See the corresponding editorial in this issue, pp 1207–1208.
J Neurosurg 111:1209–1215, 2009
Deep brain stimulation of the subcallosal cingulate gyrus for
depression: anatomical location of active contacts in clinical
responders and a suggested guideline for targeting
Clinical article
Clement Hamani, M.D., Ph.D.,1 Helen Mayberg, M.D., 2 Brian Snyder, M.D.,1
Peter Giacobbe, M.D., 3 Sidney Kennedy, M.D., 3 and Andres M. Lozano, M.D., Ph.D.1
Division of Neurosurgery, Toronto Western Hospital, and 3Department of Psychiatry, Toronto General
Hospital, University Health Network, Toronto, Ontario, Canada; and 2Departments of Psychiatry and
Neurology, Emory University School of Medicine, Atlanta, Georgia
1
Object. Deep brain stimulation (DBS) of the subcallosal cingulate gyrus (SCG), including Brodmann area 25, is
currently being investigated for the treatment of major depressive disorder (MDD). As a potential emerging therapy,
optimal target selection within the SCG has still to be determined. The authors compared the location of the electrode
contacts in responders and nonresponders to DBS of the SCG and correlated the results with clinical outcome to help
in identifying the optimal target within the region. Based on the location of the active contacts used for long-term
stimulation in responders, the authors suggest a standardized method of targeting the SCG in patients with MDD.
Methods. Postoperative MR imaging studies of 20 patients with MDD treated with DBS of the SCG were analyzed. The authors assessed the location of the active contacts relative to the midcommissural point and in relation
to anatomical landmarks within the medial aspect of the frontal lobe. For this, a grid with 2 main lines was designed,
with 1 line in the anterior-posterior and 1 line in the dorsal-ventral axis. Each of these lines was divided into 100 units,
and data were converted into percentages. The anterior-posterior line extended from the anterior commissure (AC) to
the projection of the anterior aspect of the corpus callosum (CCa). The dorsal-ventral line extended from the inferior
portion of the CC (CCi) to the most ventral aspect of the frontal lobe (abbreviated “Fr” for the formula).
Results. Because the surgical technique did not vary across patients, differences in stereotactic coordinates
between responders and nonresponders did not exceed 1.5 mm in any axis (x, y, or z). In patients who responded to
the procedure, contacts used for long-term stimulation were in close approximation within the SCG. In the anteriorposterior line, these contacts were located within a 73.2 ± 7.7 percentile distance from the AC (with the AC center
being 0% and the line crossing the CCa being 100%). In the dorsal-ventral line, active contacts in responders were
located within a 26.2 ± 13.8 percentile distance from the CCi (with the CCi edge being 0% and the Fr inferior limit
being 100%). In the medial-lateral plane, most electrode tips were in the transition between the gray and white matter
of SCG.
Conclusions. Active contacts in patients who responded to DBS were relatively clustered within the SCG. Because of the anatomical variability in the size and shape of the SCG, the authors developed a method to standardize
the targeting of this region. (DOI: 10.3171/2008.10.JNS08763)
O
Key Words • Brodmann area 25 • cingulate gyrus • cingulotomy •
deep brain stimulation • depression • psychiatry
of the most promising emerging indications
for DBS is for the treatment of MDD.5,9,12 Several
targets are being considered as of this writing, including the SCG (a region that encompasses Brodmann
area 25),9 the inferior thalamic peduncle,5 the nucleus acne
Abbreviations used in this paper: AC = anterior commissure; CC
= corpus callosum; CCa = anterior aspect of the CC; CCi = inferior
portion of the CC; DBS = deep brain stimulation; HAMD-17 =
Hamilton depression scale; MCP = midcommissural point; MDD
= major depressive disorder; PC = posterior commissure; SCG
= subcallosal cingulate gyrus; SPGR = spoiled gradient–recalled
acquisition.
