Multivoxel proton magnetic resonance spectroscopy of inflammatory and neoplastic lesions of the canine brain at 3.0 T

Krystina L. Stadler Department of Veterinary Clinical Sciences, College of Veterinary Medicine, University of Minnesota, Saint Paul, MN 55108.

Search for other papers by Krystina L. Stadler in
Current site
Google Scholar
PubMed
Close
 DVM
,
Christopher P. Ober Department of Veterinary Clinical Sciences, College of Veterinary Medicine, University of Minnesota, Saint Paul, MN 55108.

Search for other papers by Christopher P. Ober in
Current site
Google Scholar
PubMed
Close
 DVM, PhD
,
Daniel A. Feeney Department of Veterinary Clinical Sciences, College of Veterinary Medicine, University of Minnesota, Saint Paul, MN 55108.

Search for other papers by Daniel A. Feeney in
Current site
Google Scholar
PubMed
Close
 DVM, MS
, and
Carl R. Jessen Department of Veterinary Clinical Sciences, College of Veterinary Medicine, University of Minnesota, Saint Paul, MN 55108.

Search for other papers by Carl R. Jessen in
Current site
Google Scholar
PubMed
Close
 DVM, PhD

Abstract

Objective—To describe findings of 3.0-T multivoxel proton magnetic resonance spectroscopy (1H-MRS) in dogs with inflammatory and neoplastic intracranial disease and to determine the applicability of 1H-MRS for differentiating between inflammatory and neoplastic lesions and between meningiomas and gliomas.

Animals—33 dogs with intracranial disease (19 neoplastic [10 meningioma, 7 glioma, and 2 other] and 14 inflammatory).

Procedures—3.0-T multivoxel 1H-MRS was performed on neoplastic or inflammatory intracranial lesions identified with conventional MRI. N-acetylaspartate (NAA), choline, and creatine concentrations were obtained retrospectively, and metabolite ratios were calculated. Values were compared for metabolites separately, between lesion categories (neoplastic or inflammatory), and between neoplastic lesion types (meningioma or glioma) by means of discriminant analysis and 1-way ANOVA.

Results—The NAA-to-choline ratio was 82.7% (62/75) accurate for differentiating neoplastic from inflammatory intracranial lesions. Adding the NAA-to-creatine ratio or choline-to-creatine ratio did not affect the accuracy of differentiation. Neoplastic lesions had lower NAA concentrations and higher choline concentrations than inflammatory lesions, resulting in a lower NAA-to-choline ratio, lower NAA-to-creatine ratio, and higher choline-to-creatine ratio for neoplasia relative to inflammation. No significant metabolite differences between meningiomas and gliomas were detected.

Conclusions and Clinical Relevance1H-MRS was effective for differentiating inflammatory lesions from neoplastic lesions. Metabolite alterations for 1H-MRS in neoplasia and inflammation in dogs were similar to changes described for humans. Use of 1H-MRS provided no additional information for differentiating between meningiomas and gliomas. Proton MRS may be a beneficial adjunct to conventional MRI in patients with high clinical suspicion of inflammatory or neoplastic intracranial lesions.

Abstract

Objective—To describe findings of 3.0-T multivoxel proton magnetic resonance spectroscopy (1H-MRS) in dogs with inflammatory and neoplastic intracranial disease and to determine the applicability of 1H-MRS for differentiating between inflammatory and neoplastic lesions and between meningiomas and gliomas.

Animals—33 dogs with intracranial disease (19 neoplastic [10 meningioma, 7 glioma, and 2 other] and 14 inflammatory).

Procedures—3.0-T multivoxel 1H-MRS was performed on neoplastic or inflammatory intracranial lesions identified with conventional MRI. N-acetylaspartate (NAA), choline, and creatine concentrations were obtained retrospectively, and metabolite ratios were calculated. Values were compared for metabolites separately, between lesion categories (neoplastic or inflammatory), and between neoplastic lesion types (meningioma or glioma) by means of discriminant analysis and 1-way ANOVA.

Results—The NAA-to-choline ratio was 82.7% (62/75) accurate for differentiating neoplastic from inflammatory intracranial lesions. Adding the NAA-to-creatine ratio or choline-to-creatine ratio did not affect the accuracy of differentiation. Neoplastic lesions had lower NAA concentrations and higher choline concentrations than inflammatory lesions, resulting in a lower NAA-to-choline ratio, lower NAA-to-creatine ratio, and higher choline-to-creatine ratio for neoplasia relative to inflammation. No significant metabolite differences between meningiomas and gliomas were detected.

Conclusions and Clinical Relevance1H-MRS was effective for differentiating inflammatory lesions from neoplastic lesions. Metabolite alterations for 1H-MRS in neoplasia and inflammation in dogs were similar to changes described for humans. Use of 1H-MRS provided no additional information for differentiating between meningiomas and gliomas. Proton MRS may be a beneficial adjunct to conventional MRI in patients with high clinical suspicion of inflammatory or neoplastic intracranial lesions.

Neoplasia and inflammatory disease are common differential diagnoses for intracranial lesions in canine neurologic patients. In a recent study,1 investigators found intracranial neoplasms in 4.5% of dogs during necropsy. The incidence of inflammatory intracranial disease was 8% of dogs with seizures as a main clinical sign and 14% in a large cohort (n = 4,497) of animals with neurologic disease at a referral institution.2,3

Magnetic resonance imaging is commonly used in veterinary medicine for diagnosis of intracranial lesions. Conventional MRI allows for evaluation of intracranial lesions and interpretation based on inherent macroscopic properties of the lesions (location, signal intensity, contrast enhancement, edema, and hemorrhage). Investigators in other studies4,5 have found that conventional MRI findings for inflammatory and neoplastic lesions are often nonspecific, with overlapping imaging findings for the 2 conditions. Investigators in 1 study4 found that 7 MRI characteristics were significantly more common in patients with neoplasia than in patients with inflammatory and vascular disease. In a more recent study,5 investigators found that only strong contrast enhancement was significantly more common in neoplastic lesions. Despite this lack of specificity regarding individual findings, conventional MRI has reasonable overall specificity for differentiating between broad etiologic categories of intracranial disease (neoplastic, inflammatory, or cerebrovascular).6–8 The ability of clinicians to distinguish broad classes of disease with MRI is likely secondary to the summation and integration of all findings present to formulate a presumptive diagnosis.5 Nonetheless, the criterion-referenced standard for diagnosis of intracranial disease is histologic evaluation.9–11

In veterinary medicine, a presumptive MRI diagnosis is often used for prognosis and treatment recommendations in disease conditions for which a definitive diagnosis would require biopsy. Factors impacting the decision to use a presumptive diagnosis may include the invasive nature of the lesion, inoperability of the lesion, and associated cost. Thus, there is a need for improvements in noninvasive diagnostic modalities to increase confidence in the accuracy of diagnosis of intracranial lesions.

Over the past 20 years, advances in supplementary MRI sequences and MRI postprocessing software have provided additional information about intracranial lesions. Proton MRS is increasingly being used in diagnostic imaging of humans. Proton MRS provides qualitative analysis of a number of metabolites within the brain and allows for quantitative analysis if a reference with a known concentration is used. This offers physiologic and chemical information for a given ROI. The availability of physiochemical data is in contrast to the results for conventional MRI, in which anatomic and macroscopic structural change is interpreted.

In human medicine, extensive literature and clinical applications of 1H-MRS have been described for numerous intracranial lesions, including brain tumors,12–17 infectious disease,18,19 epilepsy,20,21 neurodegenerative disorders such as amyotrophic lateral sclerosis,22 dementia,23,24 and metabolic disorders.25 The main application of1H-MRS in humans is for diagnostic imaging of brain tumors;1H-MRS in human patients is considered extremely accurate (93% to 100%) for use in differentiating neoplastic from nonneoplastic lesions when used in conjunction with conventional contrast-enhanced MRI.26,27

Despite the extensive amount of data available in human medicine, limited data are available on the use of MRS in veterinary medicine. Most veterinary studies have involved the use of animals for evaluation of pharmacokinetics28 or as a model of disease for stroke,29 hypothermic circulatory arrest,30 and seizures21 in humans, rather than evaluation of clinical veterinary patients. Recently, a protocol for MRS imaging of the canine brain at 3.0 T,31 reference values for metabolite concentrations and ratios in a group of healthy laboratory Beagles,32 reference values for various age groups and brain regions in healthy dogs,33 and the report of a study34 conducted to evaluate feasibility of MRS at 3.0 and 7.0 T have been published. Investigators in 1 report35 describe MRS findings in pigs affected with neurocysticercosis; however, reports on MRS findings of naturally occurring intracranial disease in the canine brain are limited.a

The purposes of the study reported here were to describe MRS findings in dogs with confirmed intracranial lesions and compare these values with published reference ranges, to determine whether specific metabolite concentrations or ratios can be used to differentiate intracranial neoplasia from inflammatory disease, and to determine whether specific metabolite concentrations or ratios can be used to differentiate meningiomas from gliomas. We hypothesized that NAA, choline, and creatine ratios and concentrations measured by use of 1H-MRS in intracranial lesions would differ from established concentrations in brains of clinically normal dogs, NAA concentration would be decreased in both neoplastic and inflammatory conditions,36–38 and choline concentration would be increased in neoplastic but not inflammatory conditions.36–39

Materials and Methods

Animals—Medical records at the University of Minnesota Veterinary Medical Center were evaluated to identify canine clinical patients that had undergone MRI evaluation and had a confirmed diagnosis of intracranial neoplasia or a confirmed or presumptive diagnosis of intracranial inflammatory disease. Criteria for inclusion were MRI of the brain with multivoxel 1H-MRS performed at the University of Minnesota Veterinary Medical Center and a histologic diagnosis for the patient of neoplasia or inflammatory disease (via necropsy or surgical biopsy) or a presumptive diagnosis of inflammatory disease based on morphological MRI findings (including presence of multifocal intraparenchymal lesions, lesion contrast enhancement, and meningeal enhancement), clinical examination findings and CSF cytologic evaluation results compatible with a diagnosis of inflammatory disease, and positive response to appropriate treatment. Dogs were excluded if MRS was not of diagnostic quality or if the 1H-MRS ROI did not include a lesion identified by use of conventional MRI.

Procedures—Dogs were anesthetized and maintained in a plane of anesthesia that would allow mechanical ventilation during the MRI scan; the precise anesthesia protocol differed among patients. A 3.0-T imaging systemb was used for MRI and 1H-MRS of all dogs. All dogs were positioned in sternal recumbency and scanned using an extremity coil.c The protocol used for 1H-MRS was based on an optimized protocol that has been previously reported.31 In brief, standard morphological MRI sequences were acquired, and then spectroscopic acquisition that involved use of a multivoxel MRS technique (point-resolved spectroscopy) with automatic shim and water suppression (proton brain examination with point-resolved spectroscopy) was performed. A 20-cm field of view, an 18 × 18 spectroscopic phase and frequency matrix, and an echo time of 144 milliseconds were used for all dogs.

Dorsoventral and right-to-left dimensions of the prescribed multivoxel ROI differed depending on the area of brain being imaged. Care was taken to maximize the amount of brain parenchyma included while excluding the calvaria. The rostrocaudal ROI dimension (and thus the thickness of the ROI and individual voxels) was fixed at 1 cm.

Collection and evaluation of MRS data were performed with postprocessing software included with the MRI scanner.d For collection of metabolite concentrations, a multivoxel grid was applied to each ROI, and values for individual voxels containing morphologically visible pathological lesions were evaluated. The spectrum generated for the tissue within each individual voxel was evaluated, and the concentrations of NAA, choline, and creatine were recorded. Presence of abnormal peaks, including lipid (0.9 to 1.4 ppm), lactate (1.3 ppm), alanine (1.5 ppm), acetate (1.9 ppm), and succinate (2.4 ppm), was also recorded. Relative concentrations for each metabolite (ie, peak height) were manually determined from the scale of the y-axis assigned by the computer. Direct manual evaluation of peak height and chemical shift (on the x-axis) was used to minimize bias, which could have been caused by default settings and potential differences in chemical shifts between human and canine metabolites. Other data recorded for each voxel included brain lobe and whether the voxel was in the central portion of the identified lesion or at the periphery. Also recorded for each lesion was the final diagnosis and the broad diagnostic classification (neoplasia or inflammation).

Statistical analysis—All statistical analyses were performed with the aid of commercially available statistical software.e For evaluation of gliomas and meningiomas, all other types of neoplasm were removed from consideration.

Initial statistical evaluation included the concentrations of the 3 metabolites (NAA, choline, and creatine) and the 3 metabolite ratios (NAA-to-choline ratio, NAA-to-creatine ratio, and choline-to-creatine ratio), which were all evaluated separately by means of independent t tests to assess their mean values with the assumption of unequal variance among the groups. Also, nonparametric evaluation for median values was performed with the Mann-Whitney U test to determine whether the null hypotheses of these variables were statistically valid across the final broad diagnostic classifications of neoplasia versus inflammation. Independent χ2 tests were used to evaluate count data.

To determine whether any of the 3 measured metabolite concentrations or 3 ratios had a relationship to the final broad diagnostic classification of the intracranial disease (neoplasia vs inflammation) and to identify variables that may have predictive value in formulating a diagnosis, equation-based predictions were generated with a linear discriminant analysis technique. Fisher prediction equations were developed by use of the Wilks γ discriminant method. Stepwise analysis techniques were used, with the F statistic criteria set for variable entrance into the equation at a value of P ≤ 0.05 and removal at a value of P ≥ 0.10. Initially, only the 3 ratios (NAA-to-choline ratio, NAA-to-creatine ratio, and choline-to-creatine ratio) were inserted into the discriminant analysis. Discriminant analysis was performed with all data points, central data points alone, and peripheral data points alone to determine whether there were differences that depended on the location within a lesion. After this evaluation, raw values of the 3 metabolites were entered into the discriminant analysis, and the analysis was repeated with all data points (ie, central and peripheral data points were pooled).

A 1-way ANOVA was performed with all 3 ratios and all 3 metabolites, with an assumption of unequal variance and by use of the Tamhane T2 post hoc test (because of the assumption of unequal variance), to determine whether there was a relationship with a more specific diagnostic classification (glioma vs meningioma vs inflammation). The nonparametric Kruskal-Wallis test was also performed. Following this, the Mann-Whitney U test was performed to compare only gliomas and meningiomas (dogs with inflammation were removed from the analysis) to determine whether there were significant differences between the 2 tumor types. Discriminant analysis was performed with all 3 ratios and all 3 raw metabolite values to determine whether there was a relationship with a more specific diagnostic classification (glioma vs meningioma vs inflammation) or whether there was a relationship with a more specific neoplastic classification (glioma vs meningioma) when dogs with inflammation were removed.

