The survival of people with HIV infection has been dramatically prolonged by the use of highly active antiretroviral therapies (HAARTs),1,2 although life expectancy of these people is still much lower as compared with the general population.3-5 Moreover, for groups at particularly high risk, such as injection drug users (IDUs), people with a high degree of immune suppression, or with certain AIDS-defining illnesses,6-8 the survival advantage conferred by HAART is less evident.
In Italy, the yearly number of AIDS cases has been steadily decreasing in the last decade, from 5052 in 1996 to 1489 cases in 2007.9 Because the vital status of people with AIDS (PWA) is not routinely updated through the Italian national surveillance system, unbiased, population-based, survival estimates are difficult to compute and few population-based surveys could evaluate the survival of PWA. One focused on all cases diagnosed between 1990 and 1998,8 whereas 2 other investigations used, in the late 1990s, regional data on PWA in Tuscany10 or in Lazio.11
This epidemiological study intended to provide an updated population-based quantification of the survival of PWA in the HAART era in Italy. Specific aims were to compare the survival of PWA with that of the general population, to identify factors present at diagnosis that affected survival and to assess changes in the frequency of AIDS-defining and non-AIDS-defining illnesses at death.
Since 1984, AIDS cases diagnosed in health facilities throughout Italy are compulsorily reported to the national AIDS registry (RNAIDS), a national surveillance system previously described in detail.8,10 According to the 1993 revised European AIDS definition, in Italy, AIDS diagnosis is based on clinical criteria only12 and up to 6 AIDS-defining illnesses can be recorded. In addition to AIDS-defining illnesses, other information is collected at RNAIDS. Among them, age, sex, education, area of residence, date of AIDS diagnosis and first HIV-positive test, HIV transmission category, and CD4+ cell count at AIDS diagnosis were used for the aims of this analysis. AIDS-defining illnesses were categorized into: primary brain lymphoma (PBL), immunoblastic non-Hodgkin lymphoma (NHL), Burkitt NHL, invasive cervical cancer (ICC), Kaposi sarcoma (KS), Pneumocystis carinii pneumonia (PCP), other opportunistic infections (OIs), and AIDS-associated conditions other than infections and cancer (ie, wasting syndrome and encephalopathies). Cases with more than 1 AIDS-defining illness were assigned to the group at worst prognosis.8 Underreporting of PWA at RNAIDS has been estimated at about 5%,13 whereas the vital status of PWA is not routinely kept up-to-date.
To overcome this drawback, the vital status of PWA was sought for in the Italian mortality database at the Italian National Institute of Statistics (NIS), where death certificates from all over Italy are centrally collected and electronically stored. The death certificate registration is mandatory (with completeness being around 100%), and it fits the European standards.14 Since January 1999, computerized death certificates also include name and surname. Record linkage was performed using an upgraded version of Software for Automated Linkage in Italy (SALI), a software application developed and validated in Italy to match individuals while protecting anonymity.15 SALI was used since 1998 to link cancer registries with the RNAIDS in Italy16 and with the Swiss HIV Cohort in Switzerland.17 Data regarding the 13,485 PWA diagnosed from 1999 through 2006 were linked with data concerning the 4,420,498 deaths occurred in Italy between January 1, 1999, and December 31, 2006. Specifically excluded from this analysis were PWA who were (1) non-Italian citizens, to limit false-negative results of the linkage due to non-Italian names and surnames; (2) residents in the provinces of Trento (n = 50) and Bolzano (n = 58) because names and surnames were not reported in death records; and (3) those diagnosed in 2006, to allow a minimum follow-up period of 12 months. AIDS cases diagnosed solely at autopsy and pediatric cases (ie, aged less than 13 years) were also excluded, leaving 9662 adult Italian PWA who constituted the study group.
For the aim of this analysis, all conditions listed in the death certificate were taken into consideration, whether or not they were the underlying cause of death. Accordingly, illnesses listed in the death certificate were categorized as AIDS-defining or non-AIDS-defining conditions according to the clinical definition of AIDS.12 Non-AIDS-defining conditions included cancers other than KS, NHL, ICC, and other illnesses (eg, infections not included in the AIDS-defining criteria, cardiovascular diseases, liver diseases, and traumatic conditions). Conditions at death were hierarchically considered in the following order: AIDS-defining cancers, other AIDS-defining conditions, non-AIDS-defining cancers, and other non-AIDS-defining conditions.
