It is no longer believed that highly active antiretroviral therapy (HAART) can eradicate HIV within a few years [1,2], and the side-effects of lifelong treatment with antiretroviral drugs are well known [3,4]. Consequently, the issue of the optimal time to initiate HAART in chronically HIV-infected individuals is very relevant [5–12]. Ideally this should be addressed in a randomized controlled trial, but there are currently no ongoing trials considering this issue. Observational cohort studies of HIV-infected individuals are an alternative, not ideal, source of data. In general, since the introduction of HAART in 1996, few analyses of large HIV databases have been performed to examine the effect of an early start for HAART and most of the reasons put forward are based on theoretical considerations or indirect evidence from small studies. The main arguments for starting HAART in HIV-infected patients at the time of diagnosis are (i) treatment of an infectious diease ; (ii) to reduce the risk of opportunistic infections and constitutional symptoms in patients whose CD4 cell count is 200–500 × 106 cells/l, with a consequent improvement of the quality of life [13,14]; (ii) to preserve HIV-specific cellular responses, which seem to be higher if therapy is started when the CD4 cell count is relatively high [15,16]; and (iv) to reduce disease progression with its accompanying increase in HIV diversity and virulence . Some studies of previously treated and untreated patients have shown that the virological response to HAART is better in patients starting therapy at higher CD4 cell counts [18,19]. The arguments in favour of delaying the initiation of therapy are that (i) most regimens are difficult to tolerate; (ii) non-adherence may lead to the development of HIV resistance, which limits future therapy options [20,21]; (iii) considerable reconstitution of the immune system seems possible even in patients who start HAART at very low CD4 cell counts ; (iv) while the short- and medium-term side-effects of the current regimens are known, the long-term toxicity is not ; and (v) new therapies/strategies that are likely to be more effective in drug-naive patients may be available in the future [11,12].
The Italian cohort of patients naive to antiretrovirals (ICONA) is one of the largest European studies and includes patients who were antiretroviral naive at the time of enrolment and who subsequently started HAART. The database has been used to analyse factors that might be relevant to define the best time to start HAART.
Subjects and methods
Subjects of this analysis were enrolled in ICONA between May 1997 and April 1998, started antiretroviral treatment for the first time with at least three drugs simultaneously (HAART) and had a CD4 cell count available from less than 24 weeks prior to starting therapy. This analysis was performed using all the data recorded up to 30 June 2000. ICONA is a multicentre prospective cohort study of patients attending 65 infectious disease wards in Italy. All patients attended their local infectious disease ward approximately every 3 months. Informed written consent was obtained from all participants at enrolment. Demographic, clinical, immunological, virological and therapeutic characteristics were collected at enrolment and during the follow-up. Details of the study design and of data collection have been presented elsewhere . Plasma HIV RNA was measured using quantitative reverse transcriptase polymerase chain reaction (RT-PCR; Amplicor, Roche Molecular System, Pleasanton, California, USA), signal amplification branched DNA assay (Quantiplex, Chiron, Emeryville, California, USA) or nucleic acid sequence-based amplification (NASBA Organon Teknika, Boxtel, the Netherlands). The lower limit of detection of these assays was 500 copies/ml. Ultrasensitive versions (with a lower limit of detection of 50 copies/ml) were used when appropriate, starting from May 1998. CD4 cell counts were performed using standard flow cytometery techniques. No attempt was made to standardize viral loads measured with different assays. However, the analysis was repeated after adjusting the viral load levels measured using the branched DNA assay by multiplying by 2 as there is some evidence that the RT-PCR assay gives higher values .
Immunological response according to baseline CD4 cell count
Mean CD4 cell count changes were calculated every 3 months until 96 weeks from the date of initiation of HAART. A window of 3 months was chosen around a time point from which the closest estimate was taken for the CD4 cell count at that time point. For example, a window between week 6 and week 18 was chosen for the CD4 cell count at 12 weeks and the measurement closest to week 12 was used. If two measurements were equally as close, the later of the two was used. Patients were grouped according to their pre-HAART mean CD4 cell count into three groups: < 200, 201–350 and > 350 × 106 cells/l. This analysis was only descriptive, no statistical test was performed.
