Information for patients
Research is essential if we are to understand more about HIV infection and to get better and safer treatments. The research carried out by ART-CC is focussed on patients who are receiving antiretroviral therapy. Several ART-CC papers have contributed to treatment guideline development.
Patients may be particularly interested in our research summaries, VACS Index 2.0 and links to support groups.
Further information for patients can be found on the HIV treatment information base.
Below we explain the results of our recent analyses.
The life expectancy of people with HIV taking antiretroviral therapy has increased a lot over the past 25 years. Most previous studies of life expectancy were based on data from the first few years after starting antiretroviral therapy, when mortality is highest. However, many people with HIV have been successfully treated with antiretroviral therapy for many years, so, up-to-date estimates are needed for people on long-term antiretroviral therapy.
We included adults with HIV in Europe and North America who started antiretroviral therapy between 1996 and 2014 and had been on antiretroviral therapy for at least 1 year by 2015. We also included data from adults who started antiretroviral therapy between 2015 and 2019 after they had been on antiretroviral therapy for 1 year.
There were 206,891 people with HIV included in total, of whom, 5,780 died after 2015.We estimated that women with HIV at age 40 years had 35·8 years of life left if they started ART before 2015, and 39·0 years left if they started ART after 2015. For men with HIV at age 40 years, the corresponding estimates were 34·5 years and 37·0 years. The expected remaining years of life at age 40 years for women in a comparator general population was 45·8 years and 40.7 years for men.
The estimated years of life left was much lower for both men and women with CD4 counts of fewer than 49 cells, than those with CD4 counts of at least 500 - there was an estimated difference of around 20 years.
For people with HIV on ART and with high CD4 cell counts, life expectancy was only a few years lower than that in the general population. However, for people with low CD4 counts at the start of follow-up, life-expectancy estimates were much lower. This shows the continuing importance of early diagnosis and sustained treatment of HIV.
Randomized trials from low- and middle-income settings suggested early initiation of antiretroviral therapy leads to higher mortality among people with HIV who present with cryptococcal meningitis. There is limited information about impact of antiretroviral therapy timing on mortality in similar people in high-income settings.
We included data from 190 people with HIV in Europe and North America who had never been on antiretroviral therapy, but had been diagnosed with cryptococcal meningitis between 1994 and 2012. We compared the effects of starting antiretroviral early (within 14 days of being diagnosed with cryptococcal meningitis), compared with started later (14-46 days after diagnosis of cryptococcal meningitis). We did this by using complex statistical methods to mimic a randomized controlled trial.
33 people died within 6 months after diagnosis of cryptococcal meningitis. There were 13 deaths among people following starting antiretroviral therapy and 20 deaths among those starting antiretroviral therapy later. We found little evidence that starting antiretroviral therapy early versus later was associated with higher mortality among people with HIV presenting with cryptococcal meningitis in high income settings. However, due to not having data on many people, our estimates were uncertain.
Impact of Hepatitis C cure on risk of mortality and morbidity in people with HIV after ART initiation. (Publication in press)
Many people with HIV also have hepatitis C. Hepatitis C affects the liver, which, over a long period of time, causes illness and death. However, effective treatment for hepatitis C is now available. We aimed to investigate whether people with HIV who had been cured of hepatitis C had different risks of death, AIDS, and non-liver related cancers compared to people with HIV who had never had hepatitis C.
We included 62,495 people with HIV on antiretroviral therapy. Of these, 2,756 acquired hepatitis C, and 649 were then successfully treated for hepatitis C. For 582 of these people with HIV who had been cured of hepatitis C, we were able to “match” with people with HIV who had never had hepatitis C. We “matched” these people who had been cured of hepatitis C by finding people who had never had hepatitis C who had similar characteristics, including length of time spent on antiretroviral therapy.
We found that people with HIV on antiretroviral therapy who were cured of hepatitis C soon after acquiring it, did not have a higher risk of mortality compared with people with HIV on antiretroviral therapy who had never had hepatitis C.
Over the past decade, antiretroviral therapy regimens that include integrase strand inhibitors have become the mostly commonly used for people with HIV starting antiretroviral therapy. We compared rates of all-cause mortality for adults with HIV on different modern antiretroviral therapy regimens. We were particularly interested in examining differences in mortality between people with HIV starting antiretroviral therapyon regimens that included integrase strand inhibitors with people with HIV starting antiretroviral therapy on regimens that did not containintegrase strand inhibitors.
