| Abstract: |
Aims and background: There is a need for a cost-effective method to safely reduce the number of diagnostic procedures women undergo for breast cancer. We tested a new procedure for breast cancer diagnosis based on breast tissue response to low level electromagnetic incident waves. Methods: We tested 101 patients with suspicious palpable breast lesions detected by mammography or ultrasonography, who were scheduled to undergo an open biopsy. Using an electromagnetic field generator (tissue resonance interaction method probe [TRIMprob™]), we passed the TRIMprob™ over the breast area and recorded the signal variation of one or more spectral lines (dB1, dB2, dB3). The results were compared with those of a control group as well as with pathology data obtained from excisional biopsy. Results: No adverse effects of the test were observed. Pathology revealed 86 malignant breast cancers (72 invasive, 14 in situ) and 15 benign conditions. We achieved the best discrimination between normal breasts and lesions using dB1 (dB1 AUC-ROC = 0.8; dB2 AUC-ROC = 0.61; dB3 AUC-ROC = 0.76). With a specificity of 75% to 95%, the sensitivity ranged from 49% to 84%. Tumor or patient variables did not influence the results. Conclusions: The TRIMprob™ test was able to provide some degree of discrimination between normal breast tissue and lesions but not between benign and malignant lesions. The lack of influence of patient age and tumor size on test results might be advantageous in terms of early diagnosis in young women. These preliminary results need to be verified and extended in a preclinical-stage disease setting before clinical applicability can be envisaged. |
