IUCr journals

Breast cancer diagnosis using scattered X-rays

J. Synchrotron Rad. (2000) 7, 348-352 [doi:10.1107/S0909049500009973]

Breast cancer is a leading cause of cancer deaths in women and the lifetime risk of a woman developing breast cancer is as high as 10-13% in some western countries. Screening mammography is widely used in developed countries for the detection of tumors as early detection offers the best hope of treatment to improve life expectancy. Currently diagnosis is based on the so-called triple assessment - a combination of physical examination, mammography using X-rays and/or ultrasound, and fine needle aspiration cytology or core-biopsy. The histopathological assessment relies on alterations to cellular morphology and tissue architecture. Although triple assessment is effective in detecting most malignant lesions, its suboptimal specificity means that in some cases open biopsy surgery is required to exclude malignancy. There is therefore a considerable interest in developing diagnostic methods that are specific for the presence of malignant breast tissue.

Using the Daresbury SRS at beamline 2.1, Lewis, Rogers, Hall, Towns-Andrews, Slawson, Evans, Pinder, Ellis, Boggis, Hufton & Dance [J. Synchrotron Rad. (2000). 7, 348-352] carried out synchrotron X-ray diffraction studies of core-cut biopsy specimens taken from normal breast tissue, from benign tumors and from malignant tumors. They found that the diffraction patterns obtained from the three types of specimen showed significant differences in the spacing and scattered intensity of a prominent axial peak (see below). This peak arises from the collagen fibrils in the connective tissue of the breast and is associated with the regular packing of the tropocollagen molecules in the fibril. The intensity of the axial peaks is very sensitive to the degree of order in the orientation and packing of the collagen fibrils, and as might be expected, the disruption of the connective tissue associated with malignancy reduces the intensity of these peaks.

[scatter plot] Scatter plot of the fraction of scattered intensity in the diffraction peaks versus the spacing of the third-order axial reection. Ellipses are shown to illustrate the clustering.
An unexpected finding was that the measured values of the third-order spacings in benign and malignant tumors were grouped in separate clusters, as shown in the figure. The D period in tendon collagen is known to be sensitive to mechanical deformations, hydration and ionic environment, and further work would be required before these differences in spacings could be attributed with confidence to in vivo differences between the collagens.

This is an exciting development; the authors have succeeded in identifying a specific test for malignancy in breast tissue - the big challenge will be to develop a procedure suitable for routine use. In their closing remarks they state that they are currently considering the possibility of collecting diffraction data in vivo. This would have the great advantage of being non-invasive, but it is difficult to see how this could be achieved as the wavelengths of the X-rays used for diffraction studies are such that they will only penetrate a few millimeters into breast tissue before they are extinguished.

R.D.B. Fraser DSc, FAA, Noosa, Queensland, Australia
S.E. Fraser MB BS FASBP, Cairns Breast Cancer Screening Facility, Queensland, Australia

Authors' reply

The reviewers are certainly correct that the change in the collagen repeat spacing (the D spacing) could be explained by mechanical deformation or varying hydration levels. We were aware of this at the time of the experiment and took great care to maintain full hydration at all times and mounted all the samples in a similar manner. However, it is quite possible that this effect could arise from different stretching occurring for different tissues within the capillary tubes even though we did allow considerable time for relaxation between sample mounting and measurement. Nevertheless, we agree with the reviewers that there are multiple possibilities for our observation and intend to study this effect further.

The absorption problem referred to by the reviewers would be reduced if the measurement were performed in a manner similar to a localized compression view in mammography. Furthermore, we would use a technique similar to our colleagues at Philips Research (Harding & Schreiber) who have shown it is possible to identify materials in baggage scanning systems with very few scattered photons. Our current ideas for bringing the technique closer to the clinic are based around modified mammography equipment operating at doses and wavelengths commensurate with a normal mammogram. Obviously there is a very long way to go before we could see this used in the clinic, and these are only ideas at present, but sometimes we need to dream.


Harding, G. & Schreiber, B. (1999). 'Coherent X-ray scatter imaging and its applications in biomedical science and industry,' Rad. Phys. Chem. 56, 229-245.