Feasibility of targeting the cingulate gyrus using high-intensity focused ultrasound on a cadaveric specimen: illustrative case

Abstract

BACKGROUND

Cancer is commonly associated with pain. For patients with advanced cancer and intractable pain, ablative neurosurgical procedures can significantly improve pain and transition patients out of inpatient settings. These procedures are normally invasive, and this poses an important risk in this population. Cingulotomy has been reported to improve pain perception and contribute substantially to the quality of life of cancer patients with refractory pain.

OBSERVATIONS

One fresh human cadaver specimen was used for the setup. The cingulate gyrus was targeted using intraoperative magnetic resonance images, and osseous aberrations were corrected after coregistration with the preoperative head computed tomography.

After accounting for sinuses, membrane folds, and calcifications, a total of 737 elements were available for thermal ultrasound ablation. On high-power sonications, the total energy delivered reached a peak temperature of 57°C (15,050 J, 350 W, 45 seconds) in the right cingulate and 52°C (13,000 J, 405 W, 46 seconds) in the left cingulate.

LESSONS

Despite the limitations of using a cadaver model (temperature, vascularization), cingulotomy appears to be feasible using high-intensity focused ultrasound.

https://thejns.org/doi/10.3171/CASE2459

Cancer is the second leading cause of death in developing countries, and cancer pain has the most disruptive influence on the quality of life of cancer patients.^{1, 2} For many patients, pain is the first sign of cancer, and 30%–50% of all cancer patients will experience moderate to severe pain in their life,³ including up to 75%–95% of patients with metastatic or advanced-stage cancer. Despite increasing interest in and efforts to improve its management, the pain remains poorly controlled in nearly half of all patients with cancer, with little change in the past 20 years.⁴

When medical and interventional treatment options have failed, ablative procedures of the central nervous system may be utilized as a therapeutic means of pain control. Ablation techniques have been recently revisited due to the improved outcomes linked to technological advances, such as the introduction of magnetic resonance imaging (MRI)–based targeting.⁵

The anterior cingulate cortex (ACC) has been shown to be involved in the perception of pain, particularly the affective component.⁶ The ACC is located in the medial part of the cerebral hemispheres, partially circumscribing the corpus callosum. The ACC is part of the limbic system, receiving inputs from the amygdala and other limbic structures and projecting outputs to the periaqueductal gray and brainstem. Its role in emotion has led to its targeting in ablative procedures for treating obsessive-compulsive disorder, depression, and severe anxiety.⁷

The ACC has been investigated as a target for cancer pain because it contributes to numerous functions related to cognition, socio-emotional processes (like reward or empathy), and body self-perception, and it is actively engaged during pain processing. Blom et al.⁸ reported that in a mouse model, sciatic nerve injury, which causes the development of neuropathic pain, results in structural modifications of the local microcircuitry in neurons in layer 5 (L5) of the ACC. Moon et al.⁹ demonstrated that optogenetic inhibition of ACC neurons in rats with either chronic constriction of the trigeminal nerve or “spared nerve injury” (an animal model of neuropathic pain that involves creating a lesion of two of the three terminal branches of the sciatic nerve [tibial and common peroneal nerves], leaving the remaining sural nerve intact) resulted in a reduction of cold hypersensitivity, similar to what is observed after ACC lesions in rodents. Koyama et al.¹⁰ recorded single neuronal activities in the ACC of a macaque monkey while it was performing a pain-avoidance task and examined them with nociceptive cutaneous electric stimuli. The authors found that among those neurons several were involved both in nociception and in the pain anticipation response that precedes the avoidance of noxious stimuli. Similar findings have been replicated in the human ACC.¹¹

Stereotactic anterior cingulotomy was initially described as effective in improving the emotional response to pain in patients otherwise treated for psychiatric illnesses.¹² Since then, several authors have reported improved pain perception and opioid use reduction after radiofrequency (RF) cingulotomy.^{13, 14} Observed adverse events have included transient postoperative spasticity and hemiparesis, bleeding, transient postoperative confusion, ataxia, mild weight gain, transient bowel and bladder incontinence, and seizures. A noninvasive approach involving stereotactic radiosurgery is also possible, though it carries risks associated with the delivery of radiation.¹⁵