All reported significant differences were with the assumption of unequal variance between groups. All metric tests were conducted with the 2-tailed option. Values of P ≤ 0.05 were considered significant.

Results

Thirty-three dogs met the criteria for inclusion in the study. In these dogs, 19 lesions had a diagnosis of intracranial neoplasia (10 meningioma, 7 glioma or glioblastoma multiforme, 1 round-cell tumor, and 1 metastatic hemangiosarcoma; Figure 1). Inflammatory etiologies were diagnosed for lesions in 14 dogs, including 6 with histologic confirmation (4 granulomatous encephalitis or meningoencephalitis, 1 necrotizing encephalitis, and 1 blastomycosis meningoencephalitis) and 8 with presumed inflammatory disease on the basis of clinical signs and response to treatment (Figure 2).

Figure 1—
Figure 1—

Representative 1H-MRS spectrum of a histologically confirmed meningioma in the piriform lobe of a dog. The NAA peak is considerably smaller than the choline peak. Ratios calculated for this spectrum are as follows: NAA-to-choline ratio, 0.47; NAA-to-creatine ratio, 1.02; and choline-to-creatine ratio, 2.18. The NAA-to-choline ratio is lower than the reference range (0.687 to 1.568) and the choline-to-creatine ratio is higher than the reference range (0.947 to 1.797) reported for the piriform lobe in healthy dogs.32 Cho = Choline. Cr = Creatine.

Citation: American Journal of Veterinary Research 75, 11; 10.2460/ajvr.75.11.982

Figure 2—
Figure 2—

Representative 1H-MRS spectrum of histologically confirmed granulomatous meningoencephalitis in the piriform lobe of a dog. The NAA peak is smaller than the choline peak, but the difference in height between the 2 peaks is less than in neoplastic lesions. The ratios calculated for this spectrum are as follows: NAA-to-choline ratio, 0.80; NAA-to-creatine ratio, 1.08; and choline-to-creatine ratio, 1.35. All ratios calculated for this spectrum are in the reference ranges (0.687 to 1.568, 0.984 to 1.818, and 0.947 to 1.797, respectively) reported for the piriform lobe in healthy dogs.32 Notice the small lactate peak (inverted doublet at 1. 3 ppm).

Citation: American Journal of Veterinary Research 75, 11; 10.2460/ajvr.75.11.982

Lactate peaks were identified in 7 dogs (1 glioma, 1 meningioma, and 5 inflammatory lesions), and lipid peaks were identified in 14 dogs (2 gliomas, 3 meningiomas, 1 metastatic hemangiosarcoma, 1 round-cell tumor, and 7 inflammatory lesions). Alanine, acetate, and succinate peaks were not detected in any lesions. Because of the low counts for peaks present in multiple subgroups of these 5 metabolites, statistical analysis was not performed for any of these metabolites.

Separate evaluation of the metabolites and ratios by use of the Mann-Whitney U test revealed significant differences in the medians between neoplastic lesions and inflammatory lesions for 4 of 6 variables (including 1 raw value and all 3 ratios; Table 1). Significant differences were similar for comparisons of the means with the 1-way ANOVA.

Table 1—

Median and mean metabolite concentrations (arbitrary units) and metabolite ratios determined by means of 1H-MRS in intracranial lesions diagnosed as neoplasia or inflammation in dogs.

 Mann-Whitney U testIndependent t test
VariableNeoplasia medianInflammation medianP value*Neoplasia meanInflammation meanP value*
NAA1,214.52,156.00.0031,622.82,893.70.020
Choline2,666.02,580.00.9353,207.73,545.30.586
Creatine1,292.01,496.00.2881,555.82,329.40.089
NAA-to-choline ratio0.4650.850< 0.0010.4840.947< 0.001
NAA-to-creatine ratio0.9971.2230.0111.0761.3380.016
Choline-to-creatine ratio2.1061.5570.0012.5441.588< 0.001

Values of P ≤ 0.05 were considered significant

Discriminant analysis with only the 3 ratios, compared with broad diagnostic classification (neoplasia vs inflammation) in all samples (central and peripheral data were pooled), were used because the χ2 tests and t tests revealed no significant difference between central and peripheral values. Stepwise discriminant tests selected only the NAA-to-choline ratio, which yielded a predictive accuracy of 82.7% (62/75). The selection of only 1 variable (NAA-to-choline ratio) allowed a unique opportunity to make predictions based on a threshold value of that variable when there are only 2 possible groups. The Fisher equations resulted in a threshold NAA-to-choline ratio of approximately 0.74 for discrimination between neoplasia and inflammation: 44 of 50 (88%) neoplastic voxels had values less than this cutoff, and 18 of 25 (72%) inflammatory voxels had values greater than this cutoff. If all 3 ratios were forced into the discriminant analysis equation, the predictive accuracy remained 82.7%. When only voxels in the center of lesions were used, discriminant analysis yielded an accuracy of 82.4% (28/34), and when only voxels in the periphery of lesions were used, discriminant analysis yielded an accuracy of 85.4% (35/41). There was no significant difference in the values between the central and peripheral voxels.

All 6 variables were evaluated separately among the more specific diagnostic classifications (glioma vs meningioma vs inflammation), and a 1-way ANOVA revealed differences in the NAA-to-choline ratio and choline-to-creatine ratio between inflammatory conditions and both types of neoplasia, with additional differences in the NAA concentration and NAA-to-creatine ratio between meningiomas and inflammation (Table 2). Similar to the more general diagnostic classification, use of the Kruskal-Wallis test revealed significant differences in the values of 4 of 6 variables among the groups (including 1 raw value and all 3 ratios). No significant differences were found between gliomas and meningiomas (all P > 0.1) with the Mann-Whitney U test.

Table 2—

Median and mean metabolite concentrations (arbitrary units) and metabolite ratios determined by means of 1H-MRS in intracranial lesions diagnosed as meningioma, glioma, or inflammation in dogs.

 Kruskal-Wallis test1-way ANOVA and Tamhane T2 post hoc test
VariableMeningioma medianGlioma medianInflammation medianPvalue*Meningioma meanGlioma meanInflammation mean
NAA1,074.01,369.02,156.00.0021,209.8a1,900.1a,b2,893.7b
Choline2,009.02,928.02,580.00.2342,922.43,531.63,545.3
Creatine1,068.01,436.01,496.00.3941,393.61,643.22,329.4
NAA-to-choline ratio0.4550.5120.850< 0.0010.438a0.478a0.947b
NAA-to-creatine ratio0.9240.9841.2230.0051.010a1.114a,b1.338a
Choline-to-creatine ratio2.1202.1911.5570.0012.641a2.545a1.588b

Within a row, values with different superscript letters differ significantly (P ≤ 0.05).

See Table 1 for remainder of key.

Discriminant analysis was performed with all 3 ratios and all 3 metabolite concentrations in all samples (central and peripheral data were pooled), with the intention of discriminating among glioma, meningioma, and inflammation. For that analysis, both NAA concentration and the NAA-to-choline ratio were retained in the model. Diagnostic accuracy for distinguishing among these 3 conditions was 63.9% (46/72). For a discriminant analysis with all 3 ratios and all 3 metabolite concentrations for differentiation of gliomas and meningiomas, only NAA concentration was maintained in the model. This yielded an accuracy of 61.7% (29/47) for differentiating the 2 tumor types.

