HomeRadiologyVol. 298, No. 2 PreviousNext Reviews and CommentaryFree AccessEditorialRecognizing the Role of the Reticuloendothelial System in the Late Phase of US Contrast AgentsKorosh Khalili , Mostafa Atri, Tae Kyoung Kim, Hyun-Jung JangKorosh Khalili , Mostafa Atri, Tae Kyoung Kim, Hyun-Jung JangAuthor AffiliationsFrom the Department of Medical Imaging, University of Toronto, 610 University Ave, Room 3-964, Toronto, ON, Canada M5G 2M9; and Joint Department of Medical Imaging, University Health Network, Sinai Health System, Women’s College Hospital, Toronto, Canada.Address correspondence to K.K. (e-mail: [email protected]).Korosh Khalili Mostafa AtriTae Kyoung KimHyun-Jung JangPublished Online:Dec 8 2020https://doi.org/10.1148/radiol.2020203245MoreSectionsPDF ToolsImage ViewerAdd to favoritesCiteTrack CitationsPermissionsReprints ShareShare onFacebookTwitterLinked In Current US contrast agents are generally categorized into two types: agents taken up by the reticuloendothelial system (RES), which demonstrate a “postvascular phase” after an initial intravascular phase (eg, Sonazoid [Daiichi Sankyo, Tokyo, Japan]), and “blood pool” agents confined within the intravascular compartment (eg, Sonovue and/or Lumason [Bracco Imaging, Milan, Italy], Definity and/or Luminity [Lantheus Medical Imaging, North Billerica, Mass], and Optison [GE Healthcare Chicago, Ill]) (1,2). However, the behavior of the blood pool US contrast agents in various organs and liver lesions raises questions regarding the exact behavior of these agents in the late phase. In an early study of the spleen, Lim et al (3) described “splenic specificity” with Sonovue and asked, “Are [microbubbles] phagocytized by macrophages or merely trapped within the splenic parenchyma?” The late phase retention is not unique to the spleen (Fig 1). As noted by Omar et al (4), “The spleen shares the property of sequestering and retaining ultrasound contrast microbubbles with the liver.…” Blood pool US contrast agents remain within the liver parenchyma after renal enhancement visibly concludes (Fig 1). Why is there retention of contrast agent in the liver and the spleen with blood pool US contrast agents? How can this aid the diagnosis of specific lesions within the liver? Herein, we present pharmacokinetic, laboratory, veterinary, and human data that suggest the late phase behavior of US contrast agents is related to the RES and discuss the relevant clinical implications.Figure 1a: Contrast-enhanced US scans in a 32-year-old man without liver disease. (a) Transverse image of upper abdominal organs obtained 6 minutes after injection of 0.2-mL Definity shows late phase retention of the contrast agent in the liver (L), but to a lesser degree than in the spleen (S). (b, c) Sagittal images of the kidneys (K) obtained at 7 minutes demonstrate no enhancement, whereas the liver (L) and spleen (S) retain the contrast agent.Figure 1a:Download as PowerPointOpen in Image Viewer Figure 1b: Contrast-enhanced US scans in a 32-year-old man without liver disease. (a) Transverse image of upper abdominal organs obtained 6 minutes after injection of 0.2-mL Definity shows late phase retention of the contrast agent in the liver (L), but to a lesser degree than in the spleen (S). (b, c) Sagittal images of the kidneys (K) obtained at 7 minutes demonstrate no enhancement, whereas the liver (L) and spleen (S) retain the contrast agent.Figure 1b:Download as PowerPointOpen in Image Viewer Figure 1c: Contrast-enhanced US scans in a 32-year-old man without liver disease. (a) Transverse image of upper abdominal organs obtained 6 minutes after injection of 0.2-mL Definity shows late phase retention of the contrast agent in the liver (L), but to a lesser degree than in the spleen (S). (b, c) Sagittal images of the kidneys (K) obtained at 7 minutes demonstrate no enhancement, whereas the liver (L) and spleen (S) retain the contrast agent.Figure 1c:Download as PowerPointOpen in Image Viewer TerminologyHerein, we use European Federation of Societies for Ultrasound in Medicine and Biology terminology for the phases of enhancement, with the portal phase meaning 45–120 seconds after injection and the late phase meaning 2–5 minutes after injection (1).Pharmacokinetic DataUpon injection of an intravenous bolus of contrast