Survival time was calculated from date of AIDS diagnosis to date of death or to December 31, 2006, the censored date for unmatched PWA. The survival probability of PWA after AIDS diagnosis was estimated by means of the Kaplan-Meier method, and heterogeneity in survival among strata of selected variables was assessed through the log-rank test.18 The observed numbers of PWA who died within 12 months after AIDS diagnosis were compared with those expected from the general population of Italy. To this end, we used sex- and age-specific life tables published by the NIS for 2003 (http://demo.istat.it/unitav/index.html; accessed August 26, 2008), the year in the middle of this study period. The standardized mortality ratios (SMR) of observed to expected PWA, who died within 12 months after AIDS, and their 95% confidence intervals (CIs) were then computed.19
A multivariate Cox proportional hazards model was performed to identify factors present at AIDS diagnosis that were associated with the risk of death18 in the first 12 months after AIDS (ie, early mortality) or later (ie, late mortality). Multivariate hazard ratios (HRs), with their 95% CI, were computed after adjustment for factors that turned out to be statistically significant at univariate analysis (plus sex and area of residence).
The median follow-up time of the 9662 Italian PWA AIDS [males-median ages: 40 years, interquartile range (IQR): 36-47 years; females-median ages: 38 years, IQR: 33-42 years] included in this analysis was 41.0 months (IQR: 17-67 months) (data not shown). As of December 2006, 3111 deaths were recorded: 80.6% of PWA were alive after 1 year and 75.2% after 2 years, whereas 66.4% survived 5 years after AIDS diagnosis (Fig. 1). Survival was similar in males and females, and it was lower for IDUs than for other HIV transmission categories. A diagnosis of PBL conferred the shortest survival after AIDS (median survival, 4 months; 95% CI: 2 to 9), followed by immunoblastic (median survival, 16 months) and Burkitt NHL (median survival, 38 months). On the contrary, PWA with KS or PCP had the longest survival after AIDS, with 79.6% and 79.7% surviving at 5 years, respectively.
In the first year after AIDS diagnosis, the number of deaths among PWA by far exceeded the expected one. The computation of SMR indicated an overall 16.4-fold excess in the risk of death, with large variations according to sex (ie, 33.9-fold increase in women and 14.4-fold increase in men) (Fig. 2). Among women, observed to expected ratios ranged from 79.5, in those aged 20-34 years, to 10.6, in those aged 65-74 years, whereas among men, they ranged, respectively, from 31.8 to 4.0 (Fig. 2).
Table 1 shows the results of the multivariate analysis of factors associated with early or late mortality. Older PWA were at higher risk of death in both periods (χ2 for trend, P < 0.01), whereas more educated PWA turned out to be at reduced risk (χ2 for trend, P < 0.01) (Table 1). With regard to HIV transmission category, homosexual men and heterosexuals (HR = 0.6 in the late period) were at lower risks than IDUs. In both periods, the risk of death was higher for cases whose time span between first HIV-positive test and AIDS was ≥6 months (P < 0.01) compared with PWA unaware of being infected.
The risk of death was inversely related to the number of CD4+ cells at AIDS diagnosis. This association was more marked with early mortality (HR = 1.7 for <50 vs ≥200) and did not differ by time elapsed between first HIV-positive test and AIDS diagnosis (data not shown).
The number and the type of AIDS-defining illnesses at AIDS exerted a noteworthy influence on the risk of death (Table 2). With regard to early mortality, and in comparison with PWA diagnosed with PCP, we found a 9.2-fold higher risk (95% CI: 6.5 to 13.1) associated to PBL and a 4.5-fold higher risk associated with immunoblastic NHL (95% CI: 3.6 to 5.5) or Burkitt NHL (95% CI: 3.4 to 6.0) (Table 2). PWA diagnosed with OIs other than PCP had a modest elevation in risk (HR = 1.3), whereas those with HIV wasting syndrome or encephalopathies had a 2.4-fold higher risk of death (95% CI: 2.0 to 2.8). Such increase was more marked for progressive multifocal leukoencephalopathy (HR = 4.9, 95% CI: 4.0 to 6.2).
