Materials and Methods

Limb specimens—Thirty-three cadavers of adult medium- to large-breed dogs without forelimb disease were used in the study. Dogs were euthanatized by use of ketamine hydrochloride and xylazine hydrochloride administered IM followed by IV administration of pentobarbital sodium for reasons unrelated to forelimb disease. Both forelimbs were removed from each cadaver. For some cadavers, the limbs were removed within 4 hours after euthanasia. Other cadavers were frozen for variable periods immediately after euthanasia; limbs of these cadavers were removed within 6 hours after thawing. All forelimbs were severed at the elbow joint. Immediately after harvesting, all limbs were frozen (or refrozen) for variable periods before use. Limbs were allowed to thaw in a walk-in cooler for approximately 16 hours before use in the study. Limbs were thawed in 7 groups: 1 group (6 limbs from 3 dogs) was used in a pilot investigation to determine probable locations of soft tissue spaces and myofascial compartments, and 6 groups (each comprised of 10 limbs from 5 dogs) were used in the main investigation.

Contrast medium mixture—Contrast material was mixed immediately prior to the first injections of limbs in each group. The contrast medium mixture was produced by diluting 5 mL of iopromide^a and 2 mL of indigo blue acrylic ink^b in 45 mL of saline (0.9% NaCl) solution. As each group of limbs was processed, the mixture was shaken approximately every 5 minutes to keep the blue pigment in suspension.

Experimental procedures—To determine the probable locations of soft tissue spaces and myofascial compartments (based on locations of such anatomic features in humans), the manus of each of 6 limbs (obtained from 3 dogs) were used in a pilot investigation. For the purposes of the study, soft tissue spaces were defined as tissue regions bounded by bone or fascial planes that do not incorporate muscles^13,15 and myofascial compartments were defined as spaces that contain a muscle or muscles and are enclosed by bone or fascia.^4,11 Dissection to individual muscles or spaces was performed to guide injection of the contrast medium mixture into the specimens under direct observation. On the basis of results of the pilot investigation, 21 predicted soft tissue spaces and myofascial compartments were identified as injection sites for evaluation in the main investigation. These spaces and compartments were as follows: digital pads II through V, metacarpal pad (medial, central, and lateral lobes considered independently), carpal pad, midpalmar space, flexor space, extensor space, palmar subcutaneous tissues, dorsal subcutaneous tissues, interosseous muscles II through V, special muscles of digit I (abductor pollicis brevis, flexor pollicis brevis, and adductor pollicis muscles considered collectively), adductor digiti II muscle, adductor digiti V muscle, and abductor digiti V muscle.

In the main investigation, each group of 10 thawed limbs underwent CT examination simultaneously by use of a single-slice helical CT scanner.^c Three forelimbs were placed side by side with the palmar surface on the CT table and such that the distal ends of the digits entered the scanner first. Cardboard platforms were used to support 2 upper rows of 3 or 4 limbs, and the limbs in those rows were positioned in the same manner as the limbs in the lowermost row. The specific limb placed in each position was determined randomly by use of a drawing of numbers. Transverse CT images without contrast medium (precontrast images) of the 10 forelimbs in each group were obtained from the distal ends of the digits through the antebrachiocarpal joint with axial acquisitions of 1.5-mm slice thickness and 1.0-mm slice interval by use of a modified extremities protocol (200 mA, 120 kVp, 240-mm field of view, 512 × 512 matrix, and standard reconstruction algorithm). Precontrast images were obtained to confirm that there were no abnormalities in any of the limbs.

After the initial scanning, 1 predicted compartment or space in each manus was injected with contrast medium mixture directly with a 22-gauge needle via either a percutaneous approach or after dissection to the region of interest. A percutaneous approach was used for the more superficial spaces, such as the pads and subcutaneous regions, in which needle placement was certain. However, when precise localization of the needle tip needed to be verified (as for the muscle bellies and deep palmar spaces), dissection was performed to allow observation of the needle during placement. During dissection, care was taken to incise only the minimal amount of tissue required to see the anatomic part of interest and the structure itself was not cut. For injection of predicted myofascial compartments, the center of the muscle belly was injected. The volume of contrast medium mixture injected in each predicted space or compartment was based on the volume required to distend that structure in the pilot investigation. In the limbs examined in the main investigation, injection of that predetermined volume resulted in notable distension of each space or compartment or a considerable increase in resistance to injection. The injected volume of contrast medium mixture ranged from 0.5 mL in the digital pads to 3 mL in the dorsal and palmar subcutaneous spaces. The interosseous compartments, flexor space, and midpalmar space each received 1 mL of contrast medium mixture. Each of the 21 predicted locations of a soft tissue space or myofascial compartment was injected in at least 2 limbs during the study, although not every site was injected during processing of each group of limbs.

