THE ETIOLOGY AND PREDICTION OF BREAST-CANCER - FOURIER TRANSFORM-INFRARED SPECTROSCOPY REVEALS PROGRESSIVE ALTERATIONS IN BREAST DNA LEADING TO A CANCER-LIKE PHENOTYPE IN A HIGH PROPORTION OF NORMAL WOMEN

Citation
Dc. Malins et al., THE ETIOLOGY AND PREDICTION OF BREAST-CANCER - FOURIER TRANSFORM-INFRARED SPECTROSCOPY REVEALS PROGRESSIVE ALTERATIONS IN BREAST DNA LEADING TO A CANCER-LIKE PHENOTYPE IN A HIGH PROPORTION OF NORMAL WOMEN, Cancer, 75(2), 1995, pp. 503-517
Citations number
43
Categorie Soggetti
Oncology
Journal title
CancerACNP
ISSN journal
0008543X
Volume
75
Issue
2
Year of publication
1995
Pages
503 - 517
Database
ISI
SICI code
0008-543X(1995)75:2<503:TEAPOB>2.0.ZU;2-C
Abstract
Background. The authors previously have shown by gas chromatography-ma ss spectrometry that the hydroxyl radical( OH) induces alterations in the DNA base structure of the female breast, which are premalignant ma rkers of breast cancer. Fourier transform-infrared (FT-IR) spectroscop y also has a high potential for revealing a broad array of structural changes in DNA that may provide important new insight into breast canc er etiology and prediction. Methods. DNA from normal reduction mammopl asty tissue, invasive ductal carcinoma, and nearby microscopically nor mal tissue was analyzed by FT-IR spectroscopy. Statistical models base d on DNA spectral properties were developed and compared with a statis tical model previously used with base modifications. Results. Substant ial differences were found in the spectral properties of DNA from wome n with normal and cancerous breast tissue, indicating an ability to di scriminate cancerous tissue from noncancerous tissue with a sensitivit y and specificity of 83%. Most importantly, the normal population was divided into subgroups in which a nonrandom progression was identified and a cancer-like DNA phenotype that was highly correlated (r greater than or equal to 0.90) with that of the patients with cancer was exhi bited in 59% of the women. The spectral data, which also were highly c orrelated with the base-model data, were used to establish a model for predicting the probability of breast cancer. Consistent with the high cancer reoccurrence rate in the ipsilateral breast, 8 of 10 of the mi croscopically normal tissue specimens remaining after tumor excision w ere classified as cancerous using this model. Conclusions. Progressive structural changes in the DNA of the normal female breast, leading to a premalignant cancer-like phenotype in a high proportion of women, a re the basis for a new paradigm for understanding the etiology of brea st cancer and predicting its occurrence at early stages of oncogenesis . The results also suggest therapeutic strategies for potentially reve rsing the extent of DNA damage, which may be useful in disease prevent ion and treatment.