Changing Incidence of Central Nervous System Diseases in the EuroSIDA Cohort Antonella d’Arminio Monforte, MD,1 Paola Cinque, MD,2 Amanda Mocroft, PhD,3 Frank-Detlev Goebel, MD,4 Francisco Antunes, MD,5 Christine Katlama, MD,6 Ulrik Stenz Justesen, MD,7 Stefano Vella, MD,8 Ole Kirk, MD,9 and Jens Lundgren, MD,9 for the EuroSIDA Study Group A drastic decrease in incidence has been observed for most human immunodeficiency virus (HIV)–related opportunistic manifestations after use of highly active antiretroviral therapy (HAART). We assessed the trend of incidence of central nervous system (CNS) diseases in a prospective multicenter observational study involving 9,803 patients across Europe in the period 1994 to 2002 and analyzed patient and treatment variables associated with these conditions. Overall, 568 patients (5.8%) received a diagnosis of a new CNS disease. Incidence decreased significantly from 5.9 per 100 personyear in 1994 to 0.5 in 2002. Overall, the decrease was 40% per calendar year, and it was similar to that of non-CNS diseases and less evident after year 1998. In multivariable models, low CD4 cell count and high plasma viral load, but not HAART or calendar year, were significantly associated with risk to develop CNS disease, indicating that the effect of HAART was likely mediated by both improved immunological conditions and inhibition of viral replication. In contrast, use of nucleoside reverse transcriptase inhibitors, irrespective of use of protease inhibitors or non-nucleoside reverse transcriptase inhibitors, appeared to protect specifically against acquired immunodeficiency disease syndrome dementia complex, suggesting that, in this condition, therapy might have a direct, additive effect in the CNS. Ann Neurol 2004;55:320 –328 Acquired immunodeficiency disease syndrome (AIDS)– related diseases of the central nervous system (CNS) occur in late stages of human immunodeficiency virus (HIV) infection and are associated with poor prognosis. These include opportunistic infections and tumors and a spectrum of neurological disorders associated with CNS infection by HIV itself, generally referred to as AIDS dementia complex (ADC). Similar to what has been observed for extracerebral complications of HIV infection, a drastic decline of incidence and prevalence has been documented for both opportunistic CNS diseases and ADC by several retrospective or cohort studies after the introduction of highly active antiretroviral therapy (HAART).1–7 Theoretically, the effect of HAART on CNS diseases might differ from that on extracerebral complications because of the anatomical and functional properties of the CNS itself. The presence of the brain barriers might limit anti-HIV drug penetration into this com- partment and subsequent suppression of viral replication. Furthermore, unlike systemic infection, the principal CNS cells harboring HIV are macrophages and microglial cells, in which dynamics of viral replication and response to antiretroviral drugs may be different compared with peripheral lymphocytes.8 –10 These features, along with an increased patient survival induced by HAART, might expose patients to a risk of developing HIV-related complications in the CNS, especially ADC. Actually, an increased prevalence of HIV encephalopathy recently has been reported in postmortem cases associated with the use of HAART,11,12 but this finding is controversial.13 With this study, we assessed the trend of incidence of CNS diseases in Europe, including both CNS opportunistic diseases and ADC, during the period 1994 to 2002. In particular, we evaluated whether the overall trend paralleled that observed for non-CNS opportunistic diseases and whether this differed between ADC From the 1Institute of Infectious and Tropical Diseases, University of Milan; 2Clinic of Infectious Diseases, San Raffaele Hospital, Milan, Italy; 3Royal Free and University College Medical School, Department of Primary Care and Population Sciences, London, United Kingdom; 4Klinikum Innerstadt Medizinische Poliklinik Infektionsambulanz and Tagesklinik, Munich, Germany; 5Department of Infectious Diseases, Hospital Santa Maria, Lisbon, Portugal; 6Department de Medicine Tropicale, Hospital de la Pitié-Salpêtriére, Paris, France; 7Department of Infectious Diseases, Odense University Hospital, Odense, Denmark; 8Laboratory of Virology, Istituto Superiore di Sanità, Rome, Italy; and 9EuroSIDA Coordinating Centre, Copenhagen HIV Programme, Department 044, Hvidovre University Hospital, Hvidovre, Denmark. Received Jun 30, 2003, and in revised form Sep 17. Accepted for publication Sep 17, 2003. 