Kallmann’s syndrome

Subjects.

Nine xKS male patients exhibiting mirror movements were studied (mean age, 27 years; range, 19 to 48 years). All nine men had previously been examined electrophysiologically by Mayston et al.,22 corresponding to subjects K4a, K4 through 6, and K8 through 12 in that study. The diagnosis of X-linked disease was confirmed by pedigree analysis (n = 7) and/or by demonstrating a mutation or deletion at the KAL locus (n = 7). For comparison, 18 healthy men (mean age, 29 years; range, 19 to 42 years) and nine male aKS patients (mean age, 28; range, 21 to 49 years) were also studied. Autosomal disease was inferred from the absence of KAL coding exon deletions2,31 and of xKS-specific phenotypic markers. All subjects were right handed as tested by the Edinburgh Handedness Questionnaire.32 Subjects gave written informed consent to participate in the project, which was approved by the local hospital ethics committee.

MRI data acquisition.

Three-dimensional volume acquisition MR images of nine xKS patients and nine normal control subjects were obtained with a 1.0-T Picker HPQ Vista system (Cleveland, OH) using a radiofrequency spoiled volume acquisition, relatively T1-weighted to give good gray/white contrast and anatomic resolution (repetition time [TR], 21 msec; echo time [TE], 6 msec; flip angle, 35 degrees; field of view [FOV], 25 cm; 192 × 256 × 140 slices at a resolution of 1 × 1 × 1.3 mm). MR images of nine aKS patients and nine additional normal control subjects were obtained with a 2-T Siemens VISION system (Munich/Erlangen, Germany), acquiring T1-weighted images (TR, 9.5 msec; TE, 4 msec; flip angle, 12 degrees; FOV, 25 cm; 192 × 256 × 256 slices at a resolution of 1 × 1 × 1.5 mm).

Image processing and spatial normalization.

Image analysis was performed on a SUN SPARC 20 workstation (SUN Microsystems, Europe Inc., Surrey, UK) using ANALYZE33 software (version 5; Mayo Foundation, Rochester, MN) for region of interest and image editing purposes, and SPM-9634,35 software (Wellcome Dept. of Clinical Neurology, London, UK) for spatial normalization into stereotaxic space, gray/white matter segregation, smoothing, and statistical comparisons. SPM-96 uses MATLAB (version 4; Mathworks Inc., Sherborn, MA).

The MR images were interpolated to yield a 1-mm³ voxel size. They were then normalized into stereotaxic space, as defined by Talairach and Tournoux,36 using the Montreal Neurological Institute (MNI) template.37,38 The normalization was performed using the automated procedure of SPM-96 (custom affine transformations only, without nonlinear algorithms). The voxel size was set to 1 × 1 × 1 mm. The normalized images were then segregated into gray matter, white matter, and CSF spaces using the automated procedure as implemented in SPM-96. A smoothing filter of 12 mm was applied to account for interindividual variability in structural anatomy. The smoothed, segmented images can be thought of as images of gray or white matter density.

Statistical analysis.

The processed images were analyzed using the General Linear Model as implemented in SPM-96. The following categoric comparisons of pooled white matter images were carried out: 1) xKS versus normal men, 2) aKS versus normal men, and 3) xKS versus aKS patients. After specifying the appropriate design matrix, the subject effects were estimated at each and every voxel. The resulting set of voxel values for each contrast constitutes a statistical parametric map (SPM) of the t-statistic. The SPM(t) was transformed to the unit normal distribution SPM(z) and thresholded at 2.33 or p < 0.01 uncorrected. The significant foci in the SPM(z) were then characterized in terms of spatial extent and peak height (z-scores). The Talairach coordinates36 are given in millimeters for the maximally significant voxel in each area, where x defines the lateral displacement from the midline (left = negative), y defines the anteroposterior displacement relative to the anterior commissure (posterior = negative), and z defines the vertical position relative to the anteroposterior commissural line (down = negative).

Region-of-interest analysis.

