CASE REPORT
CASE REPORT

De Novo Arteriovenous Malformation Growth
Secondary to Implantation of Genetically Modified
Allogeneic Mesenchymal Stem Cells in the Brain
Makoto Nakamura, MD*
Amir Samii, MD‡
Josef M. Lang, MD*
Friedrich Götz, MD§
Madjid Samii, MD‡
Joachim K. Krauss, MD*
*Department of Neurosurgery, Hannover
Medical University, Hannover, Germany;
‡Department of Neurosurgery, International Neuroscience Institute, Hannover,
Germany; §Institute of Diagnostic and
Interventional Neuroradiology, Hannover
Medical University, Hannover, Germany
Correspondence:
Makoto Nakamura, MD,
Department of Neurosurgery,
Hannover Medical School,
Carl-Neuberg Str. 1,
30625 Hannover, Germany.
E-mail: nakamura.makoto@mh-hannover.de
Received, February 9, 2015.
Accepted, August 6, 2015.
Published Online, September 16, 2015.
Copyright © 2015 by the
Congress of Neurological Surgeons.

BACKGROUND AND IMPORTANCE: Local biological drug delivery in the brain is an
innovative field of medicine that developed rapidly in recent years. Our report illustrates
a unique case of de novo development of a cerebral arteriovenous malformation (AVM)
after implantation of genetically modified allogeneic mesenchymal stem cells in the
brain.
CLINICAL PRESENTATION: A 50-year-old man was included in a prospective clinical
study (study ID number CM GLP-1/01, 2007-004516-31) investigating a novel neuroprotective approach in stroke patients to prevent perihematomal neuronal damage. In
this study, alginate microcapsules containing genetically modified allogeneic mesenchymal stem cells producing the neuroprotective glucagon-like peptide-1 (GLP-1) were
implanted. Three years later, the patient presented with aphasia and a focal seizure due
to a new left frontal intracerebral hemorrhage. Angiography revealed a de novo left
frontal AVM.
CONCLUSION: The development of an AVM within a period of 3 years after implantation of the glucagon-like peptide-1–secreting mesenchymal stem cells suggests
a possible relationship. This case exemplifies that further investigations are necessary to
assess the safety of genetically modified cell lines for local biological drug delivery in the
brain.
KEY WORDS: AVM, GLP-1, Neuroprotection, Stem cells
Neurosurgery 78:E596–E600, 2016

DOI: 10.1227/NEU.0000000000001025

BACKGROUND AND IMPORTANCE

L

ocal biological drug delivery in the brain is
an innovative field of medicine that developed rapidly in recent years.
The use of mesenchymal stem cells (MSCs) has
been reported to elicit neuroprotective and
regenerative effects mainly through the release
of neurotrophic and immunomodulatory peptides that may well be a source of trophic support,

WHAT IS THIS BOX?
A QR Code is a matrix
barcode readable by QR
scanners, mobile phones
with cameras, and
smartphones. The QR
Code above links to
Supplemental Digital
Content from this
article.

ABBREVIATIONS: ALK1, activin-like kinase 1; AVM,
arteriovenous malformation; ENG, endoglin; GLP-1,
glucagon-like peptide-1; GLP-1R, glucagon-like
peptide-1 receptor; MSC, mesenchymal stem cell
Supplemental digital content is available for this article.
Direct URL citations appear in the printed text and are
provided in the HTML and PDF versions of this article on
the journal’s Web site (www.neurosurgery-online.com).

E596 | VOLUME 78 | NUMBER 4 | APRIL 2016

www.neurosurgery-online.com

promoting endogenous repair such as neurogenesis, angiogenesis, and synaptogenesis.1
With regard to the safety of cell therapy using
MSCs, a large meta-analysis of clinical trials
under various pathological conditions did not
show any evidence of severe side effects related to
MSC transplantation, such as acute infusionrelated toxicity, complications in peripheral
organ systems, infection, death, and tumor
formation.2 Although MSCs themselves are
not tumorigenic, migration to existing primary
tumors and modification or even stimulation of
tumor growth due to their immunomodulatory
properties cannot be completely excluded.3
Encapsulated cell biodelivery has been introduced as a novel clinical strategy for cell therapy in
the central nervous system. Encapsulation with
semipermeable hollow fibers,4 as well as spherical
polymeric microcapsules,5 protects cells

