Demographics and natural history of HIV-1-infected spontaneous controllers of viremia

Abstract

Objectives: HIV-1-infected persons spontaneously controlling viremia without treatment (SCV) are rare. Sex and race effects on prevalence and outcome are poorly defined, and it is unclear whether SCV qualitatively or quantitatively differs from typical infection. These issues are examined in this article.

Design: Medical records of 46 524 persons receiving outpatient care for HIV-1 infection were reviewed. Of these, 29 811 had adequate viremia testing for SCV screening.

Methods: SCV was defined as at least three consecutive plasma viremia measurements <50 RNA copies/ml spanning at least 1 year without treatment. SCV loss was defined as at least three consecutive viremia measurements ≥50 RNA copies/ml or one ≥1000 RNA copies/ml. Demographics of persons with SCV were compared with the total population. Viremia and blood CD4 T-cell levels during SCV were compared between demographic subgroups and persons who maintained or lost SCV during observation.

Results: In total, 53 persons (0.18%) met SCV criteria. Prevalence was higher for women versus men and blacks versus whites; these appeared independent. Loss of SCV was observed at 1.22% per year, and significantly associated with viremia 'blips'. Blip magnitudes fit log-normal distributions with means below 50 RNA copies/ml.

Conclusion: Our novel observation of higher SCV prevalence in women and blacks is consistent with prior studies of typical chronic infection. Viremia blips correspond to greater risk of loss of SCV, likely reflecting higher set-point viremia under the limit of detection. Our findings suggest that SCV represents an extreme along a continuum of HIV-1 infection, and not qualitative difference.

Figure 1. Frequencies of SCV across different demographics of HIV-1 infected persons

The frequencies of persons meeting SCV criteria in relationship to the total numbers of persons screened for each of the indicated groups are plotted. Numbers are indicated below each bar. Only Black and White race and female and male gender are shown, given the small numbers of other races and transgender persons. The bars indicate 90% confidence intervals (Jeffreys intervals).

Figure 2. Progression to viremia among 53 persons with SCV

Kaplan-Meier curves for maintenance of SCV in relationship to estimated years of infection are plotted, defining loss of SCV defined as a single plasma viremia measurement >1000 RNA copies/mL or three consecutive plasma viremia measurements >50 RNA copies/mL. Persons lost to follow-up or started on antiretroviral therapy during SCV were censored. Y-intercepts for all linear regressions were set at 100%. A. The plot for the entire group is given, with shading of the 80% confidence interval. Linear regression showed r²=0.93, slope = −1.22%/year. B. The plots for persons identified as Black (33 persons) versus White (17 persons) are given. Linear regressions showed r²=0.93, slope = −1.20%/year versus r²=0.71, slope = −1.03%/year respectively. C. The plots for the younger (26 persons, mean age 26.9 ± 4.4, range 17.4 to 34.7 years) versus older (27 persons, mean age 44.3 ± 7, range 35.1 to 62.5 years) halves of the group are given. Linear regressions showed r²=0.93, slope = −1.37%/year versus r²=0.79, slope = −0.98%/year respectively. D. The plots for the women (26 persons) versus men (24 persons) excluding transgender persons are given. Linear regressions showed r²=0.92, slope = −1.38%/year versus r²=0.29, slope = −0.53%/year respectively. E. The plots for the women including transgender men to women (26 and 3 = 29 persons) versus men (24 persons) excluding transgender persons are given. Linear regressions showed r²=0.94, slope = −1.58%/year versus r²=0.29, slope = −0.53%/year respectively. F. The plots for women (26 persons) versus men including transgender men to women (24 and 3 = 27 persons) are given. Linear regressions showed r²=0.92, slope = −1.38%/year versus r²=0.85, slope = −1.10%/year respectively. G. The plots for persons with detectable (≥50 RNA copies/mL plasma) “blips” of viremia (18 persons) versus those without (35 persons) are given. Linear regressions showed r²=0.85, slope = −2.27%/year versus r²=0.93, slope = −1.23%/year respectively (p=0.040).

Figure 3. Magnitudes of intermittently detectable viremia during SCV

Values for viremia testing demonstrating detected “blips” of viremia during SCV are plotted over estimated duration of infection for those with sustained SCV (A) versus non-sustained SCV (B) during the observation period. For those with non-sustained SCV, values are also plotted in relationship to the time of loss of SCV (C), defined as a single value >1000 or three values >50 RNA copies/mL plasma. Viremia measurements before qualification of SCV were excluded, to avoid tests before stable viremia was established in chronic infection.

Figure 4. Frequencies of “blips” of detectable viremia during SCV

A. The frequencies of detectable viremia tests (≥50 RNA copies/mL plasma) are plotted for persons with sustained (black bars) versus non-sustained (gray bars) SCV for five-year intervals after estimated time of infection. B. The frequencies of blips according to magnitudes are plotted for both groups. Viremia measurements before qualification of SCV were excluded, to avoid tests before stable viremia was established in chronic infection. *Not plotted because <6 viremia tests were available.

Figure 5. Blood CD4⁺ T cell levels in persons with SCV over time

A. Peripheral blood CD4⁺ T-cell levels are plotted over time for all persons during SCV follow-up (before loss of SCV in the case of those with non-sustained SCV) for the 40 persons with at least two years of measurements during SCV (33 with sustained SCV during observation and 7 with loss of SCV during observation). B. The distributions of observed slopes of blood CD4⁺ T-cell levels are indicated. The mean values for those with sustained versus non-sustained SCV were −3.0 and −18.7 cells/µL/year respectively.