agent, drugs rapidly mix with blood in the intravascular compartment and reach a uniform concentration. This is the distribution phase. Classic pharmacokinetic experiments have shown that the duration of the distribution phase of drugs in the intravascular compartment is typically 45–60 seconds (5). Strictly intravascular drugs, such as US contrast agents, reach an equilibrium (also known as steady state) within 1 minute after injection. Most hepatocellular carcinomas (HCCs) enhance (wash-in) wholly in the arterial/portal venous phase. Some HCCs show washout during the portal phase, but others exhibit late phase washout (Fig 2). It is difficult to explain the mechanisms of late phase washout using a strictly intravascular model of US contrast agents. Washout has been previously explained through differences between the lesion and the liver in arterial and portal venous inflow or their respective intravascular volume–to–total volume ratios (6). However, these differences should manifest well before the 2-minute mark of the late phase. With CT and extracellular MRI contrast agents, delayed enhancement and/or washout has been explained by the invocation of a second compartment, the interstitial space. But with US contrast agents, in a hypervascular mass such as HCC, late washout requires a different mechanism because the microbubbles are too large to leak into the interstitium. The only other described mechanism of pharmacokinetics for US contrast agents in vivo is through the invocation of RES (described in the next section), which can explain the late washout phenomenon.Figure 2a: US images in a 73-year-old man with hepatitis C–induced cirrhosis and late phase washout of hepatocellular carcinoma. Images were obtained in the delayed phase of the contrast-enhanced study, 2 and 4 minutes after injection of 0.2-mL Definity. (a) Contrast-enhanaed image obtained at 2 minutes (left) shows absent washout of a small hypoechoic nodule (arrow) seen on the gray-scale image (right). (b) Image obtained at 4 minutes after injection of Definity (left) shows that there is clear washout of the nodule (arrow) seen on the enhanced image (right).Figure 2a:Download as PowerPointOpen in Image Viewer Figure 2b: US images in a 73-year-old man with hepatitis C–induced cirrhosis and late phase washout of hepatocellular carcinoma. Images were obtained in the delayed phase of the contrast-enhanced study, 2 and 4 minutes after injection of 0.2-mL Definity. (a) Contrast-enhanaed image obtained at 2 minutes (left) shows absent washout of a small hypoechoic nodule (arrow) seen on the gray-scale image (right). (b) Image obtained at 4 minutes after injection of Definity (left) shows that there is clear washout of the nodule (arrow) seen on the enhanced image (right).Figure 2b:Download as PowerPointOpen in Image Viewer In a study of US contrast agent pharmacokinetics in the liver, Shunichi et al (7) measured the arterial, portal venous, hepatic venous, and parenchymal enhancement of Sonazoid. In confirmation of classic pharmacokinetics, they showed that the concentration of the agent reaches a plateau in the hepatic artery and portal and hepatic veins at approximately 50 seconds after injection; this is the equilibrium point. They also showed that at approximately 1 minute, the enhancement of the liver parenchyma surpasses that of the blood vessels and the two rapidly diverge as Sonazoid concentration drops in the vascular space, proving that a second mechanism of distribution is in effect. In case of Sonazoid, the liver and splenic sequestration has been shown to be by the RES. Parenchymal sequestration of blood pool US contrast agents is also seen in the liver and spleen. That is why in the latter stages of the late phase (4–5 minutes), the kidney, an organ with near-identical intravascular volume to the liver, lacks any enhancement while the liver remains bright (Fig 1) (8). All common US contrast agents in clinical use around the world are retained by the liver and spleen in the late phase, with their differences being the propensity for the retention and their in vivo stability (half-life). But what is the mechanism of the liver and/or spleen sequestration of the US contrast agents?In Vitro StudiesPrevious studies demonstrated Kupffer cell binding and/or phagocytosis of