www.neurosurgery-online.com

Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited

DE NOVO AVM FORMATION

transplanted into the brain from the immunological graft-vs-host
response. Because the capsules permit the free passage of nutrients,
oxygen, and smaller molecules, the cells are maintained within the
capsules and can produce and deliver therapeutic peptides to the
brain.4,5 Encapsulated cells have already been used for the treatment
of diabetes mellitus,6 amyotrophic lateral sclerosis,7,8 chronic pain,9
Huntington disease,10 and malignant brain tumors.11-13
So far, no severe side effects have been reported concerning the
use of encapsulated MSCs for biological drug delivery in the brain.
Our report illustrates a unique case of the de novo development of
a cerebral arteriovenous malformation (AVM) after implantation
of genetically modified allogeneic MSCs.

CLINICAL PRESENTATION
A 50-year-old man presented with a left frontal intracerebral
hemorrhage (Figure 1A). Cerebral magnetic resonance imaging
(Figure 1B) and angiography (Figure 1C) did not reveal any
vascular malformation. After providing informed consent, the
patient was included in a prospective clinical study (study ID
number CM GLP-1/01, 2007-004516-31; ClinicalTrials.gov
study number NCT01298830; Institutional review board
approval by the Independent Ethics Committee of Hannover
Medical School) investigating a novel neuroprotective approach in
stroke patients to prevent perihematomal neuronal destruction.
The hemorrhage was surgically evacuated. There was no evidence
of an AVM. Alginate microcapsules containing genetically
modified allogeneic MSCs producing neuroprotective glucagonlike peptide-1 (GLP-1) were implanted. To prevent migration of
the cell capsules from the implantation site, the capsules were

transplanted within a polypropylene mesh bag (1 · 1 cm in size)
with mesh pores restraining the cell capsules but allowing free fluid
convection. The patient recovered well after surgery and rehabilitation therapy. The microcapsules were removed uneventfully
after 2 weeks. Three years later, he presented with aphasia and
a focal seizure due to a new left frontal intracerebral hemorrhage
(Figure 2A). Computed tomography angiography (Figure 2B) and
digital subtraction angiography (Figures 2C and 2D; see
Video, Supplemental Digital Content, http://youtu.be/
ewsMsTQXlxc, which shows the sequence of left carotid
angiography showing the arterial and venous phases of angiography
in the anterolateral and lateral projections) revealed a de novo left
frontal AVM. During surgery, the AVM was located superficially
just at the site of the original craniotomy. The intracerebral
hemorrhage and AVM were surgically removed. The patient
recovered well and underwent another operation 1 year later for
removal of a small residual AVM nidus.

DISCUSSION
AVMs are considered to be congenital in nature, caused by
a defect in primordial capillary or venous formation. They are
characterized by vessels with thin or irregular lamina muscularis
and elastica, endothelial thickening, and hypertrophy of the tunica
media in the nidus.14 Only a very few cases are detected in utero
or in children15 compared with the vast majority of diagnoses
made during adulthood.14-16 Only 16 cases of de novo AVMs
have been reported in the literature. These patients had
undergone high-resolution baseline imaging that showed no
evidence of vascular malformations; however, only 6 of these

FIGURE 1. A, left frontal intracerebral hemorrhage on the initial computed tomography scan. B, magnetic resonance imaging on the initial presentation did not reveal any
AVM. C, left initial carotid angiography (lateral view) before initial surgery, no AVM is visible. AVM, arteriovenous malformation.

NEUROSURGERY

VOLUME 78 | NUMBER 4 | APRIL 2016 | E597

Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited

NAKAMURA ET AL

FIGURE 2. A, 3 years after surgery for hematoma removal and implantation of genetically engineered human mesenchymal
stromal cells producing glucagon-like peptide-1, the patient presented with a new left frontal intracerebral hemorrhage. B,
computed tomography angiography showed a de novo AVM superficial to the site of hemorrhage. C, left carotid angiography
(lateral view) with de novo AVM (white arrow pointing to nidus). A distal insular branch of the middle cerebral artery is
identified as the feeding artery (black arrow). D, left carotid angiography (lateral view). Superficial draining vein to the superior
sagittal sinus and transverse sinus (black arrows). AVM, arteriovenous malformation.