Immunoblastic NHL diagnosed at AIDS turned out to significantly increase the risk of late death also (HR = 2.4, 95% CI: 1.7 to 3.2), whereas Burkitt NHL and PBL did not (Table 2). Women with ICC at AIDS diagnosis were at a 3-fold higher risk of late death (HR = 2.9, 95% CI: 1.6 to 5.5), whereas their risk of early death was of borderline statistical significance. No difference in death risk emerged between cases with KS and those with PCP, both for early and late mortality (Table 2).
Information on illnesses present at death was available for 3075 of 3111 PWA (98.7%) who died during the study period. Overall, 45% of PWA had only non-AIDS-defining illnesses recorded at death. The frequency of non-AIDS-defining illnesses increased from 38.4% in 1999 to 56.9% in 2006 (χ2 for trend, P < 0.01) (Table 3). The increasing proportions of non-AIDS-defining deaths regarded both non-AIDS-defining cancers (from 3.7% in 1999 to 8.7% in 2006) (χ2 for trend, P = 0.02) and other non-AIDS-defining illnesses (from 34.7% to 48.2%) (χ2 for trend, P < 0.01) (Table 3).
Precise quantification of the survival after AIDS in the HAART era and identification of characteristics present at AIDS diagnosis offer a valuable piece of information for better understanding the natural history of HIV infection. The findings of this study indicate that nearly two thirds of Italian PWA were still alive 5 years after AIDS diagnosis, an observation that confirms survival improvements registered in industrialized countries after the introduction of HAART. According to a previous Italian study,8 66% of PWA diagnosed between 1995 and 1998 survived 2 years after AIDS, more than double a proportion of survivors recorded in the preceding quinquennia. In this study, 75% of PWA diagnosed in 1999-2005 were still alive 2 years after AIDS, with a relative increase in survival of 14%, compared with 1995-1998.8 A slight nonstatistically significant decrease in the risk of death was also noted within our study period (ie, 1999-2006), a suggestion that the positive effect of HAART on survival may have not reached its plateau yet. Notwithstanding the increasing survival, and in agreement with data from cohorts of HIV seroconverters,5 Italian PWA still experience higher rates of early post-AIDS death than expected (16-fold increased overall). SMR was noteworthily high among women (34-fold increased), pointing to the persisting heavy mortality burden due to HIV/AIDS in young adult women seen elsewhere.7,20
Several negative prognostic factors emerged from our analysis, namely, aging, injection drug use, and a lower than 200 CD4+ cells per cubic millimeter at diagnosis. In addition, duration of HIV infection is a well-established determinant of survival. Accordingly, a significant increased risk of death emerged for people known to be HIV infected for a long time before AIDS diagnosis. We have, however, shown that CD4+ cell count is a predictor of survival after AIDS, independently of the time elapsing between first HIV-positive test and AIDS.
Among the various characteristics present at AIDS diagnosis, AIDS-defining illnesses turned out to strongly affect survival, both in the early and in the late post-AIDS periods. A differential impact of AIDS-defining illness on the length of survival was already reported in both the pre- and post-HAART eras in Italy and in various international settings on people with HIV infection and AIDS.8,21 Apart from the unfavorable prognosis associated with the diagnosis of progressive multifocal leukoencephalopathy,22 the results of our analysis mainly point to a poor survival of PWA with NHL, mostly due to increased risks of death in the early post-AIDS period.
The diagnosis of PBL-recorded in less than 1% of AIDS cases-was confirmed to be the worst AIDS-defining illness in terms of survival.8,23 It has been already shown that HAART use has a different impact on morbidity and mortality of HIV-infected individuals,24 and many investigations have demonstrated that HIV-infected people treated with HAART have reduced incidence of NHL.25-27 HAART, on the other hand, seems to have decreased NHL incidence to a lesser extent than the incidence of KS,27,28 whereas no difference in survival for AIDS-associated NHL in pre-HAART and HAART eras has been noted.29 In this population-based investigation on PWA, the role of HAART on survival could not be assessed on an individual basis as history of antiretroviral treatment before AIDS was incomplete and no data were available on HAART use after AIDS.