Virological response according to baseline CD4 cell count
Standard survival analysis using Kaplan–Meier curves, log-rank test, and Cox proportional hazards regression model were used to assess the virological response to therapy according to the pre-HAART CD4 cell count group and viral load. The proportional hazards Cox regression model was stratified by the centre. The significance of the parameters was tested by comparing the parameter estimate with its standard error and referring to a χ2 (1 degree of freedom) distribution (Wald test). Patients were defined as virological failures if they did not achieve a viral load ≤ 500 copies/ml by 32 weeks after starting HAART or if they did achieve this load by 32 weeks but had a viral load that subsequently rebounded to > 500 copies/ml on two consecutive occasions. Patients who had been followed for at least 32 weeks and whose viral load remained at > 500 copies/ml were defined as failures at week 32; patients whose viral load was suppressed before week 32 but subsequently rebounded on two occasions were defined as failures at the time of the first viral load > 500 copies/ml. Follow-up was censored at the date of the last viral load measurement for patients whose viral load was suppressed before week 32 and for whom suppression was sustained and for patients who did not achieve viral suppression and their last viral load value was measured before week 32. For a more sensitive analysis, patients whose last viral load was measured before 30 June 1999, and, therefore, for whom no data were available in the final year of the study, were defined as virological failure at the date of last viral load. Analyses of time to virological failure were performed according to intention to treat (e.g., all patients were included and therapy switches were ignored).
Incidence of new AIDS events or death in patients starting HAART with a low CD4 cell count
A subset of patients started HAART with a CD4 cell count < 50 × 106 cells/l. The analysis of this set assessed whether CD4 cell count increases as a result of HAART were associated with reduced risk of AIDS and death. Starting from the date of initiation of HAART the person-years were calculated for which each patient was observed with a CD4 cell count < 50, 50–200 and > 200 × 106 cells/l. The number of new AIDS events (only AIDS-defining illnesses that were never experienced before) or death was also calculated and related to the most recent CD4 cell count. Incidence rates of AIDS or death in the three CD4 cell count categories were then calculated as number of events per person-years. Note that the same patient could contribute to several CD4 cell count groups in this analysis as his/her CD4 cell count was modified by HAART. However, none of the patients developed more than one AIDS-defining diseases so confidence intervals for the rates were calculated using the standard Poisson distribution.
The analysis includes 1421 patients who started HAART and had a CD4 cell count measured less than 24 weeks prior to initiation of therapy; all patients had at least one measure of viral load in addition to the pre-HAART viral load. The characteristics of all the patients are given in Table 1. Patients were all previously antiretroviral naive with median CD4 cell count of 272 × 106 cells/l (range, 1–1294) and a median plasma viral load of 4.84 log10 copies/ml (range, 2.70–6.72) when they started their first HAART regimen. The majority of patients (n = 1239; 87.2%) started two nucleoside reverse transcriptase inhibitors and one protease inhibitor while a smaller proportion initiated two nucleoside reverse transcriptase inhibitors and one non-nucleoside reverse transcriptase inhibitors (n = 154; 10.8%). Characteristics in the three groups were similar. However, patients with a pre-HAART CD4 cell count > 200 × 106 cells/l were more likely to be treated with two nucleoside reverse transcriptase inhibitors and one non-nucleoside reverse transcriptase inhibitor (P = 0.001), had lower viral load (P = 0.001) and were less likely to be symptomatic (P = 0.001) compared with the whole study population (Table 1).
