Prior to the availability and widespread use of highly active antiretroviral therapy (HAART), persons with AIDS commonly suffered from opportunistic infections and malignancies, which resulted in costly hospitalizations, toxic treatments, and mortality . The incidence rates of these AIDS-defining opportunistic illnesses (AOIs) have drastically declined since the introduction of HAART in 1996 and the implementation of prophylaxis against Pneumocystis jiroveci pneumonia (PCP) and Mycobacterium avium complex (MAC) in the mid 1990s [2–6]. Despite these declines, AOIs continue to be a leading cause of morbidity and mortality among HIV-infected individuals [7,8]. In 1993 in the US, the Centers for Disease Control and Prevention (CDC) expanded the AIDS surveillance case definition to include persons with laboratory confirmation of HIV infection whose CD4 cell counts were below 200 cells/μl or less than 14% of total lymphocytes [9,10], as well as those diagnosed with an AOI. Because this degree of immunosuppression generally precedes the development of an AOI, following the implementation of the change in case definition, the majority of AIDS cases in the US have been reported with severe immunocompromise but without an AOI. By 1998, the national surveillance data on AOI was so poor that the CDC stopped publishing the number and distribution of AOIs among persons reported with AIDS . The San Francisco Department of Public Health (SFDPH) has collected diagnoses of AOIs that occurred at the time of AIDS diagnosis and thereafter through periodic medical chart reviews. This has provided a longitudinal population-based registry of persons with AIDS that includes information on initial and subsequent AOIs, CD4 test results, and the use of antiretroviral therapy and prophylactic medications. We used the SFDPH AIDS case registry to examine the distribution, incidence trends, and risk factors of AOIs that occurred at the time of or after AIDS diagnosis, over a 16-year period marked by substantial improvements in the efficacy of antiretroviral therapy.
Study design, setting, and data sources
We used AIDS surveillance data to describe the distribution of AOIs and to measure incidence trends in and predictors of frequently occurring AOIs during the years 1993 through 2008. Mandatory reporting of San Francisco residents who meet the CDC AIDS surveillance case definition to the SFDPH is implemented primarily through active surveillance, in which health department personnel identify and report persons who have been diagnosed with AIDS. The medical records from all hospitals and affiliated clinics, publicly funded facilities, and all private physician offices with large HIV practices in San Francisco where the patient is known to have received care are reviewed at the time of diagnosis and approximately every 18–24 months thereafter. The information collected includes date of birth, race/ethnicity, sex, HIV risk group, country of birth, insurance status, whether the person is homeless at the time of diagnosis, the month and year of initial and subsequent AOIs, CD4 test results, the type and the month and year of initiation of antiretroviral therapy, and prophylaxis against PCP and MAC. The medical records of persons reported with AIDS in San Francisco whose care is provided outside of the city cannot be reviewed by SFDPH staff. Deaths of persons reported with AIDS are obtained through monthly computer matches of the AIDS case registry with local vital statistics and annual matches with the Social Security Death Master File and the National Death Index. The San Francisco AIDS case reporting system is evaluated annually and has consistently been found to be over 90% complete .
We analyzed data during the period 1 January 1993 through 31 December 2008. The study sample included San Francisco adult and adolescent (age ≥13) residents with AIDS who were alive as of 1 January 1993 or were diagnosed with AIDS between this date and 31 December 2008. Individuals with AIDS were reported to the health department as of 9 March 2011. We contrasted AOI incidence in three calendar periods. In the first period (1993–1995), HAART including protease inhibitors was not yet available [13–15]. In the second period (1996–2000), HAART became widely available [16,17]. The third period (2001–2008) begins with licensure of lopinavir-ritonavir, a combination protease inhibitor with especially high efficacy [18,19].
