The Regression of Hemangioma of Infancy is Regulated by Sympathetic Nerve
Date Issued
2006
Date
2006
Author(s)
Dai, Yang-Shia
DOI
zh-TW
Abstract
【Background and Purpose】 Hemangioma of infancy ( HOI）is the most common tumor in infancy, with an estimated prevalence of 1-2.6％ in neonate [Hlmdahl K et al., 1955]. It is characteristic by its typical involution of nascent, proliferating, involuting, and involuted stages. About one third of the patients are born with visible lesions, and the others are invisible until several weeks old. HOI is not a hereditary disease, and its cause is unknown. However, some risk factors have been reported, such as Caucasian, female, prematurity with birth body weight less than 1500 g, and a history of chorionic villi sampling. [Bowers RE et al., 1960; Droler BA et al., 1999 ; Mulliken JB et al., 1988].
HOI is one of capillary hemangiomas, usually located in superficial or deep dermis. It often appears on the head and neck area, and sometimes may invade the viscera。Usually, it is proliferating during the first three to nine months of age, and then becomes involuting between the age of nine and twelve months. Finally, HOI entered the involuted stage at the age of two to five years old. Approximately, 10% of the HOI patients completely resolved per year. For differential diagnosis, non-involuting congenital hemangioma (NICH) and rapidly-involuting congenital hemangioma (RICH) should be excluded by lacking of typical evolution of HOI: NICH usually does not involute and grow proportionally with increasing age [Enjolras O,et al., 2001]; RICH is often completely involuted in one year [Lopez-Gutierrez et al., 2005].
According to clinical experience, about 60% of HOI will finally resolve in a cosmetically acceptable scar tissue. Thus, the policy of treating HOI has long been “wait and see” or so-called “benign neglect”. For destructive HOI or severe complications, early intervention is necessary. Damage to the face, compression to important organs (eyes, noses, throat), and ulceration are the major indications for early treatment. The well-established therapies are corticosteroid ( topical, oral, and injection), topical imiquimod (Aldara®), pulsed dye laser, chemotherapy (cyclophosphamide or vincristine), interferon α-2a or 2b, and surgery [Garzon MC et al., 2005; Zvulunov A et al., 2002; Sidbury R et al.,2003; Welsh O et al., 2004; Hazen PG., et al. 2005; Barrio VR et al., 2005; Turner et al., 1994; Boehm DK et al., 1995; Perez PJ et al., 1995; Garmendia G et al., 2001; Rampini E et al., 2000; Tamayo L et al., 1997; Illum N et al., 1995; Chan YC et al., 2005]。Although corticosteroid is the first choice of treatment for proliferating HOI, the response rate is about 30%. Imiquimod cream enhance the local production of interferon, has been proved to be effective in recent 3 years. The second-line treatment for proliferating HOI is chemotherapy or interferon injection, however, their side effects are immunosuppression and neurological deficit, respectively [Dubois J et al., 2002; Barlow CF et al., 1998]. For laser therapy, pulsed dye laser (595nm) is most effective. But, the depth of treatment is 1-2 mm only. Early laser treatment may leave more scars or pigment change [Batta K et al., 2002]. Thus, dye laser therapy is usually recommended for children older than one year.
If we can find out the reason for mysterious involution of HOI, it would contribute a lot to novel therapy for angiogenesis inhibition. So far, there have been many studies on various aspects of HOI. From the view point of pathophysiology, HOI is capillary hemangioma. North first reported that HOI, like brain and placenta, has glucose transporter I (GLUT-1) located on the endothelial membrane of capillaries [North PE et al., 2000]. Other common benign cutaneous vascular tumor does not have GLUT-1, so GLUT-1 is a specific marker for differential diagnosis. More importantly, NICH and RICH also do not contain GLUT-1 [Leon-villapalos J et al., 2005].
