Sunday, November 8, 2009

The taxonomy of tumours

Medicine: A new technique aims to measure the activity of a tumour, and could also help provide a new way to classify cancers

ONCOLOGISTS would like to be able to classify cancers not by whereabouts in the body they occur, but by their molecular origin. They know that certain molecules become active in tumours found in certain parts of the body. Both head-and-neck cancers and breast cancers, for example, have an abundance of molecules called epidermal growth-factor receptors (EGFRs). Now a team from Cancer Research UK’s London Research Institute has taken a step towards this goal. Their technique can already identify how advanced a person’s cancer is, and thus how likely it is to return after treatment.

At present, pathologists assess how advanced a cancer is by taking a sample, known as a biopsy, and examining the concentration within it of specific receptors, such as EGFRs, that are known to help cancers spread. Peter Parker had the idea of employing a technique called fluorescence resonance-energy transfer (FRET), which is used to study interactions between individual protein molecules, to see if he could find out not only how many receptors there are in a biopsy, but also how active they are.

The technique uses two types of antibody, each attached to a fluorescent dye molecule. Each of the two types is selected to fuse with a different part of an EGFR molecule, but one will do so only when the receptor has become active.

Pointing a laser at the sample causes the first dye to become excited and emit energy. With an activated receptor, the second dye will be attached nearby and so will absorb some of the energy given off by the first. Measuring how much energy is transferred between the two dyes indicates the activity of the receptors.

Dr Parker’s idea was implemented by his colleague Banafshe Larijani. She and her colleagues used FRET to measure the activity of receptors in 122 head-and-neck cancers. They found that the higher the activity of the receptors they examined, the more likely it was the cancers would return quickly following treatment. The technique was found to be a better prognostic tool than conventional visual analysis of receptor density.

To speed things up, engineers in the same group have now created an instrument that automates the analysis. Tumour biopsies are placed on a microscope slide and stained with antibodies. The system then points the laser at the samples, records images of the resulting energy transfer and interprets those images to provide FRET scores. Results are available in as little as an hour, compared with four or five days using standard methods.

Having established the principle with head-and-neck cancer, the team hopes to extend it. They are beginning a large-scale trial to see whether FRET can accurately “hindcast” the clinical outcomes associated with 2,000 breast-cancer biopsies. Moreover, if patterns of receptor-activation for other types of cancers can be characterised, the technique could be applied to all solid tumours (ie, cancers other than leukaemias and lymphomas).

If they succeed, it will be good news for researchers who want to switch from classifying cancers anatomically to classifying them biochemically. Most cancer specialists think that patients with tumours in different parts of the body that are triggered by the same genetic mutations may have more in common than those whose tumours are in the same organ, but have been caused by different mutations. The new approach could help make such classification routine. That could, in turn, create a new generation of therapies and help doctors decide which patients should receive them, and in which combinations and doses.

Source of Information : The Economist 2009-09-05

No comments: