Radiation therapy for intracranial tumours in cats with neurological signs

Patients

The databases of the Division of Radiation Oncology of the Vetsuisse Faculty, University of Zurich (Switzerland), the Centro Oncologico Veterinario (Italy), the Royal (Dick) School of Veterinary Studies, Radiation Oncology Department, University of Edinburgh (UK) and the University of Wisconsin-Madison Veterinary Medical Teaching Hospital (USA) were searched for cats with primary, non-lymphoid intracranial space-occupying lesions with imaging diagnosis of a tumour, presented with neurological signs that had been treated with radiation therapy alone between January 2002 and June 2017. Cats that had undergone surgery (complete excision or debulking) or received any type of systemic anticancer treatment such as chemotherapy were excluded from the study. Contributing veterinary radiation oncologists were asked to collect information from the medical record of each case including signalment, neurologic signs, diagnostic tests, tumour localisation, and imaging-based diagnosis, medical management (ie, corticosteroids), radiation therapy protocol and survival time.

The space-occupying lesions in the brain were diagnosed by a board-certified radiologist either by contrast-enhanced CT or MRI. Due to the lack of histological diagnosis, the lesions were allocated to one of three groups according to their imaging characteristics: a) extra-axial tumours (consistent with meningioma, schwannoma and choroid plexus tumour), b) intra-axial tumours (consistent with glioma) and c) pituitary tumours (consistent with adenoma and adenocarcinoma of the pituitary gland).^7,14 A second grouping was made according to the location of the tumour relative to calvarial anatomy: a) rostral cranial fossa (rostral to the optic chiasm), b) middle cranial fossa (between the optic chiasm and the dorsum sellae and petrosal crests, pituitary/temporal region) and c) caudal cranial fossa (between the dorsum sellae and petrosal crests and the foramen magnum, including brainstem and cerebellum). ¹⁵ Lesions suspicious for lymphoma (based on diagnostic imaging criteria) were excluded. ¹⁴

The patient’s neurological signs at the time of diagnosis were grouped into three categories: a) patients with altered neurologic status (changes in mentation and behaviour, gait and postural abnormalities, deficits of cranial nerves including central blindness and head tilt), b) patients with only epileptic seizures and c) patients with epileptic seizures and altered neurologic status. Other information recorded included medical treatment at the time of diagnosis, findings regarding tumour staging and, if available, cerebrospinal fluid analysis.

Treatment

All cats were treated with external beam megavoltage radiation therapy. Radiation was delivered with 6 MV linear accelerators (Dynaray LA20 [ABB/Varian], Clinac DMX or Clinac iX [Varian Medical Systems], Clinac 2100 C/D [Varian Medical Systems]).

Treatment planning was performed by a board-certified veterinary radiation oncologist in all cases, on the basis of three-dimensional (3D)-CT datasets and using computerised radiation treatment planning systems (Cadplan [precursor versions of Eclipse], Eclipse External Beam Planning system version 8.6, 10.0 and 11.0; Varian Oncology Systems). Treatment planning CT scans were performed with each patient in the treatment position using a customised immobilisation device according to each institution’s standard practice. ¹⁶ The gross tumour volume (GTV) was defined as the abnormal contrast-enhancing tissue on CT or MRI images. The clinical tumour volume (CTV) was added at the radiation oncologist’s discretion and a planning target volume (PTV) was added to include a safety margin accounting for systematic and random uncertainties. Organs at risk (normal brain tissue and eyes) were contoured in all cats.

Radiation treatment was planned isocentrically using photons and 3D conformal radiation therapy. Beam shaping devices such as lead blocks, multi-leaf collimators and wedges were used to improve homogeneity and to minimise dose to organs at risk. In general, the dose was prescribed according to the International Commission on Radiation Units & Measurements.^17,18 Irradiation was delivered in various fractionated regimes.

Follow-up

Contributing veterinary radiation oncologists gathered the post-RT follow-up information from the patients’ records. If a follow-up diagnostic imaging study was available, tumour response was quantified according to the response evaluation criteria in solid tumours (RECIST) guidelines for dogs. ¹⁹ The sum of diameters of all target lesions was calculated and reported as the baseline sum diameter. Complete remission (CR) was defined as the disappearance of all target lesions. Partial response (PR) was defined as a reduction of at least 30% in the sum of diameters of target lesions from baseline. Progressive disease (PD) was defined as an increase in the sum of diameters of target lesions by at least 20% over baseline, or the appearance of new lesions. Tumours that showed neither sufficient shrinkage to qualify for PR nor sufficient increase to qualify for PD were considered to be of stable disease (SD). For cats that were not re-evaluated at one of the participating centres, the referring veterinarians and owners were queried by telephone regarding performance status, neurological signs compared with those at the time of diagnosis, disease progression and date of death by one of the authors (MKO, MPA or SCA).

