Matters of Survival

What do we know about AIDS-related OIs and survival?

by Chael Needle

Sandra Schwarcz, MD. MPH, senior HIV epidemiologist at the San Francisco Department of Public Health
Sandra Schwarcz, MD. MPH, senior HIV epidemiologist at the San Francisco Department of Public Health

[dropcap]I[/dropcap]s better HIV treatment related to improved survival after an individual is diagnosed with an AIDS-defining opportunistic illness (AIDS-OI)? That was one of the central questions of a recent research study, whose results were published in a recent issue of The Journal of Infectious Diseases (“Mortality Risk After AIDS-Defining Opportunistic Illness Among HIV-Infected Persons—San Francisco, 1981–2012,” by Djawe et al).

The short answer? Yes, in general. But while the short answer may satisfy those who only keep their eye on big-picture trends, it arguably means little to those individuals who are living with HIV and AIDS now and who need to make informed decisions about treatment and care. They deserve long, complex answers.

The researchers of this study have given them one, a long, complex answer focused not only on survival and OIs but also on how different OIs have different challenges. Even as prophylaxis and/or antiretroviral therapy prevent the development of OIs and reduce the risk of mortality, AIDS is not over. Even in the current treatment landscape, OIs still cause death. The upshot? We need to commit to finding and fine-tuning better prevention and treatment strategies for AIDS-OIs.

Finding data about OIs
Researchers reviewed HIV surveillance data of adult patients with AIDS in San Francisco who were diagnosed with one or more AIDS-OIs, estimating the probability of survival after first diagnosis of an OI. But one of the challenges researchers in general face when studying OIs over the timespan of the epidemic is data collection. Says one of the study’s coauthor’s Sandra Schwarcz, MD, MPH, senior HIV epidemiologist at the San Francisco Department of Public Health, OI data collection in San Francisco is population-based and OI-specific, and these provide key advantages if you want to analyze OI-related survival information.

Two factors of surveillance methods have stymied data collection of AIDS-OIs.

First, the definition of AIDS has backgrounded AIDS-OI information. In 1993, the method of AIDS diagnosis changed. In the early days, the definition of AIDS was dependent solely on the occurrence of an OI; in 1993, the definition was changed to include low CD4 cells, says Dr. Schwarcz. This change took into consideration that fact that a drop in a patient’s CD4 cells usually precedes but does not automatically determine the development of an OI. Obviously, this marker helped physicians and patients address OIs before the fact rather than after the fact, knowledge of which could effect better health outcomes.

So, before 1993, the reporting of an AIDS case meant the reporting of an OI, typically the first OI. After 1993, AIDS cases were reported primarily because of low CD4 cells. OIs that developed among persons who were already reported with AIDS because of low CD4 cells were not captured in the surveillance systems, says Dr. Schwarcz. In San Francisco, however, information about the development of OIs was collected as part of routine follow-up of all persons reported to the surveillance system, setting the city apart from most if not all other sites reporting AIDS cases.

Second, the data that have been collected about AIDS-OIs are frequently culled from cohort-based studies, another primary source of AIDS data collection alongside health department surveillance. However, data collection from a clinic-based cohort study, explains Dr. Schwarcz, has built-in limitations. For example, she notes, “it includes people who (1) attend the clinic; (2) volunteer to be in that study and to consent, and (3) continue in that study. So there can be biases associated with that. Some people want to be in the study; some people don’t. And there can be differences in the characteristics of those individuals, so that within a clinical study such as a clinic cohort the data may not actually be representative of the entire population.”

And while population-based surveillance will always be limited to its locality, a city like San Francisco can provide a window on the national HIV epidemic, notes Dr. Schwarcz. However, in San Francisco HIV/AIDS primarily affects men who have sex with men and, as a result, certain OIs, such as Kaposi’s sarcoma [KS], are more frequent than one would expect from other parts of the country where populations such as persons who inject drugs are more common.

Another advantage of analyzing San Francisco data is its comprehensiveness. San Francisco surveillance follows individuals diagnosed with HIV and/or an AIDS-OI throughout care, and thus can bring to light data about subsequent OIs in addition to the first OI. In addition, San Francisco collects information on HIV treatment Dr. Schwarcz points out, creating a fuller representation of the impact of treatment on mortality.

A closer look at OIs
The impetus of the study, says Dr. Schwarcz, evolved out of an earlier analysis she and others conducted based on the San Francisco Department of Public Health, HIV/AIDS Surveillance and Epidemiology’s AIDS case registry. “We looked at the incidence, that is the rate of new occurrences, of opportunistic illnesses both in initial, meaning the patient had their first opportunistic illness, as well as subsequent, meaning someone had an opportunistic illness, recovered from that illness, and then developed another one.” That study found a significant reduction in incidence rates for the eight most frequently reported AIDS-OIs from 1993 through 2008, qualitatively affirming that combination ART and increased CD4 cell counts offer protection against AIDS-OIs.

