March 4th, 2009

TaqMan HIV RNA Assay: Be Careful What You Wish For

At our hospital lab, we recently switched from the bDNA viral load assay to the new Roche TaqMan real-time PCR test.  The virologist in charge of our lab and the tech both agreed the assay was more accurate, more sensitive, and easier to do — so much so that we could increase the frequency of the test being run, a huge benefit for patient care.  I was all for it.

So what’s the problem?  As noted here in this paper from Vancouver — and anecdotally throughout the world  — there’s a new epidemic of intermittent low-level viremia picked up by this assay.  Every week, another one of my long-term successfully treated patients has a result come back at 125, 94 or — most cruelly — 49.  (The lower limit of the assay is 48.)

Plus, there’s that issue of “<48 (viral RNA detected but below the quantifiable range of the assay)” vs “Target not detected.”  Try explaining that to an obsessive patient who heretofore had been happily receiving “undetectable” viral load results through a decade of meticulous pill taking.

One view is that these are not “false positives”, but rather represent actual detection of virus though this more sensitive assay.  (The Vancouver group is not so sure — read the discussion section in their paper.)

But if these small detections of HIV RNA are real, is there a clinical significance to it?

Based on the kinds of patients in whom we are seeing these results — many on first regimens, many with no history of treatment failure — I suspect the answer will be no.  These low-level detectable results could be analagous to the low-level detection picked up by the single-copy assay — just a harmless release of virus from long-lived chronically productive cells, and not actual viral replication.

Or to quote my esteemed colleague:

I believe we will eventually come to recognize two related but distinct situations–patients may be “fully suppressed” — that is, have no actively replicating virus — but may or may not be “aviremic”.

So I feel a bit better.   Nonetheless, if you’re planning on introducing TaqMan into your clinic, prepare yourself for some significant patient education, reassurance — and stress management.

6 Responses to “TaqMan HIV RNA Assay: Be Careful What You Wish For”

  1. Shiva Saboori says:

    Thanks for your words of caution. Recently our hospital switched to a new laboratory which turns out uses this more sensitive test. We were confused by the wording of detectable but too low to quantify result and after e-mailing back and forth with the HIV/Virology contact at the new lab understood that this was due to the enhanced sensitivity of this assay!
    Lovely! What do we tell our patients?
    After discussing it amongst our group we decided to Phrase it as “< 48” rather than saying “undetectable” as we used to before. We explain that “undetectable” does not mean there is no virus around. This test is just more sensitive in “detecting” that.
    We have not presented to the patient the results of “virus detectable but not quantifiable” being different from “virus undetectable”, nor have we done any more frequent testing on the latter pts. It is nice to read that you feel there is likely no clinical diffrence between the two.
    It is an elegant balancing act, this imparting of information to fellow humans of different backgrounds, education, and pshychies.

  2. Shawn says:

    I’m a little confused. If the ability to detect lower quantities (copies?) of the HIV virus exists, doesn’t the clinical significance lie in being able to more accurately track the progression of treatment?

    In other words, if one test has a lower quantifiable range, say 10 copies and another test has a low-end of 3 copies, isnt the test that detects 3 copies of the virus more accurate and thus a better way to track ART progress?

    Maybe I missed the point entirely…

  3. Paul Sax says:

    The new tests do seem to be more accurate, and more sensitive. The question is whether this added knowledge makes monitoring of HIV therapy better for patients. So far, the vast majority of time we encounter one of these “blips” in clinical practice, we don’t change treatment since it’s not true treatment failure.

  4. Tom says:

    Well I’m one of those patients whose now experiencing VL of 60 (repeat <48),61 (repeat <48),102 (repeat at 52) and now 71. Had nearly a 15 year track record of undect. With 27 mutations (thank you monotherapy) I don't have much room to maneuver. When do readings like this indicate emerging resistance vs these more sensitive tests?

  5. Paul Sax says:

    Tom, these results area tough thing to get used to, for all of us. But the bottom line is that these tests are in fact more sensitive, so I do believe the results are “real”. The clinical relevance is likely to be small unless the result is > 200 on two consecutive measurements (a paper will be coming out on this very soon).

    One of my colleagues uses the older and less accurate bDNA assay because he just doesn’t like getting these results. Maybe he’s on to something!

  6. Shawn says:


    So, the more sensitive the test is does not necessarily mean its better for monitoring ART?

    Do you think there are any other benefits of more sensitive PCR tests?

HIV Information: Author Paul Sax, M.D.

Paul E. Sax, MD

Contributing Editor

NEJM Journal Watch
Infectious Diseases

Biography | Disclosures | Summaries

Learn more about HIV and ID Observations.