To validate SPM-96 as a method for morphometry we carried out an independent region-of-interest analysis of the corpus callosum surface in the midline. This analysis was performed on the original nonnormalized MR images. Using the semiautomatic region-of-interest function in ANALYZE, we measured the midsagittal surface of the corpus callosum in each individual and related it to the total brain surface area in the parasagittal plane, including midbrain and pons, at the most medial point of the head of the caudate nucleus. The measurements were taken blind; that is, without knowing to which group any individual belonged. Assuming a normal distribution, we performed an independent t-test using SPSS for Windows (SPSS Inc., Chicago, IL) on a personal computer. This compared the logarithms of the nine ratios of corpus callosum versus brain surface in xKS patients with the logarithms of the equivalent ratios in the normal men.

Statistical analysis (SPM).

Table 1 lists the regions with increased white matter density as determined by SPM(z) when comparing pooled white matter data for each of the three group comparisons.

xKS (with mirror movements) versus normal men.

The most extensive increases in white matter density in the xKS group were found in the middle and anterior part of the corpus callosum (figure, D). In the y-plane the differences within the corpus callosum extended from y = −21 to y = +33. Anteriorly, the bilateral regions of increased white matter density followed the rostral outline of the head of the caudate nuclei, extending into the anterior limbs of the internal capsule (figure, C). The widest extent of white matter density increase in the x-plane was from x = −24 to x = +24 (see figure, C). The widest extent in the z-plane was from z = −1 in the rostrum to z = +33 in the midbody of the corpus callosum (see figure, D). ⇓

View this table:

• View inline
• View popup

Table 1.

z-Scores and coordinates in Talairach space³⁵ for regions with significant differences (p < 0.01) when comparing pooled white matter

• Download figure
• Open in new tab
• Download powerpoint

Figure. (a–h) The left column (a–d) demonstrates the comparison of pooled white matter data for X-linked Kallmann’s syndrome (xKS) patients (with mirror movements) versus normal men. The right column (e–h) demonstrates the comparison of pooled white matter data for autosomal Kallmann’s syndrome (aKS) patients (without mirror movements) versus normal men. Regions of increased white matter density (p < 0.01) are superimposed in yellow on the Montreal Neurological Institute reference brain.37,38 (a, b) xKS: Increased white matter density in relation to the course of the corticospinal tract. Coronal section (a) at y = −24 and transverse section (b) at z = −18. (c, d) xKS: Increased white matter density in the mid and anterior sections of the corpus callosum and its radiation. Transverse section (c) at z = +11 and midsagittal section at x = 0. (e, f) aKS: Increased white matter density in the splenium of the corpus callosum but no increase in relation to the course of the corticospinal tracts. Coronal section (e) at y = −24 and transverse section (f) at z = −18. (g) aKS: Increased white matter density in the callosal radiation. Transverse section at z = +11. (h) aKS: Increased white matter density in the splenium of the corpus callosum. Midsagittal section at x = 0.

Bilateral increased white matter density was found in the posterior limbs of the internal capsule, trailing down laterally through the midbrain (figure, A and B). The extent in the z-plane was from z = +8 to z = −22. Peak changes identified by SPM(z) were located in the corticospinal tract as outlined in the stereotaxic atlas of Talairach and Tournoux,36 and onto posterior areas immediately adjacent to it.

There was one additional extensive area of increased white matter density in the posterior aspect of the left occipital lobe, localizing to the optic radiation as outlined by Talairach and Tournoux.36 In the y-plane the area extended from y = −44 to y = −90, in the z-plane the area extended from z = +13 to z = −8, and in the x-plane the area extended from x = −10 to x = −30.

Region-of-interest analysis.

The ratio of the midsagittal area of the corpus callosum versus the total brain surface area was 20% greater in the xKS group than in the group of normal men (table 2). A t-test performed on the logarithm of this ratio for xKS and normal men demonstrated that this difference was significant at p < 0.0006 (CI, 1.11 to 1.30). A t-test on the differences of the absolute values for the midsagittal callosal area was significant at p < 0.03.

Mirror movements and the corpus callosum.

