In this study the use of IA and IV contrast enhanced CT of the equine head was described using clinical cases. Several studies have described the use of contrast medium in horses [9, 13, 20–22, 26, 27], however the different routes of contrast administration have not been compared. As we used two different scanners in this study, interscanner variability needed to be considered [29].
The absolute attenuation values of the studied structures were only used to determine the structures enhancement and are not directly compared between the two scanners. The contrast medium administration protocols and scanners were consistent pairs. In the presence of interscanner or intrascanner variability, significant differences in enhancement can therefore not be explained without considering differences between the two protocols.
It was hypothesized that, the IA contrast medium administration technique would result in a similar or higher contrast enhancement using a lower volume of contrast medium compared to the IV technique. Contrast enhancement of a structure is intrinsically depending on its specific anatomy and vascular supply, extrinsically on the local dose of contrast medium and the timing of injection relative to the scan protocol. In the described study no controls were performed to exclude the influence of the individual case’s intrinsic properties on contrast enhancement of the different structures. Recognising this shortcoming, we assume differences in contrast enhancement was only influenced by the extrinsic properties.
The local contrast medium dose compared to the total dose administered and the influence of timing are different for the two techniques. The local contrast medium dose in the IV technique is influenced by the amount of contrast medium injected as a bolus and the systemic dilution (influenced mainly by the size of the individual in combination with blood pressure) of this bolus. In the IA technique the dose of contrast medium in the common carotid artery is only influenced by the local dilution with the blood in the common carotid artery at the moment of injection.
Timing of injection relative to the scan protocol for the IV technique is crucial as selective arterial enhancement, the so called “first pass”, is only seen for a relative short period. In human a 10 s time-window of selective arterial enhancement of the carotid arteries was identified 20 s after the start of the intravenous contrast medium bolus administration [30]. The results of the present study showed semi-quantitative moderate to severe enhancement of both the arterial and the venous system in all IV cases. The continuous injection of the IA protocol is assuring the visibility of contrast medium in the arterial lumen. As previously seen for the distal limb, the short delay before initiation of the scan and the time needed to scan the entire head is assuring contrast enhancement of the vascular bed, possible extravascular structures [20, 21] and as shown in the results the venous system. Based on these findings, both techniques represent a mixed (arterial and delayed)-phase CT angiography.
Our results showed for both techniques a similar semi quantitative ranking from highest to least enhancing structures. Highest enhancement and extreme outliers were seen in the maxillary veins. Due to these extreme outliers, statistical analysis was not considered useful. Streaming was the cause of these outliers in the maxillary veins. Streaming is visible as the layered appearance of highly attenuating contrast medium and blood in the vasculature due to the lack of optimal mixing. This phenomenon was detected in both the arterial and venous systems in the IA group. In the arterial system, streaming has previously been reported after intra-arterial drug administration and depends on the rate of infusion, the type of catheter used and the position of the catheter in relation to the arterial branching [31]. Streaming in the venous system has previously been reported to occur in the portal vein following direct injection of contrast medium in the splenic parenchyma in dogs [32]. However, in several cases streaming was present in the venous system and absent in the arterial system after intra-arterial contrast administration. An explanation is lacking for this observation.
The major arteries that supply blood to the head are the both common carotid arteries (left and right side of the head) and the basilar artery (brainstem, cerebellum and caudal cerebrum) [33–35]. Both protocols differ based on the arterial supply of contrast rich blood to the head. The intravenous protocol is non-selective, using both common carotid arteries and the basilar artery, compared to selective intra-arterial protocol, using only one common carotid artery. As only in the IV group a small statistical significant difference in enhancement of the brainstem was measured in this study, the ability of the IA protocol to supply contrast rich blood to the brainstem, cerebellum and caudal cerebrum should be considered a potential limitation of the protocol. However in horses, in contrast to other domestic animals and humans, the common carotid arteries are divided into three arteries (external carotid and internal carotid but also the occipital artery). Via the occipital artery, that is forming the cerebrospinal artery, contrast medium injected in the common carotid artery is supplying the basilar artery in horses [33–35]. Further research will be needed to determine to amount of contrast medium reaching the basilar artery through this connection.
Interestingly, homogenous contrast enhancement of the nose mucosa, parotis salivary gland, cerebrum, temporal muscles and masseter muscles was seen and/or measured in the IA group at the contralateral side following unilateral contrast medium injection. This simultaneous opacification is most likely the result of communications between the right and left arterial systems, through the caudal intercarotid artery and via the arterial circle of Willis [33–36]. Regardless of the side of injection, the nasal mucosa showed a significant enhancement bilaterally. As such, horses with bilateral sinonasal disorders could possibly be scanned following a single arterial injection.
