| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Medical Imaging, Hôpital Sainte-Justine and University of
Montreal, 3175 Côte Ste-Catherine, Montreal, Quebec, H3T 1C5
Canada.
2 Department of Hematology, Hôpital Sainte-Justine and University of
Montreal, Montreal, H3T IC5 Canada.
3 Department of Dermatology, Hôpital Sainte-Justine and University of
Montreal, Montreal, H3T IC5 Canada.
Received August 16, 2001;
accepted after revision December 6, 2001.
Address correspondence to J. Dubois.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Our study included 16 infants with vascular tumors, other than hemangiomas, who were to undergo biopsy. We used gray-scale sonography to identify calcifications, to evaluate the borders of the lesions to determine whether they were poorly defined or well defined, and to determine the echogenicity relative to the surrounding soft tissue. Doppler sonography served to determine the number of vessels per square centimeter and the peak arterial Doppler shift. Sonographic findings were compared with the final diagnoses established by biopsy.
RESULTS. The final diagnoses included five hemangioendotheliomas, six tufted angiomas, and five infantile myofibromatoses. Hemangioendotheliomas and tufted angiomas were ill defined compared with infantile myofibromatoses that were well defined. Only one vascular tumor, a hemangioendothelioma, fulfilled the diagnostic criteria of hemangioma. Tufted angiomas and infantile myofibromatoses were the least vascularized, with the lowest vessel density (zero to two vessels per square centimeter) and a relatively low systolic Doppler shift (0.7-1.0 kHz).
CONCLUSION. The vascular tumors—hemangioendotheliomas, tufted angiomas, and infantile myofibromatoses—were distinguishable from hemangiomas on Doppler sonography in all cases except one hemangioendothelioma. Unlike hemangiomas, these lesions should be investigated by biopsy or excision.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The sonographic examination was performed using a constant technique by a pediatric radiologist who was experienced in Doppler sonography. A 5000 scanner (ATL, Seattle, WA) with a linear 8- to 12-MHz transducer was used. Color Doppler sonograms were obtained by low-pulse frequency and a wall filter. Pulse-repetition frequency was increased only if aliasing occurred. Gray-scale sonography outlined the size, the contours, the echogenicity of the lesion and revealed the presence of visible vessels in or around the lesion. Color Doppler sonography estimated vessel density [1], counting the number of color Doppler signals in an area of 1 cm2 of the most highly vascularized part of the lesion. Pulsed Doppler sonography was also performed with maximal systolic Doppler shift being measured in kilohertz and the resistance index established.
These examinations were performed in patients who were not sedated. At the end of the study, all sonograms were reviewed by the principal investigator and compared with the final diagnosis established by biopsy or excision.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Vessel Density
The vessel density was quantified between two and four vessels per square
centimeter in three cases of hemangioendothelioma, superior to five vessels
per square centimeter in one case, and inferior to one vessel per square
centimeter in one case. Such a low vessel density was observed in the cases of
infantile myofibromatosis and of tufted angioma.
Maximum Systolic Shift
The systolic shift was higher in the five cases of hemangioendothelioma,
with measurements of 6.5 kHz, 3.9 kHz, 2 kHz, 2 kHz, and 0.6 kHz,
respectively. In the cases of infantile myofibromatosis, the maximal systolic
shift ranged from 0.8 to 1 kHz, whereas the tufted angiomas also displayed a
shift less than or equal to 0.9 kHz.
The size of the lesions varied greatly. The hemangioendotheliomas tended to be larger than the tufted angiomas and the infantile myofibromatoses. The internal architecture on gray-scale sonography was variable, except for the tufted angiomas, which tended to be hyperechogenic (Figs. 1,2,3).
|
|
|
Calcifications were reported in four of five hemangioendotheliomas but not in the cases of tufted angiomas or infantile myofibromatoses. No vessels were visible on gray-scale sonography in or around any of the lesions described.
