HIV Tests on Postmortem Specimens

“Cadaveric serum samples and premortem serum samples of 33 cornea donors were collected between November 1996 and February 1998 and tested comparatively for HIV and HCV antibodies, and hepatitis B surface antigen (HBsAg)…two HBsAg positive potential donors were identified corrrectly by the strategy of screening cadaveric sera. On the other hand, 17 or 33 serum sample pairs had divergent results in at least on of the serologic tests…In 16 of 17 potential donors with divergent results…the results of the HBsAg assay were positive in the cadaveric, but negative in the premortem serum sample. Furthermore, the HIV ELISA was positive in two cadaveric serum samples, whereas the corresponding premortem serum sample was HIV antibody negative. In one of these two cadaveric serum samples, the sample addition verification system of the HIV antibody assay indicated inappropriate conditions”

“It has long been suspected in test laboratories that considerable number of false positives are observed with samples from cadaveric donors…The FDA does not require kit manufacturers to submit data showing that HIV-1 antigen and antibody detection kits produce accurate results when applied to cadaveric blood samples. Little data exist regarding the performance of these test kits with grossly hemolyzed or cadaveric specimens…[In this study] 31 hemolyzed cadaveric blood specimens were tested for HIV-1 antigen using both the Abbot EIA [ELISA antibody test] and the Coulter EIA test kits. 27 of the sera were initially reactive [positive on the first test] and 26 [83.9%] repeatedly reactive but not confirmed [negative western blot?] when tested with the Abbott assay. Nonspecific reactivity was seen with one specimen (3.2%) using the Coulter assay [the authors later claim that, with the unspecified confirmation step all tests perform equally well, but there was no information on how the confirmation step was validated].”

“Of the 456 samples tested, HIV-antibodies were unequivocally detected in 21 samples (4.6%) by the Testpack…17 other Testpack results (3.7%) could have been considered to be positive, that is, false positive (negative in Western Blot). Usually the color-development, positive as well as negative reaction, is completed within 10-15 seconds, which is much less than the recommended two minutes. But in these 17 cases a faint vertical bar occurred beside the strong horizontal ("negative") bar after a period of 1-2 minutes. In this way of reading a HIV-positive result may be suggested. However, these results were considered to be HIV-negative, which was confirmed by Western Blot…In our opinion, a number of 3.7% inconclusive and potentially false positive Testpack results (as well as 3.1% in ELISA) is acceptable, since the consequence of false positive or inconclusive results in autopsies would lead only to special precautions of the medical staff [unless the death was claimed by some to be due to deliberate or negligent infection, in which case false positive results could have very important ramifications!]”

“Sera from 207 forensic autopsy cases were tested for the presence of antibody to the human immunodeficiency virus (HIV) using a recombinant immunoblot assay (RIBA)…Out of these cases, 172 autopsies were of individuals with no known risk factors for HIV infection, and of these, 169 had no detectable antibodies to HIV. In 2 cases, the serum reacted with p24 alone on the RIBA-HIV216 assay, but these results were not confirmed by further testing and are considered to be false positive reactions. In 1 case, the serum reacted only with gp41 on the RIBA-HIV216 test but was nonreactive with further testing. This result has been designated equivocal. Of the 35 cases at high risk of HIV infection, 4 had antibodies to HIV detected in postmortem serum samples [and only 2 had previously tested HIV-positive. No pre-death samples were tested]”