Neurology -- Correspondence for Schifitto et al., 58 (12) 1764-1768

Correspondence published:

Reply to Letter to the Editor: Giovanni Schifitto, Michael P. McDermott and Ned Sacktor (26 February 2003)

Incidence of and risk factors for HIV-associated distal sensory polyneuropathy: Sanjay C. Keswani, Ahmet Hoke (26 February 2003)

Reply to Letter to the Editor 26 February 2003

Giovanni Schifitto
University of Rochester New York,
Michael P. McDermott and Ned Sacktor
Send Correspondence to journal:
Re: Reply to Letter to the Editor

gschifitto{at}mct.rochester.edu Giovanni Schifitto, et al.

We thank Drs. Keswani and Hoke for their letter. We agree that our cohort study was not designed to definitively determine whether or not the use of dideoxynucleosides (ddN) is associated with the development of symptomatic distal sensory polyneuropathy (SDSP). Drs. Keswani and Hoke note the limitations of using baseline ddN use as a predictor and the inclusion of patients in the ddN group who had discontinued taking ddN within six months prior to study entry. Only 6 of the 74 patients classified as taking ddN discontinued these medications > 2 months before study entry, however, this is not really a significant limitation. One factor, not raised by Drs. Keswani and Hoke, that may be the most likely explanation for the finding of no increased risk of developing SDSP among ddN-exposed patients (if indeed there truly is an increased risk) is selection bias. Patients who initiated ddN therapy and developed a toxic neuropathy soon thereafter may have had this therapy withdrawn prior to entry into our study. Conversely, patients who tolerated ddN therapy upon initiation and subsequently entered our study may have thus been at relatively low risk of developing ddN-associated toxic neuropathy. Other factors such as drug adherence, which we did not measure, may have also played roles as confounders.
Certainly a randomized trial of ddN therapy would be best to assess its potential for causing toxic neuropathy, but such a trial will likely never be performed (particularly with neuropathy as a primary outcome variable). Analysis of data was done from ACTG 175/801 in a three-year trial of 2467 patients randomized to receive either zidovudine (ZDV) + didanosine (ddI), ZDV + zalcitabine (ddC), ZDV + placebo, or ddI + placebo. The incidence of neuropathy did not differ among the ZDV + placebo, ZDV + ddI, and ddI + placebo groups. Note that ZDV treatment is not known to be associated with neuropathy. There was a slightly higher incidence of neuropathy in the ZDV + ddC group (6%) than in the ZDV + placebo group (4%), however ddC is used very infrequently (no patients in our cohort were using it). A limitation of this trial was that the diagnosis of neuropathy was made by non-neurologists. Combination ddN therapy may be more likely to be associated with DSP, especially when hydroxyurea is part of the regimen [3, 4]. Non-neurologists also diagnosed DSP in these studies [3, 4]. Also in another study the data were abstracted from the Johns Hopkins AIDS database rather from a clinical trial or a formal cohort study. In our study, no patients were taking ddC and only four patients were taking ddI + stavudine (d4T), none of whom were taking hydroxyurea.
The second concern raised by Drs. Keswani and Hoke is related to our assessment of DSP. The diagnosis was based on a standardized clinical evaluation performed by neurologists. Additional assessments of small fiber function such as quantitative sensory testing (QST) or epidermal nerve fiber (ENF) density were not included. Drs. Keswani and Hoke suggest that by using QST we would have classified more "No DSP" subjects as having asymptomatic DSP (ADSP). This may or may not be the case, and this reclassification may or may not have changed our findings concerning ADSP as a risk factor for developing SDSP. The question of how much information QST adds to a thorough neurological examination in the diagnosis of ADSP is unresolved. In our ongoing North East AIDS Dementia (NEAD) cohort study, a subgroup of patients (n = 57) has undergone QST and skin biopsy to assess ENF density. None of the QST indices was as strongly associated with ENF density as sensory loss assessed by neurological examination.
We agree with the conclusions of Drs. Keswani and Hoke that a contribution by ddN drugs to the development of SDSP cannot be ruled out, and that our study has not disproved the hypothesis that ADSP is a predictor of SDSP. We never made such claims in our paper. We welcome future studies of the role of potential risk factors such as ddN exposure and ADSP in predicting the development of SDSP.
References
1. Simpson DM, Katzenstein DA, Hughes MD, Hammer SM, Williamson DL, Jiang Q, Pi J-T, AIDS Clinical Trials Group 175/801 Study Team. Neuromuscular function in HIV infection: Analysis of a placebo-controlled combination antiretroviral trial. AIDS 1998;2425-2432.
2. Rutschmann OT, Vernazza PL, Bucher HC, Opravil M, Ledergerber B, Telenti A, Malinverni R, Bernasconi E, Fagard C, Leduc D, Perrin L, Hirschel B, Swiss HIV Cohort Study. Long-term hydrozyurea in combination with didanosine and stavudine for the treatment of HIV-1 infection. AIDS 2000;2145-2151.
3. Moore RD, Wong W-ME, Keruly JC, McArthur JC. Incidence of neuropathy in HIV-infected patients on monotherapy versus those on combination therapy with Didanosine, Stavudine and Hydroxyurea. AIDS 2000;273-278.