J Neurosurg / Volume 111 / December 2009
cumbens,12 and the internal capsule. Our group has focused on the SCG, based on the premise that DBS could
modulate the increased activity observed in this region in
patients with depression.2,3,8–10,14 In our initial trial, 4 of 6
patients receiving DBS of the SCG responded to the procedure at 6 months. We have subsequently treated 14 additional patients. At 1 year, 55% of the 20 patients treated
in our center were considered to be responders, based on
a ≥ 50% improvement in their depressive symptoms.7
Because DBS of the SCG is an emerging therapy for
MDD, prognostic factors and the optimal placement of
the electrodes to achieve a good outcome have not yet
been established. In our first study, surgical targeting was
1209
C. Hamani et al.
accomplished through direct MR imaging visualization
of the subgenual cingulum. We did not rely on coordinates relative to the MCP due to the significant variability in the anatomy of the medial prefrontal cortex across
patients. One problem with targets derived directly from
imaging, however, is the potential variation that may occur across centers. As a result, we created a standardized
way of targeting the SCG. For this, we initially subdivided
our patients into 2 groups: responders and nonresponders
to therapy. Subsequently, we assessed the location of the
contacts used for long-term stimulation relative to the internal anatomy of the medial frontal lobe in patients who
responded to DBS surgery. Based on these findings, we
designed a grid to standardize the targeting of the SCG
relative to internal landmarks of the medial frontal lobe.
Methods
Patient Population
Twenty patients with MDD treated with SCG DBS at
the Toronto Western Hospital were included in this study.
To be considered a surgical candidate, the following criteria had to be met:7,9 1) a diagnosis of MDD according
to the Diagnostic and Statistical Manual (ed. IV-TR); 2)
current major depressive episode persisting for at least 1
year; 3) HAMD-17 scores4 > 20 and Global Assessment
of Function scores ≤ 50; and 4) failure to respond to 4
different antidepressant therapies, including medications
and evidence-based psychotherapy. In 17 of the patients,
electroconvulsive therapy had also failed. Exclusion criteria included suicidal plans during the recruitment phase,
other Axis I or II disorders, or neurological or clinical
conditions that could interfere with the technique or safety of the surgical procedure (that is, coagulopathy).
A significant response to DBS was considered to
have occurred when HAMD-17 scores were reduced by >
50% at 12 months. According to this criterion, 11 patients
(55%) were considered to be responders and 9 (45%) were
nonresponders to the therapy (Table 1).
Surgical Technique and Programming
A stereotactic frame (Leksell model G, Elekta, Inc.)
was placed after induction of local anesthesia, and axial
3D SPGR and T2-weighted stereotactic MR imaging sequences were acquired (Signa 1.5-T unit, General Electric, Inc.). Images were then uploaded onto a neuronavigation system with frame-based software (Framelink 4.1,
StealthStation, Medtronic, Inc.) for anatomical targeting.
The SCG was initially identified on reconstructed
sagittal images, representing the cingulate region lying
ventral to the genu of the CC (Fig. 1A). This often corresponded to a coronal section in which the initial aspect of
the anterior horns of the lateral ventricles could be appreciated (Fig. 1B). In the medial-lateral plane, the selected
target was the transition between gray and white matter
of the SCG (Fig. 1B and C). In a few cases, the region
was characterized by more complex or bifid gyri (Fig.
1B). Under these circumstances, we preferred to target
1210
the more ventral aspect of the SCG to permit maximum
electrode coverage of the region.
In the operating room, after induction of local anesthesia, bilateral bur holes were made 1–2 cm anterior to
the coronal suture and 2 cm lateral to the midline. Microelectrode recording commenced 10 mm above the target,
and was used mainly to localize the junction between
gray and white matter. This could be clearly ascertained,
because the gray matter was characterized by neuronal
activity, whereas the white matter was electrophysiologically silent.
Once the physiological target was determined, DBS
quadripolar electrodes (model 3387, Medtronic, Inc.)
were implanted in this region. Individual contacts were
tested consecutively to evaluate acute effects of stimulation,9 and were subsequently secured in place. Thereafter,
the head frame was removed and a dual-channel internal
pulse generator (Kinetra, Medtronic, Inc.) was implanted
in the right subclavicular region after induction of general
anesthesia.
