MR anatomy, anatomical variants and morphometry of hippocampal formation Текст научной статьи по специальности «Клиническая медицина»

2019

ВЕСТНИК САНКТ-ПЕТЕРБУРГСКОГО УНИВЕРСИТЕТА МЕДИЦИНА

Т. 14. Вып. 3

ANATOMY

UDC 6.61:10

MR anatomy, anatomical variants

and morphometry of hippocampal formation

N. I. Ananyeva1,2, E. V. Andreev1, L. R. Akhmerova1, T. A. Salomatina1, L. I. Wasserman1,2, R. V. Grebenshchikova1, N. G. Neznanov1, A. A. Pichikov1, N. M. Zalutskaya1

1 V. M. Bekhterev National Medical Research Center for Psychiatry and Neurology (V. M. Bekhterev NMRC PN),

3, ul. Bekhtereva, St. Petersburg, 192019, Russian Federation

2 St. Petersburg State University,

7-9, Universitetskaya nab., St. Petersburg, 199034, Russian Federation

For citation: Ananyeva N. I., Andreev E. V., Akhmerova L. R., Salomatina T. A., Wasserman L. I., Grebenshchikova R. V., Neznanov N. G., Pichikov A. A., Zalutskaya N. M. MR anatomy, anatomical variants and morphometry of hippocampal formation. Vestnik of Saint Petersburg University. Medicine, 2019, vol. 14, issue 3, pp. 235-244. https://doi.org/10.21638/spbu11.2019.306

Nowadays the question of limbic structures involvement in different types of brain pathology is much debated in literature. However, obtained results are often contradictory. This can be explained by the insufficient knowledge of normal volume and linear measurements of brain structures responsible for human emotional and cognitive functioning including different age periods. Different anatomical variants of these structures were described in literature indistinctly, often leading to misinterpretation of neuroimaging findings. Besides, hippocampal formation, being complex structure, consists of different parts, including subregions (head, body and tail) and subfields (СА1-СА4, subiculum, presubiculum, dentate gyrus), which changes depend on different psychological and psychiatric symptoms. In our study we have analyzed MRI data of mediobasal parts of temporal lobes in healthy volunteers based on literature review and our own experience. The incidence rate of different hippocampal anatomical variants in healthy population was specified in the study. We have also determined MR voxel-based morphometry as a method permitting to define and evaluate volumes of different hippocampal subfields. In our research we found out certain significant differences in hippocampus fissure volumes, parasubiculum, molecular layer of dentate gyrus, fimbria, СА3 and СА4 Brodmann areas, demonstrating that in adulthood morphofunctional connections are not finally formed, that's why volumes of molecular layers CA1-CA3 smaller in adulthood

© Санкт-Петербургский государственный университет, 2020 https://doi.org/10.21638/spbu11.2019.306

than in elder population. But hippocampal fissure become smaller in elder ages because of

atrophic changes.

Keywords: anatomical variants, temporal lobe epilepsy, hippocampus, MRI, segmentation.

Introduction

In the last decade neurology, psychiatry, medical psychology, neurophysiology as well as other specialties started to require more detailed information concerning normal and pathological individual anatomical variability of brain. Therefore, methods of neurovisualisation became more important [1-3].One of the most modern and promising method of brain anatomy research is magnetic resonance imaging (MRI), permitting to obtain intravital morphometric characteristic of examined brain structures [4; 5]. Majority of literature sources contains qualitative and, to a less degree, quantitative [4] MRI analysis of examined structures [6]. There are only few studies in literature describing individual and gender anatomical variability of their brain structures [7; 8]. This information can be useful as the base of brain structures alterations assessment in different pathologic states.

Nowadays question of limbic structures involvement in different types of brain pathology is much debated in literature [6; 7]. However, obtained results are often contradictory. This can be explained by insufficient knowledge of normal volume and linear measurements of brain structures responsible for the human emotional and cognitive functioning including different age periods. Different anatomical variants of these structures, described in literature, are indistinct, which often leads to misinterpretation of neuroim-aging findings, especially in case of cognitive-affective proportion analysis [1]. More over, hippocampal formation is very complex structure, consisting of different parts, including subregions (head, body and tail) and subfields (CA1-CA4, subiculum, presubiculum, dentate gyrus), which changes depend on different psychological and psychiatric symptoms.

