DOI: /hiv British HIV Association HIV Medicine (2013), 14, SHORT COMMUNICATION

Transcription

1 DOI: /hiv SHORT COMMUNICATION Evaluation of progressive multifocal leukoencephalopathy treatments in a Spanish cohort of HIV-infected patients: do protease inhibitors improve survival regardless of central nervous system penetration-effectiveness (CPE) score? F Fanjul, 1 M Riveiro-Barciela, 2 J Gonzalez, 3 E Delgado, 4 J Murillas, 5 A Payeras Cifré, 6 V Falcó 7 and M Riera 8 1 Department of Infectious Diseases, Hospital Universitario Son Espases, Carretera de Valldemosa SN, Palma de Mallorca, Illes Balears, Spain, 2 Infectious Diseases Department, Hospital Universitari Vall d Hebron, Universitat Autónoma de Barcelona, Spain, 3 Hospital Son Llatzer, Palma de Mallorca, Spain, 4 Hospital Universitario Son Espases, Palma de Mallorca, Spain, 5 Hospital Universitario Son Espases, Palma de Mallorca, Spain, 6 Internal Medicine, Hospital Son Llatzer, Palma de Mallorca, Baleares, Spain, 7 Infectious Diseases Department, Hospital Vall d Hebron, Barcelona, Spain and 8 Infectious Diseases, Hospital Universitario Son Espases, Palma de Mallorca, Spain Objectives The aim of the study was to investigate whether survival after progressive multifocal leukoencephalopathy (PML) diagnosis in HIV-1-infected patients was associated with central nervous system penetration-effectiveness (CPE) score and the presence or absence of protease inhibitors in the treatment regimen. Methods In the absence of treatments demonstrated to be effective for PML in HIV-1-infected patients and in the light of the controversy surrounding the use of CPE scores to make decisions on treatment after diagnosis, we determined whether there were differences in survival at 1 year depending on the type and characteristics of treatment. A multicentre retrospective observational study including three Spanish hospitals was carried out for the period from 1 January 1994 to 31 December Patients with a PML diagnosis were included in the study if they were symptomatic and met at least two of the following three criteria: (1) compatible radiological findings; (2) a positive polymerase chain reaction for John Cunningham virus (JCV) in the cerebrospinal fluid (CSF); (3) an absence of findings suggesting another infection in the central nervous system, after general CSF cultures for virus, bacteria and mycobacteria. Results A total of 98 patients were included in the study; 24.5% were diagnosed in the period , 39.8% in and 35.7% in The median follow-up time was 363 days (interquartile range days). The median CD4 count was 76 cells/ul (interquartile range cells/ul) and 62% of patients had an HIV viral load >50 HIV-1 RNA copies/ml. Thirty-eight per cent of patients received high-penetrance treatment, and 58% received treatment that included protease inhibitors. In the analysis of survival at 1 year, a higher CPE score did not result in an improvement in survival, but the presence of protease inhibitors in the regimen was associated with a statistically significant (P = 0.03) reduction in mortality (hazard ratio 0.40; 95% confidence interval ). Conclusions We consider that the lower mortality observed in the protease inhibitor group may be clinically relevant, and, if this is the case, a treatment based on protease inhibitors may be indicated for patients diagnosed with PML. Correspondence: Dr Francisco Fanjul, Hospital Universitario Son Espases, Department of Infectious Diseases, Carretera de Valldemosa SN, Palma de Mallorca, Illes Balears, Spain. Tel: (0034) ; fax: (0034) ; ffanjul@gmail.com 321

2 322 F Fanjul et al. Keywords: progressive multifocal leukoencephalopathy, HIV, protease inhibitors, survival, central nervous system penetration-effectiveness, CPE Accepted 29 October 2012 Introduction Progressive multifocal leukoencephalopathy (PML) is a demyelinating disease that occurs in immunocompromised patients. It is caused by infection with John Cunningham virus (JCV), a polyomavirus that destroys oligodendrocytes, and is associated with one of the highest morbimortality rates for an opportunistic central nervous system (CNS) infection in HIV-infected patients. During the last two decades, with the development of highly active antiretroviral therapy (HAART), the incidence of PML has decreased and the survival rate of HIV-infected patients who develop PML has significantly improved. Nevertheless, in European cohorts, PML still occurs with an annual incidence of 0.07% [1]. Furthermore, the increased use of biological treatments for a wide variety of diseases, such as psoriatic arthritis, inflammatory bowel diseases and systemic lupus erythematosus, has been associated with the development of PML in non-hiv-infected patients [2]. To date, several treatments for PML have been assessed for effectiveness (topotecan [3], cytarabine [4], cidofovir [5], interferon-a [6] and interferon-b [7]), but favourable outcomes have not been reported. In addition, knowledge obtained in vitro concerning viral action mechanisms, in particular tropism for serotonin receptors for entry into the host cell, has been used in the search for better strategies, and treatments with selective serotonin re-uptake inhibitors have been developed, but have not been found to be successful [8]. In recent years, it has been suggested that types of HAART with a high penetrance in the CNS, as observed in other CNS diseases in