Note: The following letter was rejected by Nature.

REPLY TO EMBRETSON ET AL.

SIR, Using PCR, Embretson et al (1) found "massive" HIV infection of "CD4+ lymphocytes and macrophages throughout the lymphoid system from early to late stages of infection". Pantaleo et al (2) "have demonstrated that during clinical latency, HIV accumulates in the lymphoid organs and replicates actively despite a low viral burden and low to absent viral replication in PBMC". The above findings may be used to answer some of the questions regarding the pathogenesis of AIDS by HIV.(3,4) However, if considered in conjunction with other evidence, the findings also raise several questions.

Although PCR has a high sensitivity, HIV PCR is not standardised and its reproducibility and specificity have not been determined. In the only study designed to evaluate the above parameters, and in which serological status instead of HIV itself was used as a gold standard, the PCR was found not to be reproducible and "false-positive and false-negative results were observed in all laboratories (concordance with serology ranged from 40 to 100%)." (5)

That the HIV PCR may not be specific is suggested by several studies including the following:

1. The HIV genome hybridises with the "structural genes (gag, pol and env) of both HTLV-I and HTLV-II".(6) It is accepted that about 10% of AIDS patients are infected with HTLV-I and the normal human genome contains sequences related to HTLV-I and HTLV-II.(7)

2. On Southern blotting of DNA extracted from thyroid glands of five HIV seronegative patients with Grave's disease who "were not specially selected" all samples tested positive for HIV.(8) Positive Southern blotting was also reported from PBMC of patients with "common variable hypogammaglobulinaemia".(9) Last year, evidence was published which showed "the presence of nucleotide sequences related to HIV-1 in human, chimpanzee and rhesus monkey DNAs from normal uninfected individuals".(10)

3. The finding of positive PCR in eosinophils has been interpreted to "suggest that eosinophils may act as host cells for HIV-1".(11) However, "Formaldehyde-fixed eosinophils nonspecifically bind RNA probes [HIV RNA] despite digestion with proteolytic enzymes and acetylation...When preparations are treated with amounts of ribonuclease adequate to destroy viral RNA, the eosinophilic binding remains".(12)

Taking into consideration the above evidence as well as the facts that:

(i) Pantaleo et al had no controls, that is, hyperplastic lymph nodes from HIV seronegative patients, preferably gay men or drug users who remain seronegative for long periods of time but who nonetheless have lymphadenopathies;

(ii) Embretson et al mention controls, however, the number of lymph nodes and patients examined, and the origin and physiological status of the lymph nodes is not indicated;

how does one know that the positive PCR results obtained by Pantaleo et al and Embretson et al are specific to HIV? If specific:

(a) given the fact that the lymph nodes studied by both groups had hyperplasia, and the number of patients examined were small (4 by one group and 12 by the other), how can the findings be extrapolated to all HIV seropositive or AIDS patients, considering that "The events occurring in lymph nodes in HIV-infected individuals without peripheral lymphadenopathy are unknown, since tissue from such individuals rarely becomes available for examination"?;(13)

(b) if "an extraordinarily large number" of CD4+ lymphocytes and macrophages throughout the lymphoid system are infected with active HIV which kills the infected cells, then why do the lymph nodes display little necrosis?

(c) if HIV in the lymph organs is "in a form that is likely to be infectious for other HIV target cells, especially T-cells, that circulate through these organs", "from early to late stages" of infection, why do not the number of T-cells in general and T4-cells in particular decrease, even after a long (many years) period of infection, and in fact may increase;14 and why do even "ailing people sometimes have comparatively high levels of CD4s"?(15)

(d) in Figure 2(b) in the article by Pantaleo et al, our own morphological analysis of the image shows that approximately 40% of cells possess HIV transcripts. If, on the other hand, we accept Embretson's data that HIV infected cells are six times more frequent than those displaying HIV transcripts, then there are not enough cells in that sample of lymph node to accommodate the projected number of HIV infected cells;

(e) if "lymphoid organs are solidly infected with HIV", why:

(i) in the electron-micrograph published by Pantaleo et al to show the existence of HIV particles in lymph nodes, only very few extracellular (and no budding) particles are seen? What are the particles with the same morphology reported in 13/15 (87%) of patients with "non-HIV" lymphadenopathies"? (16)

(ii) if so many HIV transcripts are present, why is PCR needed to detect them by in situ hybridisation?

Obviously much remains to be resolved. *

Eleni Papadopulos-Eleopulos (1) Valendar F.Turner (2) John M. Papadimitriou (3) David Causer (1)

(1) Department of Medical Physics, (2) Department of Emergency Medicine, Royal Perth Hospital, Perth, Western Australia; (3) Department of Pathology, University of Western Australia.

References

1. Embretson, J., et al. Nature 362, 359-362 (1993).

2. Pantaleo, G., et al. Nature 362, 355-358 (1993).

3. Temin, H.M. & Bolognesi, D.P. Nature 362, 292-293 (1993).

4. Maddox, J. Nature 362, 287 (1993).

5. Defer, C., Agut, H. & Garbarg-Chenon, A. AIDS 6, 659-663 (1992).

6. Arya, S.K., Gallo, R.C. & Hahn, B.H. Science 225, 927-930 (1984).

7. Mager, D.L. & Freeman, J.D. J Virol 61, 4060-4066 (1987).

8. Ciampolillo, A., Marini, V. & Buscema, M. Lancet I, 1096-1100 (1989).

9. Webster, A.D.B., et al Lancet I, 581-582 (1986).

10. Horwitz, M.S., Boyce-Jacino, M.T. & Faras, A.J. J Virol 66, 2170-2179 (1992).

11. Conway, B., et al. J Arch Virol 127, 373-377 (1992).

12. Fox, C.H., Kotler, D., Tierney, A., Wilson, C.S. & Fauci, A.S. J Infect Dis 159, 467-471 (1989).

13. Fox, C.H., et al. J Infect Dis 164, 1051-1057 (1991).

14. Natoli, C., et al. J Acquir Immune Defic Syndr 6, 370-375 (1993).

15. Cohen, J. Science 258, 388-390 (1992).

16. O'Hara, C.J., Groopmen, J.E. & Federman, M. Hum Pathol 19, 545-549 (1988).