Abstract Body

Background: We have previously reported a cohort of 86 HIV-1 “controllers” and defined a subset, “discord controllers (DC)” with low or declining CD4 counts (<450) despite control of plasma viral RNA. We showed that DCs are distinct both clinical and immunologically, having depletion of naïve CD4 cells and higher activation in all CD4 subsets compared with typical controllers (TC). Data are scarce on clinical management and outcome of controllers so we undertook a follow-up study of the cohort.

Methods: HIV controllers recruited into the prospective cohort were designated DC or TC depending on the geometric mean titre of the last 3 CD4 counts. Controls were HIV non-controllers and uninfected individuals. Baseline HIV-1 DNA load in peripheral blood mononuclear cells was determined by quantitative PCR and expressed as per cell equivalent.

Results: 18 DCs were recruited; 2 are lost to follow-up. DCs had higher DNA loads (13-1529, median 601 copies/106 CD4 cells) compared to TCs (0-755, median 87) (p=0.002) and were similar to those in non-controllers (27-2188, median 852). Ten DCs had received 12-66 (median 42) months of antiretroviral therapy (ART). RNA loads were 85-19837 (median 796) at ART initiation and all became undetectable on therapy. However, CD4 gain was modest; from baseline 163-308 (median 272), CD4 change was -25 – +318 (median +130) over follow-up. Those with lower nadir CD4 had lowest CD4 gain despite extended ART. Of 6 ART naïve patients, only one remained undetectable. CD4 counts in the naïve patients were unexpectedly low at 217-464 (median 304). Of note, 5 patients had declined to start ART despite low and declining CD4 counts.

Conclusions: The significantly higher DNA loads in the DC group suggest productive ongoing HIV-1 replication and are compatible with increased immune activation, poor clinical outcomes and sub-optimal CD4 response to ART as described in this, and other, controller cohorts. Initiation of ART occurred late in DCs (some declined treatment), suggesting that both clinicians and patients may feel falsely reassured by low RNA loads in the face of low CD4 counts. These results suggest that controllers may benefit from earlier ART and that clinicians should remain vigilant to this despite low or undetectable RNA loads. In addition, DNA load may be a better marker of viral replication and disease progression than RNA load and may identify controllers in whom early ART is indicated. Ongoing longitudinal follow-up of this cohort is planned.