HIV-infected patients [9,10], could be effective in treating PML by decreasing the HIV viral load at the site of action of JCV, and thus affect JCV replication through effects on the HIV-1 tat protein. To date there have been few studies performed to test this theory, and there have been contradictory reports regarding the importance of antiretroviral penetrance in the CNS for patients with PML [11]. However, use of HAART regimens with high penetrance is the current expert recommendation in the absence of further evidence [12]. In a recent publication by Gasnault et al., an improvement in survival among patients receiving intensive HAART with five drugs, one of them enfuvirtide, has been described; if confirmed, this could result in a new line of treatment for PML [13]. After reviewing the current literature, we designed this study in order to determine whether, in our hospitals, there were any survival differences between HIV-infected patients depending on the penetrance of HAART in the CNS. As secondary objectives, we also considered the possibility of survival differences associated with the use of protease inhibitors (PIs) in the antiretroviral regimen, the total number of antiretroviral drugs used, and clinical manifestations at diagnosis. Methods We carried out a multicentre retrospective observational study including HIV-infected patients diagnosed with PML from 1 January 1994 to 31 December 2009 in three Spanish hospitals: the tertiary hospitals Hospital Sons Espases (Palma de Mallorca) and Hospital Vall d Hebrón (Barcelona) and the secondary hospital Hospital Son Llázter (Palma de Mallorca). We reviewed records for all HIV-infected patients over 15 years old who had been diagnosed with PML, and included in the study all those who showed compatible clinical manifestations (we excluded asymptomatic patients) and who met at least two of following three criteria: (1) compatible radiological findings in brain computed tomography and brain magnetic resonance imaging; (2) a positive polymerase chain reaction for JCV in a cerebrospinal fluid (CSF) sample; (3) an absence of findings suggesting another infection in the CNS, after general CSF cultures for virus, bacteria and mycobacteria. These inclusion criteria are based on those of previous studies [14]. In order to classify the patients with regard to their CNS penetration-effectiveness (CPE) score, we assigned to every antiretroviral drug a value according to the modified classification proposed by Letendre et al. [15,16]: 0 points for low-penetrance drugs, 0.5 points for intermediate-penetrance drugs and 1 point for highpenetrance drugs. We then added together the scores for each individual drug in the patient s regimen, and patients whose total treatment scores were 2 were included in the high-penetrance group and those with total scores < 2 were included in the low-penetrance group. We only considered the initial treatment after PML diagnosis, not previous treatments. We grouped the most frequent clinical manifestations at diagnosis into five categories: motor deficits, language disorders, altered mental status, walking impairment and

3 Evaluation of PML treatments 323 uncoordinated movements. We also recorded less frequently presented symptoms, such as seizures and visual impairment, but we did not include them in the statistical analysis because of their low prevalence. We analysed possible associations between CPE score and the registered variables (CD4 count, period of diagnosis, HIV-1 viral load, use of protease inhibitors, previous AIDS-defining illness) using a c 2 test, and then we analysed survival by means of Kalpan-Meier curves, assessing differences between curves using a log-rank test. Values of P < 0.05 were considered statistically significant. The multivariate analysis was performed using Cox regression models, including as covariates the CD4 lymphocyte count at PML diagnosis, because of its clinical relevance and because it had been described as predictive for mortality in previous studies [17], variables that showed significant associations in the univariate analysis, and those that were significantly correlated with the CPE score category. We performed statistical analyses using SPSS v.17.0 (SPSS Inc., Chicago, IL). Results Ninety-eight patients were included in the study, with a median follow-up time of 363 days [interquartile range (IQR) days]. Of these patients, 32.7% had received antiretroviral treatment prior to PML diagnosis, and 98.4% had received it after diagnosis. Global mortality rates at 6 months and 1 year were 30.9 and 38.1%, respectively, with 8.1% of patients lost to follow-up during the first year. Patient and treatment characteristics are summarized in Table 1. Initial clinical manifestations at diagnosis were classified as follows: motor deficits, 62.2%; walking impairment, 49%; language disorders, 35.7%; altered mental status, 30.6%; uncoordinated movements, 23.5%. Visual impairment and seizures were present in < 10% of the patients. We found no differences in epidemiological or immunological variables between the high- and low-penetrance groups, with the exception of the period of diagnosis (patients received lower penetrance treatments during the first period, , compared with and ; P = 0.006) and the percentage of patients treated with PIs, this being higher in the high-penetrance group (72% vs. 42% in the low-penetrance