ABSTRACT
Background
Corticosteroids (CS) are a common treatment prescribed for dogs with brain tumors, but little data characterizing the clinical and neuroradiologic effects of CS treatment exist.
Hypotheses
Decreases in peritumoral brain edema (PBE) would be observed in dogs with brain tumors treated with CS and anticonvulsants, and decreases in edema would be accompanied by clinical improvement.
Animals
Fifty dogs with gliomas and 22 with meningiomas.
Materials and Methods
Retrospective case series. Dogs with brain tumors receiving treatments with CS and anticonvulsants underwent pre‐ and posttreatment clinical and brain magnetic resonance imaging (MRI) examinations within a 3‐week follow‐up period, and peritumoral edema and tumor volumes were calculated from each MRI study. Dogs were characterized as clinical responders or nonresponders independent of imaging results based on longitudinal changes in clinical findings and owner‐reported quality of life (QOL). Clinicopathologic and MRI variables were compared pre‐ and posttreatment by clinical response.
Results
At follow‐up, 23/50 (46%) of dogs with glioma and 15/22 (68%) with meningioma were classified as clinical responders, and 23/50 (46%) of gliomas and 14/22 (64%) of meningiomas had decreases in edema volume. Responders had significantly larger decreases in edema and mass effect than nonresponders. Decreases in tumor volumes occurred in approximately 25% of gliomas.
Conclusions and Clinical Importance
Peritumoral brain edema was decreased in 50%–60% of gliomas and meningiomas in dogs at follow‐up, and amelioration of edema often was accompanied by improved neurological signs and QOL. Corticosteroids may also clinically benefit dogs without PBE and can influence MRI surrogates used to determine glioma therapeutic responses.
Keywords: glioma, meningioma, neurooncology, prednisone
Abbreviations
- AE
adverse event
- C‐
complete response
- CEV
contrast enhancing tumor volume
- CS
corticosteroids
- FLAIR
fluid attenuated inversion recovery MRI sequence
- GLV
gross lesion volume
- KPS
Karnofsky Performance Score
- MRI
magnetic resonance imaging
- PBE
peritumoral brain edema
- PBEV
peritumoral brain edema volume
- PR
partial response
- QOL
quality of life
- ROI
region of interest
- SD
stable disease
- T1W
T1‐weighted MRI sequence
- T2W
T2‐weighted MRI sequence
- TTV
total T2W tumor volume
1. Introduction
Corticosteroids (CS) have been prescribed as a principal component of palliative care for humans and animals with brain tumors for over seven decades [1, 2, 3, 4, 5]. Peritumoral brain edema (PBE) is a well‐recognized and important cause of tumor‐associated neurological dysfunction, and CS are the primary and most effective pharmacologic intervention currently used to manage clinical signs of PBE, as well as to prevent and treat encephalopathic complications of radiation therapy [5, 6, 7]. Outside of their cytotoxic effects on malignant lymphocytes in central nervous system lymphomas, CS are not believed to have any substantial direct antineoplastic effects on other primary brain tumors [6, 8]. However, transient marked decreases to complete disappearance of contrast‐enhancing tumor burdens in brain imaging studies after treatment with CS have been rarely documented in humans with malignant gliomas [9, 10, 11, 12, 13, 14, 15, 16, 17], as well as in experimentally induced gliomas in dogs [18]. This process has been referred to as the “vanishing glioma” or steroid‐induced pseudoregression phenomenon [9, 10, 11, 12, 13, 14, 15, 16, 17].
Despite the long history of pervasive use of CS in the management of brain tumors in veterinary medicine, little evidence exists to document the frequency with which PBE is objectively decreased after CS treatment in specific tumor types, associations between decreases in PBE and improvement in neurologic signs, or dog or tumor specific variables that may contribute to the antiedema therapeutic response or lack thereof [1, 2, 3, 5]. Because some degree of PBE can be visualized in approximately 85% of imaging studies of brain tumors in dogs, such as meningiomas and gliomas, it is not surprising that existing evidence and conventional wisdom suggest that CS treatment will improve clinical signs in some dogs with PBE [1, 2, 3, 5, 18]. One study demonstrated that CS treatment was associated with quantitative decreases in PBE on serial magnetic resonance imaging (MRI) examinations in 44% of dogs with presumptively diagnosed meningiomas [5]. Limited data also exist indicating that gliomas in dogs may undergo modest volumetric decreases in tumor size and improvement in PBE after CS administration [2, 18].
The objectives of our study were to determine the frequency of acute improvements in neurological signs and MRI features in dogs with brain tumors receiving palliative treatment with CS and anticonvulsants that underwent serial clinical and neuroimaging examinations, characterize clinical and imaging phenotypes of any observed follow‐up changes, and identify variables that were associated with any observed clinical or imaging changes. We hypothesized that in dogs with intracranial gliomas and meningiomas treated with CS and anticonvulsants: (1) quantitative decreases in MRI‐derived PBE volumes (PBEV) would be observed after treatment; (2) these quantitative decreases in PBEV would be associated with improvement in tumor‐associated clinical signs; and (3) quantitative decreases in the MRI‐derived contrast‐enhancing (CEV) or total T2‐weighted tumor (TTV) volumes of some gliomas in dogs would be observed after treatment.
2. Materials and Methods
Ours was a single‐center, retrospective case series. The institutional medical records database was searched from January 2014 through December 2021 to identify dogs with histologically confirmed brain tumors. For inclusion, dogs had to have clinical signs of intracranial disease, an initial baseline brain MRI demonstrating a solitary mass lesion with imaging features consistent with a meningioma or glioma [19], a physical examination performed by a neurologist, and receive a course of CS started during or immediately after the initial MRI that was continued for at least 2 weeks. Dogs that received CS before the initial MRI tumor diagnosis were excluded. Subsequently, dogs had to undergo follow‐up clinical and brain MRI examinations performed by a neurologist within 3 weeks of the initial MRI while receiving treatments that were limited to CS and anticonvulsants. Clinical and MRI examinations obtained acutely after initiation of CS treatment were performed for planning of tumor stereotactic biopsy, surgical resection, radiotherapy treatment, or as a new baseline scan upon enrollment in investigational brain tumor clinical trials [20, 21, 22]. Upon scheduling of their follow‐up appointment, dog owners were asked to complete a quality of life (QOL) questionnaire (CanBrainQOL‐24) [23] that focused on QOL for the 2 weeks before initial diagnosis of the brain tumor. The owners completed a second QOL survey at the time of their follow‐up appointment that focused on QOL since the CS were prescribed. The follow‐up QOL assessment included an additional question that asked owners to rate their dog's overall QOL as improved, unchanged, or worse than before starting CS treatment.
Because initial brain MRI were obtained from multiple veterinary practices, image acquisition protocols were not standardized, but all initial MRI were obtained using 1.5 T systems and included at least the following sequences: T1‐weighted (T1W) pre‐ and postcontrast images in at least two planes, T2‐weighted (T2W) images in two planes, single planar fluid attenuated inversion recovery (FLAIR), and T2W gradient echo images. All follow‐up MRI were obtained at a single institution using a 1.5 T superconducting magnet (Philips Intera, Andover, MA) and a standardized image acquisition protocol [24, 25]. In selected cases, additional sequences were obtained upon the request of the attending neurologist or radiologist, or as required for specific clinical trial protocols.
From each dog's MRI at each time point, four MRI‐derived lesion volumes (Figure 1) were quantified by investigators blinded to the dog's clinical status: gross lesion volume (GLV), total T2W tumor volume (TTV), contrast‐enhancing tumor volume (CEV), and peritumoral brain edema volume (PBEV). The GLV was defined as the entire extent of all abnormal T2W/FLAIR signal intensity and was segmented using a manually drawn region of interest (ROI) on contiguous T2W transverse images (Figure 1). The GLV was calculated with the ROI‐volume function of commercial image analysis software (Osirix MD, v11.0.4, Pixmeo, Switzerland). The TTV and CEV were determined using previously described methods [21, 25]. Briefly, manually defined ROIs representing the TTV were generated on contiguous transverse T2W image slices, and TTV was calculated with the Osirix ROI‐volume function [21, 25]. The CEV was calculated using the dT1 method with the IB Neuro Delta Suite plug‐in software (v 20.10, Imaging Biometrics, Milwaukee, WI) for Osirix [25]. Because not all gliomas demonstrated contrast‐enhancement, the PBEV was calculated by subtracting the TTV from the GLV, and by subtracting the CEV from the GLV in dogs with meningiomas (Figure 1), and then manually evaluating the PBEV segmentation using simultaneous comparisons of FLAIR, T2W, and postcontrast T1W images. PBE indices (PBEV/TTV for gliomas; PBEV/CEV for meningiomas) also were calculated for each time point.
FIGURE 1.
Representative quantitative tumor habitats from dogs with a high‐grade oligodendroglioma (top two rows) and a Grade I meningioma (bottom two rows). For each tumor type, the top row of panels depicts 2D MRI sequences used to segment the various tumor volumes, and the bottom row of panels illustrate regions of interest for the gross lesion volume (GLV; light orange), total T2 tumor volume (TTV; red), peritumoral brain edema volume (PBEV; blue), and contrast‐enhancing tumor volume (CEV; green). The GLV is equivalent to the TTV + PBEV (second row bottom left panel, inset).
To account for interobserver measurement variability, an MRI volumetric parameter difference > ± 10% from the initial to follow‐up MRI was set as a conservative threshold that defined objective, detectable change [25]. Tumor therapeutic responses were classified according to volumetric criteria as described previously [21, 25]. A complete response (CR) required the elimination of all enhancing tumor or complete disappearance of all T2W/FLAIR lesions. Partial responses (PR) were characterized by ≥ 65% decreases in TTV, CEV, or both. Progressive disease (PD) was considered present if TTV, CEV, or both increased by ≥ 40%. All other responses constituted stable disease (SD) [22, 25].
In addition to the TTV, CEV, PBEV, and edema indices, other variables potentially associated with treatment responsiveness were extracted from medical records including: breed, age, sex, tumor type and grade, tumor location, CBC, and serum biochemical profile results, and days elapsed between performance of the first and second brain MRI. Tumor locations and grades were classified using published criteria specific for gliomas and meningiomas [19, 20, 21, 22, 26, 27, 28, 29]. For the initial hospital and follow‐up visits, clinical signs, Karnofsky performance scores (KPS), neurological examination abnormalities, CanBrainQOL‐24 scores, and adverse events (AE) were recorded [20, 21, 22, 23]. AE were classified and graded according to Veterinary Cooperative Oncology Group Common Terminology Criteria [30]. Clear recurrence, exacerbation, or progression of initial neurological signs were not considered AE. The types, dosages, and tapering schedules (starting dose, time to start of tapering after initial MRI, and dose and frequency at time of follow‐up MRI) of prescribed CS also were recorded. Because the routes of administration, doses, or frequency of initial in‐hospital dexamethasone treatments were not always available in medical records, dexamethasone use at the time the initial MRI was performed was recorded qualitatively (yes or no). Records containing dexamethasone dosing data (n = 20) indicated that it was administered IV at a dosage range of 0.1–0.3 mg/kg.
Each dog was categorized by a single investigator blinded to the MRI findings as a responder or nonresponder, based on a multidimensional assessment of clinical outcomes relative to baseline findings. At follow‐up, responders were required to have an overall owner QOL reported as improved and neurological examination findings that were static or improved compared to baseline, with all other dogs classified as nonresponders.
2.1. Statistical Analyses
Descriptive data utilized to define responders and nonresponders (KPS, CanBrainQOL‐24 scores, neurological examination abnormalities, and AE) were summarized but not subjected to analyses. Distributions of continuous variables were assessed using normal probability plots. Subsequently, continuous data were summarized as median (interquartile range, [IQR]) and categorical variables as counts and percentages. To test the hypothesis that quantitative decreases in PBEV would be associated with improvement in clinical signs, bivariable associations between clinical response outcome (responder or nonresponder) and prognostic variables were assessed using Fisher's exact or Wilcoxon rank sum tests. Variables identified on bivariable analysis as significant (p < 0.1) were included in a multivariable logistic regression model and subjected to a stepwise variable reduction algorithm, with the p value to enter and p value to stay set to 0.05.
To test the hypothesis that glioma volumes would be decreased on follow‐up MRI compared to the initial MRI, changes in MRI variables from baseline were tested for statistical significance while comparing between tumor types and clinical response status using linear generalized estimating equations with Bonferroni correction for multiple comparisons. Statistical significance was set to p < 0.05. All analyses were performed using SAS version 9.4 (Cary, NC, USA).
