Delineation of intermammillary relationships using magnetic resonance imaging

No study has investigated intermammillary relationships using neuroimaging modalities. This study aimed to explore them using magnetic resonance imaging (MRI). We enrolled 72 patients who underwent conventional MRI examinations, followed by constructive interference steady-state sequence in the coronal plane. The intermammillary distances (IMDs) were measured at the uppermost level of the intermammillary gap (IMDupp) and the lowest level (IMDlow) of the mammillary bodies (MBs). MBs with varying morphologies were consistently delineated. The appearance of both MBs could be classified into four patterns based on the size and relative levels, with the symmetrical type being the most common. Intermammillary relationships exhibited five patterns. In 69%, the IMDupp was discernible and measured 0.7 ± 0.4 mm, while it was not discernible in 31% due to the presence of intermammillary connection and adhesion. The age distribution did not differ between populations with and without discernible IMDupp. The IMDlow was measured 4.4 ± 0.9 mm. Although the IMDlow was not significantly different between both sexes; it was longer in subjects in their 70s. Intermammillary relationships show variable morphologies with gaps formed between both MBs. The IMDlow may become more evident in association with age-related increase in the width of the third ventricle and atrophy of the MBs.


Introduction
Mammillary bodies (MBs) are small, paired neural structures that protrude inferiorly under the floor of the third ventricle ( Fig. 1). Together with the hippocampus and fornix, these paired structures form the Papez circuit, further connecting with the limbic system [13]. Moreover, an MB projects into the thalamus as the mammillothalamic tract [16]. In endoscopic third ventriculostomy, MBs are important landmarks for perforating the ventricular floor [1,15]. Despite their small size, MBs play an important role in supporting recollective memory [5]. Damage to the anterior thalamic nucleus and mammillothalamic tract can trigger transneuronal degeneration of the ipsilateral MB, resulting in atrophy and asymmetry [4]. Despite extensive research, MBs are still not well understood. Previous investigations using magnetic resonance imaging (MRI) have documented the morphological variability of MBs [6,14]. Furthermore, MBs and the mammillothalamic tract have been suggested to show age-related conspicuity [2,10].
The human brain, involving the cerebral hemisphere and brainstem, is thought to shrink during formalin fixation [9,12].
Previous studies with cadaver specimens indicated that the intermammillary distance (IMD) could be identified in less than 15% of subjects [3,11]. However, to our knowledge, we are missing information and measurements of the intermammillary relationships using neuroimaging. The present study therefore aimed to characterize them using MRI.

Materials and methods
This retrospective study included 72 outpatients who had undergone MRI examinations following presentation to our hospital between February 2010 and December 2011.

3
These patients experienced headaches, dizziness, tinnitus, hearing disturbances, hemisensory disturbances, scintillation scotoma, and seizures. The patient population comprised 40 men and 32 women, with a mean age of 49 years (range, 9-78 years). Patients with a history of meningitis or encephalitis were excluded. Initial examinations using axial T1-and T2-weighted sequences, T2-gradient echo, fluid-attenuated inversion recovery, and diffusion-weighted sequences confirmed that none of the patients had cerebral infarction or hemorrhage, brain tumors, hydrocephalus, intracranial cysts, traumatic brain injury, considerable brain atrophy, or mesial temporal sclerosis. Subsequently, 72 patients underwent thin-sliced, constructive interference steady-state (CISS) sequence in the coronal plane involving the entire cerebral hemisphere. The CISS sequence is a gradient-echo MRI used to investigate a wide range of pathologies when routine MRI sequences do not provide adequate anatomical information. The increased sensitivity of the CISS sequence occurs as an outcome of the accentuation of T2 values between cerebrospinal fluid and pathological structures. The parameters for the CISS sequence were as follows: repetition time, 2000 ms; echo time, 311 ms; slice thickness: 2.00 mm, interslice gap, 0 mm; matrix, 320 × 274; field of view, 160 × 160 × 100 mm; flip angle, 90°; and scan duration, 5 min 30 s. All images were obtained using a 3.0-T MR scanner (Achieva R2.6; Philips Medical Systems, Best, The Netherlands). The imaging data were transferred to a workstation (Virtual Place Lexus 64. 64th edition; AZE, Tokyo, Japan) and independently analyzed by two of the authors (S.T. and H.I., both of whom have more than 15 years of experience as board-certified neurosurgeons). The MBs and IMDs were assessed on CISS images, where both MBs showed the largest dimensions. The IMDs were measured at the uppermost level (IMD upp ) of the intermammillary gap and the lowest level (IMD low ) of both MBs (Fig. 2). The Wilcoxon signed-rank test was used for statistical analyses. Differences were considered statistically not significant at p > 0.05.
This study was conducted in accordance with the guidelines of our institution for human research. Written informed consent was obtained from all patients prior to their participation.

Results
The MBs were consistently delineated in all 72 patients, revealing varying morphologies. The appearance of MB pairs could be classified into four types based on their size and relative levels: symmetrical, right-side dominant, leftside dominant, and uneven. The uneven type indicated MBs located at different levels (Fig. 3). The symmetrical type was the most frequent and was found in 40 (56%) patients, followed by the right-side dominant, left-side dominant, and uneven types, which were identified in 14 (19%), 14 (19%), and 4 (6%) patients, respectively (Table 1). In 50 patients (69%), the IMD upp was discernible and measured  (Fig. 4). Types a, b, c, d, and e accounted 17 of 72 (24%), 5 (7%), 25 (35%), 10 (14%), and 15 (21%) patients, respectively (Table 2). In the Wilcoxon signedrank test, the age distribution showed no difference between the populations with (mean age: 50 ± 17 years, range 13-78 years) and without (mean age: 44 ± 19 years, range 9-73 years) discernible IMD upp (p > 0.05). The measured IMD low was 4.4 ± 0.9 mm (range: 3.0-7.6 mm). Statistically, the length of the IMD low did not differ between both sexes (p > 0.05), but was significantly longer in those in their 70s (0.01 < p < 0.05) compared to the population younger than 20 years. Other age groups did not show any significant differences from the control population (Table 3). The frequency of type e MBs and separate MBs showed no sex difference (men, N = 7, mean age: 50 ± 24 years; women, N = 8, mean age: 48 ± 15 years, p > 0.10). Furthermore, in five patients (7%), the MB was considerably compressed by the cerebral arteries. These patients comprised three men and two women with a mean age of 50 years (range 29-62 years). The offending vessel was the left posterior cerebral artery in four patients, and the basilar artery in one patient (Fig. 5). No temporal sclerosis was found in any of these five patients.

Discussion
The IMD has been reported to be an infrequent presentation [3,11]. However, in the present study, IMDs were identified in nearly 70% of examined patients. In contrast to previous investigations with conventional MRI sequences, the present study adopted the CISS sequence, which might partially facilitate observations of the intermammillary relationships [6,14]. Furthermore, MBs are fine structures arising from the inferior wall of the cerebrospinal fluid-filled third ventricle. Therefore, following formalin fixation and efflux of the intraventricular fluid, MBs and IMDs can exhibit a degree of shrinkage and become difficult to identify in cadaver specimens [9,12].
In this study, MB pairs exhibited four patterns, with the symmetrical type being the most frequent. In contrast to a previous study documenting right-side dominance of the mammillothalamic tract, in this study, laterality was not observed in MBs [2]. The outcome needs to be further verified in a sufficiently large population.
Observations of the IMCs and IMAs revealed five distinct patterns of intermammillary relationships. The frequency of discernible IMD upp was not influenced by patient age, suggesting that IMD upp may be a consistent gap formed between normal MBs. Conversely, the IMD low was significantly longer in patients older than 71 years. We assumed that this might have resulted primarily from age-related MB atrophy and an increase in the width of the third ventricle floor in response to brain atrophy [2,8,10]. The location of the upper MBs is recognized as an important landmark during endoscopic third ventriculostomy. Also, the neuroanatomy of the MBs has been explored in association with  hypothalamic hamartomas frequently presenting with gelastic epilepsy [1,7,15]. Despite extensive studies, the pathomechanism of gelastic epilepsy remains poorly understood [7]. The IMCs found in this study have not been documented thus far, and should therefore be verified by further exploration with careful dissection. The present study had several limitations. First, the cohort comprised participants of heterogenous ages and sexes. The participants were retrospectively evaluated and not randomly assigned to the MRI examinations. The MBs and IMDs were assessed only on the coronal CISS images. The measurements on the CISS sequence were quite short, and there may be a risk of error when delineating the MB contour. Furthermore, morphological variability of the MBs and IMDs was not examined using an appropriate neuropsychological battery in relation to patients' cognitive and memory functions. In this study, the MB was   considerably compressed by the cerebral arteries in five of the 72 patients. The pathological implications of such radiological findings need to be explored in future studies. Despite these limitations, we believe that our results provide valuable insight into the understanding of MBs.

Conclusions
Intermammillary relationships show variable morphologies with gaps formed between both MBs. The IMD low may become more evident in association with age-related increase in the width of the third ventricle and atrophy of the MBs.
Author contributions ST conceived the study, performed the cadaver dissection, and wrote the manuscript. NS and HU collected the imaging data. ST and HI analyzed the imaging data.
Funding No funding was received for this study.
Availability of data and materials The data and materials used in this study are available from the corresponding author on reasonable request (Satoshi Tsutsumi).

Conflict of interest
The authors declare that they have no conflicts of interest of a financial or personal nature.
Ethical approval All procedures in this study were performed in accordance with the ethical standards of the institutional and/or national research committee, and the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards.
Informed consent Written informed consent was obtained from all participants included in the study for participation and publication of the article.