Transcranial magnetic resonance (MR)−guided focused ultrasound (MRgFUS) offers incisionless surgery and is Food and Drug Administration (FDA) approved for the treatment of essential tremor and dystonia in Parkinson’s disease.¹⁶ Recently, chronic medication-refractory neuropathic pain has been treated via central lateral thalamic nucleus thalamotomy using MRgFUS.¹⁷ This is especially relevant in patients with cancer who, after conventional surgery, have a major risk for infections and wound dehiscence linked to iatrogenic neutropenia and the increased metabolic demands for tissue healing.

Simulation experiments performed by other groups have demonstrated that the cingulate gyrus is within the theoretical treatment envelope for both the 650- and 230-kHz Exablate Neuro systems (Exablate Neuro is the brand name of the MRgFUS device; Insightec).¹⁸ However, these studies have failed to report a significant thermal rise from the baseline, which was insufficient to induce a thermal lesion. The aim of this study is to determine if the transcranial MRgFUS Exablate 650-kHz system can effectively target the anterior cingulate gyrus and to analyze technical factors implied with energy delivery and temperature rise at this target.

Illustrative Case

An 86-year-old, female, unembalmed whole cadaver was acquired within 24 hours of death due to cardiac arrest. This fresh cadaveric specimen was preserved at 2°C overnight, and about 6 hours prior to the procedure, it was thawed at room temperature (21°C). A whole cadaver was used to prevent the introduction of intracranial air, which impedes ultrasound waves. The head was closely shaved to allow efficient transduction of ultrasound energy, and a modified Cosman-Roberts-Wells (CRW) frame (Radionics) was affixed to the head with a 4-point pin connection. In accordance with standard institutional policy, care was taken to not overtighten the pins to prevent a breach of the inner table of the skull and contact of the pins with the dura.

To plan the treatment, a thin-cut 1-mm head computed tomography (CT) scan using an H60 bone filter was obtained prior to placement of the CRW frame (Siemens Somatom Definition AS+ CT scanner, spiral acquisition aligned to the anterior commissure–posterior commissure plane, kVp = 120 kV/x-ray tube current = 40 mA). The images were then uploaded to the Exablate Neuro software. After frame placement, the cadaver was brought into the MRI room. The MRI scanner used for the experiment was a 3-T GE Discovery. Three-dimensional fast imaging employing steady-state acquisition (3D FIESTA) sequences were obtained (repetition time 4.8 msec, time to echo 1.4 msec, slice thickness 0.5 mm, field of view 20 × 20 cm, matrix 352 × 192, number of excitations 4) to include the region of interest through the cingulate gyrus. A silicone membrane was applied to the scalp and filled with chilled, degassed water (dissolved oxygen less than 1.2 parts per million). Confirmatory 3D FIESTA planning images were then obtained in 3 orthogonal planes. The CT imaging bone correction algorithm was overlaid and applied to the MR images, and confirmation of accurate coregistration was performed (Fig. 1).

FIG. 1.

Axial (A), sagittal (B), and coronal (C) Kranion images showing intracranial areas that are easily accessible (green), on the periphery of accessibility (yellow), and outside the accessibility (red) of the Exablate transducer envelope. Kranion output (D) showing the transducer in relationship to the skull, with the green dots indicating the active transducer elements used and the red dots indicating the inactive elements.

At this point, we defined the targets and began the sonications. The target was chosen by direct targeting using standard measurements.¹⁹ The target location was selected 20 mm posterior to the ventral-most component of the anterior horn of the lateral ventricle in a plane parallel with the roof of the lateral ventricle in a midsagittal plane. This was then advanced rostrally in the coronal plane until at the level of the cingulate gyrus, and then final targeting was adjusted medially until the target was in the center of the cingulate gyrus using electronic steering (Fig. 2).

FIG. 2.

The cingulotomy target location of the right cingulate gyrus is demarcated by the green crosshairs (A). Map of the elements from the console (B).

A total of 3 short, low-energy sonications were performed to confirm the accurate focusing of the ultrasound waves. After appropriate targeting in all 3 planes, treatment sonications were then performed within the cingulate gyrus (Table 1). Real-time MR thermometry was implemented to assess temperature rise at the target. This process was performed bilaterally. The postprocedure T2 MRI was not performed secondary to equipment constraints, as the magnet’s table stopped working at the end of the treatment, and the specimen needed to be disposed of immediately following the experiment per the Radiology Research Office’s current protocol.

TABLE 1.

Individual sonications from the treatment export

Sonication No.	Side	Power (W)	Measured Power (W)	Duration (sec)	Energy (J)	Max Energy (J)	Aborted (cavitations)?	Focal RAS-R	Focal RAS-A	Focal RAS-S	Max Temp (°C)	Max Average Temp (°C)	Temp Rise (°C)
1	Rt	200	204.2	20	1500	1500	No	7.853032	50.92323	−4.72413	41.44114	40.127903	4.441143
2	Rt	200	204	20	1500	1500	No				42.79086	41.194866	5.790859
3	Rt	298.2717	302.1	25	3100	3100	No				47.33039	45.110111	10.33039
4	Rt	298.2717	301.5	25	3100	3100	No				44.48032	42.140072	7.480324
5	Rt	486.195	499.7	20	9723.9	12000	Yes				48.0068	45.465763	11.0068
6	Rt	544.9296	534.9	25	13,623.24	17050	Yes				38.1076	38.761318	1.107601
7	Rt	577.9543	511	25	14,250	14250	No				51.14058	48.504204	14.14058
8	Rt	342.6308	352.7	45	12,050	12050	No				52.69136	49.218662	15.69136
9	Rt	342.6308	351.9	45	15,050	15050	No				56.58853	52.222855	19.58853
10	Rt	547.7892	452.7	25	13,694.73	15000	Yes				42.22579	39.938156	5.225788
11	Lt	490.6258	295.7	35	17,000	17000	Yes	−1.7122	49.54666	−2.25513	37.30418	37.858559	0.304176
12	Lt	516.1426	358	35	14,000	14000	Yes				37	37	0
13	Lt	800	517.5	30	20,000	20000	Yes				38.76203	38.442276	1.762035
14	Lt	795.0001	651.6	30	16,700	16700	Yes				44.69089	41.341141	7.690887
15	Lt	592.5001	599	46	15,300	15300	Yes				38.75015	38.58709	1.750149
16	Lt	405	405	46	13,000	13000	No				51.72296	50.075974	14.72296
17	Lt	405	403.9	46	15,000	15000	Yes				42.84923	41.305511	5.849228
18	Lt	405	406.4	46	14,000	14000	No				49.46679	48.027828	12.46679
19	Lt	607.5001	304	40	24,300	25100	Yes				41.61691	38.80759	4.616905

RAS-A = right-anterior-superior coordinate system–anterior; RAS-R = RAS-right; RAS-S = RAS-superior; Temp = temperature.

Results

After accounting for sinuses, membrane folds, and calcifications, a total of 737 elements were available for thermal ultrasound ablation. The skull density ratio was 0.58. Adequate focusing of the ultrasound transducer was achieved for both clinically relevant sections of the cingulate gyrus. Target coordinates were respectively 36 mm anterior to the midcommissural point (MCP), 42 mm superior to the MCP, and 3 mm lateral to the MCP for each side in this case.