Since capillaries of placenta contain GLUT-1, the association between placenta and HOI was highly suspected. North found another three placental capillaries’ antigens were also expressed in HOI- FcγRII, merosin, and Lewis Y antigen (Le Y). Thus, they proposed two hypothesis on “placenta as an origin of HOI”-one was “aberrant angioblast entered the skin vessels, another was “placental thrombotic cells developed in the skin” [North PE et al., 2001]. Barnes reported that the transcriptomes of the endothelial cells of HOI and placenta have high degree of similarity [Barnes et al., 2005]. Therefore, the close resemblance of HOI and placental vessels may explain the regression of HOI.
Besides, some studies showed that the endothelial cells of HOI had characteristics similar to fetal capillaries. It implies that HOI may be a capillary tumor with the nature of fetal vessel or a product of early interruption of vessel development. Dosanjhr reported noted the level of platelet-endothelial cell adhesion molecule (PECAM-1) and von Willebrand factor (vWf) in HOI of the newborn was far less than that of normal newborn, but was similar to that of normal fetus. He also found that type I collagen was rich in newborn’s HOI and normal fetus, while type IV collagen was rich in normal newborn. [Dosanjh A et al., 2000]. Dadras found the endothelium of proliferating HOI expressed CD 34, CD31, and lymphatic endothelial hyaluronan receptor-1 (LYVE-1), while involuting HOI did not. They concluded that proliferating HOI was consisted of immature vessels, because it expressed both vascular and lymphatic markers. [Dadras SS et al., 2004].
The role of growth factors and angiogenesis stimulating factors in the pathogenesis for HOI has been noted since 1994. Proliferating HOI secrete high level of basic fibroblast growth factor (bFGF) and vascular endothelial growth factor (VEGF). At involuting stage, VEGF is normal, but bFGF is still increased at early involuting stage [Takahashi K et al., 1994]. Level of urinary bFGF was high in infants with HOI. However, it was normal in patients with arteriovenous malformation [Dosquet C et al., 1998]. Besides, some evidences showed that abnormal tissues around HOI produced angiogenesis stimulating factors-Berard found the stromal cells of HOI could secret VEGF [Berard M et al., 1997]; Bielenberg reported that proliferating HOI had hyperplastic epidermis containing abundant bFGF and VEGF [Bielenberg et al., 1999]. Other factors such as tissue inhibitor of metalloproteinase 1(TIMP-1), insulin-like growth factor 2, hypoxia inducible factor-1 alpha (HIF-1α) had been reported to be related to different stages of HOI [Takahashi K et al., 1994; Ritter MR et al., 2002; Chen D et al., 2005].
Although angiogenesis stimulating factors may explain endothelial hyperplasia of proliferating HOI, there are some further evidences about intrinsic abnormalities of HOI, for example, monoclonality of endothelial cells [Boye E et al., 2001] and loss of heterozygosity on chromosome 5q [Berg JN et al., 2001]. Boye found that endothelial cells of HOI had abnormal response to endostatin [Boye E et al., 2001]. Walter noted that HOI had mutant VEGF receptor [Walter JW et al., 2002]. Yu reported endothelial cells of HOI had increased expression of Tie2, dysregulation of angiopoietin-2, and hyperreactivity to angiopoietin-1 [Yu Y et al., 2001]. Jiang found that endothelium of various stages of HOI had different density of glucose receptor isotypes [Jiang XW et al., 2001]. Yu also found that proliferative HOI contained endothelial progenitor cells (EPCs) [Yu Y et al., 2004]. Ritter first reported the relationship between proliferative HOI and immune cells- proliferative HOI had abundant CD8+ cytotoxic T cells. Indoleamine 2, 3- dioxygenase, an enzyme associated with regulation of T cell function, is more abundant in proliferative HOI than involuting HOI. It may suggest that the regression of HOI is associated with the inhibition of T cell function [Ritter et al., 2003 et al., 1986].
A new approach to study HOI is via the viewpoint of anatomy. Waner reported “the distribution of facial hemangiomas is nonrandom”- the focal hemangiomas on the face were all located near lines of mesenchymal or mesenchymal-ectodermal embryonic fusion, while the diffuse hemangiomas showed a segmental distribution [Waner M et al., 2003]. Thus, it may imply that HOI was influenced by local abnormal nerve tissues. Moreover, HOI may contain abnormal nerve tissues. Recently we demonstrated that regulation of blood flow in proliferating HOI is impaired. Laser Doppler fluxmetry with heating provocation could hardly induce vasodilatation in proliferating HOI of 18 infants, aged 0-8 months old. It may be caused by abnormal function of local sympathetic nerves.
