Mutagen sensitivity and risk of second cancer in younger adults with head and neck squamous cell cancer: 15-year results.

PURPOSE: To evaluate the mutagen sensitivity phenotype on the risk of second primary cancer (SPC) in patients with head and neck squamous cell carcinoma (HNSCC), and to estimate the long-term rate of SPC and the outcome with SPC. METHODS: A survey was made regarding SPC among 124 younger (≤ 50 years) adults with HNSCC who were enrolled in a pretreatment mutagen sensitivity investigation during 1996-2006. Mutagen sensitivity was assessed by exposing lymphocytes to bleomycin in vitro and quantifying the bleomycin-induced chromatid breaks per cell (b/c). Patients were classified as hypersensitive (> 1 b/c) or not hypersensitive (≤ 1 b/c). RESULTS: Mean follow-up time for all patients was 68 months (range: 5-288 months), and the 15-year cancer-specific survival was 15%. Twenty patients (16%) developed a SPC (15-year estimated rate: 41%), and half of them was hypersensitive. The crude rate of SPC for hypersensitive (n = 65) or not hypersensitive (n = 59) patients were 15 and 17%, respectively (p = 0.4272). The 15-year estimated rate of SPC for hypersensitive and not hypersensitive patients was 36 and 48%, respectively (p = 0.3743). Gender, UICC stages, anatomical sites of index cancer did not prove to be a significant risk factor for SPC. Forty-five percent of SPC developed after the 10-year follow-up. The 3­year cancer-specific survival was 23% with SPC. CONCLUSION: According to our findings, mutagen hypersensitivity was not associated with an increased SPC risk in HNSCC patients. Patients are at a lifelong risk of developing a SPC. Survival with SPC is very poor.

Measurement, simulation and uncertainty assessment of implant heating during MRI.

The heating of tissues around implants during MRI can pose severe health risks, and careful evaluation is required for leads to be labeled as MR conditionally safe. A recent interlaboratory comparison study has shown that different groups can produce widely varying results (sometimes with more than a factor of 5 difference) when performing measurements according to current guidelines. To determine the related difficulties and to derive optimized procedures, two different generic lead structures have been investigated in this study by using state-of-the-art temperature and dosimetric probes, as well as simulations for which detailed uncertainty budgets have been determined. The agreement between simulations and measurements is well within the combined uncertainty. The study revealed that the uncertainty can be kept below 17% if appropriate instrumentation and procedures are applied. Optimized experimental assessment techniques can be derived from the findings presented herein.

Technical note: standardized and semiautomated Harris lines detection.

Arrest in long bone growth and the subsequent resumption of growth may be visible as radiopaque transverse lines in radiographs (Harris lines, HL; Harris, HA. 1933. Bone growth in health and disease. London: Oxford University Press). The assessment of individual age at occurrence of such lines, as part of paleopathological skeletal studies, is time-consuming and shows large intra- and interobserver variability. Thus, a standardized, automated detection algorithm would help to increase the validity of such paleopathological research. We present an image analysis application facilitating automatic detection of HL. On the basis of established age calculation methods, the individual age-at-formation can be automatically assessed with the tool presented. Additional user input to confirm the automatic result is possible via an intuitive graphical user interface. Automated detection of HL from digital radiographs of a sample of late Medieval Swiss tibiae was compared to the consensus of manual assessment by two blinded expert observers. The intra- and interobserver variability was high. The quality of the observer result improved when standardized detection criteria were defined and applied. The newly developed algorithm detected two-thirds of the HL that were identified as consensus lines between the observers. It was, however, necessary to validate the last one-third by manual editing. The lack of a large test series must be noted. The application is freely available for further testing by any interested researcher.

Interactive visuo-haptic surgical planning tool for pelvic and acetabular fractures.

