CMS Magnetic-Resonance-Guided Focused Ultrasound Surgery (MRgFUS) for Essential Tremor and Tremor Dominant Parkinson's Disease Form

Summary Of Evidence

Essential Tremor (ET)

Essential tremor (ET) is the most common movement disorder as well as one of the most treated surgically. The prevalence of ET has been estimated at approximately 3% or 10 million people in the United States. While ET does not shorten life expectancy, the associated disabling symptoms, such as hand tremor, can greatly impact quality of life (functional ADLs, work activities, mood, and socialization).

Although there are no curative therapies, symptoms of ET are well managed medically in up to 70% of patients, with surgery reserved for medication-refractory severe impairments. Current surgical options include thalamotomy with radiofrequency (RF) ablation and deep-brain stimulation (DBS); both effectively suppress tremor but require intracranial surgery. Stereotactic radiosurgery (SRS), while non-operative, suffers from delay in tremor reduction (making intraoperative validation impossible), a greater than 10% cumulative risk of adverse events, and theoretical concerns about radiation side effects.^{6, 22} DBS is currently the intervention of choice, “because of its proven efficacy, reversibility, adjustability, and durability”²², with thalamotomy “a reasonable alternative….if DBS is not available or practical”.¹ This attribute of DBS in creating an adjustable “functional lesion” causes fewer adverse events than thalamotomy^{24, 25}, and resulted in a general shift away from ablation methods.²³

Neuromodulation with ultrasound energy also required craniotomy until recently; advances in ultrasound transducer design and high-resolution magnetic resonance imaging now allow precise transcranial delivery of high-intensity focused ultrasound. The ultrasound causes a local increase in temperature in the target tissue, resulting in coagulation necrosis while sparing the surrounding normal structures. In addition to providing location guidance, MRI provides real-time clinical monitoring of treatment intensity via thermal imagery. On 1/1/16, a CPT Category III tracking code specific to MRgFUS treatment of movement disorder became effective. FDA PMA approval for the Magnetic-Resonance-Guided Focused Ultrasound Surgery System (MRgFUS) (ExAblate Model 4000, InSightec, Inc.) “for the unilateral thalamotomy treatment of idiopathic essential tremor patients with medication-refractory tremor” came on 7/11/16.³

Among the peer-reviewed clinical studies of MRgFUS for the treatment of medication-refractory ET, all but one were small, uncontrolled, pilot studies with short follow-up.^4-11 FDA approval for MRgFUS treatment of ET was based on its pivotal study, a prospective, double-blind, randomized, sham-controlled trial (RCT) of MRgFUS to create a unilateral thalamic ablation for the treatment of ET.¹² Seventy-six patients with moderate-to-severe essential tremor refractory to at least two trials of medical therapy were randomized in a 3:1 ratio to either MRgFUS or a sham procedure. The primary endpoint, the CRST at 3 months, was significantly improved in the MRgFUS group (p<0.001). Secondary outcome measures, including disability and quality of life, were also significantly improved. However, both hand and total tremor scores steadily deteriorated over the year, 23% and 38% respectively. In fact, this drop in efficacy and the limited follow-up period were cited as major concerns in the accompanying editorial which advocates for much longer follow-up (2-5 years or more) to demonstrate sustained benefit.² Another concern was persistent adverse neurologic effects in the MRgFUS group at 12 months, including gait disturbance (9%) or numbness (14%).

The editorial concludes that “A head-to-head comparison with DBS would facilitate the direct comparison of the two approaches.” Some contend that a direct comparative trial between MRgFUS and DBS will be unlikely “due to the significant differences in invasiveness of the two procedures.” Interestingly, a letter to the editor agrees a direct comparative study isn’t warranted, but apparently for the opposite ethical reason, noting “that the high rate of adverse events that is consistently reported with thalamotomy of any kind suggests that equipoise does not exist”.¹³ While it is true that MRgFUS is less invasive than DBS in terms of not requiring cranial penetration with hardware, it is more invasive than DBS in the creation of a fixed thalamic brain lesion, which can result in permanent neurologic deficit.