Statistical analysis

Statistical evaluation was performed under the supervision of a biostatistician (MR) and computed with a commercial statistical software package (IBM SPSS Statistics, Version 24). Data were used as submitted in an Excel table by the contributing veterinary radiation oncologists. Description of quantitative data characteristics, other than PFS and overall survival OS, is given by mean (± SD), unless otherwise specified. Description of qualitative characteristics is provided in absolute and relative frequencies.

PFS was defined as the interval between the first day of radiation therapy to measurable progression of disease, worsening of neurological signs compared with the neurological status before RT, or death. OS was defined as the interval between the first fraction of radiation therapy and death. Cats that were free of progression of measurable tumour and of progression of neurological signs and that were alive at the time of data collection were censored for PFS analysis. For OS, all deaths were considered events and cats that were still alive at the time of data collection or lost to follow-up were censored at the last date of contact.

Kaplan-Meier survival analysis was used and followed by Logrank test. The univariate Cox-regression analysis was used to determine whether the following factors were associated with PFS or OS: age, sex, meningioma/non-meningioma, meningioma/ pituitary tumour, tumour volume (GTV, GTV/brain ratio), treatment volume (PTV, PTV/brain ratio), tumour location, treatment institution, dose of radiation (total dose, equivalent dose in 2 gray (Gy) fractions [EQD₂]), dose of corticosteroids at first RT fraction, overall treatment time (counted from the first to the last fraction of RT). The EQD₂ was used to compare the different radiation protocols. Survival estimates at a certain time point are presented as a percentage with the corresponding 95% confidence intervals (95% CI). Median survival time estimates are complemented with 95% CI. The paired Wilcoxon test evaluated the change in corticosteroid dose from the first to the last RT fraction. Results of statistical analyses with P value <0.05 were interpreted as statistically significant.

Patient and tumour characteristics

Twenty-two cats met the inclusion criteria for this study. Nine of the cats were female (nine were spayed) and 13 were male (12 were castrated). The presented breeds included domestic shorthair (n = 18), Persian (n = 3) and Selkirk Rex (n = 1). The mean age was 12.0 years (± 2.3) and ranged from 6.5–16.4 years. The mean body weight was 4.5 kg (± 1.4) and ranged from 2.3–8.3 kg. Fifteen animals were treated at the Division of Radiation Oncology of the Vetsuisse Faculty, University of Zurich, Switzerland, six animals at the Centro Oncologico Veterinario, Sasso Marconi, Bologna, Italy, and one animal was treated at the Royal (Dick) School of Veterinary Studies Radiation Oncology Service, University of Edinburgh, UK. There were no cats that met the inclusion criteria from the University of Wisconsin-Madison Veterinary Medical Teaching Hospital, USA.

Based on imaging characteristics, tumours were diagnosed as meningioma (n = 11), pituitary tumour (n = 8), choroid plexus tumour (n = 2) or glioma (n = 1). Eleven tumours were considered extra-axial, three intra-axial and eight were located in the pituitary region. Tumours were distributed in the cranial fossa (n = 9), middle fossa (n = 8) or caudal fossa (n = 5). The pretreatment tumour volumes (GTV, absolute tumour size) had a mean of 1.78 cm³ (± 1.67) with a range of 0.15–7.70 cm³. Mean GTV/brain ratio (relative tumour size, [n = 16]) was 8.32 % (± 8.55) and the PTV/brain ratio could be derived for 16 cats with a mean of 33.18 % (± 34.00). The majority of cats presented with altered neurological status and without epileptic seizures (n = 20). One cat presented with epileptic seizures only (Table 1, case 22) and one cat with both altered neurological status and epileptic seizures (Table 1, case 21).

