No one doubts the success of Herceptin in treating advanced cases of HER2 positive breast cancer. No one questions its use in such cases. However, last month two dramatic sets of early study results published in the New England Journal of Medicine addressed its effect on early-stage breast cancers. Together they announced how Herceptin’s use in such cases may halve the chance of it recurring, prompting talk of a breast-cancer cure.

Less widely quoted was last week’s review in the Journal of the American Medical Association of randomised trials that were stopped early because the drugs worked so well it was deemed unethical to deny them to the placebo group.

Of course, positive studies have a huge influence on doctors and patients. But the reviewers, led by Professor Gordon Guyatt, of McMaster University Ontario, concluded that positive trials are too often cut short before they gather enough evidence to be conclusive. Many studies last years but, on average, those with good results stop after 13 months. So there is a high risk of overestimating benefits and underestimating side-effects.

One example involves the drug Bisoprolol, which is used for heart patients having vascular surgery. Researchers announced that it reduced the chance of death, or heart attack, by 91 per cent. But longer-term studies revealed this to be a statistical fluke.

Both October’s Herceptin studies were interim results, made public before they were completed. As The Lancet made clear, they contained no overall figures on how many women survived or developed heart problems because of the drug. The first Herceptin trials revealed a 7 per cent risk of heart failure when using the drug alone, with an increase to 28 per cent when it was used in conjunction with other drugs.

A report five years ago in the heart journal Circulation compared the risks. It pointed out that long-term survival of women with heart failure (33 per cent at six years) is equivalent to women with breast cancer that has spread to a distant site, but is worse than women with breast cancer and local spread (77 per cent at five years).

How much worse it would be for women with early breast cancer is not yet clear. With current treatments, four out of five women with early breast cancer survive. The fear of recurrence should not obscure the possibility that sufferers may be jumping from the frying pan into the fire on the basis of interim results.

Explore Health

• Health Features

• Expert Advice

• Alternative Medicine

• Mental Health

• Child Health

• Health Club