The borders of the hemangioendotheliomas and tufted angiomas were ill defined. These rather infiltrating borders made it difficult to distinguish these tumors from normal surrounding tissues, compared with the well-defined lesions of infantile myofibromatoses.
The hemangioendotheliomas were the most vascular of all the tumors. However, only one of them fulfilled the diagnostic Doppler criteria of hemangioma, with a high vessel density—more than five signals per square centimeter—and a maximum systolic Doppler shift of 2 kHz (Fig. 4). This tumor was diagnosed as a hemangioma on sonography but was distinguished clinically from a hemangioendothelioma by the presence of platelet trapping (Kasabach-Merritt phenomenon). Another hemangioendothelioma was poorly vascularized with a low vessel density—fewer than one vessel per square centimeter— and a moderate systolic shift (0.6 kHz) (Fig. 5).
|
|
Tufted angiomas and infantile myofibromatoses were the least vascularized of the three tumor types, with the lowest vessel density (zero to two vessels per square centimeter) and a relatively low systolic Doppler shift (0.7-1 kHz) (Figs. 6 and 7). The gray-scale architecture of the two lesion types was different: infantile myofibromatoses were isoechoic or hypoechoic, whereas tufted angiomas were hyperechoic. Despite the difference on gray-scale architecture, the two lesion types were not reliably distinguishable on sonography. The diagnosis was made by excision biopsy in all cases.
|
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Our previous study [1] showed that high vessel density, more than five vessels per square centimeter, and a maximal systolic Doppler shift exceeding 2 kHz permitted hemangiomas to be distinguished from other vascular lesions (Fig. 8). Here we report the results of Doppler sonography for the following three types of vascular tumors of the skin that occur in infancy: hemangioendothelioma, tufted angioma, and infantile myofibromatosis. Hemangiomas were not included in our study.
|
Kaposiform hemangioendotheliomas are uncommon. These tumors appear as tender red plaques, nodules, grouped papules, or telangiectasia. They generally occur on the trunk, the proximal portion of the limbs, and the retroperitoneum. All five cases of hemangioendotheliomas in our series were present in early infancy, similar to hemangiomas, in contrast to their later appearance reported in the literature [2]. Histologically, hemangioendotheliomas consist of dense infiltration of spindle-shaped cells along irregular lobules and sheets of cells and vessels, which form a lacy network infiltrating the dermis and subcutaneous fat, accompanied by dilated, hyperplastic, and lymphatic channels [3,4,5,6]. The lesions tend to regress spontaneously. In their active phase, hemangioendotheliomas are associated with the Kasabach-Merritt phenomenon, which consists of thrombocytopenia, microangiopathic hemolytic anemia and localized consumption coagulopathy [7], lymphangiomatosis, [4] and, occasionally, disseminated intravascular coagulation.
Tufted angiomas are rare, benign angiomatous proliferations that occur in both sexes during early childhood and are sometimes present at birth. They appear as dull red to red-brown or violaceous macules or plaques. Tufted angiomas grow slowly and sometimes cover large areas of the trunk and neck [8, 9] and do not usually regress spontaneously. They usually grow over several years. Spontaneous regression has rarely been reported [10, 11]. Histologically, tufted angiomas are composed of small capillary nodules with a cannonball distribution of numerous round lobules of vascular proliferation in the entire dermis and subcutaneous tissue, as reported in the literature [6]. Lakelike lymphatic vessels are also seen. Tufted angiomas may also present with the Kasabach-Merrit phenomenon.
Infantile myofibromatosis is a fibrous tumor localized in the extremities of newborns and young children. Infantile myofibromatosis mimics congenital hemangioma when it is red; and it follows a benign course and is easily removed surgically. Spontaneous involution has been described in the literature [6]. Histologically, the cells are immature with focal necrosis, vascular invasion, and spindle-shaped myofibroblasts [6].