OPEN IN VIEWER

Table 1

Individual patient data regarding tumour type and size, radiation protocols, progression-free survival (PFS) and overall survival (OS)

Case	Age (years)	Sex	Tumour type ‡	GTV (cm3)	Number of fractions	Treatment schedule	Dose/fraction (Gy)	Total dose (Gy)	EQD2 (Gy2)	PFS (days)	OS (days)
1	8	mc	Meningioma	4.19	10	Daily	4	40	46.7	204 *	204 *
2	13	mc	Meningioma	0.51	10	Daily	4	40	46.7	121 *	121 *
3	11	fs	Meningioma	0.86	14	Daily	3	42	45.5	740	963 †
4	11	fs	Meningioma	1.60	18	Daily	2.5	45	46.9	50	50
5	14	fs	Meningioma	0.15	14	Daily	3	42	45.5	63	63
6	12	mc	Meningioma	0.61	19	Daily	2.25	42.75	43.6	225 *	225 *
7	10	mc	Meningioma	3.49	20	Daily	2.25	45	45.9	535	540 ¶
8	10	mc	Meningioma	2.35	20	Daily	2.25	45	45.9	515	520 ¶
9	11	fs	Meningioma	7.70	20	Daily	2.25	45	45.9	510	515 ¶
10	14	fs	Meningioma	1.10	14	Daily	2.9	40.6	43.6	91	91
11	14	mc	Meningioma	2.35	20	Daily	2.25	45	45.9	174	174
12	13	mc	Pituitary gland tumour	2.55	10	Daily	4	40	46.7	839 *	839 *
13	16	mc	Pituitary gland tumour	0.72	10	Daily	4.2	42	49.7	55	55
14	13	mc	Pituitary gland tumour	1.17	14	Daily	3	42	45.5	70	70
15	12	fs	Pituitary gland tumour	0.82	10	Daily	4.2	42	49.7	644 *	644 *
16	7	m	Pituitary gland tumour	1.42	17	Daily	2.5	42.5	44.3	76 *	76 *
17	12	mc	Pituitary gland tumour	1.20	14	Daily	3	42	45.5	775	1056 †
18	12	mc	Pituitary gland tumour	2.10	14	Daily	3	42	45.5	144	144 ¶
19	11	fs	Pituitary gland tumour	1.44	20	Daily	2.25	45	45.9	99	99
20	12	mc	Choroid plexus tumour	0.17	3	MWF	8	24	36.0	558 *	558 *
21	15	fs	Choroid plexus tumour	1.60	10	Daily	4	40	46.7	265	276 ¶
22	14	fs	Glioma	1.01	10	Daily	4.2	42	49.7	575 *	575 *

m = male; mc = male castrated; fs = female spayed; GTV = gross tumour volume; EQD₂ = equivalent dose delivered in 2 Gy fractions; MWF = Monday-Wednesday-Friday

*

Cats free of progression and alive at the time of data evaluation or lost to follow-up were censored

†

Cats were re-irradiated at the time of progression and censored from analysis of OS at the start of the second course of radiation therapy

‡

Tumour type is based on advanced imaging characteristics

¶

Died of tumour- or treatment-related causes

Treatment

Twenty-one patients were treated with various definitive-intent daily-fractionated protocols (Table 1). One patient (Table 1, case 20) was treated with a highly conformal definitive-intent hypofractionated radiation protocol (3 × 8 Gy) delivered every other day and received less than 40.00 Gy of total dose. Total doses ranged from 24.00–45.00 Gy, with a mean total dose of 41.63 Gy (± 4.33). Fraction sizes ranged from 2.25–8.00 Gy, with a mean of 3.32 Gy (± 1.29). Twenty-one cats were treated using a daily (Monday-Friday) treatment schedule. Only case 20 (Table 1) was treated on a Monday-Wednesday-Friday schedule. All cats completed the full course of the planned radiation therapy. The respective EQD₂ were 36.00–49.70 Gy₂, with a mean of 45.79 Gy₂ (± 2.74). Mean overall treatment time was 19.6 days (5–29 days).

Pretreatment corticosteroid treatment with prednisolone was given to 20/22 cats with a mean dose of 0.66 mg/kg (± 0.41) q24h PO and tapered to mean doses of 0.58 mg/kg (± 0.43) q24h PO post-RT (P = 0.076). Three cats received antiepileptic treatment consisting of either phenobarbital (n = 2) or levetiracetam (n = 1) pre-, during and post-RT. Two of these cats had epileptic seizures at initial presentation and one received phenobarbital prophylactically and was tapered off the medication after RT, as no epileptic seizures were observed.

