61 Diagnostic utility of molecular and imaging biomarkers 2 distinguish thyroid carcinoma [329]. Among the individual studies, however, presented optimum ADC thresholds varied and were not externally validated [319-321, 324-328]. Only one small study had assessed DW-MRI in indeterminate thyroid nodules to date. Nakahira et al. reported a mean ADC value of 1.27 ± 0.29 *10-3 mm2/s in malignancies opposite 1.95 ± 0.24 *10-3 mm2/s in benign nodules with indeterminate cytology. These results were similar to those of their entire study population (n=42), in which a cut-off ADC value of 1.95 ±0.24 *10-3 mm2/s was 95% sensitive and 83% specific [319]. Availability and limitations of DW-MRI DW-MRI is infrequently and only experimentally used in the workup of thyroid nodules. Nonetheless, the worldwide availability and application of MRI is growing. As it uses no ionizing but only radiofrequency radiation, the associated risk to the patients is limited, provided that specific measures are taken for patients with MRI-incompatible implanted devices or metal. No MRI-contrast is necessary for DW-MRI, thus avoiding gadolinium-associated toxicity. As the spatial resolution of MRI-scanners is still improving, technical limitations of DW-MRI with regard to minimal lesion size are becoming less relevant compared to SPECT and probably also PET. Still, spatial resolution of DW-MRI sequences is less than that of conventional anatomical MRI-sequences. There are several major limitations to DW-MRI. MRI is still a rather costly technique; additional sequences such as DW-MRI adds scanner time (~5-10 min) per patient and thus further increases costs. DW-MRI methodology is not standardized yet and its optimal settings still unsettled, leading to varying ADC and b-values [319, 325, 328]. Suboptimal methodology or artifacts cause poor image quality, impede accurate interpretation and caused undesirable exclusions from already small-sized studies, with reported exclusion rates up to 28% [319, 320, 328]. Image artifacts are often caused by inhomogeneity in pathologic tissues or by their vicinity to interfaces between soft-tissues and bone or air, a source of MRI-artifacts specifically in the thyroid region. Besides viable tumour tissue, malignant tumours partly exist of components with high diffusivity, such necrosis, cystic components or intratumoral haemorrhage [319, 321]. For accurate ADC measurement, such macroscopic areas should be manually avoided when drawing a region-of-interest. However, avoiding microscopic areas of similar origin, invisible to the human eye, is an impossible task [319]. Furthermore, it is hypothesized that the substantial amounts of follicular or Hürthle cells limit the diagnostic accuracy of DW-MRI, specifically in indeterminate thyroid neoplasms. Follicular and Hürthle cell neoplasms are known for their varying colloid tissue involvement. Histologically they contain more fluid. Thus, DW-MRI would inaccurately provide a more benign image [319, 328]. These hypotheses are currently based on very limited evidence. Further prospective validation studies are desired to determine the possible diagnostic value of DW-MRI in indeterminate thyroid nodules. Future prospects also include improvements of the technique, including consensus on methodology and standardization of acquisition techniques.
RkJQdWJsaXNoZXIy MTk4NDMw