Clinical Neurology and Neurosurgery 110 (2008) 838–842

Case report

Fatal case of intracerebral hemorrhage during gamma knife
treatment for metastases
William S. Anderson a,∗ , Laurel E. Moore b,1 , Eric Ford c,2 , Daniele Rigamonti d,3
a The Johns Hopkins University School of Medicine, Department of Neurosurgery, Meyer Building 8-161,

600 North Wolfe Street, Baltimore, MD 21287, United States
b The University of Michigan, Department of Anesthesiology, 1500 E Medical Center Drive, Ann Arbor, MI 48109-0048, United States
c Department of Radiation Oncology and Molecular Radiation Sciences, The Johns Hopkins University School of Medicine,

401 North Broadway, Room 1440, Baltimore, MD 21231, United States
d The Johns Hopkins University School of Medicine, Department of Neurosurgery, Phipps Building, Room 104,

600 North Wolfe Street, Baltimore, MD 21287, United States
Received 15 January 2008; received in revised form 11 May 2008; accepted 15 May 2008

Abstract
This case report presents a patient suffering acute fatal intracranial–intratumoral hemorrhage during a gamma knife treatment session. Acute
hemorrhage during a radiosurgery session is extremely rare and a plausible cause for this case is discussed along with a literature review of
previously reported incidents. The patient was a 71-year-old male presenting with three large intracranial lesions and an underlying primary
renal cell carcinoma malignancy. Because of a severe kyphotic deformity resulting from ankylosing spondylitis, the patient was placed in a
moderate Trendelenburg position to allow his head to fit into the gamma knife unit during the radiosurgery session. The two left-sided lesions
were to be treated with 20 Gy to the 50% isodose line, and the right-sided lesion with 16 Gy to the 40% isodose line. Anesthesia was available
throughout the treatment session to aid with pain control. The gamma knife treatment was aborted because the patient suffered a generalized
seizure while in the unit. Immediate head CT of the patient revealed large acute hemorrhages into all three intracranial masses. This proved to
be a fatal complication. It is likely that this positioning contributed to the hemorrhage. The clinical history of this patient is provided as well
as a review of the literature on acute intracranial hemorrhage associated with radiosurgical therapy.
© 2008 Elsevier B.V. All rights reserved.
Keywords: Gamma knife; Intratumoral hemorrhage; Ankylosing spondylitis; Radiosurgery complication

1. Introduction
Radiosurgery is a widely used therapy for the control
of solitary and multiple metastatic brain tumors, which is
the most common use for this treatment modality [1–3].
In a large series of patients with solitary metastases of
mixed sources, radiosurgery provided a median survival
of 11 months after diagnosis [1]. Efficacy has also been
∗ Corresponding author. Tel.: +1 410 955 6405; fax: +1 410 502 5768.

E-mail addresses: wanderso@jhmi.edu (W.S. Anderson),
laurelmo@med.umich.edu (L.E. Moore), eric.ford@jhmi.edu (E. Ford),
dr@jhmi.edu (D. Rigamonti).
1 Tel.: +1 734 936 4235.
2 Tel.: +1 410 502 1477.
3 Tel.: +1 410 955 2259; fax: +1 410 955 9126.
0303-8467/$ – see front matter © 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.clineuro.2008.05.012

demonstrated for specific pathologies such as renal cell carcinoma [4] and non-small cell lung carcinoma metastases [5].
A level of safe tumor control has also been demonstrated
for repeat radiosurgery for metastases [6]. Complications
of radiosurgery for solitary metastatic lesions include the
development of neighboring radiation necrosis, increases in
surrounding edema, and tumor hemorrhage (actuarial rate at 1
and 2 years of 7.6% in mixed pathology [1]). In a small group
of patients with renal cell carcinoma (52 treated lesions),
complications were very rare and none of them suffered a
delayed hemorrhage [4], while in a larger series (75 patients
with 350 treated lesions), the delayed hemorrhage rate was
12% [7].
Acute bleeding taking place during a radiosurgery treatment session for a metastatic lesion, or in the immediate

