SENSITIVITY OF LOW‐FIELD T2* IMAGES FOR DETECTING THE PRESENCE AND SEVERITY OF HISTOPATHOLOGIC MENISCAL LESIONS IN DOGS

The sensitivity of low‐field magnetic resonance (MR) T2^* images for predicting the presence of meniscal lesions was determined in 12 dogs with naturally‐occurring cranial cruciate ligament rupture and three control dogs, using histopathology as the reference standard. Previously published grading systems were used to grade the severity of meniscal lesions on MR images, gross inspection and histopathology. Focal areas of increased signal intensity were detected in 11/12 symptomatic dogs and 3/3 control dogs. Lesions mimicking meniscal tears (pseudotears) were identified at junctions between meniscal margins and adjacent connective tissue in control dogs and dogs with naturally occurring disease. Histopathologic lesions were present in all menisci of both symptomatic and control dogs, including the menisci from two affected dogs that appeared grossly normal but were removed and submitted based on MR imaging findings. Histopathologic lesions identified included hyaline cartilage metaplasia and changes in the amount of ground substance and cellularity. The sensitivity of MR imaging for detecting the presence of meniscal histopathologic lesions was 90% in symptomatic dogs and 91% in control dogs. However, agreement between severity scores for the different tests was poor. Low‐field MR imaging is a sensitive test for predicting the presence but not severity of meniscal histopathologic lesions in dogs with naturally‐occurring cranial cruciate ligament rupture. Findings also supported previous studies indicating that histopathologic lesions can be present in dogs with grossly normal menisci. An improved grading system for comparing MR images and histopathologic severity of meniscal lesions in dogs is needed.     

MAGNETIC RESONANCE IMAGING FINDINGS IN 25 DOGS WITH INFLAMMATORY CEREBROSPINAL FLUID

To describe the signs that may be associated with intracranial inflammatory conditions, magnetic resonance (MR) images of 25 dogs that had inflammatory cerebrospinal fluid (CSF) were mixed with those of a control group of 40 dogs that had CSF negative for inflammatory disease and reviewed without knowledge of the clinical signs or diagnosis. CSF was considered inflammatory if the protein level was > 0.25 g/l and the white cell count was > 5 mm(-3). Abnormalities were found by MR imaging in 19 (76%) dogs with inflammatory CSF. Two dogs had focal lesions, 10 had multifocal lesions, and seven had diffuse lesions. Lesions affected all divisions of the brain. Mass effect was identified in seven (28%) dogs, including one that had a choroid plexus carcinoma. Lesions were hyperintense in T2-weighted images in 18 dogs and hypointense in T1-weighted images in six dogs. Multifocal or diffuse intraaxial lesions that were hyperintense in T2-weighted images were observed in 17 (68%) dogs with inflammatory CSF. Administration of gadolinium resulted in enhancement of intraaxial lesions in nine (36%) dogs and enhancement of meninges in seven (28%) dogs. Six (24%) dogs with inflammatory CSF had images interpreted as normal.     

CONTRAST AGENT Gd‐EOB‐DTPA (EOB·Primovist®) FOR LOW‐FIELD MAGNETIC RESONANCE IMAGING OF CANINE FOCAL LIVER LESIONS

Contrast‐enhanced magnetic resonance (MR) imaging with a new liver‐specific contrast agent gadolinium‐ethoxybenzyl‐diethylenetriamine penta‐acetic acid (Gd‐EOB‐DTPA; EOB·Primovist®) was studied in 14 normal beagles and 9 dogs with focal liver lesions. Gd‐EOB‐DTPA accumulates in normally functioning hepatocytes 20 min after injection. As with Gd‐DTPA, it is also possible to perform a dynamic multiphasic examination of the liver with Gd‐EOB‐DTPA, including an arterial phase and a portal venous phase. First, a reliable protocol was developed and the appropriate timings for the dynamic study and the parenchymal phase in normal dogs using Gd‐EOB‐DTPA were determined. Second, the patterns of these images were evaluated in patient dogs with hepatic masses. The optimal time of arterial imaging was from 15 s after injection, and the optimal time for portal venous imaging was from 40 s after injection. Meanwhile, the optimal time to observe changes during the hepatobiliary phase was from 20 min after injection. In patient dogs, 11 lesions were diagnosed as malignant tumors; all were hypointense to the surrounding normal liver parenchyma during the hepatobiliary phase. Even with a low‐field MR imaging unit, the sequences afforded images adequate to visualize the liver parenchyma and to detect tumors within an appropriate scan time. Contrast‐enhanced MR imaging with Gd‐EOB‐DTPA provides good demarcation on low‐field MR imaging for diagnosing canine focal liver lesions.     

