Here's a story Emma Kearns has told 50 times before, and each time she tells it she never knows how it will turn out. 'I was on my mobile phone, and I had my arm resting against the side of my breast, and I felt something there that shouldn't be. More of a mass really than a bump. I mentioned it to Dave, my partner, and my mum, and they said it would be best to get it checked out to be safe.'

This was in December 2004, when she was 27. She went to her doctor in the middle of January, and was told that the breast changes quite a lot during the menstrual cycle, so the best thing would be to come back in two weeks. 'I went back two weeks later, and by then the mass had more of a harder, rounder shape. My doctor referred me straight to the Marsden, and I was really lucky: the Marsden has got a rapid diagnostic unit which means you're tested and get the results on the same day. I made an appointment for 8 February.'

Emma Kearns was born in Long Eaton, near Nottingham. She grew up in Cambridge and Bristol, and met her boyfriend Dave at university in Bradford. She works at the Department of Education, and he's a teacher, and they now live in a small house in Sutton, Surrey. The Royal Marsden, a hospital known for its excellent cancer care, is just up the road.

'They sent me through a nice shiny booklet, explaining what to expect at the breast unit on the day. It suggested I took someone along, and so I went in with my mum. In the booklet they said they would perhaps tape the conversations I had with the staff, as some of the messages could be initially hard to take in. I thought, "I won't be needing that." I really thought they were going to tell me it was just a cyst.

'When I went in that morning they handed me a big empty folder at the reception desk, and I said, "That seems like a very big folder just for me," and they said, "Hopefully it will remain empty." The only time I'd been in hospital before was to have my wisdom teeth out. I went into a room and one of the surgeons did a brief examination, then I had a biopsy, inserting a large needle which sounded like a staple gun when it drew something out of the lump, then a mammogram. Then I waited around for the results to be collated.

'The first inkling I had that things weren't quite right was when a breast cancer specialist nurse came to get me, with a blue uniform and a special badge, rather than a white uniform and the regular badge. And then I walked into the room and there was a tape recorder on the desk.'

Each year, about 40,000 women are diagnosed with breast cancer in the UK, accounting for about a third of all cancers in women. One in nine women will get breast cancer at some stage in her life, but due to early screening and better treatments, death rates have fallen by a fifth in the last decade.

Emma Kearns didn't know much about cancer when she was diagnosed. At work she knew someone with bone cancer, and a few months before she had cheered as her mother took part in Cancer Research UK's Race For Life, but that was about it. 'So I sat in that room and I just thought, "Game Over." The head surgeon said, "I'm afraid you have breast cancer, as we suspected," and my mother was sat beside me and I remember thinking, "I'm going to look at her, and if she's crying then I really am done for." But she wasn't crying, and I thought, "OK, maybe I can get through this day."

'The first thing that came out of my mouth was, "My hair is going to fall out." I felt that I wanted to run away but couldn't. Then I said, "How can I expect Dave to stay with me now?"'

She went to a small room with her mother and a nurse, and she cried for a while. The one thing she remembers deliberately not asking was, 'How long before I die?'

'My mother said, "Let's cheer ourselves up by going shopping," but Emma thought, "Why? I'll be dead soon. It's a waste of money." I didn't understand that you could live with an illness like this.'

Emma Kearns's type of breast cancer was particularly aggressive, but unusually common among those who are diagnosed under the age of 40. Her cancer was formed at least partly by the presence of too many copies of a gene called HER2. A few years ago this knowledge would have been superfluous, and her prognosis would have been bleak. But at the beginning of 2006, our new understanding of her genetic pathways may save her life.

In 1986, when Emma Kearns was settling into primary school, a 36-year-old cancer specialist called Dennis Slamon found a new way to examine what it was that distinguished a cancer cell from a normal cell, and specifically what it was that made a cancer cell replicate at an uncontrollable rate. Slamon worked at the University of California at Los Angeles (UCLA), and spent much time attending scientific conferences. After one symposium in Denver he had met the German biologist Axel Ullrich, and the two men realised they might be able to help each other. Ullrich was an emerging star in the field of genetic cloning, and had contributed to the landmark discovery that certain growth factors and receptors in cells are related directly to cancer. This is a cornerstone of oncology we now take for granted, but when the research was first published in Naturein 1984, it not only confirmed what scientists had suspected for decades, but also held the potential for revolutionary breakthroughs in treatment. If specific cancer-causing genes could be isolated, then in time they could also be targeted and blocked and corrected. This was the great hope for the start-up biotechnology companies of the Seventies. But for a decade or so the chances of success seemed remote; the haystack was the size of a mountain with more than 100,000 genes in every human cell involved in growth rate alone.