Distribution of AIDS-defining conditions
We first tabulated initial AIDS-defining conditions included in the 1993 CDC AIDS case surveillance definition among persons who were newly diagnosed with AIDS in each of the three calendar periods. We included all AOIs that were diagnosed in the month and year of AIDS diagnosis; as a result, the total number of AOIs at diagnosis is greater than the number of AIDS diagnoses in each period. Although nearly all persons whose initial AIDS-defining condition included a single AOI, 2.7% (n = 354) were reported with more than one AOI at diagnosis. Persons who met the CDC AIDS surveillance case definition solely because of severe immunosuppression were tabulated separately.
Incidence estimates and predictors of AIDS-defining opportunistic illnesses
We estimated incidence rates for AOIs diagnosed at least 1 month after diagnosis of AIDS, focusing on the eight most common AOIs; PCP, wasting syndrome, Kaposi's sarcoma, MAC, esophageal candidiasis, HIV encephalopathy, cytomegalovirus (CMV) retinitis, and CMV. Within each calendar year, person-months and events for each AOI were summed for prevalent and newly diagnosed AIDS cases with no history of that specific AOI, from the beginning of the year or the date of AIDS diagnosis, whichever came last, until the end of the year, diagnosis of the AOI of interest, last known follow-up, or death, whichever came first. Unadjusted rates per 1000 person-years were tabulated for each AOI, first within calendar years, and then pooled within each of the three calendar periods.
We then used Poisson regression to estimate incidence rates for each AOI adjusting for age, race/ethnicity, sex, risk group, insurance, and homelessness. Adjusted rates were estimated using so-called regression standardization, in which the predicted rate was calculated for each year or period based on the overall joint distribution of covariates across all years.
The independent predictors of each incident AOI were also examined using Poisson models. We considered demographics, risk group, insurance, and homelessness, the mean of all available CD4 cell counts (calculated from the beginning of the year up until the occurrence of the AOI or end of the year), and use of HAART, and prophylaxis against PCP and MAC. For persons who were diagnosed with AIDS prior to 1996, we adjusted for the number of years between diagnosis and 1996, to account for the cumulative effects of lack of HAART. Because only the dates of treatment initiation and not dates of treatment discontinuation are collected, we coded treatment use as the percentage of time on treatment in the calendar period during which it was started and as 100% for all subsequent calendar periods.
The sample includes 18 733 persons living with AIDS, including 5788 diagnosed prior to 1993. Most were male (93%), white (66%), aged between 30 and 39 years (43%), men who have sex with men (MSM), including those who inject drugs (87%), privately insured (42%), housed at diagnosis (91%), and born in the US (88%) (Table 1). Seventy-three percent of persons met the AIDS case definition by immunologic criteria alone.
Of the 12 945 adult/adolescents diagnosed with AIDS between 1993 and 2008, 5413 were diagnosed in 1993–1995, 3706 in 1996–2000, and 3826 in 2001–2008 (Table 2). The most frequent AOIs at the time of AIDS diagnosis (esophageal candidiasis, Kaposi's sarcoma, PCP, and wasting syndrome) in the first time period continued to be the most frequent in the other time periods. The proportion of persons who were diagnosed with AIDS by virtue of low CD4 cell counts or percentages in the absence of an AOI was relatively stable during the first two calendar periods (73 and 71%, respectively) and increased in the most recent period (79%). There were a total of 1647 AOIs in the first period, 1206 in the second period, and 865 in the most recent period.
Trends in rates of opportunistic illnesses following AIDS diagnosis
Incidence of all eight AOIs following AIDS diagnosis decreased from 1993 to 2008, most notably between time periods 1993–1995 and 1996–2000, which coincided with the introduction of HAART (Table 3 and Fig. 1). In the first time period (1993–1995), the highest incidence rates (per 1000 person-years) were observed for PCP (95.0), MAC (85.2), wasting syndrome (67.8), and Kaposi's sarcoma (63.4). In the third period (2001–2008), the incidence rates for PCP and wasting syndrome remained the highest of the AOIs examined (8.4 and 6.5 per 1000 person-years, respectively), whereas the incidence rate of MAC declined to 3.2, behind the rates for esophageal candidiasis (4.7) and Kaposi's sarcoma (4.0).
