Noninvasive imaging diagnosis of sinusoidal obstruction syndrome: a pictorial review

As an important means of monitoring blood flow, US offers more information for early and differential diagnosis of SOS, especially in asymptomatic or late-onset cases [17]. Some findings of B-mode US, including hepatosplenomegaly, gallbladder wall thickening > 6 mm, portal diameter > 12 mm, hepatic vein diameter < 3 mm, and indirect signs suggest portal hypertension such as ascites and visualization of collateral circulatory, have been used as a routine criteria for SOS diagnosis. Doppler technique is a noninvasive method to improve the diagnostic efficiency of B-mode US by revealing morphologic changes and flow velocity fluctuation in hepatic vessels [18]. At present, the known characteristics of Doppler US of SOS include hepatic vein diameter < 3 mm, collateral circulatory visualization, demodulation of portal vein flow, spectral density decline, congestion index < 0.1, portal vein flow < 10 cm/s, hepatic artery resistive index > 0.75, and monophasic flow in hepatic veins. These hemodynamic changes are closely associated with the course of disease progression. At the early stage of the disease, the sinusoidal outflow tract has a smooth blood flow allowing portal vein to maintain a normal blood flow and morphology; however, as the disease progress, hepatic venous blood flow disappears and stenosis of the main hepatic veins and portal vein blood flow is reduce and reflux occurs in some severe cases [6, 19, 20]. Dynamic monitoring of changes in the above may help predict disease progression and venture a prognosis of patients with SOS. Also, three-dimensional structures of hepatic blood vessels may be viewed using multidetector CT (MDCT)/multi-planar reformation (MPR) in combination and liver CT angiography (CTA). A disordered intrahepatic vascular network, an enlarged hepatic artery, ill-defined small arteries, and invisible hepatic veins are regular features of SOS [21, 22]. Recently, some researchers have applied FDG-PET/CT to further ascertain functional changes of the liver due to altered hepatic hemodynamics. Compared with baseline images, a rise (~ 10%) in standard uptake value of the liver (SUV_liver) was observed after the onset of SOS. Presumably, sinus endothelial cell injury produced altered hemodynamics and microcirculatory obstruction, disrupting inlet/outlet flow patterns in hepatic vessels. Such changes then induce hepatic congestion and passively increase hepatic blood-pool FDG tracer activity [23].

Diffuse, patchy, or geographic heterogeneity of hepatic parenchyma, with a scattering of hypodense areas, is typical of SOS (Figs. 2, 3, and 4). Past studies indicate that parenchymal heterogeneity scores positively correlate with SOS clinical severity [13, 24, 25]. In patients with hepatic metastases of colon cancer, the efficacy of oxaliplatin-based chemotherapy is impacted by the existing SOS, and the severity of parenchymal heterogeneity is closely associated with a poor tumor response [26, 27]. Among all the heterogeneity features on enhanced CT or MRI images, cloverleaf or claw-like shapes are two classic manifestations of SOS [28]. These manifestations refer to the markedly enhanced area around the main hepatic veins which shows increased blood supply compared to the normal liver parenchyma. The appearance of characteristic shapes perhaps related to the small vascular channels surrounding the main hepatic veins. Because the initial lesion of SOS is located in hepatic sinusoid, which makes central lobular vein and sublobular vein become the first to be involved. As the disease worsens, obvious edema and necrosis of adjacent hepatocytes lead to the decrease of hepatic parenchymal enhancement; however, the small vascular channels adjacent to the main hepatic vein are kept unobstructed and result in the increase of enhancement of the liver parenchyma adjacent to the venules [12]. Recently, the hepatic pseudotumor of SOS has been noted after oxaliplatin treatment [22, 29]. Lesion of this sort is pathologically characterized by sinusoidal dilatation and congestion of hepatocytes with inflammatory cellular infiltration and areas of fibrosis [30]. On liver imaging, pseudotumor of SOS may simulate focal nodular hyperplasia, atypical HCC, or liver metastases because of the variety of enhancement patterns due to the pathological changes in the disease progression [30] (Fig. 5). On contrast-enhanced US, pseudotumor of SOS usually presents as an ill-defined hypoechoic lesion with a uniform and obvious hypervascularity in the arterial phase and a rapid wash-out appearance in the portal phase. On multi-phase dynamic-enhanced CT or MRI, an irregular lesion with peritumoral enhancement and central low attenuation/signal intensity is the most common manifestation of pseudotumor of SOS. For further improving differential diagnostic efficiency between pseudotumor of SOS and malignancy, a qualitative imaging modality, such as diffuse-weighted imaging (DWI) is recommended, which plays a role in the theoretical basis of the cellular density difference [23].

