Undetectable Without Ever Taking a Pill
Seventeen years without a single antiretroviral dose and an undetectable viral load the whole time is either a reason to leave well enough alone, or exactly the kind of unusual biology the team can't yet fully explain — and that uncertainty cuts both ways.
S.K., a 51-year-old woman, was diagnosed with HIV at 34 during a routine screening tied to a life-insurance application, a diagnosis that arrived with no symptoms and, to her own considerable surprise, no detectable viral load on repeat testing either. She lives alone in a hundred-year-old farmhouse she bought and has been slowly restoring by hand ever since, a project she describes as having taught her more patience than anything else in her life. She has worked as a public librarian for over twenty years, a career she says has given her an appreciation for how much is still genuinely unknown even in fields, like her own field of information science, that look settled from the outside — a framing she brings deliberately to her own HIV care as well. She has never taken antiretroviral therapy in the seventeen years since her diagnosis, by her own informed choice, confirmed repeatedly at every visit since. She describes herself as neither anti-medication nor waiting for a better option — simply someone who has never seen a compelling individual reason to change course.
Her viral load has remained below the limit of detection at every check across those seventeen years, and her CD4 count has stayed stable in the 800-900 cells/µL range throughout — genuine, durable elite-controller status, not a fluke of one or two lucky lab draws. What's brought this conversation back to the table is newer literature she raised herself, having read about it: the ANRS CO21 CODEX cohort found that even elite controllers carry persistently elevated markers of immune activation compared to both ART-treated patients and HIV-negative controls, and that activation has been linked to a higher rate of non-AIDS events over time. START, the trial most responsible for the current universal-treatment-recommended standard, found ART benefit essentially regardless of CD4 count — but it enrolled on CD4 count alone, with no virologic entry ceiling, and its participants arrived with a median viral load near 12,800 copies/mL against her own undetectable one. Nothing in its protocol shut her out; the cohort simply never contained anyone like her in numbers, which is a weaker fact than an exclusion criterion and, for her purposes, an equally binding one. She has no cardiovascular risk factors of her own to date, which she points out herself as a reason she isn't approaching this from a place of alarm.
HIV clinic, long-term follow-up visit
Her viral suppression has never really been the open question — it's whether suppression alone is the whole story. The ANRS CO21 CODEX cohort found elite controllers carry persistently elevated immune activation markers compared to both treated patients and HIV-negative controls, and that activation independently predicts non-AIDS events. START showed treatment benefits patients essentially regardless of CD4 count. I think that argues for offering her ART now.
The inflammation data is real, and I'm not dismissing it.
But I want to be precise about what START does and doesn't say, because the version I hear repeated is stronger than the truth. It didn't have an exclusion criterion for controllers. It enrolled on two CD4 counts above 500 and said nothing about viral load, and the median participant walked in at roughly 12,800 copies/mL. So its universal-benefit conclusion simply wasn't tested in anyone like her — Li and Sax make exactly that point when they call ART's benefit in controllers uncertain. Applying it to her is an extrapolation, not a direct reading of the trial. And CODEX describes a cohort average; her own seventeen years of durable, unmedicated control is real individual evidence about her own biology that a population-level signal doesn't automatically override. I'm genuinely uncertain here, not confidently against treatment — I just don't think either study settles it the way a direct trial would.
I don't think either of you is wrong, and I don't think this is actually a question the two of you can settle for her. Neither study directly answers it — one never enrolled anyone like her, the other describes a cohort average rather than her specific course. She's navigated this exact decision herself, with full information, for seventeen years. My role today is laying out both studies honestly, the way you each just did, and then asking her directly what she wants to do with it — not landing on a recommendation before she's even heard the full picture.
Genuinely unresolved, and left that way deliberately: the team presented both the ANRS CO21 CODEX inflammation data and START's own scope and exclusions to S.K. directly, without converging on a single recommendation first. She asked for time to think it over rather than deciding at this visit — a request the team treated as the appropriate outcome given how genuinely unsettled the underlying evidence is for someone with her specific, unusual course.