3. Results
Seventy two dogs were included, 50/72 (69%) with gliomas and 22/72 (31%) with meningiomas (Figure 2). Among dogs with gliomas, 32/50 (64%) were oligodendrogliomas (22/32 [69%] high‐grade and 10/32 [31%] low‐grade), 16/50 (32%) were astrocytomas (13/16 [81%] high‐grade and 3/16 [19%] low‐grade), and 2/50 (4%) were high‐grade undefined gliomas. Breeds of dogs with gliomas included: Boxer (14), Boston Terrier (nine), French Bulldog (six), Staffordshire Terriers (four), English Bulldog (three), mixed breeds (three), and 11 other breeds represented by a single case. Glioma location was hemispheric in 45/50 (90%), diencephalic in 3/50 (6%), and infratentorial in 2/50 (4%) cases. Hemispheric gliomas included fronto‐olfactory (25/45; 56%), parietal (9/45; 20%), temporal‐piriform (10/45; 22%), and occipital (1/45; 2%). Breeds of dogs with meningiomas included: mixed breed (three), Golden retriever (three), Labrador retriever (two), Staffordshire terrier (two), and 12 other purebred dogs. Meningioma locations included the cerebral convexities (10/22; 45%), olfactory (6/22; 27%), parasellar (2/22; 9%), cerebellopontine angle (1/22; 5%), cerebellar convexity (1/22; 5%), parasagittal (1/22; 5%), and falcine (1/22; 5%). Eighteen (82%) meningiomas were Grade 1 tumors, and 4/22 (18%) meningiomas were Grade 2. A summary of dog and tumor characteristics by tumor type at initial diagnosis is provided in Table S1.
FIGURE 2.
Flowchart of medical records review for inclusion of dogs in the retrospective study. Red colored text highlights the final population of dogs that met the inclusion criteria for the study.
Upon discharge after the initial MRI, dogs were treated with PO prednisone (70/72; 97%) or prednisolone (2/72; 3%). A total of 38/72 (53%; 95% confidence interval [CI], 41%–65%) of dogs were classified as clinical responders and 34/72 (47%; 95% CI, 35%–59%) as nonresponders to treatment (Table 1). In responders, the neurological examination abnormalities that demonstrated the highest frequencies of improvement or resolution at follow‐up included proprioceptive deficits, thalamocortical visual deficits, and gait abnormalities (Table 1). At least one AE was documented in 68/72 (94%) dogs (Table 1), and AEs probably or definitely attributable to CS or anticonvulsants accounted for all AEs in 62/68 (91%) of dogs. Grade 1 and 2 AEs probably or definitely attributable to CS or anticonvulsants occurred in 80%–90% of both responders and nonresponders, and 54/68 (79%) of dogs experienced ≥ 2 AEs attributable to prescribed medications.
TABLE 1.
Descriptive summary of clinical outcomes in 72 dogs with brain tumors following corticosteroid and anticonvulsant treatment.
| Brain tumor type (frequency, %) | ||||
|---|---|---|---|---|
| Glioma 50/72 (69%) | Meningioma 22/72 (31%) | |||
| Clinical response | Responders | Nonresponders | Responders | Nonresponders |
| Category (frequency, %) | 23/50 (46%) | 27/50 (54%) | 15/22 (68%) | 7/22 (32%) |
| Δ Karnofsky performance score from baseline (median, range) | 10 (0–40) | −10 (−20 to 0) | 10 (0–20) |
0 (−30 to 0) |
| Δ CanBrainQOL‐24 score from baseline (median, range) | −9 (−1 to −35) |
3 (−3 to 24) |
−3 (−11 to 1) |
5 (−3 to 10) |
| Experienced seizure(s) between baseline and posttreatment exam (frequency, % among all dogs with seizures) | 2/18 (11%) | 9/27 (33%) | 2/15 (13%) | 4/7 (57%) |
| Neurological abnormalities that improved or resolved (frequency, % total present in responders) | ||||
|
6/8 (75%) | 7/9 (77%) | ||
|
2/7 (29%) | NA | 3/6 (50%) | NA |
|
1/4 (25%) | 0/2 (0%) | ||
|
1/1 (100%) | 1/1 (100%) | ||
|
1/2 (50%) | 1/2 (50%) | ||
|
16/19 (84%) | 11/12 (92%) | ||
|
1/2 (50%) | 0/1 (0%) | ||
|
4/5 (80%) | 3/3 (100%) | ||
|
0/2 (0%) | 0/1 (0%) | ||
|
0/0 (0%) | 0/1 (0%) | ||
| Adverse events (frequency, %) a | ||||
|
3/23 (13%) | 1/27 (4%) | 2/15 (13%) | 0 |
|
11/23 (48%) | 16/27 (59%) | 10/15 (67%) | 5/7 (72%) |
|
9/23 (39%) | 9/27 (33%) | 3/15 (20%) | 1/7 (14%) |
|
0 |
1/27 (4%) |
0 | 1/7 (14%) |
Abbreviation: NA = not applicable.
Dogs categorized based on highest AE grade present; more than one AE could be present in each dog.
The most common Grade 1 or 2 AE attributable to CS or anticonvulsants experienced by dogs included polyuria and polydipsia (57/68; 84%), polyphagia (44/68; 65%), increases in serum alkaline phosphatase activity (39/68; 57%), increases in serum alanine aminotransferase activity (25/68; 37%), ataxia (20/68; 29%) and diarrhea (18/68; 26%). One dog with a meningioma experienced Grade 3 gastric ulceration and hemorrhage attributable to CS. On the follow‐up QOL surveys, the item that specifically evaluated owner perceptions of AE of prescribed treatment received a worse score compared to baseline in 66/72 (92%) of dogs.
Bivariable analysis (Table 2) identified dexamethasone use, tumor type, tumor location, baseline QOL score, baseline KPS score, baseline TTV, baseline CEV, baseline tumor: brain volume, baseline PBEV, baseline edema index, posttreatment change in CEV, posttreatment change in PBEV, and posttreatment edema index as significantly different between responders and nonresponders, and these confounders were included in multivariable models, in which responders were the modeled outcome. Initial modeling identified significant associations between PBEV and edema indices, and thus edema indices were removed and only PBEV included in subsequent models. The baseline PBEV and posttreatment change in PBEV were significantly different between responders and nonresponders (Table 3) in the final multivariable model, with responders having a larger baseline PBEV and larger decrease in PBEV after treatment (Table 3; Figure 3A,B). A total of 23/50 (46%) of dogs with glioma and 14/22 (64%) of dogs with meningioma had a detectable decrease in PBEV at follow‐up (Table S2).
TABLE 2.
