Safety of long-term treatment of HAM/TSP patients with valproic acid

Among 15 to 20 million human T-lymphotropic virus type 1 (HTLV-1)–infected persons worldwide, approximately 5% will develop 1 of the 2 major HTLV-1–associated diseases, HTLV-associated myelopathy/tropical spastic paraparesis (HAM/TSP) and adult T-cell leukemia (ATL).¹  HAM/TSP is a neuroinflammatory disease of the central nervous system associated with perivascular and parenchymal infiltration of HTLV-1–infected T cells and activated cytotoxic T lymphocytes (CTLs). Collateral damage results from release of proinflammatory cytokines, such as IFN-γ and TNF-α, by invading CD4⁺ and CD8⁺ infected cells and HTLV-1–specific T-lymphocytes. The immune response controlling infection may therefore also become detrimental and ultimately participate in mediating HAM/TSP progression.²  A key determinant of the development of HAM/TSP is the level of HTLV-1-infected cell burden, the proviral load (PVL) in PBMCs being 5- to 20-fold higher in patients with HAM/TSP than in asymptomatic carriers.³  A PVL > 10⁵ HTLV copies per million PBMCs constitutes a predictor of rapid progression from disease onset to Disability Status Scale (DSS) 8 (wheelchair confinement).⁴ 

In the absence of satisfactory therapy and based on preclinical trials performed in an animal model,⁵  we proposed a strategy designed to activate viral gene expression to expose virus-positive cells to the host immune response. We used valproic acid (VPA), a lysine deacetylase inhibitor that modulates chromatin condensation to alter the pattern of gene expression. We demonstrated that, in ex vivo cultures of HAM/TSP PBMCs, VPA induces hyperacetylation, activates expression of the viral core protein p19, and triggers apoptosis of CD4⁺ and CD8⁺ T lymphocytes.⁶  Real-time quantification performed in the Virology-Immunology reference laboratory of Fort-de-France indicated that the administration of VPA to HAM/TSP patients led to a decrease in their PVL a few weeks after initiation of treatment.⁷ 

We conducted a single-center, 2-year, open-label trial, with 19 HAM/TSP volunteers treated with oral doses of VPA. Objectives were to characterize ongoing mechanisms, assess clinical safety, and evaluate biologic response to VPA treatment. The following questions were addressed:

1. Is the CTL efficiency preserved and not exacerbated?

2. How does the PVL evolve in the long term?

3. What is the condition of treated patients?

VPA given orally (20 mg/kg per day) was planned to be given for 24 months when the trial started. Clinical assessment was performed every 3 months by the same observer. DSS was used as a measure of global neurologic disability, 30-meter Walking Time Test (WTT) quantified gait performance, manual muscle score, and modified Ashworth score were applied to 9 muscle groups in both legs, and urinary disability was assessed by a short questionnaire. Overall analysis of neurologic disability test was performed based on observed cases and also using the last observation carried forward method. Worsening of neurologic disability was defined as: significant mean scores or time differences between M0 and M24 or M0 and the last assessment visit before treatment discontinuation and individually by the increase DSS score of 1 or 0.5 point regarding baseline DSS (< 5.5 or ≥ 5.5, respectively) or a WTT variation rate > 20%. The study was approved by the Local and Regional Research Ethics Committee, and all procedures were carried out in accordance with the Declaration of Helsinki.

The efficiency of CD8⁺ cell–mediated lysis before and after VPA treatment was performed as described by Mosley et al.⁸  The proportions of CD4⁺ or CD8⁺ cells expressing CD38 or HLA-DR were determined by flow cytometry using the MultiTEST assay from BD Biosciences.

