Page 27 - Layout 1
P. 27
MILITARY
MEDICINE
by imaging in individual patients. The RCT described here ap- 2012). Based on these imaging observations, the authors (Fox et
plied meta-analytic connectomic imaging (meta-connectomics) al., 2004) demonstrated that TMS efficacy — its ability to mod-
to create a group-wise (population) model of the network being ulate neurons locally and remotely — is jointly determined (as a
treated; it then invidualizes this model, using structural and func- vector dot product) by E-field orientation and columnar orienta-
tional MRI acquired per patient. tion of neurons in the stimulated cortex. Optimal orientation
(maximal stimulation efficacy) is with the E-field column aligned;
TMS Navigation the least effective orientation is 90° from column aligned. This
The vast majority of TMS clinical trials have not employed principal — termed the cortical column cosine (C3) principal —
was confirmed in numerous studies by the authors (Fox et al.,
image-guided neuronavigation despite that facts that: 1) the orig- 2006; Salinas et al., 2007; Rabago et al. 2009, Narayana et al.,
inal motivation to target dorsolateral prefrontal cortex (DLPFC) 2012) and independently (Sommer et al., 2012; Krieg et al. 2013)
was based on neuroimaging studies in MDD (George et al., and is now generally accepted. The most precise and reliable man-
1997); and, 2) numerous trials have acknowledged the need to ner in which to implement image-guided TMS is with a purpose-
accommodate individual differences in structural and functional built robot (Lancaster et al, 2004), a patented implementation
neuroanatomy and have recommended imaging as the means to (Fox & Lancaster, 2006, 2010, 2015). To this end, the Research
this end (e.g., O’Reardon et al, 2007). In a meta-analytic review Imaging Institute (RII) at UT Health San Antonio UTHSCSA
of TMS trials in psychiatric disorders, Slotema (et al., 2010) ar- has developed three generations of image-guided robotic systems
gued that a large portion of inter-subject and inter-trial variability for TMS delivery, with a 4th generation in development. The 3rd
was attributable to: 1) inadequate neuroanatomical specification generation system (Figure 1) is being used in the clinical trial now
of treatment targets; and, 2) delivery imprecision, i.e. inadequate in progress.
neuronavigation. Growing awareness that TMS effects are multi-
regional and network based has lead to the recommendation that DLPFC is a large, functionally heterogeneous region with an
neuronavigation be implemented based on connectomic princi- antero/medial-to-postero/lateral functional gradient (Koechlin et
ples, i.e. based on the network properties of the brain regions al., 2003; Cieslik et al., 2012). Meta-analytic connectivity based
stimulated (Fox et al., 2012; Fox et al., 201 4; Clark et al., 2015). parcellation (CBP) of DLPFC (Figure 2A) identified two discrete
The PI and colleagues have applied neuroimaging methods in subregions: an antero/medial zone projecting to anterior cingulate
conjunction with acute and chronic TMS delivery for almost 20 cortex (ACC) and subgenual cingulate (SGC) and a postero/lat-
years and have applied this knowledge base to develop the TMS- eral region projecting to intraparietal cortex and mid-cingulate
specific neuronavigation method used in the RCT described here. motor areas (Cieslik et al., 2012). Applying this finding to per-
subject resting-state fMRI data, we use the ACC and SGC sites
Combining TMS with neuroimaging demonstrates that TMS jointly as seeds to identify the anterior subregion of DLPFC as a
effects are precisely multi-focal. That is, TMS effects are not dif- regional constraint on treatment planning (Figure 2B). Treatment
fuse local effects, spreading over a large area under the stimulating planning incorporates cortical- and scalp-surface modeling to
coil, as was previously assumed. Rather, they are highly spatially compute orientation and depth-corrected E-field values and iden-
precise at the stimulation site and propagate to remote regions tify optimal coil positioning and intensity settings.
along well-established network connections. The initial
TMS/PET study unequivocally demonstrated this property of Study Design And Rationale
precise multi-focality (Fox et al., 1997) and these results have The ongoing trial uses a 2-arm randomized, double-blind de-
been replicated and extended in numerous subsequent reports
(Fox et al., 2004, 2006; Lancaster et al., 2004; Laird et al., 2008; sign to determine the efficacy of rTMS to right DLPFC as an
Salinas et al., 2007, 2009, 2011, 2013; 2016; Narayana et al., add-on to inpatient treatment as usual for PTSD. All patients in
continued on page 28
visit us at www.bcms.org 27