320 Members of the multicenter study group on EuroSIDA are listed in the Appendix on pages 326 –327. Address correspondence to Dr d’Arminio Monforte, Institute of Infectious and Tropical Diseases, University of Milan, L Sacco Hospital, Via GB Grassi, 74, 20157 Milan, Italy. E-mail: firstname.lastname@example.org © 2004 American Neurological Association Published by Wiley-Liss, Inc., through Wiley Subscription Services and CNS opportunistic diseases and among the individual CNS opportunistic diseases. Finally, we investigated the possible influence of various factors on development of CNS diseases. Patients and Methods The EuroSIDA Study Cohort The EuroSIDA study is a prospective, European study of patients with HIV-1 infection in 70 centers across Europe also including Israel and Argentina (see appendix). Details of the study have been published.14 In brief, five cohorts of patients have been recruited. Eligible patients were older than 16 years at the time of enrolment and had a CD4 lymphocyte count of below 500/mm3 in the previous 4 months for Cohort I to III but no CD4 count restriction for Cohorts IV and V. Clinical and laboratory information was provided on a standardized data collection form at baseline and every 6 months thereafter. Dates of diagnosis of all AIDS defining illnesses were recorded using the 1993 AIDS definition from the Centers for Disease Control.15 Presumptive diagnoses of primary CNS lymphoma, progressive multifocal leukoencephalopathy (PML), or ADC were accepted; for full details on data collection and diagnostic criteria refer to http://www. cphiv.dk. Statistical Methods All patients with at least one follow-up visit after recruitment to EuroSIDA were included in this study and follow-up is to autumn 2002. The following CNS diseases have been included: toxoplasmosis, cryptococcosis, PML, primary brain lymphoma (PBL), ADC, and focal brain lesions (FBLs) without an causative diagnosis. The incidence of CNS diagnoses overall and individually was calculated using a person-years analysis. Person-years of follow-up were calculated from date of recruitment to the EuroSIDA study and ended at last clinical follow-up, date of death, or development of CNS disease (CNS-D), whichever occurred first. The follow-up was further stratified according to calendar period. Tests for changes in the incidence of CNS-D over time were performed using Poisson regression. Patients with a CNS-D at recruitment were included and were followed up until the diagnosis of their next, distinct CNS-D. Only the first event of a given disease was included. Standard Cox proportional hazards models were used to investigate progression to CNS-D and individually for ADC or all other CNS-D combined. Demographic factors, such as sex, age, and exposure category were included in models to determine their relationship with progression to CNS-D. All Cox models were stratified by center and factors which were significant in univariable analyses with p value less than 0.1 then were included in multivariable models. Because models were stratified by center, it was not possible to investigate differences between regions of Europe. Factors known to be related to progression of HIV disease (the development of AIDS and CD4 lymphocyte count) were included in Cox models as time dependent. To investigate whether the risk of CNS-D had changed over time, we modeled calendar period as a time-dependent covariate using 1998 as the reference period. Antiretroviral treatment regimen also was included as time-dependent covariates, as starting any nucleoside reverse transcriptase inhibitor (NRTI)–containing therapy, either in combination or not with antiretroviral drugs of other classes, starting a protease inhibitor (PI)–based HAART regimen, or starting a non-NRTI (NNRTI)–based HAART regimen, HAART being defined as at least triple therapy including a PI or a NNRTI. Treatment was included as an intent-totreat variable. A further Cox model was constructed to determine the effects of drugs thought to cross the blood–brain barrier (all nucleosides, NNRTIs and indinavir); this model was adjusted for the same factors as previously described, and antiretroviral treatment was included as two variables which counted the number of drugs in the regimen which do or do not cross the blood–brain barrier. Results A total of 9,803 patients of the EuroSIDA cohort were enrolled; 9,348 were free from CNS diseases at enrollment and 455 had already suffered at least from one CNS-D for a total of 479 diagnoses. Characteristics of the patients are summarized in Table 1. Median follow-up after recruitment was significantly higher in patients free from CNS diseases (42.0 months IQR 13– 65 vs 17 months IQR 6 –54; p ⬍ 0.0001). In a median follow-up of 40.9 months (IQR, 12.0 – 65.0), 568 patients (5.8%) received a diagnosis of a first new CNS-D. This was the first AIDS-defining illness in 174 patients (30.6%). ADC was the most frequent disease after recruitment, diagnosed in 190 patients (33.5%), followed by toxoplasmosis in 152 (26.8%), PML in 83 (14.6%), cryptococcosis in 64 (11.3%), PBL in 52 (9.2%), and FBL in 27 (4.8%). Figure 1 illustrates the incidence of CNS diseases considered as a whole according to calendar period. The rate of decline in CNS-D was 40% per calendar year (95% confidence interval [CI], 37– 42%) and was very similar to that of non-CNS diseases (40%; 95% CI, 30 – 41%). Starting from 1998, there was no evidence of any further decline in CNS-D (9% per year, 95% CI: 8% increase to 23% decrease; p ⫽ 0.29). The 455 patients already suffering from CNS diseases at recruitment showed a higher incidence of new diagnoses of CNS-D as