The largest declines in AOI incidence rates occurred between 1993–1995 and 1996–2000. Rates continued to decrease in 2001–2008, but more slowly. The greatest overall rate reductions between 1993–1995 and 2001–2008 were for wasting syndrome [99%, CMV retinitis (97%), MAC (96%), and CMV (95%)]. The reductions between 1996–2000 and 2001–2008 ranged from 58% for esophageal candidiasis to 94% for CMV retinitis.
Predictors of incident AIDS-defining opportunistic illnesses following AIDS diagnosis
Higher mean CD4 cell count and HAART use were independently associated with lower incidence of all AOIs, but we found no evidence for protection against PCP or MAC with prophylaxis (Table 4). We did find independent associations of age, sex, race/ethnicity, risk group, insurance status, and housing status with the development of AOIs.
Sensitivity analyses omitting prevalent AIDS cases
We included prevalent cases who were living with AIDS at the start of each calendar period, provided that they had not been diagnosed with the AOI earlier. It is possible that this introduced a survivor bias. To assess this possibility we conducted two supplemental analyses. First, we compared the socio-demographic, risk, and treatment characteristics of persons who were diagnosed with AIDS prior to the start of each calendar period with those diagnosed during each period. We also re-estimated incidence trends and risk factors for AOI restricting the analysis to persons newly diagnosed with AIDS within each time period. Although prevalent cases were more likely to have received HAART and prophylaxis, most likely reflecting increased time to have received treatment and less severe illness or slower disease progression, our findings remained essentially unchanged (data not shown).
In this population-based analysis, we found a remarkable reduction in incidence rates for all AOIs from 1993 through 2008. Our findings show the steepest declines of AOI incidence rates between 1993–1995 and 1996–2000, after the introduction of widespread use of HAART, followed by a more modest decline in 2001–2008, when more effective and better tolerated HAART regimens became available. This pattern is consistent with reported declines in AOI morbidity in the HAART era [4,5,16,20,21]. Decreases in the incidence of PCP and MAC are likely to be affected both by HAART and by increased utilization of PCP and MAC prophylaxis over time. Declines in Kaposi's sarcoma were observed many years prior to use of HAART [4,22,23] and are likely a reflection of reductions in new infections with human herpesvirus 8, the known co-factor for HIV-associated Kaposi's sarcoma [24–26]. The continued declines between 1996–2000 and 2001–2008 may reflect increased efficacy of HAART, as well as improved adherence due to the reduced toxicity and lower pill burden of more recent formulations of HAART .
Our analysis qualitatively affirmed the protective effect of HAART and higher CD4 cell counts against all AOIs. However, we likely under-estimated protective HAART effect, due to lack of information on adherence and treatment discontinuation. In addition, adjustment for CD4 cell counts, which both confound and mediate HAART effects, may attenuate the estimate for HAART, but addressing this problem would require more complicated analyses using marginal structural models , as well as better covariate information. Finally, this attenuation, as well as our inability to detect significant protection with PCP and MAC prophylaxis, may also be due to uncontrolled confounding by indication: that is, increased HAART and prophylaxis use among persons with more severe illness. Unfortunately, our database lacks the detailed clinical information that would be needed to obtain better estimates of these effects. A primary rationale for adjusting for HAART and other treatments was to reduce confounding of the other covariate effects.
We also found associations between several socio-demographic characteristics and selected AOIs. Increasing age was associated with higher incidence rates of wasting syndrome, CMV, and encephalopathy. Other studies have found that increasing age is associated with higher risk of developing an AOI [4,28–33] and may be due to HIV-related and age-related inflammation and cellular senescence [34,35]. The lower risk for Kaposi's sarcoma among women, African Americans, and persons whose risk was not male–male sex reflects the predominant distribution of Kaposi's sarcoma-associated human herpes virus 8 in MSM and is supported by other epidemiologic studies [36–38]. The higher risk of AOIs among the homeless likely reflects a number of factors in this population that contribute to worse health, including the absence of stable storage for medications, substance use, and mental illness. However, other associations we found are more difficult to interpret, given the lack of known co-factors for many AOIs. We also recognize that some nominally significant associations may be due to chance, given the large number of factors we examined.