Outcome and prognostic factors

Median follow-up time for censored cases (n = 8, still alive or lost to follow-up) was 575 days (95% CI: 464–686). Fourteen of the 22 cats died during the study period. Death was attributed to the brain tumour or possible treatment-related toxicity in 5/14 cats (35.7%). In the 64.3% of animals that died due to other causes, death was related to other neoplastic disease (n = 4, histiocytic sarcoma in the liver [Table 1, case 4], lymphosarcoma in the bladder [Table 1, case 5], oral sarcoma [Table 1, case 10], intestinal lymphoma [Table 1, case 13]), and tumour-unrelated medical problems (not accompanied with neurological symptoms) such as hyperthyroidism (Table 1, case 11), cardiac failure with dilated atria, arrhythmia and mild pericardial effusion (Table 1, case 14) or pyothorax (Table 1, case 19). In two cases (Table 1, cases 3 and 17) the cats died at home with general weakness. Tumour recurrence could not be ruled out because no necropsy was performed, so the real cause for death remains unknown in these two cases.

Based on medical records and owner communication, clinical and neurological status improved in 21/22 (95.5%) cats, compared with the pre-treatment complaints. One of the 22 cats was judged to have stable clinical signs (Table 1, case 18). In the patient in which epileptic seizures was the only presenting complaint, it was difficult to determine whether seizure control was attributable to reduction of tumour/oedema or to anti-epileptic medication.

A total of 10 post-RT diagnostic imaging examinations were performed in six patients. CT was performed four times (n = 3 cats), and MRI was performed six times (n = 3 cats). On MRI, one cat with a meningioma showed partial remission at 6 months post-RT (Table 1, case 1), a second cat with a meningioma showed stable disease at 3, 6 and 12 months post-RT (Table 1, case 6). Another cat with a pituitary gland tumour showed stable disease on MRI at 6 and 12 months post-RT (Table 1, case 12). In the CT of one cat 6 months post-RT, the tumour was in partial remission (the 12 month re-check CT of the same cat was made at another clinic and tumour volume was not measured) (Table 1, case 15). Tumours of two cats were measured 24 months post-RT; one cat still showed stable disease (Table 1, case 17) and the other cat was scanned upon re-occurrence of neurological signs and was found to have a disease progression on CT (Table 1, case 3). Overall, seven cats (31.8%) were considered to have a disease progression either of increased volume based on imaging (n = 1) or of progression of neurological signs (n = 6).

The median PFS time for all 22 cases was 510 days (95% CI: 51–969). The proportion free of progression at 1 year was 55.7% (95% CI: 33–78) (Figure 1). Of the nine cats that died within the first year after RT, seven (77.8%) died of causes unrelated to the brain tumour (Table 1, cases 4, 5, 10, 11, 13, 14 and 19) and two (22.2%) died because of the brain tumour or treatment-related toxicity (Table 1, cases 18 and 21). No difference was seen in PFS between the patients treated at the different institutions (P = 0.634). At the time of tumour progression, two cats were re-irradiated, resulting in an additional progression-free interval of 223 and 281 days (Table 1, cases 3 and 17). Both cases were censored from analysis of OS at the start of the second course of radiation therapy. Neither the patient-related (age, sex), tumour-related (meningioma/non-meningioma, meningioma/pituitary tumour, tumour size, tumour location) nor treatment-related parameters (dose of radiation [total radiation dose or EQD₂], corticosteroid dose at first RT, overall treatment time, treatment institution) were associated with PFS.

OPEN IN VIEWER

Figure 1

Median progression-free survival (PFS). The dotted line marks 1 year. Tick marks: cats that were free of progression of measurable tumour and of progression of neurological signs and that were alive at the time of data collection were censored for PFS analysis. Of the nine cats that died within the first year after radiotherapy, seven (77.8%) died of causes unrelated to the brain tumour and two (22.2%) died because of the brain tumour or treatment-related toxicity

Median OS was 515 days (95% CI: 66–964). The proportion alive at 1 year was 56% (95% CI: 33–79) and at 2 years 37% (95% CI: 15–60) (Figure 2). No difference was seen in OS between the patients treated at the different institutions (P = 0.695). None of the tested variables had a statistically detectable influence on OS.

OPEN IN VIEWER

Figure 2

Median overall survival (OS): the dotted line marks 1 year. Tick marks: cats that were still alive at the time of data collection or lost to follow-up were censored at the last date of contact