Where microbicide research is concerned, it can be useful to distinguish between studies of potentially active ingredients and studies of potential products, formulated for use by people. This section is mainly about the ingredients.

There is no shortage of ideas. By 2002, according to the Alliance for Microbicide Development, about 52 product leads were being actively developed, with at least five in Phase II or III clinical trials. Interestingly, this was about the same number as two years before. Several leads - mostly related to nonoxynol-9 - were no longer in active development, while others had been added to the list.

The list which follows sets out some of the categories of active products, and the rationale for their use. Most of these have shown some activity in laboratory studies but, as will be apparent from the previous sections, this is no guarantee that they will be of value in real life.

A more complete list, complete with progress reports on individual products, is maintained and published online by the Alliance for Microbicide Development (http://www.microbicide.org/).

• Acid buffers (Acidform, BufferGel)

The rationale for these has been explained earlier, and is also discussed in relation to lemon juice (Do we already have a microbicide?). BufferGel is proceeding towards trials as a microbicide; Acidform, developed by TOPCAD, is active against some vaginal infections (Tevi-Benissan). There were disappointing results in a trial where it was combined with N-9 and led to increased adverse effects from the N-9. It may be evaluated as a spermicide in populations at low risk of HIV.

• Antibiotics (gramicidin, squalamine, defensins - including protegrins)

Naturally occurring antibiotics that are poorly absorbed into the body and could be effective against a range of bacteria. Those which act against bacterial membranes can also damage the HIV viral envelope without harming mammalian cells. Gramicidin would be dangerous if it did enter the blood stream and also suffers from the disadvantage that it is not directly soluble in water. An alcohol-based formulation would be incompatible with condoms.

• Antiseptics (chlorhexidine, undecylinic acid)

Chlorhexidine has been studied in childbirth, with the idea that a microbicide rinse in the birth canal might protect a baby from exposure to the mother's virus during birth.

• HIV-specific binding agents (cyanovirin, monoclonal antibodies/plantibodies)

Cyanovirin was originally isolated from blue-green algae; it binds very specifically to HIV gp120. Work on cyanovirin is now focussed on engineering lactobacilli to produce it, harmless bacteria which could live in the vagina and provide enhanced protection for long periods

It is also possible to grow large quantities of human anti-HIV antibodies very cheaply in plants. A mixture of several antibodies may be more effective than a single one; there are already animal studies which have shown that SHIV infection can be prevented using such a mixture when administered by injection. It has also been shown that antibodies produced in this way will retain their activity for many hours in vaginal conditions. Unfortunately, female monkeys were not protected from vaginal challenge with an SHIV using an external application of the same monoclonal antibodies that had protected when given by injection, even when those antibodies were directly mixed with the culture used to challenge the monkeys (Lewis).

• Hydrogen peroxide (Lactobacillus suppositories, H202 gel)

Lactobacilli are a natural part of the vaginal flora: they produce hydrogen peroxide, which inhibits HIV. They also maintain vaginal acidity through the production of lactic acid. Incidentally, N9 is known to interfere with these bacteria.

• Sulphated and other charged polymers: dextrin-2-sulphate, dextrans, naphthalene sulphonate polymer (PRO 2000), carrageenans (e.g. CarraGuard = PC-515), sulphonated polysaccharides, polystyrene sulphonate, cellulose sulphate, cellulose acetate phthalate, hydroxyphthalyl-beta-lactoglobulin, heparins.

These include the three products that are most advanced towards testing in the next round of Phase III trials, as previously mentioned. Cellulose sulphate has also been identified by World Health Organization (WHO) as a priority for research. Preliminary evidence has been reported that a bacterial polysaccharide related to heparin (but without its anticoagulant properties) is effective against HIV (Vicenzi).

A chemically modified relative of carrageenan (PC-710) has been identified by Population Council researchers and appears more effective than Carraguard in protecting mice against HSV, even when administered after the mice have been exposed to the virus (Maguire).

SAMMA, a polymer of mandelic acid produced by treating it with sulfuric acid, is in pre-clinical evaluation and shows strong inhibition of HIV and HSV without damaging living cells (Herold).

• Surface Active Agents (surfactants): N9 in new formulations, N9/octoxynol-9 /benzalkonium chloride (BZK)/sodium cholate (NaCh) combinations, amphoteric surfactants (e.g. C31G), menfegol, N-docasanol, 2-octylglycerol, polybiguanides, sodium dodecyl sulfate (SDS).

These products all work on the same principle as N-9 and may suffer from the same problems. For example, a study of menfegol foaming tablets showed that women who used them experienced a high incidence of internal lesions.

A phase I trial has been carried out on a microbicide containing sodium lauryl sulfate, a detergent widely used as a foaming agent in shampoos and toothpaste. After application twice daily for 14 days by 47 women and 23 male partners, 'minor symptoms (mild or moderate) possibly related to the product use were observed in about one third of subjects. The most common events were erythema, itching, burning sensation, vaginal discharge and vaginal dryness. No genital ulceration or lesions were seen during gel use.' Whether this will be sufficiently reassuring to justify further research remains to be seen; as a detergent, SLS is in the same broad category as nonoxynol-9 (Trottier).

• Porphyrins

A class of natural and synthetic proteins of which the most famous are haemoglobin and chlorophyll. 75 different compounds have been evaluated in cell culture systems at Emory University, USA, and several have been shown to interact strongly with the HIV-1 envelope molecule, preventing cell entry (Vzorov).