Following injection of the contrast medium mixture, the limbs were returned to their previous positions on the CT couch and CT scanning of the limbs was performed (postcontrast images). The scanning protocol was similar to the precontrast CT protocol, except that a 0.5-mm slice interval was used and the scan did not include the carpus if contrast medium mixture was not observed in the proximal portion of the metacarpus.

In the main investigation, transverse CT images of the 60 limbs were reviewed by 2 board-certified veterinary radiologists (CPO and JCJ) by use of a remote diagnostic workstation and commercially available viewing software^d (digital imaging and communications in medicine [DICOM] standard). For each manus, presence of a compartment or space was based on a consensus opinion formed by use of the following criteria: sharp margination of the area of contrast medium mixture accumulation, homogeneous appearance of the contrast medium mixture, and confinement of the contrast medium mixture to a discrete region. Each injection site was graded by consensus on a scale of 0 to 4 as follows: 0 = heterogeneous appearance of contrast medium mixture with no definable margins and diffusion into surrounding soft tissues; 1 = heterogeneous appearance of contrast medium mixture with poor confinement and few well-defined margins; 2 = heterogeneous appearance of contrast medium mixture with moderate confinement and multiple poorly defined margins; 3 = fairly homogeneous appearance of contrast medium mixture, which is confined except for occasional ill-defined margins; and 4 = homogeneous appearance of contrast medium mixture, which is confined with sharp margins. Soft tissue spaces or myofascial compartments were considered to be present if the structure received a score of 3 or 4 in at least 2 of the examined manus. After grading of all spaces was completed, 3-D CT reconstructions of some manus were created by use of open-source viewing software^e to further assist in evaluation of the extent of injected spaces, but these images were not used in the grading procedures.

Immediately following postcontrast CT scanning, all forelimbs were again frozen. After the specimens were frozen, some were removed from the freezer for transverse sectioning by use of a band saw. The other forelimbs were later thawed and dissected conventionally. Selection of paws for transverse sectioning and dissection was random; each injection site had been used in at least 1 specimen that underwent transverse sectioning or dissection. The distribution of the contrast medium mixture was correlated with CT imaging findings, and the dissected limbs were used to identify the structures in each space or compartment.

Results

No abnormalities were identified in any specimen by use of the precontrast CT imaging. Of the 21 soft tissue spaces and myofascial compartments predicted on the basis of the pilot investigation findings, 18 were identified in limbs examined in the main investigation. Of the 18 spaces and compartments, 14 were assigned scores of 3, 2 were assigned scores of 3 and 4, and 2 were assigned scores of 4. The special muscles of digit I, adductor digiti II muscle, and adductor digiti V muscle did not meet the imaging criteria for presence of a myofascial compartment. Direct visual evaluation of the dissected and transversely sectioned specimens correlated well with the transverse and 3-D imaging findings; the contrast medium mixture was retained in the structures that were identified as compartments or spaces and caused distension of those structures. There was no distension of the 3 muscles that were not compartments, and the contrast medium mixture diffused throughout the surrounding soft tissues after injection of those structures. The contents of the identified soft tissue spaces and myofascial compartments were recorded (Table 1).

Soft tissue spaces of the pads—All injections into the digital pads remained within the respective digits, indicating the presence of 4 individual digital spaces. However, although occasional injections were confined to the pad stroma, most spread through the soft tissues of the digit and wrapped around the digital flexor tendons; following 2 of the 10 injections, the contrast medium mixture extended into the dorsal tissues of the digit (Figure 1). The contrast medium mixture did not extend beyond the midportion of the proximal phalanx.