compared with the whole population, but a similar decline over time (estimated 32% per year, 95% CI: 31– 41%, p ⬍ 0.0001; data not shown). Considering the incidence of individual CNS diseases (Fig 2), the rate of annual decrease for ADC (45%, 95% CI, 40 – 49%) was statistically higher than that of opportunistic CNS diseases as a whole (37%, 95% CI, 34 – 41%; p ⬍ 0.01). Differences were observed between individual CNS diseases, with the lowest annual decrease of incidence observed for PML (31%, 95% CI, 24 –38%) and the highest for PBL (48%, 95% CI, 39 –59%). Table 2 gives an overview of the CNS diagnoses in d’Arminio Monforte et al: CNS Diseases 321 Table 1. Characteristics of the Study Population at Recruitment to EuroSIDA Without Previous CNS Disease All Characteristic All Sex Men Women Risk Homosexual IDU Heterosexual Other Race White Other Region South Central North East Argentina Treatment ever taken before/at None HAART PI NNRTI Nucsa All three classes Other AIDS at recruitment No Yes Other values median (IQR) Age CD4 at recruitment VL at recruitment N % N % 9,803 100 9,348 95.4 With Previous CNS Disease N % p 455 4.6 — 7,678 2,125 78.3 21.7 7,298 2,050 78.1 21.9 380 75 83.5 16.5 0.0059 4,403 2,405 2,319 676 44.9 24.5 23.7 6.9 4,185 2,294 2,222 647 49.8 24.5 23.8 6.9 218 111 97 29 47.9 24.4 21.3 6.4 0.52 4,403 5,400 44.9 55.1 4,185 5,163 44.8 55.2 218 237 47.9 52.1 0.19 29.5 26.6 32.5 10.3 1.0 2,773 2,470 3,043 971 91 29.7 26.4 32.6 10.4 1.0 118 139 148 41 9 25.9 30.5 32.5 9.0 2.0 0.042 21.8 30.0 30.8 9.7 77.3 5.4 2,097 2,772 2,820 910 7,169 494 22.4 29.7 30.2 9.7 76.7 5.3 39 171 199 43 414 32 8.6 37.6 43.7 9.4 91.0 7.0 ⬍0.0001 0.0003 ⬍0.0001 0.84 ⬍0.0001 0.11 71.6 28.4 6,826 2,522 73.0 27.0 188 267 41.3 58.7 ⬍0.0001 32.6–43.9 16–184 2.39–4.42 0.0012 ⬍0.0001 0.30 2,891 2,609 3,191 1,012 100 recruitment 2,136 2,943 3,019 953 7,583 526 7,014 2,789 36.2 238 3.11 31.3–43.2 108–380 2.30–4.38 36.1 245 3.11 31.3–43.2 115–384 2.30–4.38 37.6 78 3.02 A total of 467 diagnoses of CNS-D were made in these 455 patients a median of 12 mo before recruitment to EuroSIDA (IQR: 6 –25 mo). These included toxoplasmosis in 217 patients (48%), ADC in 120 (26%), cryptococcosis in 88 (19%), PML in 40 (9%), FBL in 8 (2%), and PBL in 6 (1%). VL was available at recruitment to EuroSIDA for 5,169 patients (52.7%); 4,934 (52.8%) without, and 235 (51.7%) with previous CNS at recruitment ( p ⫽ 0.64, 2 test). a Any regimen containing NRTIs. CNS ⫽ central nervous system; IDU ⫽ intravenous drug user; IQR ⫽ interquartile range; VL ⫽ viral load; HAART ⫽ highly active antiretroviral therapy; PI ⫽ protease inhibitor; NNRTI ⫽ nonnucleoside reverse transcriptase inhibitor; IDU ⫽ intervenous drug user. consecutive calendar years. CD4 cell counts at the diagnosis of CNS-D were higher in later as compared with earlier years ( p ⬍ 0.0001, Wilcoxon test), although the increase was mainly among diagnoses made in 2001 or later. There was no significant change over time in the relative proportion of cases with ADC or opportunistic CNS infections ( p ⫽ 0.095, test), in the HIV-RNA copy levels at diagnosis ( p ⫽ 0.11, Wilcoxon test) or in the proportion of CNS diseases as index diseases ( p ⫽ 0.57, test, see Table 2). Possible factors predicting progression to any CNS-D were evaluated by univariable and multivariable analyses, either excluding or including viral load (Table 3). Although a very strong effect of anti-HIV 322 Annals of Neurology Vol 55 No 3 March 2004 therapy on the risk of CNS diseases was shown in the univariable analysis, this effect disappeared after adjusting for calendar year and latest CD4 count. In the first multivariable model, in which viral load was not analyzed, the latest CD4 counts were the most significant independent variable associated with CNS-D, whereas calendar year after 1998 and antiretroviral regimens were not significant risk predictors. By considering only patients with available viral load, we found that both latest CD4 count and HIV-RNA copy levels were strong predictors of CNS-D, but, again, neither therapy regimen nor calendar year after 1998 were predictive. By using the same model, we also evaluated the risk Fig 1. Incidence of central nervous system (CNS) (circles) and non-CNS (diamonds) AIDS-defining diseases in the EuroSIDA cohort, years 1994 to 2002. PYFU ⫽ person-year follow-up. of developing ADC or CNS opportunistic infections separately (Table 4). The risk of both conditions was significantly reduced in patients with higher CD4 counts; moreover, higher risk of ADC was significantly associated with age and intravenous drug usage. In addition, starting any NRTI-containing regimen was independently associated with a reduced risk of developing ADC hazard ratio (HR, 0.59; 95% CI, 0.39 – 0.90; p ⫽ 0.013), irrespective of the presence of PI or NNRTIs in a given regimen. In subset analyses, a higher latest viral load was not associated with an increased risk of ADC (HR, 1.20; 95% CI, 0.92–1.56, p ⫽ 0.17), but with an increased risk of other CNS diseases (HR 1.61 per 1 log higher 95% CI 1.37–1.89, p ⬍ 0.0001) (data not shown). In the univariable analysis, there was a 