Hepatomegaly in a patient with a history of acute myeloid leukemia

In February 2018, a 30-year-old man was referred to the outpatient clinic for recent hepatomegaly. He had a history of acute myeloid leukemia in 2009, treated with cytarabine and doxorubicin, followed a few months later by allogeneic hematopoietic stem cell transplantation. The conditioning regimen included cyclophosphamide and busulfan. The post-transplant course was uneventful. In 2010, the patient developed graft-versus-host disease of the skin and liver, which was rapidly responsive to corticosteroids and cyclosporine. Immunosuppressive agents were withdrawn in 2011. No clinical event occurred between 2011 and 2018 and the patient was not taking any medication. Physical examination revealed hepatomegaly and dilated abdominal subcutaneous veins. Blood test results were as follows: platelets 134 x 109/L, gamma-glutamyltransferase 236 IU/L (normal <60 IU/L), normal serum transaminases, alkaline phosphatase, and bilirubin levels, and prothrombin index 53 %. Abdominal ultrasonography showed an enlarged liver with a smooth liver surface, increased spleen size (13 cm in the largest axis), and mild ascites. Liver stiffness measurement using vibration-controlled transient elastography was 65 kPa. Contrast-enhanced computed tomography (Fig. 1A and B) and transjugular liver biopsy (Fig. 1C) were performed. 1.Graft-versus-host disease2.Sinusoidal obstruction syndrome3.Constrictive pericarditis4.Porto-sinusoidal vascular disorder Contrast-enhanced computed tomography showed dilated hepatic veins with normal enhancement at the venous phase, associated with a "mosaic" enhancement pattern at the arterial phase, prominent in the subcapsular area of the liver (Fig. 2). The latter is highly suggestive of hepatic sinusoidal dilatation, which was confirmed by liver biopsy examination. Liver biopsy examination also disclosed perisinusoidal fibrosis and obliteration of the lumen of the centrilobular veins (Fig. 3). Altogether, these findings suggested a heart-related hepatic venous outflow obstruction. Obliteration of the lumen of the centrilobular veins is a well-known finding in this setting and does not systematically imply a sinusoidal obstruction syndrome. An echocardiography did not find any valvulopathy, dilatation of heart cavities, nor pericardial effusion. Right heart catheterization was performed, showing a "Dip & plateau pattern" (also named "square root sign") (Fig. 4). Last, a cardiac magnetic resonance imaging demonstrated pericardial thickening. The diagnosis of constrictive pericarditis was then made. In this patient, pericarditis was regarded as secondary either to a chronic graft-versus-host disease (as described in a few case reports) or idiopathic. The patient underwent subtotal pericardiectomy. Pathological examination of the pericardium showed typical lesions of chronic pericarditis, namely abundant fibrosis including inflammatory infiltration with lymphocytes and plasma cells. Serum gamma-glutamyltransferase normalized and hepatomegaly disappeared within a few weeks. At last visit, 3 years later, the patient was doing well with no symptoms.Fig. 3Liver biopsy findings.Show full caption(A, B) Sirius red staining showed mild periportal fibrosis (black arrows) and extensive perisinusoidal fibrosis in centrilobular areas. The lumen of some centrilobular veins are not visible anymore (black asterisks). (C,D) Hematein-eosin/saffron staining showed fibrous centrilobular veins with obliterative features (black asterisks) without any sign of inflammation, and with centri- and medio-lobular sinusoidal dilatation (white arrows).View Large Image Figure ViewerDownload Hi-res image Download (PPT)Fig. 4Right heart catheterization findings.Show full caption(A) Normal right ventricular pressure. (B) "Dip & plateau pattern" (also named "square root sign"), corresponding to the equalization of end-diastolic pressures, observed in this patient (red arrow).View Large Image Figure ViewerDownload Hi-res image Download (PPT) (A, B) Sirius red staining showed mild periportal fibrosis (black arrows) and extensive perisinusoidal fibrosis in centrilobular areas. The lumen of some centrilobular veins are not visible anymore (black asterisks). (C,D) Hematein-eosin/saffron staining showed fibrous centrilobular veins with obliterative features (black asterisks) without any sign of inflammation, and with centri- and medio-lobular sinusoidal dilatation (white arrows). (A) Normal right ventricular pressure. (B) "Dip & plateau pattern" (also named "square root sign"), corresponding to the equalization of end-diastolic pressures, observed in this patient (red arrow). Sinusoidal dilatation is defined as a sinusoidal lumen that is more than one liver cell plate wide, observed in several lobules.1Marzano C. Cazals-Hatem D. Rautou P.E. et al.The significance of nonobstructive sinusoidal dilatation of the liver: impaired portal perfusion or inflammatory reaction syndrome.Hepatol Baltim Md. Sept 2015; 62: 956-963Crossref PubMed Scopus (49) Google Scholar Sinusoidal dilatation is mainly due to a hepatic venous outflow obstruction, caused by Budd-Chiari syndrome, sinusoidal obstruction syndrome, heart failure, or pericarditis. Sinusoidal dilatation can