Discussion

Accuracy for the NAA-to-choline ratio to differentiate neoplastic from inflammatory intracranial lesions by use of a cutoff value of 0.74 was 82.7%. Forced addition of the NAA-to-creatine ratio and choline-to-creatine ratio into the discriminant equation did not alter the accuracy of distinguishing neoplastic from inflammatory lesions. Given the reported overlap in conventional MRI findings between neoplasia and inflammatory lesions,4,5 differentiation of these 2 broad classes of disease has historically been a diagnostic conundrum. However, on the basis of the findings in the present study, use of 1H-MRS to derive the NAA-to-choline ratio in lesions will help clinicians differentiate these 2 classes of pathological changes, which can in turn influence the treatment plan. In humans, MRS has an accuracy of 93% to 100% for differentiation of neoplastic from nonneoplastic intracranial lesions when used in conjunction with conventional contrast-enhanced MRI.26,27 Describing and comparing the addition of conventional MRI findings was not an aim of the present study, and the stated accuracy was based solely on quantitative analysis of the metabolite ratios. Addition of 1H-MRS results to conventional MRI findings may further improve discrimination of inflammatory and neoplastic intracranial disease in veterinary patients, but further research is necessary to confirm this speculation.

Although only the NAA-to-choline ratio was valuable for differentiating intracranial neoplasia from inflammatory lesions with discriminant analysis, there were significant differences for all 3 calculated ratios (NAA-to-choline ratio, NAA-to-creatine ratio, and choline-to-creatine ratio) between the neoplastic and inflammatory lesions. Specifically, the NAA-to-choline ratio and NAA-to-creatine ratio were decreased and the choline-to-creatine ratio was increased in neoplasia relative to inflammation (Table 1). These findings are consistent with reported metabolite changes in intracranial neoplasms of humans.14,40,41 Metabolite ratios (NAA-to-choline ratio, choline-to-creatine ratio, and NAA-to-creatine ratio), rather than absolute metabolite concentrations, are used commonly in human medicine to assess choline and NAA alterations within intracranial lesions. Absolute concentrations are dependent on cellular density, which is often altered in pathological processes.16,17,36 Additionally, because metabolite concentrations are generally reported in arbitrary units, ratios are likely more clinically applicable across various scanners than are raw concentrations, inasmuch as ratios provide an internal control for variation in reported arbitrary units.

Increases in choline concentrations and decreases in NAA concentrations are detected in intracranial neoplasms in humans. The increase in choline concentration is associated with cell membrane turnover and cellular proliferation. The decrease in NAA concentration is associated with neuronal destruction and displacement. Most intracranial neoplasms originate from nonneuronal tissues and thus should have a low or absent NAA peak.36 Creatine is a marker of energy metabolism and is considered a relatively stable metabolite; thus, it is often used as an internal reference value for calculation of metabolite ratios.16,36

Magnetic resonance spectroscopy in humans has aided in the differentiation of inflammatory from neoplastic mass lesions. In humans, inflammatory masses are often associated with a decrease in NAA concentration but may lack the concurrent increase in choline concentration that typically is observed in neoplastic masses.36–38 In the study reported here, the median NAA-to-creatine ratio was decreased in both neoplastic (0.997) and inflammatory lesions (1.223), in comparison to previously published medians of reference ranges obtained with identical 1H-MRS technical parameters (median, 1.191 to 1.828 [depending on the lobe]; neoplasia, < 9/9 lobe medians; and inflammation, < 7/9 lobe medians).32 However, this is more notable for neoplasia than for inflammation because the median NAA-to-creatine ratio for neoplasia in the present study was at the low end of the reported reference range (0.984 to 2.044),32 whereas the median NAA-to-creatine ratio for inflammation in the present study was well within that range. Similar to results reported in the human literature, the median choline-to-creatine ratio in the dogs of the present study was significantly higher in the neoplastic lesions (2.106) than in the inflammatory lesions (1.557), and the median choline-to-creatine ratio in the neoplastic lesions also was higher than the upper limit for the previously reported reference range (0.828 to 1.853).32 Although the median choline-to-creatine ratio in the inflammatory lesions was slightly higher than the median values reported previously for anatomically normal brain regions, it was well within the reported reference range.32 Therefore, on the basis of these findings, 1H-MRS was useful for differentiating neoplastic lesions from inflammatory lesions, but in the absence of morphological abnormalities, 1H-MRS may not allow for differentiation of inflammatory and anatomically normal brain tissue.

In the present study, the accuracy of differentiation between neoplastic and inflammatory lesions by use of 1H-MRS was not affected by the location of voxels within the lesion. There was no significant difference in diagnostic accuracy between voxels located in the center of the lesion and those located at the periphery of the lesion. On the basis of these results, precise placement of the 1H-MRS ROI within a specific area of the lesion was not necessary and did not affect the ability to differentiate between inflammatory and neoplastic lesions. It has been reported that a large cystic area or areas of necrosis could affect the 1H-MRS spectrum appearance,42,43 but the ROIs in the study reported here typically were placed in more solid areas of lesions, so these differences were not encountered. Evaluation of 1H-MRS in a wider variety of lesion morphologies in veterinary patients would be necessary to determine the relevance of this possible spectrum variation.

Meningiomas (n = 10) and gliomas (7) were the most common types of intracranial neoplasms in the present study. Prevalence of these tumor types was similar to that previously reported in the veterinary literature.1,44 There were no significant differences in metabolite spectra between meningiomas and gliomas. Discriminant analysis also had a poor predictive value for differentiation of meningioma and glioma (accuracy, 61.7%) as well as differentiation of meningioma versus glioma versus inflammation (accuracy, 63.9%). Predictive value was poor because the 1H-MRS characteristics of gliomas typically were between those of meningiomas and inflammation, which created ambiguity (Table 2). Interestingly, the NAA concentration and NAA-to-creatine ratio were significantly different between meningiomas and inflammatory lesions, whereas gliomas had an intermediate value for both the NAA concentration and NAA-to-creatine ratio and were not significantly different from values for either of the other 2 lesion types. This pattern likely related to different amounts of neuronal tissue, and thus NAA concentrations, within the evaluated lesions. Inflammatory lesions would be expected to have the highest amount of neuronal tissue and NAA concentrations in a lesion, whereas gliomas may have an intermediate value because some neurons may persist within an infiltrative neoplasm.45 Meningiomas are expected to have no neuronal tissue because these tumors generally displace the brain parenchyma. The lack of significant differences in values between gliomas and the other 2 types of pathological lesions likely was related to the smaller magnitude of the differences and the relatively small number of dogs in each group, which resulted in inadequate power of the study to detect significant differences.

Findings regarding differentiation of tumor types in the present study were not surprising, considering that the typical 1H-MRS findings in humans with both meningiomas and gliomas include an absent or low NAA concentration and a high choline concentration. The presence of alanine (doublet peak at 1.48 ppm) was indicative of meningioma in humans; however, expression is only evident in approximately 30% to 50% of human patients.17,46 None of the canine patients in the present study had an identifiable alanine peak, which indicated that this finding may be unique to meningiomas in humans. However, use of smaller voxels (as was used in the present study) and presence of lactate peaks are associated with lack of identification of alanine peaks in human patients with meningiomas.46 Thus, further studies with optimization of 1H-MRS protocols may be necessary to recognize alanine in meningiomas of dogs.