Programming of the internal pulse generator was
started ~ 2 weeks postsurgery. Initial settings were a
mono­polar configuration (contact cathode and case anode) 2.5–3.5 V, 90 μsec of pulse width, and 130 Hz. The
contacts initially selected for stimulation were the ones
closer to the SCG, as assessed on postoperative MR imaging studies. Most commonly these included contacts 1
and 5 and 2 and 6 (0 and 4 were the most ventral contacts; 0–3 were right hemisphere and 4–7 were left hemisphere). Patients were then seen on a weekly basis and the
voltage was optimized. If clinical improvement was not
seen after a few weeks of stimulation, a new pair of homologous contacts was tested (for example, 0 and 4, 1 and
5, 2 and 6, or 3 and 7). Once the best clinically effective
contact was chosen, patients were reassessed only during regular follow-up visits. At long-term follow-up, most
patients were using monopolar stimulation at 3.0–5.0 V,
90 μsec of pulse width, and 130 Hz, with either contacts
1 and 5 or 2 and 6 serving as the cathodes and the case
as the anode.
Location of the Active Contacts Through MR Imaging
To locate the active contacts in each patient, postoperative 3D SPGR and T2-weighted axial MR images
were transferred to a workstation. Using the FrameLink
4.1 software (Mach 4.1, StealthStation, Medtronic, Inc.)
these 2 studies were fused, and coronal and sagittal
planes were reconstructed based on axial images. The AC
and PC were targeted in the axial plane and 3 additional
points were plotted in the midline. Images were then reformatted parallel to the AC-PC plane and orthogonal to
the midline. Pitch, roll, and yaw were corrected in the
StealthStation. The DBS electrodes were visualized in all
3 planes, and their tips were targeted. For this study, we
defined the tip of the electrode as the distal portion of the
electrode artifact. Even though this does not represent the
actual tip of the device, it was a standard measurement
that could be compared across the groups being studied.
Thereafter, we determined the entry point of the leads
into the cortex, the location of the tip of the electrodes,
and the distance between the MR imaging artifacts of
J Neurosurg / Volume 111 / December 2009
Targeting the subcallosal cingulate gyrus for DBS
TABLE 1: Clinical outcome in patients with major depression who responded or did not respond to DBS of the SCG*
HAMD-17 Score
Subgroup
responders
nonresponders
all
No. of Patients
Preop
Postop (1-yr FU)
% Improvement
11
9
20
24.3 ± 3.7
24.1 ± 3.2
24.2 ± 3.4
8.2 ± 2.5
18.0 ± 4.8
12.6 ± 6.2
66.4 ± 9.0
26.1 ± 13.8
48.2 ± 23.2
* Results are expressed as the mean ± SD. Abbreviation: FU = follow-up.
each of the contacts. For the purpose of our study, we
have considered the center of the sphere-shaped artifacts
corresponding to the electrode contacts to be the center of
the contacts.11,13 In this study, active contacts were defined
as those being used as the cathodes, because all patients
were receiving monopolar stimulation.
Coordinates of the contacts and tips of the electrodes
were derived from the MCP and the medial-lateral edge
of the cortical surface (x, medial-lateral plane; y, anteriorposterior plane; and z, dorsal-ventral plane). Coordinates
from the cortical surface were calculated because in some
cases we observed asymmetries in the region of the subgenual cingulum, with one of the hemispheres crossing
over the midline to the other side. In addition, we have
also calculated the euclidean distance from the active
contacts to the center of the AC. In a 3D “xyz” system,
with p1 being AC (x1, y1, z1) and p2 being the center of the
active contacts (x2, y2, z2), the euclidean distance between
p1 and p2 was represented by the �[(x1 − x2)2 + (y1 − y2)2 +
(z1 − z2)2].1
Due to the variability in the anatomy of the prefrontal cortex, we not only calculated the distance of the contacts to the MCP but also relative to internal anatomical
landmarks. To standardize the location of the electrodes
across patients, we have initially traced a projection line
abutting the most rostral aspect of the CC (the CC line)
that was perpendicular to the base of the frontal lobe (Fig.
