| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Pediatric Imaging |
1 From the Department of Diagnostic Radiology, Dalhousie University and Department of Diagnostic Imaging, IWK Health Centre, 5850/5980 University Ave, Halifax, NS, Canada B3J 3G9 (M.H.S.); Departments of Pediatrics and Health Policy Management & Evaluation, University of Toronto and Division of Hematology and Oncology, Hospital for Sick Children, Toronto, Ontario, Canada (L.S.); and Department of Diagnostic Imaging, Scarborough Hospital, General Division, Scarborough, Ontario, Canada (B.M.S.). From the 2002 RSNA scientific assembly. Received February 7, 2003; revision requested April 25; revision received June 9; accepted July 23. Address correspondence to M.H.S. (e-mail: matthias.schmidt@iwk.nshealth.ca).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: All available results of imaging studies of children who received a diagnosis of HLH between January 1985 and June 2000 were retrieved. For nine patients, abdominal US images obtained within 1 week of presentation to the hospital were reviewed retrospectively by two radiologists who were aware of the diagnosis and of the original interpretation of each study. US images were examined for evidence of splenomegaly, hepatomegaly, ascites, gallbladder wall thickening, increased periportal echogenicity, lymphadenopathy, pleural effusion, and nephromegaly. Any other abnormalities were also recorded. Differences in interpretation were resolved by consensus. The patients ranged in age from 2 months to 4
years. The male-to-female ratio was 5:4.
RESULTS: Findings at presentation included splenomegaly (in eight of the nine children), hepatomegaly (in seven children), ascites (in six children), gallbladder wall thickening (in six children), increased periportal echogenicity (in three children), lymphadenopathy (in three children), and pleural effusion (in two children). Miscellaneous findings in individual patients included coarse hepatic echotexture with a single 9-mm hypoechoic focus in the liver, multiple hypoechoic foci in the spleen, nephromegaly, gallstone, increased renal cortical echogenicity, and mural thickening of the duodenum.
CONCLUSION: In the appropriate clinical setting, the differential diagnosis of a combination of hepatosplenomegaly, ascites, gallbladder wall thickening, increased periportal echogenicity, lymphadenopathy, and/or pleural effusion should include HLH.
© RSNA, 2004
Index terms: Abdomen, US, 70.1298 • Blood, diseases, 70.659 • Ultrasound (US), in infants and children
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Primary HLH has an autosomal recessive inheritance pattern, with many instances of sporadic occurrence (2). Chemotherapy with steroids and epipodophyllotoxins can prolong survival, but bone marrow transplantation offers the only chance of cure (2). Secondary HLH may occur with malignancy; viral, bacterial, fungal, or parasitic infection; drug exposure; systemic lupus erythematosus; Still disease; Griscelli-Prunieras syndrome; and Chediak-Higashi syndrome (3,4). Survival without bone marrow transplantation is possible, provided that the underlying disorder can be treated successfully (3,4).
The median survival of patients with untreated primary HLH is 2 months; this emphasizes the importance of early diagnosis and treatment (5). Unfortunately, the diagnosis of HLH is often delayed because of its variable clinical manifestation, lack of awareness among clinicians, and lack of a single specific diagnostic marker (1). A single case report of a child who developed HLH after treatment for acute lymphoblastic leukemia described the magnetic resonance (MR) imaging finding of multiple rounded signal voids in the spleen and, to a lesser extent, in the liver that corresponded to hemosiderin deposits found at splenectomy and liver biopsy (6).
Abdominal ultrasonography (US) is more likely to be requested at the time of presentation, when the differential diagnosis is still broad. However, there is little information on abdominal US findings in HLH to date. A report of six children with HLH (two children with primary HLH and four children with secondary HLH) emphasized the US findings of gallbladder wall thickening, increased periportal echogenicity, and enlargement of lymph nodes in the porta hepatis; hepatosplenomegaly and ascites were also described (7). Another report of imaging findings in Epstein-Barr virus infection in Japanese children described two children with Epstein-Barr virus–associated HLH and imaging findings of gallbladder wall thickening, hepatosplenomegaly, ascites, pleural effusion, pulmonary edema, and cardiomegaly (8).
More recently, results of imaging studies of 25 North American children with HLH (including five children with confirmed familial HLH and eight children with definite associated conditions) were reviewed, and the following abdominal US findings were observed: hepatomegaly; splenomegaly; ascites; gallbladder wall thickening; increased periportal echogenicity; nephromegaly; increased renal cortical echogenicity; and mesenteric, periportal, perisplenic, and inguinal lymphadenopathy (9).