Given the classic appearance of most meningiomas with conventional MRI, meningiomas in humans are often diagnosed without the aid of spectroscopy.16,41 It is acknowledged in the human literature that atypical and malignant meningiomas are often challenging to distinguish from gliomas in the absence of alanine.16 Classic conventional MRI findings in meningiomas of dogs have been reported elsewhere.47,48 However, investigators in a recent study5 concluded that conventional MRI characteristics commonly associated with meningiomas and gliomas may also be seen in nonneoplastic lesions. Thus, although morphological assessment of suspected intracranial neoplasms will remain important, 1H-MRS may be useful as an adjunct to support the presumptive diagnosis in intracranial neoplastic and inflammatory lesions.

Limitations of the present study primarily related to its retrospective nature and included selection and placement of the ROI, lack of internal comparison, small sample size, and absence of histopathologic confirmation in some of the inflammatory lesions. In 1H-MRS, the ROI must be selected at the time of the scan and cannot be repositioned at a later time. Thus, some lesions with histopathologic confirmation were excluded from the study because of the absence of pathological changes within the ROI. Ideally, for internal comparison, in cases in which anatomically normal parenchyma is present, 1H-MRS should be performed on both the lesion and adjacent anatomically normal parenchyma.49 In the present study, anatomically normal tissue was often not included within the ROI and thus was not available for comparison. Histologic evaluation was not available for 8 of 14 inflammatory lesions; for these lesions, a presumptive diagnosis was based on clinical signs, conventional MRI findings, and response to immunosuppression. Excluding inflammatory lesions without histologic confirmation would drastically limit inclusion of this type of lesion in studies such as the one reported here. Previous veterinary studies9,10,50 have included use of this presumptive diagnosis method because owners often elect empirical treatment over more invasive diagnostic procedures, such as obtaining biopsy specimens. Because these patients responded to treatment, more invasive diagnostic testing (specimens obtained via biopsy or necropsy) were not elected. Use of a presumptive diagnosis based on clinical and imaging findings and response to treatment (rather than sampling) is also often used in humans for conditions such as ischemic stroke where sampling the lesion would result in unacceptable morbidity.51

Limited information on the use of 1H-MRS in veterinary medicine is available.21,28–34,a To the authors’ knowledge, the study reported here is the first that has been conducted to evaluate the use of 1H-MRS in neoplastic and inflammatory intracranial disease of dogs. A larger, prospective study would be needed to validate results of the present study. Additional applications such as spectroscopic guidelines for differentiation of low-grade from high-grade gliomas have been reported for humans.52–54 Further classification of the gliomas was beyond the scope of the present study; however, this is a future application to consider for veterinary patients.

Use of 1H-MRS was effective for differentiating inflammatory lesions from neoplastic lesions in most canine patients included in the present study. However, it could not be used to effectively differentiate between gliomas and meningiomas. Metabolite alternations were generally in agreement with those reported in human patients. Use of 1H-MRS appears to be a beneficial adjunct to conventional MRI in patients in which there is difficulty differentiating between intracranial inflammatory and neoplastic lesions.

ABBREVIATIONS

1H-MRS

Proton magnetic resonance spectroscopy

NAA

N-acetylaspartate

ROI

Region of interest

a.

Mikoloski KR, March PA, Faissler D. Diagnostic value and discriminatory ability of proton magnetic resonance spectroscopy for intracranial neoplasia in dogs (abstr). J Vet Intern Med 2012;26:804.

b.

SignaHDx, GE Medical Systems, Milwaukee, Wis.

c.

HD T/R Quad Extremity, Invivo, Pewaukee, Wis.

d.

FuncTool, GE Medical Systems, Milwaukee, Wis.

e.

SPSS, version 18, SPSS Inc, Chicago, Ill.

References

  • 1. Song RB, Vite CH & Bradley CW, et al. Postmortem evaluation of 435 cases of intracranial neoplasia in dogs and relationship of neoplasm with breed, age, and body weight. J Vet Intern Med 2013; 27:11431152.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 2. Podell M, Fenner WR, Powers JD. Seizure classification in dogs from a nonreferral-based population. J Am Vet Med Assoc 1995; 206:17211728.

    • Search Google Scholar
    • Export Citation
  • 3. Fluehmann G, Doherr MG, Jaggy A. Canine neurological diseases in a referral hospital population between 1989 and 2000 in Switzerland. J Small Anim Pract 2006; 47:582587.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 4. Cherubini GB, Mantis P & Martinez TA, et al. Utility of magnetic resonance imaging for distinguishing neoplastic from nonneoplastic brain lesions in dogs and cats. Vet Radiol Ultrasound 2005; 46:384387.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 5. Young BD, Fosgate GT & Holmes SP, et al. Evaluation of standard magnetic resonance characteristics used to differentiate neoplastic, inflammatory, and vascular brain lesions in dogs. Vet Radiol Ultrasound 2014; 55:399406.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 6. Ródenas S, Pumarola M & Gaitero L, et al. Magnetic resonance imaging findings in 40 dogs with histologically confirmed intracranial tumours. Vet J 2011; 187:8591.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 7. Wolff CA, Holmes SP & Young BD, et al. Magnetic resonance imaging for the differentiation of neoplastic, inflammatory, and cerebrovascular brain disease in dogs. J Vet Intern Med 2012; 26:589597.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 8. Leclerc MK, d'Anjou MA & Blond L, et al. Interobserver agreement and diagnostic accuracy of brain magnetic resonance imaging in dogs. J Am Vet Med Assoc 2013; 242:16881695.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 9. Zarfoss M, Schatzberg S & Venator K, et al. Combined cytosine arabinoside and prednisone therapy for meningoencephalitis of unknown aetiology in 10 dogs. J Small Anim Pract 2006; 47:588595.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 10. Adamo PF, Adams WM, Steinberg H. Granulomatous meningoencephalomyelitis in dogs. Compend Contin Educ Pract Vet 2007; 29:678690.

  • 11. Talarico LR, Schatzberg SJ. Idiopathic granulomatous and necrotising inflammatory disorders of the canine central nervous system: a review and future perspectives. J Small Anim Pract 2010; 51:138149.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 12. Shimizu H, Kumabe T & Tominaga T, et al. Noninvasive evaluation of malignancy of brain tumors with proton MR spectroscopy. AJNR Am J Neuroradiol 1996; 17:737747.

    • Search Google Scholar
    • Export Citation
  • 13. Meyerand ME, Pipas JM & Mamourian A, et al. Classification of biopsy-confirmed brain tumors using single-voxel MR spectroscopy. AJNR Am J Neuroradiol 1999; 20:117123.

    • Search Google Scholar
    • Export Citation
  • 14. McKnight TR, von dem Bussche MH & Vigneron DB, et al. Histopathological validation of a three-dimensional magnetic resonance spectroscopy index as a predictor of tumor presence. J Neurosurg 2002; 97:794802.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 15. Law M, Yang S & Wang H, et al. Glioma grading: sensitivity, specificity, and predictive values of perfusion MR imaging and proton MR spectroscopic imaging compared with conventional MR imaging. AJNR Am J Neuroradiol 2003; 24:19891998.