For further investigation about the relationship between HOI and nerve tissue, we hypothesize that nerve tissue may regulate the involution of HOI. To compare the change of nerve density at different stages of HOI, we will measure the density of nerve and apoptotic cells of HOI at various stages. Thus, it may provide primary evidence about the role of nerve tissue in HOI.
【Materials and Methods】
specimen collection
From 3 medical centers in Northern Taiwan, paraffin-embedded specimens of hemangiomas were obtained. The age at excision was 0-24 months old.
histopathology examination
5 um paraffin sections were deparaffinized in xylene for 30 minutes. After hydration with 100%, 95%, 85% alcohol, and pure water for several minutes, they were stained with Hematoxylin & Eosin (H & E). Under light microscopy, they were classified as proliferating, early involuting, and late involuting hemangiomas by morphological criteria: (1) endothelial hyperplasia, plump and multi-layered; with narrow lumens (2) single-layered endothelium, flat; with obvious, not tortuous lumens (3) single-layered endothelium, flat; with dilated and tortuous. Pyogenic granulomas were excluded.
immunohistochemistry
GLUT-1 immunohistochemical staining was done first to confirm the diagnosis of HOI. And then PGP9.5 immunohistochemical staining was performed on all HOI specimens to show nerve density. Tyrosine hydroxylase and activated caspase 3 immunohistochemistry staining were also done for further demonstration of sympathetic nerve and apoptotic cells in HOI.
image processing and data analysis
All results were photographed by digital camera under light microscope at magnification of 200 fold. The conditions for taking photographs were: (1) avoided the neighboring areas of venules, arterioles, and eccrine glands (2) focus on the area of highest nerve/apoptotic cell density, and then took four successive pictures. Photographs were then analyzed with Image Pro Plus®. The density of nerve was defined as “summation of individual areas of nerves divided by the area of the picture (unit: μm2/μm2). The density of apoptotic cells was defined as “numbers of apoptotic cells divided by the area of the picture (unit: cells/μm2).
data analysis and statistics
ANOVA and Wilcoxon rank sum test were used to analyze the difference of nerve density and apoptosis events at different stages of HOI. P<0.01 was defined as statistically significant. We also analyzed the correlation of nerve density, sympathetic nerve density, and apoptotic cell density by Pearson correlation analysis. P<0.01 was defined as statistically significant.
【Results】
specimen collection and staging result of HOI
There were103 cutaneous hemangiomas. Among them, 42 were GLUT-1 positive (Fig. 1A), diagnosing as HOI: 12 were proliferating HOI, 19 were early involuting HOI, 11 were late involuting HOI (Fig. 2). Among 61 GLUT-1 negative hemangiomas (Fig. 1B), there were 3 pyogenic granulomas (Table 1).
early involuting HOI had highest density of nerve
Nerve density at different stages of 42 HOI demonstrated by PGP9.5 immunohistochemistry showed (mean ± SD, unit: μm2/μm2): proliferating- (5.2±2.9) x 10-5, early involuting-(1016.7±173.6) x10-5, late involuting-(0.5±0.8) x10-5 (Fig. 3, Fig. 4). The difference was statistically significant by ANOVA and Wilcoxon rank sum test (Table 2).
early involuting HOI had highest density of sympathetic nerve
Tyrosine hydroxylase immunohistochemistry showed the sympathetic nerve density at different stages of HOI (mean ± SD, μm2/μm2) was: proliferating-(3.2±2.8) x10-5, early involuting-(1043.8±212.7) x 10-5, late involuting-(0.3±0.5) x 10-5 (Fig. 5, Fig. 6). The difference was statistically significant by ANOVA and Wilcoxon rank sum test (Table 3).
the majority of nerves in early involuting HOI were sympathetic nerves
Compared with density of nerves, nearly each of the sympathetic nerve density of early involuting HOI had an approximate value (Fig. 7).