Bivariable analysis of clinical and brain tumor characteristics in 72 dogs by clinical outcome after corticosteroid and anticonvulsant treatment.
| Clinical outcome | p | ||
|---|---|---|---|
| Responders 38/72 (53%) | Nonresponders 34/72 (47%) | ||
| Categorical explanatory variables (frequency, %) | |||
| Sex | 0.88 | ||
|
2/38 (5%) | 1/34 (3%) | |
|
16/38 (42%) | 16/34 (47%) | |
|
1/38 (3%) | 2/34 (6%) | |
|
19/38 (50%) | 15/34 (44%) | |
| Clinical signs | 0.17 | ||
|
10/38 (26%) | 12/34 (35%) | |
|
9/38 (24%) | 4/34 (12%) | |
|
18/38 (50%) | 18/34 (53%) | |
| Dexamethasone use | 0.001* | ||
|
17/38 (45%) | 28/34 (82%) | |
|
21/38 (55% | 6/34 (18%) | |
| Tumor type | 0.08* | ||
|
15/38 (39%) | 7/34 (21%) | |
|
23/38 (61%) | 27/34 (79%) | |
| Tumor grade | 0.82 | ||
|
18/38 (47%) | 15/34 (44%) | |
|
20/38 (53%) | 19/34 (56%) | |
| Tumor location | 0.09* | ||
|
13/38 (34%) | 20/34 (59%) | |
|
21/38 (55%) | 13/34 (38%) | |
|
4/38 (11%) | 1/34 (3%) | |
| Continuous explanatory variables (median, IQR) | |||
| Age (years) | 8 (7–10) | 8 (7–10) | 0.69 |
| Body weight (kg) | 24 (18–37) | 17.5 (11–26) | 0.17 |
| Duration of clinical signs (days) | 42 (27–59) | 48 (24–72) | 0.40 |
| Baseline CanBrainQOL‐24 score | 52.5 (45–62) | 44.5 (42–49) | 0.001* |
| Baseline Karnofsky performance score | 70 (70–80) | 90 (80–90) | 0.0004* |
| Baseline total T2W tumor volume (TTV; cm [3]) | 5.6 (3.5–10.3) | 2.4 (1.3–3.7) | 0.0001* |
| Baseline contrast‐enhancing tumor volume (CEV; cm [3]) | 2.8 (1.1–7.6) | 0.7 (0.2–1.9) | 0.001* |
| Baseline tumor; brain volume | 0.1 (0–0.1) | 0 (0–0.1) | 0.0002* |
| Baseline peritumoral brain edema volume (PBEV; cm [3]) | 2.9 (1.5–3.9) | 0.8 (0.2–1.7) | 0.001* |
| Baseline edema index | 0.4 (0.3–0.5) | 0.3 (0.1–0.5) | 0.09* |
| Prednisone starting dose | 0.8 (0.7–0.9) | 0.8 (0.7–0.8) | 0.76 |
| Timing of posttreatment MRI (days) | 15.5 (12–18) | 17 (14–19) | 0.42 |
| Posttreatment Δ TTV (%) | 2 (−0.8 to 6) | 3 (1–8) | 0.13 |
| Posttreatment Δ CEV (%) | 0 (−4 to 3) | 3 (0–9.5) | 0.01* |
| Posttreatment Δ PBEV (%) | −63 (−80 to −37) | 0 (−11 to 16) | < 0.0001* |
| Posttreatment edema index | 0.1 (0.1–0.2) | 0.3 (0.1–0.5) | 0.02* |
Abbreviation: IQR = interquartile range.
Included in multivariable model when p < 0.1.
TABLE 3.
Multivariable model of brain tumor characteristics identified as significantly associated with clinical responders after corticosteroid and anticonvulsant treatment.
| Variable | Coefficient (β) | Standard error | Wald X [2] | p | Odds ratio | 95% confidence interval |
|---|---|---|---|---|---|---|
| Intercept | −7.28 | 2.45 | ||||
| Baseline Peritumoral Brain Edema Volume (PBEV) | 0.122 | 0.048 | 6.15 | 0.01* | 1.52 | 1.05–1.22 |
| Posttreatment Δ PBEV | −0.061 | 0.012 | 18.51 | < 0.0001* | 0.75 | 0.72–0.78 |
Statistically significant, p < 0.05.
FIGURE 3.
MRI changes in brain tumors at follow‐up. The top row of panels represent initial images and the bottom row postCS treatment images. (A) American Eskimo dog classified as responder with Grade I cystic olfactory meningioma that demonstrated detectable reduction in peritumoral brain edema (arrow) volume (PBEV; −64%), with no changes in total T2W tumor volume (TTV) or contrast‐enhancing tumor volume (CEV; (insets, postcontrast T1W images) at follow‐up. (B) Mixed breed dog classified as responder with high‐grade astrocytoma in the temporal‐piriform lobe that experienced detectable reductions of TTV (−24%) and PBEV (arrow; −72%), as well as improvement in midline shift at follow‐up. (C) Staffordshire terrier classified as responder with fronto‐olfactory high‐grade oligodendroglioma that demonstrated no detectable change in TTV but an 89% reduction in CEV. (D) French bulldog with frontal high‐grade oligodendroglioma categorized as nonresponder illustrating discordant tumor responses, with increases in TTV (+12%) and PBEV (+17%), but a substantial reduction in CEV (−44%).
Detectable changes in TTV or CEV in dogs with meningiomas (Tables 4 and S2; Figure 3A) were not observed at follow‐up. Among dogs with glioma, significant differences in TTV and CEV were found between responders and nonresponders at follow‐up (Table 4). Two nonresponder dogs with glioma met objective criteria for PD at the follow‐up evaluation. On follow‐up MRI, 10 individual glioma responders and two nonresponders had detectable decreases in TTV or CEV (Table S2), and all decreases in TTV and CEV were observed in high‐grade gliomas. Modest decreases in TTV (> 10%–≤ 25% volumetric decreases) were observed in 4/23 (17%) of glioma responders (Table S2; Figure 3B) and 1/27 (4%) of glioma nonresponders. Detectable decreases in CEV were present in 7/23 (30%) of glioma responders, with the degree of CEV decrease highly variable among cases (median, −26%; range, −17% to −89%; Figure 3B,C), and a detectable decrease in CEV also was present in 1/27 (4%) glioma nonresponders. In responders, detectable decreases in follow‐up MRI tumor volumes were observed in a single variable, either TTV or CEV (Figure 3B,C) in 9/10 dogs, with one dog having a detectable decrease in both TTV and CEV. Both nonresponders with volumetric tumor changes had discordant volumetric responses, defined as a CEV or TTV that detectably decreased with a concomitant increase in TTV or CEV, respectively (Figure 3D).
TABLE 4.