Treatment of pelvic and acetabular fractures still poses a major challenge to trauma surgeons. We present a tool for intervention planning for such injuries using patient-specific models built from Computed Tomography data. The presented tool has three main parts: (1) the virtual reduction of the bone fragments, (2) the virtual adaptation of the osteosynthesis implants and (3) Finite Element Analysis (FEA) for testing mechanical behavior of the resulting intervention plan. Our tool provides an intuitive visuo-hapic interface designed to be used by trauma surgeons. The type and size of the osteosynthesis material can be determined and measurements like distances and angles relative to landmarks can be taken. First results of prospectively planned interventions show an excellent correlation and a significant gain in operation time.

4D MR imaging of respiratory organ motion and its variability.

This paper describes a method for 4D imaging, which is used to study respiratory organ motion, a key problem in various treatments. Whilst the commonly used imaging methods rely on simplified breathing patterns to acquire one breathing cycle, the proposed method was developed to study irregularities in organ motion during free breathing over tens of minutes. The method does not assume a constant breathing depth or even strict periodicity and does not depend on an external respiratory signal. Time-resolved 3D image sequences were reconstructed by retrospective stacking of dynamic 2D images using internal image-based sorting. The generic method is demonstrated for the liver and for the lung. Quantitative evaluations of the volume consistency show the advantages over one-dimensional measurements for image sorting. Dense deformation fields describing the respiratory motion were estimated from the reconstructed volumes using non-rigid 3D registration. All obtained motion fields showed variations in the range of minutes such as drifts and deformations, which changed both the exhalation position of the liver and the breathing pattern. The obtained motion data are used in proton therapy planning to evaluate dose delivery methodologies with respect to their motion sensitivity. Besides this application, the new possibilities of studying respiratory motion are valuable for other applications such as the evaluation of gating techniques with respect to residual motion.

Patient specific simulation and navigation of ventriculoscopic interventions.

In this paper a comprehensive framework for pre-operative planning, procedural skill training, and intraoperative navigation is presented. The goal of this system is to integrate surgical simulation with surgical planning in order to improve the individual treatment of patients. Various surgical approaches and new, more complex procedures can be assessed using a safe and objective platform that will allow the physicians to explore and discuss possible risks and benefits prior to the intervention. A simulation environment extends the pre-operative planning in a natural way, as it allows for direct evaluation of the surgical approach envisioned for each case. In addition, by providing intraoperative navigation based on this simulation, surgeons can carry out the previously optimized plan with higher precision and greater confidence.

Tumor growth models to generate pathologies for surgical training simulators.

Many virtual reality based surgical training simulators have been presented in the last few years. These systems promise to alleviate the lack of realistic training possibilities common to minimally invasive procedures. Virtual reality allows for riskless training on a wide range of findings in a condensed period of time. We investigated different methods for the generation of tumor models suitable for surgical training simulators. The goal of our research is a high fidelity hysteroscopy simulator which provides an individual surgical scene for every training. Emphasis was placed on the modeling of growth processes leading to the generation of macroscopically realistic findings of the most common pathologies in hysteroscopy, namely polyps and myomas found in the uterine cavity. Both a cellular automaton and a particle based tumor growth model are presented and discussed.

Generation of variable anatomical models for surgical training simulators.

The generation of variable surgical scenes is a key element for effective training with surgery simulators. Our current research aims at a high fidelity hysteroscopy simulator which challenges the trainee with a new surgical scene in every training session. We previously reported on methods able to generate a broad range of pathologies within an existing healthy organ model. This paper presents the methods necessary to produce variable models of the healthy organ. In order to build a database of uteri, a volunteer study was conducted. The segmentation was carried out interactively, also covering the establishment of an anatomically meaningful correspondence between the individual organs. The variability of the shape parameters has been characterized by principal component analysis. A new method has been developed and tested, allowing the derivation of realistic new instances based on the stochastic model and complying with non-linear shape constraints which are defined and interactively controlled by medical experts.

In vivo validation of spatio-temporal liver motion prediction from motion tracked on MR thermometry images.

PURPOSE: Magnetic resonance-guided focused ultrasound (MRgFUS) of the liver during free-breathing requires spatio-temporal prediction of the liver motion from partial motion observations. The study purpose is to evaluate the prediction accuracy for a realistic MRgFUS therapy scenario, namely for human in vivo data, tracking based on MR images routinely acquired during MRgFUS and in vivo deformations caused by the FUS probe. METHODS: In vivo validation of the motion model was based on a 3D breath-hold image and an interleaved acquisition of two MR slices. Prediction accuracy was determined with respect to manually annotated landmarks. A statistical population liver motion model was used for predicting the liver motion for not tracked regions. This model was individualized by mapping it to end-exhale 3D breath-hold images. Spatial correspondence between tracking and model positions was established by affine 3D-to-2D image registration. For spatio-temporal prediction, MR tracking results were temporally extrapolated. RESULTS: Performance was evaluated for 10 volunteers, of which 5 had a dummy FUS probe put on their abdomen. MR tracking had a mean (95 %) accuracy of 1.1 (2.4) mm. The motion of the liver on the evaluation MR slice was spatio-temporally predicted with an accuracy of 1.9 (4.4) mm for a latency of 216 ms. A simple translation model performed similarly (2.1 (4.8) mm) as the two MR slices were relatively close (mean 38 mm). Temporal prediction was important (10 % error reduction), while registration effects could only partially be assessed and showed no benefits. On average, motion magnitude, motion amplitude and breathing frequency increased by 24, 16 and 8 %, respectively, for the cases with FUS probe placement. This motion increase could be reduced by the spatio-temporal prediction. CONCLUSION: The study shows that tracking liver vessels on MR images, which are also used for MR thermometry, is a viable approach.

Abnormal immune response to brain tissue antigen in the syndrome of autism.

Cell-mediated immune response to human myelin basic protein was studied by the macrophage migration inhibition factor test in 17 autistic patients and a control group of 11 patients suffering from other mental diseases included in the differential diagnosis of the syndrome of autism. Of the 17 autistic patients, 13 demonstrated inhibition of macrophage migration, whereas none of the nonautistic patients showed such a response. The results indicate the existence of a cell-mediated immune response to brain tissue in the syndrome of autism.

Simulation of the Effect of Synapses: the Significance of the Dendritic Diameter in Impulse Propagation.

The effectiveness of synapses at various sites of the dendritic tree was studied using a segmental cable model with a program developed by Hines (Int. J. Biomed. Comput., 24, 55 - 68, 1989). The model rendered possible a high-fidelity simulation of the dendritic geometry of a frog motoneuron described in the accompanying paper (Birinyi et al., Eur. J. Neurosci., 1003 - 1012, 1992). The model was used in the passive membrane mode and the synaptic activity was simulated with current injections into large and small diameter dendrites at proximal and distal locations. Synaptic efficiency was defined by the charge transfer ratio expressed as the proportion of the injected current which appeared at the soma. The charge transfer ratio was determined with uniform and non-uniform distribution of specific membrane resistance over the soma - dendrite surface while the diameter of selected dendrite segments changed. The best charge transfer ratio was found with the largest dendrite membrane resistance, and the maximum efficiency of synaptic activity appeared at the original size of the dendrite segment stimulated. The amount of current that flowed in the proximal and distal directions from the segment stimulated depended on the diameter of that segment. The increase in diameter of proximal dendrites increased synaptic efficiency on distal dendrites, whereas the reverse caused a decline in synaptic efficiency on proximal dendrites. In addition to the diameter of dendrites, the arborization pattern also played a significant role in this mechanism. It is concluded that the cellulipetal increase in dendrite diameter greatly increases synaptic efficiency.

The Extent of the Dendritic Tree and the Number of Synapses in the Frog Motoneuron.