39% reduced risk of CNS-D (HR, 0.61; 95% CI, 0.45– 0.84; p ⫽ 0.0021) for each extra drug though to cross the blood– brain barrier, that is, all NRTIs, all NNRTIs, and indinavir.10,16,17 However, a nonsignificant reduced risk of CNS-D for each extra drug thought to cross the blood–brain barrier was observed after adjusting for exposure group, year of follow-up, age, diagnosis of a new AIDS defining illness, and CD4 both at recruitment or at last examination (HR, 0.77; 95% CI, 0.56 – 1.05; p ⫽ 0.093 and HR 0.81, 95% CI, 0.59 –1.11; p ⫽ 0.19). Discussion In this study, we observed a significant decrease of the incidence of AIDS-defining CNS diseases through Europe, after the extensive usage of HAART. This decrease was generalized to all individual CNS diseases and similar to that observed in non-CNS, HIV-related opportunistic diseases. To our knowledge, this is the largest study addressing this topic, with nearly 10,000 patients included. HAART was introduced in Europe starting from 1995 to 1996, with regional differences.16 This analysis shows that the incidence decline of CNS diseases was evident since 1996, that is, a few months after HAART introduction, similarly to what observed in other studies.2–7 Because only a small proportion of patients in the cohort had access to HAART in that year, it is possible that double combination therapy, widely used since 1995, might have been in part responsible of the initial decline. On the other hand, a slower decline in incidence was observed after 1998, with overall incidence leveling at between 0.3 and 0.5 per 100 persons per year of follow-up. A similar, slowing rate of decline recently has been reported for both global AIDSrelated events and deaths during the same period.17,18 It will be essential to observe whether incidence figures increase in the coming years, as a result of the increasing number of treatment failures.19 As already observed for all AIDS-defining events, CD4 counts at the diagnosis of CNS-D increased over time.2,14,20,21 It is possible that the initially HAARTinduced increase in circulating CD4 cells was not protective against infections in the first months of therapy.22 Alternatively, this finding might reflect the proportion of events occurring in patients with relatively high CD4 counts, due to a cohort effect.14 Nevertheless, by multivariable analysis, CD4 counts, together with viral load, but not treatment regimen, were predictors of risk to develop CNS-D. This observation can be explained by the effect of antiretroviral therapy on viral replication, resulting in restoration of the immune system and prevention of AIDS events. In fact, excluding latest CD4 counts from the multivariable model, the protective effect of HAART is maintained (data not shown). Fig 2. Incidence of individual central nervous system (CNS) diseases in the EuroSIDA cohort, years 1994 to 2002. ADC ⫽ AIDS-dementia complex; TOX ⫽ brian toxoplasmosis; CCOC ⫽ cryptococcosis; PML ⫽ progressive multifocal encephalopathy; PBL ⫽ primary brain lymphoma; FBL ⫽ focal brain lesions; CI ⫽ confidence interval. PYFU ⫽ person-year follow-up. d’Arminio Monforte et al: CNS Diseases 323 Table 2. Number of CNS Diagnoses, CD4 Cell Counts, and HIV-RNA Copy Levels at the Time of Diagnosis and Rate of CNS Disease as AIDS Index Disease over Calendar Time CNS diagnoses, N (%) CNS-OI ADC CNS disease as ID CD4 at diagnosis, median (IQR) VL at diagnosis (range), median (IQR) Patients with VL data, N (%) 1994 1995 1996 1997 1998 1999 2000 2001 and after 103 164 141 54 28 32 22 24 101 (61.6) 63 (38.4) 48 (29.3) 89 (63.1) 52 (36.9) 46 (32.6) 42 (77.8) 12 (22.2) 19 (35.2) 20 (71.4) 8 (28.6) 10 (35.7) 27 (84.4) 5 (15.9) 11 (34.4) 16 (72.7) 6 (27.3) 9 (40.9) 27 (10–69) 20 (7–70) 21 (8–71) 46 (8–101) 30 (10–77) 64 (21–185) 30 (9–68) 109 (20–156) 3.08 (—) 4.80 (4.67–5.00) 5.29 (4.66–5.64) 4.92 (3.60–5.60) 4.70 (3.92–5.30) 4.62 (2.60–5.32) 4.46 (2.60–5.13) 4.87 (2.48–5.60) 3 (1.8) 30 (21.3) 36 (66.7) 27 (96.4) 31 (96.9) 65 (63.1) 38 (36.9) 27 (26.2) 1 (1.0) 18 (75.0) 6 (25.0) 4 (16.7) 21 (95.5) 24 (100) CNS ⫽ central nervous system; HIV ⫽ human immunodeficiency virus; AIDS ⫽ acquired immunodeficiency syndrome; OI ⫽ opportunistic diseases; ADC ⫽ AIDS dementia complex; ID ⫽ index disease; IQR ⫽ interquartile range; VL ⫽ viral load. It has been hypothesized that, because of poor penetration of some drugs into the CNS, HAART could be less efficient in preventing CNS than non-CNS diseases. This was particularly a concern for ADC, in which failing to achieve therapeutical antiretroviral drug levels in the brain might theoretically prevent efficient viral suppression locally.10 Furthermore, HIV has been shown in vitro to transactivate several opportunistic agents, including JC virus, Mycobacterium tu- berculosis, or Cryptococcus neoformans,23–25 and therefore lack of HIV inhibition in the CNS might theoretically favor the local development of opportunistic manifestations. However, the observation of a similar incidence trend observed in CNS and non-CNS diseases argues against a lower neuroprotective effect of HAART. Furthermore, we found no significant association between