patients were initially evaluated with catheter angiography. The
diagnosis of the de novo AVM was confirmed after an interval
ranging from 2 to 17 years, and patients were 6 to 68 years old at
the time of de novo AVM diagnosis (mean, 26.4 years).16
The pathogenesis of AVMs is not sufficiently understood. It
may involve an interaction between genetic susceptibility and
environmental (acquired) factors. Several studies of patients with

E598 | VOLUME 78 | NUMBER 4 | APRIL 2016

hereditary hemorrhagic telangiectasia, an autosomal dominant
disease that includes AVMs at multiple sites, have revealed that the
gene coding for endoglin (ENG) or activin-like kinase 1 (ALK1) is
mutated. Both ENG and ALK1 are proteins involved in signaling
pathways critical for angiogenesis and inflammation.17 A possible
genetic influence in the formation of AVMs was also suggested by
reports of familial clustering of sporadic AVMs.18

www.neurosurgery-online.com

Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited

DE NOVO AVM FORMATION

Some previous studies have shown that cytokines such as
vascular endothelial growth factor, which has a key role in
angiogenesis, may foster the development of cerebral AVMs.15,19
However, the exact triggering event that leads to increased and
disorganized angiogenesis ultimately leading to AVM formation
has not been identified. In a mouse model with ALK1 deficiency,
the development of de novo AVMs was observed after angiogenic
stimulation.20 The question remains whether the presence of
a genetic abnormality alone is enough to trigger AVM formation
or a second angiogenic stimulus is required.21
GLP-1 is an endogenous insulinotropic peptide secreted from
the gastrointestinal tract in response to food. It enhances
pancreatic islet beta-cell proliferation and glucose-dependent
insulin secretion and lowers blood glucose and food intake in
patients with type 2 diabetes mellitus. GLP-1 and its receptor
(GLP-1R) are also found in the central nervous system. GLP-1Rs
are expressed throughout the brain and mediate important effects
on neuroprotection and cognition.22 Stimulation of the GLP-1R
in the brain is associated with neurotrophic functions that,
additionally, can protect cells against excess glutamate and other
toxic insults23 and has been shown to improve learning and
memory in GLP-1R–deficient mice.22
With regard to the role of GLP-1 in angiogenesis, in vitro
studies have suggested that GLP-1 may promote endothelial cell
proliferation and angiogenesis, which is mediated through the Akt,
PKC, and src pathways.24
The delivery of GLP-1 via encapsulated genetically modified
MSCs in an animal model of myocardial infarction provided
a prolonged distribution of GLP-1 and paracrine stem cell factors,
which improved left ventricular function and reduced epicardial
infarct size. This was associated with increased angiogenesis and an
altered remodeling response. These results suggested that encapsulated genetically modified MSCs would be beneficial for
recovery after myocardial infarction.25
MSCs or multipotent stromal cells have the capability of secreting
a multitude of bioactive molecules. Numerous angiogenic factors,
growth factors, and cytokines have been discovered in the MSC
secretome; all have been demonstrated to alter endothelial cell
behavior in vitro and induce angiogenesis in vivo. As a consequence,
MSCs have been widely explored as a promising treatment strategy
in disorders caused by insufficient angiogenesis such as chronic
wounds, stroke, and myocardial infarction.26
In the current study, intracerebral hemorrhage was chosen as
a disease model to investigate the safety of encapsulated MSC
biodelivery of GLP-1 in a phase I/II trial. Microencapsulated
allogeneic MSCs were transplanted in the brain tissue cavity after
neurosurgical evacuation of the hematoma. The objective of this
approach is to improve the outcome after surgery for intracerebral
hemorrhage; the local, neuroprotective, and anti-inflammatory
cell therapy was targeting the secondary neuronal injury in the
perihematomal area occurring in the first weeks after bleeding.27
Previous studies have shown that encapsulated MSC–based
delivery of GLP-1 has the capability to achieve concentrations
in the cerebrospinal fluid that are pharmacologically active.27,28