Materials and methods

We have examined 101 healthy volunteers aged 17-50 years without any neurologic or psychopathologic symptoms, 75 (70.3 %) of them were female, 30 (29.3 %) — male.

I. We used the following exclusion criteria:

1) younger than 17 years and elder than 50 years at the moment of study, which allowed to limit the influence of the age on the clinical findings and neuroimaging;

2) psychotic drug, alcohol or narcotic abuse;

3) severe decompensated somatic or neurologic disorders.

II. We examined 2 groups of normal volunteers for accurate analysis of subfields and subregions of hippocampus in young and old ages.

1. 10 healthy volunteers 13-21 years old;

2. 10 healthy volunteers 55 years and older.

All of the participants underwent quantitative evaluation of depression symptoms using HamiltonDepression Scale (HAM-D) and Montgomery-Asberg Depression Rating Scale (MADRS). Severity of anxiety symptoms was evaluated using Hamilton Anxiety scale (HAM-A). Total statistical analysis of the study results was performed using statistical packages "Statistica 6.0 for Windows" and SSPS 25.

MRI was performed using 1.5 T AtlasExelartVantageXGV (Toshiba, Japan) with 8-channel head coil. Standardized brain MRI protocol included fast spin echo sequences (FSE) for obtaining Т1 weighted images (T1-WI), T2-WI, and T2-FLAIR. For detailed examination of mediobasal parts of temporal lobes we used additional exam protocol, which included T2 FLAIR and Real IR in the oblique coronal plane with 2.2 mm slice thickness, obtained perpendicular to long hippocampus axis. These sequence structures of temporal lobes mediobasal parts include entorhynal cortex, head, body and tail of the hippocampus, temporal horns of lateral ventricles, and basal cisterns.

T2 FLAIR was performed using the following parameters: TR = 8000, TE = 105, FOV = 22.0, MTX = 30, ST = 2.2, GAP = 0.6, FA = 90/180.

To obtain REALIR we used the following parameters: TR = 3450, TE = 18, FOV = 22, MTX = 320, ST = 2.2, GAP = 0.6, FA = 90/160.

MR-images of the mediobasal parts of temporal lobes obtained in the oblique coronal plane were used for evaluation of hippocampal shape and degree of rotation; we measured hippocampal volume and level of head height, body and tail.

Hippocampal anatomical variants were assessed according to N. Bernasconi criteria [9]. We performed visual analysis of hippocampal shape and measurement of parahippocampal gyrus angle (for the assessment of vertical orientation of the hippocampus), and the distance between the third ventricle and fimbria (for the assessment of medial position).

At next stage, 3D MPRAGE sequence was obtained using the following parameters: TR = 12, TE = 5, FOV = 25.6, MTX = 256, ST = 2.0, FA = 20. For evaluation of brain structures volume we performed post-processing and voxel volumetry was obtained automatically (software environment FreeSurfer), partly automatically and manually (program package DISPLAY) [6].

FreeSurfer- is a program package initially developed only for the segmentation of brain cortex, however, was later upgraded to full value instrumental segmentation and visualization of structural and functional elements (FreeSurfer/Massachusetts General Hospital. URL: http://surfer.nmr.harvard.edu). We used FreeSurfer 6.0 also for accurate analysis of subregions and subfields hippocampal formation.

MINI-Display represents a program, developed for object visualization and manipulations in three dimensions, generally for cortex surfaces. This program is able to reflect and to segment MRI, PET and other imaging modalities, and has many additional functions. User interface of the MINI-Display in nonconventional and menu-orientated system is based on the keypress and mouse manipulation (Montreal Neurological Institute, Quebec, Canada).

Results

Mean height of hippocampus in healthy volunteers equals 8.56 mm at head level, 6.34 mm at body level and 5.12 mm at tail level.

Hippocampal volumes of healthy volunteers measured using voxel morphometry with FreeSurfer post-processing program are represented in Table 1.

We have not found significant age correlation of the hippocampal volumes in healthy volunteers aged 18-50 years. But we did have indicated significant age correlation, that is, right and left hippocampus in healthy men was remarkably higher than in healthy women with high confidence (Table 2).