2). Thereafter, we drew 2 straight lines perpendicular to
the projection line described above. The first connected
the center of AC and the CC line and was called the ACCCa line. The second connected the center of AC to the
center of the active contact (AC-Ct). The percentage obtained with the formula AC-Ct/AC-CCa × 100 represent-
ed the relative distance of the center of the active contact
from AC and CCa. In the dorsal-ventral plane, a line perpendicular to the base of the frontal lobe was drawn from
the CCi to the edge of the frontal lobe (abbreviated “Fr”
for the formula). A second line was then drawn parallel
to CCi-Fr from the inferior aspect of CC to the center of
the active contact (CCi-Ct). The percentage obtained with
the formula CCi-Ct/CCi-Fr × 100 represented the relative
distance of the center of the active contact to CCi and Fr.
Statistical Analysis
A Student t-test was used to compare the following
variables: 1) the location of the contacts and tip of the
electrodes relative to the MCP and medial-lateral edge of
the cortical surface; 2) the location of the contacts relative to the AC-CCa and CCi-Fr lines; and 3) the euclidean
distance between AC and the active contacts. Correlation
analysis was used to assess whether there was a relationship between the location of the active contacts and outcome.
For the Student t-test, a probability value < 0.05 was
considered statistically significant. A correlation was
considered to be strong when r ≥ 0.75, moderate when r ≥
0.5, weak when r ≥ 0.25, and nonexistent when r < 0.25.
All values in the text are expressed as the mean ± SD.
Results
Electrode Location in Responders and Nonresponders to
DBS of the SCG
A summary of our results may be found in Table 2.
Because no differences in electrode placement were no-
Fig. 1. The SCG as visualized on MR imaging. A: Sagittal T2-weighted section showing the region of the cingulate gyrus
that lies underneath the genu of the CC (arrow). B: Coronal T2-weighted section in which the most rostral aspect of the caudate nucleus can be appreciated. The black arrow points to the transition between gray and white matter of the SCG; the white
arrowhead depicts a more complex bifid gyrus. C: Axial 3D SPGR imaging section showing the transition between gray and
white matter in the SCG (arrow).
J Neurosurg / Volume 111 / December 2009
1211
C. Hamani et al.
responders to DBS of the SCG were as follows: 6.3 ± 2.0
lateral (x [5.6 ± 1.3 from the edge of the cortical surface]);
34.0 ± 2.7 anterior (y); and 2.6 ± 3.1 inferior (z). The euclidean distance from the center of the active contacts to
the center of AC in these patients was 22.9 ± 2.2.
A schematic representation of the location of the active contacts relative to internal anatomical landmarks in
responders to surgery is presented in Fig. 3B. On average,
these contacts were located in the 73.2 ± 7.7 percentile of
the AC-CCa line in the anterior-posterior axis (0% being
the center of the AC and 100% the line crossing the CCa)
(Fig. 3C). In the dorsal-ventral plane, active contacts in
responders to the therapy were located in the 26.2 ± 13.8
percentile of the CCi-Fr line (0% being the inferior edge
of the CC and 100% the inferior limit of the frontal cortex) (Fig. 3C).
Discussion
Fig. 2. Standardized method used to assess the location of the
electrode contacts in patients who underwent DBS of the SCG. A:
A projection line (dotted line) perpendicular to the base of the frontal
lobe was initially traced abutting the most rostral aspect of the CC (CCa
line). Thereafter, we drew 2 straight lines perpendicular to this projection line. The first connected the center of the AC and the CC line, and
was called the AC-CCa line. The second connected the center of the
AC to the center of the active contact (AC-Ct). B: In the dorsal-ventral
plane, a line perpendicular to the base of the frontal lobe was drawn
from the CCi to the edge of the Fr. A second line was then drawn parallel to the CCi-Fr line from the inferior aspect of CC to the center of the
active contact (CCi-Ct).
ticed when the right and left hemispheres were compared,
data from both sides were grouped for the analysis.