W.S. Anderson et al. / Clinical Neurology and Neurosurgery 110 (2008) 838–842

839

Fig. 1. T1 post-gadolinium contrast axial MRI images demonstrating the location of the three treated metastases.

post-treatment period, is extremely rare. We describe in this
case report the occurrence of simultaneous hemorrhage in
several metastatic lesions in one patient during a gamma knife
treatment session. This individual had a history of ankylosing
spondylitis with a severe cervical kyphotic deformity, and had
to be placed in Trendelenburg position to allow the head and
frame to fit correctly into the gamma knife unit. It is possible
that this positioning contributed to the patient’s hemorrhagic
complication. Two previous patients have been described as
having acute onset intratumoral hemorrhage within 24 h after
radiosurgical treatment, one at 15 min post-treatment [8], and
one at 2 h post-treatment [9]. A description of the current
patient suffering a hemorrhage during gamma knife treatment
as well as a review of the previous two patients is presented.
Discussion of the possible role Trendelenburg positioning
and its alteration in venous flow dynamics might have played
in this unusual complication is also included.

2. Case report
A 71-year-old gentleman with a history of ankylosing
spondylitis presented to an outside institution approximately
1 month prior to treatment with new onset left hand weakness. On exam in the emergency room, this consisted of
poor hand intrinsic muscle strength, 2/5 biceps weakness,
C5 dermatomal hypoalgesia, and a decreased biceps reflex.
An MRI performed at that time demonstrated degenerative osteophytic disease from C5-C7 with pronounced cord
impingement. Ten days later the patient underwent a C5-6
anterior cervical discectomy and fusion with plated stabilization and allograft with demineralized bone matrix. A
brief course of perioperative dexamethasone was instituted.

Immediately after surgery, the patient noted a subjective
improvement in left upper extremity strength, but after 3 days
this had reversed its course and both the left arm and leg began
exhibiting pronounced weakness.
Approximately 3 weeks after presentation, a CT of the
brain was performed which revealed three intracranial mass
lesions: a right frontal mass measuring 17 mm × 21 mm in
greatest axial dimension, a left inferior frontal 6-mm diameter mass, and a second left premotor frontal mass of 11-mm
diameter. At that time, an MRI was difficult to obtain because
of the patient’s severe cervical kyphosis related to the ankylosing spondylitis. After diagnosing these intracranial mass
lesions, the patient exhibited a dramatic functional improvement while on dexamethasone. The patient was also placed
on phenytoin, and at no time exhibited seizures prior to treatment. A full oncologic staging diagnostic evaluation was
performed which additionally revealed a 3-mm right pulmonary nodule and a large left renal mass. A fine needle
aspiration of the pulmonary nodule was performed, which
revealed metastatic renal cell carcinoma, positive for CD10.
A plan was proposed to treat these various malignant lesions,
including radiosurgery for the intracranial lesions, possible
radiofrequency ablation for the lung nodule, and possible
laparoscopic treatment of the renal mass.
A system of sloped pads was slipped under the patient’s
back to orient the patient’s head collinear to the MRI head
coil. The preoperative planning images to be used for radiosurgical treatment were obtained (Fig. 1). The same system
of pads was also used on the radiosurgery (gamma knife)
treatment table, effectively putting the patient into 15–20◦ of
Trendelenburg positioning. The patient had significant cervical discomfort at baseline and an anesthesia team was present
to help with positioning and pain management. Because of

840

W.S. Anderson et al. / Clinical Neurology and Neurosurgery 110 (2008) 838–842

Fig. 2. Gamma plan screen snapshots and dose–volume histograms for the (A) right premotor lesion, (B) left inferior frontal lesion, and (C) left premotor
lesion.