US contrast agents, but the study by Yanagisawa et al (9) performed a comprehensive analysis of five US contrast agents using cultured and in vivo rat Kupffer cells. They demonstrated that Kupffer cells phagocytized four of five US contrast agents. Although Sonazoid and Optison demonstrated the highest proportion of phagocytosis (99%), Sonovue also showed a weak rate of phagocytosis (7%). Furthermore, Yanagisawa et al noted that Sonovue bubbles had attached to the Kupffer cells (without phagocytosis) but some later detached. In addition, the authors quantified in vivo hepatic contrast agent retention at 10 minutes without and with RES inactivation (by means of GdCl3). With RES inactivation, there was a drop in the enhancement of the liver by 13% for Sonovue as compared with 47% for Sonazoid. To summarize, the study by Yanagisawa et al showed that Sonovue microbubbles attach to the RES in rats and some become phagocytized, whereas the inactivation of the RES results in a drop in the late phase liver retention.The study by Yanagisawa et al is cited as evidence for differentiating US contrast agents into strict blood pool and dual phase agents, rather than recognizing the similar behavior of the agents albeit to variable degree. Sonovue is not nearly as stable as Sonazoid, and it is possible that repeating the experiment at an earlier timepoint (eg, 5 minutes after injection rather than 10 minutes after injection) would have prevented a much higher proportion of Sonovue retention by means of RES inactivation. The varying affinity of the diverse surface receptors of the RES to the shell of the US contrast agents may explain their different rate of binding (9). Nevertheless, given the sensitivity of contrast-enhanced US (which allows visualization of individual bubbles using modern equipment), a low rate of binding and/or phagocytosis may still affect enhancement behavior, as happened in the clinical studies discussed in the subsequent sections.Yanagisawa et al did not study Definity. However, Miller and Dou (10) performed experiments investigating cell injury by ultrasound exposure with Definity and Optison. They found that both contrast agents became attached to monolayers of cultured mouse macrophage-like phagocytic cell lines but not endothelial cells (10).Veterinary StudiesThe spleen and RES-containing splenic lesions show retention of blood pool US contrast agents in dogs. In one large clinical series, normal canine spleens (four of four) and canine spleens with reactive and nodular hyperplasia (18 of 18) all showed late phase (5 minute) retention of Sonovue (11). Extramedullary hematopoiesis, a common splenic mass in dogs and inclusive of reticuloendothelial cells, also demonstrated late phase retention of Sonovue in four of six lesions. Conversely, hemangiosarcomas (which do not have RES activity) appeared hypoechoic to the normal spleen in the delayed phase in 10 of 10 lesions (11).Human StudiesExtrahepatic ApplicationsThe late phase retention of the contrast agent is the specific diagnostic feature in normal splenic tissue for both Sonovue and Definity, and the 2017 European Federation of Societies for Ultrasound in Medicine and Biology guidelines recommend the use of US contrast agents to characterize accessory spleens and splenosis (2). However, late phase retention is not unique to the normal splenic parenchyma. Multiple reports (12–14) have shown that splenic hamartomas, which have distinctly different architecture to the normal splenic parenchyma (but also contain RES), retain Sonovue in the late phase. We have also noticed late phase enhancement of splenic hamartoma using Definity, as proven with a sulfur colloid scan. Extramedullary hematopoiesis and myelolipoma, both having variable degrees of RES activity, have been studied with US contrast agents, and some reports demonstrate late phase retention of Sonovue (15,16). In the review of the literature, we did not find reports of late phase retention of US contrast agents in organs not typically associated with the RES, such as the pancreas, kidneys, prostate, uterus, ovaries, and bowel.Hepatic ApplicationsAll hepatic lesions with a documented presence of Kupffer cells have shown late phase binding and/or sequestration of blood pool US contrast agents. The archetypal example