    • Search Google Scholar
    • Export Citation
  • 16. Law M. MR spectroscopy of brain tumors. Top Magn Reson Imaging 2004; 15:291313.

  • 17. Sibtain NA, Howe FA, Saunders DE. The clinical value of proton magnetic resonance spectroscopy in adult brain tumours. Clin Radiol 2007; 62:109119.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 18. Rémy C, Grand S & Laï ES, et al. 1H MRS of human brain abscesses in vivo and in vitro. Magn Reson Med 1995; 34:508514.

  • 19. Burtscher IM, Holtås S. In vivo proton MR spectroscopy of untreated and treated brain abscesses. AJNR Am J Neuroradiol 1999; 20:10491053.

    • Search Google Scholar
    • Export Citation
  • 20. Sandok EK, O'Brien TJ & Jack CR, et al. Significance of cerebellar atrophy in intractable temporal lobe epilepsy: a quantitative MRI study. Epilepsia 2000; 41:13151320.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 21. Neppl R, Nguyen CM & Bowen W, et al. In vivo detection of postictal perturbations of cerebral metabolism by use of proton MR spectroscopy: preliminary results in a canine model of prolonged generalized seizures. AJNR Am J Neuroradiol 2001; 22:19331943.

    • Search Google Scholar
    • Export Citation
  • 22. Bowser R, Turner MR, Shefner J. Biomarkers in amyotrophic lateral sclerosis: opportunities and limitations. Nat Rev Neurol 2011; 7:631638.

  • 23. Miller BL, Moats RA & Shonk T, et al. Alzheimer disease: depiction of increased cerebral myo-inositol with proton MR spectroscopy. Radiology 1993; 187:433437.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 24. Huang W, Alexander GE & Chang L, et al. Brain metabolite concentration and dementia severity in Alzheimer's disease: a 1H MRS study. Neurology 2001; 57:626632.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 25. Bianchi MC, Tosetti M & Battini R, et al. Proton MR spectroscopy of mitochondrial diseases: analysis of brain metabolic abnormalities and their possible diagnostic relevance. AJNR Am J Neuroradiol 2003; 24:19581966.

    • Search Google Scholar
    • Export Citation
  • 26. Sorby WA. An evaluation of magnetic resonance imaging at The Royal North Shore Hospital of Sydney, 1986–1987. Med J Aust 1989; 151:811, 1418.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 27. Rand SD, Prost R & Haughton V, et al. Accuracy of single-voxel proton MR spectroscopy in distinguishing neoplastic from nonneoplastic brain lesions. AJNR Am J Neuroradiol 1997; 18:16951704.

    • Search Google Scholar
    • Export Citation
  • 28. Lee SH, Kim SY & Woo DC, et al. Differential neurochemical responses of the canine striatum with pentobarbital or ketamine anesthesia: a 3T proton MRS study. J Vet Med Sci 2010; 72:583587.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 29. Kang BT, Jang DP & Lee JH, et al. Detection of cerebral metabolites in a canine model of ischemic stroke using 1H magnetic resonance spectroscopy. Res Vet Sci 2009; 87:300306.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 30. Barreiro CJ, Williams JA & Fitton TP, et al. Noninvasive assessment of brain injury in a canine model of hypothermic circulatory arrest using magnetic resonance spectroscopy. Ann Thorac Surg 2006; 81:15931598.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 31. Ober CP, Warrington CD & Feeney DA, et al. Optimizing a protocol for 1H-magnetic resonance spectroscopy of the canine brain at 3T. Vet Radiol Ultrasound 2013; 54:149158.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 32. Warrington CD, Feeney DA & Ober CP, et al. Relative metabolite concentrations and ratios determined by use of 3-T regionspecific proton magnetic resonance spectroscopy of the brain of healthy Beagles. Am J Vet Res 2013; 74:12911303.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 33. Ono K, Kitagawa M & Ito D, et al. Regional variations and age-related changes detected with magnetic resonance spectroscopy in the brain of healthy dogs. Am J Vet Res 2014; 75:179186.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 34. Martin-Vaquero P, da Costa RC & Echandi RL, et al. Magnetic resonance spectroscopy of the canine brain at 3.0 T and 7.0 T. Res Vet Sci 2012; 93:427429.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 35. Chawla S, Gupta RK & Husain N, et al. Prediction of viability of porcine neurocysticercosis with proton magnetic resonance spectroscopy: correlation with histopathology. Life Sci 2004; 74:10811092.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 36. Bertholdo D, Watcharakorn A, Castillo M. Brain proton magnetic resonance spectroscopy: introduction and overview. Neuroimaging Clin N Am 2013; 23:359380.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 37. Kingsley PB, Shah TC, Woldenberg R. Identification of diffuse and focal brain lesions by clinical magnetic resonance spectroscopy. NMR Biomed 2006; 19:435462.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 38. Pivawer G, Law M, Zagzag D. Perfusion MR imaging and proton MR spectroscopic imaging in differentiating necrotizing cerebritis from glioblastoma multiforme. Magn Reson Imaging 2007; 25:238243.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 39. Nagar VA, Ye J & Xu M, et al. Multivoxel MR spectroscopic imaging—distinguishing intracranial tumours from non-neoplastic disease. Ann Acad Med Singapore 2007; 36:309313.

    • Search Google Scholar
    • Export Citation
  • 40. Gupta RK, Cloughesy TF & Sinha U, et al. Relationships between choline magnetic resonance spectroscopy, apparent diffusion coefficient and quantitative histopathology in human glioma. J Neurooncol 2000; 50:215226.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 41. Lehnhardt FG, Röhn G & Ernestus RI, et al. 1H- and 31P-MR spectroscopy of primary and recurrent human brain tumors in vitro: malignancy-characteristic profiles of water soluble and lipophilic spectral components. NMR Biomed 2001; 14:307317.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 42. Chang KH, Song IC & Kim SH, et al. In vivo single-voxel proton MR spectroscopy in intracranial cystic masses. AJNR Am J Neuroradiol 1998; 19:401405.

    • Search Google Scholar
    • Export Citation
  • 43. Rock JP, Hearshen D & Scarpace L, et al. Correlations between magnetic resonance spectroscopy and image-guided histopathology, with special attention to radiation necrosis. Neurosurgery 2002; 51:912919.

    • Search Google Scholar
    • Export Citation
  • 44. Snyder JM, Shofer FS & Van Winkle TJ, et al. Canine intracranial primary neoplasia: 173 cases (1986–2003). J Vet Intern Med 2006; 20:669675.

    • Search Google Scholar
    • Export Citation
  • 45. Tong ZY, Yamaki T, Wang YJ. Proton magnetic resonance spectroscopy of normal human brain and glioma: a quantitative in vivo study. Chin Med J (Engl) 2005; 118:12511257.

    • Search Google Scholar
    • Export Citation
  • 46. Yue Q, Isobe T & Shibata Y, et al. New observations concerning the interpretation of magnetic resonance spectroscopy of meningioma. Eur Radiol 2008; 18:29012911.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 47. Kraft SL, Gavin PR & DeHaan C, et al. Retrospective review of 50 canine intracranial tumors evaluated by magnetic resonance imaging. J Vet Intern Med 1997; 11:218225.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 48. Sturges BK, Dickinson PJ & Bollen AW, et al. Magnetic resonance imaging and histological classification of intracranial meningiomas in 112 dogs. J Vet Intern Med 2008; 22:586595.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 49. Cecil KM. Proton magnetic resonance spectroscopy: technique for the neuroradiologist. Neuroimaging Clin N Am 2013; 23:381392.

