Keywords
deep seated - exoscope - metastatic brain tumors - minimally invasive approaches -
tubular retractors
Introduction
Metastatic brain tumors are the most common central nervous system tumors in adults.[1]
[2]
[3]
[4]
[5] The most common primary tumors include the lung, breast, kidney, skin, and gastrointestinal
systems.[1]
[2]
[3]
[4]
[5] The treatment of this disease involves some combination of surgery and/or radiation
therapy.[1]
[2]
[3]
[4]
[5] Surgery is typically reserved for patients with good neurologic function, solitary
and accessible lesions, symptomatic lesions, and/or those with good systemic control
of their primary cancer.[2]
[3]
[6]
[7]
A subset of patients with metastatic brain cancer develop deep-seated lesions within
critical white matter tracts, basal ganglia, and/or thalamic structures.[2]
[3]
[6]
[7] These lesions are typically treated with stereotactic radiosurgery (SRS) because
of the potentially increased risk of surgical morbidity associated with accessing
and resecting them.[6]
[7] SRS, unlike surgery, does not improve mass effect, is less effective for larger
lesions, is associated with a delayed treatment effect, and does not provide tissue
for diagnosis and molecular analyses.[6]
[7] However, minimally invasive approaches may decrease the morbidity associated with
conventional surgical approaches for these deep-seated lesions.[8]
[9]
[10]
[11]
[12]
[13]
[14] The use of this approach was primarily limited to studies combining metastatic brain
tumors with other pathology.[10]
[15]
[16] In this study, we summarize our personal experience with these approaches for deep-seated
metastatic brain tumors.
Material and Methods
Patient Selection
Before the start of this study, institutional review board approval was obtained to
collect patient data prospectively. All patients who were operated on by the senior
author were identified prospectively, and patient information and outcomes were collected.
From January 2016 to December 2017, 15 consecutive patients were operated on for a
single deep-seated metastatic brain lesion using minimally invasive approaches. “Deep
seated” was defined as any subcortical location below the deepest adjacent sulcus
in close proximity to the basal ganglia and/or thalamus. “Minimally invasive” was
defined as the use of tubular retractors with exoscopic visualization, as previously
described.[8]
[10]
[11]
[12] The variables that were prospectively collected included age, sex, tumor location,
pre- and postoperative neurologic function, Karnofsky Performance Score (KPS), pre-
and postoperative tumor volume, and length of hospital stay.
General Treatment Strategy
The surgical goal for each case was gross total resection without neurologic deficit.
The patients with deep-seated lesions who were typically offered surgery were those
with solitary and/or symptomatic lesions with significant size (> 2.5 cm) and mass
effect with good recursive partitioning analysis (RPA) scores (class 1–2).[2] In cases where patients presented with RPA class 3 (KPS < 70), surgery was offered
when lesions caused significant mass effect and/or neurologic deficits. Patients with
lesions, especially deep seated ones ≤ 2.5 cm, were typically offered radiation therapy
to minimize the potential morbidity associated with surgical intervention unless there
was significant mass effect and symptoms associated with the lesion. Minimally invasive
approaches were chosen when the lesions were deep seated below the adjacent sulcus
in eloquent regions including critical white matter projection and association tracts,
basal ganglia, and/or thalamus.
The surgical approach was previously described.[10]
[11]
[12] Patients who were offered surgery with the use of minimally invasive approaches
typically underwent high-resolution magnetic resonance imaging (MRI) with and without
contrast and diffusion tensor imaging within 48 hours of surgery ([Figs. 1] and [2]). The planned surgical approach was chosen based on minimizing morbidity to eloquent
cortical and subcortical structures. The patient was taken to the operating room on
the day of surgery and induced and intubated per routine. After intubation, the patient's
head was fixated in a Mayfield skull clamp. If possible, the head was kept neutral
to maintain orientation. Under stereotactic guidance, a small 3-cm skin incision was
then made overlying the planned trajectory, followed by a ∼ 2-cm craniotomy. The craniotomy
was centered on the desired sulcus, thus allowing for a transsulcal, parafascicular
approach to minimize potential injury to adjacent eloquent cortical and subcortical
areas.
Fig. 1 Left basal ganglia non–small cell lung cancer metastasis. This 49-year-old male patient
presented with headaches and confusion, and he underwent gross total resection using
minimally invasive techniques through a left frontal trans-sulcal approach. Preoperative
(a) axial T2-weighted fluid-attenuated inversion recovery (FLAIR) and (b) coronal and (c) sagittal T1-weighted magnetic resonance imaging (MRI) with contrast. Postoperative
(d) axial T2-weighted FLAIR and (e) coronal T1-weighted MRI with contrast.