of such lesions is focal nodular hyperplasia. The majority of focal nodular hyperplasias have absent washout in the late phase as a characteristic feature. Studies describing focal nodular hyperplasia enhancement features report sustained portal phase enhancement equal to or greater than that of the normal liver to 5 minutes (17). Studies have noted sustained late phase enhancement in focal nodular hyperplasia in 75.2%–90.6% of lesions using Sonovue, 86%–91% of lesions using Definity, and 93.5% of lesions using Sonazoid (17–20)Ple.Although hepatocellular adenomas often have diminished Kupffer cells, a proportion of hepatocellular adenomas demonstrate functioning RES to the same degree as background liver (21). Studies have noted sustained late phase enhancement in adenoma in 33.3%–46% of hepatocellular adenomas using Sonovue and 47%–63% using Definity (17,18,22–25). We found no series describing a hepatocellular adenoma enhancement pattern using Sonazoid.Kupffer cells occur in most HCCs. Their population (as compared with adjacent parenchyma) is maintained in about two-thirds of well-differentiated tumors but drops to 20%–25% of normal livers in moderately to poorly differentiated tumors (26). The differentiation of HCC correlates with the Kupffer cell phase of Sonazoid, with moderately and poorly differentiated tumors demonstrating hypoenhancement while the majority of well-differentiated tumors demonstrate isoenhancement (27,28) in the Kupffer cell phase. Moreover, well-differentiated tumors are more likely to demonstrate washout in the late phase (at 5 and 10 minutes) than in the early phase (28). The differentiation of HCC correlates with the presence and timing of washout for Sonovue and Definity. In a recent series of 276 patients studied with Sonovue, 21.5% of well-differentiated HCCs showed iso-echogenicity in the late phase, compared with 10.2% and 3.4% of moderately and poorly differentiated HCCs, respectively (29). Washout at 1 minute occurred in 36.6% of well-differentiated HCCs and in 61.2% and 74.7% of moderately and poorly differentiated tumors, respectively. The same patterns were first described for Definity by Jang et al (30).Contrast-enhanced US patterns of intrahepatic nonhepatocyte type lesions with RES activity, namely extramedullary hematopoiesis and myelolipoma, are rare. Two reports of extramedullary hematopoiesis showed late phase hyperenhancement with Sonovue (31,32).RES activity can also explain the late phase washout of hemangioma observed with blood pool US contrast agents, as with Sonazoid (Fig 3). In one series using Definity, washout occurred in six of 29 hypervascular (rapidly enhancing) hemangiomas (21%) (33). The washout occurred after 3 minutes in five of these six hemangiomas. A strictly intravascular model of US contrast agents cannot explain how a lesion fed by an artery fills up and subsequently washes out of contrast agent in the late phase while the same does not happen in the adjacent liver parenchyma.Figure 3a: Contrast-enhanced US images in 50-year-old man with MRI-proven hemangioma obtained after injection of 0.2-mL Definity. (a) Image in the early portal venous phase demonstrates typical peripheral nodular enhancement (arrow). (b) Image obtained 2 minutes after contrast agent injection shows that much of the lesion is filled with contrast material (arrow). (c) Image obtained 9 minutes after injection demonstrates that the lesion (which was hyperechoic on gray-scale image [not shown]) shows washout (solid arrow). Note enhancement of the background liver with clearing of contrast agent in the hepatic vein (dashed arrows).Figure 3a:Download as PowerPointOpen in Image Viewer Figure 3b: Contrast-enhanced US images in 50-year-old man with MRI-proven hemangioma obtained after injection of 0.2-mL Definity. (a) Image in the early portal venous phase demonstrates typical peripheral nodular enhancement (arrow). (b) Image obtained 2 minutes after contrast agent injection shows that much of the lesion is filled with contrast material (arrow). (c) Image obtained 9 minutes after injection demonstrates that the lesion (which was hyperechoic on gray-scale image [not shown]) shows washout (solid arrow). Note enhancement of the background liver with clearing of contrast agent in the hepatic vein (dashed