One theory now being investigated is that people who have received tissue transplants, such as kidneys, may have a higher level of resistance to the disease than normal.
This could mean that antibodies generated against other human cells are more effective than those against entirely foreign cells such as HIV.
However, any vaccine will not be available before the end of the century, and could lead to social problems in the countries which need it most, said Ann Rees, of the Jeferiss Research Trust laboratories.
Speaking yesterday, she said: "If we did develop a vaccine, there would be a question of whether the countries that needed it most could be persuaded to to accept its use."
Ninety per cent of HIV infection occurs in developing countries, she said. "It would be a vaccine against a disease which is, in those countries, sexually transmitted - that's socially difficult for some people to accept. But it's important that they are, for the greater good.
"Efforts to develop an HIV vaccine have been continually frustrated since the virus was first identified in the 1980s. This is because it mutates very rapidly and interpolates itself into the infected person's immune system."
There are already three trial vaccines against SIV, the monkey form of the disease, which are being tested on macaque monkeys.
None though has been tested on humans.
But this year scientists in Britain and the United States have made rapid strides in understanding how the virus attaches itself to body cells.
They have also found that about one in 100 Caucasians have a genetic mutation which seems to make them naturally immune to the virus.
Professor Thomas Lehner, of Guy's Hospital, said: "I am more optimistic that there will be a vaccine against HIV, particularly because we now know much more about its mechanism. The whole climate of opinion has changed in the past six months. It has been a revolution."
He said that the new work on vaccines in monkeys had led to a number of new areas of study. Special interest has focused on one vaccine, which extracts the antibodies generated against SIV from one monkey and injects them into another, uninfected monkey. Experiments have demonstrated that these monkeys are more resistant to SIV than normal. This is reckoned to be caused by the uninfected monkey producing "allo-antigens", which are antibodies against closely related cells. The same types of antibody are produced by patients who have organ transplants, and can lead to the rejection of the donated organ. But in the case of SIV, and possibly HIV, it appears to confer resistance.
"There's a possibility that the allo-antigens are a key to it," said Prof Lehner. "Certainly, it seems that the protection isn't associated with the virus, but with the human antigen. However, the mechanism is still a huge question."