The purpose of our study was to evaluate the US findings at presentation in a series of children who received a diagnosis of HLH at a single institution.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
years. The male-to-female ratio was 5:4. All patients were found to have hemophagocytosis (Table 1). Splenomegaly was diagnosed clinically in all patients, and hepatomegaly was diagnosed clinically in all but one patient (patient 2). Lymphadenopathy was present at physical examination in two patients (patients 1 and 7), and one patient (patient 7) had a skin rash. Institutional review board approval was granted for review of patient records and images. Informed consent was not required by the institutional review board.
|
Image Review
Results of abdominal US studies performed within 1 week of presentation were reviewed retrospectively by two radiologists (M.H.S. and B.M.S., with differences resolved by consensus), with particular attention paid to the presence of the following findings: splenomegaly, hepatomegaly, ascites, gallbladder wall thickening, increased periportal echogenicity, lymphadenopathy, pleural effusion, and nephromegaly.
Splenomegaly was judged to be present if the bipolar diameter of the spleen in a coronal image exceeded 5 cm in infants 3 months of age or younger, 6 cm in infants 6 months of age or younger, 7 cm in infants 12 months of age or younger, and 9 cm in children 1–5 years of age (10). Hepatomegaly was judged to be present if the liver edge extended beyond the lower pole of the right kidney (11). Nephromegaly was judged to be present if measurements of the bipolar diameter exceeded the 95th percentile according to age-correlated nomograms (12). Gallbladder wall thickening and increased periportal echogenicity were subjectively assessed. Lymphadenopathy was deemed to be present if lymph nodes in any region had a short-axis diameter greater than 1 cm (13).
Any other abnormalities identified on the US images were recorded. Radiologists were aware of the diagnosis and of the original interpretation of the results of each imaging study. All of the US examinations had been performed because of clinical indications. Clinical findings gleaned from a chart review (L.S.) and imaging findings recorded by the radiologists (M.H.S. and B.M.S.) were compared by one of the authors (M.H.S.).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
A number of findings were observed in individual patients. These included coarse hepatic echotexture with a single 9-mm hypoechoic focus in the liver (patient 1), multiple hypoechoic foci in the spleen (patient 2), gallstone (patient 2), nephromegaly (patient 1), increased renal cortical echogenicity (patient 5), and mural thickening (to 5 mm) of the duodenum (patient 8).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
All of the patients in our study met the diagnostic criteria for primary HLH at the time of diagnosis. One patient (patient 6) had a family history of HLH. This can be considered strong evidence of primary HLH. Another patient (patient 5) had Epstein-Barr virus infection at the time of the diagnosis of HLH. However, this does not rule out primary HLH in this patient, since an infection such as Epstein-Barr virus infection can trigger primary HLH in a susceptible individual (1). One patient (patient 2) was readmitted within 1 year of treatment for HLH and found to have anaplastic large cell lymphoma. This patient was subsequently categorized as having malignancy-associated HLH.
The understanding of the pathophysiologic mechanisms of HLH is still incomplete, but many changes occurring at the cellular and molecular levels have been revealed (2). There is widespread dysregulation of the immune system, with activation of macrophages, defective T-cell function, low or absent natural killer cell activity, and elevation of circulating cytokine levels (14,15).
Histopathologic findings at autopsy have been described in a series of 27 children with primary HLH (16). In the spleen, the red pulp is expanded and infiltrated with histiocytes; the white pulp is diminished and depleted of lymphocytes (16). In the liver, lymphocytes and histiocytes infiltrate the portal tracts and, to a much lesser extent, the hepatic lobules (16). In lymph nodes, histiocytes accumulate in T-cell areas and sinuses; the follicles become sparse, and lymph nodes are depleted of lymphocytes in some patients (16). These findings can help explain the abnormalities seen at US in the spleen, liver, portal tracts, and lymph nodes.
Chateil et al (7), while recognizing that these findings are nonspecific, have speculated that the increased periportal echogenicity and gallbladder wall thickening seen in many patients with HLH may be related to portal lymphohistiocytic infiltration and subsequent lymphatic and/or portal venous stasis. The description by Öst et al (16) of periportal lymphohistiocytic infiltration observed at autopsy in children with primary HLH supports this hypothesis. However, it should be noted that Fitzgerald and McClain (9) found no histopathologic correlate at autopsy for the increased periportal echogenicity seen at US in nine of the patients in their study.
