LONGITUDINAL DEVELOPMENT OF THE SPINE AND SPINAL CORD IN HUMAN FETUSES

Objective: The spinal cord extends from the foramen magnum to the sacrum in the human fetus at the beginning of the 2 nd quarter. However, the medullary cone is located at or above the level of the second lumbar vertebra at birth. The objective is to determine the difference between the rates of longitudinal growth of the spinal cord and the spine in human fetuses from the 13 th to the 22 nd week of gestation (WoG) using magnetic resonance imaging (MRI). Methods: Descriptive observational cross-sectional study of 24 stillbirths (13 ♂ , 11 ♀ ), between the 13 th and 22 nd WoG, using spinal MRI. We recorded spine and spinal cord lengths in millimeters from the foramen magnum to the coccyx for the former and to the medullary cone for the latter. We identified the position of the medullary cone according to vertebral level and its correlation with the gestational age and the literature. Results: The spinal cord increased in length from 50 to 93 mm, the spine from 57 to 137 mm, and the medullary cone rose from S1 to L2. The rate of growth was 1.2 mm/day for the spine and 0.6 mm/day for the spinal cord. Conclusions: Discordance in the rate of normal longitudinal growth of the spine and spinal cord caused the medullary cone to rise from S1 level to L2 in the second trimester of pregnancy. These results allow an understanding of normal development and certain congenital malformations. Level of evidence IV; Case series. descriptivo observacional transversal en 24 mortinatos (13 ♂ , 11 ♀ ), con edades entre la 13ª y 22ª SG, mediante RM de CV. Se registró la longitud de CV y ME, en milímetros, desde el foramen mágnum al coxis en la primera y hasta el cono en la segunda. Se identificó la posición del cono según el nivel vertebral y su correlación con edad gestacional y literatura. Resultados: La ME incrementó su longitud de 50 a 93 mm, la CV de 57 a 137 mm y el cono medular ascendió desde S1 hasta L2. El ritmo de crecimiento fue de 1.2 mm/día para la CV y de 0.6 mm/día para la ME. Conclusiones: La discordancia en el ritmo de crecimiento longitudinal normal de la CV y ME determinó que el cono medular ascienda desde el nivel S1 hasta L2 en el segundo trimestre de gestación. Los resultados permiten comprender el desarrollo normal y ciertas malformaciones congénitas.


INTRODUCTION
The development of fetal ultrasonography and magnetic resonance imaging significantly improved the accuracy of intrauterine diagnosis of open and closed defects of the neural tube. 1 The hypertrophy and shortening of the filum, also known as the tight filum terminale, is a simple spinal dysraphism without subcutaneous masses; 2 it can cause tethered spinal cord syndrome in the child preventing the normal ascent of the medullary cone. [3][4][5] This condition can give symptoms in childhood, or even in adulthood, linked to neurological deficit from spinal cord traction, but also as repeated episodes of meningitis or spinal abscesses. 1 Structural defects of this type arise from an error in the normal morphogenesis of the filum terminale and the medullary cone 6 during the "retrogressive differentiation" phase, in which the caudal medulla atrophies, forming the filum and the coccygeal ligament, and the cone ascends progressively. 1,3 Now, the spinal cord extends from the foramen magnum to the sacrum in the human fetus at the beginning of the second trimester and subsequently ascends. At birth or at two months of life, it is located at the level of the second lumbar vertebra or above (L1-L2), or in any case above the L2-L3 disc, (Figure 1) a position that it maintains throughout life. 1,4,7,8 In fact, during childhood, once the spinal cord has reached its definitive level, growth of the spinal cord and the spine continue at the same rate and rhythm until adulthood. 8 The normal rate of medullary ascent during the fetal period until reaching its final level has not been fully clarified. It has been shown that after thirty five weeks the cone is located near its final level 7,9 and there are previous studies that have demonstrated the progressive ascent of the cone during the fetal period, relating it especially with the length of the fetus and with the vertebral level each week, 7,8,10 though none has measured the differential growth speeds between the spinal cord and the spine. Some authors 7 refer to the individual rate of spinal cord and spine growth, suggesting that it would determine the "relative ascent" of the cone, but they have not measured these speeds. Defining the normal and proportional rate could help to understand the pathological changes of this part of morphogenesis, which leads to the genesis of disorders like tight filum terminale.
The objective of our study was to determine the differences in the rates of longitudinal growth of the spinal cord and the spine in human fetuses from the 13th to the 22nd week of intrauterine gestation (WG) using magentic resonance (MR).

METHODS
A quantitative basic science field investigation was conducted through a descriptive observational non-probabalistic cross-sectional study of 24 stillbirths, ( Figure 2) with ages between the 13th and 22nd weeks of gestation using MRs of the spine. These specimens were collected from the Developmental Anatomy Laboratory of the Facultad de Medicina de la Universidad Nacional del Nordeste (UNNE), in Corrientes, Argentina. Since the study was conducted on specimens from a Museum of Anatomy, most of which had been archived there for many years, approval by an Institutional Review Board was neither requested nor required. There were also no data that linked the samples to any living or missing person. All the fetuses were in a good state of preservation and without malformations evident under microscopic inspection. For the MR study, a 1.5 Tesla (General Electric, model Brivo MR 355 1.5 T) apparatus with a dedicated 8-channel coil system was used. A volumetric T1 FSPGR (fast spoiled gradient echo) sequence was acquired with TR=8.7, TE=3.1, angle=20º, 2 acquisitions, field of vision 160 mm, matrix = 448x224 obtaining between 164 and 180 1 mm sagittal images, depending on the size of the fetus, to visualize the entire anatomy. The images were processed at a workstation (General Electric Advantage Workstation 4.6).
The gestational age of each specimen was determined by the fetal biometry tables commonly used in ultrasound. Thus, the femoral length and the biparietal diameter were recorded in millimeters. Both parameters were obtained two times and averaged. For the estimation of the age, the 50th percentile of the tables prepared by Kurmanavicius J. et al. 11,12 were considered. In addition, automatic age determination was peformed with the same measurements using ultrasound equipment (Esaote, MyLab 60) without finding differences between the two determinations.
The length of the spine and spinal cord were then recorded in millimeters from sagittal images, from the foramen magnum to the coccyx and from the foramen magnum to the cone, respectively. ( Figure 3) In addition, the position of the medullary cone was identified according to the vertebral level. The data were correlated with gestational age.   The speed of growth of the spine and spinal cord were calculated arithmetically using the mean lengths and the difference in days between the youngest and the most developed specimens. The Mann-Whitney U test was used to verify the homogeneity of the samples in terms of sex and the length measurements performed and Fisher's exact test was used to assess the differences between the highest and lowest cone termination levels by sex.

RESULTS
The results are summarized in Table 1. First, the entire fetal anatomy was evaluated through images and specimens with macroscopic malformations of any organs were eliminated, thus ensuring that all the fetuses to be included in the study were without malformations or dysmorphies. Ten (10) specimens had the cone at level L3, 5 at level L4, 4 at level S1, 3 at level L5, and 1 each at levels L2 and S2. The Fisher's test resulted in no statistical differences (p = 0.2445) for the comparison between the sexes of the highest and lowest levels of the termination of the medullary cone.
The spinal cord increased in length from 50 to 93 mm, the spine from 57 to 137 mm, and the medullary cone ascended from S1/ S2 to L2 between the 13 th and 22 nd weeks of gestation ( Figure 4). There were statistically significant differences between the lengths of the spine in male and female fetuses (p = 0.4884) with the Mann-Whitney test, but not for the length of the spinal cord (p = 0.6432). The mean growth rate observed was 1.2 mm/day for the spine and 0.6 mm/day for the spinal cord. The mean ascent verified during the entire trimester was four (4) vertebral segments (i. e., 4 bodies and 4 discs).

DISCUSSION
Different malformations, disruptions, and dysplasias can occur on the spine and spinal cord of the fetus in formation, secondary to intrinsic, extrinsic, and mixed factors. 13 The formation of the spinal cord develops in three phases: gastrulation, primary neurulation, and secondary neurulation. 9,14,15 After the closure of the neural groove at approximately 28 days of gestation, the masses of cells caudal to the posterior neuropore tunnel down and form a channel that gives origin to the most distal portions of the spinal cord: the filum terminale and the medullary cone. 6 In the third and final phase, at around 11 weeks, "retrogressive differentiation" occurs, in which the caudal structures undergo a programmed and orderly necrosis leaving the filum, the coccygeal ligament, and the terminal ventricle of the cone. 1,9,10,16,17 It has been shown that only after this process does the disproportion between the growth rates of the spinal cord and the spine begin to be apparent. 8 Coincidentally, in this study we found a growth rate of 1.2 mm/day for the spine and 0.6 mm/day for the spinal cord. (Table 2) On the contrary, most spinal malformations originate during somitogenesis, especially in the segmentation phase, that is, during the embryonic period. 18 However, the phenomenon of progressive caudal reduction of the spinal cord is a phenomenon of natural morphogenesis that is also seen to some extent in somites, although in a different period. The initial number of post-sacral somites decreases from eight (±2) to four (±1). 19,20 Therefore, sacral vertebral anomalies as well as the most severe forms of lumbosacral agenesis, can find their pathogenic explanation either in the alteration of this phase 6,19,20 or in the defects of the segmental formation of the notochord. 21     Table 3. Position of the medullary cone by weeks of gestation.

Weeks of Gestation
Position of the Cone L2 L3 L4 L5 S1 S2 The level of the medullary cone in the spine of the fetus is quite variable, but its normal position in the newborn is at the level of disc L2-L3 and it ascends to its final position at the level of or above disc L1-L2 between 2 months and 1 year of extrauterine life 9 . (Figure 1) Our data coincided with the findings of Zalel et al. in their prospective ultrasonographic study of 110 gestations with fetuses between 13 and 40 weeks, 22 in which almost all the fetuses under 18 weeks of gestation had the medullary cone at or below the level of vertebral segment L4. (Table 3) In the fetuses under 35 weeks of gestation, the cone was positioned between levels L2 and L5, while in those over 35 weeks of gestation the cone was positioned above level L3. 7,9 At 39 weeks, the cone was in this location in 94% of the fetuses 7 . (Table 3) Some authors 7 maintain that the ascent of the cone is constant throughout the life of the fetus. Others have demonstrated the rapid ascent of the medullary cone before 35 weeks of gestation, [8][9][10] as well as the appearance of disproportionate growth rates between the spine and the spinal cord during this period. 8 Although Vettivel 10 relates the growth and ascent of the medullary cone to the length of the fetus in his study, the latter can be easily correlated with gestational age 23 . (Table 4 and Figure 5) Our data confirm these findings, given that the speed of growth of the spine is double that of the spinal cord during the second trimester of pregnancy ( Figure 6). However, in the sample examined here, the cone has a definitive position at 22 weeks of gestation, which is temporally prior to the bibliography consulted. In his study, Arthurs found only 23% of the fetuses with the cone positioned at level L3 or above. 7 This difference between our findings and those of Arthurs et al. may be related to the formaldehyde preservation time of the specimens used in our study. (Figure 4) Unlike what happens in open spinal dysraphism, occult dysraphic anomalies are usually not obvious at birth. 1,3 And although a tethered spinal cord sindrome has also been described with the cone in the normal position, 3,24 the fact that the ascent of the cone is almost complete during the first two trimesters of pregnancy is important 25,26 . (Table 3) The mean ascent of four (4) vertebral segments (4 bodies and 4 discs) during the second trimester that we observed in this study contrasts strongly with the mean ascent of one (1) to one and a half (1.5) vertebral bodies described by Arthurs 7 for the third trimester and may be hypothetically due to the biological variability of the samples.
However, both the findings of our study and of others 7,9,22 indicate that with ultrasonographic or magnetic resonance studies during pregnancy, it can be assumed that if after 35 weeks the medullary cone is below L2-L3, the spinal cord is in an abnormally low position and some caudal malformation may exist.
In a similar study, although with a larger case series, the authors could not establish either the sex or the race of the specimens, so demographic conclusions could not be drawn. 7 Other studies 10 have observed racial differences with respect to the vertebral level of the termination of the spinal cord, but no statistical differences between the sexes have been reported. 10 In this study we found a difference between the sexes in the length of the spine but not of the spinal cord.
In any case, our work has some limitations to be noted. Among them, the prolonged time that the specimens remained in Figure 6. Correlation between the speeds of growth of the spinal cord and the spine. formaldehyde, possibly causing retraction of the central nervous system and distorting the acquired data, and the small number of specimens evaluated compared to other reports using the same technique. 7

CONCLUSION
The discordance between the normal longitudinal growth of the spine (1.2 mm/day) and the spinal cord (0.6 mm/day) indicates that the medullary cone ascends from level S1 to level L2 from the 13 th to the 22 nd week of gestation. The results help to understand normal development and some congenital malformations, such as tethered spinal cord.