Frog motoneurons were intracellularly labelled with cobaltic lysine in the brachial and the lumbar segments of the spinal cord, and the material was processed for light microscopy in serial sections. With the aid of the neuron reconstruction system NEUTRACE, the dendritic tree of neurons was reconstructed and the length and surface area of dendrites measured. The surface of somata was determined with the prolate - oblate average ellipsoid calculation. Corrections were made for shrinkage and for optical distortion. The mean surface area of somata was 6710 microm2; lumbar motoneurons were slightly larger than brachial motoneurons. The mean length of the combined dendritic tree of brachial neurons was 29 408 microm and that of lumbar neurons 46 806 microm. The mean surface area was 127 335 microm2 in brachial neurons, and 168 063 microm2 in lumbar neurons. The soma - dendrite surface area ratio was 3 - 5% in most cases. Dendrites with a diameter of 600 microm from the soma. This suggests that about two-thirds of the synapses impinged upon distant dendrites >600 microm from the soma. The efficacy of synapses at these large distances is investigated on model neurons in the accompanying paper (Wolf et al., Eur. J. Neurosci., 4 1013 - 1021, 1992).

Topography and organization of cranial nerve nuclei in the sand lizard, Lacerta agilis.

Cobaltic-lysine complex compound was used to label cranial nerves of the ventrolateral (branchiomotor) and dorsomedial (somatomotor) nuclear columns in the sand lizard, Lacerta agilis. The dendritic arborizations and axonal trajectories of neurons of the respective nuclei were reconstructed from serial sections. A fairly uniform neuronal morphology was found in the nuclei of the ventrolateral column: a spindle-shaped perikaryon gave rise to dorsomedial and ventrolateral dendritic trees, the latter arborizing in a characteristic broomlike manner within a narrow region in the lateral white matter. Axons of all neurons converged upon the medial longitudinal fasciculus and after making a hairpin turn formed the corresponding motor roots. A group of small neurons constituted a separate subnucleus within the V motor nucleus. The VII and IX nuclei were fused into a single nuclear complex. The nucleus ambiguus was found dorsal to the XII nucleus and lateral to the dorsal vagal nucleus. The latter nucleus extended rostrally to the caudal pole of the VI nucleus, and its neurons sent axons to the VII, IX, and X nerves. The term "dorsal visceromotor column" designates the extended dorsal vagal nucleus. A number of small polygonal neurons lying scattered in the lateral part of the medulla were labeled via the VII, IX, and X nerves. This loose aggregate of labeled neurons was termed the "lateral visceromotor area." On the basis of nuclear topography and cellular morphology, the existence of a bulbar XI nucleus was excluded. Three different types of neurons could be distinguished in the dorsomedial nuclear column. Neurons with oval or spherical perikarya and radially oriented dendrites constituted the nuclei innervating external eye muscles. Except for the IV nucleus, axons followed a ventral trajectory. The accessory VI nucleus was composed of a second type of neuron with elongated soma and dorsoventral dendrite orientation; the dorsally directed axon turned ventrally at the VI nucleus. The XII nucleus contains a third type of neuron with strongly decussating dendrites. The distinct differences in the neuronal morphology did not support the classical assumption that all of the nuclei of the dorsomedial motor column supply muscles derived from somitic mesoderm. Sensory fibers of the trigeminal nerve formed the familiar spinal tract, which partially decussated in the medullospinal transition zone and could be followed as far as the lumbar segments on the ipsilateral side of the spinal cord. Neurons of the mesencephalic root of the trigeminal nerve were localized in the optic tectum; their descending fibers joined the medial aspect of the spinal tract.(ABSTRACT TRUNCATED AT 400 WORDS)

Organization of the ambiguus nucleus in the frog (Rana esculenta).