Fig. 2 Left thalamic non-small lung metastasis. This 64-year-old man presented with right
hemiparesis and underwent gross total resection using minimally invasive techniques
through a left superior parietal trans-sulcal approach. Preoperative (a) T2-weighted fluid-attenuated inversion recovery (FLAIR) and (b) axial diffusion tensor imaging; (c) postoperative T2-weighted FLAIR with motion artifact; (d) intraoperative view through the exoscope, and (e) the view through the tubular retractor after gross total resection of the lesion.
After the craniotomy, the dura was opened in a cruciate fashion. The sulcus was then
opened sharply under exoscopic visualization and magnification (0- or 90-degree exoscope;
Vitom, Karl Storz, El Segundo, California, United States). After the sulcus was opened
to its depth, a preselected tubular retractor (NICO BrainPath, Indianapolis, Indiana,
United States) was then passed to the superficial surface of the lesion with the aid
of stereotactic navigation guidance. For smaller and/or deeper lesions, ultrasound
(Hitachi Aloka, Wallingford, Connecticut, United States) was used to help with trajectory
planning and lesion targeting. Once the superficial aspect of the lesion was reached,
the tubular retractor was either held manually or fixated with a Leyla retractor system
(B. Braun Medical Inc., Bethlehem, Pennsylvania, United States).
A pseudo en bloc resection was then performed involving internal debulking with the
use of a combination of tumor forceps, suction, tissue-biting device (Myriad, NICO)
and/or bipolar cautery, followed by delivering the capsule of the tumor circumferentially
in one piece. During internal debulking, the tumor cells were confined to the tubular
retractor that was docked onto the tumor surface to minimize potential disseminated
disease. In cases of larger tumors, the tubular retractor was manipulated and toggled
to access the entire tumor. After resection, the retractor was withdrawn, and the
dura, bone, and skin were closed in standard fashion. A postoperative MRI was done
within 48 hours. The patient then underwent postoperative stereotactic radiosurgery
within 3 weeks of the surgery, with the radiation focused on the resection cavity.
Results
Fifteen consecutive patients underwent surgical resection of a deep-seated metastatic
brain tumor during the reviewed period ([Table 1]). The average ± standard deviation age was 63 ± 12 years. Seven (47%) were female.
Tumors were located in the centrum semiovale in seven (47%), basal ganglia in three
(20%), thalamus in two (13%), and cerebellum in three (20%). Among the lesions located
in the centrum semiovale, three involved the corticospinal tract, two the superior
longitudinal fasciculus, one the visual tract, and one the inferior frontal occipital
fasciculus. The median preoperative KPS was 60 (IQR: 50–90). The most common presenting
symptoms were headaches in seven (47%), confusion/mental status changes in 6 (40%),
weakness in five (33%), and seizures in two (13%). The primary tumor was a non–small
cell lung cancer in eight (53%), bladder cancer in in two (13%), gastrointestinal
cancer in one (7%), melanoma in one (7%), renal cell cancer in one (7%), and small
cell lung cancer in one (7%). For the patient with the small cell lung cancer, biopsy
of the lung lesion revealed mixed adenoma and small cell carcinoma.
Table 1
Patients who underwent resection of deep-seated brain metastasis using minimally invasive
approaches and techniques
|
Patient no.
|
Location
|
Presenting symptoms
|
Pathology
|
Resection, %
|
Preop versus postop KPS
|
Length of hospital stay, d
|
|
1
|
Centrum semiovale (SLF)
|
Seizures
|
GI
|
100
|
Stable
|
5
|
|
2
|
Centrum semiovale (visual tract)
|
Seizures
|
NSCLC
|
100
|
Stable
|
4
|
|
3
|
Basal ganglia
|
Headaches, personality change
|
NSCLC
|
100
|
Stable
|
5
|
|
4
|
Cerebellar hemisphere
|
Confusion
|
Renal cell
|
100
|
Improved
|
2
|
|
5
|
Cerebellar vermis
|
Lethargy
|
NSCLC
|
100
|
Improved
|
2
|
|
6
|
Centrum semiovale (SLF)
|
Headaches
|
SCLC
|
100
|
Stable
|
3
|
|
7
|
Thalamus
|
Headaches, weakness
|
NSCLC
|
100
|
Worsened
|
2
|
|
8
|
Centrum semiovale (CST)
|
Headaches, weakness
|
NSCLC
|
100
|
Improved
|
2
|
|
9
|
Basal ganglia
|
Headaches, confusion
|
Bladder
|
100
|
Improved
|
2
|
|
10
|
Cerebellar vermis
|
Headaches, confusion
|
Salivary
|
100
|
Stable
|
2
|
|
11
|
Centrum semiovale (CST)
|
Weakness
|
NSCLC
|
95.7
|
Stable
|
2
|
|
12
|
Centrum semiovale (IFOF)
|
Headaches
|
Melanoma
|
100
|
Stable
|
2
|
|
13
|
Basal ganglia
|
Lethargy
|
NSCLC
|
100
|
Improved
|
2
|
|
14
|
Thalamus
|
Weakness
|
NSCLC
|
100
|
Improved
|
4
|
|
15
|
Centrum semiovale (CST)
|
Weakness
|
Bladder cancer
|
100
|
improved
|
3
|
Abbreviations: CTS, corticospinal tract; GI, gastrointestinal; IFOF, inferior frontal
occipital fasciculus; NSCLC, non–small cell lung cancer; SCLC, small cell lung cancer;
SLF, superior longitudinal fasciculus.
