Editor—The use of hair as a biopsy tissue has been considered for some time. For instance, in the case of breast cancer, raised zinc levels in head hair have been reported.1 Besides, x ray diffraction patterns of hair are rich and have attracted much attention for 70 years.2 However, its potential use as a diagnostic indicator of disease was only suggested a short time ago.3 Most recently, James et al 4 reported that x ray diffraction of hair taken from women diagnosed with breast cancer (and those at high risk by virtue of a provenBRCA1/BRCA2mutation) showed a diffuse ring. They claimed a 100% correlation with the disease, advocating the use of pubic hair as a simple non-invasive screening method for breast cancer. The use of pubic hair was suggested in view of possible damage to the head hair from cosmetic treatments. Despite this note of caution, the study of James et al 4 was based on 12 pubic hair samples with only eight from cancer affected subjects. Here, we report a detailed double blind study from 109 women belonging to five clinically distinct groups as well as a normal population group and show that there is no correlation between the diffuse ring and breast cancer or breast cancer predisposition..

The present work was initiated in November 1998 to provide an independent double blind study on clinically well controlled samples because of our concerns with the study of Jameset al 4 for which a major proportion of the samples was provided by two of us (AH and DGRE). Both diffraction and x ray fluorescence data have been obtained. Diffraction data were collected on station 2.1 with the multiwire area detector and the fluorescence data on station 7.1 of the UK Synchrotron Radiation Source (SRS). The use of a multiwire detector allowed on line alignment of hair within seconds and enabled efficient diffraction data collection from a large number of samples. Six groups of subjects were selected including a normal population group. For each subject, head and pubic hair were collected and reference numbers assigned in a random manner. The identities of samples were kept blinded until all data were analysed. Each woman was asked to provide information on whether they had had any hair treatment (perm, dye, etc) or were on medication. No hair treatment was reported for pubic hair and thus here only results from pubic hair (108 samples) are discussed. Results of head hair are included in table 1 for completeness. The groups consisted of 27 unaffected controls aged 26-60 years, 21 isolated cases of breast cancer (<31 years) who had been screened negative forBRCA1/BRCA2 mutations, and three cases aged over 60 years not so tested. The remaining groups came from a set of 43 families with proven BRCA1/BRCA2mutations: 25 affected mutation carriers, 10 unaffected mutation carriers, and 23 unaffected close female relatives who tested negative for the known mutations in the family.

Diffraction patterns could be grouped into two categories: one characterised by a diffuse ring at 4.78 ± 0.10 nm and one with no ring present (fig 1). The occurrence of this ring is well known fromx ray patterns of both keratins and muscle and is ascribed to lipid crystals resulting from degradation processes.5 Contrary to the observation of Jameset al,4 the pubic hair of only 12 of 49 (25%) women with breast cancer showed the diffuse ring. This is only slightly larger than in the normal population group, where about 20% of the pubic hair samples showed the ring. In the case of the affected group who tested positive for aBRCA1 or BRCA2mutation, 56% of the pubic hair samples showed no ring and only 24% of the pubic hair samples showed the ring. Table 1 provides a detailed summary of diffraction results for all six groups. A statistical analysis for pubic hair samples was performed to determine whether there was an association between the presence of the diffuse ring and breast cancer. There were 49 samples from women with breast cancer and 59 from unaffected women. A χ² value of 0.86 was obtained. For 1% significance, a value of 6.63 and for 5% significance, a value of 3.84 is required. Thus, it can be concluded that there is no measurable association between the diffuse ring and breast cancer. The trace element (Zn, Cu, Fe, and S) analysis of intact hair showed no correlation with the ring structure in the diffraction pattern or with the subjects' group. The women in the normal population group whose hair had shown the diffuse ring were examined and shown not to have breast cancer.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

X ray diffraction patterns from the pubic hair of women diagnosed with breast cancer. The left half represents the pattern with no ring, the right half shows the diffuse ring. The first strong meridional reflection (arrow 1) is used for normalising the patterns, arrow 2 shows the equatorial reflection, which gets intensified in cases when the diffuse ring appears.  

Our x ray diffraction data do not support the recent claim that hair from breast cancer patients or those at high risk (BRCA1/BRCA2 mutation carriers) show a distinct diffuse ring. This conclusion for breast cancer diagnosis was also reached on a much smaller study of head hair only.6In our study, diffraction patterns from 75% (37 of 49) of the breast cancer patients do not show this ring. Moreover, the χ²test shows no association between the diffuse ring and breast cancer and, as such, the claim that x ray diffraction of pubic hair can be used as a screening method for breast cancer or breast cancer predisposition is invalid.