Detecting breast cancer with non-contrast MR imaging: combining diffusion-weighted and STIR imaging.

We combined diffusion-weighted (DWI) and short TI inversion recovery (STIR) imaging to evaluate the diagnostic capability of non-contrast magnetic resonance (MR) imaging to detect breast cancer. Seventy women patients underwent mammography and MR imaging with combined DWI (b factor: 1000) and STIR that revealed malignancy, and postoperative pathological examination confirmed breast cancer. Interpreted images were evaluated for sensitivity, false negative rate (FN), sensitivity by pT, and sensitivity by background density of the mammary gland. Of the 70 cases, 68 were diagnosed as cancer by DWI and STIR (sensitivity, 97% [68/70]; FN, 2.9% [2/70]). Sensitivities by pT were: pTis, 67% (4/6); pT1, 100% (33/33); and pT2-4, 100% (31/31). No significant differences were observed in sensitivity between pT1 and pT2-4 (P<0.001). Sensitivities by background density of mammary gland were: fatty/scattered fibroglandular tissue, 95% (20/21) and heterogeneous fibroglandular tissue/mostly fibroglandular tissue, 98% (48/49). No significant differences were observed (P<0.001). Two cases, an intraductal and an apocrine carcinoma, were incorrectly diagnosed by MR imaging. Precise diagnosis of breast cancer is possible with combined DWI and STIR, even in non-contrast MR imaging, regardless of the diameter or background density of mammary gland. It is hoped that non-contrast MR imaging that combines DWI and STIR will become an established clinical screening method.