A total of 3 initial targeting sonications followed by 7 lesioning sonications were performed on the right cingulate target. This was followed by a total of 8 lesioning sonications on the left cingulate target. During high-power sonications, the total energy delivered resulted in the highest temperature peak of 57°C (15,050 J, 350 W, 45 seconds) in the right cingulate and 52°C (13,000 J, 405 W, 46 seconds) in the left cingulate. Table 1 reports the temperature rise for each sonication assuming a baseline temperature of 37°C, as we did not measure the actual specimen temperature. We did not observe a significant difference in target heating efficiency between sides. Fluctuations in energy delivery and power were noted during several of the high-power sonications. This variability is likely due to the nature of the cadaveric specimen compared to an in vivo specimen, as previously reported.²⁰ With higher-powered lesioning, it should be noted that the tissue condition of the cadaver impacted the ease with which a thermal increase in brain tissue could be achieved, as reported in Table 1. The cadaver specimen also had a large brain mass, which had been present prior to the death of the female donor, that was poorly demarcated on the preprocedure and treatment images (Fig. 3). The mass was outside the treatment field but may have affected the targeted tissue secondary to different tissue densities.²¹

FIG. 3.

The composite figure shows the ill-defined lesion that was present within the cadaveric specimen.

Patient Informed Consent

Patient informed consent was not required in this study.

Discussion

Observations

In this cadaveric study, we demonstrated that the anterior cingulate gyrus can be effectively targeted using the Exablate 650-kHz transducer system and that the temperature rise at the focus is within the range previously described to create an ablative lesion within the thalamus.²² The cingulate cortex is a bilateral parafalcine structure made up of the cingulate gyrus and the cortical gray matter lining the superior and inferior borders of the cingulate sulcus, which is more superficially located than the current FDA-approved thalamic ventral intermediate nucleus target. In a usual Cartesian coordinate system, the cingulate gyrus is typically 2–10 mm lateral to the midline and anywhere from 16 to 24 mm posterior to the anterior horn of the lateral ventricle along a plane parallel to the roof of the lateral ventricle. For this study, a midpoint of 20 mm was chosen posterior to the anterior horn of the lateral ventricle bilaterally. The center of the cingulum in that plane was then used as the final target for the center for the ablation. This target is knowingly quite anterior within the allowable treatment envelope for the Exablate transducer.

Based on the current literature, there is much heterogeneity regarding which portion of the cingulate gyrus is targeted, with distances ranging from 16 to 24 mm posterior to the anterior horn of the lateral ventricle.^{23, 24} There is no standard consensus on the ablation volume necessary for pain relief; however, one single-center review found a mean ablation volume of 13.3 ml.³

Anterior cingulotomy has proven to be a promising treatment for cancer-associated pain for several decades.²⁵ In a study following patients who had undergone RF anterior cingulotomy (n = 10), 60% reported “fair” or “good” pain relief at 3 months postoperatively.¹¹ This result is durable when contextualized with an earlier finding (n = 12) in which 58.3% reported “meaningful” or “significant” pain relief 3 months postoperatively.¹⁴ Similar pain alleviation is observed when anterior cingulotomy is performed via laser interstitial thermal therapy (LITT). In a meta-analysis of patients who had undergone either LITT or RF cingulotomies for neoplasm-associated pain, 68% (n = 97) reported postoperative relief.²⁶

The field of MRgFUS is a rapidly emerging area of interest, and recent trials have shown the procedure’s potential for the treatment of essential tremor, Parkinson’s disease, obsessive-compulsive disorder, neuropathic pain, and brain tumors.²⁷

One main advantage over other viable surgical alternatives, mainly LITT and RF ablation, is the avoidance of a surgical incision. MRgFUS therapy offers an incisionless solution that can potentially limit some of these complications with regard to wound healing and infection and even which patients may be appropriate for surgical candidacy. Furthermore, susceptibility to infections among cancer patients is well represented in the literature. In a single-surgeon study employing LITT for the treatment of metastatic disease, 2 of 9 patients developed complications, one a superficial wound infection and the other a cerebrospinal fluid leak.²⁸

Given the possibility to perform multiple ablative sonications within the transducer envelope and the ability to intraoperatively discuss the patient’s pain levels, it would be theoretically possible to place 2–4 successive ablations in the cingulate gyrus to fully cover the 16- to 24-mm span that has been reported in the literature.

