Wilbrand Knee Revisited

Supplemental Digital Content is Available in the Text.

Abstract

Background:

Over a century ago, German ophthalmologist Hermann Wilbrand reported inferonasal crossing fibers within the chiasm curve anteriorly into the contralateral optic nerve. This anatomic bend, “Wilbrand knee,” is classically cited as the explanation for the “junctional scotoma,” a contralateral superotemporal visual field defect associated with lesions affecting the optic nerve at its junction with the chiasm. More recent reports have called into question the existence of Wilbrand knee or suggested that it may simply be an artifact.

Methods:

Four human optic chiasms (obtained from cadaver donors with no reported premortem visual pathology) and 2 monkey chiasms were fixed and thin sectioned (40 µm), then examined using anisotropic scattering imaging, a novel technique that takes advantage of the fact that light reflects off well-defined linear structures (i.e., axonal tracts) in a predictable manner based on their orientation. Using this technique, tissue structures oriented in different directions can be distinguished at high resolution without the need for tissue staining.

Results:

In all 4 human optic chiasms, thin fiber tracts consistent with, but less prominent than, those Wilbrand had described were observed. No such tracts were found in the monkey chiasms.

Conclusions:

Wilbrand knee exists in humans but is modest in its anterior projection. Wilbrand knee does not seem to be present in monkeys, however, which may explain conflicting reports in the literature regarding its existence.

In the late 19th century, German ophthalmologist Hermann Wilbrand studied 2 human optic chiasms after unilateral enucleation using postmortem myelin staining and reported that inferonasal crossing fibers within the chiasm curved anteriorly into the contralateral optic nerve.^1,2 This anatomic bend (now known as “Wilbrand knee”) is classically cited as the explanation for the contralateral superotemporal visual field defect that may develop when a lesion affects the optic nerve at its junction with the chiasm (the so-called “junctional scotoma”).

More recent reports, however, have called into question the existence of Wilbrand knee or suggested that it may simply be an artifact of monocular enucleation.^3,4 Horton studied 8 rhesus monkeys and 3 squirrel monkeys using [³H] proline labeling, found no evidence of Wilbrand knee in monkeys, and killed immediately after radiolabeling.³ In monkeys who had been enucleated, however, Horton found that crossing fibers within the chiasm were pulled forward into the atrophic optic nerve. This distortion became more prominent with the passage of time as optic nerve atrophy increased. Horton also studied 3 human chiasms using myelin staining, months to years after unilateral enucleation, and found that crossing fibers entered the contralateral optic nerve to a modest degree. Assuming anatomic homology between monkeys and humans, Horton concluded that Wilbrand knee does not exist in humans in the absence of enucleation.³

Subsequently, Lee, et al reported 3 patients with optic nerve pathology (2 patients with optic nerve sheath meningiomas and one patient with sarcoid) who underwent resection of the optic nerve at the optic nerve–optic chiasm junction.⁴ All 3 patients had normal postoperative visual fields, with no junctional scotoma in the contralateral eye, implying that all crossing fibers had been spared.

Both Horton and Lee, et al concluded that the anterior projection of the crossing fibers described by Wilbrand was most likely an artifact related to postenucleation atrophy of the optic nerve,^3,4 although neither study was able to assess the anatomy of the human optic chiasm in a nonpathological state.

Although the localizing value of the junctional scotoma persists regardless of the debate over the existence of Wilbrand knee, the question remains an interesting one. We describe the use of a novel tool, anisotropic scattering imaging (ASI; see Supplemental Digital Content, Figure, http://links.lww.com/WNO/A792), to look for the presence or absence of Wilbrand knee in normal human and monkey optic chiasms to reconcile the discrepancy between Wilbrand findings and those of Horton and Lee, et al.

