An ongoing dialogue on HIV/AIDS, infectious diseases,
February 3rd, 2012
More on Low (but Detectable) Viral Loads — Is Knowing This Useful?
I have a very smart, very experienced colleague — clue, his initials are CC, and he doesn’t pitch for the Yankees — who continues to use bDNA testing for HIV viral load monitoring. You know, the assay with a lower limit of detection of 75 copies.
He knows that bDNA is less sensitive than PCR.
He knows that it’s more expensive than PCR.
He even knows (I think) that it’s less accurate than PCR.
So why does he use it? Because he detests the added anxiety and aggravation that these periodic low-level results — usually 20-200 copies, or “<20 but target detected” — give his patients. More importantly, he’s not convinced it provides him with any useful information.
I certainly get that. And suspect he’s not alone in using this (ancient) test, which is probably best known for a study that put viral load testing on the map way back when.
But I can’t bring myself to use an inferior and more expensive test, so I’ve switched whole-hog to PCR. As a result, I’ve been forced to learn a whole new speech to give to patients when these results come back. It usually goes something like this:
Gerald [not his real name], the result came back at 43… Yes, this is detectable, but remember our old test only went down to 50, so in fact this is a great result — keep up the good work … No, I’m not worried … If you’re worried we can repeat it, but I can assure you I won’t recommend changing meds even if it comes back the same … Yes I’m sure … You’re welcome.
Now it’s been a few years since we’ve had these tests, and several studies (like this one) thus far didn’t even suggest any significance to these low-level detectable results.
Until now.
Over in Journal Watch AIDS Clinical Care, Helmut Albrecht summarizes a recent study that compares the likelihood of virologic rebound in 1247 patients who had viral loads measured by the RealTime PCR assay. Three groups were defined: those with results between 40-49 (Group A), < 40 but detectable (Group B), and truly undetectable (Group C):
The proportion of patients who experienced viral rebound to >50 copies/mL was 34.2% for group A, 11.3% for group B, and 4.0% for group C. The proportion with rebound to >400 copies/mL was 13.0%, 3.8%, and 1.2%, respectively. These associations were independent of adherence levels.
Based on these data, it does seem that the lower the viral load, the better — but wait!
Is there more resistance among those who rebounded? No.
Could this just be a proxy for duration of virologic suppression? Highly likely.
Is there a different clinical outcome among those who rebounded? Not commented on, but highly doubtful.
Is there any evidence that our management should be changed based on these fascinating results, however biologically plausible they may be? Emphatically NO. Or at least, not yet. Good summary of these issues in Raj Gandhi’s accompanying editorial.
Which is why CC can continue to use the bDNA — it’s defensible — and while I’ll press on with PCR and all it’s telling us about low-level viremia. Knowledge is power, after all.
I’m just not sure what to do with it yet.
One Response to “More on Low (but Detectable) Viral Loads — Is Knowing This Useful?”
Paul E. Sax, MD
Associate Editor
NEJM Clinician
Biography | Disclosures & Summaries
Learn more about HIV and ID Observations.
FROM NEJM — Recent Infectious Disease Articles- Minocycline-Induced Hyperpigmentation April 2, 2026A 68-year-old woman with rosacea presented with a 6-week history of dark patches on the skin of her arms and legs. Two weeks before the onset of the skin changes, she had started taking minocycline daily.
- Legislating Medicine — Directed Donation and the Politics of Patient Choice April 1, 2026A Tennessee bill focused on directed blood donation exemplifies a pattern of efforts to legislate medical practice in ways that override scientific consensus while invoking the language of autonomy.
- Probable Japanese Encephalitis Virus Transmission through Organ Transplantation March 26, 2026Infection with Japanese encephalitis virus, a mosquito-borne flavivirus, was identified in a patient who received a liver transplant from a donor who resided in California.
- Communicating about Vaccines in a Politically Contentious Climate March 26, 2026Recent changes in U.S. vaccine policy are sowing confusion and threatening the health of children and the wider population. How should pediatric clinicians communicate with parents to stem the damage?
- Eczema Herpeticum March 26, 2026A 33-year-old pregnant woman with previously controlled atopic dermatitis presented with 3 days of fever and an itchy, painful rash. Erythematous vesiculopapular lesions were noted on the face, neck, chest, and arms.
- Minocycline-Induced Hyperpigmentation April 2, 2026
So if the point of the virus suppression threshold is to prevent resistance – as is summarized in the editorial – these data support that all three of these thresholds accomplish this task equally well. And I’d predict that if they included a bDNA comparator – those who are <75 have a similarly low resistance rate. If true – it strengthen the case that going for undetectable isn't the goal – durable suppression is the goal. With whatever numeric predictor there is that helps discriminate who is most likely on that path. As assays evolve, it has again become important to ensure we know which values predict durable suppression. It could be interesting to see if inflammatory markers differ between these degrees of suppression. And there may yet be differences between drug classes and their ability to maintain suppression over the long term, despite periodic low level viremia. Nevertheless, the current relevance of this endpoint is manifest by the European Protea trial assessing monoRx with a boosted PI vs more standard regimens – it has the primary endpoint as loss of future drug options, as opposed to historic virologic criteria that predicts that event.