Table 1. Hippocampal volumes of healthy volunteers measured using voxel morphometry with FreeSurfer post-processing program

Structure Volume (mm3)

Right Left

Hippocampus 4297.08 ± 413 4403.29 ± 191

Amygdala 1577±206 1504±171

Table 2. Gender correlation of hippocampal volumes

Gender Right hippocampus p Left hippocampus p

Men 4584.41 ± 316.625 0.000 4631.56 ± 280.232 0.000

Women 4163.83 ± 385.626 4297.42 ± 388.50

Different anatomical variants were observed in 43.5 % of cases in the following proportions: choroidal fissure asymmetry was found in 7.9 % of healthy volunteers (8); deep vertical collateral fissure was observed in 8.9 %.

Different morphological variants were observed in 43.5 % cases in the following proportions: choroidal fissure asymmetry — 7.9 % (8); deep vertical collateral fissure — 8.9 % (9); asymmetry of the poles in temporal horns of lateral ventricles — 8.9 % (9); round shape of hippocampus solely — 8.9 % (9); round shape associated with deep vertical collateral sulcus — 8.9 % (9).

Pair-wise comparison proved that the volume of the round shape hippocampus is reliably lower than the volume of the hippocampus with typical morphology.

Comparative analysis with Mann-Whitney criteria of young and old volunteers groups determined significant differences in hippocampus fissure volumes, parasubicu-lum, molecular layer of dentate gyrus, fimbria, CA3 and CA4 Brodmann areas.

In case of old age volunteers, volume of hippocampal fissure was larger on 28 % in right and 27 % in left hemisphere.

But volumes of CA3 and C4 fissure were larger on 14 % and 13 % in right and 9 % and 4 % in left hemisphere, respectively, in volunteers of old age, than in young volunteers.

Also molecular layer of dentate gyrus was higher by 13 % in right and 4 % in left hemisphere in volunteers of old age, than in young volunteers.

Volume of parasubiculum was larger on 12 % in right and on 3 % in left side, volume of fimbria — on 3 % in right and 11 % in left side in case of young age volunteers.

Discussion

Tremendous development of science requires more detailed neuro morphological study of the brain structures with due account for individual variability. Nowadays neuro-imaging methods, such as magnetic resonance imaging (MRI) of human brain allow accurate evaluation of different brain structures. However, in spite of the 25 years of experience using MRI for the identification of structural changes of human brain, many questions of morphometry and anatomical variants of different brain parts remain unclear. This also relates to limbic system and to mediobasal parts of temporal lobes in particular.

Therefore, detailed study of the hippocampal development is highly important for the understanding of normal and pathological hippocampal morphology.

Hippocampus obtains curved shape during early ontogenesis. Before the 10-th week of prenatal development dentate gyrus and cornu Ammonis represented rudimentary structures located in line one by one along posteromedial wall of lateral ventricle [8]. Its anterior margin is located close to medial part of perforated substance. From the 10th week dentate gyrus thickens was followed by formation of hippocampal fissure between the dentate gyrus and cornu Ammonis. On the 12-14-th weeks dental gyrus thickening leads to its rotation towards cornu Ammonis, hippocampal fissure becomes deeper and more differentiated, and orientates between the walls of cornu Ammonis and parahip-pocampal region (which includes parahippocampal gyrus and subiculum). Due to growth of dentate gyrus, medial surface of the hemisphere starts to press on lateral ventricle. In the brain of three-month fetus, opened from the lateral side, one can observe that hippocampus arches from the rostral and dorsal sides of the interventricular foramen. Further hippocampus turns back and forms ventromedial part of the pallidum. Along its ventral border vascular fissure is developing, containing vessels forming choroid plexus of lateral ventricle. At 15-16 week of the pregnancy this fissure appears thinner and to 18-21 weeks of pregnancy it obliterates together with pia mater and small vessels. However, residual cavities can remain, which can be visualized on MRI as hippocampal cystic lesions. Hippocampus at 24-th week appears the same as in the brain of the adult.

