Image-guided stereotactic radiotherapy in 4 dogs with intracranial neoplasia

Abstract

The purpose of this study was to describe the use, and side effects, of a novel stereotactic radiotherapy protocol using TomoTherapy^® in 4 dogs with confirmed or suspected primary extra-axial intracranial neoplasia. Three fractions of 8 Gy were prescribed. Acute side effects were noted in 1 dog; no late effects were noted.

Stereotactic refers to the use of a three-dimensional system that allows individuals to precisely locate a lesion using diagnostic imaging. Stereotactic radiosurgery (SRS) and stereotactic radiotherapy (SRT) are image-guided treatment modalities that provide accurate delivery of high dose radiation to tumors in fewer fractions compared to traditional fractionated radiation (1,2). Stereotactic radiosurgery and SRT techniques for animals have been adapted from human protocols. Stereotactic radiosurgery delivers 1 dose of radiation to a brain tumor. Stereotactic radiotherapy is an amalgam of SRS and conventional multiple fractionated radiotherapy and usually consists of 3 to 5 fractions (3). The Radiation Therapy Oncology Group (human organization) has established tolerated doses of single fraction SRS for different tumor volumes.

Stereotactic radiation has several proposed advantages over traditional fractionated radiation therapy including reduced side effects and fewer anesthetic events with less required food withholding (4,5). The purpose of this study was to describe the use, and side effects, of a novel SRT protocol in dogs with confirmed or suspected primary extra-axial intracranial neoplasia.

The medical records of dogs diagnosed with confirmed or presumptive primary intracranial neoplasia and treated with stereotactic radiotherapy at the University of Wisconsin-Madison Veterinary Medical Teaching Hospital (UW-VMTH) were reviewed. Dogs treated between 2011 and 2014 were included. To be included in the study dogs must have i) had a primary extra-axial intracranial neoplasia confirmed by histologic examination or diagnosed by magnetic resonance imaging (MRI); and ii) received SRT using TomoTherapy^® (Hi-Art System; Accuray/TomoTherapy, Sunnyvale, California, USA) at the University of Wisconsin Veterinary Care Hospital. Extra-cranial tumors with local invasion into the brain (e.g., skull osteosarcoma), and pituitary tumors, were excluded. Dogs were permitted to receive glucocorticoid therapy and anti-seizure medications during treatment. A presumptive diagnosis was based on imaging characteristics previously reported for canine primary brain tumors (4,6). Magnetic resonance imaging was performed with either a 1.0 T (Signa Advantage; GE Healthcare, Milwaukee, Wisconsin, USA) or 1.5 T magnet (Signa Horizon LX; GE Healthcare). Transverse T1-weighted pre- and post-contrast, T2-weighted, gradient echo and fluid attenuated inversion recovery (FLAIR) sequences, sagittal T2-weighted and sagittal and dorsal plane, T1-weighted post contrast sequences were obtained in all dogs. A board-certified radiologist reviewed all studies at the time of diagnosis.

Treatment planning was based on a virtual planning study in human patients specifically designed for helical TomoTherapy^® machines (7). Three fractions of 8 Gy (total dose 24 Gy) were prescribed in all dogs. The MR images (T1-weighted post contrast) were fused with non-contrast simulation computer tomography (CT) in order to facilitate target and organ at risk (OAR) contouring. Iodinated contrast medium was administered after the original non-contrast scan was obtained. Slice thickness for MRI and CT scans were 4 mm and 2.5 mm, respectively. This did not affect the quality of image set fusion. The gross tumor volume (GTV) was defined by contrast enhancement on the MR T1-weighted post gadolinium image; and the planning tumor volume (PTV) was a 2-mm expansion from GTV. Previous work with our immobilization system determined that 2 mm is an appropriate margin (8). A 3-mm ring peripheral to the target was created to facilitate dose fall-off in adjacent normal tissue. A 2.5-mm diameter central sub-volume (CVS) was placed at the target center. An inner sub-volume (SV) having a volume greater than 1 voxel, but less than 2 mL was added when tumor volumes were larger than 2 mL. Using inverse planning, all plans were calculated on a fine dose grid giving a dose voxel size of 1.41 × 1.41 × 1.25 mm. A field width of 10 mm was used to maximize rate of dose fall-off at cranial and caudal target boundaries. The PTV maximum dose (PTVmax) was 120% of the prescribed dose of 24 Gy, delivered in 3 fractions. Minimum dose of 24 Gy was prescribed to 95% of the PTV. Three doses of 8 Gy were delivered on a Monday, Wednesday, Friday or a Wednesday, Friday, Monday schedule in all dogs. TomoTherapy^® allows back calculation to confirm that the dose delivered matches the planning dose. This was performed in all 4 cases and the delivered dose matched the planning dose. Dogs were set up in their immobilization device (made during CT simulation) prior to each treatment (Figure 1) and a mega-voltage CT (MVCT) was obtained using the linear accelerator. This MVCT was aligned with the planning kVCT in 6 degrees of freedom. Registration parameters were then applied and treatment was delivered. General anesthesia, including mechanical ventilation, was used for each treatment; however, anesthetic protocols varied depending on patient needs and anesthesiologist preference.