the number of drugs thought to cross the blood– brain barrier and the development of CNS-D. Table 3. Relative Risk of Any CNS Disease after Recruitment to the EuroSIDA Study Multivariablea Univariable Risk Homosexual IDU Heterosexual Other Yearc 1994 1995 1996 1997 1998 1999 2000 2001 or later Age (per 10 yr older) AIDSc (Tdc) Latest CD4c (50% higher) Started nucsd (Tdc) Started PI-HAARTc (Tdc) Started NNRTI-HAARTc (Tdc) VLc (log higher) Multivariableb RH 95% CI p RH 95% CI p RH 95% CI p 1.00 1.06 0.85 0.60 — 0.85–1.39 0.68–1.08 0.38–0.93 — 0.59 0.18 0.024 1.00 1.12 0.99 0.56 — 0.88–1.42 0.78–1.26 0.35–0.88 — 0.35 0.93 0.011 1.00 1.41 1.18 0.49 — 0.88–2.25 0.77–1.80 0.21–1.12 17.17 14.27 9.12 2.02 1.00 0.94 0.50 0.55 1.04 1.84 0.60 1.08 0.36 9.21–32.02 8.43–24.16 5.43–15.30 1.19–3.45 — 0.49–1.79 0.24–1.05 0.25–1.20 0.94–1.14 1.16–2.93 0.58–0.62 0.86–1.35 0.29–0.45 ⬍0.0001 ⬍0.0001 ⬍0.0001 0.0096 — 0.84 0.065 0.13 0.47 0.0098 ⬍0.0001 0.51 ⬍0.0001 7.11 4.23 4.23 1.66 1.00 1.13 0.68 0.82 1.10 0.93 0.63 0.95 0.81 3.45–14.65 2.34–7.64 2.38–7.55 0.96–2.89 — 0.58–2.19 0.32–1.45 0.36–1.84 1.00–1.22 0.58–1.50 0.61–0.65 0.75–1.20 0.61–1.09 ⬍0.0001 ⬍0.0001 ⬍0.0001 0.070 — 0.72 0.32 0.63 0.055 0.76 ⬍0.0001 0.65 0.86 5.66 5.68 2.65 1.13 1.00 1.21 0.86 1.08 1.15 1.12 0.57 0.94 0.93 0.30–105.44 0.75–42.85 1.12–6.29 0.58–2.19 — 0.58–2.53 0.38–1.96 0.45–2.61 0.96–1.37 0.50–2.51 0.53–0.62 0.62–1.43 0.65–1.35 0.25 0.092 0.027 0.72 — 0.62 0.72 0.87 0.13 0.78 ⬍0.0001 0.76 0.71 0.59 2.29 0.42–0.82 2.02–2.60 0.0016 1.04 ⬍0.0001 — 0.69–1.57 — 0.86 — 0.77 1.43 0.48–1.22 1.24–1.65 0.26 ⬍0.0001 — 0.15 0.45 0.090 CNS ⫽ central nervous system; CI ⫽ confidence interval; IDU ⫽ intravenous drug user; PI ⫽ protease inhibitor; HAART ⫽ highly active antiretroviral therapy; NNRTI ⫽ nonnucleoside reverse transcriptase inhibitor; VL ⫽ viral load; RH ⫽ relative hazard. a Multivariable model excluding viral load. Multivariable model including viral load. Fitted as time-dependent covariate (tdc). d Nucs is any regimen containing NRTIs. b c 324 Annals of Neurology Vol 55 No 3 March 2004 Table 4. Multivariable Relative Risk of Any AIDS Dementia or Other CNS Disease after Recruitment to the EuroSIDA Study AIDS Dementia Complex Risk Homosexual IDU Heterosexual Other Yeara 1994 1995 1996 1997 1998 1999 2000 2001 and after Age (per 10 yr older) AIDSa (Tdc) Latest CD4a (50% higher) Started Nucsa (Tdc) Started PI-HAARTa (Tdc) Started NNRTIHAARTa (Tdc) RH 95% CI 1.00 1.84 1.18 0.71 — 1.23–2.77 0.78–1.77 0.34–1.46 8.48 7.63 3.89 1.10 1.00 0.54 0.63 0.36 1.38 0.72 Other CNS Disease p RH 95% CI p — 0.0032 0.43 0.35 1.00 0.87 1.06 0.48 — 0.65–1.16 0.81–1.39 0.28–0.84 — 0.87 0.67 0.011 2.23–32.33 2.62–22.28 1.35–11.24 0.38–3.15 — 0.13–2.17 0.17–2.31 0.07–1.82 1.17–1.63 0.31–1.68 0.0017 0.0002 0.012 0.87 — 0.39 0.48 0.22 ⬍0.0001 0.44 6.03 3.04 4.01 1.85 1.00 1.30 0.61 0.85 1.06 1.18 2.61–13.94 1.55–5.96 2.08–7.72 0.99–3.45 — 0.62–2.72 0.25–1.48 0.34–2.13 0.94–1.19 0.71–1.97 ⬍0.0001 0.0012 ⬍0.0001 0.053 — 0.48 0.27 0.73 0.36 0.52 0.64 0.59 0.61–0.68 0.39–0.90 ⬍0.0001 0.013 0.62 1.06 0.60–0.65 0.80–1.40 ⬍0.0001 0.69 0.83 0.48–1.41 0.48 0.81 0.59–1.12 0.20 0.96 0.43–2.15 0.91 1.01 0.64–1.59 0.98 AIDS ⫽ acquired immunodeficiency syndrome; CNS ⫽ central nervous system; CI ⫽ confidence interval; IDU ⫽ intravenous drug user; PI ⫽ protease inhibitor; HAART ⫽ highly active antiretroviral therapy; NNRTI ⫽ nonnucleoside reverse transcriptase inhibitor; RH ⫽ relative hazard. a Fitted as time-dependent covariate (tdc). Any regimen containing NRTIs. b However, considering the individual CNS diseases, we observed variation in the incidence decline, with decrease rates higher for ADC and PBL and slower for PML. The relatively slow incidence decline of PML is consistent with recent prevalence data, showing a relatively high frequency of PML cases in the HAART era, second only to toxoplasmosis as more frequent opportunistic CNS-D.26 This observation might partly be explained by considering the proportion of PML cases that become manifest in the context of HAARTinduced immune reconstitution.27,28 With regard to ADC, literature data on the effect of HAART on this disease are contrasting. Whereas some authors showed a positive correlation between HAART usage and reduction of ADC or HIV-related lesions,3,6,13 others reported the persistence of a relevant number of HIVinduced CNS disease cases despite HAART.2,11,12,29 In this regard, our findings show that the effect of HAART on ADC incidence was at least as important, if not superior to that on other CNS diseases. The multivariable model showed that having started NRTIs in any regimen was associated with a 41% reduction of the risk of ADC, over and above the protective effect of starting any HAART regimen. Because higher CD4 counts were significantly associated with a reduced risk of both ADC and opportunistic CNS diseases, the protective effect of NRTIs in ADC appears to lay beyond that exerted through the CD4 cells increase. It is known that even zidovudine alone is effective in both prevention and treatment of HIV-related encephalopathy.30,31 A decline of ADC incidence and of prevalence