Our study is subject to several limitations. Most important, information from the surveillance database is incomplete in at least three ways. First, whereas evaluation of our surveillance system has demonstrated that reporting is consistently above 90%, follow-up on persons who moved or transferred care outside of San Francisco cannot be obtained, and while we conduct follow-up chart reviews at all clinical sites that serve large numbers of HIV-infected patients, we may be missing follow-up data on a few patients seen at smaller private practices. We attempt to review medical records at least every 2 years but recognize that review of records for 2008 may be incomplete. However, the very strong trends we report are unlikely to be explained by bias from this source, which adjustment for age and treatments as time-dependent variables should have reduced. Second, our analysis of predictors of selected AOIs is limited to variables routinely collected by chart review as part of AIDS surveillance, and by the lack of more direct measures of being in care. We cannot determine the extent to which more extensive clinical and adherence information would have altered those findings. Third, our estimates of incidence may be biased low because AOI recurrences after the first postdiagnosis occurrence are not captured by surveillance procedures. This is most likely to affect our estimates of incidence rates in 1993–1995, when AOIs were much less well controlled, suggesting that the observed declines might appear even stronger if recurrences had been captured. Due to this surveillance constraint, AIDS cases do not contribute person-time after a first postdiagnosis AOI occurrence, correctly reflecting the fact that recurrences are not captured.
An additional limitation is that relatively accurate and complete dates of HIV diagnosis have only been available since 2006, when name-based HIV case reporting was mandated in California. As a result, we were unable to assess the effects of delayed entry into care on AOI incidence, or to calculate AOI incidence rates among persons with HIV infection who had not progressed to AIDS. Thus our incidence rates are not comparable to estimates based on cohort samples including this group.
Finally, our data are from a single city in which over 65% of AIDS cases are white MSM. As such, the findings may not be generalizable to areas where the AIDS case population differs from ours. However, MSM currently account for 61% of all new HIV infections in the US . Our results shed light on clinical trends among AIDS cases arising from this large risk group.
To our knowledge, ours is the only AIDS case registry that has collected longitudinal data of AOI diagnoses that occurred at the time of AIDS diagnosis or thereafter and has done so since the first AIDS case was reported in 1981.The continued development of AOIs is a reminder of the need for early HIV diagnosis, prompt entry into and retention in care, and initiation of and adherence to HAART and AOI prophylaxis. Ongoing surveillance of AOIs offers one method to measure control of HIV disease and the impact of changes in HIV treatment.
Author contributions: L.S. contributed to the conception and design of the study, the analysis and interpretation of the data, and wrote the initial draft of the manuscript.
M.-J.C. contributed to the conception and design of the study, the acquisition, analysis and interpretation of the data, and reviewed and edited the manuscript.
E.V. contributed to the conception and design of the study, developed the analysis plan, contributed to the analysis and interpretation of the data, and reviewed and edited the manuscript.
L.H. contributed to the conception and design of the study, the acquisition, analysis and interpretation of the data, and reviewed and edited the manuscript.
S.S. contributed to the conception and design of the study, the acquisition, analysis and interpretation of the data, and reviewed and edited the manuscript.
The authors wish to thank Drs Kate Buchacz and Nicholas Moss for their thoughtful review of the manuscript.
Conflicts of interest
There are no conflicts of interest.
1. Jones JL, Hanson DL, Dworkin MS, Alderton DL, Fleming PL, Kaplan JE, et al. Surveillance for AIDS-defining opportunistic illnesses, 1992–1997
. MMWR CDC Surveill Summ
2. Masur H, Kaplan JE, Holmes KK. Guidelines for preventing opportunistic infections among HIV-infected persons: 2001
. Ann Internal Med
3. Dore GJ, Li Y, McDonald A, Ree H, Kaldor JM. Impact of highly active antiretroviral therapy on individual AIDS-defining illness incidence and survival in Australia
. J Acquir Immune Defic Syndr
4. Buchacz K, Baker RK, Palella FJ Jr, Chmiel JS, Lichtenstein KA, Novak RM, et al. AIDS-defining opportunistic illnesses in US patients, 1994–2007: a cohort study
5. Brodt HR, Kamps BS, Gute P, Knupp B, Staszewski S, Helm EB. Changing incidence of AIDS-defining illnesses in the era of antiretroviral combination therapy
6. Ives NJ, Gazzard BG, Easterbrook PJ. The changing pattern of AIDS-defining illnesses with the introduction of highly active antiretroviral therapy (HAART)in a London clinic
. J Infect
7. Ledergerber B, Egger M, Erard V, Weber R, Hirschel B, Furrer H, et al. AIDS-related opportunistic illnesses occurring after initiation of potent antiretroviral therapy: the Swiss HIV Cohort Study
8. Mofenson LM, Brady MT, Danner SP, Dominguez KL, Hazra R, Handelsman E, et al. Guidelines for the Prevention and Treatment of Opportunistic Infections among HIV-exposed and HIV-infected children: recommendations from CDC, the National Institutes of Health, the HIV Medicine Association of the Infectious Diseases Society of America, the Pediatric Infectious Diseases Society, and the American Academy of Pediatrics
. MMWR Recomm Rep
9. CDC. 1993 Revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults
. Morb Mortal Wkly Rep (RR)
10. CDC. 1993 Revised classification system for HIV infection and expanded surveillance case definition for aids among adolescents and adults
. Atlanta, GA: Center for Disease Control and Prevention; 1992.
11. CDC. HIV/AIDS surveillance report, 1998.Morb Mortal Wkly Rep
12. San Francisco Department of Public Health S. HIV/AIDS epidemiology 2009 annual report
. San Francisco, CS: San Francisco Department of Public Health; 2009.
13. Fischl MA, Richman DD, Grieco MH, Gottlieb MS, Volberding PA, Laskin OL, et al. The efficacy of azidothymidine (AZT) in the treatment of patients with AIDS and AIDS-related complex. A double-blind, placebo-controlled trial
. N Engl J Med
14. Nightingale SD, Cameron DW, Gordin FM, Sullam PM, Cohn DL, Chaisson RE, et al. Two controlled trials of rifabutin prophylaxis against Mycobacterium avium complex infection in AIDS
. N Engl J Med
15. Hughes WT, Kuhn S, Chaudhary S, Feldman S, Verzosa M, Aur RJ, et al. Successful chemoprophylaxis for Pneumocystis carinii pneumonitis
. N Engl J Med
16. Detels R, Tarwater P, Phair JP, Margolick J, Riddler SA, Munoz A. Effectiveness of potent antiretroviral therapies on the incidence of opportunistic infections before and after AIDS diagnosis
17. Hammer SM, Squires KE, Hughes MD, Grimes JM, Demeter LM, Currier JS, et al. A controlled trial of two nucleoside analogues plus indinavir in persons with human immunodeficiency virus infection and CD4 cell counts of 200 per cubic millimeter or less. AIDS Clinical Trials Group 320 Study Team
. N Engl J Med
18. Walmsley S, Bernstein B, King M, Arribas J, Beall G, Ruane P, et al. Lopinavir-ritonavir versus nelfinavir for the initial treatment of HIV infection
. N Engl J Med
19. Murphy RL, Brun S, Hicks C, Eron JJ, Gulick R, King M, et al. ABT-378/ritonavir plus stavudine and lamivudine for the treatment of antiretroviral-naive adults with HIV-1 infection: 48-week results
