During the early days of the AIDS epidemic, results of autopsy findings indicated that death in the majority of AIDS patients was attributable to opportunistic infections affecting the respiratory tract and nervous system [1–4] Through the year 1986  75% of autopsies on Aids patients showed single infections, particularly Pneumocystis carinii pneumonia (PCP); however, from 1987, 72% of autopsies presented with multiple infectious agents including PCP, Mycobacterium avium complex (MAC), cytomegalovirus (CMV), and various fungal agents  Studies focusing on determining causes of death and distribution of organ involvement in AIDS following the era of antiretroviral (e.g., zidovudine) and prophylactic anti-opportunistic infection treatment showed that although PCP remained the most common cause of death, its frequency and number of deaths declined over time [5,6] Deaths from bacterial sepsis, CMV, MAC infection and toxoplasmosis also declined during this period, but mortality from fungal infections, tuberculosis, encephalopathy, and causes unrelated to AIDS increased  The death rate from non-Hogdkin's lymphoma (NHL) remained high and Kaposi's sarcoma (KS) continued to present more frequently in patients whose risk factor for HIV infection was homosexuality or bisexuality, nonetheless the death rate from KS was greatest for patients with a risk factor of blood exposure to HIV  Survival was shorter and deaths from tuberculosis more common in patients with a history of injecting drug use  Overall, survival for patients in other AIDS risk groups increased over time, particularly in those treated with antiretroviral therapy  The distribution of organ involvement in major opportunistic infections and neoplasms has remained virtually the same throughout the years of the study 
Taken together, these studies have indicated that although effective prevention and treatment has helped to ameliorate the development of certain AIDS-related conditions, it also allowed other illnesses to quickly become the cause of death. Supporting this notion, recent studies have shown an increased incidence of HIV-induced brain lesions in AIDS patients with long-term survival  This study showed a 40% incidence of HIV encephalitis during the first years of the epidemic, although survival was short during this period  Although the incidence of HIV encephalitis fell markedly around the time zidovudine was introduced and remained low in patients using zidovudine until death [7–9] the rate of HIV encephalitis has increased during the later years, the new cases occurring mainly in patients who had discontinued zidovudine use [7–9] In addition, variations in the frequency of AIDS-related central nervous system (CNS) pathology might be associated with appearance of drug-resistant pathogens, injecting drug use and with severity of the systemic disease [7–9] and the introduction in 1996–1997 of new antiretroviral treatments based on the use of protease inhibitors  In this context, the main objectives of the present study were to analyze changes in frequency of systemic AIDS pathology over time and its relationship with variations in CNS pathology.
Materials and methods
We reviewed autopsy reports from 390 AIDS patients who underwent examination at University of California at San Diego/Medical Center (UCSD-MC) during the period of 1982–1998. Patients included for this study had confirmation of HIV presence during life either by ELISA or Western blot and met diagnostic criteria for AIDS according to the Centers for Disease Control definitions  In all cases complete autopsy data were available and, in order to assess the changing trends in autopsy pathology, cases were divided into groups according to the year in which the autopsy was performed. A database was constructed taking into account the patient demographic information, final anatomic diagnosis, organ distribution, and pathological cause of death. Frequency of opportunistic infections (viral, fungal, bacterial and protozoon), neoplasms (KS, NHL) and other disorders was calculated as a percentage of the total AIDS autopsies for that year. Additional analysis was performed of the relationship between the severity of the systemic disease and neuropathological findings. For this purpose, following the criteria of Kibayashi  cases were subdivided according to the HIV-associated neuropathological findings into three groups: (i) no alterations; (ii) early alterations; and (iii) late alterations. For each group, the frequency of systemic opportunistic infections alone, opportunistic infections plus neoplasms, neoplasms alone and no brain alteration was determined for each year included in the study. To determine the statistical significance of time trends for factors in the study, linear regression analysis was performed.