of HIV-related brain lesions was indeed observed in the years after the introduction of this drug in clinical practice.30,32 Furthermore, more recent data indicate that dual NRTI combinations preserve or improve neurological performance.33 Whereas the benefit of single or dual NRTIs regimens on systemic immunological and virological markers is usually mild and transitory,16 these same regimens might be effective in the CNS because of the peculiarity of the brain compartment for HIV replication, and their use have contributed substantially to the marked decline of ADC incidence. Our study has some limitations. First, the actual number of events probably was underdiagnosed, because we did not consider the diagnosis of some CNS diseases, such as cytomegalovirus (CMV) encephalitis or mycobacteriosis. Although mycobacteriosis is uncommon in the CNS, CMV encephalitis used to be d’Arminio Monforte et al: CNS Diseases 325 very frequent in the years immediately preceding the introduction of HAART, being reported in approximately 20% of postmortem AIDS cases.11 In addition, we examined only the incidence of the first diagnoses of CNS diseases after recruitment, thus excluding both recurrences of the same CNS event and new events in the same patient. Second, plasma viral load was not available from most of the patients during the first years of EuroSIDA, thus reducing the possibility of examining this marker as risk predictor. Finally, data on primary prophylaxis for opportunistic CNS diseases, namely, toxoplasmosis and cryptococcosis, were not collected in detail and therefore not included in this analysis. Nonetheless, we briefly examined the use of primary prophylaxis for toxoplasmosis and verified that it remained fairly stable over calendar time approximately 70 to 80% among patients with severe immune deficiency, that is, with CD4 counts less than 100/l (data not shown). Thus, changes in use of prophylaxis did not explain the decreasing incidence of toxoplasmosis. On the other hand, an increasing proportion of patients stopped specific prophylaxis after HAARTinduced increase of CD4 cell counts,34 and thus changes in use of prophylaxis was unlikely to explain the decreasing incidence of toxoplasmosis. In conclusion, this study showed a reduction of incidence of CNS diseases across Europe after the introduction of HAART. High CD4 counts and low plasma viral load were strongly associated with a lower risk of CNS-D, indicating that the immune reconstitution is the main factor responsible for this incidence decline. However, an additive, direct effect of antiretroviral therapy on virus replication in CNS might have accounted for the marked incidence decline observed in ADC. This study also showed a stabilization of CNS disease incidence after 1998. This finding, together with the consideration that an increasing proportion of HIV-infected patients undergo multiple treatment failure, underscores the importance of continuing surveillance for CNS disease manifestations. Appendix The EuroSIDA Study Group The multicenter study group on EuroSIDA (national coordinators in parentheses) included the following: Argentina: (M. Losso), A. Duran, Hospital J. M. Ramos Mejia, Buenos Aires; Austria: (N. Vetter) Pulmologisches Zentrum der Stadt Wien, Vienna; Belgium: (N. Clumeck) P. Hermans, B. Sommereijns, Saint-Pierre Hospital, Brussels; R. Colebunders, Institute of Tropical Medicine, Antwerp; Czech Republic: (L. Machala) H. Rozsypal, Faculty Hospital Bulovka, Prague; Denmark: (J. Nielsen) J. Lundgren, T. Benfield, O. Kirk, Hvidovre Hospital, Copenhagen; J. Gerstoft, T. Katzenstein, A.-B. E. Hansen, P. Skinhøj, 326 Annals of Neurology Vol 55 No 3 March 2004 Rigshospitalet, Copenhagen; C. Pedersen, Odense University Hospital, Odense; Estonia: (K. Zilmer) Tallinn Merimetsa Hospital, Tallinn; France: (C. Katlama) M. De Sa, Hôpital de la Pitié-Salpétière, Paris; J.-P. Viard, Hôpital Necker-Enfants Malades, Paris; T. Saint-Marc, Hôpital Edouard Herriot, Lyon; P. Vanhems, University Claude Bernard, Lyon; C. Pradier, Hôpital de l’Archet, Nice; Germany: (M. Dietrich) C. Manegold, Bernhard-Nocht-Institut for Tropical Medicine, Hamburg; J. van Lunzen, H.-J. Stellbrink, Eppendorf Medizinische Kernklinik, Hamburg; V. Miller, S. Staszewski, J. W. Goethe University Hospital, Frankfurt; F.-D. Goebel, Medizinische Poliklinik, Munich; Bernd Salzberger, Universität Köln, Cologne; J. Rockstroh, Universitäts Klinik, Bonn; Greece: (J. Kosmidis) P. Gargalianos, H. Sambatakou, J. Perdios, Athens General Hospital, Athens; G. Panos, I. Karydis, A. Filandras, 1st IKA Hospital, Athens; Hungary: (D Banhegyi) Szent Lásló Hospital, Budapest; Ireland: (F. Mulcahy) St. James’s Hospital, Dublin; Israel: (I. Yust) M. Burke, Ichilov Hospital, Tel Aviv; S. Pollack, Z. Ben-Ishai, Rambam Medical Center, Haifa: Z. Sthoeger, Kaplan Hospital, Rehovot; S. Maayan, Hadassah University Hospital, Jerusalem; Italy: (S. Vella, A. Chiesi) Istituto Superiore di Sanita, Rome; C. Arici, Ospedale Riuniti, Bergamo; R. Pristerá, Ospedale Generale Regionale, Bolzano; F. Mazzotta, A. Gabbuti, Ospedale S. Maria Annunziata, Florence; R. Esposito, A. Bedini, Università