Immunological response according to baseline CD4 cell count
Patients were followed for a median of 81 weeks (25–75% percentiles 48–107) from the date of initiation of HAART. Median duration of follow-up was similar in patients in the three groups who started HAART: CD4 cell count ≤ 200 (82 weeks), 201–350 (77 weeks) and > 350 × 106 cells/l (73 weeks). Figure 1 shows the 12-weekly crude mean increase in CD4 cell count over 24 months of therapy according to the level of pre-HAART CD4 cell count. There was no apparent relationship between the rise in CD4 cell counts and the starting level. On average, CD4 cell counts had increased by at least 180 × 106 cells/l at week 96, irrespective of the pre-HAART CD4 cell count level. Importantly, the absolute rise in CD4 cell count appeared to be greater in patients with a lower starting level (e.g., 281 × 106 cells/l in patients starting with 201–350 × 106 cells/l and 186 × 106 cells/l in those starting with > 350 × 106 cells/l;Fig. 1). Follow up for at least 90 weeks was achieved for 617 patients (43.4%) and a measure of CD4 cell count was available at 90–102 weeks for 419 (67.9%). Overall in these patients, there was a mean change in CD4 cell count of 228 × 106 cells/l (range, −548–881) by 96 weeks of starting HAART and only a minority (n = 32; 9.2%) had a CD4 cell count < 200 × 106 cells/l at this time: 29 of 154 (18.8%) had started HAART at ≤ 200 × 106 cells/l; 1 of 83 (1.2%) started at 201–350 × 106 cells/l; and 2 of 111 (1.8%) started at > 350 × 106 cells/l CD4 cell count.
Virological response according to baseline CD4 cell count
Viral response to HAART in patients starting therapy at different CD4 cell counts was assessed using survival analysis and defining virological failure as failure to suppress viral load to < 500 copies/ml by 32 weeks or viral rebound on at least two occasions. Virological failure occurred in 536 (37.7%) (Fig. 2): 238 of 558 (42.7%) whose pre-HAART CD4 cell count was ≤ 200 × 106 cells/l, 110 of 312 (35.3%) at 200–349 × 106 cells/l and 188 of 551 (34.1%) at ≥ 350 × 106 cells/l. Figure 2 shows the Kaplan–Meier estimates of the probability of virological failure in these three groups. By 96 weeks, the probability of virological failure was 51.0% [95% confidence interval (CI), 46.4–55.6], 43.8% (95% CI, 37.6–50.4) and 44.0% (95% CI, 39.0–49.0) for patients starting HAART with a CD4 cell count of ≤ 200, 201–350 > 350 × 106 cells/l, respectively. The log-rank test comparing the three survival curves was statistically significant (log-rank χ-square 7.8;P = 0.02). A direct comparison of patients in the 201–350 and > 350 × 106 cells/l groups using the Cox proportional hazard regression model adjusted for pre-HAART viral load (fitted as categorical with cut-offs 5000 and 30 000 copies/ml) showed no difference in risk [adjusted relative hazard (RH) 1.01; 95% CI, 0.79–1.29;P = 0.93;Table 2]. Patients in the lowest CD4 cell group (≤ 200 × 106 cells/l) did show an increased hazard of virological failure compared with patients in the highest starting CD4 cell count group (< 350 × 106 cells/l): adjusted RH 1.15 (95% CI, 0.93–1.42;P = 0.20;Table 2). The risk of failure in patients with a viral load > 30 000 copies/ml was 30% higher than that in patients with viral load ≤ 5000 copies/ml although the association was not significant (P = 0.19;Table 2) even when using > 10 000 copies/ml as the higher group (data not shown). Pre-HAART viral load (fitted as categorical) was not independently associated with the risk of virological failure. Similar results were obtained if the viral loads measured using a branched DNA assay were multiplied by a factor of 2 and if the 234 patients (16.5%) who were lost to follow-up were defined as virological failures at the date of their last viral load measurement.