The common root of the glossopharyngeal, vagal, and accessory nerves and the individual branches of the vagus complex were labeled with cobalt, and the organization of the ambiguus nucleus was studied. The cell column labeled through the common root extended from the upper part of the medulla to the rostral spinal cord over a distance of about 3,500 microns. The labeling of individual branches revealed four subdivisions. 1) The pharyngomotor subdivision occupied the rostral 800 microns of the cell column. It gave origin to the innervation of the pharyngeal muscles. 2) The visceromotor subdivision, consisting of small and medium-sized cells labeled by way of the visceral branches of the vagus, was found in the rostrocaudal extent of the medulla. 3) the laryngomotor subdivision extended in the obex region over a distance of more than 1,000 microns. It supplied the sphincter muscles of the larynx. The dilator laryngeal muscle was represented in the rostral part of the visceromotor subdivision. 4) The accessory nerve subdivision was located in the lower medulla and the rostral spinal cord. From the results, the following conclusions are drawn. 1) The basic organization of the frog ambiguus nucleus is comparable to that of the rat, differences in nuclear organization reflecting differences in peripheral structures. 2) The cytoarchitectonic structure of the four subdivisions innervating different peripheral targets characteristically differ from each other. 3) On the basis of its characteristic neuronal morphology, the accessory nerve nucleus is regarded as an independent structure.

The accessory motor nuclei of the trigeminal, facial, and abducens nerves in the rat.

The cobalt labeling technique was applied to the trigeminal, facial, and abducens roots in order to investigate the location and morphology of motoneurons of the main and accessory motor nuclei of these cranial nerves in the rat. On the basis of soma-dendritic morphology, axonal trajectory, and tegmental location, three different accessory nuclei were distinguished in the brainstem. Neurons of the accessory nuclei of V and VII form a common club-shaped nucleus extending between the main motor nuclei of V and VII. The dendritic tree of these neurons is oriented in the dorsoventral direction and the axon invariably follows a dorsolaterally directed bent course. The accessory VI nucleus is located ventral and lateral to the main VI nucleus in the lateral tegmentum. The neurons have elongated perikarya, strong ventral and weak dorsal dendritic arbors, and the bent axonal course is directed dorsomedially. Neurons of the main VI nucleus have rounded perikarya and radial dendritic arborizations. Muscle groups innervated by the accessory and main motor nuclei have distinctly different functions.

Tectal neurons of the frog: intracellular recording and labeling with cobalt electrodes.

With the aid of cobalt-filled intracellular electrodes, responses evoked by electrical and visual stimulation were recorded from tectal neurons. Diffuse light and "edge" and "spot" stimuli were used for visual stimulation; electrical stimuli were delivered by means of bipolar electrodes to the optic tract. The cells were subsequently filled with cobalt for a correlative histological study. The following types of neurons were distinguished: (1) Pear-shaped neurons with restricted and broad dendritic fields occurring in tectal layers 8, 6, 4, and 2. In one group, the dendrites bore beads and small protrusions and the short axon arborized within, or close to, the dendritic arbor. They responded mostly with inhibitory postsynaptic potentials to the various stimulations. This type of neuron was interpreted to be a tectal interneuron. In the second group the dendrites were smooth, the dendritic arbor was large, and the axon could be followed to layer 7 in a few cases. Excitatory activities with frequent spike discharges prevailed in their responses. They were regarded as efferent neurons of the tectum. (2) Large ganglionic neurons showed a greater variation in form than has been described in previous Golgi studies. Their responses to the different stimulations were also very variable. The axons of five neurons could be followed to the dorsal part of the thalamus, to the postoptic commissural system, and to the direct and crossed tectobulbar tract. These axons emitted several collaterals to tectal, pretectal, and tegmental structures. (3) A single stellate neuron was impaled in layer 9. It responded only to electrical stimulation with excitation. (4) Deep bipolar neurons were not mentioned in earlier works. Their somata were found in layer 6; the vertically oriented dendrites arborized in plexiform layers 7, 5, and 3, over a distance of more than 1 mm. These cells were sensitive only to electrical stimulation.

The dorsal column nuclei of the frog.

