Infectious Disease Postdoc/Clinician
Department of Pediatrics, University Hospital, Syracuse NY
Send response to journal:
In reply to Julian Turningheart's point, perhaps the answer lies in the question. It seems to me that the "capraciousness" of the HIV antibody test is anything but. Nucleotide tests are also possibly affected by non-specific binding, and are more affected by accidental contamination. I'm not aware of any one modality being far superior to another in this regard. The individual test components may be imperfect, but no more so than any other diagnostic test. This is why there are multiple stages involved in coming to a definitive diagnosis. To turn the concept around, since serology, culture and PCR correlated almost 100% in the Jackson and Ho papers previously quoted, how can we decide which to use? There were two seropositive patients of note in the cohorts studied: one was PCR+ culture- and the other was culture+ PCR-. If anything, serology seems superior.
As such there is no pressing need to develop general diagnostic testing kits based on nucleotide detection. Since no tests have been developed, none have FDA approval, hence it is forbidden by Federal Law. Certain DNA-based tests are permitted for confirmation of infection - HIV RNA-based monitoring tests have been shown to produce rare false-positives if used as a diagnostic test, so they suffer from more specific warnings about being used inappropriately.
In the field of blood donor screening there is a need for more rapid detection of HIV infection, since anti-HIV antibodies may not appear for several weeks post-infection. This area clearly has a need for nucleotide -based tests. Such tests have been developed, validated and are FDA- approved.
There is always a gap between what can be physically done in the lab and what can be legally permitted in the clinical setting. A procedure (DNA provirus detection) is not the issue at stake, but whether a particular kit manufactured by a particular manufacturer has been created, validated and approved for clinical use. Identical reagents sold for "research purposes" may not be used for clinical diagnostics in any way, as stated in their package insert. This is to safeguard the patient, rather than make a statement about the validity of the technique itself. Several non-FDA approved diagnostic kits are on the market using technology similar to the other serological tests. They are unapproved not because serology testing is unreliabe, but because that particular kit has not been validated.
A list of HIV (and HTLV and Hepatitis C) testing kits currently approved by the FDA can be found at:
The list includes 8 HIV nucleic-acid based tests of which 3 are for donor screening, with the remaining 5 being only permitted for use as monitoring tools (viral load and mutation analysis). Much of the test paperwork is available on the website, including false-positive validation.
A FAQ covering the use of unapproved HIV tests is at:
It includes links to articles describing the successful prosecution of an indivdual who distributed a useless non-validated unapproved test kit.
With the growing use of rapid nucleotide-based diagnostic methods in the hospital setting, especially with regard to on-site validation of novel diagnostic procedures, it may well be that a new market will open up in molecular diagnostics for HIV. For now however there is no market and no kit has been developed, to my knowledge. If you want to allow non-FDA approved/exempt research equipment and reagents to be used in clinical diagnosis then that's your choice, but not one of many I suspect!
Almost certainly not the choice of the doctor who may get sued for an incorrect diagnosis, since by using an unapproved test he would have no defense.
I appreciate that the examples given here are based on the American system, but the guidelines used by NICE in the UK are similar, and the questioner is from Tennessee.
Nick Bennett firstname.lastname@example.org
Competing interests: None declared