Thankfully, science has always been populated by mavericks and true believers. At UCLA, Dennis Slamon had amassed a large collection of breast, lung, liver and colon tumours, a collection which some considered a ghoulish passion, but Slamon regarded as a research bank. For a few years it wasn't quite clear what purpose the bank would yield, but in the summer of 1986 the collection found its use. He asked Axel Ullrich to send him DNA samples from his growth factor genes - oncogenes - and Slamon and his assistant would attempt to match them against the DNA extracted from his tumours. After a while, Slamon got a hit from his breast and ovarian cancers. Ullrich revealed that the oncogene in question was called HER2, first identified a few years earlier (the initials indicated it was the human epidermal growth factor receptor 2).

This was a momentous leap, but it didn't answer all the questions. The scientists found that HER2 was linked to breast cancer growth when copies of the normal HER2 gene mutated and 'overexpressed' its protein. This protein appeared as a receptor on a cell's surface, and sent a signal within the cell for it to grow and divide. A healthy breast cell has about 50,000 HER2 receptors on its lining, but on a mutated gene this may increase to more than a million.

But Slamon's own collection of breast cancer tissues (from about 30 tumours) had no case histories attached to them, and it was impossible for him to tell what sort of breast cancer HER2 mutation caused. So Slamon turned to another collaborator who had oncogene protein samples linked to many patients' case notes, and from these it was possible to detect that HER2 disease was not only aggressive and common in younger patients, but was linked to the frequent recurrence of tumours. Ullrich and Slamon published their findings in the magazine Science. At first, fellow researchers couldn't believe what they read. It all seemed so simple. All they had to do now was find a way for their discovery to benefit patients.

'My tumour was initially 3cm by 4cm,' Emma Kearns says. 'And so the first chemo was an attempt to shrink it, which hopefully would mean only a lumpectomy rather than a mastectomy.'

The treatment worked, and her breast was saved. When the tumour was removed it measured only 8mm, and the 11 lymph nodes cut out at the same time contained only dead cells.

Her first course of chemotherapy consisted of a combination of two drugs known as AC - Adriamycin and Cytoxan. She did lose a lot of her hair at the front and top, but the use of a helmet resembling a horse-riding hat froze and saved some follicles at her crown. She wore the occasional wig and head scarf, 'and in the summer I just looked like a boho chick'.

The second chemotherapy after surgery in June 2005 was Taxotere, whose active ingredient is derived from the yew tree. She underwent radiotherapy at the same time, and the whole process made her nails turn white and snap off. 'After some treatments I was OK,' Kearns says, 'and after some I was heaving before I even got to the car at the hospital.'

Compared to the promises of biotechnology, familiar cancer therapies can sometimes seem like something from a darker age, best summed up by an American doctor as a combination of 'slash, poison and burn'. The surgical removal of breast tumours is still regarded as the best line of defence. Chemotherapy has advanced greatly since the blunt instrument first wielded in the late Fifties and early Sixties, but the great range of treatments designed for the various types and stages of cancer are inexact weapons, as is radiotherapy: the side effects - severe nausea, bone marrow disease, loss of hair, fevers, diarrhoea - are the clearest indication that in the process of killing diseased cells, immune-boosting and healthy growth cells perish, too. In the longer-term, infertility would be another blow.

The commercial use of tamoxifen in the mid-Seventies signalled a new way forward for breast cancer care. This therapy, manufactured originally by ICI under the trade name Nolvadex before its mechanism was fully understood, blocked oestrogen production by adhering to the oestrogen receptors in breast cancer cells, thus interrupting one of the pathways of tumour growth. Initially licensed for late-stage cancer treatment, seven years ago it was also approved as a preventative treatment for those considered at high risk from breast cancer, not least if there is a family history of the disease. Tamoxifen is taken in pill form over two to five years, and it can have many side effects, not least those commonly associated with the menopause.