NEUROSURGERY

The interim evaluation of the currently described study in stroke
patients revealed that as many as 30% of the transplanted MSCs
survived the 2-week implantation period and were still secretorily
active after explantation.27
Our report describes the case of a patient with de novo
development of a cerebral AVM after implantation of genetically
modified allogeneic MSCs in the brain after surgical removal of an
intracerebral hemorrhage. The imaging studies (computed tomography, magnetic resonance imaging, and angiography) at the time
of initial presentation did not reveal any vascular lesion, which may
have been the cause of the intracerebral hemorrhage. Although an
angiographically occult AVM due to thrombosis or compression in
the setting of the hemorrhage cannot be ruled out completely with
imaging studies alone, the absence of any superficial AVM could
be confirmed intraoperatively. The AVM was detected 3 years later
just underneath the craniotomy site. Therefore, even an angiographically occult AVM would have been detected easily during
the first surgical evacuation of the intracerebral hemorrhage.
Histological examination of the specimen confirmed with
certainty the diagnosis of an AVM. A hypervascular tumor was
excluded.
The potential cause of de novo AVM formation in our patient
remains hypothetical. The encapsulation of genetically modified
MSCs prevented any direct migration of MSCs into the brain
parenchyma; however, local secretion of angiogenic factors and
cytokines with an influence on endothelial proliferation and
angiogenesis cannot be excluded.
As recently reported, GLP-1 itself may promote endothelial cell
proliferation and angiogenesis.24 Further studies are needed to
confirm whether GLP-1 may also directly trigger the formation of
de novo AVMs in the brain. It also needs to be considered that
a foreign body reaction against the polypropylene mesh may have
been a possible pathomechanism for de novo AVM formation.
The clinical trial investigating GLP-1–secreting MSCs for the
prevention of secondary neuronal after intracerebral hemorrhage
was terminated after treatment of 11 patients. Although no safety
concerns emerged during the study period and the efficacy results
were promising, the reduced cell viability attributed to the
implantation site led to the conclusion that the design of the
containment for GLP-1 Cell Beads (BTG International Germany
GmbH, Alzenau, Germany) needed to be re-engineered. Because no
additional knowledge was expected to be gained from continuation
of the study, it was decided to terminate the study early.

CONCLUSION
De novo formation of a cerebral AVM has been observed only in
16 cases so far.16 The development of an AVM within 3 years
after implantation of the GLP-1–secreting CellBeads suggests
a possible relationship. Possible pathomechanisms include a local
response to potentially angiogenetic peptides secreted by the
MSCs or the GLP-1 itself, which may promote endothelial cell
proliferation and angiogenesis. A foreign body reaction against
the polypropylene mesh may also need to be considered.

VOLUME 78 | NUMBER 4 | APRIL 2016 | E599

Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited

NAKAMURA ET AL

This case exemplifies that further investigations are necessary to
assess the safety of genetically modified cell lines for local biological
drug delivery in the brain.
Disclosure
The authors have no personal, financial, or institutional interest in any of the
drugs, materials, or devices described in this article.

REFERENCES
1. Savitz SI. Introduction to cellular therapy: the next frontier for stroke therapeutics.
Stroke. 2009;40(3 suppl):S141-S142.
2. Lalu MM, McIntyre L, Pugliese C, et al. Safety of cell therapy with mesenchymal
stromal cells (SafeCell): a systematic review and meta-analysis of clinical trials.
PLoS One. 2012;7(10):e47559.
3. Lazennec G, Jorgensen C. Concise review: adult multipotent stromal cells and
cancer: risk or benefit? Stem Cells. 2008;26(6):1387-1394.
4. Tseng JL, Aebischer P. Encapsulated neural transplants. Prog Brain Res. 2000;127:
189-202.
5. Murua A, Portero A, Orive G, Hernandez RM, de Castro M, Pedraz JL. Cell
microencapsulation technology: towards clinical application. J Control Release.
2008;132(2):76-83.
6. Soon-Shiong P, Heintz RE, Merideth N, et al. Insulin independence in a type 1
diabetic patient after encapsulated islet transplantation. Lancet. 1994;343(8903):
950-951.
7. Aebischer P, Schluep M, Deglon N, et al. Intrathecal delivery of CNTF using
encapsulated genetically modified xenogeneic cells in amyotrophic lateral sclerosis
patients. Nat Med. 1996;2(6):696-699.
8. Zurn AD, Henry H, Schluep M, et al. Evaluation of an intrathecal immune
response in amyotrophic lateral sclerosis patients implanted with encapsulated
genetically engineered xenogeneic cells. Cell Transpl. 2000;9(4):471-484.
9. Buchser E, Goddard M, Heyd B, et al. Immunoisolated xenogenic chromaffin cell
therapy for chronic pain. Initial clinical experience. Anesthesiology. 1996;85(5):
1005-1012; discussion 29A-30A.
10. Bachoud-Levi AC, Maison P, Bartolomeo P, et al. Retest effects and cognitive
decline in longitudinal follow-up of patients with early HD. Neurology. 2001;56
(8):1052-1058.
11. Bjerkvig R, Read TA, Vajkoczy P, et al. Cell therapy using encapsulated cells
producing endostatin. Acta Neurochir Suppl. 2003;88:137-141.
12. Read TA, Stensvaag V, Vindenes H, Ulvestad E, Bjerkvig R, Thorsen F. Cells
encapsulated in alginate: a potential system for delivery of recombinant proteins to
malignant brain tumours. Int J Dev Neurosci. 1999;17(5-6):653-663.
13. Visted T, Lund-Johansen M. Progress and challenges for cell encapsulation in
brain tumour therapy. Expert Opin Biol Ther. 2003;3(4):551-561.

