Send response to journal:
In response to Dr Bennett :
You state : " There is no reason to suspect that the AIDS-risk groups suffer from hypergammaglobulinemia"
However, I could find some evidence of the contrary (I link directly to the PubMed entries for more convenience) :
In studying 60 intravenous drug addicts, HIV+, the authors found that 86% of them had hypergammaglobulinemia. They also state : "The numerous serologic findings in these patients fundamentally express the existence of a chronic polyclonal stimulation of B cells which may be initiated by the action of the drug itself".
In studying 117 intravenous drug users of which 45 had PGL, the authors found : "Hypergammaglobulinemia and an inverted T helper/suppressor ratio and lymphopenia were found in 78%, 43%, and 20% of the patients, respectively; the presence of HTLV-III antibodies was demonstrated in 12 (46%) of the 26 tested patients"
78% of hypergammaglobulinemia and only 46% HIV+... Therefore , a substantial proportion of hypergammaglobulinemia cases cannot be linked to HIV (and hypergammaglobulinemia, as far as I am aware, is not considered an AIDS-defining condition - you should agree with this since you state "Hypergammaglobulinemia can occur in some people post-HIV infection", which means it is rather infrequent)
In conclusion, there is at least strong suggestion that hypergammaglobulinemia is vastly prevalent in intravenous drug users, regardless of HIV status.
Here are my comments about the studies you cited :
Ref 1 : Indeed, I can understand why this study is cited by the dissidents. This study shows that there is a high correlation between a positive HIV test and the presence of Plasmodium falciparum antibodies in a sample of patients from Zaire. The authors state that this correlation cannot be found in American patients, ruling out a possible cross-reactivity with Plasmodium falciparum antibodies. However, this does not rule out in any way the possibility that other immunological abnormalities in the African patients would explain the correlation. In support of this is the fact that "Immune-complex levels, but not IgG, IgM, or IgE levels, also correlated strongly with seropositivity" The authors themselves conclude that "Possible explanations include (...) enhanced antibody production due to the effect of malaria on the immune system; (...) false-positive reactivity in the ELISA assay due to cross- reactive antibodies or other unknown factors" In no way this study gives any convincing evidence of the "general" specificity of the ELISA test on this group of people - on the contrary, it gives strong suggestion about the possibility of this not being the case. It is also worth noting that no comparison with actual culture and isolation of the virus has been done.
Ref 2 : Like the previous study, this one shows there is no cross-reactivity between the HIV ELISA test and Plasmodium falciparum antibodies. However, my concern was not really with possible *specific* cross- reactivity (specific antibodies like the ones against Plasmodium falciparum being cross-reactive with HIV antigens), but rather with the fact that African malaria patients are often malnourished and often have other health problems including immune abnormalities that would make the ELISA test to react (whatever the reason) The only way to assess the specificity of this test, in this case, would be to perform them on such a group of people (African malnourished patients with many diseases) and compare the results with virus culture and isolation (since this is the best gold standard we have). Would you please show me such a study ? Unless such a study has been done, I cannot understand how we are able to diagnose malnourished African people as HIV- infected and infer these tremendous statistics we can see everywhere.
Ref 3 : I do not see in any way how this study can show anything about the specificity of the HIV antibody tests. The only thing it shows is that in a majority of HIV+ people diagnosed with TB, a commercially available TB test is unable to detect any antibody against lipoarabinomannan in most patients. I would say the reason of this result may be because (1) the antibodies are absent, because (a) the TB diagnosis was wrong or (b) their immune system did not respond to the infection ; or (2) the antibodies are present and the test is worthless, at least in this case. If this is the only study you can show me when I ask for a study showing the specificity of the HIV antibody tests in TB patients, I would say the dissidents are not completely foolish.
Refs 4,5,6 :
- I cannot see any reference to a study published in a peer-reviewed journal that would present the clinical data and the findings presented in these package inserts. Maybe I missed them ?
- The "gold standards" used are just *other* already licensed *antibody* tests (usually WB) ! Unless you can show me data about the specificity of these "gold standards" in the various populations I am interested in (and to do this, the only way is to compare with actual virus isolation and culture), I do not see in any way how I could be convinced about the specificity of the given tests in the various groups I have cited in my previous post (and even in any group at all).
I am sure you completely agree with me that the only way to measure the specificity of a test is to compare it with a gold standard, which is in our case actual virus isolation and culture. In none of the references given there is such a study. One of this study (ref 1) even suggests (not proves, I agree) a poor specificity. Moreover, even I, as a non-specialist, could find quite easily 2 studies strongly suggesting that hypergammaglobulinemia is vastly prevalent in intravenous drug users. This calls even more for high-quality studies of the specificity of antibody tests in populations at risk for AIDS. Since AIDS has been one of the most well-funded diseases in the history of medical research, I would be astonished that no specific serious clinical study of the specificity of the antibody tests on various AIDS-risk groups, by comparison with actual virus isolation and culture, had ever been done.
Competing interests: None declared