During the period 1982–1998 a total of 1970 adult autopsies were performed at UCSD-MC of which 390 were AIDS cases (360 males and 30 females). The ages of these individuals ranged from 20 to 60 years with a mean age 38.9 ± 0.9 (Table 1). The age range was 24–75 years (mean, 37.7 years) for the male group and 23–67 years (mean, 38.3 years) for the female group. Of the 390 cases, 274 had homosexuality as a risk factor for HIV, 101 had a history of injecting drug use and 15 had both homosexuality and history of injecting drug use. Due to the relatively low numbers of cases for the years 1982–1983, linear regression analysis was performed with cases from 1984 to 1998. Linear regression analysis showed that there was a significant increase in the mean age at the time of autopsy (r = 0.628, P = 0.012) (Table 1). However, it is worth noting that the numbers of AIDS autopsies has declined in the last few years relative to the peak seen in 1994 (Table 1). This might be attributable to the introduction in 1996–1997 of new anti-retroviral therapies based on the use of protease inhibitors  The lung (83.9%) was the organ most frequently affected by AIDS-related disorders followed by the CNS (62.7%) and the adrenal glands (43.9%) (Table 2). During the period of the study, opportunistic infections were the most frequent finding (67.9%), followed by combined opportunistic infections and neoplasms (33.7%), neoplasms alone (4.6%), and no alterations (1.0%). As to the opportunistic infections CMV was the most common (50.3%) infection, followed by bacterial infections (40.7%), fungal infections (29.6%), PCP (26.4%) and MAC (20.7%) (Table 3). Throughout the period of study, CMV and bacterial infections remained a common finding (Table 3). Overall, the frequency of PCP (r = –0.739, P = 0.0017), CMV (r = –0.724, P = 0.0023), and MAC (r = –0.682, P = 0.0051) decreased (Table 3), the frequency of bacterial infections increased (r = 0.527, P = 0.0043), and the frequency of fungal infections remained unchanged (r = –0.31, P = 0.261).
As to malignant neoplasms, compared to NHL (14.1%), KS (27.1%) was more frequently found during the total period of the study (with year 1989 being the exception), and its relative frequency was comparable over time (Table 3). Overall, no significant trend was observed as to KS (r = –0.429, P = 0.11) or NHL (r = 0.256, P = 0.356) frequency in autopsy cases over the years of this study.
The nervous system showed alterations in 63% of the autopsied AIDS patients whereas 37% did not show significant brain pathology. During the total period of the study, HIV-related neuropathology (28.3%) was the most common finding followed by CMV (18.0%), fungi (4.9%), progressive multifocal leukoencephalopathy JC virus (3.4%), toxoplasmosis (2.5%) and MAC (1.6%) (Table 4). The frequency of HIV encephalitis has fluctuated from year to year with no clear upward or downward trend. The frequency of CMV has shown a significant downward trend over time (r = 0.57, P = 0.03). The frequency of fungi involvement of the CNS has also fluctuated considerably over time (Table 4) and as to the less commonly seen toxoplasmosis, PML/JC, and MAC the frequency has remained low over time (Table 4).
In order to understand better the effect of systemic pathology over time on the AIDS-related CNS alterations, cases were subdivided into three groups that would reflect the progression of the neuropathology  (i) with no brain alterations; (ii) with early brain alterations; and (iii) with late brain alterations. For the total period of the study, 30% of the AIDS autopsy cases did not show brain alterations, whereas 10% of the cases showed early brain alterations, and 60% displayed late brain alterations. Of the cases with no brain alterations, systemic opportunistic infections were present in 24.8% of the cases, neoplasms in 2.5%, neoplasms and opportunistic infections in 10.4% with only 0.6% of cases showing no systemic disease. In these cases, the frequency of opportunistic infections with and without neoplasms did not vary significantly over time (Table 5). Of the cases with early brain alterations, systemic opportunistic infections were present in only 5.9% of the cases, neoplasms in 0.5%, and neoplasms and opportunistic infections in 1.7%. In this second group, the frequency of opportunistic infections with and without neoplasms did not vary significantly over time (Table 5). Finally, in the group of cases with late brain alterations, systemic opportunistic infections with (21.6%) or without (37.2%) neoplasms were more frequent than in the other two groups (Table 5). In contrast, the frequency of neoplasms alone was very low (0.8%). The frequency of systemic opportunistic infections decreased over time (r = –0.626, P = 0.01), while the frequency of combined opportunistic infections and neoplasms fluctuated over time without showing a definitive upward or downward trend.