APPEARANCE OF THE CANINE MENINGES IN SUBTRACTION MAGNETIC RESONANCE IMAGES

The canine meninges are not visible as discrete structures in noncontrast magnetic resonance (MR) images, and are incompletely visualized in T1‐weighted, postgadolinium images, reportedly appearing as short, thin curvilinear segments with minimal enhancement. Subtraction imaging facilitates detection of enhancement of tissues, hence may increase the conspicuity of meninges. The aim of the present study was to describe qualitatively the appearance of canine meninges in subtraction MR images obtained using a dynamic technique. Images were reviewed of 10 consecutive dogs that had dynamic pre‐ and postgadolinium T1W imaging of the brain that was interpreted as normal, and had normal cerebrospinal fluid. Image‐anatomic correlation was facilitated by dissection and histologic examination of two canine cadavers. Meningeal enhancement was relatively inconspicuous in postgadolinium T1‐weighted images, but was clearly visible in subtraction images of all dogs. Enhancement was visible as faint, small‐rounded foci compatible with vessels seen end on within the sulci, a series of larger rounded foci compatible with vessels of variable caliber on the dorsal aspect of the cerebral cortex, and a continuous thin zone of moderate enhancement around the brain. Superimposition of color‐encoded subtraction images on pregadolinium T1‐ and T2‐weighted images facilitated localization of the origin of enhancement, which appeared to be predominantly dural, with relatively few leptomeningeal structures visible. Dynamic subtraction MR imaging should be considered for inclusion in clinical brain MR protocols because of the possibility that its use may increase sensitivity for lesions affecting the meninges.     

CANINE MENINGEAL DISEASE: ASSOCIATIONS BETWEEN MAGNETIC RESONANCE IMAGING SIGNS AND HISTOLOGIC FINDINGS

In order to compare the accuracy of MR sequences for diagnosis of meningeal disease, MR images of the brain, and histopathologic specimens including the meninges of 60 dogs were reviewed retrospectively by independent observers in a cross‐sectional study. MR images included T1‐weighted pre‐ and postgadolinium images, subtraction images, T2‐weighted images, and T2‐weighted fluid‐attenuated inversion‐recovery (FLAIR) images. Pathologic changes affected the pachymeninges in 16 dogs, leptomeninges in 35 dogs, and brain in 38 dogs. The meninges were normal in 12 dogs. Meninges were classified histopathologically as normal (grade 0), slightly or inconsistently affected (grade 1), or markedly affected (grade 2). When applying relaxed pathologic criteria (grades 0 and 1 considered normal), the results of ROC analysis (area under curve, AUC) were: T1‐weighted postcontrast images 0.74; subtraction images 0.7; T2‐weighted images 0.68; FLAIR images 0.56. The difference in AUC between T1‐weighted postgadolinium images and FLAIR images was significant (P = 0.04). AUC for FLAIR images was not significantly different from 0.5. When applying strict pathologic criteria (only grade 0 considered normal), none of the MR sequences had AUC significantly different from 0.5. On the basis of T1‐weighted postgadolinium images and subtraction images, correct anatomic classification of lesions occurred more often for pachymeningeal than leptomeningeal lesions (P < 0.001). Overall, MR imaging had low sensitivity for diagnosis of meningeal pathology in dogs, particularly for changes affecting the leptomeninges. Subtraction images had similar accuracy to T1‐weighted postgadolinium images for meningeal lesions in dogs. T2‐weighted FLAIR images appear to have limited diagnostic utility for meningeal lesions.     