Overall, we found no significant differences in the
location of the electrodes relative to the MCP in responders and nonresponders to DBS of the SCG (Table 2). In
addition, no meaningful correlation (for example, r ≥ ±
0.5) was found between the location of the electrodes and
the percentage of change in HAMD-17 scores for each
patient (Table 3). The only statistically significant difference between the groups was that electrodes in patients
who responded to surgery were in a slightly more ventral
location relative to anatomical landmarks of the medial
prefrontal lobe (Fig. 3A). This difference, however, was in
the order of 1–2 mm and probably of no clinical significance (see Discussion for details).
Location of the Active Contacts in Patients who Responded
to DBS of the SCG
Coordinates of the active contacts relative to MCP in
1212
The main objective of our study was to develop a
standardized way of targeting the SCG. Because the
anatomy of the medial prefrontal lobe is quite variable,
the placement of DBS electrodes based on coordinates
relative to the AC, PC, and MCP is often inadequate. In
our first article, we reported that a suitable location for
the electrodes would be the midpoint between the most
anterior edge of the CC and the AC.8 In the patients who
underwent operations subsequently, however, we realized
that the target selected through direct visualization of the
SCG on MR images was often more anterior than the one
we previously published.
To develop a standardized way of targeting the SCG,
we undertook 2 main steps. The first was to subdivide our
patients into responders (a decrease in HAMD-17 scores
of > 50% at 12 months) and nonresponders to DBS. Subsequently, we assessed the location of the contacts used
for long-term stimulation relative to the internal anatomy
of the medial frontal lobe in patients who responded to
surgery.
In the first part of the analysis, we initially attempted
to assess whether there were any differences in the anterior-posterior location of the electrodes between responders and nonresponders to DBS of the SCG. The premise
was that the placement of electrodes in distinctive regions
along this axis could lead to variable outcomes due to
the recruitment of different fiber systems by stimulation.6
Our results did not corroborate this hypothesis, because
the only difference we found was that the electrode array in nonresponders was significantly more dorsal than
the one in patients who responded to the procedure. This
difference however, was on the order of millimeters and
not likely to be of clinical significance. In addition, active
contacts in responders were relatively clustered within the
SCG and not in more ventral aspects of the frontal lobes.
Because all nonresponders had at least 1 electrode contact placed within the SCG, it is unlikely that the small
dorsal-ventral differences observed in our analysis were
responsible for the lack of clinical response in these patients.
A few caveats of our analysis have to be taken into account. We have only explored a small portion of the subJ Neurosurg / Volume 111 / December 2009
Targeting the subcallosal cingulate gyrus for DBS
TABLE 2: Location of the electrode contacts relative to the MCP and anatomical landmarks within the medial frontal lobe in patients who
did or did not respond to DBS of the SCG*
Electrode Contacts &
Subgroup
0&4
responders
nonresponders
p value
1&5
responders
nonresponders
p value
2&6
responders
nonresponders
p value
3&7
responders
nonresponders
p value
Contact Position ‡
Coordinate†
x
xcx
y
z
ED From AC
AC-ct (%)
CCi-ct (%)
5.6 ± 2.2
6.2 ± 2.3
0.4
5.0 ± 1.3
6.1 ± 1.8
0.06
34.0 ± 2.9
33.1 ± 1.8
0.2
–5.8 ± 3.0
–4.9 ± 3.3
0.3
22.9 ± 4.0
22.2 ± 4.2
0.4
72.3 ± 7.6
71.4 ± 5.6
0.7
42.0 ± 19.1
26.6 ± 29.9
0.07
5.6 ± 3.5
6.7 ± 2.4
0.2
5.6 ± 1.3
6.0 ± 1.7
0.4
34.2 ± 2.6
33.0 ± 1.9
0.1
–3.0 ± 3.1
–1.9 ± 3.3
0.3
22.5 ± 4.6
21.9 ± 4.3
0.4
73.2 ± 7.7
71.1 ± 5.6
0.3
28.8 ± 14.1
15.9 ± 20.3
0.02
6.9 ± 2.0
7.3 ± 2.4
0.6
6.0 ± 1.3
6.7 ± 1.7
0.1
34.4 ± 2.5
33.0 ± 2.0
0.06
–0.1 ± 3.1
0.9 ± 3.4
0.3
21.6 ± 3.3
20.6 ± 3.8
0.4
73.6 ± 8.2
70.8 ± 5.5
0.2
17.2 ± 10.9
6.8 ± 14.2
0.02
7.5 ± 2.0
6.5 ± 5.4
0.5
6.6 ± 1.9
7.2 ± 2.2
0.4
34.5 ± 2.4
33.1 ± 2.1
0.06
2.8 ± 3.1
3.9 ± 3.5
0.3
35.7 ± 3.1
34.0 ± 6.1
0.6
74.3 ± 8.5
71.0 ± 5.6
0.2
5.6 ± 9.3
–5.9 ± 18.5
0.02
* Values represent the mean ± SD. No differences in electrode placement were noticed when the right and left hemispheres were compared, and
therefore, data from both sides were grouped for the analysis. Probability values were obtained with a t-test comparing responders and nonresponders
to treatment. Abbreviation: ED = euclidean distance.