The present study showed that over the course of the past 15 years the frequency of opportunistic infections in AIDS autopsy cases has decreased. However, while the frequency of fungal infections remained unchanged, the frequency of bacterial infections increased. The decrease in PCP frequency is consistent with recent studies showing that although PCP continues to be an important cause of death in AIDS patients, its frequency in autopsies has declined significantly  In a study conducted in an inner-city population (Bronx, New York, USA)  the finding of PCP at autopsy declined from 63.7% to 14.5% in the course of a 12-year period (1982–1993) and a study conducted in Los Angeles (California, USA)  reported a decline in PCP from 59.5% to 49.1%. For a similar period, the present study showed a reduction in the cases of PCP from 40% to 22% and in the last 3 years it has decreased even further. In fact, the study by Klatt et al. showed that decreased frequency in deaths from PCP was more closely associated with antiretroviral therapy than with prophylaxis therapy. In this context, the continued decrease in PCP frequency in autopsy material might also be related to the added effects of therapies such as the highly active antiretroviral therapy (HAART), which typically involves a combination of two nucleoside analogs and an HIV-1 protease inhibitor 
Consistent with previous studies [5,6] the present study showed that in spite of aggressive therapy, infection by other agents remained unchanged or was increasing. Particularly in the UCSD cohort, bacterial and fungal infections combined to a lesser extent with CMV have continued to be a frequent finding. The appearance of more aggressive bacterial strains or antibiotic resistant strains is a probable explanation of these findings; however, it is also possible that the added presence of CMV in these patients might play a role. Some studies have suggested a direct mechanism involving dysregulation of the B-cell response by CMV  which could, in turn, render the host more vulnerable to opportunistic infections including bacterial and fungal agents. In addition, it has also been proposed that CMV might act as a co-factor that enhances HIV pathogenicity 
Further supporting the possibility of high prevalence of active HIV infection despite antiretroviral therapy, the present study showed that after the lungs, the CNS is the most commonly involved organ in AIDS autopsy cases. This category includes CNS damage directly related to HIV (HIV encephalitis) as well as opportunistic brain infections. In the present study HIV encephalitis was present in 26% of the cases, while opportunistic infections were present in 36%, no brain alterations was observed in 37% of the AIDS autopsy cases. Consistent with previous studies [7–9] there was an overall trend towards an increase in HIV encephalopathy from 1982 to 1993. Furthermore, deaths from this condition also increased and those deaths were associated with antiretroviral therapy. Development of antiretroviral drug resistance might result in selection of neurotrophic HIV populations accounting as pointed out by Klatt et al. for an increase in the cases of encephalopathy in patients with longer survival. A recent autopsy study in a population in Oslo, Norway has confirmed the observation that HIV-induced brain lesions increased for those cases with longer survival  However, this study also showed that, when maintained, antiretroviral therapy can reduce HIV encephalopathy and when it is discontinued HIV encephalitis increases. Other studies  have also observed a decline in HIV involvement of the CNS since the establishment of antiretroviral therapy. For the present study our group, as well as others [7–9] have observed a decline in HIV encephalitis as an autopsy finding in 1987–1989; however, since then there has been a trend toward an increase in the numbsent study our group, as well as others [7–9] have observed a decline in HIV encephalitis as an autopsy finding in 1987–1989; however, since then there has been a trend toward an increase in the number of these cases. One possible explanation for these differences might be related to the severity and duration of the disease in various populations. In order to determine the impact of systemic disease in HIV neuropathogenicity and how this has changed overt time, we compared the frequency of opportunistic infections and neoplasms in AIDS cases divided according to the nature of HIV-associated neuropathology. A previous study has shown that this is a valuable approach to determining the stage of HIV involvement of the CNS  This study showed that chronic lymphocytic meningitis, which has been regarded as an indicator of early disease, in fact could be found in various stages of AIDS, but that the frequency was higher in asymptomatic carriers. As to HIV encephalitis, the frequency of this condition was unrelated to the stage of AIDS, suggesting that HIV encephalitis occurs before opportunistic infections and neoplasms  In the present study the frequency of opportunistic infections and neoplasms was lower in those cases with chronic lymphocytic meningitis than in cases with no brain alterations. Cases with advanced brain alterations had the highest frequency of opportunistic infections and neoplasms. Over time there has been a tendency for opportunistic infections and neoplasms to decrease in patients with HIV-associated CNS damage. This apparent discrepancy could be explained on the basis that if HIV involvement of the CNS is a very early process then opportunistic infections will be less frequent in such cases; probably there is a phase where no apparent brain lesions are found but extensive dissemination of opportunistic infections appear. At this stage HIV could be in a quiescent state in the brain. Once systemic opportunistic infections and neoplasms had disseminated then HIV in the brain is reactivated, and more severe lesions are observed. This suggests that after an initial phase of HIV infection to the brain–in the absence of widespread systemic opportunistic infections–HIV might become less active in the brain followed by a phase of dissemination of systemic disease with subsequent reactivation of HIV-related CNS disease.
