Accepted Manuscript

Why Do Benign Tumors Hemorrhage?

Gregory Albert, MD, MPH

PII: S1878-8750(16)30723-9
DOI: 10.1016/j.wneu.2016.08.056
Reference: WNEU 4467

To appear in: World Neurosurgery

Received Date: 11 August 2016

Accepted Date: 12 August 2016

Please cite this article as: Albert G, Why Do Benign Tumors Hemorrhage?, World Neurosurgery (2016),
doi: 10.1016/j.wneu.2016.08.056.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT

Why Do Benign Tumors Hemorrhage?

Hemorrhage of intracranial tumors is a well-known phenomenon. The presentation is often

dramatic with headache, nausea, vomiting, etc. and may be the first sign that the patient has an

PT
intracranial mass lesion. Tumoral hemorrhage is most commonly associated with high-grade
lesions such as metastases and high-grade gliomas. However, low-grade tumors such as low-
grade gliomas, pituitary adenomas, and meningiomas are also known to hemorrhage. Often,

RI
mechanisms such as rapid tumor growth, necrosis, blood vessel infiltration by tumor or tumor

SC
thrombus, etc. are implicated in the occurrence of intratumoral hemorrhage.[1] However,
these are factors that often do not apply to low grade tumors such as meningiomas and
therefore other pathophysiologic mechanisms must be sought.

U
AN
Hemorrhage within meningiomas is a rare occurrence. These hemorrhages may be
intratumoral, intracerebral, subdural, subarachnoid, or a combination of these. In 2004,
M

Kuzeyli, et al. reported on 11 patients with intratumoral hemorrhage. This was out of a total of
126 patients with meningiomas whom they operated on. However, only two of these patients
D

had radiographically recognized macroscopic hemorrhages, for an incidence of 1.5%. These

authors postulated that the presence of microcysts and/or micronecrosis may increase the risk
TE

of intratumoral hemorrhage in patients with meningiomas.[2] Niiro, et al. reviewed 6 patients

whom they had cared for who presented with hemorrhagic meningiomas. This represented
EP

2.0% of the meningiomas treated at that institution over the same time period. These authors
hypothesized that higher proliferative index may increase the risk of hemorrhage.[3] Other
C

mechanisms of hemorrhage that have been proposed include thin-walled tumor vessels,
AC

erosion of intracranial blood vessels by a growing tumor, tumor necrosis, and stretching of
bridging veins.[4]

In this month’s issue of World Neurosurgery, Wang, et al. offer a histological analysis of
hemorrhagic meningiomas. In their manuscript, they review the pathology of six hemorrhagic
meningiomas and compare them to twelve non-hemorrhagic meningiomas. They identified
 ACCEPTED MANUSCRIPT

both undifferentiated and differentiated blood vessels in specimens from each group.
Interestingly, the hemorrhagic meningiomas had a higher percentage of undifferentiated blood
vessels and fewer numbers of pericytes. The authors propose that these findings are
implicated in the hemorrhagic events.[5]

PT
The authors utilized two immunohistochemical markers to identify blood vessels and categorize
that as differentiated versus undifferentiated. CD31 is found in the junctions between

RI
endothelial cells. CD34 is also found in endothelial cells. The differential staining pattern

SC
whereby CD31 staining is present in all blood vessels whereas CD34 staining is present in only
well-differentiated blood vessels was first described by Poblet, et al. in a study of
angiosarcomas.[6] Since that time, this technique has been applied to other tumor types, as
referenced by the authors of the current study.
U
AN
Using these techniques, the authors of the current study have been able to demonstrate that
M

the overall vascularity of hemorrhagic versus non-hemorrhagic tumors is the same (CD31
staining). However, hemorrhagic tumors contain a greater number of undifferentiated vessels
D

(as indicated by a lower number of CD34-positive blood vessels). These authors then went a
step further and used SMA to stain for pericytes. They were able to demonstrate that fewer
TE

pericytes were present around undifferentiated blood vessels. The undifferentiated nature of
the blood vessels and paucity of pericytes may contribute to the propensity of these tumors to
EP

hemorrhage.[5]
C

Pericytes play critical roles in normal cerebrovascular physiology. They are a critical component
AC

of the neurovascular unit and play a role in angiogenesis. They not only help to promote blood
vessel growth, but they also help stabilize the basement membrane of the mature blood vessel
and promote survival of the endothelial cells.[7] Given the role that pericytes play in vascular
development, it would be interesting to know if the paucity of pericytes identified by these
authors is a cause or consequence of the vascular immaturity in the hemorrhagic meningiomas.
 ACCEPTED MANUSCRIPT