To our knowledge, histopathologic findings in the gallbladder in patients with HLH have not been described. It is reasonable to suppose that lymphohistiocytic infiltration of portal tracts may extend to the gallbladder in some patients. However, in our study, gallbladder wall thickening and increased periportal echogenicity were not always both present. Gallbladder wall thickening may be reactive in the presence of hepatitis, and it may represent edema in the presence of ascites. In this context, it is interesting to note that gallbladder wall thickening occurred without ascites, but with a coarse hepatic echotexture, in one of our patients (patient 1).
It is difficult to know the importance of most of the findings that occurred in individual patients in our study. There was no biopsy of the small solitary hypoechoic liver lesion in patient 1. Multiple hypoechoic foci in the spleen of patient 2 could have been related to HLH or to anaplastic large cell lymphoma, which was diagnosed after treatment for HLH was completed. The nephromegaly observed in patient 1 at the time of presentation coincided with the presence of histiocytes in the urine, and hemophagocytosis was observed at renal biopsy performed after the patient received a 6-day course of cyclosporin. The increased renal cortical echogenicity observed in patient 5 at the time of presentation was not investigated further and therefore remains of uncertain importance.
Fitzgerald and McClain noted the findings of nephromegaly and increased renal cortical echogenicity relatively more frequently than we did, but they did not state explicitly how often these findings were present in treated versus untreated patients (9). In our experience, the agents used for chemotherapy of HLH are nephrotoxic and can produce nephromegaly and increased renal cortical echogenicity. The findings of a gallstone in patient 2 and mural thickening of the duodenum in patient 8 remain of uncertain importance.
Hepatosplenomegaly, ascites, and gallbladder wall thickening were the most common findings at presentation in our patients with HLH. Increased periportal echogenicity and lymphadenopathy, which were among the most frequent findings in the series of Chateil et al (7), were less common among our patients. The relative frequency of findings in our study more closely resembles that reported by Fitzgerald and McClain (9). Lymphadenopathy, which Chateil et al (7) observed to occur primarily in the porta hepatis, was more variable in distribution in our study and in the series reported by Fitzgerald and McClain (9).
In the present study, the agreement between US findings and physical examination findings of hepatosplenomegaly was generally good. However, discrepancies were noted in some patients. This suggests that US and physical examination are complementary in their ability to reveal hepatosplenomegaly at presentation in patients with HLH. US can additionally help identify gallbladder wall thickening, increased periportal echogenicity, and intraabdominal lymphadenopathy.
Our study had several limitations. The number of patients was small. The study was retrospective. The authors were not blinded with respect to the clinical diagnosis or to the original interpretation of the US images. Lack of blinding introduces potential bias into the subjective assessment of findings such as gallbladder wall thickness and periportal echogenicity. To assess spleen size, we used standards that are in common clinical use at our institution (10). To assess liver size, we determined the relative position of the liver edge and the right renal lower pole on US images; this is the method currently used at our institution (11). The use of age-correlated nomograms for each of these organs would have permitted more precise quantitation.
Knowledge of the US features of HLH at the time of presentation is important because abdominal US is likely to be performed early in the diagnostic work-up of children with clinical findings caused by HLH. None of the US findings described by Chateil et al (7), Moritani et al (8), Fitzgerald and McClain (9), or us is specific. These findings can be produced by hepatitis (17). Very similar findings were recently reported in a series of Korean children with infectious mononucleosis who apparently did not have HLH (18). The substantial overlap in the appearances of primary HLH, Epstein-Barr virus– associated HLH, and infectious mononucleosis raises the possibility of shared pathophysiologic mechanisms. Other diagnoses that should be considered include Henoch-Schönlein purpura (19), leukemia, and Langerhans cell histiocytosis.
MR imaging may prove helpful in narrowing the differential diagnosis and should be used when HLH is considered in the differential diagnosis (6). However, the findings reported by Zilkha et al (6) in a single patient with HLH who also had leukemia and infection at the time of presentation will need to be replicated before the usefulness of MR imaging can be known.
Although the individual findings are nonspecific, the combination of US findings of hepatosplenomegaly, ascites, gallbladder wall thickening, increased periportal echogenicity, lymphadenopathy, and/or pleural effusion should raise the suspicion of HLH in the appropriate clinical setting.
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, M.H.S.; study concepts and design, M.H.S., L.S., B.M.S.; literature research, M.H.S., L.S.; clinical studies, L.S.; data acquisition, M.H.S., L.S.; data analysis/interpretation, M.H.S., L.S., B.M.S.; manuscript preparation, M.H.S.; manuscript definition of intellectual content, editing, revision/review, and final version approval, M.H.S., L.S., B.M.S.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