† Definitions of coordinates: x, medial-lateral coordinate of the contacts relative to the MCP; xcx, medial-lateral coordinate of the contacts relative to
the cortical surface; y, anterior-posterior coordinate of the contacts relative to the MCP; z, dorsal-ventral coordinate of the contacts relative to the MCP.
‡ Definitions of contact positions: AC-ct, the anterior-posterior line connecting the AC to the contact; CCi-ct, the dorsal-ventral line connecting the
inferior edge of the CC to the contact (see text for details).
genual cingulum. Almost all contacts in our series were
within 65–80% of the AC-CCa line (0% being the center
of AC and 100% the projection line crossing the CCa). It
is possible that electrodes placed more anteriorly or posteriorly than our general target could yield different results. By moving the electrodes more anteriorly we would
probably affect more fibers from the cingulate bundle,
whereas by moving it more posteriorly we would theoretically affect a more complete set of projections to and
from the SCG (including not only the cingulate bundle
but also those involving the uncinate fascicle as well as fibers innervating subcortical and brainstem structures, the
thalamus, and hypothalamus).6 The problem of targeting
more posterior aspects of the SCG is surgical safety, be-
cause the anterior cerebral arteries are very close to that
region. What we can conclude based on our findings is
that within the small targeted region of the SCG, the location of the electrode contacts did not determine outcome.
Also important to consider for anatomical, and eventually
physiopathological reasons, is whether electrodes were in
the white or gray matter of the SCG. In both responders
and nonresponders, most contacts were located in the border gray/white matter. Moreover, contacts that were not in
this anatomical region were only 1–2 mm into either the
SCG gray or white matter. The only exceptions to this
rule were the upper contacts in some patients, which were
more dorsal and closer to the white matter of the CC.
Prognostic features associated with a good outcome
TABLE 3: Correlation between electrode contacts 1 and 5 and stereotactic coordinates relative to the MCP and
anatomical landmarks within the medial frontal lobe*
Coordinate
Parameter
% improvement in HAMD-17 scores
Lines (%)
x
xcx
y
z
AC-CCa
CCi-Fr
–0.27
–0.18
0.17
–0.12
0.23
0.39
* Values represent the Pearson correlation coefficient (r). We have chosen to demonstrate results of the correlation analysis only
for contacts 1 and 5 for the sake of clarity. Nevertheless, similar r values were obtained when the homologous contacts 0 and 4,
2 and 6, and 3 and 7 were considered. See Table 2 for definitions of the coordinates. Abbreviations: AC-CCa = anterior-posterior
line connecting AC to the anterior edge of the CC; CCi-Fr = dorsal-ventral line connecting the inferior edge of the CC to the base
of the frontal lobe.