View Full Size

Representative transverse CT images illustrating digital spaces in the manus of a dog cadaver obtained following injection of contrast medium mixture (prepared by dilution of 5 mL of iopromide and 2 mL of indigo blue acrylic ink in 45 mL of saline [0.9% NaCl] solution) into the second digital pad. Medial is to the left of the image in both panels. A—Image at the level of the distal phalanx of digit II. Notice that the contrast medium mixture is mostly contained by the stroma of the digital pad (arrow). B—Image at the level of the second proximal interphalangeal joint. In this image, the contrast medium mixture extends into the palmar and dorsal soft tissues of the digit (arrowheads).

Citation: American Journal of Veterinary Research 71, 2; 10.2460/ajvr.71.2.138

• Download Figure
• Download figure as PowerPoint slide

Following injection of the medial and lateral lobes of the metacarpal pad, the contrast medium mixture was prevented from extending into the central lobe by septae. The contrast medium mixture extended dorsally between the respective digits (II and III or IV and V) at the level of the metacarpophalangeal joints and was present in the palmar tissues adjacent to the manica flexoria of the flexor tendons and the palmar annular ligaments (Figure 2). A similar pattern of distribution was evident following injection of the central lobe of the metacarpal pad; the stromal component of the contrast medium mixture was located in the central lobe, and the remainder extended dorsally to a point between the metacarpophalangeal joints and proximal phalanges of the third and fourth digits. The contrast medium mixture extended a small distance proximally and distally from the level of the pad following injections of each of the 3 lobes of the metacarpal pad. The proximal margin of the central metacarpal pad space was the midpalmar space. Following injection of the carpal pad, all of the contrast medium mixture remained within the pad stroma.

View Full Size

Representative transverse CT images illustrating the metacarpal pad spaces in the manus of a dog cadaver obtained following injection of contrast medium mixture into lobes of the metacarpal pad. These spaces extend dorsally between the metacarpophalangeal joints and proximal phalanges. Medial is to the left of the image in all panels. A—Image at the level of the third and fourth metacarpophalangeal joints. After injection into the lateral lobe, contrast medium mixture is located within the lateral metacarpal pad space (arrow). Notice the septum (arrowhead) that separates the medial and central lobes of the metacarpal pad. B—Image obtained at a level more proximal than that in panel A. Contrast medium mixture is located within the lateral metacarpal pad space. C—Image at the level of the third and fourth metacarpophalangeal joints. After injection into the central lobe, contrast medium mixture is located within the central metacarpal pad space (arrow). D—Image obtained at a level more proximal than that in panel C. Contrast medium mixture is located within the central metacarpal pad space.

Citation: American Journal of Veterinary Research 71, 2; 10.2460/ajvr.71.2.138

• Download Figure
• Download figure as PowerPoint slide

Soft tissue spaces of the palmar tissues—Injection into the midpalmar space resulted in distribution of contrast medium mixture in a pattern of multiple spokes extending primarily around interosseous muscles III and IV and adductor digiti II and V, with a small portion of the space extending adjacent to interosseous muscles II and V (Figure 3). The midpalmar space was on the midline of the manus immediately deep to the deep digital flexor tendons and immediately proximal to the central metacarpal pad space. Although most of the midpalmar space was proximal to the metacarpal pad, the most distal spoke-like extension of the space continued to the level of the metacarpophalangeal joints of the third and fourth digits; it also extended dorsally between metacarpal bones III and IV at the level of and just proximal to the metacarpophalangeal joints. Proximally, the midpalmar space extended to the level of the proximal metacarpal bones (metacarpal bases).

View Full Size

Computed tomographic images illustrating the midpalmar space in the manus of a dog cadaver obtained following injection of contrast medium mixture. Medial is to the right of the image in panels A through C. A—Transverse image at the level of the distal portions of the third and fourth metacarpal bones. The distal extent of the midpalmar space (arrow) is at approximately the same level as the proximal extent of the central metacarpal pad space. B—Transverse image obtained at a level more proximal than that in panel A, at the level of the distal portions of the second and fifth metacarpal bones. Notice that the midpalmar space extends around the interosseous muscles (asterisk). C—Transverse image obtained at a level more proximal than that in panel B, at the level of the base of the first distal phalanx. Notice that multiple spoke-like extensions of the midpalmar space are evident at this level. The adductor digiti V muscle and palmar margin of the fifth interosseous muscle are denoted by the + symbol. D—Three-dimensional reconstructed image of the manus. The manus is viewed from a palmaroproximomedial direction. The midpalmar space is colored orange in this reconstruction.