also be observed without any obstruction, then named "nonobstructive sinusoidal dilatation"; particularly in an inflammatory setting (e.g. pyelonephritis, systemic diseases).1Marzano C. Cazals-Hatem D. Rautou P.E. et al.The significance of nonobstructive sinusoidal dilatation of the liver: impaired portal perfusion or inflammatory reaction syndrome.Hepatol Baltim Md. Sept 2015; 62: 956-963Crossref PubMed Scopus (49) Google Scholar On contrast-enhanced cross-sectional imaging, sinusoidal dilatation typically appears as a heterogeneous and reticular enhancement of the liver, also called "mosaic" enhancement pattern, at the late hepatic arterial phase.2Brancatelli G. Furlan A. Calandra A. et al.Hepatic sinusoidal dilatation.Abdom Radiol N Y Août. 2018; 43: 2011-2022Crossref Scopus (35) Google Scholar This aspect is typically more visible in the peripheral liver area. At histology, sinusoidal dilatation is mainly located in the centrilobular area, even in the absence of an outflow block.1Marzano C. Cazals-Hatem D. Rautou P.E. et al.The significance of nonobstructive sinusoidal dilatation of the liver: impaired portal perfusion or inflammatory reaction syndrome.Hepatol Baltim Md. Sept 2015; 62: 956-963Crossref PubMed Scopus (49) Google Scholar It can be associated with sinusoidal congestion (lumen filled with erythrocytes), perisinusoidal fibrosis, atrophy and/or regenerative changes of hepatocytes, and cardiac cirrhosis in case of prolonged hepatic venous outflow obstruction. Additional obliteration of hepatic and portal vein branches may also develop. When sinusoidal dilatation is diagnosed, Budd-Chiari syndrome should be systematically searched for. When absent, right heart causes should be investigated, including constrictive pericarditis. Constrictive pericarditis is due to inflammation and fibrosis of the pericardium.3Syed F.F. Schaff H.V. Oh J.K. Constrictive pericarditis--a curable diastolic heart failure.Nat Rev Cardiol Sept. 2014; 11: 530-544Crossref PubMed Scopus (99) Google Scholar The subsequent impaired diastolic cardiac filling can lead to heart failure. Tuberculous pericarditis is the most common cause in developing countries, while idiopathic and the cardiac surgery-induced forms are more prevalent in developed countries.3Syed F.F. Schaff H.V. Oh J.K. Constrictive pericarditis--a curable diastolic heart failure.Nat Rev Cardiol Sept. 2014; 11: 530-544Crossref PubMed Scopus (99) Google Scholar The diagnosis is sometimes challenging since dyspnoea can be absent, as well as a pericardial effusion. Hepatomegaly is present in about 50 % of patients at diagnosis, and ascites in one third of them. When echocardiography and heart magnetic resonance imaging remain inconclusive, right heart catheterization should be performed to look for an elevation of central venous and intracardiac pressures, a right ventricular pressure tracing showing a "Dip & plateau pattern" (also named "square root sign") – reflecting the sudden cessation, due to the pericardial constriction, of the early diastolic filling of the right ventricle – and an end-diastolic pressure ≥1/3 of the systolic pressure in the right ventricle. Except for rare cases of transient constrictive pericarditis, a surgical pericardiectomy is needed and is associated with very good outcomes.3Syed F.F. Schaff H.V. Oh J.K. Constrictive pericarditis--a curable diastolic heart failure.Nat Rev Cardiol Sept. 2014; 11: 530-544Crossref PubMed Scopus (99) Google Scholar Several differential diagnoses could have been discussed in our patient. First, histological lesions could be consistent with sequelae of sinusoidal obstruction syndrome. Yet, no clinical or laboratory feature suggesting sinusoidal obstruction syndrome, namely jaundice, painful hepatomegaly, weight gain or ascites, occurred in the weeks following allogeneic hematopoietic stem cell transplantation in this patient. Second, portosinusoidal vascular disorder could also be considered based on the observation of portal vein branches at histology. Yet, the main histological lesions were in the centribolular area and not in the portal tracts. Third, graft-versus-host disease of the liver could be considered in the context of allogeneic hematopoietic stem cell transplantation. However, liver graft-versus-host disease usually occurs earlier after transplantation and is commonly associated with skin graft-versus-host disease. Importantly, histological lesions of liver graft-versus-host disease are quite different from those observed here, involving mainly bile ducts. The authors received no financial support for this article. The authors declare that they have no competing interest. Please refer to the accompanying ICMJE disclosure forms for further details. LM and PER conceived and wrote the manuscript; VP provided histological images and their description; OB provided CT imaging and their description; DV reviewed the manuscript. We would like to express our gratitude to Prof. Alain Cohen-Solal for his sound advice on heart related liver diseases and to Professor Gérard Socié for sharing his expertise on allogeneic hematopoietic stem cell transplantation. The following are the supplementary data to this article: Download .pdf (.34 MB) Help with pdf files Multimedia component 1