The pituitary gland, the nose septum and mucosa and the parotid salivary gland are highly vascularised structures, which showed marked enhancement on most post-contrast images. Previous reports described the accuracy of delineating the pituitary gland of adult horses, using 250 mL MD-76 (370mgI/mL) [27]. The higher dose of IV contrast medium used in this study, 400 mL Iobitridol (350mgI/mL), will probably not improve one’s ability to detect the pituitary gland and the other marked enhanced structures. However, for the nose septum and mucosa, two patterns of contrast enhancement were observed. In our experience the patchy pattern is more often visible with an increasing dose of contrast medium. Due to a local higher contrast medium concentration, this pattern of enhancement is most likely highlighting the small calibre blood vessels running in these structures. Interestingly, no significant enhancement of the nose septum in the IV group was detected, review of the data showed almost no enhancement in several cases and moderate enhancement in the other cases. Similar results were seen for the temporal and masseter muscles. The absence of significant enhancement in these structures is most likely caused by the variation in contrast dosage (due to the differences in bodyweight) between the individuals included in this group.
The remaining structures had none to very mild significant contrast enhancement. In none of these structures, the enhancement was appreciated while reviewing the cases. Increasing the local dose of contrast medium would theoretically result in higher enhancements in these structures. Based on our results the absence of visible contrast enhancement in these structures after contrast medium administration using one of the above-described protocols has to be considered normal.
A conspicuous finding in these un-enhancing structures is several higher attenuation measurements on pre-contrast images compared to post-contrast images. This is most likely due to small movements of the horses in between scans. The copied ROI’s incorporated therefore not exactly the same tissue samples of the structures. This intra-patient variability has to be considered if abnormal contrast enhancement is diagnosed.
Comparing both techniques, the results display different enhancements in three structures: the eye, pituitary gland and rectus capitis muscle. Although all three structures show a significant difference, for the eye and pituitary gland, the median difference was very small (1.1 and 0.4 HU respectively). For the rectus capitis muscle the difference was higher (6.9 HU). This muscle receives its blood supply directly from the common carotid arteries. A possible explanation for the detected difference could be the streaming causing a local exceptionally high concentration of the contrast medium in the blood supplied to the rectus capitis muscles.
Significant interobserver differences were only detected for the measurements performed on the brain and brainstem. Placement of the ROI’s on these structures differed between the two observers. Reviewing the ROI’s placed on the brain did not allow to detect a difference in attenuation (median interobserver difference of 4.5 HU), this in accordance with previous reports [4]. In contrary depending on the WW and WL setting used to review the brainstem, the grey and white matter could clearly be distinguished (median interobserver difference of 9.2 HU). Depending on the distance to the cranium for the brain and the central or peripheral localisation for the brainstem, an attenuation measurement is made of the more attenuating grey matter (observer 2) or less attenuating white matter (observer 1).
In this study, mild or marked abnormal contrast enhancement was seen in several cases. Three different abnormal enhancement patterns where detected in the included cases: enhancement due to the strong vascularisation of a soft tissue mass, rim enhancement of an encapsulated structure and increased enhancement due to inflammation of an anatomical structure. Diagnostically these findings help in the characterisation of a lesion and in the delineation of the margins between normal soft tissue and soft tissue lesions that are not always evident on pre-contrast images [7, 13, 19]. Especially contrast enhancement in structures, that are normally not or only very mild enhancing, as previously described for cerebral lesions [13, 15] and seen in case 5 and 24, have to be considered abnormal. In none of the cases with abnormal contrast enhancement both techniques were simultaneously performed, therefor-direct comparison between the two protocols in demonstrating specific lesions is not possible based on this study.
Some concerns may be raised for both techniques. In the first case were catheterization of the common carotid artery was attempted this leaded to hematoma formation. The wall of the common carotid artery is thick and the introduction of the catheter should therefor be done with a fast and strong movement to penetrate the vessel wall. This procedure had a steep learning curve considering the fact that only the first catheterization was followed by the formation of a hematoma. A specific concern for intra-carotid drug administration in humans is cerebral embolism due to air emboli [31]. Carefully removing air from the pressure injector and prefilling the extention set prior to attaching to the catheter are provisions to prevent air emboli. In human, a transient but clinically tolerable increase in intra-arterial pressure of the internal carotid and vertebral artery has also been observed, following prolonged intra-arterial injection of contrast medium [37]. In horses, an elevated mean arterial blood pressure or heart rate were seen in 5 % of the cases after intra-arterial iodinated contrast medium administration without requiring intervention [38]. Intra-arterial pressure was not monitored quantitatively in any of the horses included in the study. Although volatile agent-induced hypotension is well known and a concern during inhalation anaesthesia in horses [39]. Our main consideration not to monitor intra-arterial pressure was to keep the total anaesthesia time as short as possible.
A second concern is the contrast medium induced anaphylactic reactions, reported for several species including horses [38, 40–42]. Mild reactions as an elevated heart rate, changes in blood pressure, urticarial and oedema are the most often seen symptoms. Excluding these reactions in the study population solely based on the clinical records is difficult, as for these conditions no treatment or intervention is considered necessary [38, 42] and blood pressure is not standardly recorded in our institutes during CT studies. Moderate and severe anaphylactic reactions require treatment or intervention [38, 42]. No such reactions have been described in the anaesthesia and clinical records of the included cases.