concern regarding the stability of the patient’s cervical spine
and to accommodate the patient’s wishes, he received only
small doses of opiate and was awake and conversant throughout the 1-hour MRI and 3-hour radiosurgery treatment.
The planned dosage was for 20 Gy to the 50% isodose line
for both left-sided lesions, and 16 Gy to the 40% isodose line
for the right-sided mass. An additional right frontal white
matter (<1 cm) lesion discovered during the planning MRI
was also to be treated for 20 Gy to the 50% isodose line (maximum dose 40.0 Gy), and a final small (<1 cm) right superior
frontal lesion was to be treated for 20 Gy to the 50% isodose
line (maximum dose 24.8 Gy). The maximum doses to the
three largest lesions were left inferior frontal lesion, 33.5 Gy;

left premotor, 33.3 Gy; right premotor frontal, 31.9 Gy. No
unusually excessive “hot spots” were noted in the pretreatment gamma knife plan. The whole brain maximum dose
was 40.0 Gy. Screenshots of the Gamma Plan (Elekta AB,
Stockholm, Sweden) software during treatment planning as
well as dose–volume histograms for the three large lesions
are presented in Fig. 2.
During the treatment for the posterior portion of the right
frontal lesion and the left superior frontal lesion, the patient
suffered a focal motor seizure involving his left foot. The
patient became acutely hypertensive during this focal seizure
as well. As the patient was pointing toward his foot during
treatment he lost consciousness and suffered a generalized

W.S. Anderson et al. / Clinical Neurology and Neurosurgery 110 (2008) 838–842

Fig. 3. Head CT without contrast demonstrating intratumoral hemorrhages
in the metastatic lesions.

seizure. The patient was immediately removed from the treatment couch and an emergent head CT examination revealed
large hemorrhagic lesions at the site of the metastases (Fig. 3).
The patient never regained consciousness from this generalized seizure, and his neurologic status gradually declined.
Comfort care measures were instituted after extensive discussions between staff and family and he died shortly afterward.
No pretreatment coagulation profile was available for the
patient, but immediately after the post-treatment CT examination, the prothrombin time was 10.3 s, with an activated
partial thromboplastin time of 18.8 s (normal 23.5–34.0 s).

3. Discussion
This case report describes a patient with three intracranial
metastatic tumors from a renal cell carcinoma source who
suffered concurrent hemorrhages in the masses during radiosurgical treatment. A previous case report described by Izawa
et al. [8] presents a 46-year-old woman who was being treated
for multiple brain metastases from lung cancer. The largest of
these metastases had been removed surgically from the left
frontal lobe without complication. On postoperative day 4
she underwent gamma knife treatment for the remaining nine
intracranial lesions (20 Gy to the 50% isodose line). Per Izawa
et al. [8], 15 min post-treatment the patient became unresponsive and the head CT demonstrated a large intraparenchymal
hemorrhage in the left cerebellar hemisphere where one of the
treated metastases was located. The authors monitored this

841

patient’s blood pressure during treatment, and there were no
abnormal values recorded. One case report in the Japanese
literature is similar, describing the case of a patient undergoing LINAC radiosurgical therapy for intracranial metastases
from a breast cancer primary [8,9]. This patient suffered peritumoral hemorrhage 2 h after treatment. Additionally, in the
large series of Wowra et al. [7], one patient with a treated
renal cell metastases did suffer an acute hemorrhage 5 h posttreatment. No further details were provided. A comparative
table of these four cases is presented in Table 1.
One concerning aspect of our patient’s treatment plan
involved his positioning on the gamma knife table. He was
placed into a relative Trendelenburg position by approximately 15–20◦ using a system of pads placed under the
lower back and legs. This position enabled the head with
frame attached to fit appropriately into the collimator helmet, which would otherwise have been impossible due to his
severe kyphotic deformity. It is conceivable that this positioning created an adverse venous pressure gradient making the
intratumoral vascular system more likely to bleed even during
the radiosurgery treatment. Internal jugular venous distention occurs during Trendelenburg positioning, with only 10◦
of positioning needed to increase the mean internal jugular
diameter by 2 mm [10]. In a study 20 healthy volunteers, 25◦
of Trendelenburg positioning produced the maximum amount
of distention [10], which is closer to the angle used during
the gamma knife treatment. There are analogous rises in ICP
with Trendelenburg positioning as well as decreases with the
reverse Trendelenburg positioning, which is often used for
its clinical effect [11,12]. We have no definitive evidence that
this positioning was the cause of our patient’s simultaneous
hemorrhage in all three lesions, but given that acute hemorrhages taking place either during the treatment or within
a few hours after it have been extremely rare, we feel the
positioning was a very likely contributor.
It is not unusual for intracranial neoplasms to undergo
spontaneous hemorrhage, whether they are intrinsic brain
tumors or metastatic lesions. Prospective series indicate that
brain tumors present as hemorrhagic lesions 2–3% of the time
[13–16]. Similarly, in prospective series, a given presenting
spontaneous intracerebral hemorrhage can be associated with
an intracranial tumor approximately 7% of the time [14].
Renal cell carcinoma metastatic lesions in particular have
a propensity to undergo spontaneous hemorrhage [17,18].
But they do seem to hold up fairly robustly to gamma knife
treatment, with one large prospective series of 350 metastases demonstrating a delayed bleed rate of 12% of treated
patients [7]. This further highlights the temporal correla-