Incidence of new AIDS events or death in patients starting HAART with a low CD4 cell count
The CD4 cell count restored by HAART was evaluated as a prognostic indicator for a new AIDS event or death. Figure 3 shows the person-years at risk, number of AIDS events or deaths, and the event rate according to the most recent CD4 cell count during HAART therapy in 218 patients (15.3% of the total) starting HAART with a CD4 cell count < 50 × 106 cells/l. While there was a relatively high rate of AIDS or death during the time in which the patients’ CD4 cell count remained < 50 × 106 cells/l (0.19/person-year; 95% CI, 0.10–0.33; 13 events in 67 person-years), the rate was very small when the CD4 cell count increased to 50–199 × 106 cells/l (0.05/person-year; 95% CI, 0.02–0.10; 9 events in 175 person-years) and > 200 × 106 cells/l (0.03/person-year; 95% CI, 0.005-0.07; 5 events in 144 person-years) (Fig. 3).
When to start antiretroviral therapy in asymptomatic HIV-infected individuals remains a crucial question [5–12]. Ideally, this question should be studied in a randomized controlled trial as in non-randomized studies the decision to initiate antiretroviral therapy may be influenced by factors such as the patients’ potential adherence to therapy or preferences . Nonetheless, the ICONA study provides some evidence to help clinicians and patients to decide when to initiate therapy. Our data showed that there was a similar rise in CD4 cell count in patients starting HAART at a high CD4 cell count compared with patients starting at a lower level. After 24 months of therapy, patients experienced a very similar CD4 cell count rise irrespective of their level of pre-HAART CD4 cell count (i.e. 180 × 106 cells/l). The immune system has much redundancy and reserve and, consequently, dramatic perturbations seem necessary to place a person at risk of AIDS or death. In the natural history of HIV infection such a risk is low in individuals with a CD4 cell count > 200 × 106 cells/l, or even > 100 × 106 cells/l . There is evidence that this is still true even if this count has been restored by HAART [22,26]. In ICONA, patients who started HAART with CD4 cell count of < 50 × 106 cells/l and had a CD4 cell count rise to > 200 × 106 cells/l had an extremely low risk of subsequent occurrence of an AIDS-defining illness or death [5 events per 144 person-years]. This was also observed in a similar large clinical observational database . The majority of our patients (90.8%) for whom results were available after 96 weeks had achieved > 200 × 106 cells/l by 2 years of therapy irrespective of their pre-HAART CD4 cell count. There was no evidence that starting therapy at > 350 × 106 cells/l instead of at 201–350 × 106 cells/l offered any increased immunological benefit at 2 years after starting HAART.
The rate of virological failure by 96 weeks of therapy was significantly higher in patients starting with a CD4 cell count ≤ 200 × 106 cells/l (Fig. 2) but there was no difference in the Kaplan–Meier estimate of virological failure in patients starting at 201–350 or at > 350 × 106 cells/l (Fig. 2). This was confirmed by a Cox regression analysis, which was also adjusted for pre-HAART viral load levels (Table 2). In our study, pretherapy viral load was not associated with the risk of virological failure (Table 2). This is in conflict with other previous reports [14,27]. However, these other studies were not restricted to antiretroviral-naive patients and their definition of the virological endpoint and the statistical modelling used were different.
There are a number of limitations to this study. In an observational setting, it is possible that patients starting therapy at a low CD4 cell count are intrinsically different from those starting at a higher value. For example, the former might be perceived by medical staff to be less likely to adhere to therapy . Unfortunately, adherence to treatment was assessed (via self-reported questionnaire) only on a subset of enrolled patients and no data are available as yet. Additionally, the quality of life of patients after starting HAART was not assessed. Viral load was assayed using different methods and no attempt was made to standardise these values. Since there is evidence that RT-PCR measurements for a specimen are about twofold higher than those found by the branched DNA assay  this could be source of bias. However, our results were not changed if a correction was applied for this (data not shown). The Cox regression analysis was stratified by centre and these tended to use the same assay for all their samples. Similarly, the ‘time to virological failure’ analysis could be affected by how patients who are lost to follow-up were handled. Again, the conclusions were unchanged when the survival analysis defined patients lost to follow-up as failures.
Our study does not directly address the question of what is the risk of developing an AIDS-defining illness for a patient who decides to delay the initiation of therapy until his/her CD4 cell count has fallen to < 200 × 106 cells/l compared with another patient who decides to start HAART at approximately 350 × 106 cells/l. This is an important point as developing AIDS-defining illnesses is associated with poor quality of life and, for example in lymphomas, with uncertainty regarding the possibility of successfully treatment. Unfortunately, ICONA currently lacks the power for this analysis and it will occur when longer follow-up and a larger number of clinical events are available.