Incidence of and risk factors for HIV-associated distal sensory polyneuropathy 26 February 2003

Sanjay C. Keswani
Johns Hopkins Hospital Baltimore MD,
Ahmet Hoke
Send Correspondence to journal:
Re: Incidence of and risk factors for HIV-associated distal sensory polyneuropathy

skeswani{at}jhmi.edu Sanjay C. Keswani, et al.

We read with interest the article by Schifitto et al, [1] that noted a 36% 1-year incidence rate of symptomatic distal sensory polyneuropathy (SDSP) in a pre-highly active antiretroviral therapy cohort. Conclusions reached by the authors were that dideoxynucleoside (ddX) drug usage and the presence of asymptomatic DSP (ADSP) were not significant risk factors for SDSP. However, we are concerned that the study was not adequately designed to accurately assess these putative risk factors, for the following reasons:
1. It is known from previous studies that the timing of onset of toxic neuropathy after ddX initiation is short, ranging from 1 week to 6 months depending on the ddX agent and the dose administered. [2] Thus I found it surprising that in order to evaluate whether ddX use was a risk factor for SDSP development; the authors studied the association between dideoxynucleoside use at baseline prior to entering the study and the time to development of SDSP within a subsequent 30-month period of follow-up. Even more inexplicable was the fact that patients who had discontinued taking dideoxynucleosides up to 6 months prior to cohort entry were still included in the ddX group. It would have been far more appropriate to investigate whether patients who actually took ddX drugs over the period of follow-up developed SDSP. Perhaps this data was not available to the authors, as the adherence of patients to ddX therapy was not assessed.
2. No objective measures of peripheral nerve function, such as Quantitative Sensory Testing (QST), were used in the study to help distinguish between 'no DSP' and ADSP. This classification was made on the basis of patient histories and examinations performed by a neurologist at each of three sites. It is known that subclinical nerve damage, as detected by QST, is present in many people with HIV who have no signs or symptoms of DSP. [3] Thus, we consider it highly likely that the proportion of individuals classified as having ADSP was underestimated, and conversely the number of 'no DSP' subjects was overestimated. This would have resulted in ADSP being under appreciated as a risk factor for SDSP.
We therefore think that it is premature to rule out a contribution by ddX drugs to the development of SDSP, particularly as much evidence exists of peripheral neurotoxicity by these agents. [4] Furthermore, the hypothesis that ADSP is a predictor of SDSP has not been disproved by this study.
References
1.Schifitto G, McDermott MP, McArthur JC, et al. Incidence of and risk factors for HIV-associated distal sensory polyneuropathy. Neurology 2002;58:1764-1768.
2.Simpson DM, Tagliati M. Nucleoside analogue-associated peripheral neuropathy in human immunodeficiency virus infection. J Acquir Immune Defic Syndr Hum Retrovirol 1995;9:153-161.
3.Gulevich SJ, Kalmijn JA, Thal LJ, et al. Sensory testing in human immunodeficiency virus type 1-infected men. HIV Neurobehavioral Research Center Group. Arch Neurol 1992;49:1281-1284.
4.Keswani SC, Pardo CA, Cherry CL, Hoke A, McArthur JC. HIV- associated sensory neuropathies. Aids 2002;16:2105-2117.