20. Hanna DB, Gupta LS, Jones LE, Thompson DM, Kellerman SE, Sackoff JE. AIDS-defining opportunistic illnesses in the HAART era in New York City
. AIDS Care
21. Palella FJ Jr, Delaney KM, Moorman AC, Loveless MO, Fuhrer J, Satten GA, et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. HIV Outpatient Study Investigators
. N Engl J Med
22. Rutherford GW, Schwarcz SK, Lemp GF, Barnhart JL, Rauch KJ, Warner WL, et al. The epidemiology of AIDS-related Kaposi's sarcoma in San Francisco
. J Infect Dis
23. Katz MH, Hessol NA, Buchbinder SP, Hirozawa A, O’Malley P, Holmberg SD. Temporal trends of opportunistic infections and malignancies in homosexual men with AIDS
. J Infect Dis
24. Whitby D, Howard MR, Tenant-Flowers M, Brink NS, Copas A, Boshoff C, et al. Detection of Kaposi sarcoma associated herpesvirus in peripheral blood of HIV-infected individuals and progression to Kaposi's sarcoma
25. Gao SJ, Kingsley L, Hoover DR, Spira TJ, Rinaldo CR, Saah A, et al. Seroconversion to antibodies against Kaposi's sarcoma-associated herpesvirus-related latent nuclear antigens before the development of Kaposi's sarcoma
. N Engl J Med
26. Martin JN, Ganem DE, Osmond DH, Page-Shafer KA, Macrae D, Kedes DH. Sexual transmission and the natural history of human herpesvirus 8 infection
. N Engl J Med
27. Hernan M, Brumback B. Robins J. marginal structural models to estimate the causal effect of zidovudine on the survival of HIV-positive men
28. Gallafent JH, Buskin SE, De Turk PB, Aboulafia DM. Profile of patients with Kaposi's sarcoma in the era of highly active antiretroviral therapy
. J Clin Oncol
29. d’Arminio Monforte A, Cozzi-Lepri A, Castagna A, Antinori A, De Luca A, Mussini C, et al. Risk of developing specific AIDS-defining illnesses in patients coinfected with HIV and hepatitis C virus with or without liver cirrhosis
. Clin Infect Dis
30. Mocroft A, Oancea C, van Lunzen J, Vanhems P, Banhegyi D, Chiesi A, et al. Decline in esophageal candidiasis and use of antimycotics in European patients with HIV
. Am J Gastroenterol
31. Sorkin JD, Bolton PA, Greenblatt J, Sithisarankul P, Vlahov D, Graham NM. Age, gender, and other predictors of the wasting syndrome among HIV-1-infected injecting drug users
32. Bhaskaran K, Mussini C, Antinori A, Walker AS, Dorrucci M, Sabin C, et al. Changes in the incidence and predictors of human immunodeficiency virus-associated dementia in the era of highly active antiretroviral therapy
. Ann Neurol
33. Valcour V, Shikuma C, Shiramizu B, Watters M, Poff P, Selnes O, et al. Higher frequency of dementia in older HIV-1 individuals: the Hawaii aging with HIV-1 cohort
34. Kuller LH, Tracy R, Belloso W, De Wit S, Drummond F, Lane HC, et al. Inflammatory and coagulation biomarkers and mortality in patients with HIV infection
. PLoS Med
35. Singh T, Newman AB. Inflammatory markers in population studies of aging
. Ageing Res Rev
36. CDC. CDC surveillance summaries
. Morb Mortal Wkly Rep
37. d’Arminio Monforte A, Sabin CA, Phillips A, Sterne J, May M, Justice A, et al. The changing incidence of AIDS events in patients receiving highly active antiretroviral therapy
. Arch Intern Med
38. Greenberg AE, Thomas PA, Landesman SH, Mildvan D, Seidlin M, Friedland GH, et al. The spectrum of HIV-1-related disease among outpatients in New York City
39. CDC. HIV in the United States: at a glance. March 2012 ed. http://www.cdc.gov/hiv/resources/factsheets/PDF/HIV_at_a_glance.pdf
: CDC, Division of HIV/AIDS Prevention; 2012. pp. 1–2.