The median percentage tumor resection was 100% (IQR: 100–100%). Only one patient underwent
subtotal resection due to a decline in motor-evoked potentials during the case. Following
surgery, the median KPS was 80 (IQR: 70–90); seven (47%) had improved KPS, seven (47%)
had stable KPS, and one (7%) had worse KPS. The patient with worse KPS had worsened
leg weakness secondary to involvement in the internal capsule of the thalamic lesion.
However, at the postoperative visit, the leg weakness had nearly resolved. No patients
had notable postoperative hemorrhages, strokes, seizures, wound infections, and/or
other local complications. One (7%) had a new postoperative deep vein thrombosis requiring
anticoagulation. The median length of stay following surgery was 2 days (IQR: 2–3.5
days). All patients underwent postoperative SRS. The median follow-up time was 6 months
(IQR: 2–12 months). No patients had carcinomatosis as of most recent follow-up.
Discussion
In this prospective consecutive case series, 15 patients with deep-seated metastatic
brain tumors involving eloquent cortical and subcortical structures underwent resection
using a minimally invasive approach consisting of tubular retractors and exoscopic
visualization. Although the lesions in this series were in eloquent or deep-seated
locations, 14 (93%) had improved or stable neurologic function after surgery; 14 (93%)
underwent gross total resection. No significant local complications including strokes,
hemorrhages, and infections occurred, and all patients were functional enough to undergo
postoperative SRS. This is the largest individual case series evaluating the use of
this approach for deep-seated metastatic brain tumors.
There are ∼ 200,000 new metastatic brain tumor cases in the United States each year,
which is ∼ 10 times more prevalent than gliomas.[1]
[2]
[3]
[4]
[5] Approximately 30 to 70% of patients with primary cancers develop metastatic brain
tumors.[1]
[2]
[3]
[4]
[5] This number is expected to rise as more patients develop primary cancers, imaging
technology becomes more widely available, and the lack of progress continues in finding
ways to prevent cancers from metastasizing to the brain.[1]
[2]
[3]
[4]
[5] When patients develop metastatic brain tumors, the treatment typically involves
some combination of surgery and/or radiation therapy.[1]
[2]
[3]
[4]
[5] This management is partially dictated by their RPA class, consisting of age, KPS,
primary tumor control, and the presence of systemic disease.[1]
[2]
[3]
[4]
[5] Patients in RPA class I (age < 65 years, KPS ≥ 70, controlled primary cancer, and
no extracranial metastases) and many in class II (age > 65 years, KPS ≥ 70, uncontrolled
primary cancer and/or extracranial metastases) are typically offered surgery if their
lesion is of significant size causing mass effect is surgically accessible and causing
symptoms.[1]
[2]
[3]
[4]
[5] Most surgeons define accessibility as superficially located.[1]
[2]
[3]
[4]
[5] For patients with lesions not meeting these criteria, they are typically offered
nonsurgical and/or palliative options including radiation and/or steroid therapy.[1]
[2]
[3]
[4]
[5]
Deep-seated lesions are surgically challenging and therefore typically treated with
palliative options because of their relative inaccessibility.[10]
[11]
[12] Moreover, these deep-seated lesions are typically associated with significant mass
effect and/or neurologic deficit because they are near the confluence of the ventricular
system (i.e., third ventricle), condensed white matter tracts (i.e., internal capsule),
and eloquent gray matter nuclei in close proximity to each other (i.e., basal ganglia
and thalamus), and therefore they usually present with a poor KPS as was the case
in this series. The conventional approach for managing deep-seated brain tumors is
to perform a standard needle biopsy to obtain tissue for diagnosis when diagnosis
is unclear, followed by radiation therapy.[1]
[2]
[3]
[4]
[5] This is because the standard surgical approach involves a large craniotomy, extensive
white matter dissection, and then resection, associated with significant morbidity.[9]
[10]
[12] Because most metastatic brain tumors undergo postoperative radiation therapy, the
risk of neurologic injury can delay radiation therapy.[1]
[2]
[3]
[4]
[5] Therefore, to minimize potential morbidity with these high-risk deep-seated lesions,
most surgeons do not pursue surgical resection.[1]
[2]
[3]
[4]
[5] This withholding of surgical resection means the potential benefits of surgery cannot
be realized that include surgical debulking to make adjuvant therapy more effective,
relief of mass effect, symptomatic improvement, and/or more tissue for molecular analyses.[10]
[12]
Minimally invasive approaches are being used for a variety of intracranial lesions.[8]
[10]
[11]
[12]
[14]
[16]
[17]
[18]
[19] Tubular retractors have been used for decades.[8]
