As many as 5.7 million Americans were infected with the H1N1 virus between April and late July, U.S. researchers said on Thursday, offering the clearest picture yet of how quickly and widely swine flu can spread.

Researchers used computer models to estimate the number of people who have contracted swine flu, which began infecting Americans in April.

They estimated that 1.8 million to 5.7 million cases of swine flu occurred between April and July 23, sending between 9,000 and 20,000 people to the hospital.

About 6 percent of people who were hospitalized with the virus died, the team, led by Carrie Reed at the U.S. Centers for Disease Control and Prevention, reported in the journal Emerging Infectious Diseases.

This suggests that as many as 1,300 people died from their infections between April and July. Officially, 1,000 U.S. deaths have been attributed to H1N1 since April.

Dr. Anne Schuchat of the CDC said on Thursday the agency does not have an update beyond July 24.

“We do believe many millions of people have already contracted this virus in the United States,” Schuchat said.

“It’s probably now well more than 20,000 hospitalizations,” she said. “Really, the priority is to minimize illness and death.”

Part of the U.S. plan to do that was through widespread vaccinations, but manufacturing delays have stalled those efforts. “We had all hoped to have more vaccine now than we have,” Schuchat said.

Earlier government estimates had suggested there would be as many as 40 million vaccine doses available for state and local health authorities to distribute by the end of October.

Schuchat said 24.8 million doses of the H1N1 vaccine are available, 1.6 million more doses than on Wednesday.

The United States has ordered up to 250 million doses of H1N1 vaccine from five companies — MedImmune, a unit of AstraZeneca, Sanofi-Aventis, Australia’s CSL, GlaxoSmithKline and Novartis.

Except for MedImmune, all had problems making vaccine at first and are still struggling to make the virus grow in eggs, the first step to manufacturing influenza vaccine.

Schuchat said state and local health departments have had to adapt their vaccination plans to cope with the delays, and dole out a limited number of doses to people at greatest risk of developing severe disease from H1N1, including people with underlying health conditions and women who are pregnant.

Several studies released at the meeting of the Infectious Diseases Society of America in Philadelphiaon Thursday showed that vaccinating pregnant women protected their babies, also.

They said babies were less likely to be premature and were bigger if their mothers were vaccinated against flu. A separate study showed that people who had been taking cholesterol lowering drugs called statins were less likely to die from flu. By Julie Steenhuysen, National Post.

Cancer is at its root a disease of the genes. Whether a tumour affects the lungs, brain or breast, it is the result of a cell that has acquired the capacity to divide unchecked because of mutations in its DNA.

These mutations can be inherited, they can be the result of environmental carcinogens such as cigarette smoke or they can accumulate through errors copied during cell division — all three routes, indeed, are often involved.

This genetic background to all cancers offers an exciting route to better therapy, because the precise pattern of DNA defects that drives a tumour will influence its growth, its spread and the way it responds to treatment.

Drugs such as Herceptin, for breast cancer, are already taking healthcare towards more personalised medicine. Herceptin is highly effective but only against breast tumours with a particular genetic profile, and it must be prescribed accordingly.

Oncologists will be able to select the most appropriate treatment for almost any patient by sequencing the DNA of his or her tumour, and then using the combination of drugs that is likely to be most effective.

Two major hurdles stand in the way, however. The first is the cost of sequencing the genome of a tumour — about $50,000 (£31,362), although this will likely drop to about $1,000 within two years.

The second is a lack of understanding about which mutations and combinations of mutations drive tumours, and how they affect susceptibility to drugs. This is where projects such as the drug database at St George’s, University of London, will be pivotal.

The aim is to help researchers to “mix and match” drugs and cancer mutations, to determine combinations of defects that predict whether a given medicine is likely to work.

Last year, The Times revealed the launch of a similar initiative from the Sanger Institute near Cambridge and Massachusetts General Hospital in Boston. It will examine 1,000 colonies of cancer cells with known genetic defects, which will be exposed to 400 chemical agents to identify which colonies are susceptible to which drugs.

Such studies also have the potential to “rescue” drugs that have failed clinical trials for broad-spectrum use, either because they do not work for enough patients, or cause serious side-effects for a minority.

Some of these may prove to be both safe and effective for patients with a particular genetic profile. It is interesting to note that one “rescued” drug — thalidomide — is among those on the St George’s database.

What is happening now for cancer is also likely, in the longer term, to change the way that many other medical conditions are treated. Cancer may be the most obvious common disease with a genetic root, but it is far from the only one: disorders such as type 2 diabetes and heart disease also have a strong genetic component, and the DNA you inherit also affects the way your body metabolises drugs.

As more of these genetic variations become understood, doctors will increasingly start to practise “pharmacogenomics” for all manner of diseases, precribing drugs and other treatments according to their patients’ DNA profiles. The days when computerised tools like the St George’s database are used routinely in healthcare are probably not far off.  By Mark Henderson, The Times.