CONTRAST‐ENHANCED MAGNETIC RESONANCE ANGIOGRAPHY FOR DIAGNOSIS OF PORTOSYSTEMIC SHUNTS IN 10 DOGS

Computed tomography angiography, sonography, scintigraphy, and portography can be used to evaluate the portal vasculature to evaluate for a portosystemic shunt (PSS). Time‐of‐flight magnetic resonance angiography (TOF‐MRA) and contrast‐enhanced MRA (CE‐MRA) are other potentially useful techniques. The aim of this study was to evaluate CE‐MRA in 10 dogs suspected of having a PSS. Noncontrast MR images of the abdomen were obtained using a Siemens Symphony MR‐scanner (1.5 T) and a T1‐weighted FLASH‐3D sequence with a very short scan time (about 20 s). After injection of contrast medium, the initial sequence was repeated five times. The sequence with the best contrast medium filling of the portal vasculature was selected subjectively, subtracted from the initial survey image series, and a maximum intensity projection (MIP) of the subtraction data, in multiple views, was created. The cross‐sectional and MIP images were evaluated for abnormal portosystemic vasculature. A single PSS was identified and confirmed at surgery in all dogs. A portocaval shunt was found in five dogs, a portophrenic shunt in three dogs, a portoazygos shunt in one, and a central divisional intrahepatic shunt in one other dog. Based on our results, CE‐MRA is a useful tool for imaging abdominal and portal vasculature and for the diagnosis of a PSS.     

Effect of acquisition time and chemical fat suppression on meningeal enhancement on MR imaging in dogs

Our purpose was to characterize meningeal gadolinium enhancement on magnetic resonance (MR) imaging in dogs with inflammatory and neoplastic diseases, and to assess interobserver variability and the impact of delayed acquisition and chemical fat saturation on its conspicuity. Transverse T1-weighted FLAIR images were acquired prior to, and immediately following gadolinium injection (T0), and at 5 (T5) and 15-20 min delay (TD), with and without fat suppression, in 155 consecutive dogs imaged for suspected brain disease. The agreement on meningeal enhancement was globally substantial (kappa = 0.61) and the likelihood of obtaining a definite diagnosis was significantly increased with the use of fat suppression (P < or = 0.004). Meningeal enhancement was judged definitively present by consensus in 46 of 155 (30%) dogs. Of these, meningeal enhancement was characterized qualitatively and quantitatively in 30 dogs with a clinical diagnosis (18 inflammatory, 11 neoplastic, 1 infarct), and image sequences were compared. Meningeal enhancement was more often diffuse and leptomeningeal in animals with inflammation versus neoplasia (50% vs. 42%, and 69% vs. 48%, respectively), but significant associations were not found. Meningeal thickness and contrast ratio were higher with neoplasia (P < or = 0.02), but results did not vary significantly between series for either group. Yet, images with fat suppression were most useful 50% of the time for definite diagnosis and/or characterization of meningeal enhancement. While delayed image acquisition following gadolinium injection does not improve characterization of meningeal enhancement in dogs, fat suppression is beneficial qualitatively.     

MAGNETIC RESONANCE IMAGING FEATURES OF INTRACRANIAL GRANULAR CELL TUMORS IN SIX DOGS

Magnetic resonance (MR) imaging characteristics of intracranial granular cell tumors (GCTs) have been previously reported in three dogs. The goal of this retrospective study was to examine a larger number of dogs and determine whether distinctive MR characteristics of intracranial GCTs could be identified. Six dogs with histologically confirmed intracranial GCTs and MR imaging were included. Tumor location, size, mass effect, T1‐ and T2‐weighted signal intensity, and peritumoral edema MR characteristics were recorded. In all dogs, GCTs appeared as well‐defined, extra‐axial masses with a plaque‐form, sessile distribution involving the meninges. All tumors were located along the convexity of the cerebrum, the falx cerebri, or the ventral floor of the cranial vault. All tumors were mildly hyperintense on T1‐weighted images, and iso‐ to hyperintense on T2‐weighted images. A moderate‐to‐severe degree of peritumoral edema and mass effect were evident in all dogs. Findings indicated that, while several MR imaging characteristics were consistently identified in canine cerebral GCTs, none of these characteristics were unique or distinctive for this tumor type alone.     