Surface potentials and field potentials recorded from the medulla in response to cutaneous and mixed nerve stimulation revealed that fibers of the dorsal white column project onto two separate neuron groups in the medulla representing the dorsal column nuclei. The hind-limb was represented in the medial group and the fore-limb in the lateral group in the projection. The fore-limb nerves projected to a region extending from the medulla to the 4th segment of the spinal cord. Primary afferent depolarization and depression of synaptic activities were shown by direct and indirect stimulation of dorsal column fiber terminals in the dorsal column nuclei. Conditioning volleys set up in hind-limb nerves had no effect on test responses evoked by stimulation of forelimb nerves, and vice versa. Slow negative potentials with decreasing latencies were recorded from the posterocentral nucleus of the thalamus in response to stimulation of the 2nd dorsal root, the dorsal column and the dorsal column nuclei, respectively. The physiological results were correlated with histological observations using the cobalt labelling method. It was concluded that the amphibian dorsal column-medial lemniscus system is closely comparable with that found in the mammalian brain.

Spatial distribution of pre- and postsynaptic sites of axon terminals in the dorsal horn of the frog spinal cord.

Axon terminals which could be interpreted as dorsal root boutons, were photographed from a series of 98 ultrathin sections with a Jeol 100B electron microscope. A total of 13 boutons were recovered for computer reconstruction. Two of them were terminal boutons, eight en passant boutons and three boutons were only partially recovered. All boutons contained multiple synaptic sites (maximum 33 and minimum seven) at which axodendritic and axoaxonic synapses were established. Axodendritic synapses were of the asymmetric type and they were directed toward adjacent dendrites. In axoaxonic synapses, which were of the symmetric type, the boutons were invariably on the postsynaptic side. Among the presynaptic profiles axons with spherical and pleomorphic vesicles and dendrites with flattened vesicles could be discerned. On average, each 2.67-microns2 bouton surface area contained one presynaptic site at which an axodendritic synapse was established, and each 7-microns2 surface area contained one postsynaptic site for an axoaxonic (or dendroaxonic) contact. A tendency of grouping of synaptic sites was observed. Distance measurements between the closest neighbours of all synaptic sites were made in four combinations in boutons with the original and with a random distribution of synaptic sites. The arithmetic mean of distances measured between the presynaptic and the closest postsynaptic sites was almost twice as big as that measured in the reverse direction. The difference between these values became greatly reduced in the case of random distribution. The arithmetic mean of distances between the closest neighbours of presynaptic sites was about the same as that between the closest neighbours of postsynaptic sites. This latter value was considerably increased with randomly distributed synaptic sites. The results suggest a non-random distribution of synaptic sites on the surface of boutons. The analysis of cluster formation of synaptic sites performed with a numerical taxonomy technique revealed that the majority of the 153 synaptic sites were comprised in 27 clusters containing both pre- and postsynaptic sites within the 1-micron similarity level. All postsynaptic sites were within 1 micron of one or more presynaptic sites. On the basis of the assumption that the postsynaptic sites are occupied by inhibitory axoaxonic synapses, it is suggested that the transmitter release from the presynaptic sites can be individually controlled in this structural arrangement. A probable mechanism of this function may be the passive invasion of the bouton by the impulse propagating actively along the dorsal root fibre.

Investigation of the dendritic geometry of brain stem motoneurons with different functions using multivariant statistical techniques in the frog.