Figure 1.

Photograph of a dog attached to the bite-block array and deformable mattress for immobilization prior to radiotherapy. Asterisk indicates the caudal indexing bar.

Two Radiation Therapy Oncology Group (RTOG) indices were applied to each case to evaluate the quality of SRT plans (9). Conformity Index (CI) reported dose coverage of the target volume (TV) with a CI of 1–2 indicating ideal. The Homogeneity Index (HI) indicated dose homogeneity across the TV. A HI ≤ 2 indicated treatment complied with the intended protocol (9). Acute and late radiation morbidity was scored according to the Veterinary Radiation Therapy Oncology Group (VRTOG) toxicity criteria for each dog (10). Routine follow-up was recommended at 14 d, 1 and 6 mo, then yearly or sooner if clinical signs returned following SRT for all dogs. A Quality of Life Index (QOLI) survey was given once by telephone follow-up to owners of the surviving dogs. Owners were asked to rate their dog’s quality of life after SRT on a scale of 1–5. A score of 1 indicated a poor, and a score of 5 indicated a good, quality of life compared to before radiation therapy.

Four dogs met the inclusion criteria. One each of the following breeds was included: Labrador retriever mix, golden retriever, Maltese mix, and Dalmatian mix. The median age on presentation was 9.75 y (range: 8.6 to 11.8 y). Three dogs were spayed female; 1 was a castrated male dog. Thoracic radiographs (n = 4) and abdominal ultrasound examinations (n = 3) were performed prior to MRI, and were within expected parameters. Two dogs had histopathological confirmation of meningioma, the remaining 2 dogs were diagnosed on imaging characteristics with presumptive meningioma. Histopathologic diagnoses were attained following transfrontal craniotomgy with biopsy in 1 dog, and necropsy in 1 dog. Stereotactic radiotherapy was performed at a median of 15 d (10 to 53 d) after the imaging diagnosis. Dog 1 had SRT delayed 53 d because a transfrontal craniotomy for lesion debulking was performed 27 d after imaging diagnosis; SRT was started 25 d after surgery. For dog 1, the pre-surgery MRI was fused with the post-surgery CT scan. The GTV included the original tumor before surgery. No CTV was contoured for this or the other dogs, reported herein. Chemotherapy was not administered to any patient prior to, during, or following SRT. Median planning TV was 4.26 cc (range: 0.23 to 8.09 cc). Optimal CI was obtained in 3 of 4 dogs; the 4th dog had a CI of 0.97, indicating a slight decrease in TV coverage. Median CI was 1.11 (range: 0.97 to 1.53). Homogeneity index was < 2 in all dogs. Median maximum dose was 28.3 Gy (range: 27.5 to 29.4 Gy) (Table 1).

Table 1.