The vasculature of the germinal matrix is comprised of immature blood vessels with few
pericytes. This reduction in pericytes is thought to contribute to a weakening of the
microvasculature and increased propensity for germinal matrix hemorrhage, a common
complication of prematurity well known to neurosurgeons. In fact, deletion of Smad4 in

PT
endothelial cells, which causes detachment of the pericytes from the blood vessels, recreates
the germinal matrix hemorrhagic phenotype in experimental animals.[7]

RI
Pericytes also help maintain the integrity of the blood-brain barrier and contribute to the

SC
regulation of cerebral blood flow in response to hemodynamic changes. Either of these could
contribute to intratumoral hemorrhage. Pericytes have been found to be deficient in cerebral
cavernous malformations, which are, of course, known to be hemorrhagic lesions. The

U
complexity of the signaling pathways between pericytes and endothelium in these situations is
AN
beyond the scope of this commentary.[8]
M

In addition to the reduced number of pericytes, Wang, et al. also found that the hemorrhagic
meningiomas tended to have more dilated blood vessels with thin walls, more evidence of
D

venous thrombosis, and more tumor necrosis. However, these findings did not reach statistical
significance.[5] Thin-walled dilated blood vessels are certainly at greater risk for hemorrhage.
TE

In addition, venous obstruction is a well-known cause of hemorrhage in a variety of situations

such as sinus thrombosis and arteriovenous malformations. Other authors, as noted above,
EP

have implicated tumor necrosis in the hemorrhage of intracranial tumors. However, it is

conceivable that while the presence of necrosis could predispose to hemorrhage, it can also be
C

caused by hemorrhage.
AC

This preliminary finding that it is not the overall vascularity of the tumor, but rather the lack of
differentiation of the blood vessels and reduced number of pericytes that place patients at risk
for intratumoral hemorrhage is an interesting finding. It would be interesting to know if this
same pathophysiology applies to other tumor types such as pituitary adenomas with their risk
for pituitary apoplexy, high-grade gliomas, and brain metastases.
 ACCEPTED MANUSCRIPT

Although this study contains a small number of patients, the authors are to be congratulated
for their contribution to the literature. Intratumoral hemorrhage into meningiomas is an
exceedingly rare event. Therefore, this study is likely to have little impact on the care of

PT
patients with meningiomas. However, it is one additional step to our understanding of the
etiology of intratumoral hemorrhage. Further research needs to be done to determine what
other mechanisms lead to this potentially devastating event in patients with all tumor types.

RI
U SC
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT

References

[1] Yuguang L, Meng L, Shugan Z, Yuquan J, Gang L, Xingang L. Intracranial tumoural

haemorrhage – a report of 58 cases. Journal of Clinical Neuroscience 2002;9(6):637-9.

PT
[2] Kuzeyli K, Çakir E, Usul H, Karaarslan G, Yazar U, Baykal S, Reis A, Cobanoglu Ü.
Intratumoral haemorrhage: a clinical study. Journal of Clinical Neuroscience
2004;11(5):490-2.

RI
[3] Niiro M, Ishimaru K, Hirano H, Yunoue S, Kuratsu J. Clinico-pathological study of
meningiomas with haemorrhagic onset. Acta Neurochirurgica 2003;145(9):767-72.

SC
[4] Kim DG, Park CK, Paek SH, Choe GY, Gwak HS, Yoo H, Jung HW. Meningioma manifesting
intracerebral haemorrhage: a possible mechanism of haemorrhage. Acta
Neurochirurgica 2000;142(2):165-8.

[5]
U
Wang HC, Wang BD, Chen MS, Li SW, Chen H, Xu W, Zhang JM. An underlying
AN
pathological mechanism of the meningiomas with intratumoral hemorrhage:
undifferentiated microvessels. World neurosurgery 2016.

[6] Poblet E, Gonzalez-Palacios F, Jimenez FJ. Different immunoreactivity of endothelial

M

markers in well and poorly differentiated areas of angiosarcomas. Virchows Archiv

1996;428(4-5):217-21.
D

[7] Winkler EA, Bell RD, Zlokovic BV. Central nervous system pericytes in health and disease.
Nature neuroscience 2011;14(11):1398-405.
TE

[8] Sweeney MD, Ayyadurai S, Zlokovic BV. Pericytes of the neurovascular unit: key
functions and signaling pathways. Nature neuroscience 2016;19(6):771-83.
C EP
AC