In agreement with other studies that have shown improvements in survival of PWA with KS in industrialized countries,8,30 our study confirms a very long survival for PWA with KS, with 85% still alive after 2 years, in comparison with less than 40% recorded between 1995 and 1998.8 The reduced survival (56.5% at 5 years) of women in whom ICC was the AIDS-defining illness, as compared to women with ICC in the general population (ie, approximately 75%),31 provides additional evidence that cervical cancer screening of HIV-infected women in Italy needs substantial improvements in quality and coverage.32,33
All individuals included in this analysis were diagnosed with AIDS-defining illnesses,12 but only 55% of those who died during the study period had 1 or more AIDS-defining illnesses recorded in the death certificate. This observation mainly reflects the decreased impact of AIDS-associated OIs and-to a lesser extent-AIDS-defining illnesses other than cancer on mortality. This finding is in agreement with observations from similar studies, showing increasing proportions of non-AIDS diseases at death of PWA or people with HIV infection.34
Although some selections of study participants were necessary, completeness was the main advantage of this population-based investigation where 2 databases covering the whole Italian population were used.
In conclusion, this study documented the prolonged survival of Italian PWA from 1999 to 2006 and showed that the proportions of survivors several years after AIDS improved in comparison with those recorded in previous years. However, early mortality by far exceeded that seen in the general population, and NHL-although present in a small proportion of PWA-was still a strongly negative prognostic factor. Moreover, the frequency of non-AIDS-defining illnesses at death increased over the study period.
Drs. Frova L, Grippo F, Marchetti S, Pace M, Pappagallo M, Cinque S, Di Fraia G, and Pennazza S, from ISTAT, Roma, Italy; Dr. Susanna Conti from Istituto Superiore di Sanità, Roma, Italy; and Drs. Polesel J and Lise M from Centro di Riferimento Oncologico, Aviano, Italy, are greatly acknowledged for their contribution to the study. The authors thank Mrs Luigina Mei for editorial assistance.
1. Palella FJ Jr, Delaney KM, Moorman AC, et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. N Engl J Med
2. Detels R, Munoz A, McFarlane G, et al. Effectiveness of potent antiretroviral therapy on time to AIDS and death in men with known HIV infection duration. JAMA
3. Keiser O, Taffé P, Zwahlen M, et al. All cause mortality in the Swiss HIV cohort study from 1990 to 2001 in comparison with the Swiss population. AIDS
4. Wang C, Vlahov D, Galai N, et al. Mortality in HIV-seropositive versus seronegative persons in the era of highly active antiretroviral therapy: implications for when to initiate therapy. J Infect Dis
5. Bhaskaran K, Hamouda O, Sannes M, et al. Changes in the risk of death after HIV seroconversion compared with mortality in the general population. JAMA
6. Wood E, Hogg RS, Dias Lima V, et al. Highly active antiretroviral therapy and survival in HIV-infected injection drug users. JAMA
7. Poundstone KE, Chaisson RE, Moore RD. Differences in HIV disease progression by injection drug use and by sex in the era of highly active antiretroviral therapy. AIDS
8. Conti S, Masocco M, Pezzotti P, et al. Differential impact of combined antiretroviral therapy on the survival of Italian patients with specific AIDS-defining illnesses. J Acquir Immune Defic Syndr
9. Suligoi B, Boros S, Camoni L, et al. Aggiornamento dei casi di AIDS notificati in Italia e delle nuove diagnosi di infezione da HIV
. Roma, Italy: Istituto Superiore di Sanità; 2007.
10. Pezzotti P, Napoli PA, Acciai S, et al. Increasing survival time after AIDS in Italy: the role of new combination antiretroviral therapies. AIDS
11. Porta D, Rapiti E, Forastiere F, et al. Changes in survival among people with AIDS in Lazio, Italy, from 1993 to 1998. AIDS
12. Ancelle Park RA. Expanded European AIDS case definition. Lancet
13. Conti S, Farchi G, Galletti A, et al. La notifica della mortalità per AIDS in Italia (1992):qualità della certificazione e sottonotifica. G Ital AIDS
14. Istat-Istituto Italiano di Statistica. Death certificate and certification practices: an international comparison. Essays
. 2004, no. 14.