More recently, follow-up on this same cohort of seventy-six patients with refractory moderate-to-severe essential tremor has been reported on sixty-seven of the patients continued with monitoring for two years. The improvement in tremor was durable at 1 year (53%; 8.9 ± 4.8; 70 patients) and at 2 years (56%; 8.8 ± 5.0; 67 patients).  Disability score improved throughout this period, none of the adverse effects worsened, two resolved and there were no new delayed complications.²⁷

A recently published meta-analysis is meant to provide “an approximation of an RCT” head-to-head comparison between MRgFUS, DBS, and SRS; the authors claim an actual RCT is unlikely.²² Pre- and postoperative tremor-related disability scores were collected from 32 studies involving 83 MRgFUS, 615 DBS, and 260 SRS cases. MRgFUS thalamotomy resulted in significantly higher utility scores (defined as quality of life and derived from percent change in functional disability) compared with DBS (P < 0.001) or SRS (P < 0.001). The authors conclude that “preliminary experience with MRgFUS supports its broad adoption for medically refractory ET.”

A retrospective analysis of 59 patients who underwent unilateral treatment for drug-resistant ET with RF thalamotomy (n=17), DBS (n=19), and MRgFUS (n=23) showed no statistical differences in tremor severity improvement at 1 month or 1 year follow-up.²³ However, MRgFUS had a significantly lower complication rate (p < 0.01) at 1 year (4.4%) compared with RF (11.8%) and DBS (21.1%). The authors conclude that “MRgFUS is a promising therapy with the potential to replace DBS for patients who cannot tolerate DBS, the standard surgical treatment for ET,” but that “the long-term effects of MRgFUS should be systematically evaluated in a future prospective, randomized study in order to demonstrate whether MRgFUS provides superior management of ET symptoms.”

Tremor Dominant Parkinson’s Disease (TDPD)
Efficacy Outcomes
The efficacy of MRgFUS for the treatment of TDPD was reported in 3 evidence syntheses and one uncontrolled clinical trial, which was not included in any review:

• Lin, et al. (2021) conducted a systematic review and network meta-analysis of RCTs and NRSI.³ The review included 2 RCTs and 6 NRSI (total n=144), where MRgFUS was indirectly compared to deep brain stimulation (DBS) for Parkinsonian tremor. PD-related disability and impairment were measured using the Unified Parkinson Disease Rating Scale (UPDRS). Tremor (medication-off), Tremor (medication-on), Total (medication off/on) were also primary outcomes. The reviewers concluded MRgFUS was an efficacious intervention for improving parkinsonian tremor and presented indirect evidence in identifying that MRgFUS is not inferior to DBS in parkinsonian tremor suppression. These conclusions should be considered with caution due to the lack of long-term assessment of efficacy and indirect comparison between DBS and MRgFUS.
• Xu, et al. (2021) systematically reviewed RCTs and NRSI for the efficacy of MRgFUS for TDPD in adults (mean age >65 yrs.).⁵ Tremor quantification (UPDRS III) was assessed in 10 studies. Timing to follow-up was < 6 months in eight of the studies. Five studies reported statistically significant improvement of performance on UPDRS-III after MRgFUS, whereas in one study the improvement was not statistically significant. Although improvement of performance on UPDRS-III was also reported by the other studies, a statistical result was not available due to the lack of clear report or study design (case report with only one sample). The reviewers described several limitations. Most included studies were observational studies with low quality. Only two studies included a double-blinded, randomized, and controlled phase which only lasted for 3 months, and then turned to the single-arm, open-label phase. The sample size of included studies was small, and the follow-up was short. The high heterogeneity among these studies precluded a quantitative synthesis of the outcome.
• Ge, et al. (2021) performed a meta-analysis of 2 RCTs (N=67; 27, 40) that measured the efficacy of MRgFUS compared to sham in the treatment of Parkinson’s disease.⁷ The blinded phase lasted for 4 months in one experiment²¹ and up to 3 months in the other.¹² The MRgFUS group showed significant improvement in limb tremor on the treated side (SMD: − 1.20; 95% CI: − 2.06, − 0.34). In addition to only two studies, small sample sizes, and short-term data, the authors noted the need for large multicentered RCTs.
• A single uncontrolled clinical trial that was not included in an evidence synthesis was identified for this review.²² Yamamoto, et al. (2021) prospectively investigated the 1-year outcomes of ventral intermediate nucleus (VIM) thalamotomy with MRgFUS for eleven patients (mean age 71.6; 58-79) with medication-refractory TDPD. Tremor symptoms and disability were assessed using the Clinical Rating Scale for Tremor (CRST) at baseline and at 1, 3, and 12 months. All patients experienced significant improvement in tremor during the procedure. The median improvement in CRST scores of the hand contralateral to the treated VIM between baseline and 12 months postoperatively was 87.9% (70.5–100.0). The median scores of the treated upper extremity showed the improvement to be 66.7% (50.0–100.0), 100.0% (100.0–100.0), and 100.0% (100.0–100.0) in resting, postural, and action tremors, respectively. Furthermore, the median improvement in total tremor scores on the CRST was 65.3% (55.7–87.7) from baseline to 12 months, and the median improvement of functional disability in Part C on the CRST was 66.7% (15.5–85.1). The small sample size, absence of a comparator group, and the short-term outcome measures were the primary limitations of this study.