A major concern with VPA is that it may impair the anti–viral immune response. Indeed, the rate of CD8⁺ cell-mediated lysis of Tax-expressing cells ex vivo is halved in short-term cultures complemented with a 5mM dose of VPA.⁸  Although this concentration is indeed above the levels that can be achieved in patients (estimated ∼ 1-2mM), the risk of long-term treatment with lower doses cannot be predicted. Because the CTL response is an important factor in the immune control of HTLV-1 infection,²  we addressed this question directly by measuring the efficiency of CD8⁺ cell-mediated lysis. The cytotoxic response was evaluated by an assay in which the number of Tax⁺ cells surviving autologous CD8⁺ killing is measured by FACS after overnight culture.⁹  The percentages of surviving Tax⁺ CD4⁺ cells were then plotted against the proportion of CD8⁺ cells present (duplicate assays are shown in Figure 1A ■ and □). Lysis efficiency, which was estimated by fitting a mathematical model to these data,⁹  was calculated and expressed as the proportion of Tax-expressing CD4⁺ cells killed per CD8⁺ cell per day (Figure 1B). To assess whether these variations were within the normal range observed in untreated patients, the initial rate of lysis efficiency was plotted against the median absolute change in lysis efficiency between consecutive time points. Lysis efficiency variations observed during VPA treatment (□) were not significantly different from those in untreated patients (Figure 1C ♦, P = .072, Wilcoxon-Mann-Whitney 2-tailed test). We conclude that CD8⁺ cell-mediated lysis efficiency fluctuates throughout treatment but remains within the normal range, indicating that the subject's CTL immune response against HTLV-1 is not significantly suppressed by VPA. Likewise, the CTL response is not significantly exacerbated and must not be a concern regarding HAM/TSP progression.

To support this conclusion, we next assessed T-cell activation as defined by expression of CD38 and HLA-DR. PBMCs from 5 patients before and 3 months after initiation of VPA treatment were analyzed by FACS (illustrated for patient 4 on the plots of Figure 1D and recapitulated for 5 patients on Figure 1E). It appeared that there is no significant difference in the proportions of CD4⁺ or CD8⁺ cells expressing CD38 and HLA-DR.

The initial objective of VPA treatment in HAM/TSP patients was to permanently reduce the PVL with the aim of attenuating collateral damages to the central nervous system. However, long-term administration of VPA did not achieve this goal. The PVL determined by real-time PCR¹⁰  after 12 and 24 months or the last quantification before treatment discontinuation was similar to before treatment (Figure 2A). Based on last observation carried forward analysis, no significant differences were found in mean neurologic scores and WTT (Figure 2B). No patient's DSS score increased significantly over the study. Eight of 19 patients stopped VPA treatment before M24 (Figure 2C). Three patients were withdrawn because of a significant increase in the rate of variation of WTT. WTT improved rapidly after treatment discontinuation. Gait impairment worsening was probably related to severe drowsiness and tremor side effects experienced by 3 patients. The main clinical side effects of VPA were drowsiness (52%), tremor (47%), digestive symptoms (37%), vertigo (26%), and alopecia (10%), and their frequencies tended to decrease over the trial course. No significant biologic side effect was documented. Walking deterioration in a few patients may be associated with VPA side effects and is quickly reversible after treatment discontinuation.

In conclusion, we have shown that CTL response is preserved on VPA treatment of HAM/TSP patients. Although clinical symptoms are not improved, our data clearly indicate that long-term treatment with VPA is safe. This report is important in view of the recent evidence in STLV-1 infected baboons in which combined treatment with VPA and azidothymidine efficiently decreases PVL.¹¹  If effective in human, long-term administration of VPA and azidothymidine may reduce the risk of HAM/TSP. Safety is also crucial for maintenance therapy of acute ATL where molecular clearance was observed with VPA combined with azidothymidine plus IFN-α.¹² 

This work was supported by the Fonds National de la Recherche Scientifique, the Télévie, the Belgian Foundation against Cancer, the Sixth Research Framework Program of the European Union (project INCA LSHC-CT-2005-018704), the Neoangio excellence program and the Partenariat Public Prive PPP INCA of the Direction générale des Technologies, de la Recherche et de l'Energie/DG06 of the Walloon government, the Action de Recherche Concertée Glyvir of the Communauté française de Belgique, the Fonds spéciaux pour la Recherche of the University of Liège, the Synbiofor program of GxABT (University of Liège), the Centre anticancéreux près University of Liège, and the Plan Cancer of the Service Public Fédéral. M.B., N.G., and S.R. (Postdoctoral Researchers) and L.W. (Research Director) are members of the FRS-FNRS.

Contribution: S.O., A.S., and D.S. performed clinical evaluation of VPA treatment; G.B., O.V., and A.L. measured proviral loads; N.G. and C.B. performed CTL lysis assay; M.B. performed the FACS analysis of CD38 HLA-DR; B.A. modeled data; S.O., A.L., S.R., R.C., and L.W. designed, coordinated, and supervised studies; and S.O. and L.W. wrote the paper.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Luc Willems, Molecular and Cellular Epigenetics, Interdisciplinary Cluster for Applied Genoproteomics of University of Liège, avenue de l'Hôpital B34 Sart-Tilman, 4000 Liège, Belgium; e-mail: luc.willems@ulg.ac.be.