Three recent report meetings addressed the issue of when it would be the optimal time to start HAART in chronic HIV infection [28–30]. One study showed that a higher proportion of patients achieved a viral load ≤ 50 copies/ml if they started HAART at a CD4 cell count ≥ 500 than if they waited until the CD4 count was < 400 × 106 cells/l . A second study showed evidence for more durable undetectable viral load (< 400 copies/ml) in patients started with a CD4 cell count > 350 × 106 cells/l compared with patients starting with < 200 × 106 cells/l; comparison with patients starting at 200–350 × 106 cells/l was not performed . A third study, based on mathematical modelling instead of real data, indicated that immediate initiation of therapy would be recommendable only if this strategy would guarantee a minimum reduction of 15–20% in the rate of virological failure compared with a wide range of strategies implying a delay of therapy . In our study, the difference in the rate of failure at 96 weeks between patients starting at a CD4 cell count of 201–350 and one of > 350 × 106 cells/l was less than 1%.
Although our analysis of an observational clinical database cannot provide categorical answers, it does indicate that patients starting HAART with a CD4 cell count ≤ 200 × 106 cells/l may experience a worse virological response to HAART than those starting with > 350 × 106 cells/l. There is no clear advantage in terms of subsequent virological and immunological response in starting HAART at a CD4 cell count > 350 × 106 cells/l rather than at 201–350 × 106 cells/l. However, further studies are needed to assess whether it is safe to defer HAART until the CD4 cell count reaches 200 × 106 cells/l, as opposed to 350 × 106 cells/l, as currently suggested in the UK HIV treatment guidelines .
1. Finzi D, Blankson J, Siliciano JD. et al
. Latent infection of CD4+T cells provides a mechanism for lifelong persistence of HIV-1 even in patients on effective combination therapy. Nat Med 1999, 5: 512 –517.
2. Furtado MR, Callaway DS, Phair JP. et al
. Persistence of HIV-1 transcription in peripheral-blood mononuclear cells in patients receiving potent antiretroviral therapy. N Engl J Med 1999, 340: 1614 –1622.
3. Carr A, Samaras K, Burton S. et al
. A syndrome of peripheral lipodistrophy, hyperlipidaemia and insulin resistance in patients receiving HIV protease inhibitors. AIDS 1998, 12: F51 –F58.
4. Carr A, Miller J, Law M, Cooper DA. A syndrome of lipoatrophy, lactic acidaemia and liver dysfunction associated with HIV nucleoside analogue therapy: contribution to protease inhibitor-related lipodystrophy syndrome. AIDS 2000, 14: F25 –F32.
5. Bassetti S, Battegay M, Furrer H. et al
. Why is highly active antiretroviral therapy
(HAART) not prescribed or discontinued? Swiss HIV Cohort Study. J Acquir Immune Defic Syndr Hum Retrovirol 1999, 21: 114 –119.
6. Ho DD. Time to hit HIV, early and hard. N Engl J Med 1995, 333: 450 –451.
7. Walker BD, Basgoz N. Treat HIV-1 infection like other infections – treat it. JAMA 1998, 280: 91 –93.
8. Feinberg M. Hidden dangers of incompletely suppressive antiretroviral therapy. Lancet 1997, 349: 1408 –1409.
9. Burman WJ, Reeves RR, Cohn DL. The case for conservative management of early HIV disease. JAMA 1998, 280: 93 –95.
10. Buckingham W, Gebbie KM. Treatment policy: gold standard or golden calf? AIDS Reader 1998, 8: 160 –215.
11. Volberding P. When and where to start: guidelines for the initiation of antiretroviral therapy. AIDS Reader 2000, 10: 150 –155.