Because it is rarely effective in women under 35, Emma Kearns was not a good candidate for tamoxifen. But now there was something new. She was copied into all the letters the Royal Marsden sent to her doctor, and her correspondence was invariably headed with the code HER2+, which put her in a group of between 25 to 30 per cent of women with breast cancer (rising to 40 per cent of women younger than 40). Until a few years ago most women with breast cancer would not be tested for the HER2 gene, and not just because the technology was complex and expensive. There was no point because there was no treatment.

This position had changed by the time Emma Kearns presented herself. 'They were saying that there is this research going on,' she remembers, '"and we don't know how it will affect you, but you are a suitable candidate. We'll let you know, because the results of the research aren't in yet."'

That was in April 2005, and the results of the research were but days away. But the trials in question were a specific inquiry into the efficacy of a drug for use in early-stage breast cancer. The same drug's dramatic impact on late-stage and metastatic breast cancer (cancerous cells that have spread beyond the breast), had already been proven, and it had far exceeded the wildest hopes of Dennis Slamon and Axel Ullrich.

In the late Eighties, the two scientists took their work on HER2 one stage further - into mice. Slamon worked in his lab at UCLA, while Ullrich developed his research with his colleagues at Genentech, the biotechnology company that had been founded in 1976 and had later adopted the corporate slogan 'In Business For Life'. The goal was to develop a monoclonal antibody effective against mutating HER2 - a synthetic drug, made in large quantities, designed specifically to block the receptors on the cell surface. The progress was swift. In 1989, an HER2 antibody was shown to work in human breast tumour cells; in 1991 the first anti-HER2 antibody that had initially been developed in mice was tested in humans, and reported a few months later in the journals. This phase 1 trial showed no damaging side effects and enabled the researchers to work out a suitable dosage. The phase II trials, testing the drug's effectiveness in a small group of patients recruited by Dennis Slamon at UCLA and other sites, commenced in 1993 and ended a year later, and the results were very encouraging. About 15 per cent of the women benefited from the drug (their tumour either shrunk or didn't increase), but this gave Slamon confidence that the biological theories were at least correct. Much of the money at UCLA came from the cosmetics company Revlon, which contributed about $13m to the university's cancer work under Slamon's leadership between 1989 and 1997. As it entered the phase II trials, the drug finally had a name: trastuzumab. And Genentechhad also registered a trade name for it: Herceptin.

'There was little doubt in our minds that it would work,' Dennis Slamon told me one morning before he left home for his office at UCLA's Jonsson Cancer Centre. 'We knew that if we understand what's broken, and target that specifically, then in theory we'll have more effective and less toxic therapies. But there was a lot of doubt in the field, because monoclonal antibodies had traditionally failed.'

In the phase II studies, Slamon says the 15 per cent success rate was encouraging because he was seeing patients who'd failed with everything else. 'Everyone now believes this is the crest of the wave,' he says. 'Before the Herceptin data it was only a big promise, and after it, it became a reality.'

But the big test was yet to come. The phase III trials, which had great difficulty recruiting just a few hundred women to such a novel experiment, began in 1995, and would be the first big test of Herceptin delivered alongside traditional chemotherapy drugs. The results were revealed publicly in 1998.

The day before I talked to Dr Slamon, I had been to see Mark Sliwkowski, staff scientist and director of translational oncology at Genentech in San Francisco. The company was so big it had its own street: DNA Way. Dr Sliwkowski joined Genentech in 1991, not long after it had been bought by Roche Holding Ltd, the Swiss company that owned the giant pharmaceutical company Roche. He was one of the first at Genentech to learn the results of the phase III Herceptin trials in November 1997. 'It was unbelievable,' he says. 'I knew that it worked, but I was told not to tell anybody. In fact, my wife worked here, and I had to keep the data from her for a month.'