Susceptibility-weighted imaging (SWI) is a quantitative imaging method that exploits differences in magnetic sensitivity to evaluate tissue densities. At present, SWI is advantageous in assessing various pathologies, particularly hemorrhages, micro-bleeds, or iron deposition in the brain [31]. The pivotal event (i.e., endothelial cell injury) in SOS ultimately eventuates in erythrocytic decomposition and a glut of hemosiderin within the space of Disse, potentially detectable by SWI. Some researchers have tested this theory by adding SWI sequences to routine MRI studies. The patchy low-signal intensities of resultant SWI images seem to approximate a portal venous-phase distribution, thus confirming these pathologic underpinnings and affording a new and noninvasive means for early diagnosis of SOS [32]. MRI studies are frequently enhanced by superparamagnetic iron-oxide nanoparticles (SPION-MRIs) or gadolinium ethoxybenzyl diethylenetriamine pentaacetic acid (Gd-EOB-DTPA) to assess the function of hepatocytes. SPION-MRIs provide a quantitative gauge of damage to Kupffer cells in the course of SOS, exhibiting fair agreement between severity of SOS and its corresponding histopathology [33]. The hepatobiliary phase (HBP) of a Gd-EOB-DTPA-enhanced MRI is a better index of actual hepatocytic function. Previous investigations have shown that Gd-EOB-DTPA-enhanced MRIs deliver high specificity and good interobserver consistency in evaluating the progression of SOS [34]. Furthermore, Yoneda et al. have reported a relation between organic anion transporting polypeptides (OATP) and functionality of hepatocytes [13]. They have documented SOS-related functional damage, showing that low-signal intensities on HBP images correlate positively with degrees of hepatocytic injury. Of note, gradual hepatic functional recovery after interruption of chemotherapy is verifiable by continuous HBP scanning.

Fibroscan testing and acoustic radiation force impulse (ARFI) imaging, the most popular noninvasive transient elastography technologies, have been widely used to assess ongoing hepatic fibrosis and cirrhosis. Liver stiffness measurements (LSMs) determined by Fibroscan correspond with changes in the mechanical properties of the liver. Clinical increases in LSMs are seen in many conditions, including inflammation, fibrosis, congestion, cholestasis, and portal hypertension [35‐37]. The injury to sinusoidal endothelial cells inherent in SOS promotes endothelial cell embolization and sinusoidal obstruction, leading to subendothelial fibrosis of sinusoids and venules, and eventual congestive fibrosis of the liver. Thus, LSM elevations may be useful preclinical and surveillance biomarkers of SOS [38]. Indeed, earlier data suggest a pre-transplant LSMs cut-point of 8 kPa in predicting liver toxicity after HSCT, thus supporting the implementation of quantitative SOS diagnostics [39]. In addition, an animal experiment has indicated that liver shear-wave velocity (SWV), measured by ARFI imaging, bears a close relation to histologic scores of SOS, as well as degrees of hepatic lobular inflammation. Observed SWI outcomes were significantly higher in rats with SOS, as opposed to a matched control group [40]. Finally, hepatic SWV has been used in some studies (with promising results) as a quantitative biomarker to evaluate the treatment responses of patients with SOS [41].

Hepatosplenomegaly, ascites, a thickened gallbladder wall, and periportal edema are the chief imaging features of SOS [42] (Fig. 2). Such manifestations are hallmarks of portal hypertension, stemming from obstructed intrahepatic venous drainage, and the events that follow (i.e., congestion of the liver and spleen, loss of liver compliance, and hepatic injury). Because many diseases are similar in terms of clinical and laboratory parameters, Erturk et al. consider narrowed hepatic veins and collateral aspects of portal hypertension to be the features distinguishing SOS from other conditions [43].