  • 50. Granger N, Smith PM, Jeffery ND. Clinical findings and treatment of non-infectious meningoencephalomyelitis in dogs: a systematic review of 457 published cases from 1962 to 2008. Vet J 2010; 184:290297.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 51. Perkins CJ, Kahya E & Roque CT, et al. Fluid-attenuated inversion recovery and diffusion- and perfusion-weighted MRI abnormalities in 117 consecutive patients with stroke symptoms. Stroke 2001; 32:27742781.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 52. Negendank WG, Sauter R & Brown TR, et al. Proton magnetic resonance spectroscopy in patients with glial tumors: a multi-center study. J Neurosurg 1996; 84:449458.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 53. Isobe T, Matsumura A & Anno I, et al. Quantification of cerebral metabolites in glioma patients with proton MR spectroscopy using T2 relaxation time correction. Magn Reson Imaging 2002; 20:343349.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 54. Pirzkall A, Nelson SJ & McKnight TR, et al. Metabolic imaging of low-grade gliomas with three-dimensional magnetic resonance spectroscopy. Int J Radiat Oncol Biol Phys 2002; 53:12541264.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Figure 1—

    Representative 1H-MRS spectrum of a histologically confirmed meningioma in the piriform lobe of a dog. The NAA peak is considerably smaller than the choline peak. Ratios calculated for this spectrum are as follows: NAA-to-choline ratio, 0.47; NAA-to-creatine ratio, 1.02; and choline-to-creatine ratio, 2.18. The NAA-to-choline ratio is lower than the reference range (0.687 to 1.568) and the choline-to-creatine ratio is higher than the reference range (0.947 to 1.797) reported for the piriform lobe in healthy dogs.32 Cho = Choline. Cr = Creatine.

  • Figure 2—

    Representative 1H-MRS spectrum of histologically confirmed granulomatous meningoencephalitis in the piriform lobe of a dog. The NAA peak is smaller than the choline peak, but the difference in height between the 2 peaks is less than in neoplastic lesions. The ratios calculated for this spectrum are as follows: NAA-to-choline ratio, 0.80; NAA-to-creatine ratio, 1.08; and choline-to-creatine ratio, 1.35. All ratios calculated for this spectrum are in the reference ranges (0.687 to 1.568, 0.984 to 1.818, and 0.947 to 1.797, respectively) reported for the piriform lobe in healthy dogs.32 Notice the small lactate peak (inverted doublet at 1. 3 ppm).

  • 1. Song RB, Vite CH & Bradley CW, et al. Postmortem evaluation of 435 cases of intracranial neoplasia in dogs and relationship of neoplasm with breed, age, and body weight. J Vet Intern Med 2013; 27:11431152.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 2. Podell M, Fenner WR, Powers JD. Seizure classification in dogs from a nonreferral-based population. J Am Vet Med Assoc 1995; 206:17211728.

    • Search Google Scholar
    • Export Citation
  • 3. Fluehmann G, Doherr MG, Jaggy A. Canine neurological diseases in a referral hospital population between 1989 and 2000 in Switzerland. J Small Anim Pract 2006; 47:582587.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 4. Cherubini GB, Mantis P & Martinez TA, et al. Utility of magnetic resonance imaging for distinguishing neoplastic from nonneoplastic brain lesions in dogs and cats. Vet Radiol Ultrasound 2005; 46:384387.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 5. Young BD, Fosgate GT & Holmes SP, et al. Evaluation of standard magnetic resonance characteristics used to differentiate neoplastic, inflammatory, and vascular brain lesions in dogs. Vet Radiol Ultrasound 2014; 55:399406.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 6. Ródenas S, Pumarola M & Gaitero L, et al. Magnetic resonance imaging findings in 40 dogs with histologically confirmed intracranial tumours. Vet J 2011; 187:8591.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 7. Wolff CA, Holmes SP & Young BD, et al. Magnetic resonance imaging for the differentiation of neoplastic, inflammatory, and cerebrovascular brain disease in dogs. J Vet Intern Med 2012; 26:589597.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 8. Leclerc MK, d'Anjou MA & Blond L, et al. Interobserver agreement and diagnostic accuracy of brain magnetic resonance imaging in dogs. J Am Vet Med Assoc 2013; 242:16881695.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 9. Zarfoss M, Schatzberg S & Venator K, et al. Combined cytosine arabinoside and prednisone therapy for meningoencephalitis of unknown aetiology in 10 dogs. J Small Anim Pract 2006; 47:588595.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 10. Adamo PF, Adams WM, Steinberg H. Granulomatous meningoencephalomyelitis in dogs. Compend Contin Educ Pract Vet 2007; 29:678690.

  • 11. Talarico LR, Schatzberg SJ. Idiopathic granulomatous and necrotising inflammatory disorders of the canine central nervous system: a review and future perspectives. J Small Anim Pract 2010; 51:138149.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 12. Shimizu H, Kumabe T & Tominaga T, et al. Noninvasive evaluation of malignancy of brain tumors with proton MR spectroscopy. AJNR Am J Neuroradiol 1996; 17:737747.

    • Search Google Scholar
    • Export Citation
  • 13. Meyerand ME, Pipas JM & Mamourian A, et al. Classification of biopsy-confirmed brain tumors using single-voxel MR spectroscopy. AJNR Am J Neuroradiol 1999; 20:117123.

    • Search Google Scholar
    • Export Citation
  • 14. McKnight TR, von dem Bussche MH & Vigneron DB, et al. Histopathological validation of a three-dimensional magnetic resonance spectroscopy index as a predictor of tumor presence. J Neurosurg 2002; 97:794802.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 15. Law M, Yang S & Wang H, et al. Glioma grading: sensitivity, specificity, and predictive values of perfusion MR imaging and proton MR spectroscopic imaging compared with conventional MR imaging. AJNR Am J Neuroradiol 2003; 24:19891998.

    • Search Google Scholar
    • Export Citation
  • 16. Law M. MR spectroscopy of brain tumors. Top Magn Reson Imaging 2004; 15:291313.

  • 17. Sibtain NA, Howe FA, Saunders DE. The clinical value of proton magnetic resonance spectroscopy in adult brain tumours. Clin Radiol 2007; 62:109119.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 18. Rémy C, Grand S & Laï ES, et al. 1H MRS of human brain abscesses in vivo and in vitro. Magn Reson Med 1995; 34:508514.

  • 19. Burtscher IM, Holtås S. In vivo proton MR spectroscopy of untreated and treated brain abscesses. AJNR Am J Neuroradiol 1999; 20:10491053.

    • Search Google Scholar
    • Export Citation
  • 20. Sandok EK, O'Brien TJ & Jack CR, et al. Significance of cerebellar atrophy in intractable temporal lobe epilepsy: a quantitative MRI study. Epilepsia 2000; 41:13151320.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 21. Neppl R, Nguyen CM & Bowen W, et al. In vivo detection of postictal perturbations of cerebral metabolism by use of proton MR spectroscopy: preliminary results in a canine model of prolonged generalized seizures. AJNR Am J Neuroradiol 2001; 22:19331943.

    • Search Google Scholar
    • Export Citation
  • 22. Bowser R, Turner MR, Shefner J. Biomarkers in amyotrophic lateral sclerosis: opportunities and limitations. Nat Rev Neurol 2011; 7:631638.