di Modena, Modena; A. Chirianni, E. Montesarchio, Presidio Ospedaliero A.D. Cotugno, Naples; V. Vullo, P. Santopadre, Università di Roma La Sapienza, Rome; P. Narciso, A. Antinori, P. Franci, M. Zaccarelli, Ospedale Spallanzani, Rome; A. Lazzarin, R. Finazzi, Ospedale San Raffaele, Milan; A. D’Arminio Monforte, Osp. L. Sacco, Milan; Latvia: (L. Viksna) Infectology Centre of Latvia, Riga; Lithuania: (S. Chaplinskas) Lithuanian AIDS Centre, Vilnius; Luxembourg: (R. Hemmer), T. Staub, Centre Hospitalier, Luxembourg; Netherlands: (P. Reiss) Academisch Medisch Centrum bij de Universiteit van Amsterdam, Amsterdam; Norway: (J. Bruun) A. Maeland, V. Ormaasen, Ullevål Hospital, Oslo; Poland: (B. Knysz) J. Gasiorowski, Medical University, Wroclaw; A. Horban, Centrum Diagnostyki i Terapii AIDS, Warsaw; D. Prokopowicz, A. Wiercinska-Drapalo, Medical University, Bialystok; A. Boron-Kaczmarska, M. Pynka, Medical Univesity, Szczecin; M. Beniowski, Osrodek Diagnostyki i Terapii AIDS, Chorzow; H. Trocha, Medical University, Gdansk; Portugal: (F. Antunes) Hospital Santa Maria, Lisbon; K. Mansinho, Hospital de Egas Moniz, Lisbon; R. Proenca, Hospital Curry Cabral, Lisbon; Romania: D. Duiculescu, Spitalul de Boli Infectioase si Tropicale Dr Victor Babes, Bucarest; A. Streinu-Cercel, Institute of Infectious Diseases, Bucarest; Slovakia: (M. Mokrá) D. Staneková, M. Hábeková, V. Mayer, Dérer Hospital, Bratislava; Spain: (J. González-Lahoz) B. Diaz, T. Garcı́a-Benayas, L. Martin-Carbonero, V. Soriano, Hospital Carlos III, Madrid; B. Clotet, A. Jou, J. Conejero, C. Tural, Hospital Germans Trias i Pujol, Badalona; J. M. Gatell, J. M. Miró, Hospital Clinic i Provincial, Barcelona; Sweden: (A. Blaxhult) Karolinska Hospital, Stockholm; A. Karlsson, Södersjukhuset, Stockholm; P. Pehrson, Huddinge Sjukhus, Stockholm; Switzerland: (B. Ledergerber) R. Weber, University Hospital, Zürich; P. Francioli, A. Telenti, Centre Hospitalier Universitaire Vaudois, Lausanne; B. Hirschel, V. Soravia-Dunand, Hospital Cantonal Universitaire de Geneve, Geneve; H. Furrer, Inselspital Bern, Bern; Ukraine: (N. Chentsova) Kyiv Centre for AIDS, Kyiv; United Kingdom: (S. Barton) St. Stephen’s Clinic, Chelsea and Westminster Hospital, London; A. M. Johnson, D. Mercey, Royal Free and University College London Medical School, London (University College Campus); A. Phillips, M. A. Johnson, A. Mocroft, Royal Free and University College Medical School, London (Royal Free Campus); A. Pinching, J. Parkin, Medical College of Saint Bartholomew’s Hospital, London; J. Weber, G. Scullard, Imperial College School of Medicine at St. Mary’s, London; M. Fisher, Royal Sussex County Hospital, Brighton; R. Brettle, Western General Hospital, Edinburgh. Virology Group C. Loveday, B. Clotet (Central Coordinators) plus ad hoc virologists from participating sites in the EuroSIDA Study. Steering Committee Francisco Antunes, Anders Blaxhult, Nathan Clumeck, Jose Gatell, Andrzej Horban, Anne Johnson, Christine Katlama, Bruno Ledergerber (chair), Clive Loveday, Andrew Phillips, Peter Reiss, Stefano Vella. Coordinating Center Staff J. Lundgren (project leader), I. Gjørup, O. Kirk, N. Friis-Moeller, A. Mocroft, A. Cozzi-Lepri, L. Paddam, D. Mollerup, M. Nielsen, A. Hansen, D. Kristensen, L. Kolte, L. Hansen, J. Kjær. This study was sponsored by the European Commission BIOMED 1 (CT94-1637), BIOMED 2 (CT97-2713), and the 5th framework (QLK2-2000-00773) programs and grants from Bristol-Myers Squibb, GlaxoSmithKline, Roche, and Boehringer-Ingelheim. The participation of centers from Switzerland was supported by a grant from the Swiss Federal Office for Education and Science. References 1. Bacellar H, Munoz A, Miller EN, et al. Temporal trends in the incidence of HIV-1 related neurologic diseases: multicenter AIDS Cohort Study. Neurology 1994;44:1892–1900. 2. Dore GJ, Correll PK, Li Y, et al. Changes to AIDS dementia complex in the era of highly active antiretroviral therapy. AIDS 1999;13:1249 –1253. 3. d’Arminio Monforte A, Duca PG, Vago L, et al. Decreasing incidence of CNS AIDS-defining events associated with antiretroviral therapy. Neurology 2000;54:1856 –1859. 4. Ammassari A, Cingolani A, Pezzotti P, et al. AIDS-related focal brain lesions in the era of highly active antiretroviral therapy. Neurology 2000;55:1194 –1200. 5. Kirk O, Pedersen C, Cozzi-Lepri A, et al. Non-Hodgkin lymphoma in HIV-infected patients in the era of highly active antiretroviral therapy. Blood 2001;98:3406 –3412. 6. Sacktor N, Lyles RH, Skolasky R, et al. HIV-associated neurologic disease incidence changes: multicenter AIDS Cohort Study, 1990 –1998. Neurology 2001;56:257–260. 7. Maschke M, Kastrup O, Esser S, et al. Incidence and prevalence of neurological disorders associated with HIV since the introduction of highly active antiretroviral therapy (HAART). J Neurol Neurosurg Psychiatry 2000;69:376 –380. 8. Sommadossi JP. Comparison of metabolism and in vitro antiviral activity of stavudine versus other 2⬘,3⬘-dideoxynucleoside analogues. J Infect Dis 1995;171(suppl 2):S88 –S92. 