law M. Lactate concentrations distinguish between nucleoside neuropathy and HIV neuropathy. Aids, 2003. 17(7): p. 1094-6.
  16. Dalakas MC, Semino-Mora C and Leon-Monzon M. Mitochondrial alterations with mitochondrial DNA depletion in the nerves of AIDS patients with peripheral neuropathy induced by 2’3′-dideoxycytidine (ddC). Lab Invest, 2001. 81(11): p. 1537-44.
  17. Keswan SC et al. Mechanisms of gp120 peripheral neuro toxicity. Programme and abstracts of the 10th Conference on Retroviruses and Opportunistic Infections;10-14 February, 2003; Boston, Mass. Abstract 679.
  18. Dworkin RH et al. Advances in neuropathic pain: diagnosis, mechanisms, and treatment recommendations. Arch Neurol, 2003. 60(11): p. 1524-34.
  19. Breitbart W et al. The undertreatment of pain in ambulatory AIDS patients. Pain, 1996. 65(2-3): p. 243-9.
  20. Breitbart W, Kaim M and Rosenfeld B. Clinicians’ perceptions of barriers to pain management in AIDS. J Pain Symptom Manage, 1999. 18(3): p. 203-12.
  21. Breitbart W et al. Patient-related barriers to pain management in ambulatory AIDS patients. Pain, 1998. 76(1-2): p. 9-16.
  22. Simpson DM et al. Lamotrigine for HIV-associated painful sensory neuropathies: a placebo-controlled trial. Neurology, 2003. 60(9): p. 1508-14.
  23. Lynn B. Capsaicin: actions on nociceptive C-fibres and therapeutic potential. Pain, 1990. 41(1): p. 61-9.
  24. Simpson DM et al. Novel high-concentration capsaicin patch for the treatment of painful HIV-associated distal symmetrical polyneuropathy: results of an open label trial. 11th CROI 2004, Abstract 490.
  25. Robbins WR et al. Treatment of intractable pain with topical large-dose capsaicin: preliminary report. Anesth Analg, 1998. 86(3): p. 579-83.

Links to other websites are current at date of posting but not maintained.