group; P = 0.005). When we analysed the differences in survival in the univariate model, we found no statistically significant differences in mortality between patients receiving higher and lower penetrance treatments (1-year mortality 34.7% vs. 40.5%, respectively; log-rank P = 0.95). However, we did find differences between patients receiving PIs and those not receiving PIs, with global mortality rates of 27.8% vs. 32.4% at 6 months and 31.5% vs. 45.9% at 1 Table 1 Patients and treatments characteristics Univariate survival analysis, Kaplan-Meier model Multivariate survival analysis, Cox regression model Variable Value (n = 98) HR 95% CI P HR 95% CI P Sex (% male) (for female patients) Age (years) [mean (SD)] 39.1 ( 7.3) Period of PML diagnosis (%) CD4 count (cells/ml) [median (IQR)] 76 (30 166) Previous AIDS-defining illness (%) Plasma HIV viral load (%) Detectable 62.2 Undetectable 19.4 Unknown 18.4 Parenteral drug abuse (%) Past Not known 41.8 Active 5.1 CPE score 2 (%) Protease inhibitors (%) Number of ARV drugs 3 (%) Previous ART (%) ART, antiretroviral therapy; ARV, antiretroviral; CI, confidence interval; CPE, central nervous system penetration-effectiveness; HR, hazard ratio; IQR, interquartile range; PML, progressive multifocal leucoencephalopathy; SD, standard deviation.

4 324 F Fanjul et al. Fig. 1 Multivariate 1-year survival analysis. The dotted line represents 0.5 cumulative survival. year, respectively, but these differences were not statistically significant (1-year log-rank P = 0.21). Also, the more symptomatic forms of PML presentation, in terms of the number of symptoms recorded, were associated with higher mortality rates, with differences being statistically significant (P = 0.038) when three categories were compared (0-1, 2-3 and 4-5 symptoms). The hazard ratio (HR) for death in patients presenting more than two symptoms (of those we categorized) at diagnosis was 2.1 [95% confidence interval (CI) ]. In a survival Cox regression model, we found no differences for CPE category using as covariates the presence/absence of PIs in the treatment, the CD4 lymphocyte count (categorized as or < 100 cells/ul), the initial symptomatology (coded as above) and the period of diagnosis. We found statistically significant differences in survival (P = 0.03) between groups receiving and not receiving PIs, with lower mortality (HR 0.40; 95% CI ) in the PI group even after adjusting for symptomatology, CD4 lymphocyte count and CPE category as covariates (Fig. 1). Discussion We carried out this study in order to determine whether there were any differences in mortality rates in patients diagnosed with PML related to the penetrance of their treatment in the CNS. Our results did not show any such significant differences. We found only small differences in the univariate survival model that were not statistically significant. In terms of our secondary objectives, we found differences in survival between patients depending on their initial clinical symptoms and whether they received treatment with PIs, regardless of the global penetrance score in the CNS. Initially, we found a difference between the group receiving PIs and that not receiving PIs in the multivariate analysis, but not in the univariate analysis. When we compared the two groups in terms of symptoms, we found that there was a difference in the initial number of symptoms that each group presented, with those patients receiving PIs presenting more symptoms. This difference was statistically significant (P = 0.038) in the c 2 test and we believe that this may explain our results. These findings are also consistent with those of previous studies [18]. The design of the study has some weaknesses. We considered only the first antiretroviral treatment after diagnosis, and not subsequent regimen switches. We used only survival as the endpoint, with no follow-up of clinical symptoms or neurological long-term consequences. Another weakness is that we did not consider the effectiveness of the initial antiretroviral treatment in terms of improvements in CD4 lymphocyte count or reductions in HIV viral load. We tried to minimize these possible biases by only considering survival differences up to 1 year, as the initial choice of treatment would have had the greatest influence in this period. The study also has some strengths. It is one of the largest studies of PML to be carried out in Spain, representing nearly 275 patient-years and with only 8.1% of patients lost to follow-up in the first year. Also, it is one of the first studies to investigate initial clinical presentation in relation to the survival of patients. Regarding our results for treatment with PIs, we believe that these findings could be related to the specific antiapoptotic effect of PIs [19]. This conclusion is supported by recent virology publications that have related the protapoptotic effect of the agnoprotein produced by JCV to sensitization of oligodendrocyte cells to oxidative stress, which is responsible at least partly for cell destruction and demyelinization [20]. Conclusions We believe that the improvement in survival in the PI group may be clinically relevant. If so, this suggests that treatment based on PIs may be indicated for patients diagnosed with PML, in the absence of better clinical evidence regarding effective treatments. Of course, further studies are required to confirm these findings.