METHODS

Anisotropic scattering is an optical phenomenon used in this study to dissect specific fiber tracts traveling among other fiber tracts. Anisotropic scattering refers to direction-specific reflection of light by certain tissues, as opposed to isotropic (i.e., diffuse) scattering by tissue such as cerebral gray matter and homogeneous substances such as milk. Myelinated axons can be believed as long thin reflective cylinders that are highly organized in orientation. Using a bundle of thin optical fibers to model axons differences in light reflection and scattering is illustrated for 2 possible angles of illumination (Fig. 1A). When these fiber bundles are systematically illuminated from up to 72 different angles, the angle-dependent light reflections can be plotted, with different fiber bundle orientations associated with unique patterns (Fig. 1B).

FIG. 1.

Anisotropic scattering imaging. Fiber bundles scatter light differently, based on different angles of illumination (A). These angle-dependent light reflections can be mapped, with different orientations associated with unique patterns (B). A thin unstained section of human brain imaged sequentially while being illuminated from 3 different angles (C).

White matter in the human brain is highly anisotropic. A thin unstained coronal section of the human brain at the level of the red nucleus and subthalamic nucleus was imaged sequentially while being illuminated from 3 different angles (Fig. 1C). Note that the appearance of the fiber tracts changes although the camera and the tissue remained stationary, a consequence of different reflections under the 3 different angles of illumination. When the 3 black and white images are pseudo-colored in red, green, and blue, then combined, individual fiber tracts are more easily appreciated (Fig. 1C). The latter procedure produces a “qualitative” display that was used to reveal Wilbrand knee for the first time in the absence of unilateral enucleation.⁵

The qualitative approach can be modified to extract and display “quantitative” 3D information about fiber orientation. In brief, 72 separate directions of illumination—12 azimuth directions for each of 6 inclinations—can be used to create 72 images of the same tissue. A “fingerprint” of the directional pattern of light scattering is obtained for each pixel of the image. The observed pattern of light scattering from the tissue is compared with calibration patterns obtained from reference fiber bundles with predefined orientations. For quantitative imaging, a computer algorithm was used to pattern match each pixel of the sample image to a reference calibration pattern.

An example of a color-coded fiber orientation display in spherical coordinates, with the azimuth and pitch displayed separately, from a single horizontal section from the dorsal aspect of a monkey optic chiasm is shown in Figure 2. Note that this image appears to confirm the interdigitated organization of fiber tracts first described by Hoyt and Luis 60 years ago.⁶

FIG. 2.

Monkey optic chiasm, color-coded in spherical coordinates, displayed by azimuth and pitch.

To perform these studies, primate tissue was obtained from rhesus macaque monkeys and human tissue was obtained from the Maryland Department of Health State Anatomy Board. We requested that the human cadaveric tissue be obtained from donors with no known premortem visual pathology.

RESULTS

Using the qualitative method described above, we had previously shown that a tract consistent with Wilbrand knee can be observed (Fig. 3A) in axial sections from the human optic chiasm.⁵ We have now confirmed this earlier finding in 3 additional human optic chiasms using the quantitative method to track the orientation of their fiber tracts:

FIG. 3.

Qualitative and quantitative anisotropic scattering imaging applied to a human optic chiasm sectioned in the axial plane. Wilbrand knee visualized qualitatively (A) and quantitatively (white arrows), color-coded by azimuth (B).

If Wilbrand knee exists, we would expect its fibers to be located in front of the anterior notch of the chiasm, traveling in a path perpendicular to those of optic nerve fibers. In the axial plane, axons of Wilbrand knee should differ from optic nerve fibers in “azimuth”. In the coronal plane, axons of Wilbrand knee should differ from those of the optic nerve in “pitch”. In all human chiasms studied, a perpendicular tract was observed anterior to the optic nerve–chiasm junction in ventral (but not central or dorsal) sections, consistent with Wilbrand knee, although this tract does not project anteriorly as dramatically as Wilbrand depicted.