12. Deeks S. Optimizing antiretroviral therapy: strategies and regimens.Part 1: antiretroviral strategies.HIV/AIDS Clinical Management
13. Ledergerber B, Egger M, Opravil M. et al
. Clinical progression and virological failure on highly active antiretroviral therapy
in HIV-1 patients: a prospective cohort study. Lancet 1999, 353: 863 –868.
14. Ledergerber B, Egger M, Erard V. et al
. AIDS-related opportunistic illnesses occurring after initiation of potent antiretroviral therapy. JAMA 1999, 282: 2220 –2226.
15. Appay V, Nixon DF, Donahoe SM. et al
. HIV-specific CD8+ cells produce antiviral cytokines but are impaired in cytolytic function. J Exp Med 2000, 192: 63 –75.
16. Altfeld M, Rosenberg ES, Eldridge RL et al.Increase in breadth and frequency of CTL responses following structured therapy interruptions in individuals treated with HAART during acute HIV-infection.Seventh Conference on Retroviruses and Opportunistic Infections.
San Francisco, January–February 2000 [abstract 357].
17. Bonhoeffer S. Models of viral kinetics and drug resistance in HIV-1 infection. AIDS Patient Care STD 1998, 12: 769 –774.
18. Miller V, Staszewski S, Sabin CA. et al
. CD4 lymphocyte count as a predictor of the duration of highly active antiretroviral therapy
-induced suppression of human immunodeficiency virus load. J Infect Dis 1999, 180: 530 –533.
19. Deeks S, Hecht FM, Swanson M. et al
. HIV RNA and CD4 cell count
response to protease inhibitor therapy in an urban AIDS clinic: response to both initial and salvage therapy. AIDS 1999, 13: F35 –F43.
20. Molla A, Korneveva M, Gao Q. et al
. Ordered accumulation of mutations in HIV protease confers resistance to ritonavir. Nat Med 1996, 2: 760 –766.
21. Condra JH, Schleif WA, Blahv OM. et al
. In vivo
emergence of HIV-1 variants resistant to multiple protease inhibitors. Nature 1995, 374: 569 –571.
22. Miller V, Staszewski S, Nisius G. et al
. Risk of new AIDS diseases in people on triple therapy. Lancet 1999, 353: 463. 463.
23. d'Arminio Monforte A, Cozzi Lepri A, Rezza G. et al
. Insights into the reasons for discontinuation of the first highly active antiretroviral therapy
(HAART) regimen in a cohort of antiretroviral naive patients. AIDS 2000, 14: 499 –507.
24. Nolte FS, Boysza J, Thurmond C, Clark WS, Lennox JL. Clinical comparison of an enhanced-sensitivity branched-DNA assay and reverse transcription-PCR for quantitation of human immunodeficiency virus type 1 RNA in plasma. J Clin Microbiol 1998, 36: 716 –720.
25. Cozzi Lepri A, Pezzotti P, Dorrucci M. et al
. HIV disease progression in 854 women and men infected through injecting drug use and heterosexual sex and followed for up to nine years from seroconversion. BMJ 1994, 309: 1537 –1542.
26. El-Sadr WM, Burman WJ, Bjorling Grant L. et al
. Discontinuation of prophylaxis against Mycobacterium avium
complex disease in HIV-infected patients who have a response to antiretroviral therapy. N Engl Med 2000, 342: 1085 –1092.
27. Grabar S, Pradier C, Le Corfec E. et al
. Factors associated with clinical and virological failure in patients receiving a triple therapy including a protease inhibitor. AIDS 2000, 14: 141 –149.
28. Meibohm A, Alexander JR J, Robertson M et al.Early versus late initiation of indinavir (IDV) in combination with zidovudine (ZDV) and lamivudine (3TC) in HIV-infected individuals.Seventh Conference on Retroviruses and Opportunistic Infections.
San Francisco, January–February 2000 [abstract 521].
29. Moore R, Keruly J, Bartlett J, and Chaisson R. Start HAART early (CD4 < 350 cells/mm3) or later? Evidence for greater effectiveness if started early.Seventh Conference on Retroviruses and Opportunistic Infections.