Table 1
Comparison of four reported acute (<24 h) hemorrhage cases after radiosurgical treatment
Patient

Pathology

Post-treatment time to neurologic decline

Outcome

71-Year-old male
46-Year-old female [8]
44-Year-old female [9]
NA [7]

Renal cell carcinoma
Lung adenocarcinoma
Breast cancer
Renal cell carcinoma

Occurred during treatment
Acute neurologic decline 15 min after treatment
2 h post-treatment
5 h post-treatment

Fatal
Fatal
Recovery to baseline function
NA

842

W.S. Anderson et al. / Clinical Neurology and Neurosurgery 110 (2008) 838–842

tion of when our patient’s hemorrhage occurred, which we
believe supports the patient’s positioning as contributing to
this morbidity.
It is conceivable that this patient had an underlying coagulopathy which contributed to his acute hemorrhage. No
coagulopathic workup was performed for the patient before
gamma knife therapy. The prothrombin time and activated
partial thromboplastin times were normal for two successive values after the hemorrhage, but this does not rule out
all bleeding disorders. Additionally, the patient’s cause of
hemorrhage might truly have been multifactorial given the
predisposition for renal cell metastatic lesions to hemorrhage,
the unusual position the patient was placed in, and his concurrent radiosurgical therapy. Some combination of these factors
could explain this unfortunate outcome. Preoperative or preprocedure embolization of renal cell carcinoma metastases
has been described, particularly in cases of bony metastases
or spinal compressive lesions [19,20]. Preoperative embolization is also sometimes used for intracranial hypervascular
dural-based masses like meningiomas [21]. This has never
been described specifically for intracranial renal cell metastases, and most likely would not have been useful in this
patient with multiple lesions involving eloquent brain.
4. Conclusion
Given this patient’s severe kyphotic deformity, an alternative radiosurgery treatment device would probably not have
helped change the positioning. The patient would still have
been placed in Trendelenburg position to mate well with the
facemask in the case of a linear accelerator based system.
Another possibility would have been the newer gamma knife
unit currently being marketed (Perfexion, Elekta AB, Stockholm, Sweden). Its more versatile treatment volume might
have been a better choice for our patient if it had been available at our institution, and might not have required such
a significant degree of Trendelenburg angle. Patient positioning has always been an extremely important aspect of
every neurosurgical operative case, and on its own can produce or protect from surgical morbidity. We believe that this
case shows the importance of positioning for radiosurgical
procedures as well.
References
[1] Flickinger JC, Kondziolka D, Lunsford LD, Coffey RJ, Goodman
ML, Shaw EG, et al. A multi-institutional experience with stereotactic
radiosurgery for solitary brain metastasis. Int J Rad Oncol Biol Phys
1994;28:797–802.
[2] Kondziolka D, Patel A, Lunsford LD, Kassam A, Flickinger JC. Stereotactic radiosurgery plus whole brain radiotherapy versus radiotherapy
alone for patients with multiple brain metastases. Int J Rad Oncol Biol
Phys 1999;45:427–34.

