Three-Dimensional Microscopic Characteristics of the Human Uterine Cervix Evaluated by Microtomography.

Despite its seemingly simple anatomy, the uterine cervix (UC) has complex functions throughout a woman’s life. From the isolation of the uterine cavity from vaginal microorganisms beginning in the prepubertal period, to sexual functions that can be extended after menopause, the UC is an organ that connects two different environments. To fulfill its functions, the UC changes according to the age of the woman, the cyclical hormones, and the presence of an implanted ovum [ 1 ]. The anatomical study of this simple, multifunctional organ has been documented since the second century when it was called the neck of the uterus by Soranus [ 2 ]. Over the centuries, the anatomy of the UC has been iconographically represented by drawings, as well as macroscopic, and microscopic studies. In the 20th century, three-dimensional (3D) medical imaging provided a spatial perspective of the anatomy of the female genital tract, from in vivo and ex vivo specimens [ 3 , 4 ]. However, artifacts and low-resolution images reduce the possibility of adequate perceiving the spatial relationship of different tissues on the cervix [ 3 ]. Cervical cancer is the fourth most common cancer among women worldwide [ 5 ], and 85% of cervical cancers occur in women living in low– to middle-income countries [ 6 ]. The origin of most cervical cancers occurs during the transformation of intracervical epithelium in extracervical epithelium, on the squamous-columnar junction [ 7 ]. Recently, human ex vivo specimens have been studied using microtomography (microCT), which allows the visualization of human organs in 3D and improves the perception of 3D relationships between tissues at the microscopic scale [ 8 ]. The purpose of this study is to analyze the three-dimensional anatomy of the human UC at a microscopic scale, the characteristics of the external and internal cervical epithelium, and the effects of metaplastic transformation on both the epithelium and squamocolumnar junction. Additionally, this study aims to evaluate the tridimensional characteristics of the cervical stroma, and its relation to internal and external os, analyzed by two-dimensional (2D) and 3D images provided by microCT.

This analysis is part of a study using microCT to image human tissues and organs, approved by the Ethics Committee of University of Vassouras (CAAE 56031916.0.0000.5290). A prospective case selection was performed and patients undergoing hysterectomy for benign gynecological conditions (uterine fibroids, adenomyosis, and abnormal uterine bleeding) were included in this study. Patients were invited to participate in this study and signed an informed consent form. Specimens were fixed in 10% formalin solution for more than 24 h at room temperature to ensure proper preservation of cellular and tissue architecture [ 9 ]. The specimens were washed twice with water and, for contrast enhancement between different soft tissues, the specimens were immersed in a 10% formalin and 10% Lugol solution diluted in physiological solution for more than 72 h. The specimens were washed of staining solution to reduce excessive saturation on the surface. The specimens were fixed on Styrofoam for support during image acquisition to reduce artifacts. In the course of the imaging capture, the specimen was rotated on its biggest axis, thus allowing capture in different angles. After image acquisition, the specimens were immersed in 10% formalin solution to remove the iodine concentration and were submitted to traditional histologic studies. The General Electric Phoenix VTomeX (General Electric; Wunstorf, Germany) was used for imaging acquisition. The parameters for acquisition, including energy, current, and exposure time, were adapted for each specimen for better contrast between the cervical soft tissues. The image acquisition parameters were energy between 50 and 130 kV (median 98 kV), exposure time between 220 and 480 ms (median 333 ms), current between 220 and 480 μA (median 307 μA), scan duration between 44 and 72 min (median 56 min), and voxel size between 22 and 39 μm (median 32 μm) ( Table 1 ). After the capture of the images of the specimens, the software NRECON version 1.7.4.5 (Bruker; Kontich, Belgium) was used for gathering the 2D X-ray images in a volumetric image for 3D analysis. With the volumetric image formed, the multiplanar analysis of the 3D volumetric image demonstrating the 2D appearance of the specimens in all the 3 planes was performed using the DataViewer (version 1.5.6.2 64-bit 2017; Bruker; Kontich, Belgium). The analysis and 3D image reconstruction were performed using CTVox version 1.16 (Bruker; Kontich, Belgium).

A total of 10 human uterine cervixes from 10 patients who agreed to participate in this study were analyzed. The age of the patients ranged from 39 to 48 years (median 45 years). Demographic information of the patients is presented in Table S1 . In nine cases, the uterine cervixes were anatomically normal. In one case, however, mechanical compression and distortion of the cervix caused by multiple Naboth cysts were observed. Surgery was suggested due to a lack of clinical response to conservative treatment and compressive effects of the enlarged uterus due to fibroids. All cervixes were removed through abdominal incisions after total hysterectomy. The extrafascial technique was used in all cases. In all specimens, the uterine cervix presented the vaginal segment (ectocervix), the entire abdominal part of the cervixes, including internal os. Laterally, the cervixes were removed, preserving the parametrium to avoid intra- and post-operative complications. In some specimens, the uterine segment was also examined. For proper contrast between different soft tissues, the iodine solution must penetrate the tissue throughout its volume. In eight specimens, the contrast was able to adequately penetrate the UC. In two specimens, the contrast was unable to differentiate the medial segment of the subepithelial stroma. Sufficient impregnation of contrast in the specimens allowed the study of the microscopic anatomy of the UC in 2D and 3D images. The 2D images could be viewed in any selected plane and in any direction. Using the DataViewer software (version 1.5.6.2 64-bit 2017; Bruker; Kontich, Belgium), 2D images of specimens could be virtually sliced in any direction with simultaneous visualization of the specimen in different planes. The section of interest could also be visualized in all three planes simultaneously ( Figure S1 ). The contrast of the images could be adjusted to better visualize the region of interest for the observer. The 3D images were generated using the CTVox software version 1.16 (Bruker; Kontich, Belgium). The UC is the lowest part of the uterus and has the histologic characteristics of a transition from a fibromuscular structure in the cranial part adjacent to the uterine segment to a fibrous structure in the portio vaginalis. The UC is divided into two parts: the supravaginal part, lying above the vagina in the abdominal segment, and the vaginal part protruding into the vagina [ 10 ]. The stratum corneum can be seen in 2D microCT images as the external part of the endocervix. This feature also correlates with the traditional microscopic image in 2D ( Figure 1 ). Similar to in vivo specimens, iodine was more intensely present in this part of the cervix, more radiopaque in the microCT X-ray image. However, the limitation of microCT becomes apparent when it is noticed that the scale of the image pixels generated by microCT cannot show the images in the same scale as the traditional optical microscopy. In addition, the X-ray images produce images according to the X-ray opacity of the tissue and its ability to absorb contrast. Metaplasia is the transformation of cells or tissues into other cells or tissues. Transformation of mature endocervical epithelium into mature squamous epithelial cells is a physiological process of the UC when columnar epithelium is exposed to the vaginal environment. After the transformation of the columnar epithelium, the metaplastic areas appear as normal ectocervical squamous epithelium. This transformation can be seen by sectioning the UC when the residual crypts form the Naboth cysts. The Naboth cysts were easily visualized in 2D and 3D images obtained by microCT. In one specimen, the presence of multiple Naboth cysts distorted the anatomy of the cervical canal. The cyst pushed the endocervical epithelium caudally, creating an irregular cavity in the cervical lumen ( Figure 2 ). The metaplastic process was well demonstrated in all ten specimens. The transformation zone could be visualized in 2D and 3D images, and the border between the squamous and glandular epithelium was clearly demonstrated. The squamocolumnar junction was well visualized, especially in the 3D images. Small round openings on the surface of the squamous epithelium demonstrated the opening of the endocervical epithelium under the epithelium under metaplasia. The specimen presented the main features of UC at colposcopy, with the squamocolumnar junction completely visible (type 1), partially visible (type 2), and not visible (type 3) ( Figure 3 ). The 3D images of microCT allowed the sectioning of the UC in all planes, allowing the visualization of the squamocolumnar junction in 2D or 3D images ( Figure 4 ). The endocervical epithelium showed different characteristics in the specimens, mostly influenced by the metaplastic transformation. MicroCT images allowed the 3D visualization of the different types of squamocolumnar junctions and the corresponding 3D anatomy of the endocervical epithelium. The epithelium presenting outside the external os has a “grape” appearance. This appearance is created by the compressed folds of the epithelium, forcing the epithelium in a parallel direction with its outermost extremity of polypoid epithelium, creating the “grape” appearance visualized in colposcopic examinations (arrow, Figure 5 A). The epithelial folds represent a thick portion of the external os (arrowhead, Figure 5 A). This interesting pattern of parallel folds can be seen as metaplasia progresses. However, there is a significant reduction in the number of epithelial folds and their thickness ( Figure 5 B). The epithelium becomes scarce, and the stroma occupies the external os of the UC ( Figure 5 C). In the final stage of metaplasia, there is no visible cervical endothelium on colposcopic examination. The portio vaginalis is completely covered by metaplastic squamous epithelium. The columnar epithelium lost its thickness and some folds with a cobblestone appearance were visible in 2D and 3D images ( Figure 5 D). The specimens showed different patterns of endocervical canal folding and consequently different sizes and complexities in the lumen. The natural history of trauma secondary to labor and natural epithelial obstruction with cystic formations can be demonstrated by the deformation of the luminal space in some specimens ( Video S1 ). In other specimens, however, the presence of cysts in the cervical epithelium, even larger cysts, had little effect on the cervical lumen and, most importantly, no apparent change in the cervical folds ( Figure 6 , Video S2 ). The cervical mucus could be observed in some specimens. The thick and dense mucus filled the glands and remained inside the cervix even after the period of immersion in formalin and iodine solution. The mucus did not absorb iodine due to its characteristic water concentration (95–99%). The presence of mucus created an interface between the lumen and the epithelium, facilitating the 3D visualization of the structural appearance of the cervical epithelium from the luminal perspective, without the necessity of mechanical dilatation. The epithelium presents crypts on 2D images, following a regular pattern of size and spacing between epithelial layers. The crypts tend to be perpendicular to the lumen or have their layers facing the external os. The size of the epithelial crypts is continuous, decreasing in size with proximity to the internal os. The 3D images of the epithelial lining inside the endocervix show an unpredictable and complex pattern. The gathering of the epithelial lining creates folds that have their greatest length parallel to the cervical axis. The folds have their axis toward the exocervix ( Figure 7 ) and are reduced in size near the internal os, where parallel folds can be seen, and become sparse and thin in advanced metaplastic cervixes. The internal os is characterized by a subtle transition between uterine tissues, with a complex endocervical epithelium becoming more organized and predictable to a thin endometrial gland that is smaller in size compared to the endocervical epithelium ( Figure 8 ). No other differences were observed in the subepithelial stroma. The Video S3 shows the main anatomical features of the UC. Using 2D images, the cervical stroma shows an alteration in the orientation of the directions of the collagen fibers. The stromal fibers appear to be parallel to the endocervical lumen along their entire length. However, in the internal os, there is an apparent change to a perpendicular direction ( Figure 9 ).

This case series demonstrates aspects of human UC anatomy at the microscopic level, in 2D and 3D images provided by microCT. Throughout the last century, other X-ray technologies have imaged the cervix in vivo, but the improvement of soft tissue contrast in ex vivo biological specimens has allowed the acquisition of cervical images with high resolution [ 11 ]. MicroCT is a non-destructive technology, allowing the acquisition of images without destroying specimens for further histologic study [ 12 , 13 ]. The endocervix produces mucus at rate of 20–60 mg/day in reproductive age, and this production increases 10-fold in the preovulatory phase of the cycle, reaching 700 mg/day [ 14 ]. Traditional studies consider the presence of 100 “mucus-secreting glandular-like units” in the cervical canal [ 15 ]. In this series, the 3D images allowed the visualization of the crypts with details due to the presence of mucus. Also, the traditional studies for estimating the number of crypts were based on 2D histologic slides, which made the estimation of crypts inaccurate. Figure 10 shows the spatial distribution of glands in a 2D image. The connective tissue of the cervical stroma is divided into two segments: the superficial zone and the subepithelial stroma, which is rich in interstitial fluid and may be infiltrated by adjacent inflammatory exudates. The superficial zone is deeper and has a dense collagenous layer [ 16 , 17 , 18 ]. The possibility of visualizing the endocervical epithelium during colposcopic examination with optical instruments such as hysteroscopy is well known, but these instruments require additional distance to the inner part of the epithelium [ 19 , 20 ]. Creating a true distance for visualization requires the use of mechanical force, such as forceps, or increasing the intraluminal pressure with saline solution or gas, which compresses and distorts the structures. In this series, the presence of mucus in the microtomographic image of the cervix allowed the visualization of the 3D structure of the folds and crypts of the cervical epithelium without any additional mechanical effect. MicroCT has been used for the study of human specimens since the early 1980s, but the development of methods to enhance contrast between different soft tissues of biological specimens allowed microCT to be used for the study of human soft tissues. From brain tissue [ 21 ] to fetal heart morphology [ 22 ], microCT has gained increasing interest in the three-dimensional study of microscopic anatomy. The addition of iodine solution as a soft tissue contrast agent has facilitated the specimen preparation process [ 23 ]. In human gynecological specimens, the 3D microscopic anatomy of ectopic pregnancy [ 24 ] and tubal changes [ 25 ] has been described using microCT. The normal characteristics of the microanatomy of the fallopian tube were described, with detailed anatomy of the tubal mucosa [ 12 ]. The high definition of the tubal mucosa allowed segmentation of the tubal mucosa and creation of a volumetric image of the tubal lumen, allowing virtual navigation within the normal fallopian tube [ 26 ]. The high definition of the endocervical lumen was also observed in this series case. The microtomographic images allowed visualization of the folds created by the epithelium without external effects for luminal dilatation. A study evaluating the anatomy of uterus and cervix of rats demonstrated the accuracy of microtomographic images when compared to classic histology methods. The traditional microscopy can achieve cellular and sub cellular at high-resolution when compared to microCT. However, in rats, microCT was very effective in the study of muscular fibers of the uterus [ 27 ]. In humans, the orientation of the cervical collagen–cellular fibers has been described based on 2D images for centuries, and recent studies have demonstrated the circumferential orientation of the uterine components in the internal os [ 28 , 29 ].

In summary, this study demonstrated the main anatomical features of the human UC from ex vivo specimen images provided by microCT, with special attention to the exhibition of the cervical lumen and its characteristics according to the stage of cervical metaplasia. In addition, the large field of view of microCT images allows visualization of the cervical stroma in 3D images. We believe that 3D microCT images may be an important tool for studying the orientation of cervical stroma fibers in human specimens, as recently described in animal studies [ 23 ].