Dennis et al.23-25 proposed a lack of callosally mediated interhemispheric inhibition to explain mirror movements. They hypothesized that, under normal circumstances, there is an underlying propensity for both motor cortices to become active even when voluntary movements are intended with one hand, but that the motor cortex contralateral to the side of intended movement exerts a tonic inhibitory activity on the other motor cortex. Indeed, there is functional evidence for the existence of transcallosal inhibition.41-43 Interhemispheric connections of the primary motor cortices pass through the corpus callosum.44,45 The rostral half of the body of the corpus callosum contains predominantly interhemispheric fibers to the motor cortex and supplementary motor cortex, whereas the dorsal and ventral sectors of the caudal part of the genu contain fibers to the lateral premotor regions.

In the current study, the middle and anterior parts of corpus callosum were significantly larger in the xKS patients than in the normal control subjects. However, an increase in callosal size was also observed when comparing the aKS patients with the normal control subjects, albeit in more posterior regions. Direct comparison between xKS and aKS patients revealed no significant differences in callosal size between the two groups. It follows that increased callosal size does not segregate consistently with the presence or absence of mirror movements in Kallmann’s syndrome.

Mirror movements and the corticospinal tracts.

The alternative hypothesis is that mirror movements in xKS patients result from an abnormal complement of corticospinal tract fibers that project ipsilaterally.21 Muller et al.46 studied the maturation of ipsilateral corticospinal tract projections in normal children age 3 through 11 years using focal transcranial magnetic brain stimulation. They elicited ipsilateral MEPs in two-thirds of the children examined. These occurred more often in proximal than in distal arm muscles and not in children older than 10 years. The latency of the MEPs was approximately 12 to 14 msec longer than the usual contralateral response, indicating a slow conduction velocity. Muller et al.46 hypothesized that the disappearance of ipsilateral MEPs in older children was related to the development of tonic transcallosal inhibition as the corpus callosum became fully myelinated. However, the physiologic ipsilateral corticospinal projections they identified are clearly not analogous to the fast-conducting corticospinal pathways, innervating predominantly distal forearm muscles, demonstrated in xKS patients by Mayston et al.22

The current study has demonstrated two symmetric tracts of increased white matter density, extending from the lower limb of the internal capsule and laterally through the midbrain toward the pons. Looking at these tracts as coregistered onto the MNI template (see figure, A and B), they might appear to lie just posterior to the conventional anatomic course of the corticospinal tract in humans. However, careful examination of the peak white matter differences in the xKS group shows that more than 50% correspond to the course of the corticospinal tract as delineated by Talairach and Tournoux.36 Moreover, these peaks lie largely within the confidence limits for the course of the corticospinal tract shown in the autopsy study of 30 human brainstem specimens by Afshar et al.47 It should also be noted that this application of SPM-96 to morphometric analysis18 depends on the normalization and smoothing of data derived from a group of nine individuals. Thus the confidence limits for precise anatomic localization are not clear. Finally, it is possible that an abnormal development of corticospinal tract fibers occurs in xKS at the borders of the normal course of the corticospinal tract. Taking all these factors into account, the data point to bilateral hypertrophy of the corticospinal tract in patients with xKS exhibiting mirror movements.

Differences in structure do not necessarily entrain differences in function, and vice versa, such that a physiologic alteration in transcallosal inhibition might not necessarily be reflected in a change in size of the corpus callosum. Therefore, this study cannot exclude entirely a callosal contribution to the etiology of mirror movements. Nevertheless, the data constitute strong anatomic evidence to support the physiologic data of Mayston et al.,22 suggesting that mirror movements in xKS are due to abnormal, fast-conducting ipsilateral corticospinal pathways. It is thus likely that anosmin-1, the protein encoded by KAL, plays a role in the development of the corticospinal tracts—specifically by inhibiting the establishment of fast-conducting ipsilateral pathways directly innervating those anterior horn cells that supply distal forearm muscles. This theory is consistent with the identification of KAL messenger RNA in embryonic human spinal cord tissue at 45 days postfertilization by reverse-transcriptase PCR.48