Quantitative MRI tumor volumetric characteristics of 72 dogs treated with corticosteroids and anticonvulsants by tumor type and clinical response.
| Brain tumor type (frequency, %) | ||||||
|---|---|---|---|---|---|---|
| Glioma 50/72 (69%) | Meningioma 22/72 (31%) | |||||
| Clinical response | Responders | Nonresponders | p | Responders | Nonresponders | p |
| Category (frequency, %) | 23/50 (46%) | 27/50 (54%) | 15/22 (68%) | 7/22 (32%) | ||
| Quantitative MRI parameters | ||||||
| Baseline total T2W tumor volume (TTV cm [3]; median, IQR) | 2.9 (2.1–4.6) | 3.3 (1.9–5.0) | 0.74 | 5.3 (2.8–8.8) | 6.1 (3.1–10.4) | 0.69 |
| Baseline contrast‐enhancing tumor volume (CEV cm3; median, IQR) | 1.28 (0.46–1.9) | 0.89 (0.29–1.61) | 0.59 | 5.3 (2.6–8.7) | 5.9 (2.9–10.2) | 0.63 |
| Baseline number of tumors with midline shift (frequency, %) | 15/23 (65%) | 21/27 (77%) | 0.71 | 12/15 (80%) | 5/7 (71%) | 0.66 |
| Posttreatment number of tumors with midline shift | 8/23 (35%) | 20/27 (74%) | 0.03* | 7/15 (47%) | 5/7 (71%) | 0.04* |
| Posttreatment Δ% TTV (median, IQR) | −0.3 (−14 to 2) | 4 (1–8) | 0.02* | 1.3 (0.2–9.4) | 1.2 (−0.7 to 4.1) | 0.22 |
| Number of contrast‐enhancing tumors (frequency, %) a | 20/23 (87%) | 19/27 (70%) | 0.29 | 15/15 (100%) | 7/7 (100%) | 1.0 |
| Posttreatment Δ% contrast‐enhancing tumor volume (CEV) (median, IQR) | −13 (−29 to 3) |
9.5 (−1.6 to 33) |
0.04* | 1.5 (−0.9 to 6.6) | 2.5 (−1.6 to 3.8) | 0.84 |
The number of contrast‐enhancing tumors in each group was unchanged between pre‐and posttreatment MRI examinations.
Statistically significant, p < 0.05.
4. Discussion
In dogs with brain tumors, CS are frequently prescribed to treat clinical signs associated with brain edema, which contributes to mass effect and intracranial hypertension [2, 3, 5]. Our results support previous observations and our hypothesis that PBE is decreased in many dogs with meningiomas and gliomas, the two most common types of primary brain tumors in dogs that receive treatment with CS and anticonvulsants [3, 5]. Tumor‐bearing dogs classified as responders that experienced improvements in clinically relevant outcomes, such as neurological function and caregiver reported QOL at follow‐up, had larger edema burdens at diagnosis and experienced significantly larger decreases in quantitative indicators of PBE and qualitative mass effect than nonresponders. Although dramatic and complete disappearance of tumor volumes on follow‐up MRI, such as have been reported in “vanishing glioma” cases in humans, were not observed in dogs in our study, modest decreases in TTV and modest to substantial decreases in CEV, consistent with the tumor pseudoregression phenomenon, did occur in selected cases of high‐grade glioma [9, 10, 11, 12, 13, 14, 15, 16, 17, 18]. Our study also demonstrated that acute treatment with CS and anticonvulsants had no substantial effect on the TTV or CEV of dogs with meningiomas, which also has been reported in humans [8].
Peritumoral brain edema was present on initial MRI in 42/50 (84%) of dogs with gliomas and 19/22 (86%) of dogs with meningiomas, similar to the frequencies that previous studies have identified PBE in these tumor types [19, 26, 27]. In dogs with PBE on initial MRI, objective decreases in the edema burden in 23/42 (55%) of dogs with gliomas and 74% (14/19) of dogs with meningiomas were observed at follow‐up. Furthermore, the majority of dogs with glioma (20/23; 87%) and meningioma (13/15; 90%) classified as responders had substantial decreases of PBEV, often accompanied by resolution of midline shifts, whereas progression of PBE was noted in 10/27 (37%) glioma and 1/7 (14%) meningioma dogs considered nonresponders. In humans with brain tumors, discordant and variable results are observed with respect to the relative efficacy of CS at quantitatively decreasing PBE, with some studies indicating that CS treatment has no significant effect on PBEV in people with gliomas, meningiomas, or brain metastases [31, 32, 33], and others observing quantitative decreases in edema burdens of glioma and brain metastases, but not meningiomas [34, 35].
Among responders, the neurological deficits present in several dogs with each tumor type that demonstrated the highest frequencies of improvement included proprioceptive deficits, gait abnormalities, and thalamocortical visual dysfunction. One possible explanation for this observation is that the projection tracts for general proprioception and the visual pathway are located in the internal capsule and optic radiations, respectively, subcortical white matter structures that are commonly affected by vasogenic PBE associated with forebrain tumors [2, 3, 19]. Conversely, neurological deficits that had lower observed frequencies of or no improvement, such as head and body postural abnormalities and subcortical visual deficits, often were attributed to the presence of tumor compression or infiltration of brain structures involved in those neuroanatomic pathways, such as the vestibular nuclei and caudal cerebellar peduncle in the case of a cerebellopontine angle meningioma, or the optic tract with a diencephalic glioma.
Our study and others illustrate the challenges associated with systematically evaluating and determining the clinical relevance of various phenotypic manifestations of brain tumors across diverse dog and human caregiver populations, and particularly the contribution of structural epilepsy to the holistic disease burden and objective therapeutic response assessment [1, 2, 5, 23]. Although > 80% of the dogs reported here had structural epilepsy, the acute follow‐up period inherent to the study design and uncontrolled use of anticonvulsant drugs precluded robust and objective evaluation of endpoints of seizure treatment success such as seizure freedom and prolongation of the interictal interval. This issue may have resulted in misclassification bias of dogs as nonresponders, particularly in cases in which the sole clinical manifestation of the brain tumor was seizures. We attempted to mitigate this bias by assessing multidimensional aspects of health that included clinician and owner derived measures of clinical response.
Among nonresponders, 42% of dogs with glioma and 72% of dogs with meningiomas had seizures as their only clinical sign, compared to approximately 10% of responders. Additionally, 33% of glioma and 57% of meningioma nonresponders with structural epilepsy experienced seizures between the initial and follow‐up visits, compared with ≤ 13% of responders, despite the relatively short interval between the two assessments. It is known that seizures in and of themselves can result in interictal MRI signal changes in the brain that have been attributed to the presence of cytotoxic or vasogenic brain edema [36]. Thus, it is possible that seizures occurring between the initial and follow‐up MRI could have exacerbated PBEV in some dogs, although ictal‐induced changes in the brain of dogs often result in bilateral and symmetrical changes in cortical gray matter, the hippocampus, and thalamic nuclei, which were not observed in our study [36]. Results of our study further emphasize that seizure control is a substantial contributor to owner‐perceived QOL of dogs with structural epilepsy, and there is a critical need for anticonvulsants that are effective in dogs with tumor‐associated epilepsy [1, 2, 3, 23, 37].