We give an account of an effort to make quantitative morphological distinctions between motoneurons innervating functionally different muscles in the trigeminal and facial motor nuclei of the frog. Six groups of neurons were considered in the two nuclei on the basis of their peripheral targets. One group consisted of neurons (n = 7) innervating the levator bulbi muscle, which separates the orbital cavity from the oral cavity. In the second, third and fourth groups, motoneurons (n = 27) innervating jaw closer muscles (temporalis, masseter, pterygoideus) were studied. Neurons (n = 6) innervating the submaxillary muscle comprised the fifth group. This muscle forms the muscular floor of the mouth. It is active in deglutition and contributes to the opening of the mouth. The sixth group is formed by neurons of the facial nucleus (n = 7), which innervate the depressor mandibulae muscle. This is the main opener of the mouth. Neurons were selectively stained by cobalt labelling through the muscle nerves and the morphometric values of successfully labelled neurons were fed into a IBM AT 386 computer through a digitizing tablet for three-dimensional reconstruction. Four neurons labelled directly through the motor root of the trigeminal nerve but innervating unidentified muscles were added to the investigation. Two sets of quantitative measurements were taken from the neurons. In the first set (neurometric data), 17 quantitative variables were measured in the perikaryon and the dendritic arbor. In the second set, 15 variables concerned with the orientation and shape of the dendritic tree, the relation of the perikaryon to the dendritic tree and the spatial expansion of dendrites were measured in the three dimensions of Cartesian space (product-moment data). The data were subjected to multivariant statistical analysis. First, they were partitioned with cluster analysis. The average linkage between groups algorithm and the cosine of vectors of variables, or the Pearson correlation similarity coefficients were used. Neurometric data and product-moment data were analysed separately and in combination, and six to seven clusters were considered. In each case, the majority of neurons innervating jaw closer muscles were grouped into clusters different from neurons innervating jaw opener muscles. The best separation of functionally different neurons was achieved with the neurometric data set. The groups of neurons obtained from cluster analysis were subjected to non-parametric discriminant analysis with the eight nearest-neighbour classification criterion, and the results were checked with a cross-validation technique.(ABSTRACT TRUNCATED AT 400 WORDS)

The efferent system of cranial nerve nuclei: a comparative neuromorphological study.

A number of inconsistencies and controversies are inherent in the classification of cranial nerve nuclei based on the concepts of the various head-theories. The assumption of head segmentation, which is common to these theories, serves as the basis for designating the dorsomedial nuclei as the somatomotor column, although they innervate striated muscles of a viscus and a specific sense organ. The ventrolateral nuclei are called the specific visceromotor column; they innervate striated muscles in the branchiogenous area, but many of these muscles insert on skeletal elements. A series of comparative neuromorphological studies investigating the dendritic arborization pattern and axonal trajectory in the frog, lizard, and rat suggests a much more delicate classification in which nine morphologically and functionally different neuron groups can be discerned: 1. The hypoglossal nucleus appears coincidentally with the muscular tongue in amphibia. The spindle-shaped perikaryon, the bipolar dendritic arborization, and the straight ventral trajectory of the axon are characteristic morphological features in all three animal species investigated. 2. The oculomotor, trochlear, and abducens nuclei present a remarkably conservative topography and organization in all vertebrates with a moving eye. With their oval-shaped or polygonal perikarya and radiating dendritic arborization, these neurons distinctly differ from hypoglossal neurons. The ipsilateral axons follow a straight ventral course, the contralateral axons form a dorsal loop before crossing the midline, and the crossing is not consequence of neuron migration to the contralateral side. 3. The accessory abducens nucleus is present in tetrapods except apes and human. The elongated perikaryon and the dorsoventral dendritic orientation distinctly distinguish these neurons from other cranial motoneurons, the nucleus is found in the lateral part of the reticular formation. The neurons differentiate in situ, they do not migrate from the main abducens nucleus. 4. In the submammalian trigeminal and facial nuclei, two basic neuron types can be distinguished on the basis of their morphology. The first type is larger and accumulates in the rostral part of the trigeminal nucleus. This type innervates the jaw closer muscles. The second type is found in the caudal part of the trigeminal nucleus and in the facial nucleus. These neurons innervate the muscular floor of the mouth and the facial contingent supplies the jaw opener muscle. A very characteristic feature in the axonal trajectory is an initial medial course and a hairpin turn, or dorsal loop, at the lateral aspect of the medial longitudinal fasciculus. In addition to the two types of neurons, there is a third type in the frog trigeminal nucleus. This innervates an orbital muscle.(ABSTRACT TRUNCATED AT 400 WORDS)