In conclusion, the present study showed that despite beneficial effects of antiretroviral and anti-opportunistic infection therapy, involvement of the brain by HIV continues to be a frequent finding in the UCSD series.
1. Welch K Finkbeiner W Alpers CE et al Autopsy findings in the acquired immunodeficiency syndrome.
JAMA 1984 2521152–1159.
2. Hui AN Koss MN Meyer PR Necropsy findings in acquired immunodeficiency syndrome: A comparison of premortem diagnoses with postmortem findings.
Hum Pathol 1984 15670–676.
3. Guarda LA Luna MA Smith JL Mansell PWA Gyorkey F Roca AN Acquired immune deficiency syndrome: Postmortem findings.
Am J Clin Pathol 1984 81549–557.
4. Reichert CM O'Leary TJ Levens DL Simrell CR Macher AM Autopsy pathology in the acquired immune deficiency syndrome.
Am J Pathol 1983 112357–382.
5. Sehonanda A Choi YJ Blum S Changing patterns of autopsy findings among persons with acquired immunodeficiency syndrome in an inner-city population.
Arch Pathol Lab Med 1996 120459–464.
6. Klatt EC Nichols L Noguchi TT Evolving trends revealed by autopsies of patients with the acquired immunodeficiency syndrome.
Arch Pathol Lab Med 1994 118884–890.
7. Maehlen J Dunlop O Liestol K Dobloug JH Goplen AK Torvik A Changing incidence of HIV-induced brain lesions in Oslo, 1983–1994: effects of zidovudine treatment.
AIDS 1995 91165–1169.
8. Davies J Everall IP Weich S McLaughlin J Scaravilli F Lantos PL HIV-associated brain pathology in the United Kingdom: an epidemiological study.
AIDS 1997 111145–1150.
9. Bacellar H Munoz A Miller EN et al Temporal trends in the incidence of HIV-1-related neurologic diseases: Multicenter AIDS cohort study, 1985–1992.
Neurology 1994 441892–1900.
10. Ioannidis JP O'Brien TR Goedert JJ Evaluation of guidelines for initiation of highly active antiretroviral therapy in a longitudinal cohort of HIV-infected individuals.
AIDS 1998 122417–2423.
11. Hammett TA Ciesielski CA Bush TJ Fleming PL Ward JW Impact of the 1993 expanded AIDS surveillance case definition on reporting of persons without HIV risk information.
J Acquir Immune Defic Syndr Hum Retrovirol 1997 14259–262.
12. Kibayashi K Mastri AR Hirsch CS Neuropathology of acquired immunodeficiency virus infection at different stages.
Hum Pathol 1996 27637–642.
13. Dore GJ Marriott DJ Duflou JA Clinico-pathological study of cytomegalovirus (CMV) in AIDS autopsies: under-recognition of CMV pneumonitis and CMV adrenalitis.
Austr New Zeal J Med 1995 25503–506.
14. Portegies P Enting RH de Gans J et al Presentation and course of AIDS dementia complex: 10 years of follow-up in Amsterdam, The Netherlands.
AIDS 1993 7669–675.