  • 23. Miller BL, Moats RA & Shonk T, et al. Alzheimer disease: depiction of increased cerebral myo-inositol with proton MR spectroscopy. Radiology 1993; 187:433437.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 24. Huang W, Alexander GE & Chang L, et al. Brain metabolite concentration and dementia severity in Alzheimer's disease: a 1H MRS study. Neurology 2001; 57:626632.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 25. Bianchi MC, Tosetti M & Battini R, et al. Proton MR spectroscopy of mitochondrial diseases: analysis of brain metabolic abnormalities and their possible diagnostic relevance. AJNR Am J Neuroradiol 2003; 24:19581966.

    • Search Google Scholar
    • Export Citation
  • 26. Sorby WA. An evaluation of magnetic resonance imaging at The Royal North Shore Hospital of Sydney, 1986–1987. Med J Aust 1989; 151:811, 1418.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 27. Rand SD, Prost R & Haughton V, et al. Accuracy of single-voxel proton MR spectroscopy in distinguishing neoplastic from nonneoplastic brain lesions. AJNR Am J Neuroradiol 1997; 18:16951704.

    • Search Google Scholar
    • Export Citation
  • 28. Lee SH, Kim SY & Woo DC, et al. Differential neurochemical responses of the canine striatum with pentobarbital or ketamine anesthesia: a 3T proton MRS study. J Vet Med Sci 2010; 72:583587.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 29. Kang BT, Jang DP & Lee JH, et al. Detection of cerebral metabolites in a canine model of ischemic stroke using 1H magnetic resonance spectroscopy. Res Vet Sci 2009; 87:300306.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 30. Barreiro CJ, Williams JA & Fitton TP, et al. Noninvasive assessment of brain injury in a canine model of hypothermic circulatory arrest using magnetic resonance spectroscopy. Ann Thorac Surg 2006; 81:15931598.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 31. Ober CP, Warrington CD & Feeney DA, et al. Optimizing a protocol for 1H-magnetic resonance spectroscopy of the canine brain at 3T. Vet Radiol Ultrasound 2013; 54:149158.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 32. Warrington CD, Feeney DA & Ober CP, et al. Relative metabolite concentrations and ratios determined by use of 3-T regionspecific proton magnetic resonance spectroscopy of the brain of healthy Beagles. Am J Vet Res 2013; 74:12911303.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 33. Ono K, Kitagawa M & Ito D, et al. Regional variations and age-related changes detected with magnetic resonance spectroscopy in the brain of healthy dogs. Am J Vet Res 2014; 75:179186.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 34. Martin-Vaquero P, da Costa RC & Echandi RL, et al. Magnetic resonance spectroscopy of the canine brain at 3.0 T and 7.0 T. Res Vet Sci 2012; 93:427429.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 35. Chawla S, Gupta RK & Husain N, et al. Prediction of viability of porcine neurocysticercosis with proton magnetic resonance spectroscopy: correlation with histopathology. Life Sci 2004; 74:10811092.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 36. Bertholdo D, Watcharakorn A, Castillo M. Brain proton magnetic resonance spectroscopy: introduction and overview. Neuroimaging Clin N Am 2013; 23:359380.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 37. Kingsley PB, Shah TC, Woldenberg R. Identification of diffuse and focal brain lesions by clinical magnetic resonance spectroscopy. NMR Biomed 2006; 19:435462.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 38. Pivawer G, Law M, Zagzag D. Perfusion MR imaging and proton MR spectroscopic imaging in differentiating necrotizing cerebritis from glioblastoma multiforme. Magn Reson Imaging 2007; 25:238243.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 39. Nagar VA, Ye J & Xu M, et al. Multivoxel MR spectroscopic imaging—distinguishing intracranial tumours from non-neoplastic disease. Ann Acad Med Singapore 2007; 36:309313.

    • Search Google Scholar
    • Export Citation
  • 40. Gupta RK, Cloughesy TF & Sinha U, et al. Relationships between choline magnetic resonance spectroscopy, apparent diffusion coefficient and quantitative histopathology in human glioma. J Neurooncol 2000; 50:215226.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 41. Lehnhardt FG, Röhn G & Ernestus RI, et al. 1H- and 31P-MR spectroscopy of primary and recurrent human brain tumors in vitro: malignancy-characteristic profiles of water soluble and lipophilic spectral components. NMR Biomed 2001; 14:307317.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 42. Chang KH, Song IC & Kim SH, et al. In vivo single-voxel proton MR spectroscopy in intracranial cystic masses. AJNR Am J Neuroradiol 1998; 19:401405.

    • Search Google Scholar
    • Export Citation
  • 43. Rock JP, Hearshen D & Scarpace L, et al. Correlations between magnetic resonance spectroscopy and image-guided histopathology, with special attention to radiation necrosis. Neurosurgery 2002; 51:912919.

    • Search Google Scholar
    • Export Citation
  • 44. Snyder JM, Shofer FS & Van Winkle TJ, et al. Canine intracranial primary neoplasia: 173 cases (1986–2003). J Vet Intern Med 2006; 20:669675.

    • Search Google Scholar
    • Export Citation
  • 45. Tong ZY, Yamaki T, Wang YJ. Proton magnetic resonance spectroscopy of normal human brain and glioma: a quantitative in vivo study. Chin Med J (Engl) 2005; 118:12511257.

    • Search Google Scholar
    • Export Citation
  • 46. Yue Q, Isobe T & Shibata Y, et al. New observations concerning the interpretation of magnetic resonance spectroscopy of meningioma. Eur Radiol 2008; 18:29012911.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 47. Kraft SL, Gavin PR & DeHaan C, et al. Retrospective review of 50 canine intracranial tumors evaluated by magnetic resonance imaging. J Vet Intern Med 1997; 11:218225.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 48. Sturges BK, Dickinson PJ & Bollen AW, et al. Magnetic resonance imaging and histological classification of intracranial meningiomas in 112 dogs. J Vet Intern Med 2008; 22:586595.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 49. Cecil KM. Proton magnetic resonance spectroscopy: technique for the neuroradiologist. Neuroimaging Clin N Am 2013; 23:381392.

  • 50. Granger N, Smith PM, Jeffery ND. Clinical findings and treatment of non-infectious meningoencephalomyelitis in dogs: a systematic review of 457 published cases from 1962 to 2008. Vet J 2010; 184:290297.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 51. Perkins CJ, Kahya E & Roque CT, et al. Fluid-attenuated inversion recovery and diffusion- and perfusion-weighted MRI abnormalities in 117 consecutive patients with stroke symptoms. Stroke 2001; 32:27742781.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 52. Negendank WG, Sauter R & Brown TR, et al. Proton magnetic resonance spectroscopy in patients with glial tumors: a multi-center study. J Neurosurg 1996; 84:449458.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 53. Isobe T, Matsumura A & Anno I, et al. Quantification of cerebral metabolites in glioma patients with proton MR spectroscopy using T2 relaxation time correction. Magn Reson Imaging 2002; 20:343349.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 54. Pirzkall A, Nelson SJ & McKnight TR, et al. Metabolic imaging of low-grade gliomas with three-dimensional magnetic resonance spectroscopy. Int J Radiat Oncol Biol Phys 2002; 53:12541264.

    • Crossref
    • Search Google Scholar
    • Export Citation

Advertisement