E600 | VOLUME 78 | NUMBER 4 | APRIL 2016

14. Choi JH, Mohr JP. Brain arteriovenous malformations in adults. Lancet Neurol.
2005;4(5):299-308.
15. Stevens J, Leach JL, Abruzzo T, Jones BV. De novo cerebral arteriovenous
malformation: case report and literature review. AJNR Am J Neuroradiol. 2009;30
(1):111-112.
16. Morales-Valero SF, Bortolotti C, Sturiale CL, Lanzino G. Are parenchymal AVMs
congenital lesions? Neurosurg Focus. 2014;37(3):E2.
17. Kim H, Marchuk DA, Pawlikowska L, et al. Genetic considerations relevant to
intracranial hemorrhage and brain arteriovenous malformations. Acta Neurochir
Suppl. 2008;105:199-206.
18. van Beijnum J, van der Worp HB, Schippers HM, et al. Familial occurrence of
brain arteriovenous malformations: a systematic review. J Neurol Neurosurg
Psychiatry. 2007;78(11):1213-1217.
19. Kilic T, Pamir MN, Kullu S, Eren F, Ozek MM, Black PM. Expression of
structural proteins and angiogenic factors in cerebrovascular anomalies. Neurosurgery. 2000;46(5):1179-1191; discussion 1191-1192.
20. Chen W, Sun Z, Han Z, et al. De novo cerebrovascular malformation in the adult
mouse after endothelial Alk1 deletion and angiogenic stimulation. Stroke. 2014;45
(3):900-902.
21. Leblanc GG, Golanov E, Awad IA, Young WL. Biology of vascular malformations
of the brain. Stroke. 2009;40(12):e694-e702.
22. During MJ, Cao L, Zuzga DS, et al. Glucagon-like peptide-1 receptor is involved
in learning and neuroprotection. Nat Med. 2003;9(9):1173-1179.
23. Greig NH, Mattson MP, Perry T, et al. New therapeutic strategies and drug
candidates for neurodegenerative diseases: p53 and TNF-alpha inhibitors, and
GLP-1 receptor agonists. Ann N Y Acad Sci. 2004;1035:290-315.
24. Aronis KN, Chamberland JP, Mantzoros CS. GLP-1 promotes angiogenesis in
human endothelial cells in a dose-dependent manner, through the Akt, Src and
PKC pathways. Metabolism. 2013;62(9):1279-1286.
25. Wright EJ, Farrell KA, Malik N, et al. Encapsulated glucagon-like peptide-1producing mesenchymal stem cells have a beneficial effect on failing pig hearts.
Stem Cells Transl Med. 2012;1(10):759-769.
26. Bronckaers A, Hilkens P, Martens W, et al. Mesenchymal stem/stromal cells as
a pharmacological and therapeutic approach to accelerate angiogenesis. Pharmacol
Ther. 2014;143(2):181-196.
27. Heile A, Brinker T. Clinical translation of stem cell therapy in traumatic brain
injury: the potential of encapsulated mesenchymal cell biodelivery of glucagon-like
peptide-1. Dialogues Clin Neurosci. 2011;13(3):279-286.
28. Glage S, Klinge PM, Miller MC, et al. Therapeutic concentrations of glucagon-like
peptide-1 in cerebrospinal fluid following cell-based delivery into the cerebral
ventricles of cats. Fluids Barriers CNS. 2011;8:18.

Supplemental digital content is available for this article. Direct URL citations
appear in the printed text and are provided in the HTML and PDF versions of this
article on the journal’s Web site (www.neurosurgery-online.com).

www.neurosurgery-online.com

Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited