The Ictal SPECT Neuroimaging Program within the RUSH Epilepsy Center was established in 2002 by Dr. Marvin Rossi. A span of a decade prior was spent harnessing drugs labelled with radioactive isotopes for studying rodent models of epilepsy, and age-related memory loss (Rossi et al, 2005).
Historically, autoradiography is a technique that laid the foundation for modern nuclear medicine imaging in living animals, including humans.
An autoradiograph of a cross section of rat brain is shown here. This image is actually an imprint on an x-ray film of radioactive drug-labelled
binding sites distributed throughout the brain.
So, the film or emulsion is apposed to the radioligand-labelled tissue section to obtain the autoradiograph.
The colors represent different concentrations of brain receptors with an affinity for a specific radioisotope-labelled drug (in this case, tritium is the radioisotope, also known by the designation [3H]).
The autoradiograph to the left demonstrates [3H]-pirenzepine binding to acetylcholine receptors.
Numerous radioligands have been employed in our lab as internal beacons adhering to brain cell machinery implicated in epilepsy and memory impairment.
Visualization of intact tissue autoradiography requires that the tissue containing the radioligand-receptor complexes is subtracted from a background tissue comparison, where non-specific (also known as non-saturable) labelling of the radioligand has occurred (the radioligand can also adhere non-specifically to other binding sites). The 'auto-' prefix indicates that the radioactive emissions emanate from the drug molecule WITHIN the brain section itself. This technique is distinguished from x-ray imaging of the body, such as in a CT scanner, where an EXTERNAL radioactive source is employed.
Autoradiography may be used to determine the tissue localization, concentration and affinity of specific brain receptor binding sites. Examples of such binding sites are found on the brain cell surface, inside the cell, in metabolic pathways, or within blood vessels. This autoradiographic approach contrasts to techniques, such as positron emission tomography (PET) and single positron emission computed tomography (SPECT), ONLY IN THAT a calculation is required for determining the exact 3-dimensional localization of emissions (photons) from an INTERNAL radiation source. This calculation technique involves careful use of what is called 'coincidence counting'. In other words, the photon detectors from the PET or SPECT cameras provide information about the distribution of radioactivity emitted from inside of the patient's body.
Ictal SPECT has gained considerable momentum over the last decade as a non-invasive functional imaging tool. This powerful scanning tool is capable of mapping the location of the seizure-onset by imaging transient blood flow changes. Blood flow changes must be captured during the early phase of a focal-onset seizure. A small dose of a radioactive isotope-labelled drug, also known as a radiotracer (usually Technetium-99 labelled HMPAO (Ceretec) or ECD (Neurolite)) is injected intravenously within seconds of the seizure-onset. These radiotracers are fat soluble substances that cross the blood brain barrier and have a long retention time in the brain. These tracers are retained in cells after an enzymatic conversion to charged compounds. The chemical conversions prevent the radiotracer from 'back diffusing' out of the brain. The photon emissions from these agents produce a photograph or static snapshot of brain function within a timeframe of about a minute. So, no blood flow changes will be captured even if the patient has recurrent seizures beyond several minutes after injection of the radiotracer. Both Ceretec (99mTc-HMPAO), and Neurolite (99mTc-ECD) are stable between 3 to 6 hours.
The development of SPECT post-processing techniques (that is, analyses performed after scanning), has complemented standard SPECT scanners (also called gamma cameras) developed in the 1980s. SISCOM is a post-processing technique originally innovated at the Mayo Clinic. SISCOM analysis is similar to that performed for autoradiography. A baseline blood flow SPECT (captured when the patient is not seizing for about 24 hours) is subtracted from the seizure-onset SPECT study. Since the injection of baseline and ictal (seizure-onset) SPECT studies with radiotracer may occur at different time points of isotope decay, the intensities of each imaging study are often different. So, signal normalization is performed. A digital subtraction of the two normalized SPECT studies is then completed. This final subtraction study is then matched to the patient's high resolution brain MRI. The likelihood of localizing the seizure-onset zone has been shown to increase to about 80-90% by using the SISCOM technique, compared to about 40% when visually comparing side-by-side ictal and baseline SPECT studies.
In order to reliably visualize the blood flow region associated with the ictal onset zone (also known as the seizure-onset zone), several criteria are essential. The illustration below depicts EARLY regional cerebral blood flow (rCBF) changes (from 0 to 2 minutes) and LATE rCBF changes (beyond 2 minutes) in a temporal lobe seizure compared to its interictal baseline brain SPECT. The radioactive tracer must be injected intravenously very near the seizure-onset. If we inject the radiotracer too late then it is less likely that the seizure-onset region can be identified.
Beginning in 2005, three novel validation blood flow-related neuroimaging strategies were developed in our laboratory. These strategies or techniques are called (1) Ratio Ictal SPECT Co-registered to MRI (RISCOM) (Jain et al, 2014; Rossi et al, 2015), (2) Subtraction Activated SPECT (SAS) (Rossi et al, 2006, 2010), and (3) Dynamic SPECT (dSPECT) Imaging (Rossi et al, 2011; RUSH IRB ORA # 10031201). SAS and dSPECT image processing are performed following implantation of intracranial neurostimulation therapy leads for the purpose of validating targeted direct stimulation influencing the predicted neural circuit that includes the seizure-onset network. These techniques were optimized for a novel portable high-resolution SPECT scanner (Samsung-NeuroLogica, Corp). The unique 'clamshell detectors' housed within the scanner's gantry are shown below.
Historically, mechanically rotating 2 and 3 gamma camera systems have been used to acquire SPECT studies. At one time the RING bore photomultiplier tube(PMT)-based SPECT scanner was also used in the United States. New technologies are reviving this scanner type, with the aim of dramatically improving SPATIAL resolution. This figure from Samsung-Neurologica Corp demonstrates a new 72 PMT system FDA-approved in 2010. This system employs a strategy known as SPIRAL SCANNING FOCUSED COLLIMATION. This technology is dramatically different from the mechanically rotating camera head detectors. This technology can improve SPECT resolution that approaches that of PET (about 3 mm). The resolution of standard SPECT is about 7-12 mm.
Our lab has developed long-term comparison studies at RUSH University Medical Center comparing SPECT scanner technologies (Siemens vs NeuroLogica), and post-processing techniques. These ongoing studies are integrated into our RUSH Epilepsy Center clinical workflow for each patient. Our novel post-processing technique is called, Relative (Ratio) Ictal SPECT-related blood flow changes CO-registered to MRI (RISCOM). Again, SISCOM is a straight subtraction, after image normalization and fitting. In contrast, RISCOM generates a modified ratio of ictal to baseline image voxel data. We are finding that RISCOM often yields more complete information over a wider dynamic range in the imaging signal. So, we consistently demonstrate significantly enhanced imaging of ictal onset-related blood flow changes compared to SISCOM. RISCOM processing using the 72 detector SPECT scanner system (Inspira, Samsung-NeuroLogica) potentially facilitates identifying the seizure-onset in extensive epileptic circuits not visible with the standard 2 detector SPECT scanner (Siemens). More data are required to determine if the extent of resection overlapping early ictal-related RISCOM blood flow changes is associated with treatment outcome post-resection (Jain et al, 2014). Application of the 72-detector SPECT scanner system coupled with RISCOM may help better clarify the extent of the seizure-onset zone for strategic placement of intracranial electrodes for resective surgery.
SAS imaging and its utility are described as follows. An implanted recently FDA-approved direct cortical neurostimulator (NeuroPace, Inc) is used to deliver focal stimulation
without causing a focal seizure, also called an after-discharge. Radiotracer is injected at onset of stimulation to identify grey matter-related transient blood flow
changes both near and distant to the stimulated intracranial electrodes.
White matter pathways are used like 'on ramps' such that axons can propagate electrical current distant from the source of stimulation. This information likely represents the extent of cortical activation for a given set of stimulation parameters passed through electrode contacts placed in human cortex. SAS can be used as a technique to validate presurgical electrode placement planning for responsive direct neurostimulation therapy (NeuroPace, Inc).
Novel dynamic SPECT technology using a 72-detector SPECT scanner system (Samsung-NeuroLogica, Corp) improves upon conventional static SPECT scanner technology by demonstrating near real-time alterations in blood flow during propagation of stimulation current through an epileptic circuit. Dynamic SPECT imaging visualized transient blood flow changes sequentially as 30 second epochs during time-lapsed visualization of propagation of stimulation current activating the neural circuit. In contrast, a static SPECT captures the entirety of the seizure-onset collapsed in time as a single snapshot.
The figure above shows time-lapse imaging of a very thick SPECT slice every 30 sec at the onset of injecting a 5ml radiotracer bolus concurrently with delivering direct stimulation therapy. Dynamic blood flow changes are seen (A) 30 sec during tracer injection while stimulating, (B) 60 sec after injection during ongoing delivery of stimulation therapy, and (C) 90 sec after injection of radiotracer while delivering stimulation therapy without triggering seizure activity. The voxels are brightest, where blood flow related activity is highest (hyper-perfused), the voxels are darker than the background, where the activity is less (hypo-perfused).
See Jalota et al (2015) for details of our SISCOM experience at RUSH University Medical Center.
1. Rossi MA, deToledo-Morrell L, Mash D (2005). Spatial memory in aged rats is related to PKCgamma-dependent
G-protein coupling of the M1 receptor. Neurobiology of Aging 26:53-68.
2. Rossi MA et al (2006).
3. Rossi MA et al (2008).
4. Jain M, Krug K, Balaguera P, Millan C, Jalota A, Pylypyuk V, Byrne R, Rossi MA (2014). Moving Towards New Techniques In The Evaluation Of The Ictal Onset Zone: Ratio Ictal SPECT (RISCOM) Using a 72-Detector Focused Collimator Ring SPECT Scanner System. AES Abstr. 2.241. Seattle WA.
5. Rossi MA, Pylypyuk V, Krug K (2011). Time-lapsed transient brain blood flow changes demonstrated during delivery of direct stimulation therapy through depth leads implanted at the hippocampal grey-white matter junction. AES Abstr. 2.197.
6. Rossi MA, Stebbins G, Murphy C, Greene D, Brinker S, Sarcu D, Tenharmsel A, Stoub T, Stein MA, et al. (2010) Predicting white matter targets for direct neurostimulation therapy. Epilepsy Res 91(2-3):176-186.
7. Jalota A, Rossi MA, Pylypyuk V et al (2015). Resecting Critical Nodes in an Epileptogenic Circuit in Refractory Focal-Onset Epilepsy Using Subtraction Ictal SPECT Co-registered to MRI (SISCOM). Journal of Neurosurgery (accepted).