Deep Brain Frameless Radiosurgery for Drug Resistant Invalidating Tremor. Dose Escalation Pilot Study

BACKGROUND The stereotactic lesioning of thalamus and basal ganglia for the treatment of tremor is a well-known procedure which, before the introduction of deep brain stimulation ( DBS), was usually performed using stereotactic surgical procedures.[16] Consistent positive experience in functional radiosurgery using the Gamma Knife or linear accelerators was reported since the first report by Leksell in 1951.[4, 8, 11, 13, 14, 17, 18, 19, 21, 24, 25, 26, 27] CyberKnife (CK), (Accuray Inc., Sunnyvale, California, USA) allows a real frameless stereotactic radiosurgery of recognizable intracranial targets such as arteriovenous malformations (AVMs), tumours, and trigeminal nerve for the treatment of trigeminal neuralgia.[1, 2, 3, 7, 9, 23] When compared to the frame-based radiosurgery , frameless radiosurgery is a pain free procedure which offers the advantage of a better patient's compliance by avoiding local anaesthesia and the discomfort due to wearing the frame for the period of time needed for the whole procedure.

The present literature on functional frame-based radiosurgery (mainly based on Gamma Knife treatments) shows that, while a mean dose of 140 Gy has been effectively used, toxicity increases with higher doses. Moreover the minimal effective dose has never been reported [8, 10, 15, 25 ].

Based on the fact that the 3D dose distribution with Gamma Knife shows some subtle but substantial differences with the Cyberknife, the investigators chose a lower prescription dose compared to the literature data when performing our first attempts to treat movement disorders.

Particularly the prescription dose as the minimum effective dose for a functional disorder such as trigeminal neuralgia was identified.

At that time no previous CyberKnife experiences had been reported. The first two patients treated at our Institution by using these lower treatment doses (75 and 90 Gy respectively) [28] had remarkably positive results thus bringing us to consider the real relative effectiveness of low doses.

The minimum effective dose to perform a thalamotomy was chosen accordingly , by drawing a dose escalation clinical trial rather than a de-escalation dose protocol.

Particularly, the protocol started with a dose of 75 Gy and reached a dose of 90 Gy.

The analysis of the results from this trial highlighted the absence of tremor control but also confirmed the absolute safety of the selected doses (no reported toxicity).

For this the aim of the present study is to continue the experience to define the minimum effective dose when performing a safe frameless radiosurgical thalamotomy.

Moreover radiosurgery of invisible targets to treat movement disorders and intractable pain are so far still the domain of frame-based procedures, due to the need of a solid reference system registered to the anterior commissure-posterior commissure (AC-PC) line, which allows the use of stereotactic atlases.

In this study the investigators want to confirm the precision of the previously defined mathematical method (that uses atlas-derived stereotactic coordinates) to perform safe frameless radiosurgery of invisible targets.

STUDY DESIGN Longitudinal, interventional study. The project is a dose escalation study (phase II study). The effect of the treatment in term of tremor control will be evaluated.

Particularly, the differences between pre and post-treatment tremor item score will be evaluated.

The radiological (MRI/CT) appearance of a radiation necrosis at the treatment site as well as the toxicities will be also considered.

The study will finish when the 4 dose levels will be delivered.

PRE-TREATMENT EVALUATION Prior medical history and physical examination. Complete history, physical examination including a detailed neurological examination and evaluation of Karnofsky Performance Status will be performed.

The Fahn Tolosa Marin tremor rating scale (FTMTRS) will be assessed for all patients.

The unified Parkinson disease rating scale motor score will be utilize to define the severity of symptoms only in case of Parkinson's disease Unresponsiveness to conservative treatment will be verified. Any prior surgery, prior radiation therapy and/or radiosurgery of the brain will be recorded.

Prior surgery at the site of the lesion, prior radiation therapy and/or radiosurgery of the brain will not prevent the patient from participating in the protocol.

Risk analysis of surgery procedure will be express by a skilled neurological surgeon.

PRE-TREATMENT IMAGING PROCEDURE A 1 mm thickness high quality CT Scan and a dedicated MRI will be always acquired.

When possible a 3tesla MRI will be utilized. The multimodality images will be fused and then exported to the CK treatment planning system (TPS).

CYBERKNIFE RADIOSURGERY The target selection. The preferentially selected target of treatment will be the VoO-VoP nuclei of the Thalamous The choice of VoA (12-13 mm lateral to the midline, 2 mm anterior to the midcommissural point and 2 mm superior to the commissural plane ) was intended to lessen the risk of anatomical variability in these patients. In fact, this target is relatively far from the motor fibres running in the posterior limb of the internal capsule and far from other "eloquent" nuclei.