Summary data for stereotactic radiotherapy treatment protocol using the TomoTherapy^® system. Planning target volume (PTV), conformity index, homogeneity index, and maximum dose (Gy) are reported

	Case
	1	2	3	4	Median
Planning target volume (cc)	5.53	8.45	3.18	0.23	4.3
Conformity index	1.53	1.01	0.97	1.22	1.1
Homogeneity index	1.23	1.15	1.20	1.16	1.2
PTV maximum dose (Gy)	29.44	27.53	28.82	27.82	28.3
PTV minimum dose (Gy)	19.22	22.53	19.63	23.32	25.4
PTV median dose (Gy)	25.58	26.01	26.31	25.67	25.8
Location of tumor	Right olfactory bulb	Right ventral brainstem	Ventral thalamus	Left cavernous sinus	N/A
Tumor size (cm)	1.1 × 1.7 × 1.1	1.7 × 1.5 × 1.6	1.0 × 1.3 × 1.1	0.9 × 0.6 × 0.6	N/A

N/A — Not available.

Median follow-up time was 287 d (range: 65 to 558 d). Acute side effects of SRT were noted in 1 dog. The acute side effect was limited to erythema of the pinnae, which was responsive to topical steroid administration. A second dog had a self-limiting leukocytosis and lethargy, which was attributed to side effects of prednisone; however, side effects from anesthesia or other unidentified etiology cannot be ruled out. No late side effects were noted for any dog; however, late side effects may occur months to years following radiation. Median owner-assessed QOLI was 3 (range: 3 to 4). Six months following SRT, dog 3 had repeat MRI. Using Response Evaluation Criteria in Solid Tumors (RECIST), a partial response was assigned based on a 55% reduction in tumor size (10).

This is the first study which describes SRT using TomoTherapy^® and reports complications in a pilot group of dogs. TomoTherapy^® is unique because it combines a 6 MV linear accelerator and a helical CT imaging system to enable daily image guidance to ensure accuracy of patient position before each treatment delivery and confirmation of dose delivery. TomoTherapy^® delivers dose by intensity-modulated radiotherapy with a multileaf collimator that shapes and modulates the beam. The radiation protocol of 8 Gy × 3 fractions for a total dose of 24 Gy was based on 2 papers that discussed IMRT, SRS, SRT in humans and normal tissue tolerance doses and Biological Effective Dose (BED) calculations (7,11,12). Both the CyberKnife and the Gamma Knife have also been used for stereotactic radiosurgery. The CyberKnife is a frameless robotic radiosurgery system that uses a linear accelerator to deliver radiotherapy, with the intention of targeting treatment. The Gamma Knife typically contains 201 cobalt-60 sources of approximately 30 curies (1.1 TBq), each placed in a circular array in a heavily shielded assembly. The device aims gamma radiation at the target. An ablative dose of radiation is sent to the target in 1 treatment session, while surrounding tissues are relatively spared. TomoTherapy^® delivers dose in a helical fashion, 360° delivery, which differs from traditional linear accelerators that only deliver from a few angles. TomoTherapy System can deliver SRS, SRT and fractionated radiotherapy. Adding specific software to traditional linear accelerators with MLCs allows rotational delivery equal to TomoTherapy^®. However, TomoTherapy^® provides daily image-guidance with sub-millimeter accuracy in translational and roll dimensions and allows adaptive radiotherapy, which is unique to this system.

Side effects noted in this pilot group of dogs were limited to acute side effects in 1 dog. Acute side effects were not noted in a recent study by Griffin et al (13), which reported outcome and side effects in dogs with presumptive intracranial meningioma following SRT. However, in that report 14 dogs were reported to have side effects within the first 6 mo. Due to the short follow-up time in our study, long-term side effects cannot be compared with that study.

Conformal Index (CI) and Homogeneity Index (HI) are commonly used to evaluate stereotactic treatment plans in human patients. Homogeneity index is used to compare the dose gradient within the target and CI is used to evaluate the quality of target coverage, where the isodose line that covers the target is ≥ to 90%. These indices were used to illustrate similar plan outcomes with respect to target dose gradient and coverage in these dogs.

In summary, SRT with TomoTherapy^® was delivered to 4 dogs with primary intracranial extra-axial neoplasia. Treatment was delivered over 3 treatments, thereby reducing the number of anesthetic events compared to traditional fractionated radiation. Minimal short-term, and no long-term side effects were noted in these dogs; however, longer follow-up in a larger cohort of dogs is needed to further evaluate the significance of this finding. CVJ