Corticosteroids are known to decrease PBE by upregulation of gene expression and improved function of endothelial tight and adherens junctions that regulate blood–brain barrier integrity and permeability [6]. Although our study substantiates evidence that PBE is an important contributor to clinical morbidity, the observation that two nonresponders also experienced > 50% decrease in PBE also suggests that brain tumor associated neurological deficits can be mediated by multiple pathophysiologic events, such as neuroinflammation, activation of pain pathways, or altered neuronal‐glial signaling [3, 5]. Responders with both tumor types in our study, as well as in another recent investigation of dogs with extra‐axial brain tumors, included dogs with no evidence of PBE [5]. This observation suggests CS may exert beneficial effects via modulation of multiple mechanisms that can participate in tumor‐associated neurological dysfunction, and thus also should be considered for use in symptomatic dogs without imaging evidence of PBE [5, 6]. However, given that 85% of dogs in our study and other studies experienced at least one AE attributable to CS that often negatively impacted their QOL, CS should not be prescribed indiscriminately to dogs with brain tumors [1, 37].
Although marked decreases in TTV or complete disappearance of CEV consistent with the vanishing glioma phenomenon described in people were not observed, we did identify several dogs with glioma that experienced some degree of tumor volumetric decrease at follow‐up, consistent with tumor pseudoregression [8, 9, 10, 11, 12, 13, 14, 15]. This finding has important implications for the management of dogs with glioma, because changes in TTV or CEV are used as objective indicators of glioma therapeutic response, and the presence of tumor enhancement repeatedly has been shown to be an imaging feature significantly associated with high‐grade glioma [3, 19, 21, 22, 25, 27]. Decreases in TTV were observed in 5/50 (10%) of dogs with glioma, although volumetric decreases in these cases were modest (≤ 25%) and did not meet the threshold for PR. Decreases in CEV were seen in 8/50 (16%) of dogs, were generally of larger magnitude than those observed for TTV, and the change was of sufficient magnitude (−89%) in one dog to technically meet criteria for PR [25]. Based on these observations and consensus management practices in humans with glioma, to account for the potential confounding effects of CS on tumor volumes, it is recommended that a new baseline MRI scan be obtained 2 weeks after starting CS and before the administration of other cytoreductive treatments in dogs with glioma [38, 39]. Therapeutic responses then should be assessed by comparison of subsequent MRI scans to the new posttreatment MRI examination.
Our study had some limitations, including the presumption that observed changes in clinical signs, tumor features, and volumes, and PBE burdens were attributable to CS treatment. In humans and dogs, PBE can naturally wax and wane over time and even resolve in the absence of treatment [5]. The CEV of gliomas is also a temporally dynamic imaging tumor surrogate that can increase or decrease spontaneously or secondary to treatment [8, 21, 38, 39]. Thus, causation between CS treatment and changes in neurologic signs, CEV, TTV, or PBEV cannot be established. Because initial MRI were obtained from multiple veterinary practices using different image acquisition protocols, this factor could introduce additional biases in quantitative imaging analyses. We attempted to mitigate this possibility by setting conservative thresholds defining quantitative responses that accounted for interobserver variability. Our study design also did not allow for identification of the ideal CS formulation, dose, dosing frequency, and duration of treatment to mitigate PBE or other tumor‐associated pathophysiologic phenomena, because these factors were not controlled or investigated. Despite widespread use of CS in humans with brain tumors for over 75 years, standardized CS treatment protocols for PBE based on tumor type are lacking, and CS prescribing practices vary widely among physicians managing brain tumors [38, 40, 41]. Finally, tumor molecular variables such as vascular endothelial growth factor, aquaporins, cyclooxygenase‐2, and matrix metalloproteinases that may mediate PBE formation, were not analyzed [6, 42]. Because CS treatment can alter expression of these molecules, analyses ideally would be performed on treatment‐naïve or paired pre‐ and posttreatment tumor samples, and only posttreatment tissues were available in our study [6, 38, 42].
In conclusion, decreases in PBE were observed at follow‐up in approximately 50%–60% of gliomas and meningiomas in dogs, respectively. Although the extent of PBE amelioration at follow‐up was highly variable, dogs with brain tumors that demonstrated improvements in neurological signs and owner‐reported QOL had larger PBEV at initial diagnosis and significantly larger decreases in edema burden and mass effect than dogs that did not improve clinically. Notably, clinical and QOL improvements at follow‐up also were observed in some dogs without PBE or with imaging evidence of worsening PBE. Adverse events attributable to CS and anticonvulsants occurred in approximately 85% of dogs, and although the majority of these AE were expected and objectively graded as minor, their presence contributed negatively to owner‐perceived QOL in > 90% of cases. Palliative treatment with CS and anticonvulsants can influence MRI surrogates commonly used to determine tumor therapeutic responses, and 12/50 (24%) of dogs with glioma in our study experienced decreases in TTV or CEV consistent with some degree of tumor pseudoregession. Given the contribution of PBE to clinical morbidity and the relatively high proportion of dogs with PBE that were nonresponders, there is a need for more effective CS dosing protocols or alternative antiedema treatments.
Disclosure
Authors declare no off‐label use of antimicrobials.
Ethics Statement
Approved by the Institutional Animal Care and Use Committee of Virginia‐Maryland College of Veterinary Medicine, protocols 15–103, 17–204, and 20–057. Authors declare human ethics approval was not needed.
Conflicts of Interest
The authors declare no conflicts of interest.
Supporting information
Table S1. Summary of clinical and brain tumor characteristics of 72 dogs at initial examination and MRI by tumor type.
Table S2. Summary of clinical and brain tumor characteristics of 72 dogs treated with corticosteroids and anticonvulsants by tumor type and clinical response.
Table S3. The STROBE‐Vet statement checklist.
Funding: This work was supported by the National Cancer Institute (P01CA207206, R01CA256285, R01CA276233).
