Skip Navigation

Breast cancer drug gets helpmate

August 23, 2004 By Helen Pearson This article courtesy of Nature News.

Remedy suggested for patients that do not respond to Herceptin.

Please log in to rate this page.

View Comments

Researchers have discovered why the cancer drug Herceptin fails in many patients, and may have hit on a way to improve the therapy. The finding is a step towards personalized cancer treatments, in which doctors prescribe cocktails of therapies tailored to individual patients.

Herceptin, whose generic name is trastuzumab, was one of the first cancer drugs to specifically target a rogue molecule, and was widely applauded at its 1998 launch. It is used to tackle a particularly aggressive form of breast cancer that makes up 25-30% of cases and from which patients have a very poor chance of survival. In these women, the tumour ramps up production of a protein called ErbB2.

Herceptin curbs the growth of the tumour cells by binding ErbB2. But despite its dramatic impact in some patients, only around 35% respond, and all those who take it risk severe side effects such as heart attacks.

Now Dihua Yu at the University of Texas, Houston, and her colleagues have improved our understanding of how Herceptin works, by examining tumour cells treated with the drug.

Not enough PTEN

When it binds ErbB2, Herceptin also boosts the activity of another protein called PTEN, which is known to slow cancer cell growth, the researchers report in Cancer Cell1.

After examining nearly 50 patient samples, the team went on to show that women who do not respond well to Herceptin tend to lack a normal quantity of PTEN protein in their tumours. This could be one reason why the drug does not work properly in these patients.

You could almost promise a patient they would respond.
Dihua Yu
University of Texas, Houston
The result suggests that breast cancer patients could be screened for the presence of PTEN in their tumours, to determine who is a good candidate for the drug and who should avoid taking the expensive and potentially dangerous treatment. "We are very, very excited about this," says Yu. "You could almost promise a patient they would respond."

However, screening for PTEN might still miss a few patients who will not be helped, cautions Matthew Ellis, who directs the breast cancer programme at Washington University in St Louis, Missouri. Patients might, for example, have mutations in another gene that have the same effect as missing PTEN.

Yu's team has also come up with a way to help Herceptin work in these patients who lack PTEN, using another drug that mimics the effects of the PTEN protein. When they used this therapy alongside Herceptin in mice with breast cancer, the tumours shrank to nearly half the size of those in groups treated with only one of the two drugs.

Personal cocktail

Neither of the new results is ready to be used in the clinic. First, doctors must examine whether testing for PTEN does indeed predict how a woman will respond to Herceptin. And the new drug will have to be tested alongside Herceptin in clinical trials.

The finding follows two studies published earlier this year, which showed why another cancer drug called Iressa (gefitinib) only works in 10% of patients with lung cancer. The teams involved found that those who do respond have specific mutations in the gene targeted by the drug23.

Such studies highlight the hope of doctors for future cancer treatment: that they might be able to diagnose a cocktail of drugs based on the characteristics of an individual patient's tumour. "It's a good example of how we're making progress in personalized medicine," Ellis says.


  1. Nagata Y., et al. Cancer Cell, 6. 117 - 127 (2004).
  2. Paez J.G., et al. Science, 304. 1497 - 1599 DOI: 10.1126/science.1099314 (2004).
  3. Lynch T.J., et al. NEJM, 350. 2129 - 2139 (2004)


User Tools [+] Expand

User Tools [-] Collapse

Pinterest button


Please log in to add this page to your favorites list.

Need Assistance?

If you need help or have a question please use the links below to help resolve your problem.