The data showed one thing above all others: Herceptin could delay the recurrence of breast cancer and extend the life of patients. Women who had been diagnosed with metastatic disease could expect to enjoy 18 to 24 months of cancer-free life more than those on the trial who had not been given Herceptin. The Food and Drug Administration, and then European agencies, swiftly licensed the use of Herceptin in late-stage breast cancer, and a new world opened up. The drug would encourage an early HER2 screening for all women with breast cancer, and it would place a large financial burden on patients and healthcare providers. But above all it would ask a challenging question: if Herceptin proved effective in late-stage breast cancer, for how long could it extend the lives of women who were still in the early stages of the disease? 'We always thought that's where the big bang would be,' Mark Sliwkowski says, 'and once the late-stage results came in we immediately started planning for it.' The new trial began recruiting in 2000, and the results were revealed in April 2005. About 5,000 women would take part in the United States, and another 5,000 in Europe.

'I was in my car when I got a call with the results,' Pamela Klein, a senior director of oncology at Genentech, told me. 'And I just pulled up to the side of the road and cried.'

The chief UK investigator on the European early-stage Herceptin trial - known as HERA - is Ian Smith, professor of Cancer Medicine at the Royal Marsden Breast Unit in the Fulham Road. The day I visited him in December, he had just arrived back from the annual breast cancer conference in San Antonio, where he had heard Dennis Slamon deliver yet more good news about Herceptin. This time it was about side effects. Herceptin did not cause nausea, hair loss or bone-marrow disease, but previous trials had shown a small risk of cardiotoxicity, which manifested itself in a shortness of breath caused by a reduction in the pumping action of the left ventricle. In San Antonio, Dr Slamon had shown that it was possible to reduce this risk to almost nothing by a change in the chemotherapy that accompanied the delivery of Herceptin.

The HERA trial was due to present its results in San Antonio, but scientific rivalries changed the plan. 'The results were produced in a great rush, Professor Smith concedes. 'It is very competitive - we're all working for the common good, but we also want to beat the guy down the road. We obviously wanted to be the first to present our results, but there were also these two American trials, and we heard that the two American trials were going to present at ASCO.' ASCO, the American Society of Clinical Oncology, was the big annual cancer convention, held last May in Florida. 'So we decided to speed things up. The analysis of the HERA trial was originally going to take place over the summer. But we asked our data management team what we could do, and they said, "If we work round the clock we could get the results for you earlier."'

The HERA paper and the results of the US trials were presented in a special session attended by 6,000 people. 'You couldn't get in,' Smith says. 'The Americans presented their results first, and there was a gasp in the hall. There was thunderous applause at the end. And then we presented our data and the response was the same. People had tears in their eyes.'

The HERA trial showed that the use of Herceptin and standard chemotherapy in early-stage breast cancer reduced the risk of recurrence of cancer by 46 per cent. The American trial suggested the risk reduction was 52 per cent.

The New England Journal of Medicine, which published the findings, called the results 'simply stunning', but an editorial a few weeks later in The Lancetsuggested caution, and stressed that Herceptin had still to undergo stringent safety checks for use in early-stage care.

But Mark Sliwkowski says he will never forget the response from his fellow cancer specialists who had worked all their lives for news like this. 'The new data said that if you're 40, you have a really good chance to live to be 85. It's the first time - and we have never used this word - that you can actually begin to talk about a cure.'

The HERA trial looked not only at efficacy but also at the optimum period during which Herceptin should be given. The guidelines recommend one treatment - delivered intravenously through a drip for 90 minutes - to be given every three weeks for a year, but the ongoing part of the HERA trial will detect what effect two years' treatment will have. It may show that a year is too short for the best results. But another trial, also presented in San Antonio in December, suggested that a year's treatment could be quite superfluous. This Finnish trial, known as FinHer, suggested that a course of three treatments over nine weeks might be enough to produce the same results as a course lasting a year. This was perhaps the most dramatic news of the conference, but it comes with a strong caution. The number of patients taking Herceptin was less than 250, a small enough group to produce erratic results. A larger trial is now under way.

As soon as he returned to London from Orlando, Professor Smith noticed that the mood at the Royal Marsden had changed. 'We had a very difficult time in the clinics - everyone rushing around, the media getting excited, everyone trying to get hold of it. For the patients it was obviously very emotive - they were told they had a high-risk cancer and there was now only a 50 per cent chance of it coming back.'

And there was a major problem: unless you were in the control arm of the HERA trial, the drug was difficult to get hold of. 'I took the view early on that it was an exceptional result and that it was inconceivable that the licensing authorities wouldn't license it,' Professor Smith says, 'but in an ideal world we all should have shut up about it. But no one's going to do that, least of all Genentech and Roche.'