arrows).Figure 3b:Download as PowerPointOpen in Image Viewer Figure 3c: Contrast-enhanced US images in 50-year-old man with MRI-proven hemangioma obtained after injection of 0.2-mL Definity. (a) Image in the early portal venous phase demonstrates typical peripheral nodular enhancement (arrow). (b) Image obtained 2 minutes after contrast agent injection shows that much of the lesion is filled with contrast material (arrow). (c) Image obtained 9 minutes after injection demonstrates that the lesion (which was hyperechoic on gray-scale image [not shown]) shows washout (solid arrow). Note enhancement of the background liver with clearing of contrast agent in the hepatic vein (dashed arrows).Figure 3c:Download as PowerPointOpen in Image Viewer Clinical Implication of Acknowledging a RES Phase for All US Contrast Agents in Clinical UseBecause the strength of the RES phase retention and/or sequestration may be different for each agent, it may affect the recognition of washout in the late phase. The absence of washout in the late phase for HCC has been noted in 1.4%–9% of studies using Definity and 2%–14% of studies using Sonazoid (27,28,30,34). However, reported rates of absent washout for Sonovue are more variable and distinctly different, ranging from 12.0% to 55.3% (29,35–37). This greater heterogeneity may be because Sonovue has been the subject of more studies with more patients compared with the other agents. Or it may be due to Sonovue’s shorter half-life. But a lower affinity for the RES could also explain Sonovue’s reported higher rate of absent washout in HCC. In one large series of 2 cm or smaller nodules studied with Sonovue (35), 100% of the 190 arterially enhancing nodules were HCC; only 44.7% showed washout. When the proportion of HCCs showing absent washout is reportedly much higher for Sonovue, should they be placed in the same management category (Liver Imaging Reporting and Data System category 4) as those showing absent washout with Definity? With cirrhosis, the RES shows diminished function in the liver, a well-known phenomenon noted on sulfur colloid scans called a “colloid shift” (38). Historically, most studies of Sonovue originated from Europe, where hepatitis B, the major cause of noncirrhotic HCC, is much less prevalent. It is possible that the rates of absent washout differ based on the RES activity, which is markedly reduced in patients with advanced cirrhosis but not in patients with hepatitis B who do not have cirrhosis. Unless compared with other agents in the same patient (or at least the same patient population) to prove concordant behavior, diagnostic algorithms relying on rates of washout, such as the contrast-enhanced US Liver Imaging Reporting and Data System, may not be uniformly applicable to Sonovue and Definity or to patients with or patients without cirrhosis. In addition, recognizing the RES action allows for the use of US contrast agents in a wider range of applications.In conclusion, the evidence presented earlier supports the contribution of the RES in the late phase of US contrast agents, in particular Sonovue and Definity. Pharmacokinetically, RES action explains the positive and negative enhancement of lesions in the late phase of US contrast agents. In the laboratory, the mentioned agents have shown direct attachment and RES-associated phagocytosis. Moreover, inactivation of the RES leads to reduced late phase enhancement. In both humans and animals, late phase parenchymal uptake of US contrast agents occurs in the liver and spleen. All types of lesions with variable RES activity (whether in the liver or spleen) demonstrate late phase enhancement. However, organs and lesions with no RES activity do not show this late enhancement. Finally, for most benign and malignant hepatic (and nonhepatic) lesions there is concordance of enhancement patterns between Sonovue and/or Definity and Sonazoid, the latter a confirmed RES contrast agent. Recognizing the role of the RES in the late phase for all US contrast agents can help us realize and expand the potential of this important clinical tool.Disclosures of Conflicts of Interest: K.K. disclosed no relevant relationships. M.A. disclosed no relevant relationships. T.K.K. disclosed no relevant relationships. H.J.J. disclosed no relevant relationships.