Some authors consider that abnormality of the hippocampal formation during prenatal development is associated with hippocampal anatomical variants [9]. N. Bernasconi distinguished following variants of hippocampal morphology:

— medial position in respect to the temporal horn of lateral ventricle (in this case anterior part of hippocampus is located close to cerebral peduncle (on the level of hippocampal head and anterior part of the body), and posterior part of hippocampus is located close to quadrigeminal (posterior part of the body and tail);

— round (globular) shape and vertical orientation of hippocampus;

— empty choroid fissure;

— dystopia of fimbria;

— deep and vertical collateral fissure;

— bulging of the collateral fissure towards the empty choroidal fissure; in appropriate location of subiculum;

— partial reduction of parahippocampal gyrus (our data submitted in Fig. 1, N. Ananyeva).

abc

Fig. 1. Anatomical variants of mediobasal parts of temporal lobe: round shape of hippocampi (a), empty choroidal fissure (b), vertical collateral fissure (c). (Ananyeva N.)

a b

Fig. 2. Medial position (a) and vertical orientation (b) of hippocampus on coronal images on the level of hippocampal body. (Ananyeva N.)

Medial position was estimated as distance between the midline and fimbria, which forms medial border of the hippocampus. Taking into account possible differences of the individuals this measurement was compared with the distance between the midline and the border of the temporal lobe (our data submitted in Fig. 2a, N. Ananyeva).

Vertical orientation of the collateral fissure was determined by measuring of parahip-pocampal gyrus angle (angle between ascend and descent parts of parahippocampal gyrus (our data submitted in Fig. 2b, N. Ananyeva)).

When hippocampus is normally orientated, parahippocampal gyrus angle approaches 180°. Anatomical variants of hippocampus and round shape in particular are highly debatable item in literature. Initially a term "malrotation" was proposed. However, nowadays term "malrotation" is considered disputable and term incomplete hippocampal inversion is preferred, because it implicates incomplete inversion of hippocampus during prenatal period [10].

Moreover, some authors consider this anatomical variant only as a combination of the key criteria (round shape) with other abnormalities, such as vertical collateral fissure, unilateral involvement of the whole hippocampus, change of fornix position, enlargement of temporal horn of lateral ventricle with no signs of hippocampal MR signal abnormalities or brain congenital anomalies [10].

Other authors proposed to assess by incomplete criteria. Then incomplete hippocampal inversion is defined in 18-37 % of healthy volunteers [7; 11]. According to our studies, analyzing solely round shape of the hippocampus, it was observed in 20.4 % of examined persons. Combined round shape of the hippocampus was observed in 20.4 % of healthy volunteers. However, it is important to add that we have indicated only a combination of the round shape of hippocampus and vertical collateral sulcus, but did not observe all of the criteria proposed by P. Barsi [10].

There are some studies reporting that the volume of left hippocampus is smaller than the right, however the volume difference is insignificant. In our study we did not received such findings [11].

Fig. 3. Identification of hippocampal head, body and tail (N. Malykhin)

Changes of the hippocampal volume can be seen in many types of the brain pathology, however proximity of these structures makes them hard-separable on MR images, as a result there is a risk of few slices exclusion from the analysis when performing MR morphometry. Conversely, combined analysis of amigdalo-hippocampal complex can be performed. Beyond that point differentiation of hippocampal tail and thalamus can be challenging. Therefore, appeared a need in reliable method of hippocampus and amigdala volume analysis (DISPLAY, Montreal Neurological Institute, Quebec, Canada), that is also used for the evaluation of intracranial volume and performing 3-dimensional reconstructions [6].

Hippocampal analysis should be performed using coronal images starting from the hippocampal tail [6]. It can be difficult to differentiate the borders of the hippocampal regions. The border between the tail and body can be seen on coronal images where the columns of fornix are fully visualized. The border between the body and the head can be seen on images where the top of hippocampal hook is clearly visualized. The most important structure that can be used for lateral, anterior and inferior borders identification of hippocampus is falciform recess of inferior horn of lateral ventricle (Fig. 3).

Volumes of hippocampal regions in healthy volunteers measured using post-processing software DISPLAY are shown in the Table 3.