In the UK, Herceptin costs about £21,000 per patient per year, and this was not something the NHS had budgeted for. A press release from Roche last May suggested that the early use of Herceptin could save the lives of 2,800 patients a year, but the drug still has many months of safety checks before it is approved. In October, health secretary Patricia Hewitt took the unprecedented step (for an unlicensed drug) of announcing that no woman should be refused Herceptin on cost grounds alone, but this was something some Primary Care Trusts (PCTs) found hard to accept. Or at least they found it hard to accept until the bolder and more eloquent patients who had been refused Herceptin took their cases to the media and the courts.

In addition, Professor Smith had noticed a new trend in the world of private health insurance. Before the great results in May, most companies had been willing to fund Herceptin; after the trials created an increased demand, they were more reluctant, and began suggesting Herceptin was a preventative measure rather than treatment, which anyone who had seen the trial results regarded as nonsense. Professor Smith and his colleagues found that their time was increasingly taken up by arguments over short-term/long-term cost benefits. Herceptin was expensive now, but it was cheap compared to the potential cost of caring for a patient with recurrent tumours. They also found they were making many more pleas to PCTs to grant patients Herceptin under 'exceptional circumstances', and at the end of last summer one of these concerned Emma Kearns.

'When the trial results came out I could tell that my doctors were getting more and more excited about it,' she remembers. 'But they didn't want to get my hopes up. They said I had already benefited from a lot of good treatment.'

The Royal Marsden applied to her local PCT, Sutton and Merton. Her doctors claimed she was an ideal candidate, but there was no reply for several weeks. While she waited she ran the Race For Life with 49 friends, raising almost £10,000 for Cancer Research UK. 'Then one day the hospital called me up to say they had heard, and my application had been rejected. They never put anything in writing to me, but we talked to the PCT, and it turned out that the application had to go to an exceptional circumstances panel, who have to be given a reason why I, above any other woman, should be funded to receive Herceptin. Of course, that's absolute nonsense. Why should I have it and other women not? So it's not going to get through. But I should have it, and so should every other woman who tests positive for the protein.'

The Marsden appealed, and Paul Burstow, Kearns's Liberal Democrat MP, used his position as a member of the House of Commons Health Committee to raise Kearns's predicament with Patricia Hewitt. The health secretary repeated her wish that all suitable patients should receive Herceptin if their doctors recommended it, irrespective of cost and the lack of a report from the National Institute for Clinical Excellence (NICE). Kearns and the Royal Marsden again heard nothing from her PCT for weeks, but then her mother called up to find out what was happening.

'They told her they had decided to fund it,' Kearns says. 'I was massively thrilled. But then I went in to start the treatment and I was really disappointed. It was just a small bag of clear liquid I had fought for, and it looked like water.' The Herceptin was delivered in the same way as her chemo, from a bag held on a long pole that dripped into her left hand as she sat in a chair. The first time she received it she began to feel very cold and then very hot, a common reaction, and the treatment was suspended for a few minutes. She has now had three courses - the fourth is due this week - with no other side effects.

When she visits the Royal Marsden these days she finds that she is treated as a success story. Her oncologists talk among themselves about new regimens and new trials, and of how Herceptin has changed the lives of so many patients they know. Soon, the idea of Herceptin - that is, the biology behind it - may soon change the entire cancer landscape, and other areas of medicine will benefit as much.

Emma Kearns's hair has grown back, but now there are new dilemmas. 'The initial treatments have worked, but now I'm thinking about trying to keep the disease away, and worrying if it will come back and are there any bits left, and I find I'm not as strong. All people say is, "Are you better now?" and I say, "Yes I am," but you can never be sure. The fact that it was in my lymph glands showed that obviously it can spread, and there may still be lots of sites hiding things. Which is why the Herceptin is so important now.'

She's getting married in August. 'Before the cancer I used to worry if Dave really loved me, but now I just know he does.' She says she is wondering whether she should just carry on as she was before she became unwell, or whether she should use her time in a different way. She may now have a much longer life in which to make up her mind.

· www.cancerresearchuk.org