Citation: American Journal of Veterinary Research 71, 2; 10.2460/ajvr.71.2.138

• Download Figure
• Download figure as PowerPoint slide

The flexor space contained the tendons of the superficial and deep digital flexor muscles and was separated from the midpalmar space by a thick fascial plane (Figure 4). Although the space was predominantly located on the midline, in 2 of the 4 injected specimens, it also had a lateral recess in the proximal half of the metacarpus through which the tendon of abductor digiti V passed. The flexor space extended proximally to the level of the metacarpal bases, although only the deep digital flexor tendon was incorporated at the most proximal extent. The flexor space extended distally to the level of the distal third of the diaphysis of metacarpal bones III and IV; this corresponded to the level of the second and fifth metacarpophalangeal joints, although the space did not extend as abaxially as the second or fifth metacarpal bones in this distal location.

View Full Size

Transverse CT images and photograph of an anatomic section illustrating the flexor space in the manus of a dog cadaver obtained following injection of contrast medium mixture. Medial is to the left of the image in all panels. A—Computed tomographic image. Notice that contrast medium mixture is located within the flexor space (arrow) in the midmetacarpus region. B—Computed tomographic image obtained at a level more proximal than that in panel A. The contrast medium mixture filling defects in the flexor space are caused by the deep digital flexor tendons (black arrowhead) and superficial digital flexor tendons (white arrowhead). Notice also the defect in the palmarolateral skin attributable to the incision made prior to injection. C—Computed tomographic image obtained at a level more proximal than that in panel B (ie, at the level of the metacarpal bases). Notice that the flexor space surrounds only the deep digital flexor tendon at this level. D—Photograph of an anatomic section of the manus of another dog; the section represents approximately the same level as that in the image in panel C. The lateral recess surrounds the abductor digiti V muscle (arrow) at the myotendinous junction.

Citation: American Journal of Veterinary Research 71, 2; 10.2460/ajvr.71.2.138

• Download Figure
• Download figure as PowerPoint slide

The palmar subcutaneous space was located in the most superficial palmar tissues of the manus and extended from the carpal pad to the distal portions (heads) of metacarpal bones III and IV (Figure 5). On the midline of the manus, its dorsal boundary was a fascial plane that separated it from the flexor space and superficial digital flexor tendons. Medially and laterally, the palmar subcutaneous space extended dorsally near the palmarolateral cortex of metacarpal bone V and the palmar cortex of metacarpal bone I, but it did not extend to the dorsum of the manus. At this location, the skin of the manus was closely associated with the metacarpal bones and there was minimal subcutaneous connective tissue, although no discrete structure that prevented the extension of the contrast medium mixture dorsally was identified.

View Full Size

Transverse and sagittal CT images illustrating the palmar subcutaneous space in the manus of a dog cadaver obtained following injection of contrast medium mixture. A—Transverse CT image. Medial is to the left of the panel. The palmar subcutaneous space extends nearly to the cortices of the first digit and fifth metacarpal bone. B—Sagittal CT image. Dorsal is to the left of the panel. The sharply marginated space extends between the carpal pad (asterisk) and the metacarpal pad (+).

Citation: American Journal of Veterinary Research 71, 2; 10.2460/ajvr.71.2.138

• Download Figure
• Download figure as PowerPoint slide

Soft tissue spaces of the dorsal tissues—The extensor space was located on the dorsum of the manus and contained the common digital extensor and lateral digital extensor tendons (Figure 6). However, there was minimal extension of the contrast medium mixture around the dorsal aspect of these tendons because the dorsal surface of the tendons was intimately associated with the fascial plane that separated the extensor space from the dorsal subcutaneous space. The extensor space extended from the bases of the metacarpal bones proximally to the second through fifth metacarpophalangeal joints distally. At the distal margin of the space, the extensor space extended around the abaxial cortices of the metacarpal bones.

View Full Size

Transverse CT images illustrating the extensor space in the manus of a dog cadaver obtained following injection of contrast medium mixture. Medial is to the right of the image in both panels. A—The distal extent of the extensor space (arrow) is visible in this image obtained at the level of the distal portion of metacarpal bone II. Notice the extension of the space around the abaxial cortices of metacarpal bones III and IV. B—In the midmetacarpal region, the contrast medium mixture in the extensor space is in direct contact with the dorsal cortices of metacarpal bones II through V.

Citation: American Journal of Veterinary Research 71, 2; 10.2460/ajvr.71.2.138

• Download Figure
• Download figure as PowerPoint slide

The dorsal subcutaneous space was located in the most superficial dorsal tissues of the manus, extending from the middle of the metacarpus to the metacarpophalangeal joints or proximal phalanges of the second through fifth digits. Its palmar (deep) margin was the fascial plane associated with the dorsal surface of the extensor tendons. The dorsal subcutaneous space extended medially and laterally to the dorsomedial and dorsal margins of metacarpal bones II and V, respectively; at this location, the dense fascia associated with the extensor space appeared tightly bound to the skin, likely preventing extension of contrast medium mixture farther in a palmar or abaxial direction (Figure 7).

View Full Size

Transverse CT image illustrating the dorsal subcutaneous space at the level of the distal third of the metacarpus in the manus of a dog cadaver obtained following injection of contrast medium mixture. Medial is to the right of the image. Notice the unfilled space between the contrast medium mixture and the metacarpal bones, which represents the extensor space.

Citation: American Journal of Veterinary Research 71, 2; 10.2460/ajvr.71.2.138

• Download Figure
• Download figure as PowerPoint slide

Myofascial compartments—Each of the 4 interosseous compartments (interosseous II through V) contained only the corresponding interosseous muscle (Figure 8). However, although distension of the muscle was evident following injection, filling with contrast medium mixture was incomplete in multiple instances; in some limbs, the filling was heterogeneous or there was the appearance of subcompartmentalization of interosseous muscles II and V. It should also be noted that although the interosseous muscles are almost entirely located palmar to the corresponding metacarpal bones, a small triangular extension of the interosseous muscles extended between the metacarpal bones.

View Full Size

Transverse CT images illustrating interosseous compartments in the manus of dog cadavers obtained following injection of contrast medium mixture. Medial is to the left of the image in both panels. A—After injection into interosseous muscle III, contrast medium mixture is located within the third interosseous compartment. Notice the relatively homogeneous filling and sharp margins of the injected material. B—After injection into interosseous muscle II, contrast medium mixture is located within the second interosseous compartment. Although the filling of the injected material is homogeneous and its margins are rather sharp, the filled area represents only approximately half that expected for this muscle. The unfilled portion of the muscle is indicated by an asterisk, although the precise boundaries cannot be seen because of the contrast resolution constraints of CT.

Citation: American Journal of Veterinary Research 71, 2; 10.2460/ajvr.71.2.138

• Download Figure
• Download figure as PowerPoint slide

The abductor digiti V compartment primarily enclosed the muscle belly of abductor digiti V and, accordingly, extended from the accessory carpal bone to the level of the metacarpal bases. A small amount of tissue immediately surrounding the muscle was also included in this compartment, and medially, the caudal (palmar) interosseous artery and ulnar nerve passed through the compartment (Figure 9).

View Full Size

Transverse CT image illustrating the abductor digiti V compartment (arrow) at the level of the proximal carpal row in the manus of a dog cadaver obtained following injection of contrast medium mixture. Medial is to the right of the image. The heterogeneous filling defect in the contrast medium mixture represents the muscle belly that is partially filled with injected material.

Citation: American Journal of Veterinary Research 71, 2; 10.2460/ajvr.71.2.138

• Download Figure
• Download figure as PowerPoint slide

Discussion

Results of the present study indicated that there are at least 18 individual soft tissue spaces and myofascial compartments in the manus of dogs. Despite the use of relatively large volumes of contrast medium for the small spaces studied, distention of spaces without leakage of contrast medium was the norm. This finding indicates that, at least in the acute stages, disease processes such as infection may remain relatively confined to the initial site of injury. For example, injection of contrast medium mixture in the digital pads resulted in retention of material within the soft tissues of the digit, and injection into the carpal pad resulted in retention of material solely within the pad stroma.

The results of injection of contrast medium mixture into the metacarpal pad and midpalmar space were perhaps more interesting with regard to future clinical applications in mapping paths of disease spread. Injection of contrast medium mixture into each lobe of the metacarpal pad revealed that dense connective tissue prevented extension of the material into the other lobes of the pad. However, the contrast medium mixture extended dorsally from each of the 3 lobes and terminated between the associated metacarpophalangeal joints. Similarly, the midpalmar space was located predominately within the palmar tissues in the center of the manus, but 1 spoke-like projection of the space extended dorsally and terminated between the distal aspect of the third and fourth metacarpal bones. Thus, it is plausible that disease in any of these spaces could be directed to the dorsum of the manus, which could easily lead to incomplete resection of diseased tissue during a surgical procedure. In contrast, the subcutaneous spaces had more regular shapes. A large volume of the contrast medium mixture was injected into the dorsal and palmar subcutaneous spaces, and the material remained confined yet caused marked distention of the tissues, indicating a large capacitance of these tissues. Additionally, communication between the dorsal and palmar subcutaneous tissues was not identified.

The flexor space in the canine manus appears to be analogous to the ulnar bursa in the human hand, given that it extended around the flexor tendons.²⁰ However, the ulnar bursa in humans is a serous sheath that generally continues distally as the flexor tendon sheath for the fifth digit,⁷ whereas this space in dogs appeared to be composed only of connective tissue and was bounded by fascial planes. No serous structure in the palmar metacarpal region in dogs has been described,²¹ and the flexor space did not communicate with the flexor tendon sheaths of the digits in the present study. An interesting finding in the flexor space in 2 of the 4 injected specimens in our study was the presence of a lateral recess that encompassed the proximal aspect of the abductor digiti V tendon. This likely represents an anatomic variant; variation in the precise morphology of soft tissue spaces and myofascial compartments in human hands has been reported.^5,7,15,22 Given the small number of dogs' limbs in which this space was injected, the prevalence of the lateral recess cannot be estimated.

The cause of the variable appearance of the digital spaces is unknown. Following most of the digital pad injections in the present study, the contrast medium mixture spread throughout the soft tissues of the digit. Occasionally, however, the injection was associated with a very high backpressure and the injected material remained within the stroma of the digital pad. In all injected limbs, the tip of the needle was placed just deep to the keratinized dermal tissue of the pad. In the instances in which contrast medium mixture was retained in the pads themselves, it is likely that the needle was not fully extended through the dermis into the pad stroma, thereby making injection difficult and preventing spread of the contrast medium mixture. When the needle was fully within the pad stroma, the contrast medium mixture could easily spread throughout the digital soft tissues. However, an alternate theory is that each digital pad is a discrete space separate from the digital soft tissue space and that the needle was positioned deep to the pad in those injections where the contrast medium mixture spread more freely. Nonetheless, this alternate theory is not a likely explanation because all needle tips remained superficial, and the stromal area of the pad filled in all cases, which would not be expected if the 2 compartments were separate and injections were made into the deeper tissues.

The limited proximal extent of the dorsal subcutaneous space was surprising, and a reason for the limited extent despite the relatively large volume of contrast medium mixture administered was not identified. It is possible that the subcutaneous tissues in the region of the dorsal portion of the metacarpus are diminished in volume or that the skin is more tightly adhered to the deeper tissues, thus acting as a barrier to spread. It is also possible that administration of an even larger volume of contrast medium mixture would have allowed more proximal spread because spread was inhibited by the relatively large capacitance of the more distal tissues into which the contrast medium mixture was injected.

The heterogeneous and incomplete filling of some interosseous muscles with contrast medium mixture in the present study was understandable because the injected material had to dissect between muscle fibers. However, in 1 instance each, only the more abaxial portion of interosseous muscles II and V were filled with contrast medium mixture. The margination in both of these specimens was sharp, suggesting the existence of a subcompartment within the interosseous muscle. This was not evident in all injected specimens and could represent an anatomic variant or may indicate the presence of a weaker boundary that was not sufficient to withstand the pressure of injection in some limbs. Further evaluation would be required to determine the commonality and importance of this finding.

Although we chose to refer to the flexor space as a soft tissue space, one could argue for its inclusion in the list of myofascial compartments because it contains the lumbrical, interflexorius, and flexor digitorum brevis muscles. Its categorization as a soft tissue space was selected for 2 reasons. First, the midpalmar space in human hands contain the lumbrical muscles, and numerous publications include reference to this region as a space,^{7,13,15,20,22–24} whereas it is classified as a compartment in only 1 report,⁴ to our knowledge. Second, an underlying premise for the recognition of a myofascial compartment is that it could be a location of compartment syndrome, which is induced by increased intramuscular pressure (often a result of muscular swelling). Because the muscle bellies in the flexor space represent only a miniscule percentage of the overall volume of the space, their contribution to compartment syndrome is doubtful.

The adductor digiti II and V muscles were not considered myofascial compartments because the contrast medium mixture leaked out of all sides of the muscle bellies into the surrounding tissues, even with a small injection volume, and no distention of the muscles was visible following injection. Upon close examination, there was minimal to no fascial covering of these muscles to prevent diffusion of the contrast medium mixture. Similar to the adductor digiti muscles, the special muscles of digit I were not considered a compartment on the basis of CT or anatomic findings. The fascia surrounding these muscles is also very thin, and the contrast medium mixture spread throughout the surrounding tissues. Leakage of contrast medium mixture from the needle tract could have confounded interpretation and resulted in classification of a true space or compartment as not a true space or compartment. However, leakage from the needle tract appeared as a focal linear accumulation of contrast medium mixture in several CT images (generally resulting in assignment of a score of 3 for the space or compartment) and appeared considerably different from the diffuse spread of contrast medium mixture associated with the adductor digiti II and V muscles and special muscles of digit I.

Because of the strict definition of a score of 4 on the 5-point scale of contrast medium confinement (homogeneous and confined [with sharp margins] distribution of contrast medium mixture) applied in the present study, few of the spaces and compartments were assigned a score of 4. Only the flexor space, dorsal subcutaneous space, palmar subcutaneous space, and interosseous muscle III would have been considered soft tissue spaces or myofascial compartments if at least 1 score of 4 was required for classification; only the dorsal and palmar subcutaneous spaces received a score of 4 on all CT evaluations. Scores of 3 and 4 were used to classify structures as spaces or compartments in our study. A score of 3 was selected to avoid misclassification of structures in which there was slight leakage of the contrast medium mixture out of the injection site or slightly ill-defined margins caused by partial volume averaging (both of which are imaging artifacts). The use of a 5-point scale was analogous to the common use of 5-point scales for scoring the certainty of a lesion in imaging studies; in such studies,^25–27 the 2 highest scores are designated as degrees of a positive finding, the middle score is considered equivocal, and the lowest 2 scores are designated as negative findings.

One limitation of the present study was the use of limbs from medium- to large-breed dogs. Limbs from small-breed dogs were excluded because of the small size of the soft tissue spaces and myofascial compartments and the difficulties in injecting those small structures without additional damage to the surrounding tissues. However, similar results would be expected in dogs of any size. Additionally, if the present study were repeated to examine the soft tissue spaces and myofascial compartments of the pes, similar results also would be expected because the anatomic features of the metatarsus are quite similar to those of the metacarpus in dogs.²¹ Indeed, such a study of the pes would be useful because injuries to the manus and the pes are both common. However, because of limited resources and the recognition of differences in compartments of hands and feet of humans,^5,7,10,28 we chose to focus on only the manus of dogs in the present study.

We acknowledge that a study identifying soft tissue spaces and myofascial compartments in limbs of cadaveric dogs cannot reveal the actual clinical importance of these structures in living dogs. Although the contrast medium mixture remained confined under the conditions of our experiment, processes such as inflammation could cause degradation of fascial barriers and potentially allow further spread of disease. Moreover, although development of tissue distention following injection of contrast medium mixture indicated increased pressure within the tissues, we cannot be certain whether that pressure mimics the amount of pressure that may develop under pathological conditions. Higher pressures may cause failure of some of the fascial boundaries that remained intact during our study. Pressures within the soft tissue spaces and myofascial compartments were not measured during our study because of the small size of the spaces; accurate pressure readings would require insertion of a second needle into the space or compartment, which was not possible without causing undue trauma to the structure being studied. Nevertheless, results of the present study indicated that there are at least 18 discrete soft tissue spaces and compartments in the canine manus, and those structures may influence the spread of disease in the distal portion of the forelimb in dogs. However, further investigation to evaluate their importance in clinical disease is required.