Safety Outcomes
Adverse events during sonification and/or after MRgFUS were reported in three of the evidence syntheses and one uncontrolled clinical trial that was not included in an evidence review:

• Schreglmann, et al. (2018) included RCTs and NRSI in a systematic review and meta-analysis.⁴ The relative safety of MRgFUS was compared to lesional neurosurgical interventions in the treatment of tremor due to Parkinson's disease. These procedures included gamma knife of the ventral intermediate nucleus (GK vim), radiofrequency of the globus pallidus internus (RF GPi), radiofrequency of the subthalamic nucleus (RF STN), radiofrequency of the ventral intermediate nucleus (RF vim). Mean rates of persistent side effects after unilateral lesions in Parkinson’s disease were 12.8% (RF vim), 13.6% (RF STN), 9.2% (RF GPi), 0.7% (GK vim) and 7.0% (MRgFUS vim).
• Xu, et al. (2021) reported most adverse events associated with the use of MRgFUS for the treatment of Parkinsonian tremors were mild and transient. Further, there were no statistical differences existed in adverse events between the active and the sham groups. During the procedure, headache was the most common adverse event. Posttreatment ablation-related adverse events included ataxia (35-60%; mild/transient to 1-yr), paresthesia (19-39%; mild transient to 1-yr). Behavioral changes were reported in a single study and resolved within one month.
• Ge, et al. (2021) reported no serious side effects/adverse events with MRgFUS for Parkinson’s disease.⁷ Data obtained from 2 small RCTs found dizziness (OR: 4.68; 95% CI: 1.20, 18.23) was more common in the treatment group, with no group differences in the other adverse events.
• Yamamoto, et al. (2021) Assessed adverse events in a prospective uncontrolled trial involving 11 participants diagnosed with medication-refractory TDPD.²² Most adverse events were mild and transient, and improved within 12 months. Headache was the most common adverse event and occurred only during sonication, disappearing immediately after the sonication was terminated. Notably, no delayed adverse events were observed.

Measures of Function and/or Well-being of the Patient

Two evidence syntheses evaluated other outcomes, which included patient-reported measures [e.g., quality of life (QoL), ability to perform activities of daily living (ADLs)] and neurocognitive functioning:

• Lennon, et al. (2021) employed a systematic review of RCTs and NRSI in assessing pre- and post-operative cognitive functioning in medication-refractory TDPD patients undergoing MRgFUS. Based on two studies that only utilized cognitive screeners, MRgFUS appeared to be a procedure without significant concerns related to cognitive outcomes in the perioperative or follow-up periods up to 12 months. QoL and ADLs also seemed to be generally preserved in an overwhelming majority of patients This review highlighted the limitations in the ability to speak on these outcomes. The small number of published articles, assessment methods (cognitive screening tools in lieu of comprehensive neuropsychological data), short-term follow-up, and participant attrition did not permit strong conclusions.
• A meta-analysis of two small RCTs found significant improvement in the ability of patients with Parkinson’s disease, who received MRgFUS, to perform daily activities (SMD: − 0.86; 95% CI: − 1.41, − 0.32) compared to the sham group. There were no significant group differences in other indicators.⁷

ECRI

EXECUTIVE SUMMARY

• Evidence is inconclusive - too few data on outcomes of interest
• Evidence from 1 small randomized controlled trial (RCT) and 3 small case series suggests that MRgFUS can safely reduce tremor and improve QOL in patients with PD; however, these studies are too small and at too high a risk of bias to be conclusive. Larger RCTs comparing MRgFUS to other treatments and reporting on tremor and QOL at follow-up times longer than 1 year are needed to validate results and determine comparative and longer-term efficacy; ongoing trials may partially address some evidence gaps.
• Evidence limitations: The RCT is at risk of bias because of its small size and unbalanced MRgFUS and sham control groups. All included case series are at a high risk of bias due to small size, single-center focus, and lack of control groups. One case series evaluated patients at different follow-up times. Overall, studies included patients with different subtypes of PD and who underwent different surgeries; results from 1 group of patients and 1 type of surgery may not generalize to other patient groups or procedures. Finally, no studies reported on the amount of time patients spent on-medication compared with time off-medication, which is needed to give a clearer picture of MRgFUS's effectiveness, especially over time.

Analysis of Evidence

In summary, MRgFUS is a promising new treatment approach that has attributes, positive and negative, distinct from both traditional thalamotomy and DBS. However, long-term effectiveness and safety remain uncertain^1,23 and warrant a direct comparison with DBS, the current surgical standard. Widespread non-coverage by both Medicare ^14-17 and commercial payers ^18-21 supports this interpretation.

However, given the support for traditional thalamotomy, generally, as an alternative “if DBS is not available or practical”, and the support for MRgFUS thalamotomy, specifically, as an alternative in patients “who are not a candidate for DBS” by the American Association of Neurological Surgeons (AANS), Congress of Neurological Surgeons (CNS) and the American Association of Stereotactic and Functional Neurosurgery (ASSFN), Noridian considers MRgFUS reasonable and necessary in that context. Patient selection criteria will largely mirror those used in the pivotal study (see Coverage and Limitations section for details).

This LCD addresses use of Magnetic-Resonance-Guided Focused Ultrasound Surgery (MRgFUS) for the treatment of idiopathic essential tremor (ET) patients including Tremor Dominant Parkinson's disease (TDPD) patients with medication-refractory tremor.

MRgFUS unilateral thalamotomy is considered medically reasonable and necessary in patients with all four of the following criteria:

1. medication refractory ET (defined as refractory to at least two trials of medical therapy, including at least one first-line agent)

2. moderate to severe postural or intention tremor of the dominant hand or another nationally accepted clinical measure of tremor severity

3. disabling ET (defined by a score of ≥2 on any of the eight items in the disability subsection of the CRST or another nationally accepted clinical measure of tremor severity)

4. not a candidate for DBS (e.g., advanced age, anticoagulant therapy, surgical comorbidities, or has failed Deep Brain Stimulation (DBS), but has no retained cranial implants)

Limitations (not covered):

1. Treatment of head or voice tremor
2. Bilateral thalamotomy
3. An advanced neurodegenerative condition
4. Unstable cardiac disease
5. Depression sufficiently severe to compromise beneficiary’s ability to provide informed consent and limit likely clinical benefit of the treatment
6. Severe cognitive impairment (such as may be defined by a score of <24 on the Mini–Mental State Examination)
7. A skull density ratio (SDR) (the ratio of cortical to cancellous bone) <0.40
8. MRI contraindicated