We included 62,500 adults with HIV in Europe and North America starting antiretroviral therapy for the first time. 1243 (2.0%) died. Mortality rates were similar for people with HIV starting antiretroviral therapyon regimens containing dolutegravir, elvitegravir, rilpivirine, darunavir, or efavirenz. However, mortality was higher for people with HIV starting antiretroviral therapy on a regimen containingraltegravir compared with people with HIV starting antiretroviral therapy on regimens containing dolutegravir, elvitegravir, rilpivirine, darunavir, or efavirenz.
We found there was little evidence that rates of death were different between people with HIV starting antiretroviral therapy on most types of regimens. However, people with HIV starting antiretroviral therapy on regimens containingraltegravir had higher mortality than people starting on most other regimens.
UNAIDS wants each to have 90% of people with HIV on antiretroviral therapy to have the HIV virus suppressed. This is measured using HIV viral load and is defined by UNAIDS as having as less than 1000 HIV-1 RNA copies per millilitre. However, different countries use different thresholds to report viral suppression. Therefore, a method to adjust estimates from countries that use different thresholds to the <1000 threshold is needed.
We used data from people with HIV in Europe from the Antiretroviral Therapy Cohort Collaboration (ART-CC) and in seven global regions from the International epidemiology Databases to Evaluate AIDS (IeDEA) collaboration. We looked at the different distributions of viral loads for people with HIV on antiretroviral therapy. We developed different statistical models to try to capture these distributions and then checked whether they predicted the distributions in separate sets of data – called validation data.
We included data from 2010-2019 on 921,157 adults with HIV and 37,431 children with HIV. We found that the shapes of the distributions of the viral load results differed between regions. Overall, a model called the “reverse Weibull” was decided to best fit the data. For example, if a country reports that 80% of people with HIV on antiretroviral therapy were virally suppressed at <200 copies/mL, then 88.3% of people on antiretroviral therapy would be virally suppressed at the UNAIDS threshold of <1000 copies/mL.
This reverse Weibull model has been included in UNAIDS’s modelling software to allow people to use different viral suppression thresholds.
Information is lacking on survival outcomes and causes of death after cancer diagnosis among people with HIV.
We investigated causes of death within 5 years of cancer diagnosis among people with HIV on antiretroviral therapy in Europe and North America. Among 83,856 adults with HIV that started antiretroviral therapy between 1996 and 2015, 4,436 were subsequently diagnosed with cancer. Cancers were grouped as AIDS-defining malignancies, viral non-AIDS-defining malignancies, and non-viral non-AIDS-defining malignancies. We calculated the reasons of death among people with HIV after diagnosis of different cancers. We compared these 5-year survival rates with those in UK and French populations without HIV.
Of 603 deaths after diagnosis of an AIDS-defining malignancy, 292 (48%) were due to an AIDS-defining malignancy. Of 847 deaths after diagnosis of a non-viral non-AIDS defining malignancy, 467 (55%) were due to a non-AIDS-defining malignancy. This figure for viral non-AIDS malignancies was 74/189 (39%) of deaths. Estimated 5-year survival for people with HIV diagnosed with liver (around 29%), lung (around 18%) and cervical (around 75%) cancer was similar to in the general population. Survival after Hodgkin's lymphoma diagnosis in people with HIV (around 75%) was lower than in the general population.
Among people with HIV on antiretroviral therapy diagnosed with cancer, mortality rates and causes of death varied by cancer type, with mortality highest for liver and lung cancers. Deaths within 5 years of diagnoses of non-AIDS-defining malignancies were more likely to be from cancer than AIDS.
UNAIDS, which is part of the World Health Organization, uses a set of models to allow policy makers within countries to model the HIV epidemics in their countries. The assumptions behind these models need to be regularly tested to check that they are valid. We used data from the Antiretroviral Therapy Cohort Collaboration (ART-CC)’s European cohorts to validate mortality rates and estimates of excess mortality for people with HIV on antiretroviral therapy in high-income European countries in UNAIDS’s models.
We included 94,026 adults with HIV, of whom 5,515 died. All-cause annual mortality rates in UNAIDS's models for 2000–2003 were 0.0121, reducing to 0.0078 in 2012–2015, whilst the ART-CC's corresponding annual mortality rates were 0.0151 reducing to 0.0049. So, there was a steeper decline in mortality in the ART-CC than seen in UNAIDS's models.
