Departments of aNeurology and bHIV Medicine, St Vincent's Hospital, Darlinghurst, NSW 2010, Australia; and cNational Centre in HIV Epidemiology and Clinical Research, Darlinghurst, NSW 2010, Australia.
Received: 30 August 2002; revised: 19 September 2002; accepted: 12 November 2002.
We prospectively assessed serum lactate and plasma HIV viral load in 20 patients with stavudine-related nucleoside neuropathy, 10 with HIV-related distal sensory polyneuropathy (DSPN), 20 receiving stavudine but without neuropathy and 23 not receiving stavudine and without neuropathy. Raised serum lactate levels discriminated between stavudine nucleoside neuropathy and DSPN with 90% sensitivity and 90% specificity (P = 0.001). DSPN occurred despite an undetectable viral load, making it a poor discriminator.
Peripheral neuropathy, termed ‘nucleoside neuropathy', may complicate the use of antiretroviral drugs such as stavudine, didanosine and zalcitabine . Clinically, it is a distal symmetrical sensory neuropathy often associated with pain, and essentially indistinguishable from the symmetrical distal sensory polyneuropathy (DSPN) caused by HIV. The only differentiating factors are the timing of nucleoside commencement and an improvement after nucleoside withdrawal, although the latter may be confounded initially by an exacerbation of the pain . Apart from pain relief, the only therapies for nucleoside neuropathy and DSPN are nucleoside withdrawal and possibly more intensive antiretroviral therapy . Nonetheless, patients are often left with persistent disabling pain. It is currently thought likely that nucleoside neuropathy is related to nucleoside-induced mitochondrial dysfunction , whereas DSPN is related to the plasma HIV viral load – at least the risk of developing nucleoside neuropathy and its severity [4,5].
Clinically, it is important to be able to distinguish between nucleoside neuropathy and DSPN. We therefore assessed whether nucleoside neuropathy and DSPN could be distinguished by elevated serum lactate concentrations (as a marker of mitochondrial function) in the former and raised plasma HIV viral load in the latter.
We prospectively studied all patients developing either nucleoside neuropathy or DSPN over a one year period in the HIV medicine outpatient department of a tertiary referral hospital. Only those with a history of symptoms of 4 weeks or less were included in the nucleoside neuropathy or DSPN groups to exclude those patients who might have developed irreversible nerve damage. All patients had venous lactate measured after at least 10 min of rest. Blood was collected without a tourniquet or fist clenching, and samples were transported immediately for analysis (Hitachi 917 Analyser using Roche lactate reagents; Roche, Branchburg NJ, USA). Lactate concentrations were repeated within 24 h and the lower value was taken for further analyses. The upper limit of the normal range in our laboratory is 2.2 mmol/l. The plasma HIV viral load was also measured (Roche lower limit 400 copies/ml).
The diagnosis of nucleoside neuropathy was made after the fulfilment of three major criteria: the development of a neuropathy consistent with nucleoside neuropathy (painful neuropathy characterized by distal symmetrical sensory loss with absent ankle reflexes developing after the recent commencement of stavudine, didanosine or zalcitabine); the exclusion of other likely causes (diabetes, vitamin B12 and red cell folate deficiency, thyroid dysfunction, hepatitis B and C, paraproteinemias, any other potentially neurotoxic drug); and an improvement in the symptoms after cessation of the nucleoside. Patients who were considered probably to have nucleoside neuropathy on the latter criteria were clinically assessed every 2–4 weeks for 8 weeks to determine the response to the cessation of the nucleoside. Only those patients who had symptom improvements were finally considered to have nucleoside neuropathy. At that point, the serum lactate concentration was assessed and repeated within 24 h as described previously.
The diagnosis of DSPN was made in patients after two major criteria were fulfilled: the development of a neuropathy consistent wi with DSPN (the same clinical features as nucleoside neuropathy, except that no patient was taking stavudine, didanosine or zalcitabine); and other likely causes had to have been excluded, as described above.
Patients without neuropathy who were otherwise well with CD4 cell counts of less than 200 cells/μl and who were taking highly active antiretroviral therapy were also examined and had lactate and HIV viral loads assessed. A CD4 cell count below 200 cells/μl was chosen as both nucleoside neuropathy and DSPN are more likely to occur in this range.
The mean and standard deviation for the plasma HIV viral load and CD4 cell count were calculated. The extent to which elevated serum lactate concentrations (> 2.2 mmol/l) and detectable HIV viral loads could be used to discriminate between stavudine-associated nucleoside neuropathy and DSPN was assessed using logistic regression.
The data are summarized in Table 1. No patient was taking zalcitabine. In the 20 nucleoside neuropathy patients it was considered to be related to stavudine, but five were also taking didanosine. In all the stavudine neuropathy patients a switch to another nucleoside, either zidovudine or abacavir, led to an improvement in the severity of the neuropathy and normalization of the serum lactate concentration. In the DSPN patients with an undetectable viral load, a review of their previous viral load results revealed that they had had suppressed viral loads for the preceding year on average, and that the neuropathy had developed in the context of the suppressed viral load. There was no significant difference in ages or CD4 cell counts between the groups.
An elevated serum lactate concentration was 90% sensitive and 90% specific in discriminating between stavudine nucleoside neuropathy and DSPN (odds ratio 81, P = 0.001). Although in univariate analyses a detectable viral load was significantly associated with DSPN (odds ratio 13.5, P = 0.008), adding a detectable viral load to elevated serum lactate did not improve the discrimination between stavudine nucleoside neuropathy and DSPN compared with elevated serum lactate alone (P = 0.187, sensitivity and specificity unchanged at 90%).
These results suggest that elevated serum lactate concentrations can be useful in the diagnosis of nucleoside neuropathy (at least that related to stavudine) and its distinction from DSPN, whereas the plasma HIV viral load is not, and indeed 40% of DSPN patients had a plasma HIV viral load below 400 copies/ml.
1.Brew BJ (editor). Medication-related and nutrition-related peripheral neuropathy.
In: HIV neurology
. New York: Oxford University Press; 2001. pp. 209–217.
2.Markus R, Brew BJ. HIV-1 peripheral neuropathy and combination antiretroviral therapy. Lancet
3.Keswani S, Hasan C, McArthur J, Griffin J, Hoke A. An in vitro model of antiretroviral toxic neuropathy in dorsal root ganglion sensory neurons.
In: 9th Conference on Retroviruses and Opportunistic Infections.
Seattle, 24–28 February 2002 [Abstract 70].
4.Childs EA, Lyles RH, Selnes OA, Chen B, Miller EN, Cohen BA, et al. Plasma viral load and CD4 lymphocytes predict HIV-associated dementia and sensory neuropathy. Neurology
5.Simpson DM, Haidich A-B, Schifitto G, Yiannoutsos CT, Geraci AP, McArthur JC, et al. Severity of HIV-associated neuropathy is associated with plasma HIV-1 RNA levels. AIDS