EVALUATION OF STANDARD MAGNETIC RESONANCE CHARACTERISTICS USED TO DIFFERENTIATE NEOPLASTIC,INFLAMMATORY, AND VASCULAR BRAIN LESIONS IN DOGS

Magnetic resonance (MR) imaging characteristics are commonly used to help predict intracranial disease categories in dogs, however, few large studies have objectively evaluated these characteristics. The purpose of this retrospective study was to evaluate MR characteristics that have been used to differentiate neoplastic, inflammatory, and vascular intracranial diseases in a large, multi‐institutional population of dogs. Medical records from three veterinary teaching hospitals were searched over a 6‐year period for dogs that had diagnostic quality brain MR scans and histologically confirmed intracranial disease. Three examiners who were unaware of histologic diagnosis independently evaluated 19 MR lesion characteristics totaling 57 possible responses. A total of 75 dogs with histologically confirmed intracranial disease were included in analyses: 51 with neoplasia, 18 with inflammatory disease, and six with cerebrovascular disease. Only strong contrast enhancement was more common in neoplasia than other disease categories. A multivariable statistical model suggested that extra‐axial origin, T2‐FLAIR mixed intensity, and defined lesion margins were also predictive of neoplasia. Meningeal enhancement, irregular lesion shape, and multifocal location distinguished inflammatory diseases from the other disease categories. No MR characteristics distinguished vascular lesions and these appeared most similar to neoplasia. These results differed from a previous report describing seven MR characteristics that were predictive of neoplasia in dogs and cats. Findings from the current study indicated that the high performance of MR for diagnosing canine intracranial diseases might be due to evaluator recognition of combinations of MR characteristics vs. relying on any one MR characteristic alone.     

MAGNETIC RESONANCE ARTHROGRAPHY OF THE CANINE STIFLE JOINT TECHNIQUE AND APPLICATIONS IN ELEVEN MILITARY DOGS

The stifle joints of eleven military working dogs were evaluated using conventional magnetic resonance (MR) imaging and MR arthrography. A protocol optimizing MR imaging of the canine stifle joint is discussed, as well as potential uses for administration of intra-articular gadolinium. The technique for performing MR arthrography is described, and post-contrast image findings are reviewed. MR arthrography was performed by using an intra-articular injection of diluted gadolinium. Consistently good quality images were obtained, and no complications were clinically detected following MR arthrography. Cranial cruciate ligament abnormalities were seen in six dogs, meniscal abnormalities were visualized in nine menisci, and synovitis and medial ligament strain were seen in eight dogs. Surgical and post-mortem confirmation of these findings is discussed in seven dogs. Although MR arthrography adds an invasive procedure to conventional MR imaging, it can provide useful information on pathologic changes in the canine stifle joint.     

POSITRON EMISSION TOMOGRAPHY FEATURES OF CANINE NECROTIZING MENINGOENCEPHALITIS

A Yorkshire terrier and a Chihuahua were referred for acute onset, generalized tonic‐clonic seizures and were suspected to have meningoencephalitis based on magnetic resonance (MR) imaging findings. Brain lesions appeared hyperintense with T2‐weighted imaging and hypointense with T1‐weighted imaging, and were characteristic of necrotizing meningoencephalitis. Both dogs were diagnosed with necrotizing meningoencephalitis based on pathologic findings. Fluorine‐18 fluorodeoxyglucose positron emission tomography (FDG‐PET) was performed on both animals before euthanasia with the permission of the owner. In FDG‐PET images, these lesions seen in MR images were characterized by multifocal or diffuse hypometabolism. Our FDG‐PET results provided evidence of glucose hypometabolism in areas of necrosis and cavitation associated with necrotizing meningoencephalitis. FDG‐PET has the potential to provide valuable diagnostic information in dogs with suspected necrotizing encephalitis.     