15. Dal Maso L, Braga C, Franceschi S. Methodology used for “software for automated linkage in Italy” (SALI). J Biomed Inform
16. Franceschi S, Dal Maso L, Arniani S, et al. Risk of cancer other than Kaposi's sarcoma and non-Hodgkin's lymphoma in persons with AIDS in Italy. Cancer and AIDS Registry Linkage Study. Br J Cancer
17. Clifford GM, Polesel J, Rickenbach M, et al. Cancer risk in the Swiss HIV Cohort Study: associations with immunodeficiency, smoking, and highly active antiretroviral therapy. J Natl Cancer Inst
18. Armitage P, Berry G, Matthews JNS. Statistical Methods in Medical Research
. 4th ed. Malden, MA: Blackwell Science; 2002.
19. Breslow NE, Day NE. Statistical Methods in Cancer Research. Volume II-The Design and Analysis of Cohort Studies
. Lyon, France: International agency for research on cancer; 1987. IARC Scientific Publication no. 82.
20. Hessol NA, Kalinowski A, Benning L, et al. Mortality among participants in the Multicenter AIDS Cohort Study and the Women's Interagency HIV Study. Clin Infect Dis
21. CASCADE Collaboration. Effective therapy has altered the spectrum of cause-specific mortality following HIV seroconversion. AIDS
22. Lewden C, May T, Rosenthal E, et al. Changes in causes of death among adults infected by HIV between 2000 and 2005: the “Mortalité 2000 and 2005” Survey (ANRS EN19 and Mortavic). J Acquir Immune Defic Syndr
23. Dal Maso L, Rezza G, Zambon P, et al. Non-Hodgkin lymphoma among young adults with and without AIDS in Italy. Int J Cancer
24. International Collaboration on HIV and Cancer. Highly active antiretroviral therapy and incidence of cancer in human immunodeficiency virus-infected adults. J Natl Cancer Inst
. 2000;92:1823- 1830.
25. Carrieri MP, Pradier C, Piselli P, et al. Reduced incidence of Kaposi's sarcoma and of systemic non-Hodgkin's lymphoma in HIV-infected individuals treated with highly active antiretroviral therapy [letter]. Int J Cancer
26. Serraino D, Piselli P, Busnach G, et al. Risk of cancer following immunosuppression in organ transplant recipients and HIV-positive individuals in southern Europe. Eur J Cancer
27. Polesel J, Clifford GM, Rickenbach M, et al. Non-Hodgkin lymphoma incidence in the Swiss HIV Cohort Study before an after highly active antiretroviral therapy. AIDS
28. Franceschi S, Dal Maso L, Rickenbach M, et al. Kaposi sarcoma incidence in the Swiss HIV Cohort Study before and after highly active antiretroviral therapy. Br J Cancer
29. Matthews GV, Bower M, Mandalia S, et al. Changes in acquired immunodeficiency syndrome-related lymphoma since the introduction of highly active antiretroviral therapy. Blood
30. Biggar RJ, Engels EA, Ly S, et al. Survival after cancer diagnosis in persons with AIDS. J Acquir Immune Defic Syndr
31. AIRTUM Working Group. I tumori in Italia, rapporto 2007: Sopravvivenza. Epidemiol Prev
. 2007;31(Suppl 1):58.
32. Franceschi S, Dal Maso L, Suligoi B, et al. Evidence for lack of cervical cancer screening among HIV-positive women in Italy [letter]. Eur J Cancer
33. Serraino D, Pavia C, Navarra A, et al. Knowledge and use of Papanicolaou test among HIV-positive women [letter]. Int J Cancer
34. Palella FJ, Baker RK, Moorman AC, et al. Mortality in the highly active antiretroviral therapy era. Changing causes of death and disease in the HIV Outpatient Study. J Acquir Immune Defic Syndr