A ventral axial section from a human chiasm is shown in Figure 3B, in which the “azimuth” of the axons is color-coded to reveal the path of Wilbrand knee. The green regions identify the 135° (”2 o'clock/8 o'clock”) orientation of the axons comprising the right optic nerve. The fibers within the red arc are traveling perpendicular to the optic nerve fibers at approximately 45°.

The same procedures were applied to the optic chiasm from a rhesus monkey. Horizontal sections, 30 µm thick, were imaged using both qualitative and quantitative approaches. An example from a ventral section comparable with the level where Wilbrand knee was observed in human chiasms is shown in Figure 4. No crossing fibers traveling perpendicular to fibers from the optic nerve were seen in front of the anterior notch of the chiasm.

FIG. 4.

Qualitative and quantitative anisotropic scattering imaging applied to a monkey optic chiasm sectioned in the axial plane. No evidence of Wilbrand knee visualized qualitatively (left) or quantitatively (right).

An additional monkey optic chiasm and human optic chiasm were sectioned in the coronal plane, then imaged using the quantitative method to assess “pitch.” No perpendicular fibers were seen anterior to the chiasm in the monkey sections (Fig. 5A). By contrast, perpendicular fibers (red) were present anterior to the chiasm in the human sections (Fig. 5B). Note that these perpendicular fibers were located only ventrally, consistent with Wilbrand knee.

FIG. 5.

Quantitative anisotropic scattering imaging applied to monkey and human optic chiasms sectioned in the coronal plane, color-coded by pitch. No evidence of Wilbrand knee in the monkey (A). Evidence of Wilbrand knee in the human (red), located ventrally, anterior to the chiasm (B).

DISCUSSION

Wilbrand knee exists in humans but not in monkeys. Anterior projection of crossing fibers into the optic nerve was found in all 4 human chiasms studied but in neither of the monkey chiasms studied. As expected, when present, Wilbrand knee is limited to the most ventral portion of the optic chiasm, but its anterior projection is more modest than that described by Wilbrand.

Can these observations be reconciled with the work of Horton and Lee, et al? Our findings confirm the original observation by Horton that Wilbrand knee is absent in monkeys (at least in rhesus and squirrel monkeys). Our finding that Wilbrand knee is present in the normal human chiasm is not contradicted by any of the findings by Horton in normal and postenucleation monkey chiasms and postenucleation human chiasms. We agree with Horton that postenucleation optic nerve atrophy may distort the path of crossing fibers, as documented in both monkeys and humans. This phenomenon likely explains the exaggerated appearance of Wilbrand knee in his original descriptions. The observation that optic nerve atrophy may “pull forward” crossing fibers, however, does not preclude some anterior projection of crossing fibers in the absence of pre-existing pathology.

We would suggest that, because Wilbrand knee is modest in its anterior projection, it is possible that it might have been spared during the optic nerve resections described by Lee, et al. It is also possible that the optic nerve pathology (tumors or inflammation) that necessitate the surgical procedures may have distorted the architecture of the optic nerve–optic chiasm junction enough to shift Wilbrand knee posteriorly. (Perhaps rather than being “pulled forward” by optic nerve atrophy, Wilbrand knee was “pushed back” by mass effect from the optic nerve lesions.)

Both the qualitative and quantitative methods of ASI are novel hybrid optical-computational imaging techniques that offer the ability to identify anatomical structures without relying on tissue staining or labelling. It is simple in concept yet provides high spatial resolution and contrast. Here, we have applied the techniques to shed some light on a controversial visual pathway model. We believe that ASI will be useful in other contexts as well.

STATEMENT OF AUTHORSHIP

Conception and design: R. K. Shin, C.-M. Tang; Acquisition of data: J. Kachhela, C.-M. Tang; Analysis and interpretation of data: R. K. Shin, C.-M. Tang. Drafting the manuscript: R. K. Shin, C.-M. Tang; Revising the manuscript for intellectual content: R. K. Shin, J. Kachhela, C.-M. Tang. Final approval of the completed manuscript: R. K. Shin, J. Kachhela, C.-M. Tang.