Hippocampus is a part of hippocampal formation, which include also dentategyrus, subiculum, presubiculum, entorhinal cortex. Hippocampus (cornu Ammonis) is a tight

type of cells, which is extended in anterior-posterior direction along medial wall of inferior horn of lateral ventricle [12]. Main nervous cells of hippocampus include pyramidal neurons and polymorphic cells. As ancient cortex hippocampus consists of 3 main layers: stratumoriens, stratumpyramidale and stratumradiatum + stratumlacunosum-moleculare (Fig. 4).

Table 3. Examination of subfields and subregions of the hippocampal in young and old age

Structure Young healthy volunteers Old healthy volunteers Level of significance

Left hemisphere Right hemisphere Left hemisphere Right hemisphere

Whole hippocampus 3350.38211 3256.182089 3445.861117 3414.173636

Hippocampus tail 561.9416047 544.6546328 541.5249371 527.6189509

Subiculum 402.64394 381.5967118 435.7437506 415.3908504

CA1 584.0629195 599.7761962 621.8925157 627.5692734

Hippocampal fissure 168.64971 164.4610933 213.692648 211.3207803 P < 0.01

Presubiculum 342.3967863 318.7739042 330.956494 301.8749004

Parasubiculum 67.755132 62.355125 68.42057929 55.22181486 P < 0.05

Molecular layer of the hippocampus 537.55853 521.2902828 562.1268349 559.7053719

Molecular layer of dentate gyrus 292.0825065 276.5951105 303.3211026 312.4577006 P < 0.05

CA3 191.4176447 193.1031418 209.4775464 221.8378734 P < 0.05

CA4 250.6389882 236.3428308 261.986464 267.5831651 P < 0.05

Fimbria 66.46761617 67.28637933 54.18618571 65.40572986 P < 0.01

HATA 54.4164435 54.407774 56.22470671 59.50800443

Layer on ventral surface, alveus, consisting of myelinated axons of pyramidal neurons, basal dendrites and initial segments of axons, located in polymorphic layer. They are followed by pyramidal neurons layer, then stratum-radiatum, consisting of trunks of apical dendrites and stratum-lacunosum-moleculare, consisting of pre-terminal and terminal branching of apical dendrites. Such organization of hippocampus exists on all his frontocaudal extent (laminar hippocampal organization) [6; 12].

Specificity of hippocampal organization of pyramidal layer is the basis for its dividing on 4 main areas; СА1 — СА4. Main areas of hippocampus — СА1 and СА3. СА1 area consists of 2 small tight layers of pyramidal neurons, large neurons of CA3 and axons of pyramidal neurons of CA3 named "Shaffar collaterals", contacting with apical dendrites of CA1 area. These connections are 2 main associative paths, connecting its main elements. So, we can review hippocampus as a package of morphofunctional segments series, which can function relatively independent. And CA1-CA3 area — convergence point of information traffic from associative cortex and phylogenetic ancient parts of the brain.

In our research we found out significant differences in volumes of hippocampus fissure, parasubiculum, molecular layer of dentate gyrus, fimbria, СА3 and СА4 Brodmann areas, which demonstrate that in adulthood morphofunctional connections are not finally formed, that's why volumes of molecular layers CA1-CA3 smaller in adulthood than in

Medial Corpus geniculatum laterale

(латеральное коленчатое тело)

Lateral

Plexus Choroideus Cauda nuclei

caudate Плотная

Fimbria hippocampi Gurus dentatus

Presubiculum /ч

Ствол мозга

Eminentia collateralis

Рыхлая нейрональная полоска Аммониева рога

> # < Сектор Сомьера

Шестислойная кора обонятельного мозга

Fig. 4. Scheme of hippocampus [13]

old population. But hippocampal fissure become smaller in older ages because of atrophic changes.

Table 4. Voxel morphometry (DISPLAY). Hippocampal volumes in healthy volunteers.

Hippocampal region Right hippocampus Left hippocampus

Head 2314±526 2484 ± 526

Body 1052±193 987±210

Tail 326±125 304 ± 79

Conclusion

Neuroimaging methods are permanently upgrading and improving. This allow to feel reliance in the further success in diagnosis and understanding of different types the genesis in brain pathology that can be reached using these methods in the nearest future [14].

However, in spite of the great achievements and advantages of the instrumental methods of the inter vivo brain studies many questions of the special aspects of morphology and anatomical variants of the mediobasal part of temporal lobe remain unclear.

References

1. Wasserman L. I., Ananyeva N. I., Gorelik A. L., Ezhova R. V. et al. Affective-cognitive disorders: methodology of analysis of the structural-functional correlations using the model of temporal epilepsy. Bulletin of the South-Urals State University. Seria: Psychology, 2013, vol. 6, iss. 1, pp. 67-71. (In Russian)

2. Ivanov M. V., Ananyeva N. I., Wasserman L. I., et al. Complex diagnosis of endogeneous depressive disorders using neuroimaging and neurocognitive correlations. Obozrenie psihiatrii i medicinskoi psihologii im. V. M. Bekhtereva, 2014, no. 2, pp. 39-44. (In Russian)

3. Kissin M. Ya., Ananyeva N. I., Shmeleva L. M. et al. Neuromorphological features of anxious and depressive disorders in temporal lobe epilepsy. Obozrenie psihiatrii i medicinskoi psihologii im. V. M. Bekhtereva, 2012, no. 2, pp. 11-17. (In Russian)

4. Trofimova T. N., Ananyeva N. I., Nazinkina Yu.V., Karpenko A. K., Khalikov A. D. Neuroradiology. Ed. by T. N. Trofimova. St. Petersburg, SPbMAPO Publ., 2005. (In Russian)

5. Trofimova T. N., Medvedev Yu. A., Ananyeva N. I. et al. Using of the postmortem MRI of the brain in pathomorphological analysis. Arkhiv patologii, 2008, vol. 70, iss. 3, pp. 23-28. (In Russian)

6. Malykhin N. et al. Three-dimensional volumetric analysis and reconstruction of amygdala and hippocampal head, body and tail. Psychiatry Research: Neuroimaging, 2007, vol. 155, iss. 2, pp. 155-165.

7. Humphrey T. The development of the human hippocampal fissure. J. Anat., 1967, vol. 101, pp. 655-676.

8. McLean J. The investigation of hippocampal and hippocampal subfield volumetry, morphology and metabolites using 3TMRI. Thesis for the degree of PhD. Glasgow, University Glasgow, 2012. 354 p.

9. Bernasconi N., Kinay D., Andermann F. et. al. Analysis of shape and positioning of the hippocam-pal formation: an MRI study in patients with partial epilepsy and healthy controls. Brain, 2005, vol. 128 (Pt 10), pp. 2442-2452.

10. Barsi P., Kenéz J., Solymosi D., Kulin A. et al. Hippocampal malrotation with normal corpus callosum: a new entity? Neuroradiology, 2000, vol. 42, iss. 5, pp. 339-345.

11. Raininko R., Bajic D. "Hippocampal malrotation": No real malrotation and not rare. AJNR Am. J. Neuroradiology, 2010, vol. 31, iss. 4, p. 39.

12. Afanas'ev Yu.I., Yurina N. A., Kotovsky E. F. Histology, embriology, citology. 6th ed. Minsk, 2006. 786 p.

13. Duus P. Topical diagnosis in neurology. Thieme Stuttgart. 4th edition. New York, 2005. 531 p.

14. Bogdanov A., Degtyarev A., Guschanskiy D., Lysov K., Ananyeva N., Zalutskaya N., Neznanov N. Analog-digital approach in human brain modeling. 201717th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing, CCGRID 2017. Bk: Proceedings, 2017, no. 17, pp. 807-812.

Received: April 17, 2019 Accepted: November 6, 2019

Authors' information:

Natalia I. Ananyeva — MD, PhD, Full Prof; [email protected]

Evgeniy V. Andreev — [email protected]

Linara R. Akhmerova — [email protected]

Tatiana A. Salomatina — [email protected]

LudvigI. Wasserman — MD, PhD, Full Prof.; [email protected]

Ruslana V. Grebenshchikova — MD; [email protected]

Nikolay G. Neznanov — MD, PhD, Prof.; [email protected]

Alexey A. Pichikov — MD, PhD; [email protected]

Natalya M. Zalutskaya — MD, PhD; [email protected]