There are indeed challenges with translating the sonication parameters directly from this cadaveric model to in vivo applications. Reaching lesioning-range temperatures in an in vivo setting may theoretically require more energy than in this cadaver model, as previously reported,²⁹ due to the presence of blood flow. Nevertheless, this should not be interpreted as the impossibility of creating a lesion at the cingulate target. As a matter of fact, previous experience with cadaveric models similar to the one implemented in this experiment has failed to demonstrate a lesioning range (> 18°C) temperature rise even for the thalamic target currently used in essential tremor patients.³⁰ This suggests that further studies are required to assess the optimal sonication parameters to create a large cingulate lesion comparable with RF or laser ablation.

Lastly, Chauvet et al.³⁰ have previously used fresh cadaveric models for their simulation studies. However, their implementation differs from the one used in our paper. Based on the recommendation of the anatomy laboratory at our institution, we used a specimen with the head attached to the torso in order to minimize manipulation of the specimen and ultimately avoid intracranial air. According to their experience, minimal surgical manipulation coupled with using a fresh specimen were the most important factors in limiting the egress of cerebrospinal fluid and a significant decrease in blood volume in the brain, thereby preventing pneumocephalus.

Lessons

When performing the treatment sonications, we encountered several issues due to the presence of cavitations. This is likely attributable to several factors. The cadaver used was a fresh frozen specimen provided within 24 hours following the death of the donor. Additionally, there was a preexisting brain lesion that contributed to the mass effect, likely altering brain density and causing cerebral edema, which would have impacted the ability to achieve higher temperatures secondary to gas formation and cavitation within the brain tissue. In targeted tissues, frictional heat generated by acoustic energy can reach temperatures of up to 60°C, forming well-defined areas of coagulation necrosis. Meanwhile, building acoustic pressure causes gaseous nuclei to expand and contract, leading to acoustic cavitation as cellular and subcellular structures collapse.³¹ Previous studies have found that for treatment-level sonications, the optimal system settings that achieve 100% delivery of energy were 12,000–15,000 J, 350–400 W, and a 45-second duration of delivery. Using these settings, we achieved a maximum temperature of 57°C, assuming a baseline temperature of 37°C. However, cumulative dose estimation was hindered by the frequent cavitations, though not reported in this study. As with previous cadaver studies, due to the absence of blood flow, high-frequency focused ultrasound induced a rapid increase in tissue temperature, which is not so important in vivo due to blood perfusion, and can result in locally altered tissue architecture, ultimately facilitating the presence of cavitations.³² Lastly, we were not able to obtain confirmation (based on imaging or histology) of the ablated volume secondary to equipment constraints.

In conclusion, the cingulate gyrus is a feasible target for ablation using MR-guided high-intensity focused ultrasound. This could lead to another treatment option for severe cancer-associated pain. More in vivo studies are needed to understand the optimal parameters for ablation in this area, using the experience with RF and LITT as a reference. Additionally, a limited clinical trial with advanced cancer patients with intractable pain using focused ultrasound could be performed, as cingulotomy has been used before with success in this population.

Acknowledgments

Dr. Dalm reported grants from the National Institutes of Health outside of the submitted work.

Disclosures

Dr. Dalm reported personal fees from Medtronic for service as an instructor for pain therapies, personal fees from Surgical Information Sciences for consulting, and personal fees from Varian for service as a Functional Medical Advisory Board member outside of the submitted work.

Author Contributions

Conception and design: Sammartino, Dalm; Acquisition of data: Sammartino, Mossner, Reddy, Dalm; Analysis and interpretation of data: all authors; Drafting the article: all authors; Critically revising the article: all authors; Reviewed submitted version of manuscript: Sammartino, Stecko, Dalm; Approved the final version of the manuscript on behalf of all authors: Sammartino; Statistical analysis: Sammartino, Reddy; Administrative/technical/material support: Sammartino, Reddy; Study supervision: Sammartino.

Correspondence

Francesco Sammartino: The Ohio State University, Columbus, OH. francesco.sammartino@osumc.edu.