The percentage of deaths that were AIDS-related in UNAIDS's models was 74.7% in 2000–2003, dropping to 43.6% in 2012–2015. In the ART-CC, AIDS-related mortality constituted 45.3% of mortality in 2000–2003 and 26.7% between 2012 and 2015, much lower than in UNAIDS's models. Excess mortality in the ART-CC was broadly similar to UNAIDS's models, dropping from 75.3% in 2000–2003 to 30.7% in 2012–2015.
Due to these results, it was decided that the all-cause mortality assumptions behind UNAIDS's models should be changed for people with HIV on antiretroviral therapy in high-income European countries. They were adjusted to be higher in 2000-03 and to decline more quickly after that.
People with HIV experience higher mortality compared with people without HIV. This is despite high levels of viral suppression for people with HIV on antiretroviral therapy. The Veterans Aging Cohort Study (VACS) Index was developed to include clinical biomarkers of general health with age, CD4 cell count, and HIV-1 RNA to predict mortality risk in a variety of populations of people with HIV. We investigated whether adding further biomarkers would improve how well the index predicts mortality.
We used data from the VACS cohort in the US to test which biomarkers to add. We then used data from the Antiretroviral Therapy Cohort Collaboration (ART-CC) in Europe and North America to test that these results could apply to different populations. For each person with HIV included (28,390 from VACS and 12,109 from the ART-CC), we took laboratory values from a randomly selected visit from 2000 to 2014. This visit was at least 1 year after antiretroviral therapy initiation.
7293 people with HIV from VACS died and 722 from the ART-CC. Including biomarkers for the highest CD4 cell count, CD8 cell count, and CD4:CD8 ratio before follow-up start did not improve how well the VACS Index predicted mortality. However, the addition of albumin, white blood count, and body-mass index, improved prediction of mortality in both VACS and the ART-CC. Results were consistent in nine ART-CC cohorts that were included and for different lengths of follow-up and all subgroups.
VACS Index 2.0 has now been developed, which added albumin, white blood count, and body-mass index to VACS Index version 1.0.
An increasing number of people with HIV now start antiretroviral therapy with high CD4 cell counts. We investigated whether this makes restoring CD4 and CD8 cell counts and the CD4:CD8 ratio to median levels seen in HIV-uninfected individuals more likely for people on antiretroviral therapy.
We determined median and values for CD4 and CD8 cell counts and their ratio using data from 2,309 individuals without HIV. We used measurements from 60,997 HIV-positive individuals from the Antiretroviral Therapy Cohort Collaboration.
When CD4 cell counts among people with HIV at the start of antiretroviral therapy were higher, higher long-term CD4 cell counts and CD4:CD8 ratios were reached. The highest long-term CD4 cell counts were observed in middle-aged individuals. During the first 2 years of antiretroviral therapy, median CD8 cell counts among people with HIV moved towards those seen in people without HIV. However, the changes after 2 years of antiretroviral therapy were small and long-term CD8 cell count levels were higher than median values seen in individuals without HIV. Median CD4:CD8 values did not reach the values seen in people with HIV, even when people with HIV started antiretroviral therapy with CD4 counts over 500 cells/mm.
Starting antiretroviral therapy with a CD4 cell count of ≥500 cells/mm makes reaching median reference CD4 cell counts seen in individuals without HIV more likely. However, median CD4:CD8 ratio values among people with HIV remained below the median levels of HIV-negative individuals. This was because of high CD8 cell counts.
Adherence to antiretroviral therapy is critical for successful treatment of HIV. However, comparisons across settings are difficult because adherence is measured in different ways. We examined using different adherence measures for identification of patients at risk of viral failure. Eight cohorts from Europe and North America contributed data from pharmacy refills or self-report questionnaires collected between 1996 and 2013.
For pharmacy data, we examined associations between the percentage of adherence during the 1st year of antiretroviral therapy with viral failure one-year after starting antiretroviral therapy. For self-report data, we examined 28-day adherence with viral failure based on closest viral load measure within 6 months after questionnaire date. Since adherence differed markedly by measurement type, we defined different cut-off points for pharmacy (lower <45%, medium 45–99%, higher 100%) and self-report (lower ≤95%, medium 96–99%, higher 100%) data.
Using the pharmacy data, people with lower and medium adherence had higher chance of viral failure compared with people with higher adherence. Using the self-reported adherence data, people with lower (but not medium) adherence had higher chance of viral failure than people with higher adherence. Both types of measure were strongly associated with viral failure. Although adherence measurements over longer time-frames are preferable for prediction, they are less useful for intervention.
Having HIV can lead to chronic inflammation and to an increased risk of death due to reasons other than AIDS. We wanted to investigate whether for people with HIV on antiretroviral therapy, AIDS-defining events were associated with increased non-AIDS related mortality.
We included 124,587 people with HIV who started antiretroviral therapy from 1996 to 2014, of whom 11,280 subsequently died. Of these deaths, 4,051 (36%) were due to non-AIDS related mortality.
People with HIV who had an AIDS-defining event had around twice the chance of dying from non-AIDS-related mortality than people with HIV who had not had an AIDS-defining event. Similar patterns were seen when splitting out the non-AIDS-related mortality into types of non-AIDS-related mortality. However, the risk of non-AIDS-related mortality differed depending on which AIDS-defining event a person had had (tuberculosis, pneumocystis, or non-Hodgkin Lymphoma).
These findings suggest that a common pathway may be independently driving both AIDS-related and non-AIDS-related mortality.
Antiretroviral pill count and clinical outcomes in treatment-naive patients with HIV infection.
Treatment guidelines recommend single-tablet regimens for people with HIV starting antiretroviral therapy. These regimens might be as effective and cost less if taken as separate drugs. We assessed whether the one pill once a day combination of efavirenz, emtricitabine and tenofovir reduces the risk of disease progression compared with multiple-pill formulations of the same regimen.
We selected people with HIV starting antiretroviral therapy on one-, two- or three-pill formulations of this regimen.
Among 11,739 people with HIV starting antiretroviral therapy, there were 386 AIDS events and 87 deaths. People often switched to the same regimen with fewer pills, after which we did not include their data in our analyses. After the first month, two pills rather than one was associated with an increase in the risk of AIDS or death, but three pills rather than two did not appreciably add to that increase. The analyses estimated that 77 people with HIV would need to be exposed to a one-pill regimen rather than a three-pill regimen for 1 year to avoid one additional AIDS event or death.
This particular single-tablet regimen is associated with a modest decrease in the risk of AIDS or death relative to multiple-pill formulations.
Health care for people living with HIV has improved substantially in the past two decades. We examined changes in 3-year survival and life expectancy of patients starting combination antiretroviral therapy in four calendar periods (1996-99, 2000-03, 2004-07, and 2008-10).
We included 88,504 people with HIV starting antiretroviral therapy. 2,106 died in the first year after starting antiretroviral therapy, whilst 2,302 died during the second and third years after starting antiretroviral therapy.
People who started antiretroviral therapy in 2008-10 had lower all-cause mortality in the first year after starting antiretroviral therapy than patients starting ART in the comparator period of 2000-03. A similar result was also seen for mortality in the second and third years after starting antiretroviral therapy.Life expectancy in 20-year-old patients starting ART increased by about 9 years in women and 10 years in men between 1996 and 2010.
Between 1996 and 2013, survival of people living with HIV in the first 3 years since antiretroviral therapy initiation improved substantially.Please see here for further details: http://www.bristol.ac.uk/news/2017/may/hiv-life-expectancy-.html
There is a lot of evidence that CD4 cell counts are useful to predict mortality among people with HIV, However, there is less evidence about whether CD8 cell counts and the CD4:CD8 cell count ratio are able to predict mortality. Therefore, we investigated whether CD4:CD8 ratios and CD8 counts were prognostic for all-cause, AIDS, and non-AIDS mortality in people with HIV on antiretroviral therapy that had high CD4 counts and were virally suppressed.
We included 49,865 people with HIV on antiretroviral therapy. Of these, 1834 died, 249 of AIDS-related mortality, 1076 of non-AIDS-related mortality, and 509 of these deaths were due to an unknown cause or were unclassifiable.
We found little evidence that CD4:CD8 ratio was prognostic for all-cause mortality after adjusting for other factors. CD8 count had a U-shaped association with all-cause mortality – people with low or high CD8 counts, had higher mortality than people with CD8 counts that were in the middle third. AIDS-related mortality declined as the CD4:CD8 ratio increased. AIDS-related mortality declined as CD8 counts decreased. There was little evidence of an association between non-AIDS-mortality and either CD4:CD8 ratio or CD8 count.
Overall, the size of the associations between mortality and both the CD4:CD8 ratio or CD8 count was too small for them to be useful as prognostic markers for people with HIV on antiretroviral therapy.
When a clinical treatment fails or shows suboptimal results, the question of when to switch to another treatment arises. Treatment switching strategies are often dynamic because the time of switching depends on the evolution of an individual's characteristics over time. Dynamic strategies can be directly compared in randomized trials. For example, people with HIV receiving antiretroviral therapy could be randomized to switching therapy within 90 days of their HIV-1 RNA crossing above a threshold of either 400 copies/ml (tight-control strategy) or 1000 copies/ml (loose-control strategy).
We reviewed an approach to emulate a randomized trial of dynamic switching strategies using observational data. We estimated the comparative effect of tight-control vs. loose-control strategies on death and AIDS or just death.
Of 43,803 individuals who started antiretroviral therapy, 2,001 were included in the mortality analysis and 1,641 were included in the AIDS or death analysis. There were 21 deaths and 33 AIDS or death events in the tight-control group, and 28 deaths and 41 AIDS or death events in the loose-control group. Compared with tight control group, the loose control group did not have very different changes of death or a combined measure of AIDS and death.
People with HIV have different subtypes of the virus. We aimed to compare mortalityfor people with these different subtypes of HIV. We investigated this among adults who started antiretroviral therapy between 1996 and 2012 using data from eight European and three Canadian cohorts of people with HIV.
Of the 20784 people with HIV included, the most common subtypes were B (15 419; 74%), C (2091; 10%), CRF02AG (1057; 5%), A (873; 4%), CRF01AE (506; 2.4%), G (359; 1.7%), and D (232; 1.1%). These subtypes followed patterns of region of origin for that person and how they acquired HIV. 1172 of these people with HIV died.
We found that people with HIV with subtype A had worse prognosis, an observation which may be complicated by socio-demographic factors.
A person's CD4 count when they start antiretroviral therapy is strongly associated with short-term survival. However, the association of CD4 count with longer-term survival is less well characterized among people with HIV. We included 37,496 people with HIV from 18 European and North American HIV cohorts who started antiretroviral therapy between 1996 and 2001. We estimated mortality rates by time since starting antiretroviral therapy and by CD4 counts.
Of the 37,496 people included, 6,344 died. The mortality rates were highest during the first 6 months on antiretroviral therapy, dropping by half to the period 5-9.9 years after starting antiretroviral therapy. It then dropped slightly further in the period 10 years after starting antiretroviral therapy. CD4 count when starting antiretroviral therapy was strongly predictive of mortality during the first year of antiretroviral therapy. However, this association became weaker the longer the period from having started follow-up.
After surviving 5 years of antiretroviral therapy, the rates of mortality of people who started antiretroviral therapy with low baseline CD4 counts became quite similar to the rates of mortality of people with HIV who started antiretroviral therapy with intermediate and high baseline CD4 counts.
People with HIV are now surviving for many years on antiretroviral therapy. Among people with HIV who started antiretroviral therapy between 1996 and 1999 who then survived for more than 10 years, we estimated mortality rates and what factors predicted mortality in the period after these 10 years. We also looked at what was the cause of death for people who died.
We included 13,011 people with HIV, of whom 656 died in the period beginning 10 years after first starting antiretroviral therapy.
CD4 counts measured when starting antiretroviral therapy did not predict mortality from 10 years after starting antiretroviral therapy. However, CD4 counts measured around 10 years after starting antiretroviral therapy did predict subsequent mortality. Similarly, detectable viral replication around 10 years after starting antiretroviral therapy also predicted subsequent mortality, as did older age (versus younger age), being male (versus female), having acquired HIV through injecting drug use (versus other modes of acquisition), and having had a prior AIDS event (versus not).
60% of people with HIV that were included had a five-year risk of death that was less than 5%. 30% of people aged 60 years of olderhad a five-year risk of death less than 5%.The most frequent causes of death (among 340 classified) were non-AIDS cancer, AIDS, cardiovascular, and liver-related disease.