[10]
[11]
[12]
[14]
[16]
[17]
[18]
[19] In the 1980s, Kelly used metal tubes of different sizes attached to the stereotactic
frame to access thalamic tumors with the aid of computer-assisted resection.[13]
[20] These early channel-based retractors were then developed into transparent ovular
retractors with inner stylets to resect pathology in deep-seated locations that included
intraventricular lesions such as colloid cysts, high-grade gliomas, and abscesses,
among others.[17]
[18] These larger retractors were typically limited to transcortical approaches through
noneloquent cortical regions such as the anterior frontal lobe.[13]
[16]
[17]
[18]
[21] Circular instead of ovular retractors were then developed that allowed transsulcal
access, and therefore approaches could be used in eloquent cortical regions.[10]
[11]
[12]
[14] These circular retractors were initially used to access deep-seated intracranial
hemorrhages involving the basal ganglia, insula, and thalamus,[14] and later applied for excisional biopsies and resections of deep-seated gliomas[10]
[11]
[12] and intraventricular lesions.[19] The application of this technique in a strictly metastatic brain tumor population
is limited.[8]
[16]
Previous studies using minimally invasive approaches for metastatic brain tumors are
few and limited.[15]
[16]
[22] Hong and colleagues demonstrated the use of minimally invasive methods for 20 patients
with deep-seated lesions; 8 had metastatic brain tumors.[16] Day described the same approach in 49 patients; 23 had gliomas, 20 had metastatic
tumors, and 6 had intracranial hemorrhages.[15] In this study, five (10%) had a new deficit, one had a stroke (2%), and one (2%)
died.[10]
[16] More recently, Bakhsheshian and colleagues demonstrated the use of this approach
for metastatic brain tumors for 25 patients from six centers over a 6-year period.
Gross total resection was achieved in 80%, with one patient each (4% each) developing
new weakness and a deep vein thrombosis.[8]
This is the first study that focuses on a single surgeon's use of a minimally invasive
approach for deep-seated metastatic brain tumors. This is important because the management
and surgical approach for these lesions differ than other pathology including gliomas,
abscesses, and demyelinating disorders, among others. This study also studies the
application of this approach via a transsulcal route to minimize potential injury
to eloquent cortical and subcortical tracts. Prior studies primarily used a transcortical
route.[13]
[16]
[18]
[21] Lastly, this study also used the most modern optics with exoscopic visualization
to maximize optics and ergonomics, whereas other earlier studies used primarily microscopic
visualization.[13]
[16]
[18]
[21] The exoscope is similar to an endoscope, but instead of being inside the body cavity,
it hovers outside of the body similar to a microscope. The advantages of an exoscope,
as compared with a microscope, is that it provides greater magnification, has wider
focal lengths, and offers better ergonomics. Nonetheless, the surgeries in this series
could have been done with a microscope without the advantages just described.
This study, however, has some limitations. The metastatic lesions in this study are
diverse and therefore not applicable to one type.[3] It is well known that the lesion's consistency varies with the primary pathology.[3] Moreover, the number in this study limits its generalizability even though this
is the largest single surgeon's experience in the literature. A significant portion
of patients in this study presented with a poor KPS, and therefore RPA class 3 was
likely because of the deep-seated location of their brain metastases and therefore
represents the poorest prognostic subset of patients with metastatic brain cancer.
Additionally, the use of tubular retractors is limited to deep-seated locations below
the sulcal boundaries in eloquent locations. It does not apply to more superficial
lesions and those located in noneloquent areas. Nevertheless, we believe this study
documents the applicability of this approach to metastatic tumors arguably in the
most dangerous locations by being deep in the brain in eloquent locations involving
critical gray matter structures and white matter tracts.
Conclusions
A subset of metastatic brain tumors occurs in deep-seated locations. Historically
patients with these lesions were only offered palliative options including stereotactic
needle biopsy, radiation therapy, and medical therapy because aggressive surgical
resection was associated with significant morbidity. We demonstrate in the largest
single surgeon's experience the use of tubular retractors via transsulcal approaches
to deep-seated metastatic brain tumors with exoscopic visualization. This approach
can be used to achieve extensive resection with minimal morbidity for arguably the
highest risk metastatic brain tumors.