early involuting HOI had highest density of apoptotic cells
Activated caspase 3 immunohistochemistry showed the apoptotic cell density at different stages of HOI (mean ±SD, cells/μm2): proliferative- (0.0±0.0) x 10-5，early involuting-(112.7±25.0) x 10-5，late involuting-(1.0±1.8) x 10-5(Fig. 8, Fig. 9). The difference was statistically significant by ANOVA test (Table 4).
density of nerve, sympathetic nerve, and apoptotic cells at different stages of HOI had strong correlation
It showed that early involuting HOI showed an extraordinary high density of nerves, sympathetic nerves, and apoptotic cells. By Pearson’s correlation test, the coefficient of nerve density and sympathetic nerve was 0.97094, of nerve density and apoptotic cell density was 0.95329, and of apoptotic cell density and sympathetic nerve density was 0.95252. They were all statistically significant by Pearson correlation analysis (Table 5).
【Discussion】
diagnosis and staging of HOI
Although GLUT-1 has been a well-known specific marker for pathological diagnosis of HOI [North PE et al., 2000], However, most of the studies on HOI did not confirm the diagnosis of HOI by GLUT-1 immunohistochemical staining. They usually made the diagnosis by clinical observation. In this study, we used GLUT-1 immunohistochemistry to confirm the diagnosis of HOI. We found that only 41% of capillary hemangiomas were HOI (42%, if pyogenic granulomas were excluded). It was far less than the percentage reported in previous study [North PE et al., 2000], in which the percentage of HOI was 97%. Thus, the importance of GLUT-1 in diagnosis and histological study of HOI should be kept in mind.
The controversy of clinical staging of HOI is mainly on the early stage of involution. The regression of HOI was evidenced by reduction of the size and change of the color (became dark red or grayish white). However, the initiation of regression may originate from deeply central part of the tumor, which is invisible. Thus, it may need quantification of urinary bFGF or immuno- histochemical staining of tissue markers to identify the early change of HOI. We may expect that molecules related with cell proliferation/death or angiogenesis regulation may be useful markers [Chen D et al., 2005; Ritter MR et al., 2002; Takahashi K et al., 1994]. Previous studies did find several tissue biomarkers at various stages of HOI, however their staging system was based on clinical observation and age of the patients. Thus, cellular markers for staging of HOI still need further investigation.
Moreover, our study created strict morphological criteria for pathological staging of HOI. We also define the difference between early and late involuting HOI, in order to find the change that may be related to the regression of HOI.
early involuting HOI had highest density of nerves, sympathetic nerves, and apoptotic cells
It has been well known that PGP9.5 is a pan-neuronal marker, the best target for immunohistochemical staining of nerve fibers [Reilly DM et al., 1997]. However, among previous studies on nerve of HOI, only one used PGP9.5 immunohistochemistry [Adegboyega PA et al., 2005; Jang YC et al., 2000]. They neither used GLUT-1 immunohistochemistry to confirm the diagnosis of HOI, nor did precise staging for HOI. Thus, the power of their conclusion about the nerve density at various stages of HOI was not strong enough.
The reason that PGP9.5 immunohistochemistry was rarely used in studies on nerves of HOI is technical difficulty. PGP9.5 can be easily stained with fresh frozen tissues; however, it is poorly stained with paraffin sections. Our specimens were old paraffin sections, which needed time-consuming repeated testing procedures to define the optimal conditions. Our success at demonstrating PGP9.5 immunohistochemistry in this collection of HOI paraffin-embedded specimen has a great contribution to the reliability of nerve density of HOI.
Our study showed that early involuting HOI had highest density of nerves, which was far more than that of proliferating and late involuting HOI. The nerve density of proliferating and late involuting HOI was extremely low. In previous studies that used S-100 protein or PGP9.5 immunohistochemistry to evaluate the nerve density of HOI, they found no significant difference between proliferating/ involuting HOI and normal dermis [Adegboyega PA et al., 2005; Jang YC et al., 2000]. Our results were quite different from previous studies. It may be due to different criteria of staging. On the other hand, the proliferating HOI of previous studies might include “early involuting” HOI, so they had similar nerve density to normal dermis. Because our study strictly defined the entity of early involuting HOI, we discovered a new finding that had never been noted.