J Neurosurg / Volume 111 / December 2009
1213
C. Hamani et al.
Fig. 3. Location of the electrode contacts relative to anatomical landmarks of the medial prefrontal lobe. A: Schematic
representation of the location of the electrode contacts in patients who responded (red circles) and those who did not respond
(blue circles) to DBS of the SCG. Each circle represents the average location of homologous electrode contacts (left and right
hemispheres together) across patients. B: Schematic representation of the active contacts used for long-term stimulation in
patients who responded to DBS of the SCG. Each circle represents 1 electrode contact, and each color represents 1 patient (the
procedure was done bilaterally, so there are 2 circles with the same color for each patient). C: Schematic representation of
the average location of the active contacts in patients who responded to DBS of the SCG. The central circle represents average
results and the outer circle denotes SDs.
after DBS of the SCG are still unclear. Because the location of the electrode contacts was similar in responders
and nonresponders to DBS, one could hypothesize that
differences in outcome could be related to variations in
the clinical characteristics of depression across patients.
Against this assumption, however, baseline total HAMD17 scores and symptom cluster subscores in our series
were similar in responders and nonresponders to surgery.7
Yet, one caveat in this regard is that our study so far has
only included 20 patients. It is possible that future multicenter trials recruiting a larger number of patients may
yield different results. In summary, at present we cannot
ascertain why certain patients did not respond to DBS of
the SCG.
above. One question that may emerge is whether one type
of MR imaging sequence is better than others for targeting. At present, we are mainly using 3D SPGR images for
preoperative targeting. In this study, we have chosen to
illustrate our targeting method only in T2-weighted sections because the contrast between gray and white matter
can be better appreciated.
Finally, we would like to emphasize that the strategy
discussed above only represents a guideline for targeting. Refinements should be conceived and implemented
for each individual patient as the surgeon visualizes the
Proposed Method for Targeting the SCG
Based on the location of the active contacts in patients who responded to surgery in our series, we conceived a more standardized method for targeting the SCG
compared with our initial description.9
The first step is to visualize the subgenual cingulum
on a midline sagittal section (Fig. 4A). Thereafter, a line
parallel to the base of the frontal lobe and a perpendicular line abutting the CCa should be traced. The next step
would be to draw AC-CCa and CCi-Fr lines parallel and
perpendicular to the base of the frontal lobe, respectively.
Both lines should be divided into quartiles (25, 50, 75,
and 100%) (Fig. 4B and C). Axial and/or coronal planes
of a region corresponding to 70–75% in AC-CCa and
25–30% in CCi-Fr should then be selected to define the
anatomical gray/white matter junction of the SCG (Fig.
4D). In most of our cases, this region corresponded to a
coronal plane that was 4–8 mm posterior to the rostral
aspect of the anterior horns of the lateral ventricles. In addition, this plane often corresponded to the last section in
which we did not see basal ganglia structures or in which
a small crescent of caudate nucleus could be appreciated.
According to the results in our 20 initial patients, at least
1 contact should ideally be placed in the region described
1214
Fig. 4. Magnetic resonance images demonstrating the standardized
method for targeting the SCG. A: Sagittal section showing the SCG.
(Panels B and C are magnified images that correspond to the square in
A.) B and C: To target the SCG, a line is initially traced parallel to the
base of the Fr. Thereafter, AC-CCa and CCi-Fr lines are to be drawn
parallel and perpendicular to the base of the frontal lobe, respectively,
and divided into quartiles. D: Axial and/or coronal planes of a region
corresponding to 70–75% in AC-CCa and 25–30% in CCi-Fr should
then be selected to define the anatomical gray/white matter junction
of the SCG.
J Neurosurg / Volume 111 / December 2009
Targeting the subcallosal cingulate gyrus for DBS
region. In summary, this was the first article attempting
to standardize the targeting of the SCG. Our method provides a general guideline and has to be corroborated by
other authors performing psychiatric surgery for depression as well as in future prospective studies.
Conclusions
Because of the anatomical variability in the size and
shape of the medial prefrontal lobe, we developed a standardized method for targeting the SCG. Based on the location of the active contacts in patients who responded
to surgery in our series, we designed a grid that takes
into account internal anatomical landmarks for each individual patient.
Disclosure
Drs. Mayberg, Giacobbe, Lozano, and Hamani are consultants
for, Dr. Ken­nedy receives research support from, and Dr. Mayberg
holds a patent licensed to Advanced Neuromodulation Systems, Inc.