FAT‐SUPPRESSED SPOILED GRADIENT‐RECALLED IMAGING OF EQUINE METACARPOPHALANGEAL ARTICULAR CARTILAGE

The purpose was to evaluate the capacity of 1.5 T magnetic resonance (MR) imaging to assess articular cartilage in racehorses with naturally occurring metacarpophalangeal joint osteoarthritis. A sagittal, three‐dimensional spoiled gradient‐recalled echo (SPGR) with fat saturation (FS) sequence was acquired ex vivo on 20 joints. Following joint dissection, specific areas on the third metacarpal condyle were designated for subsequent sampling for histologic cartilage thickness measurement and modified Mankin scoring. Cartilage thickness was measured and cartilage signal intensity was also graded (0–3) on MR images at these selected metacarpal sites. Cartilage structure was graded (0–3) macroscopically and on MR images by two examiners in defined subregions of the proximal phalanx, third metacarpal, and proximal sesamoid bones. There was good precision (mean error 0.11 mm) and moderate correlation (r=0.44; P<0.0001) of cartilage thickness measurements between MR images (0.90±0.17 mm) and histology (0.79±0.16 mm). There was moderate correlation between modified Mankin histologic score and signal intensity of cartilage (r=0.36; P<0.01) or MR cartilage structure assessment (r=0.49, P>0.001) on SPGR‐FS. The sensitivity to detect full‐thickness cartilage erosion on MR was only moderate (0.56), and these lesions were often underestimated, particularly when linear in nature. However, the specificity to detect such lesions on MR was high (0.92). While few limitations were identified, the use of a clinically applicable SPGR‐FS sequence allows a reasonably accurate method to assess structural changes affecting the articular cartilage of the equine metacarpophalangeal joint.     

MAGNETIC RESONANCE IMAGING FEATURES OF MULTILOBULAR OSTEOCHONDROSARCOMA IN 3 DOGS

Three dogs with multilobular osteochondrosarcoma of the skull were evaluated using magnetic resonance (MR) imaging. Spin echo T1, T2, proton weighted and post contrast T1W images were obtained with a 1.5 Tesla magnet. The MR imaging findings were similar in all three dogs with mixed signal intensities in the T1W, T2W and proton weighted images and fairly large areas of contrast enhancement in the post contrast T1W images. The extent of brain and soft tissue involvement were well delineated and provided useful information concerning surgical planning. MR imaging provided a useful method of evaluating dogs with skull tumors.     

MAGNETIC RESONANCE IMAGING OF OCULAR AND ORBITAL DISEASE IN 5 DOGS AND A CAT

Magnetic resonance (MR) images were acquired in five dogs and one cat with ocular and orbital disease. MR images were obtained in the dorsal or oblique dorsal, and oblique sagittal planes. Pathologic changes identified in MR images included inflammatory lesions, cystic structures, and neoplasms. All abnormalities were readily apparent in TI-weighted images. MR findings in affected animals were often similar in signal intensity, location, and growth pattern to those found in people with comparable diseases. Although no MR changes were considered pathognomonic for a given disease, MR imaging provided detailed information on the homogeneity, extent and invasiveness of the lesions.     