[3] Rutigliano MJ, Lunsford LD, Kondziolka D, Strauss MJ, Khanna V,
Green M. The cost effectiveness of stereotactic radiosurgery versus
surgical resection in the treatment of solitary metastatic brain tumors.
Neurosurgery 1995;37:445–55.
[4] Mori Y, Kondziolka D, Flickinger JC, Logan T, Lunsford LD. Stereotactic radiosurgery for brain metastasis from renal cell carcinoma. Cancer
1998;83:344–53.
[5] Kim YS, Kondziolka D, Flickinger JC, Lunsford LD. Stereotactic radiosurgery for patients with nonsmall cell lung carcinoma metastatic to the
brain. Cancer 1997;80:2075–83.
[6] Bhatnagar A, Heron DE, Kondziolka D, Lunsford LD, Flickinger JC.
Analysis of repeat stereotactic radiosurgery for progressive primary
and metastatic CNS tumors. Int J Rad Oncol Biol Phys 2002;53:
527–32.
[7] Wowra B, Siebels M, Muacevic A, Kreth FW, Mack A, Hofstetter A.
Repeated gamma knife surgery for multiple brain metastases from renal
cell carcinoma. J Neurosurg 2002;97:785–93.
[8] Izawa M, Chernov M, Hayashi M, Kubota Y, Kasuya H, Hori T.
Fatal intratumoral hemorrhage immediately after gamma knife radiosurgery for brain metastases: case report. Minim Invas Neurosurg
2006;49:251–4.
[9] Uchino M, Kitajima S, Miyazaki C, Otsuka T, Seiki Y, Shibata I. Peritumoral hemorrhage immediately after radiosurgery for metastatic brain
tumor. No Shinkei Geka 2003;31:911–6.
[10] Clenaghan S, McLaughlin RE, Martyn C, McGovern, Bowra J. Relationship between Trendelenburg tilt and internal jugular vein diameter.
Emerg Med J 2005;22:867–8.
[11] Tankisi A, Larsen JR, Rasmussen M, Dahl B, Cold GE. The effects of
10◦ reverse Trendelenburg position on ICP and CPP in prone positioned
patients subjected to craniotomy for occipital or cerebellar tumours.
Acta Neurochir 2002;144:665–70.
[12] Tankisi A, Rasmussen M, Juul N, Cold GE. The effects of 10◦ reverse
Trendelenburg position on subdural intracranial pressure and cerebral
perfusion pressure in patients subjected to craniotomy for cerebral
aneurysm. J Neurosurg Anesthesiol 2006;18:11–7.
[13] Kondziolka D, Bernstein M, Resch L, Tator CH, Fleming JF, Vanderlinden RG, et al. Significance of haemorrhage into brain tumours:
clinicopathological study. J Neurosurg 1987;67:852–7.
[14] Schrader B, Barth H, Lang EW, Buhl R, Hugo HH, Biederer J, et
al. Spontaneous intracranial haematomas caused by neoplasms. Acta
Neurochir 2000;142:979–85.
[15] Wakai S, Yamakawa K, Manaka S, Takakura K. Spontaneous intracranial haemorrhage caused by brain tumour: its incidence and clinical
significance. Neurosurgery 1982;10:437–44.
[16] Zuccarello M, Pardatscher K, Andrioli C, Fiore DL, Lavicoli R. Brain
tumours presenting as spontaneous intracerebral haemorrhage. Zentralbl Neurochir 1981;42:1–8.
[17] Mandybur TI. Intracranial hemorrhage caused by metastatic tumors.
Neurology 1977;27:650–5.
[18] Wronski M, Arbit E, Russo P, Galicich JH. Surgical resection of
brain metastases from renal cell carcinoma in 50 patients. Urology
1996;47:187–93.
[19] Schirmer CM, Malek AM, Kwan ES, Hoit DA, Weller SJ. Preoperative
embolization of hypervascular spinal metastases using percutaneous
direct injection with n-butyl cyanoacrylate: technical case report. Neurosurgery 2006;59:431–2.
[20] Forauer AR, Kent E, Cwikiel W, Esper P, Redman B. Selective palliative transcatheter embolization of bony metastases from renal cell
carcinoma. Acta Oncol 2007;46:1012–8.
[21] Shi Z.-S., Feng L., Jiang X.-B., Huang Q., Yang Z., Huang Z.S. Therapeutic embolization of meningiomas with onyx for delayed
surgical resection. Surg Neurol 2008 Feb7. [Epub ahead of print.],
doi:10.1016/j.surneu.2007.05.031.