San Francisco, January–February 2000 [abstract 522].
30. Tebas P, Henry K, Nease R et al.When should we start? Use of Markov modelling and decision analysis to evaluate the long-term implications of antiretroviral therapy.Seventh Conference on Retroviruses and Opportunistic Infections.
San Francisco, January–February 2000 [abstract 523].
31. British HIV Association (BHIVA) guidelines for the treatment of HIV-infected adults with antiretroviral therapy.HIV Medicine
I.CO.N.A. study group in Italy. Ancona: M. Montroni, G. Scalise, A. Costantini, M. S. Del Prete; Aviano (PN): U. Tirelli, G. Nasti; Bari: G. Angarano (scientific committee), L. M. Perulli; Bergamo: F. Suter, C. Arici; Bologna: F. Chiodo (scientific committee), F. M. Gritti, V. Colangeli, C. Fiorini, L. Guerra; Brescia: G. Carosi (scientific committee), G. P. Cadeo, F. Castelli, C. Minardi, D. Vangi; Busto Arsizio: S. Caprioli, G. Migliorino; Cagliari: P. E. Manconi, P. Piano; Catanzaro: T. Ferraro, L. Cosco; Chieti: E. Pizzigallo, F. Ricci; Como: G. M. Vigevani, L. Pusterla; Cremona: G. Carnevale, A. Pan; Cuggiono: P. Viganò, G. C. Ghiselli; Ferrara: F. Ghinelli, L. Sighinolfi; Florence: F. Leoncini, F. Mazzotta, S. Ambu, S. Lo Caputo; Foggia: B. Grisorio, S. Ferrara; Galatina (LE): P. Grima, P. Tundo; Genoa: G. Pagano, N. Piersantelli, A. Alessandrini, R. Piscopo; Grosseto: M. Toti, Chigiotti; Latina: F. Soscia, L. Tacconi; Lecco: A. Orani, G. Castaldo; Lucca: A. Scasso, A. Vincenti; Mantova: A. Scalzini, F. Alessi; Milan: M. Moroni (study coordinator), A. Lazzarin (scientific committee), A. Cargnel, F. Milazzo, L. Caggese, A. d'Arminio Monforte, V. Testori, F. Delfanti, B. Carini, B. Adriani, S. Garavaglia, C. Moioli; Modena: R. Esposito, C. Mussini; Naples: N. Abrescia, A. Chirianni, O. Perrella, M. Piazza, M. de Marco, V. Montesarchio, E. Manzillo, S. Nappa; Padua: P. Cadrobbi, R. Scaggiante Palermo: A. Colomba, T. Prestileo; Pavia: G. Filice, L. Minoli, F. A. Patruno Savino, R. Maserati; Perugia: S. Pauluzzi, A. Tosti; Piacenza: F. Alberici, M. Sisti; Pisa: F. Menichetti, A. Smorfa; Potenza: C. de Stefano, A. La Gala; Ravenna: T. Zauli, G. Ballardini; Reggio Emilia: L. Bonazzi, M. A. Ursitti; Rimini: R. Ciammarughi, M. Arlotti; Rome: L. Ortona (scientific committee), F. Dianzani (scientific committee), A. Antinori, G. Antonucci, S. D'Elia, G. Ippolito, P. Narciso, N. Petrosillo, G. Rezza, V. Vullo, A. de Luca, A. Del Forno, M. R. Capobianchi, M. Zaccarelli, P. de Longis, M. Ciardi, E. Girardi, G. D'Offizi, F. Palmieri, P. Pezzotti, M. Lichter; Sassari: M. S. Mura, M. Mannazzu; Turin: P. Caramello, A. Sinicco, M. L. Soranzo, D. Giacobbi, M. Sciandra, B. Salassa; Varese: D. Torre; Verbania: A. Poggio, G. Bottari; Venice: E. Raise, S. Pasquinucci; Vicenza: F. de Lalla, G. Tositti; Taranto: F. Resta, A. Chimienti.