Reliability of high‐ and low‐field magnetic resonance imaging systems for detection of cartilage and bone lesions in the equine cadaver fetlock

Reasons for performing study: To determine the reliability of 2 magnetic resonance imaging (MRI) systems for detection of cartilage and bone lesions of the equine fetlock. Objectives: To test the hypotheses that lesions in cartilage, subchondral and trabecular bone of the equine fetlock verified using histopathology can be detected on high‐ and low‐field MR images with a low incidence of false positive or negative results; that low‐field images are less reliable than high‐field images for detection of cartilage lesions; and that combining results of interpretation from different pulse sequences increases detection of cartilage lesions. Methods: High‐ and low‐field MRI was performed on 19 limbs from horses identified with fetlock lameness prior to euthanasia. Grading systems were used to score cartilage, subchondral and trabecular bone on MR images and histopathology. Sensitivity and specificity were calculated for images. Results: High‐field T2*‐weighted gradient echo (T2*W‐GRE) and low‐field T2‐weighted fast spin echo (T2W‐FSE) images had high sensitivity but low specificity for detection of cartilage lesions. All pulse sequences had high sensitivity and low–moderate specificity for detection of subchondral bone lesions and moderate sensitivity and moderate–high specificity for detection of trabecular bone lesions (histopathology as gold standard). For detection of lesions of trabecular bone low‐field T2*W‐GRE images had higher sensitivity and specificity than T2W‐FSE images. Conclusions: There is high likelihood of false positive results using high‐ or low‐field MRI for detection of cartilage lesions and moderate–high likelihood of false positive results for detection of subchondral bone lesions compared with histopathology. Combining results of interpretation from different pulse sequences did not increase detection of cartilage lesions. MRI interpretation of trabecular bone was more reliable than cartilage or subchondral bone in both MR systems. Potential relevance: Independent interpretation of a variety of pulse sequences may maximise detection of cartilage and bone lesions in the fetlock. Clinicians should be aware of potential false positive and negative results.     

RELIABILITY OF T2‐WEIGHTED SAGITTAL MAGNETIC RESONANCE IMAGES FOR DETERMINING THE LOCATION OF COMPRESSIVE DISk HERNIATION IN DOGS

Magnetic resonance imaging is used commonly to diagnose intervertebral disk herniation in dogs. It is common to locate areas of suspected compression on sagittal T2‐weighted (T2‐W) images and then obtain limited transverse images in these areas to reduce the acquisition time (a step‐by‐step approach). Our objective was to assess the frequency of correct localization of spinal cord compression due to disk herniation using only the sagittal images. The results from isolated readings of the sagittal T2‐W images alone or combined with a single‐shot fast spin echo (SSFSE) slab in 118 dogs were compared with a gold standard, based on a consensual reading of all images available, including complete transverse images across the entire spinal segments under study. The sites of compression were localized correctly from the sagittal images in 89.8% of dogs. If only the most significant lesions were accounted for, the percentage increased up to 95.2%. In 54.9% of the readings with incorrect localization, the actual compressive site was immediately adjacent to the one suspected from review of the sagittal images. The frequency of correct localization was higher in the cervical region, and was increased by examination of the SSFSE slab. The most common cause of disagreement was the presence of multiple degenerate bulging disks. Based on these results we recommend obtaining transverse images across the entire segment when multiple bulging disks are present. It is also recommended to obtain transverse images across the spaces immediately adjacent to the suspected site of herniation from review of the sagittal images.     

IMAGING DIAGNOSIS—MALIGNANT PERIPHERAL NERVE SHEATH TUMOR PRESENTING AS AN INTRA‐AXIAL BRAIN MASS IN A YOUNG DOG

A 3‐year‐old Labrador retriever was presented with acute onset seizures. Magnetic resonance imaging demonstrated an intra‐axial mass affecting the right temporal lobe of the brain. Surgical resection and histopathological findings were most consistent with a malignant peripheral nerve sheath tumor. After initial recovery, deterioration 3 months post surgery prompted euthanasia. Post‐mortem revealed a mass protruding from the ventral surface of the temporal lobe, encroaching upon the optic chiasm and invading the brain. Histopathology findings were again consistent with malignant peripheral nerve sheath tumor. Although rare, this tumor should be included as a possible differential diagnosis for intra‐axial brain masses in dogs.