References
- 1. Hu H., Barker A., Harcourt‐Brown T., and Jeffery N., “Systematic Review of Brain Tumor Treatment in Dogs,” Journal of Veterinary Internal Medicine 29, no. 6 (2015): 1456–1463, 10.1111/jvim.13617. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Dickinson P. J., “Advances in Diagnostic and Treatment Modalities for Intracranial Tumors,” Journal of Veterinary Internal Medicine 28, no. 4 (2014): 1165–1185, 10.1111/jvim.12370. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Miller A. D., Miller C. R., and Rossmeisl J. H., “Canine Primary Intracranial Cancer: A Clinicopathologic and Comparative Review of Glioma, Meningioma, and Choroid Plexus Tumors,” Frontiers in Oncology 9 (2019): 1151, 10.3389/fonc.2019.01151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Ingraham F. D., Matson D. D., and Mc Laurin R. L., “Cortisone and ACTH as an Adjunct to the Surgery of Craniopharyngiomas,” New England Journal of Medicine 246, no. 15 (1952): 568–571, 10.1056/NEJM195204102461502. [DOI] [PubMed] [Google Scholar]
- 5. Poirier V. J., Gieger T., James F. M. K., et al., “Peritumoral Edema in Canine Extra‐Axial Brain Tumours: Effect of Steroids,” Veterinary and Comparative Oncology 23, no. 1 (2025): 73–81, 10.1111/vco.13030. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Kotsarini C., Griffiths P. D., Wilkinson I. D., and Hoggard N., “A Systematic Review of the Literature on the Effects of Dexamethasone on the Brain From In Vivo Human‐Based Studies: Implications for Physiological Brain Imaging of Patients With Intracranial Tumors,” Neurosurgery 67, no. 6 (2010): 1799–1815, 10.1227/NEU.0b013e3181fa775b. [DOI] [PubMed] [Google Scholar]
- 7. Strasberg J. R., Rossmeisl J. H., Kelsey K. L., Yoshikawa H., Gieger T. L., and Nolan M. W., “A Prospective Evaluation of Succinct Prednisone Tapering After Brain Tumor Irradiation in Dogs,” Journal of Veterinary Internal Medicine 38, no. 5 (2024): 2571–2577, 10.1111/jvim.17163. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Hatam A., Bergström M., Yu Z. Y., Granholm L., and Berggren B. M., “Effect of Dexamethasone Treatment on Volume and Contrast Enhancement of Intracranial Neoplasms,” Journal of Computer Assisted Tomography 7, no. 2 (1983): 295–300, 10.1097/00004728-198304000-00017. [DOI] [PubMed] [Google Scholar]
- 9. Romano A., De Giorgi S., Romano A., et al., ““Vanishing” Glioblastoma: A Case Report and Review of the Literature,” Radiology Case Reports 19, no. 8 (2024): 3276–3282, 10.1016/j.radcr.2024.04.040. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. D'Elia A., Maiola V., La Pira B., et al., “Vanishing Glioblastoma After Corticosteroid Therapy: Does This Occurrence Modify Our Surgical Strategy?,” Clinical Neurology and Neurosurgery 115, no. 4 (2013): 490–494, 10.1016/j.clineuro.2012.06.010. [DOI] [PubMed] [Google Scholar]
- 11. Goh J. J., See S. J., Ang E., and Ng W. H., “Vanishing Glioblastoma After Corticosteroid Therapy,” Journal of Clinical Neuroscience 16, no. 9 (2009): 1226–1228, 10.1016/j.jocn.2008.10.029. [DOI] [PubMed] [Google Scholar]
- 12. Zaki H. S., Jenkinson M. D., Du Plessis D. G., Smith T., and Rainov N. G., “Vanishing Contrast Enhancement in Malignant Glioma After Corticosteroid Treatment,” Acta Neurochirurgica 146, no. 8 (2004): 841–845, 10.1007/s00701-004-0282-8. [DOI] [PubMed] [Google Scholar]
- 13. Buxton N., Phillips N., and Robertson I., “The Case of the Disappearing Glioma,” Journal of Neurology, Neurosurgery, and Psychiatry 63, no. 4 (1997): 520–521, 10.1136/jnnp.63.4.520. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Hasegawa H., Pal D., Ramirez R., Ismail A., and Marks P., “Glioblastoma Multiforme Fades on CT Imaging After Dexamethasone Therapy,” Journal of Clinical Neuroscience 16, no. 12 (2009): 1707–1708, 10.1016/j.jocn.2009.02.024. [DOI] [PubMed] [Google Scholar]
- 15. Cairncross J. G., Macdonald D. R., Pexman J. H., and Ives F. J., “Steroid‐Induced CT Changes in Patients With Recurrent Malignant Glioma,” Neurology 38, no. 5 (1988): 724–726, 10.1212/wnl.38.5.724. [DOI] [PubMed] [Google Scholar]
- 16. Watling C. J., Lee D. H., Macdonald D. R., and Cairncross J. G., “Corticosteroid‐Induced Magnetic Resonance Imaging Changes in Patients With Recurrent Malignant Glioma,” Journal of Clinical Oncology 12, no. 9 (1994): 1886–1889, 10.1200/JCO.1994.12.9.1886. [DOI] [PubMed] [Google Scholar]
- 17. Mazur M. D., Nguyen V., and Fults D. W., “Glioblastoma Presenting With Steroid‐Induced Pseudoregression of Contrast Enhancement on Magnetic Resonance Imaging,” Case Reports in Neurological Medicine 2012 (2012): 816873, 10.1155/2012/816873. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. R. L. Galloway, Jr. , Maciunas R. J., and Failinger A. L., “Factors Affecting Perceived Tumor Volumes in Magnetic Resonance Imaging,” Annals of Biomedical Engineering 21 (1993): 367–375, 10.1007/BF02368629. [DOI] [PubMed] [Google Scholar]
- 19. May J. L., Garcia‐Mora J., Edwards M., and Rossmeisl J. H., “An Illustrated Scoping Review of the Magnetic Resonance Imaging Characteristics of Canine and Feline Brain Tumors,” Animals 14, no. 7 (2024): 1044, 10.3390/ani14071044. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Shinn R. L., Kani Y., Hsu F. C., and Rossmeisl J. H., “Risk Factors for Adverse Events Occurring After Recovery From Stereotactic Brain Biopsy in Dogs With Primary Intracranial Neoplasia,” Journal of Veterinary Internal Medicine 34, no. 5 (2020): 2021–2028, 10.1111/jvim.15885. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Rossmeisl J. H., Herpai D., Quigley M., et al., “Phase I Trial of Convection‐Enhanced Delivery of IL13RA2 and EPHA2 Receptor Targeted Cytotoxins in Dogs With Spontaneous Intracranial Gliomas,” Neuro‐Oncology 23, no. 3 (2021): 422–434, 10.1093/neuonc/noaa196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Rossmeisl J. H., King J. N., Robertson J. L., Weger‐Lucarelli J., and Elankumaran S., “Phase I/II Trial of Urokinase Plasminogen Activator‐Targeted Oncolytic Newcastle Disease Virus for Canine Intracranial Tumors,” Cancers (Basel) 16, no. 3 (2024): 564, 10.3390/cancers16030564. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Weiske R., Sroufe M., Quigley M., Pancotto T., Werre S., and Rossmeisl J. H., “Development and Evaluation of a Caregiver Reported Quality of Life Assessment Instrument in Dogs With Intracranial Disease,” Frontiers in Veterinary Science 7 (2020): 537, 10.3389/fvets.2020.00537. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Packer R. A., Rossmeisl J. H., Kent M. S., J. F. Griffin, IV , Mazcko C., and LeBlanc A. K., “Consensus Recommendations on Standardized Magnetic Resonance Imaging Protocols for Multicenter Canine Brain Tumor Clinical Trials,” Veterinary Radiology & Ultrasound 59, no. 3 (2018): 261–271, 10.1111/vru.12608. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Garcia Mora J. K., Robertson J., Hsu F. C., et al., “Comparison of Linear and Volumetric Criteria for the Determination of Therapeutic Response in Dogs With Intracranial Gliomas,” Journal of Veterinary Internal Medicine 36, no. 3 (2022): 1066–1074, 10.1111/jvim.16406. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Sturges B. K., Dickinson P. J., Bollen A. W., et al., “Magnetic Resonance Imaging and Histological Classification of Intracranial Meningiomas in 112 Dogs,” Journal of Veterinary Internal Medicine 22, no. 3 (2008): 586–595, 10.1111/j.1939-1676.2008.00042.x. [DOI] [PubMed] [Google Scholar]
- 27. José‐López R., Gutierrez‐Quintana R., de la Fuente C., et al., “Clinical Features, Diagnosis, and Survival Analysis of Dogs With Glioma,” Journal of Veterinary Internal Medicine 35, no. 4 (2021): 1902–1917, 10.1111/jvim.16199. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Koehler J. W., Miller A. D., Miller C. R., et al., “A Revised Diagnostic Classification of Canine Glioma: Towards Validation of the Canine Glioma Patient as a Naturally Occurring Preclinical Model for Human Glioma,” Journal of Neuropathology and Experimental Neurology 77, no. 11 (2018): 1039–1054, 10.1093/jnen/nly085. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Belluco S., Marano G., Lurier T., et al., “Standardization of Canine Meningioma Grading: Validation of New Guidelines for Reproducible Histopathologic Criteria,” Veterinary and Comparative Oncology 21, no. 4 (2023): 685–699, 10.1111/vco.12932. [DOI] [PubMed] [Google Scholar]
- 30. LeBlanc A. K., Atherton M., Bentley R. T., et al., “Veterinary Cooperative Oncology Group‐Common Terminology Criteria for Adverse Events (VCOG‐CTCAE v2) Following Investigational Therapy in Dogs and Cats,” Veterinary and Comparative Oncology 19, no. 2 (2021): 311–352, 10.1111/vco.12677. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Kural C., Atac G. K., Tehli O., et al., “The Evaluation of the Effects of Steroid Treatment on the Tumor and Peritumoral Edema by DWI and MR Spectroscopy in Brain Tumors,” Neurologia i Neurochirurgia Polska 52, no. 4 (2018): 495–504, 10.1016/j.pjnns.2018.03.002. [DOI] [PubMed] [Google Scholar]
- 32. Ostergaard L., Hochberg F. H., Rabinov J. D., et al., “Early Changes Measured by Magnetic Resonance Imaging in Cerebral Blood Flow, Blood Volume, and Blood‐Brain Barrier Permeability Following Dexamethasone Treatment in Patients With Brain Tumors,” Journal of Neurosurgery 90, no. 2 (1999): 300–305, 10.3171/jns.1999.90.2.0300. [DOI] [PubMed] [Google Scholar]
- 33. Hatam A., Yu Z. Y., Bergström M., Berggren B. M., and Greitz T., “Effect of Dexamethasone Treatment on Peritumoral Brain Edema: Evaluation by Computed Tomography,” Journal of Computer Assisted Tomography 6, no. 3 (1982): 586–592, 10.1097/00004728-198206000-00025. [DOI] [PubMed] [Google Scholar]
- 34. Andersen C., Astrup J., and Gyldensted C., “Quantitative MR Analysis of Glucocorticoid Effects on Peritumoral Edema Associated With Intracranial Meningiomas and Metastases,” Journal of Computer Assisted Tomography 18, no. 4 (1994): 509–518, 10.1097/00004728-199407000-00001. [DOI] [PubMed] [Google Scholar]
- 35. Skjoeth J. and Bjerre P. K., “Effect of Glucocorticoids on ICP in Patients With a Cerebral Tumour,” Acta Neurologica Scandinavica 96, no. 3 (1997): 167–170, 10.1111/j.1600-0404.1997.tb00261.x. [DOI] [PubMed] [Google Scholar]
- 36. Nagendran A., McConnell J. F., De Risio L., et al., “Peri‐Ictal Magnetic Resonance Imaging Characteristics in Dogs With Suspected Idiopathic Epilepsy,” Journal of Veterinary Internal Medicine 35, no. 2 (2021): 1008–1017, 10.1111/jvim.16058. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. J. H. Rossmeisl, Jr. , Jones J. C., Zimmerman K. L., and Robertson J. L., “Survival Time Following Hospital Discharge in Dogs With Palliatively Treated Primary Brain Tumors,” Journal of the American Veterinary Medical Association 242, no. 2 (2013): 193–198, 10.2460/javma.242.2.193. [DOI] [PubMed] [Google Scholar]
- 38. Arvold N. D., Armstrong T. S., Warren K. E., et al., “Corticosteroid Use Endpoints in Neuro‐Oncology: Response Assessment in Neuro‐Oncology Working Group,” Neuro‐Oncology 20, no. 7 (2018): 897–906, 10.1093/neuonc/noy056. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Ellingson B. M., Wen P. Y., and Cloughesy T. F., “Evidence and Context of Use for Contrast Enhancement as a Surrogate of Disease Burden and Treatment Response in Malignant Glioma,” Neuro‐Oncology 20, no. 4 (2018): 457–471, 10.1093/neuonc/nox193. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Jessurun C. A. C., Hulsbergen A. F. C., Lamba N., RDS N. T., Smith T. R., and MLD B., “Practice Variation in Perioperative Steroid Dosing for Brain Tumor Patients: An International Survey,” World Neurosurgery 159 (2022): e431–e441, 10.1016/j.wneu.2021.12.067. [DOI] [PubMed] [Google Scholar]
- 41. Lieberman A., LeBrun Y., Glass P., et al., “Use of High Dose Corticosteroids in Patients With Inoperable Brain Tumours,” Journal of Neurology, Neurosurgery, and Psychiatry 40, no. 7 (1977): 678–682, 10.1136/jnnp.40.7.678. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Dickinson P. J., Sturges B. K., Higgins R. J., et al., “Vascular Endothelial Growth Factor mRNA Expression and Peritumoral Edema in Canine Primary Central Nervous System Tumors,” Veterinary Pathology 45, no. 2 (2008): 131–139, 10.1354/vp.45-2-131. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1. Summary of clinical and brain tumor characteristics of 72 dogs at initial examination and MRI by tumor type.
Table S2. Summary of clinical and brain tumor characteristics of 72 dogs treated with corticosteroids and anticonvulsants by tumor type and clinical response.
Table S3. The STROBE‐Vet statement checklist.