The constitution of nerves of HOI had not been further investigated in previous reports; however, we found that early involuting HOI had large population of sympathetic nerves. This is novel and unexpected, because there is no sympathetic nerve innervation beyond terminal arteriole in normal dermis. It may imply that abnormal sympathetic nerves appeared in early involuting HOI have a role in inducing regression of HOI.
density of nerves, sympathetic nerves, and apoptotic cells at different stages of HOI had strong correlation
Our results showed that density of nerves, sympathetic nerves, and apoptotic cells at different stages of HOI were highly associated. That is, while the density of nerve peaks, sympathetic nerves and apoptotic cells have the highest density. It also means that the main constituent of the nerves of early involuting HOI is sympathetic nerves. At the same time, apoptotic cells appeared with greatest amount. This is an indirect evidence for apoptosis of HOI induced by sympathetic nerve.
reliability of morphological criteria in this study for staging of HOI
In 1998, Razons et al used terminal deoxynucleotidyl transferase- mediated deoxyuridine triphosphate nick end labeling (TUNEL) to detect apoptotic cells in HOI. They found scanty apoptotic cells in proliferating stage, but many in involuting stages [Razon MJ et al., 1998]. Our study used activated caspase 3 (AC3) to detect early apoptotic events in HOI, and the result was similar. We found that the density of apoptotic cells was highest at early involuting stage, and very low in proliferating and late involuting stages. We successfully demonstrated not only that AC3 could sufficiently detect apoptosis in HOI, but also our morphological criteria had a great reliability.
nerve tissue inhibit angiogenesis
It has been noted that one of the prognostic factors of neuroblastoma is pathological pattern-higher population of Schwann cells will result in better outcome. Huang successfully demonstrated that Schwann cells could secret tissue inhibitor of metalloproteinase 2 (TIMP 2) to inhibit angiogenesis [Huang D et al., 2000]. Crawford also found that Schwann cell produced pigment epithelium-derived factor (PEDF) [Crawford SE et al., 2001]. Chlenski reported that Schwann cell could synthesize secreted protein acidic and rich in cysteine (SPARC) [Chlenski A et al., 2002]. Thus, Schwann cells had multiple factors contributory to inhibition of angiogenesis.
Schwann cell was seen in both myelinated and non-myelinated nerves. So, the Schwann cell accompanying nerves in early involuting HOI may produce some angiogenesis inhibitors to induce regression of HOI. So far, there is no known report about sympathetic axon producing angiogenesis inhibitor. However, we know that sympathetic nerve is distributed along with vessels. There have been some mutually positive feedback regulating factors reported in the literature [Honma Y et al., 2002; Mukouyama TS et al., 2004]. We may reasonably hypothesize they have also negative control factors. In some condition, sympathetic nerve may produce angiogenesis inhibitors.
【Conclusion】We conclude that abnormal density of nerve is evident at different stages of HOI. Especially, very high density of sympathetic nerve and high activity of apoptosis appeared simultaneously in the early involuting HOI, which provides a strong pathological evidence of nerve’s role in involution of HOI. These abnormally located sympathetic nerves may have important effect on regression of HOI. Since there have been no reports on the suppressive effect on vessels, our finding is unique.
Subjects
嬰兒型血管瘤
第一型葡萄糖運輸蛋白
鹼性纖維母細胞生長因子
血管內皮細胞生長因子
免疫組織化學染色法
蛋白質基因產物9.5
酪胺酸羥化酶
活化凋亡蛋白3
交感神經
hemangioma of infancy
glucose transporter I
basic fibroblast growth factor
vascular endothelial growth factor
immunohistochemistry
protein gene product 9.5
tyrosine hydroxylase
activated caspase 3
sympathetic nerve
Type
other
File(s)![Thumbnail Image]()
Loading...
Name
ntu-95-P92421012-1.pdf
Size
23.31 KB
Format
Adobe PDF
Checksum
(MD5):c35198637a089102f43e3d73e65b711f