Differential diagnoses of a red lesion in an infant include other tumors that may be clinically indistinguishable from hemangiomas, such as hemangioendotheliomas (kaposiform), tufted angiomas, infantile myofibromatoses, sarcomas, and metastatic neuroblastomas.
Gray-scale and color Doppler sonography have proved useful in outlining the size of soft-tissue lesions; showing their subcutaneous portions, which may not be evident at clinical examination; in distinguishing hemangiomas from other vascular masses [1, 12].
In a previous study [1] of 136 children, we established the concept of vessel density estimation on Doppler sonography by counting the number of color Doppler signals per square centimeter in the most vascularized area of the lesion, using constant and strictly-controlled technical factors (8- to 12-MHz transducer), low-repetition frequency, low-wall filter, and high Doppler gain. Pulsed Doppler sonography served both to distinguish color Doppler artifacts from the blood-flow signal and to establish the maximal systolic Doppler shift. We chose to express this Doppler shift in kilohertz rather than in centimeters per second as a velocity measurement because it is often difficult to determine the angle between the Doppler beam and small, tortuous vessels in a highly vascularized tumor. An attempt was made to use an angle of less than 60° between the sampled vessel and the Doppler beam. When estimating vessel density in vivo, we found a real difficulty in distinguishing a single tortuous vessel, which gives rise to several Doppler signals, from many distinct vessels. The same problem arises when measuring vessel density at microscopy. However, despite these limitations, we found that hemangiomas were present sonographically with a high vessel density estimate and a Doppler shift exceeding 2 kHz. When these two criteria are met, the diagnosis of hemangioma is made reliably [1].
We used the same criteria to describe three other types of soft-tissue masses occurring in infants. Hemangioendotheliomas are ill defined and contain foci of calcium, rarely seen in hemangiomas, tufted angiomas, or fibromatoses. Three of the five patients had a moderate vessel density estimate, and the Doppler shift exceeded 2 kHz in four of the five patients. One lesion was highly vascularized and was falsely diagnosed sonographically as a hemangioma. However, the presence of platelet trapping in this patient distinguished the lesion from a hemangioma.
Tufted angiomas are superficial, with a maximal depth of 1 cm surrounded by subcutaneous fat that is more echogenic than usual, perhaps because of the presence of lymphatics with the wall interface.
Although infantile myofibromatoses were the only well-defined lesions in the series and tufted angiomas the only hyperechoic ones, the diagnosis was made only after excision. Both infantile myofibromatoses and tufted angiomas were poorly vascularized.
This brief description of the three types of soft-tissue masses in infancy underlines the features distinguishing them from the usual hemangiomas. Our criteria of high vessel density and maximal Doppler shift exceeding 2 kHz proved reliable in differentiating these lesions from hemangiomas, with the exception of a case of kaposiform hemangioendothelioma associated with the Kasabach-Merritt phenomenon.
In terms of rare malignant soft-tissue tumors, we have encountered two cases of rhabdomyosarcoma, two cases of metastatic neuroblastoma, one case of T-cell lymphoma, and one case of dermatofibrosarcoma in the last 10 years. None of these fulfilled the previously mentioned Doppler criteria, and a biopsy was therefore performed immediately.
In conclusion, the three types of soft-tissue lesions were distinguishable from hemangiomas by Doppler sonography except for one case of hemangioendothelioma. Once this distinction has been established, patients with these atypical soft-tissue lesions should be examined further by biopsy or excision, contrary to hemangiomas, which are treated conservatively in most patients.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  J.M. Chang, B.J. Kwon, M.H. Han, H.S. Kang, and K.H. Chang Kaposiform hemangioendothelioma arising from the internal auditory canal. AJNR Am. J. Neuroradiol., April 1, 2006; 27(4): 931 - 933. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S Ostlere and R Graham Imaging of soft tissue masses Imaging, December 1, 2005; 17(3): 268 - 284. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