So the question of mortality in the context of declining AIDS-OIs was one issue that deserved further exploration.

“It’s very clear that treatment for HIV infection has had a dramatic effect on survival, as well as a dramatic effect on comorbidities associated with HIV infection, and prominently opportunistic illnesses. That’s pretty well established—and that is fabulous,” notes Dr. Schwarcz. However, attention to OIs, particularly their relation to mortality has somewhat waned as a topic of research, and this provided a jumping-off point for the new study.

In this study, mortality risk was reviewed across three time periods over thirty years: 1981–1986, 1987–1996, 1997–2012. Researchers found that the most recent era, when combination antiretroviral therapies became available and widely accessible, sharply contrasts with previous eras. The probability of survival of at least five years after an initial OI diagnosis was significantly higher by leaps and bounds in the 1997–2012 period compared with 1981-1986, prior to effective medication: A seven percent chance of survival in the pre-ART era leapt to eighteen percent in the mono/dual-ART era (which had the highest annual number of diagnoses of all eras), and then bounded up to sixty-five percent in the combination ART era. Across these three eras, the mortality rate of patients with AIDS-OIs who died by the end of observation (17,099 of the 20, 858 patients analyzed) was 98.4, 89.2, and 41.9 percent, respectively.

Information about different AIDS-OIs was also teased out. For the entire timespan of the study, the ten AIDS-OIs that were most frequently diagnosed, from most to least, were pneumocystis pneumonia (PCP), KS, HIV wasting syndrome, esophageal candidiasis, Mycobacterium avium complex (MAC), HIV encephalopathy, extrapulmonary cryptococcosis, chronic cryptosporidiosis, immunoblastic lymphoma, and cytomegalovirus disease other than retinitis. The bottom five included invasive cervical cancer, recurrent Salmonella septicemia, disseminated coccidioidomycosis, chronic isosporiasis, and disseminated histoplasmosis.

Survival rates by specific OIs differed from one another although survival improved for each of these OIs in most recent time period compared to earlier years. For some OIs, like MAC and immunoblastic lymphoma, the median survival was less than five years even in the combination therapy era. Survival was markedly better for other OIs such as PCP, KS, and cryptosporidiosis all of which had median survival of greater than fifteen years. Compare this with the median survival time in the previous eras—less than five years—and you can see the progress. Note however that, even as the number of OIs have declined the same OIs’ ranking by frequency of diagnosis has not changed significantly since 1981.

The numbers also show that many patients in the combination therapy era were prescribed combination therapy as well as prophylaxis against PCP and MAC. But changes across the board were not uniform. Compared to PCP, and after adjusting for age, transmission category, CD4 cell count, use of any ART, and prophylaxis against MAC and PCP, progressive multifocal leukoencephalopathy (PML) and brain lymphomas, as well as persons with multiple OIs had worse mortality while the highest survival was observed among patients with KS and chronic cryptosporidiosis.

The authors noted, too, that San Francisco survival rates might be higher than other places around the country because of programs that help patients enter and remain in care.

Surveillance means survival
The study authors note that there is room for improvement when you consider there is a thirty-five percent mortality rate within five years of the initial AIDS-defining OI in the current treatment era.

In response to a question about whether or not research into OI-specific treatments and/or prophylaxis is needed, Dr. Schwarcz responded: “If those could be identified those could be extremely beneficial. And certainly that was seen with prophylaxis against pneumocystis pneumonia, where that prophylaxis was really beneficial. In fact if you look at the trends in death you can see that before we even had good antiretroviral therapy—meaning, highly effective—the deaths in San Francisco plateaued between ’92 and ’94…and that’s attributed to PCP prophylaxis. [You could really see] it had a marked impact on death.” The study authors noted that individuals with a first AIDS-OI diagnosed in the combination therapy era were more likely to die if they were not prescribed PCP or MAC prophylaxis, or antiretroviral therapy. Other factors influencing survival—such as injection drug use and access to healthcare—were also discussed.

Alongside bringing to light new areas of needed research, Dr. Schwarcz believes the study results will hopefully alert clinicians to be be more aware of opportunistic illnesses when treating individuals living with HIV/AIDS. She also points out that early diagnosis, early treatment, and adherence to medications are all standard-of-care strategies that generally confer the best survival benefit.

Funding surveillance could be addressed as well. Notes Dr. Schwarcz: “Conducting surveillance costs money. And it’s tax dollars. And it’s really important I think for individuals to recognize the value of conducting surveillance. It’s a good use, I believe, of taxpayer dollars and it’s something that needs to continue to be supported if we’re going to be able to continue to have these sorts of data, that are really important. Lots of people don’t recognize that it’s critical, and, like everything else, it costs money.”


Chael Needle wrote about Man Up Mondays in the June issue of A&U.