Differential diagnosis of axillary inflammatory and metastatic lymph nodes in rabbit models by using diffusion-weighted imaging: compared with conventional magnetic resonance imaging

Abstract

Objective: This experiment aims to determine the diagnostic value of diffusion-weighted imaging (DWI) in the differentiation of axillary inflammatory lymph nodes from metastatic lymph nodes in rabbit models in comparison with conventional magnetic resonance imaging (MRI).

Materials and methods: Conventional MRI and DWI were performed at 4 weeks after successful inoculation into the forty female New Zealand white rabbits' mammary glands. The size-based and signal-intensity-based criteria and the relative apparent diffusion coefficient (rADC) value were compared between the axillary inflammatory lymph nodes and metastatic lymph nodes, with histopathological findings as the reference standard. Receiver operating characteristic (ROC) curve analysis was performed to evaluate the diagnostic performance of the aforementioned criteria and rADC value in differentiating the axillary inflammatory lymph nodes from metastatic lymph nodes.

Results: Thirty-two axillary inflammatory lymph nodes and 46 metastatic ones were successfully isolated and taken into pathological analysis. The differences of the aforementioned criteria between the two groups were not statistically significant (p > 0.05). However, the rADC value of the inflammatory lymph nodes (0.9 ± 0.14) was higher than that of metastatic ones (0.7 ± 0.18), with significant difference (p = 0.016). When the rADC value was chosen as 0.80, the area under the ROC curve is greater than all other criteria, and the sensitivity, specificity, accuracy, positive predictive value, and negative predictive value for differentiating two groups were 86.2%, 79.3%, 81.2%, 84.2%, and 85.6%, respectively.

Conclusion: Diffusion-weighted imaging is a promising new technique for differentiating axillary inflammatory lymph nodes from metastatic lymph nodes. Compared with routine magnetic resonance sequences, DWI could provide more useful physiological and functional information for diagnosis.

Keywords: Animal model; Axillary lymph node; Diffusion-weighted imaging; Relative ADC.

Fig. 1

Manifestations of gross specimen and histopathological examination of axillary inflammatory and metastatic lymph nodes. Gross specimen of axillary inflammatory lymph node show that they have smoothly pinkish-grey surfaces and abundant vessels (A), while axillary metastatic lymph nodes appear to have shaggy pale-grey surface with fish like texture (B). Histopathological examination (H & E, × 200) shows inflammatory cell infiltration and cells arrangement sparsely with intercellular substance edema (C), while tumor cells deposit within metastatic lymph node, and enlarged nuclei, hyperchromatism, high nuclear-to-cytoplasmic ratio, and pathological mitotic figure are observed (D).

Fig. 2

Appearance of conventional MRI, diffusion-weighted imaging and apparent diffusion coefficient map of axillary inflammatory lymph node. Inflammatory lymph node in right axillary fossa on axial T2WI shows iso-intensity with unclear borderline (arrow) (A). Axial T2WI shows high signal intensity (arrow) (B). Corresponding diffusion-weight imaging demonstrates homogeneous high signal intensity (arrow), and conspicuity in identification of same node is higher in this image compared with T2WI (C). Node is depicted as area of green in most part, with scattered yellow dots (arrow) on pseudo-color apparent diffusion coefficient map, and relative apparent diffusion coefficient value of this node is 0.86 (D).

Fig. 3

Manifestation of diffusion-weighted imaging and apparent diffusion coefficient map of axillary metastatic lymph nodes. Two axillary metastatic lymph nodes in bilateral axillary fossa on axial diffusion-weighted imaging show obviously homogeneous high signal intensity with clear margin (arrow) (A). Both of them demonstrated as areas of blue on corresponding apparent diffusion coefficient map (arrow), and relative apparent diffusion coefficient value of left and right nodes is 0.60, 0.55, respectively (B).

Fig. 4

Graph shows that receiver operating characteristic curve for rADC value is superior to size-based criteria (short-diameter, long-axis diameter, and long/short-axis ratio) and signal intensity on T2WI and DWI for differentiation of inflammatory from metastatic lymph nodes. DWI = diffusion-weighted imaging, rADC = relative apparent diffusion coefficient