Dr. Lozano owns stock in Medtronic. The authors report no other
conflicts of interest concerning the materials or methods used in this
study or the findings specified in this paper.
References
1. Andrade-Souza YM, Schwalb JM, Hamani C, Eltahawy H,
Hoque T, Saint-Cyr J, et al: Comparison of three methods of
targeting the subthalamic nucleus for chronic stimulation in
Parkinson’s disease. Neurosurgery 56:360–368, 2005
2. Damasio AR, Grabowski TJ, Bechara A, Damasio H, Ponto
LL, Parvizi J, et al: Subcortical and cortical brain activity during the feeling of self-generated emotions. Nat Neurosci 3:
1049–1056, 2000
3. George MS, Ketter TA, Parekh PI, Horwitz B, Herscovitch P,
Post RM: Brain activity during transient sadness and happiness in healthy women. Am J Psychiatry 152:341–351, 1995
4. Hamilton M: A rating scale for depression. J Neurol Neuro­
surg Psychiatry 23:56–62, 1960
5. Jimenez F, Velasco F, Salin-Pascual R, Hernandez JA, Velasco
M, Criales JL, et al: A patient with a resistant major depression disorder treated with deep brain stimulation in the inferior thalamic peduncle. Neurosurgery 57:585–593, 2005
J Neurosurg / Volume 111 / December 2009
6. Johansen-Berg H, Gutman DA, Behrens TE, Matthews PM,
Rush­worth MF, Katz E, et al: Anatomical connectivity of the
subgenual cingulate region targeted with deep brain stim­u­
lation for treatment-resistant depression. Cereb Cortex 18:
1374–1383, 2008
7. Lozano AM, Mayberg HS, Giacobbe P, Hamani C, Craddock
RC, Kennedy SH: Subcallosal cingulate gyrus deep brain
stimulation for treatment-resistant depression. Biol Psychia­
try 64:461–467, 2008
8. Mayberg HS, Liotti M, Brannan SK, McGinnis S, Mahurin
RK, Jerabek PA, et al: Reciprocal limbic-cortical function and
negative mood: converging PET findings in depression and
normal sadness. Am J Psychiatry 156:675–682, 1999
9. Mayberg HS, Lozano AM, Voon V, McNeely HE, Seminowicz D, Hamani C, et al: Deep brain stimulation for treatmentresistant depression. Neuron 45:651–660, 2005
10. Pardo JV, Pardo PJ, Raichle ME: Neural correlates of selfinduced dysphoria. Am J Psychiatry 150:713–719, 1993
11. Pollo C, Villemure JG, Vingerhoets F, Ghika J, Maeder P,
Meuli R: Magnetic resonance artifact induced by the electrode Activa 3389: an in vitro and in vivo study. Acta Neuro­
chir (Wien) 146:161–164, 2004
12. Schlaepfer TE, Cohen MX, Frick C, Kosel M, Brodesser D,
Axmacher N, et al: Deep brain stimulation to reward circuitry
alleviates anhedonia in refractory major depression. Neuro­
psy­chopharmacology 33:368–377, 2008
13. Starr PA, Turner RS, Rau G, Lindsey N, Heath S, Volz M,
et al: Microelectrode-guided implantation of deep brain stim­
ulators into the globus pallidus internus for dystonia: techniques, electrode locations, and outcomes. J Neurosurg 104:
488–501, 2006
14. Talbot PS, Cooper SJ: Anterior cingulate and subgenual prefrontal blood flow changes following tryptophan depletion
in healthy males. Neuropsychopharmacology 31:1757–1767,
2006
Manuscript submitted June 19, 2008.
Accepted October 22, 2008.
Please include this information when citing this paper: published
online May 29, 2009; DOI: 10.3171/2008.10.JNS08763.
Address correspondence to: Andres M. Lozano, M.D., Ph.D.,
Di­vis­ion of Neurosurgery, Toronto Western Hospital, 399 Bathurst
Street, WW 4-447 Toronto, Ontario M5T2S8 Canada. email: lozano
@uhnres.utoronto.ca.
1215