9. Perno CF, Newcomb FM, Davis DA, et al. Relative potency of protease inhibitors in monocytes/macrophages acutely and chronically infected with human immunodeficiency virus. J Infect Dis 1998;178:413– 422. 10. Enting RH, Hoetelmans RMW, Lange JMA, et al. Antiretroviral drugs and the central nervous system. AIDS 1998;12: 1941–1955. 11. Masliah E, DeTeresa RM, Mallory ME, Hansen LA. Changes in pathological findings at autopsy in AIDS cases for the last 15 years. AIDS 2000;14:69 –70. 12. Neuenburg JK, Brodt HR, Herndier BG, et al. HIV-related neuropathology, 1985 to 1999: rising prevalence of HIV encephalopathy in the era of highly active antiretroviral therapy. J Acquir Immune Defic Syndr 2002;31:171–177. 13. Vago L, Bonetto S, Nebuloni M, et al. Pathological findings in the central nervous system of AIDS patients on assumed antiretroviral therapeutic regimens: retrospective study of 1597 autopsies. AIDS 2002;16:1925–1928. 14. Lundgren JD, Phillips AN, Vella S, et al. Regional differences in use of antiretroviral agents and primary prophylaxis in 3122 European HIV-infected patients. J Acquir Immune Defic Syndr 1997;16:153–160. 15. 1993 Revised Classification System for HIV infection and Expanded Surveillance Case Definition for AIDS among adolescents and adults. Morb Mort Wkly Rep 1992;41:19 –23. 16. Kirk O, Mocroft A, Katzenstein T, et al, for the EuroSIDA study group. Changes in use of antiretroviral therapy in regions of Europe over time. AIDS 1998,12:2031–2039. 17. Josefson D. Rate of decline in US AIDS cases is slowing. BMJ 2001;323:417. 18. Mocroft A, Ledergerber B, Katlama C, et al for the EuroSIDA Study Group. Continued decline in the AIDS and death rates in the EuroSIDA study: 1994 –2002. Lancet 2003;362:22–29. 19. Mocroft A, Phillips AN, Friis-Moller N, et al. Response to antiretroviral therapy among patients exposed to three classes of antiretrovirals: results from the EuroSIDA study. Antivir Ther 2002;7:21–30. 20. Ledergerber B, Egger M, Erard V, et al . AIDS-related opportunistic illnesses occurring after initiation of potent antiretroviral therapy: the Swiss HIV Cohort Study. JAMA 1999;15:282: 2220 –2226. Comment in JAMA 2000;283:2653–2654. 21. d’Arminio Monforte A, Testa T, Adorni F, et al. Clinical outcome and predictive factors of failure of highly active antiretroviral therapy in antiretroviral-experienced patients in advanced stages of HIV-1 infection. AIDS 1998;12:1631–1637. d’Arminio Monforte et al: CNS Diseases 327 22. Li TS, Tubiana R, Katlama C, et al. Long-lasting recovery in CD4 T-cell function and viral-load reduction after highly active antiretroviral therapy in advanced HIV-1 disease. Lancet 1998; 351:1682–1686. 23. Chowdhury M, Taylor JP, Chang CF, et al. Evidence that a sequence similar to TAR is important for induction of the JC virus late promoter by human immunodeficiency virus type 1 Tai J Virol 1992;66:7355–7361. 24. Pettoello-Mantovani M, Casadevall A, Kollmann TR, et al. Enhancement of HIV-1 infection by the capsular polysaccharide of Cryptococcus neoformans. Lancet 1999;339:21–23. 25. Lederman MM, Georges DL, Kusner DJ, et al. Mycobacterium tuberculosis and its purified protein derivative activate expression of the human immunodeficiency virus. J Acquir Immune Defic Syndr 1994;7:727–733. 26. Antinori A, Cingolani A, Lorenzini P, et al. Clinical epidemiology and survival of progressive multifocal leukoencephalopathy in the era of highly active antiretroviral therapy: data from the Italian Registry Investigative Neuro AIDS (IRINA). J Neurovirol 2003;9(suppl 1):47–53. 27. Cinque P, Bossolasco S, Boschini A, et al. The effect of HAART-induced immune reconstitution on development and outcome of progressive multifocal leukoencephalopathy. Study on 43 cases with review of the literature. J Neurovirol 2003; 9(suppl 1):73– 80. 328 Annals of Neurology Vol 55 No 3 March 2004 28. Pasquier RA, Koralnik IJ. Inflammatory reaction in progressive multifocal leukoencephalopathy: harmful or beneficial? J Neurovirol 2003;9(suppl 1):25–31. 29. Langford TD, Letendre SL, Marcotte TD, et al. Severe, demyelinating leukoencephalopathy in AIDS patients on antiretroviral therapy. AIDS 2002;16:1019 –1029. 30. Portegies P, Enting RH, de Gans J, et al. Presentation and course of AIDS dementia complex: ten years of follow-up in Amsterdam, the Netherlands. AIDS 1993;7:669 – 675. 31. Sidtis JJ, Gatsonis C, Price RW, et al. Zidovudine treatment of the AIDS dementia complex: results of a placebo-controlled trial. Ann Neurol 1993;33:343–349. 32. Vago L, Castagna A, Lazzarin A, et al. Reduced frequency of HIV-induced brain lesions in AIDS patients treated with zidovudine. J Acquir Immune Defic Syndr 1993;6:42– 45. 33. Price RW, Yiannoutsos CT, Clifford DB, et al. Neurological outcomes in late HIV infection: adverse impact of neurological impairment on survival and protective effect of antiviral therapy. AIDS Clinical Trial Group and Neurological AIDS Research Consortium study team. AIDS 1999;13:1677–1685. 34. Kirk O, Reiss P, Uberti-Foppa C, et al. Safe interruption of maintenance therapy against previous infection with four common HIV-associated opportunistic pathogens during potent antiretroviral therapy. Ann Intern Med 2002;137:239 –250.