HYPERICIN RESULTS: COMMUNITY RESEARCH ALLIANCE STUDY

Background

Hypericin is an antiviral found in St. John's wort, a plant long used as a medicinal herb. AIDS researchers at New York University have studied hypericin's activity against HIV, and also against other retroviruses in animals (Lavie and others, 1989; Meruelo and others, 1988). Hypericin works in an entirely different way than AZT, and was better than AZT against the retroviral infections in animals.

The researchers have now developed methods for synthesizing large amounts of hypericin; with luck they will begin early "phase I" human trials next summer. Meanwhile the Community Research Alliance, a San Francisco research organization started by San Francisco's PWA Coalition, began its own observational study of hypericin herbal extracts as an AIDS/HIV treatment last summer. Because the study volunteers obtained their own extracts from buyers' clubs or health-food stores, the Community Research Alliance did not administer any drug and therefore did not need to go through the FDA to obtain an "IND" (Investigational New Drug approval), which for various reasons would have been impossible in this case. The study complied with legal and ethical requirements by being approved by an Institutional Review Board, which the Community Research Alliance had previously organized in accordance with Federal regulations. The organization's Scientific Advisory Committee had also approved the study.

For more information on the Community Research Alliance, see AIDS TREATMENT NEWS #85, August 11, 1989. This writer is a co-founder of the Community Research Alliance, and co-author of the formal report of the study, which has been submitted to the Sixth International Conference on AIDS. The study's principal investigator, and principal author of the report, is William C. Cooper, M. D.

Methods

The Community Research Alliance study lasted four months; a total of 33 volunteers were enrolled. The only entry criteria was to be HIV positive, and not to have used hypericin extracts within the last six months (so that baseline data could be obtained). Extensive blood tests were given at baseline and at four additional monthly visits; other data collected included medical history, physical examinations before and after the four-month test period, and the Merieux Multitest of skin immune response. All testing and other expenses (except for the cost of the herbal extracts, which the research organization could not legally provide) were paid by the Community Research Alliance, which financed the study primarily through donations from individuals; a grant from the American Foundation for AIDS Research (AmFAR) paid salary for a medical director.

While four months of data were collected, the study actually ran six months, as resource limitations prevented enrollment of all the volunteers at once. There was no problem recruiting people to join the study. Volunteers were not asked to use any particular brand or dose of hypericin extracts, and they were not asked to make any changes in the treatments they would have used anyway, for the sake of the study. They were, of course, asked to report all treatments they used. All volunteers kept daily logs, and compliance with the protocol is believed to have been very good.

A separate treatment database now being developed may be able to collect long-term followup information from the volunteers in the future.

Results

Of the 33 patients enrolled, 26 completed the four months. Of the seven who did not complete the study, six were hospitalized for AIDS complications and had to drop out; one was dropped for non-compliance with the protocol.

A first look at the data showed no clear trends. About as many of the volunteers had T-helper decreases as increases. But a closer look showed a pattern which had not been expected. Of the 26 volunteers who completed the study, ten had never used any AZT, while ten others had been on AZT throughout (the remaining six had either started or stopped AZT during the four months). Of the ten "AZT virgins," eight had T-helper increases; the mean increase for the ten was 12 percent, which occurred in the first month and was sustained for the four months for which data are available. But of the ten on AZT throughout the study, eight had T-helper decreases, with the mean decrease for the ten of 13 percent. (The mean T-helper count increased nine percent in the first month, then declined.)

These results do not show that AZT worked poorly with hypericin. Those who were on AZT were more seriously ill, with mean T-helper count of 189 at baseline (before they started hypericin) than those who had never used AZT (mean T-helper count 558 at baseline). The difference in outcome is not surprising, since other anti-HIV treatments, such as AZT and ddI, are known to work better in patients who begin with higher T-helper counts.

The 12 percent T-helper increase in the ten who had never used AZT may be a meaningful trend, since without any treatment the average T-helper count would tend to decrease. However the increase in this small group was not enough to be statistically significant, meaning that it might have occurred by chance. The fact that the increase occurred in the first month, and then was sustained for several months, is consistent with results of other antivirals.

P24 antigen test results were inconclusive -- although they did show that there was no dramatic antiviral effect of the hypericin-extract treatment. Only six of the 26 who completed the study were p24 positive at the beginning. Two became p24 negative, but they were on AZT throughout the study, so the improvement might have been due to AZT. Of the other four, two went up and two went down. No one changed from negative to positive.

What about safety? There have been several anecdotal reports that some people's liver enzyme values have risen after use of St. John's wort, suggesting possible toxicity. In this study, five of the 26 volunteers showed rises in liver-function test values, and were advised to discontinue the herbal extract. All returned to normal after 30 days off the treatment, except for one who had earlier liver problems and high liver-function values before starting the study. This effect (rise in liver function tests in some patients) is suspected to be due not to hypericin but to some other component in the herbal extract, since animal studies have shown no effect on liver enzymes even when very large doses of pure hypericin were given.

Much other data was collected and has not yet been analyzed. But principal investigator Dr. William Cooper has not seen anything striking or unusual in a first look.

Comments

A small, uncontrolled observational study like this one is not designed to prove that a treatment does or does not work. This study provides no proof either way, although the trends observed were in the direction which would be expected if there was some effect.

One reason for doing this study was the hope that there might have been benefits dramatic enough (for some groups of patients at least) that they would be unmistakable even in this limited trial. No such benefits were found. We should remember, however, that the available herbal extracts contain only small amounts of hypericin, and there have been doubts as to whether there is enough to be effective. It is still possible that pure hypericin, or perhaps better extracts, could be of major benefit.

Even before this study began, it was clear that the chance was slight that the herbal extracts would have so dramatic an effect that a small observational study would show it unmistakably. But the chance was not zero; and a positive result would have been of such great importance that we believe the chance justified doing this study, and justifies future studies of other available treatments which have a good scientific rationale.

References

Lavie, G. and others. Studies of the mechanisms of action of the antiretroviral agents hypericin and pseudohypericin. Proceedings of the National Academy of Sciences USA, volume 86, pages 5963-5967, August 1989.

Meruelo, D. and others. Therapeutic agents with dramatic antire-troviral activity and little toxicity at effective doses:
Aromatic polycyclic diones hypericin and pseudohypericin. Proceedings of the National Academy of Sciences USA, volume 85, pages 5230-5234, July 1988.