The VIM nuclei can be considered because the similarity of the main features. Target definition and stereotactic atlas registration of the CT images. During CT scanning it is critical that the patient's head remains in a fixed position in order to avoid movement artefacts. For the patients high-quality images will be obtained by restraining the patient's head using a standard thermoplastic mask and acquiring the images very rapidly, always less than 40 seconds with the CT equipment in our possession (Light Speed Ultra, General Electric, Fairfield, Connecticut, USA). Mild sedation may be necessary in some cases with head tremor. In more severe cases, in which body and/or head movements could prevent an optimal image acquisition, administration of low-dosage of Midazolam under anaesthesiological control is mandatory. If the head is immobile the CT gantry behaves like a solid reference system with fixed relationships to the brain structures. In other words the CT screen may be seen as a bi-dimensional stereotactic frame and each pixel of the CT screen represents a discrete part of the brain identified by X lateral and Y antero-posterior coordinates to the screen origin. The slice containing the anterior commissure (AC) is arbitrarily assigned to depth = 0 (Z coordinate); the depth of each slice is calculated relative to this point (the slices are 1.25 mm thickness). The system calculate the AC X, Y, and Z coordinates (Z = 0) and the coordinates of the posterior commissure (PC) where Z is the distance in mm from slice zero. In cases in which the AC and PC lie on the same slice, AC and PC Z coordinates are both equal zero and the calculations are easier. Finally the values in pixels are converted into millimetres based on the matrix/FOV ratio of the CT screen. In other words X and Y value of each pixel of the brain image on the CT screen are obtained and Z is derived as the depth of the slice measured as the vertical distance from the slice 0. Finally the investigators calculate the coordinates of the AC-PC midpoint, which is the origin of the stereotactic atlas, and a simple rototranslation between the origin of the screen and the origin of the stereotactic atlas allows us to obtain atlas-registered X, Y, and Z coordinates in millimeters of each point on the CT axial brain slices (Fig. 1).

Target coordinates of Voa/Vop complex ( X = +/-12 mm , Y = 2 mm , Z = 2 mm) derived from the stereotactic atlases registered to the mid-commissural point are easily transposed onto the corresponding CT slice and the target is drawn on the treatment planning system (Multiplan, Accuray Inc.). In other words, the axes roto-translation between the CT screen and the commissural system of the patient allows the use of atlas-derived stereotactic coordinates to make the invisible functional target visible. The CT images will be also be fused with MRI to obtain more anatomic details about the anatomical structures surrounding the estimated target.

High quality control of CT couch movements is of course mandatory for the above described procedure and possible undesired movements of the CT couch during the examination could affect the precision of the Z coordinate; even if the Voa/Vop complex is relatively close (slice + 2mm) to the slice containing the anterior commissure (slice 0) , possible errors must be taken into account.

A MRI (1 mm slice, T1 sequence ) will be performed and merged with TC images . Radiosurgical planning. After than the target volume will be identified and the critical normal structures will be contoured, in conjunction with a physicist, inverse planning using the CyberKnife Treatment planning system (TPS) will be performed to yield a treatment plan.

The treatment plan used for each treatment will be based on an analysis of the volumes and doses including dose-volume histogram (DVH) analyses of the PTV and critical normal structures.

Doses selection The aim of this procedure is to cause a lesion confined to the estimated target. The first 2 patient will be treated with a dose of 100 Gy to the 100% isodose line, the 3rd and the 4th patient will be escalated with 120 Gy to the 100% isodose line, the 5th and 6th patient will be treated with 130 Gy to the 100% isodose line and then the dose will be escalated to140 Gy will be utilized (7th and 8th).

Treatment will be delivered as a single shot radiosurgery.

PATIENT ASSESSMENTS Evaluation During Treatment Patients will be seen and evaluated during CyberKnife radiosurgery with documentation of tolerance, including acute reaction.

Evaluation Following Treatment Patients will be followed as follows 3, 6, 12, 18 and 24 months after radiosurgery Images (CT and/or MRI) will be performed and the modification of target size will be measured.

The FTMTRS and/or UPDRS score will be assessed. At each follow-up visits, toxicity will be evaluated and documented Criteria for tremor response The tremor control assessment will be based on the quantitative/qualitative analysis of pre and post-treatment tremor rating scale score/graphical features.

The quantitative analysis will be based on the numerical differences between pre and post-treatment FTMTRS score.

The qualitative analysis was mainly based on the visual judgement of the Jancovich spiral drawing from the FTMTRS..

UPDRS motor will be also evaluated for patients affected by Parkinson Disease (PD).

DISEASE RELATED DRUG ADMINISTRATION DURING THE TREATMENT AND FOLLOW-UP PERIOD Levodopa equivalent daily dosage (LEDD) will be tentatively maintained during all the observation period. LEDD could be modified for a relevant worsening of symptoms (akinesia and/or rigidity not for tremor).

STATISTICAL CONSIDERATION Descriptive statistic (mean, percentage, frequency, etc.) will be performed on quantitative variables

ADVERSE EVENTS AND RADIOSURGERY TOXICITY Adverse events will be graded and registered according to National Cancer Institute, Common Terminology Criteria for Adverse Events (NCI-CTCAE) version 4.

Serious Adverse Event stopping criteria. The study will be closed if more than 2 patients will show a serious adverse events related to or suspected for a relation with the treatment.

DATA COLLECTION Patients will be allocated a number and their data will be collected on a Case Report Forms. Data will include information from each protocol visit and will be completed on a timely manner.