5 Evaluation of PML treatments 325 We also propose that the statistically significant relationship between initial clinical manifestations and mortality should be further investigated in order to determine whether a clinical prognosis score should be developed. References 1 d Arminio Monforte A, Cinque P, Mocroft A et al. Changing incidence of central nervous system diseases in the EuroSIDA cohort. Ann Neurol 2004; 55: Carson KR, Evens AM, Richey EA et al. Progressive multifocal leukoencephalopathy after rituximab therapy in HIV-negative patients: a report of 57 cases from the Research on Adverse Drug Events and Reports project. Blood 2009; 113: Royal W 3rd, Dupont B, McGuire D et al. Topotecan in the treatment of acquired immunodefi ciency syndrome-related progressive multifocal leukoencephalopathy. J Neurovirol 2003; 9: Hall CD, Dafni U, Simpson D et al. Failure of cytarabine in progressive multifocal leukoencephalopathy associated with human immunodefi ciency virus infection. N Engl J Med 1998; 338: De Luca A, Ammassari A, Pezzotti P et al. Cidofovir in addition to antiretroviral treatment is not eff ective for AIDS-associated progressive multifocal leukoencephalopathy: a multicohort analysis. AIDS 2008; 22: Geschwind MD, Skolasky RI, Royal WS, McArthur JC. The relative contributions of HAART and alpha-interferon for therapy of progressive multifocal leukoencephalopathy in AIDS. J Neurovirol 2001; 7: Ath A, Venkataramana A, Reich DS, Cortese I, Major EO. Progression of progressive multifocal leukoencephalopathy despite treatment with beta-interferon. Neurology 2006; 66: Cettomai D, McArthur JC. Mirtazapine use in human immunodeficiency virus-infected patients with progressive multifocal leukoencephalopathy. Arch Neurol 2009; 66: Patel K, Ming X, Williams PL, Robertson KR, Oleske JM, Seage GR 3rd. Impact of HAART AND CNS-penetrating antiretroviral regimens on HIV encephalopathy among perinatally infected children and adolescents. AIDS 2009; 23: Marra CM, Zhao Y, Clifford DB et al. Impact of combination antiretroviral therapy on cerebrospinal fluid HIV RNA and neurocognitive performance. AIDS 2009; 23: Lanoy E, Guiguet M, Bentata M et al. Survival after neuroaids: association with antiretroviral CNS Penetration- Effectiveness score. Neurology 2011; 76: Gasnault J, Lanoy E, Bentata M, Guiguet M, Costagliola D. Intracerebral penetrating ART are more effi cient on survival of HIV+ patients with progressive multifocal leucoencephalopathy (ANRS CO4 FHDH). 15 th Conference on Retroviruses and Opportunistic Infections. Boston, MA, February 2008 [Abstract 385]. 13 Gasnault J, Costagliola D, Hendel-Chavez H et al. Improved survival of HIV-1-infected patients with progressive multifocal leukoencephalopathy receiving early 5-drug combination antiretroviral therapy. PLoS One 2011; 6: e Berger JR, Pall L, Lanska D, Whiteman M. Progressive multifocal leukoencephalopathy in patients with HIV infection. J Neurovirol 1998; 4: Letendre S, Marquie-Beck J, Capparelli E et al. Validation of the CNS penetration-eff ectiveness rank for quantifying antiretroviral penetration into the central nervous system. Arch Neurol 2008; 65: Letendre S, Ellis R, Deutsch R et al. Correlates of Time-to-Loss-of-Viral-Response in CSF and Plasma in the CHARTER Cohort. CROI 2010 Poster Marzocchetti A, Tompkins T, Clifford DB et al. Determinants of survival in progressive multifocal leukoencephalopathy. Neurology 2009; 73: Tassie JM, Gasnault J, Bentata M et al. Survival improvement of AIDS-related progressive multifocal leukoencephalopathy in the era of protease inhibitors. Clinical Epidemiology Group. French Hospital Database on HIV. AIDS 1999; 13: Vlahakis SR, Bennet S, Whitehead S, Badley AD. HIV protease inhibitors modulate apoptosis signaling in vitro and in vivo. Apoptosis 2007; 12: Merabova N, Kaminski R, Krynska B, Amini S, Khalili K, Darbinyan A. JCV agnoprotein-induced reduction in CXCL5/LIX secretion by oligodendrocytes is associated with activation of apoptotic signaling in neurons. J Cell Physiol 2012; 227: