Cephalometric predictors of abnormal sella turcica among individuals with Sheehan Syndrome: A case-control study

The present study aimed to analyze cephalometric predictors of abnormal sella turcica (ST) in Sheehan Syndrome (SS) patients. An observational case-control study was performed with SS volunteers from the Endocrinology and Diabetology Service of the Walter Cantídio University Hospital (Brazil). The sample was composed of 32 patients allocated in two groups: case group (16 adult women diagnosed with SS) and control group (16 healthy individuals matched by sex and age). Analyzes of the linear dimensions (length, diameter, and depth) of the ST on lateral teleradiographies were made using the Radiocef Studio 2 software. The morphological pattern (anterior oblique wall, double floor contour, bridged, irregularities of the dorsal face, pyramidal) was also assessed. The subjects mean age was 65.47 ± 10.19 years. SS patients had lower mean length (p <0.001), width (p <0.001) and height (p = 0.033) compared to the control group. The presence of morphological alteration of the ST was statistically significant (p = Research, Society and Development, v. 10, n. 5, e52210515316, 2021 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v10i5.15316 2 0.009) compared to the controls. The most frequent morphological changes were irregularities of the dorsal face (37.5%; p = 0.018), oblique anterior wall (12.5%), double floor contour (6.3%), and bridged appearance (6.3%). Our study found smaller dimensions and morphological variations of ST in Brazilian SS individuals, highlighting the importance of ST-related imaging screening.

0.009) compared to the controls. The most frequent morphological changes were irregularities of the dorsal face (37.5%; p = 0.018), oblique anterior wall (12.5%), double floor contour (6.3%), and bridged appearance (6.3%). Our study found smaller dimensions and morphological variations of ST in Brazilian SS individuals, highlighting the importance of ST-related imaging screening. Keywords: Hypopituitarism; Sella Turcica; Radiography dental digital; Maxillofacial abnormalities.

Introduction
Sheehan's syndrome (SS) is an endocrine disorder known since 1937 when researcher Harold L. Sheehan observed a pattern of partial or total necrosis of the pituitary gland of women who had suffered severe blood loss during or after childbirth as a consequence of heavy bleeding or hypovolemic shock (Sheehan, 1937). During the gestational period, the pituitary gland may double its volume, mainly due to the hyperplasia of the adenohypophysis (Gibbins, Albright & Rouse, 2013). Thus, it is more susceptible to ischemia damage, as it has a deficient arterial micro-irrigation, with a predominance of venous microirrigation (Karaca et al., 2016).
SS is characterized as a pituitary insufficiency. The loss of gland function decreases, over time, the secretion of hormones such as the adrenocorticotrophic, thyroid-stimulating, luteinizing, follicle-stimulating, cortisol, estradiol, and prolactin. These hormones are associated with bone metabolism, such as growth hormone, which is responsible for maintaining bone mass in adults, regulating bone remodeling (Gokalp et al., 2009). Hormones can change the pattern of bone trabeculation, sometimes causing osteopenia and osteoporosis (Bolanowski, Halupczok & Jawiarczyk-Przybyłowska, 2015).
The research on the dimensional parameters and morphological pattern of ST in patients with SS through lateral teleradiographies is highly relevant and unprecedented. This study hypothesizes that ST-related cephalometric aspects may predict altered dimensions and morphological characteristics in the SS. Thus, this investigation aimed to analyze cephalometric predictors of abnormal ST in SS patients.

Study design
A case-control study was performed with patients previously diagnosed with SS, matched by sex and age with individuals without the disease. In this type of methodological design, a group of patients who have a disease of interest (cases) and a group of individuals without that disease (controls) are recruited for evaluation (Freire & Pattussi, 2018). In addition, this study followed the Strengthening the Reporting of Observational Studies in Epidemiology guidelines for observational investigations (https://www.strobe-statement.org/).

Context
The subjects with SS were attended at the Endocrinology and Diabetology Service of the Walter Cantídio University Hospital (Brazil). Patients referred to dental examination at the Faculty of Pharmacy, Dentistry, and Nursing of the Federal University of Ceará were invited to participate. All volunteers signed an informed consent document, and the research was in accordance with the Ethics Committee of the Federal University of Ceará (approval number # 983.022).

Volunteers
The For each case, a volunteer without SS or systemic or endocrinopathy changes that could interfere bone metabolism was matched by sex and age. All selected patients already had lateral teleradiographies, as they were patients from the medical/dental complex of the Walter Cantídio University Hospital (Brazil).

Variables
The variables assessed in this study were the linear measurements and morphological analysis of ST. Linear measures (length, diameter, and depth) were measured using the methods of Silverman (1957) and Kisling (1966) (Silverman, 1957;Niswander, 1967). All reference lines used in the present study were located in the mid-sagittal plane.  Source: Authors.

Data sources /Measurements
The digital lateral teleradiographies were obtained using the Kodak K9000 3D equipment (Kodak Dental Systems, Carestream Health, Rochester, NY, USA) belonging to the UFC Dentistry course, with a 14-bit grayscale (16384 tons), kilovoltage, and milliamperage adjusted according to the size of each patient with an exposure time ranging from 0.4 to 1.6 seconds. For radiographic taking, patients were positioned upright, and their heads were positioned correctly on the cephalostat and oriented horizontally in the Frankfurt plane, with the teeth in Maximum Habitual Intercuspidation (MHI). The teleradiographies were exported in Tagged Image File Format -Image File Format with Tags (TIFF) and evaluated in a dark and quiet room, using the Adobe Photoshop CS5 program (Adobe Systems Incorporated, USA), where they were calibrated at 300 dpi. TS morphology analyzes were performed for each patient. The teleradiographies were exported to the Radiocelf Studio 2 software (Radiomemory, Belo Horizonte-MG), where they were calibrated at 300 dpi. Linear measurements of ST were performed for each SS or control patient.

Reliability
All analyses were performed by two examiners during three consecutive days and repeated within 15 days to ensure the reliability and agreement of the measurements. To assess reproducibility errors, (Harris & Smith, 2009) the following analyzes were performed: (1) Cohen's kappa test for categorical data; (2) intraclass correlation coefficient (ICC) to assess systematic errors regarding numerical data; (3) Dahlberg's formula to assess casual errors in the measurements made.

Statistical analysis
Data were submitted to the Kolmogorov-Smirnov normality test and analyzed using Student's t-test, unpaired or paired ANOVA followed by the Bonferroni post-test and correlation of Pearson (parametric data) or Chi-square/Fisher's exact (categorical data). After categorizing patients, data were compared between the two groups using Fisher's exact test and Pearson's chi-square test or Mann-Whitney test (nonparametric data) using the SPSS 20.0 software (p <0.05). In addition, a ROC curve was obtained for the diagnosis of SS based on the measures of the ST, with cutoff points to estimate sensitivity, specificity, positive and negative predictive values, and accuracy.
The mean age of the individuals was 65.47 ± 10.19 years. SS patients had lower mean length (p <0.001), diameter (p <0.001) and height (p = 0.033) compared to the control group. In both groups, the width and height did not differ significantly but were greater than the length (p <0.001) (Figure 2). Regarding cephalometric analysis-based skeletal classification in individuals with SS, Class I patients presented the width and height of the ST significantly higher than the length (p = 0.024). In Class II patients, there was no difference between measures (p = 0.198), with ST showing a more spherical shape characteristic in these SS patients. In Class III patients, only the width was significantly higher than the length (p = 0.004), which resulted in an oval ST for this skeletal classification (Table 1). Research, Society andDevelopment, v. 10, n. 5, e52210515316, 2021 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v10i5.15316 6 In patients with SS, ST width (p = 0.012) and height (p = 0.015) were directly correlated with the area of the sella. Furthermore, ST width and height (p = 0.001) correlated with each other measurement (Table 2). The establishment of cutoff points to estimate SS based on the measurements of the ST showed high areas under the ROC curve that were statistically significant for sella area (p <0.001), length (p <0.001), width (p <0.001), and height (p = 0.035). Cut points of 50 cm² (area), 6.5 cm (length), 8.5 cm (width), and 6.5 cm (height) were defined as suggestive points for SS diagnosis according to ST analysis on lateral cephalometric exams since they showed reasonable values of sensitivity, specificity, and accuracy (>80%) ( Table 3). Compared to controls, the presence of morphological alteration of the TS of individuals with SS was statistically significant (p = 0.009). The most frequent morphological alterations of ST in women with SS were irregularities of the dorsal face (p = 0.018). Ponticulos ponticus was found in 37.5% of SS individuals, equally distributed between incomplete and complete types (Table 4).

Discussion
Cephalometric analyzes are essential tools in the craniofacial description. In the field of dentistry-related imaging science, this is the first study with SS patients that evaluated morphologic and morphometric aspects among these patients.
The size of the pituitary gland, which controls the body's endocrine functions, is also affected by the morphometry of the ST (Kjaer et al., 1999). Dimensional changes in the ST are associated with several developmental anomalies (Kjaer et al., 1999;Kjaer, 2015). Any change in the development of the pituitary gland can affect the size and shape of the ST (Kjaer, 2015).
The growth of ST is completed in early childhood, and several studies use dimensions and morphology of ST to assess facial direction and growth (Kjaer et al., 1999;Abdel-Kaderm, 2007). Thus, the shape and size of ST gain importance under some clinical conditions, such as SS, Type I Diabetes, Turner Syndrome, Fragile X Syndrome, Trisomy 21, Kallmann Syndrome, Type I Neurofibromatosis, Meckel -Gruber Syndrome, and Williams Syndrome (Kjaer, 2015;Diri H et al., 2014;Friedrich et al., 2017;Kjaer et al., 1998).
Regarding the linear dimensions of ST, several studies have shown that there are variations in its morphology throughout 8growth and development (Björk & Skieller, 1983;Axelsson et al., 2005;Alkofide, 2007 (Suri, Tompson & Cornfoot, 2010;Korayem & AlKofide, 2015). In analyzes using lateral radiographs, as in our approach, normality of ST was found less frequently. The most frequent abnormality was an oblique anterior wall. Irregularity in the dorsal wall and pyramidal shape was present in some individuals with DS (Korayem & AlKofide, 2015). Irregularity of the dorsal face (37.5%; p = 0.018) and an oblique anterior wall (12.5%) were morphological patterns found in our study. A previous investigation performed a case review on the normal and pathological development of ST and pituitary gland in humans. The shape of the ST of individuals with Chondrodystrophy, enlarged and with an irregular internal contour, and Hydrocephalus subjects, who had a wide and abnormal posterior wall, did not present any connection between the anterior and posterior walls, was described (Kjaer et al., 1997). Unlike the results of our study, individuals with SS had statistically significant variations in the morphological pattern.
Despite the correlation between the pituitary gland and the ST in several syndromes, there are still controversies about how this interaction occurs. Smaller ST size plays a critical risk factor to the occurrence of SS. During the gestational period, the pituitary increases the sella size due to intense hormonal activity. Then, the compression and ischemia of the ST, due to lack of space, may predispose to SS development. Thus ST could be a predictor of the syndrome instead of an etiological fator (Bakiri et al., 1991).

Conclusion
The present study is innovative in analyzing cephalometric predictors of ST abnormalities in patients with SS. This investigation found smaller linear dimensions of length, diameter, depth, and morphological variations for patients with SS compared to healthy patients, highlighting the importance of teleradiographic evaluation in individuals with SS since ST is easily identified in such exams commonly requested in Dentistry. In addition, we strongly recommend future studies for the imaging evaluation of maxillomandibular repercussions from endocrine diseases, such as those related to the ST.