HIV and the immune response

Until the advent of protease inhibitors in the mid-1990s, estimates of the proportion of people with HIV who would develop AIDS and the time this would take to occur varied from 50 to 100% and from 18 months to 25 years. This wide range reflected individual response to HIV; knowledge concerning HIV transmission, pathogenesis, and natural history; and the availability and access to testing, monitoring, and treatment options. 

The monitoring of large cohorts over many years enables researchers to observe trends in disease progression and to compare predictions with actual rates of disease progression. In the era of highly active antiretroviral therapy (HAART), the following factors are associated with a poor prognosis after starting HIV treatment:

• CD4 cell count less than 200 cells/mm³.

• Viral load above 100,000 copies/ml.

• Age above 50 years.

• Injecting drug use.

Although the life expectancy for HIV patients in developed countries on HIV treatment had steadily increased since 1996, it is still roughly 10 years less than if the person were uninfected. Someone beginning antiretroviral treatment in 2003 or later, with >200 cells/mm³, can expect to live another 50 years as compared to 60 more years in the general population. A 35 year old beginning HIV treatment would be expected to live till age 72. If the person started treatment with <200 cells/mm³, those numbers would drop to 63 and 67 respectively. Naturally, there are differences between subgroups of people, specifically those who are injecting drug users.¹

Results from the SMART study also indicate that patients with lower CD4 cell counts also have a higher risk of serious illnesses that are not HIV-related.²

A French study, using data from nearly 2500 HIV-positive individuals, predicted that those people starting HAART with CD4 cell counts over 500 cells/mm³ would have the same rate of mortality as that found in the general population.³

One other study suggested that in the past several years, the rate of decline in mortality may have slowed as a result of late HIV diagnosis, treatment side-effects, poor adherence, and drug-resistance.⁴ Using participants from the CASCADE study, as compared to the general population in industrialised countries, showed that HIV infection itself did not increase the expected mortality rate in the first five years of infection. There is an excess of deaths in the first ten years for those aged 15 to 24 years.⁵

Scientific understanding in regard to immunological, virological, genetic, and other causes of disease progression, as well as a range of markers to help determine status and predict prognosis, broadens each year. Increasing use of this information is being applied clinicially, both in developed and resource-limited areas.⁶  

Various studies have reported contradictory findings regarding a possible acceleration of HIV disease progression among people infected more recently. There is some evidence that people infected in Italy during the 1990s had a more rapid rate of disease progression than those infected in the 1980s.⁷

Viral fitness (the rate at which HIV replicates) was reduced in a study of viral isolates taken between 2002 and 2003 and earlier isolates taken between 1986 and 1989.⁸ One thought was that the virus was mutating and becoming more aggressive. British researchers, looking at the same question, did not find any evidence of an increase in the rate of disease progression in the United Kingdom.

Patients who develop drug resistance are advised to continue antiretroviral therapy to prevent the re-emergence of wild-type virus. Drug-resistant virus is probably less fit than wild-type virus and therefore less pathogenic.

This recommendation is supported by the lack of an association between the number of resistance mutations and the incidence of opportunistic infections or death.⁹ People with drug-resistant virus also have less T-cell activation than people with wild-type virus not on therapy.¹⁰

The rate of CD4 cell count decline in people who become infected with a drug-resistant strain of HIV was similar to that of people with wild-type virus who were not on ARV therapy.¹¹ However, one study did find that use of any ARV drug before starting HAART was a risk factor for disease progression in the short-term.¹²

See Drug resistance for further discussion of drug-resistant virus.

The World Health Organization (WHO), British HIV Association (BHIVA), and US Department of Health and Human Services (US DHHS) guidelines base their recommendations on when to start ARV treatment on CD4 cell count, if available. CD4 T-cells direct immune response; their depletion in HIV infection limits the ability of the individual to fight off disease progression. They are, according to numerous studies, the key predictor of disease progression and mortality.

Certain specific subtypes of CD4 and CD8 cells appear to be much stronger indicators of the rate of disease progression than others. In a study of recent seroconverters among US military personnel, investigators found no relationship between disease progression and either the number or proportion of CD4 and CD8 memory cells. However, patients who had higher levels of naive CD8 cells were significantly less likely to progress to AIDS or die, even after taking into account CD4 cell count. Survival of CD8 memory cells, which was assessed by monitoring a marker called CD127, was also shown to be associated with a better prognosis (p < 0.003).¹³

Viral load is the other key predictor of disease progression. Higher viral loads during primary infection have been linked to faster decline of CD4 T-cells, as are the number of symptoms at seroconversion.¹⁴ The viral load 'set point', usually achieved within a year of seroconversion is also a fairly reliable predictor of long-term disease progression.

Standard viral load is a measure of the amount of HIV (as indicated by viral RNA) in the blood plasma. Studies have also investigated viral load levels purely within blood cells, as measured by cell-associated viral DNA. Several have found that viral load within cells also functions as an independent predictor of disease progression. In the study of US military personnel mentioned above, patients with higher levels of cells that carried viral DNA had significantly faster disease progression. This remained the case even after adjusting for other factors such as CD4 cell count, viral load and age.¹³ Another Dutch research team found that patients with significant increases in viral load within cells were significantly more likely to have CD4 cell count declines, while plasma viral load remained unassociated with CD4 cell counts.¹⁵

The question of whether patients with high HIV-1 DNA levels should start antiretroviral therapy earlier than other patients has been raised, but not definitively answered.¹⁶

The course of HIV disease appears to be strongly linked to Immune activation, the degree to which the immune system is actively engaged in fighting infection. Investigators are exploring several markers of the degree of immune activation to see which may be most predictive. One such marker, high levels of CD38 expression by CD8 T-cells, has been found strongly prognostic of disease progression, possibly more so than viral load. These findings suggest that T-cell activation is a key determinant of survival time.¹⁷

Another study supporting an association between immune activation and disease progression found that in people not receiving ARV treatment, a low dose of the immunosuppressive drug prednisolone (which suppresses T-cell activation) slightly increased CD4 cell counts over three years. Patients not receiving prednisolone in this study experienced a substantial drop in CD4 cell counts over the same time period.¹⁸  

HIV-specific T-cell responses may play a role in determining the rate of disease progression. Slower falls in CD4 cell counts have been associated with stronger HIV-specific CD4 and CD8 T-cell responses.¹⁹

A study of recent serocoverters in the US military found that higher levels of another marker of immune activation, the Ki-67 marker in CD4 and CD8 cells, was associated with a significantly higher risk of disease progression.¹³

Researchers in Paris compared a number of immune function markers in long-term non-progressors. While they found that HIV p24-specific CD4 and CD8 T-cell responses were negatively correlated with the amount of HIV integrated into T-cells, the two immune factors that predicted if a person would still be a long-term non-progressor after nine years of infection were the presence of HIV p24-specific CD4 T-cell responses and presence in the blood of HIV-gp41 antibodies belonging to IgG2 subtype.²⁰

Throughout the continuum of HIV disease, R5-tropic virus (M-tropic virus using the CCR5 co-receptor) is likely to predominate, certainly in the period before a diagnosis of AIDS is made. R5-tropic virus is estimated to cause 90% of HIV-1 infections and produces much more virus per infected cell than does X4-tropic virus (T-tropic virus using the CXCR4 co-receptor), probably one reason why it predominates for much of the course of infection.^{21 22} 

However, virus that uses the CXCR4 co-receptor can become dominant and this development is associated with a much more rapid loss of CD4 T-cells. This co-receptor switch is accompanied by a change from non-syncytium-inducing (NSI) to syncytium-inducing (SI) virus, resulting in increased cell death in large numbers of uninfected CD4 T-cells. NSI strains preferentially infect macrophages; SI strains are T-tropic and use the CXCR4 receptor. One study published in 1998 found that, in a six-year period, NSI virus was associated with a 78% survival rate and SI virus with a 21% survival rate.²³ 

In 2007, a US study indicated that the presence of dual/mixed (DM or R5/X4)-tropic virus versus exclusive R5-tropic virus sped disease progression by a factor similar to a ten-fold (1 log) increase in viral load for participants who began the study with a median duration of HIV infection of four years and who were followed for an additional 52 months. Trofile assay results for over 300 of the participants indicated that at baseline, 90% had pure CCR5-tropic virus, 10% dual/mixed tropic virus, and none purely X4-tropic. Baseline CD4 count was 450 cells/mm³ or more. There was twice the likelihood of progressing to the study endpoint (either decline in CD4 count to <350cells/mm³, ARV initiation, or death) in those with dual/mixed R5/X4-tropic virus.²⁴ The rate of progression between the two groups following ARV initiation is not noted. 

Similarly a British review of viral tropism in 400 patients found that in the 12 months prior to starting antiretroviral therapy, patients infected with X4 or dual/mixed-tropic virus experienced a significantly greater decrease in CD4 cell count and more clinical events than did patients with R5-tropic virus. However, in the two years after the start of ARV therapy, time to viral suppression and the proportion of patients achieving viral suppression were similar at 6, 12, and 24 months, as were CD4 cell count increases, regardless of tropism. The group infected with X4- or D/M-tropic virus did have a 2.56-increased relative risk of experiencing a clinical event in the 24-month follow-up period.²⁵

These findings would seem to indicate a role for early tropism testing, not only to guide ARV choice, but also to alert clinicians to a possible need to either initiate ARV therapy earlier in patients with D/M-tropic virus or schedule more frequent viral load and CD4 cell count monitoring. 

Disease progression varies with the subtype of HIV-1. In women, AIDS incidence rates of 3.5 per 100 person-years have been reported with subtype A, 16.0 for subtype D, and 14.5 for subtype C.

A Ugandan study found that individuals infected with subtype D were 29% more likely to die of AIDS or experience CD4 cell count declines than those infected with subtype A.²⁶ 

Similarly a study of ARV-naive Kenyan women reported in 2007 indicated that women with HIV subtype D had more than twice the risk of death of women with subtype A over a six-year period. Of the nearly 150 women who seroconverted during the study period (1993 to 2004), 78% had subtype A, 14% subtype D, and 7% subtype C. There was no significant difference in viral load at any time, but CD4 counts declined faster and there was a 2.7 increased relative risk of mortality among the women with subtype D virus versus those with subtype A. Researchers commented that a previous study had found that subtype D more easily shifted tropism from the CCR5 to the CXCR4 co-receptor.²⁷ 

In a study of co-receptor use in 68 ARV-naive pregnant Ugandan women, nearly half of the group had subtype A or A/D recombinant virus and no-one had CXCR4-tropic virus. Nine of the 25 subtype D viruses studied showed dual or mixed tropism (36%), a much higher percentage than is found in individuals with subtype A or B virus. This finding has clinical implications for treatment of individuals with subtype D infection.²⁸ 

Variations or mutations in HIV’s genes also affect the speed of disease progression. Some cases of long-term non-progression and slow progression have been linked to alterations in the genes, env, vpr, and nef.^{29 30 31}

Infection with more than one strain of HIV from the same or a different subgroup can occur. This is known as dual infection or superinfection. Cases of superinfection have been definitively established, but there is limited evidence about their frequency. Documented instances of superinfection have been associated with as much as a threefold increase in disease progression as well as failure of antiretroviral therapy. Of the documented cases of superinfection, most occurred during primary infection, but some up to five years after that time.^{32 33}  

A sudden viral load increase seen in the first year or two after primary infection is more suggestive of superinfection than viral increases seen later.^{34 35}

Recent evidence suggests high levels of neutralising antibodies may give some protection against superinfection. Individuals who are repeatedly exposed to new strains of HIV and who have high levels of broadly cross-neutralising antibodies seem to avoid superinfection.³⁶