Contrast-free MRI protocol for pituitary assessment in children with growth or puberty disorders: a practical approach.

Objectives Most of the pituitary MRI examinations in children with growth or puberty disorders (GPD) might not require gadolinium-based contrast agent (GBCA) administration. Methods Retrospective re-analysis of contrast-enhanced 567 pituitary MRIs of children with GPD. Two sets of sequences were created from each MRI examination: Set1 - common sequences without contrast administration and Set2 - common pre- and post-contrast sequences. The differences in the visibility of pituitary lesions between sets were statistically analyzed.


KEY WORDS
Magnetic Resonance Imaging; Pituitary Gland; Child; Contrast Media, Gadolinium

KEY POINTS
• In most cases, non-contrast MRI of pituitary gland in children with growth or puberty disorders is sufficient for correct gland assessment and further clinical management.
• The antero-posterior pituitary dimension (the cut-off value = 7.5 mm) was the most significant factor in determining the appropriateness of using gadolinium-based contrast agent.
• A practical approach to MRI assessment of the pituitary gland in children with growth or puberty disorders will optimize the management of patients before hormone therapy. changes exclusion is one of the most important elements of diagnostic management of growth and puberty disorders (GPD) in children. [1], [2] The method of choice for pituitary assessment is magnetic resonance imaging (MRI). [2] Actually, the MRI protocol of the pituitary gland includes images before and after administration of gadolinium-based contrast agent (GBCA), [2], [3] which according to current knowledge is not neutral to the human body. Apart from the commonly known gadolinium-related side effects, such as allergic reactions as well as nephrogenic systemic fibrosis (in kidney failure), [4], [5] recently scientific literature alerts about the possible accumulation of gadolinium compounds in some brain structures (globus pallidus and dentate nucleus). [6] These observations are of concern to the medical community and the questions are self-evident: will the developing brains of children have an increased susceptibility to long-term, potentially neurotoxic effects of GBCA deposits [7], [8] and whether it is possible to limit the use of GBCA in pituitary MRI in children. [9], [10] Additionally, the omission of GBCA administration would shorten the examination time and the duration of general anesthesia, which is necessary in the youngest patients. [2] Over the past decade, the concern about possible harmful effects of anesthetics and sedatives used in neonates and young children have increased. [11]- [14] Studies on animals demonstrate that repeated or prolonged administration of these drugs is potentially neurotoxic and may cause delayed cognitive development. [14]- [16] The aim of the study was to assess the diagnostic value of non-contrast MRI of the pituitary gland in children with GPD and to determine the criteria indicating the necessity to perform post-contrast examination.
We hypothesized that most of the pituitary MRI examinations in children with GPD do not require GBCA administration.
To the best of our knowledge, this is the first study focusing on performing MRI of the pituitary gland without gadolinium administration in children.

Study Participants
For this retrospective study, after approval by the local ethics committee (opinion No. 117/2019), a total of 579 young patients from the Department of Developmental Endocrinology and Diabetology who met several inclusion and exclusion criteria were selected. Inclusion criteria were as follows: contrast-enhanced pituitary MRI examination carried out in our Radiology Department, age below 18 years, and growth or puberty disorders. Exclusion criteria were: inadequate MRI due to artifacts (n=1) or absence of contrast agent (n=11). Consecutive pituitary MRIs of 567 children between January 2007 and May 2020, were retrospectively reevaluated (Table 1).
For the purposes of statistical analysis, two sets of sequences ( Figure 1) were created from each pituitary MRI examination of every patient: Set1 including common sequences without contrast administration and Set2 which included common pre-and post-contrast sequences (conventional MRI examination). Set1 was the reference ("control") group for Set2 in each patient. Each pair of sets (Set1 and Set2) from each patient was assessed separately and independently, and the evaluating neuroradiologist was blinded to patient information.
The patients were deliberately not divided into age groups, because they were individuals with hormonal disorders and in their case, according to our observations (the head of the project has 20 years of experience in pituitary imaging) and reports of other research centers, [17] the dimensions of the pituitary gland are not directly proportional to age, which is in a healthy population.

Technical Details and Image Analysis
The MR examinations were carried out using two devices: GE 1.5T MR Signa HDx (n=425) and 3T Philips Ingenia (n=142), using 16-channel coils dedicated to the head and neck area.
T1 weighted images (T1-WI) and T2 weighted images (T2-WI) were taken in the coronal and sagittal planes using thin 3 mm slices before and after intravenous administration of the macrocyclic GBCA. The contrast dose was 0.1 mmol/kg body weight (0.2 ml/kg BW). General anesthesia was additionally used in younger patients (below age of 7 years) to exclude movement artifacts.
The size (gland volume and three dimensions) and the morphology of the pituitary gland, the possible presence and then location, signal pattern (SP) and potential occurrence of contrast enhancement of focal lesions were retrospectively reassessed on the GE ADW 4.6 as well as the Philips IntelliSpace workstations.

Statistical analysis
The computation was performed in the R statistical platform. [18] The receiver operating characteristic (ROC) curve analysis was performed using the "pROC" R package. [19] According to the statistical analysis of the differences in the visibility of focal lesions between Set1 (common pituitary MRI sequences without contrast administration) and Set2 (conventional MRI examination with common pre-and post-contrast sequences), we were able to assess the rightness of using GBCA in children with GPD. The cut-offs for the presumed risk factors were estimated using ROC analysis, to distinguish between the negative effect (which is the unnecessary use of GBCA) and positive effect (correct GBCA administration) of the post-contrast examination in conventional pituitary MRI.

Pituitary dimensions
Logistic regression revealed that only the AP dimension of the pituitary gland had statistically significant (P<0.05) chances for diagnostic incorrectness, with the following results: OR=2.23, 95%CI, 1.35 to 3.71, p-value=0.002 (Table 2) The results of these authors indicated that the pathology frequency was as follows: RCC in 0.03%, EPP in 0.02% and adenomas in 0.04% of all patients.
In our study of 567 pediatric patients with GPD, three types of focal pituitary lesions were found: RCC, EPP and MA (11.6%, 3,5% and 0.9% of all patients, respectively). The frequency of focal lesions in our patient group is significantly higher than in the cited study on generally healthy Moreover, our observations showed that in common pituitary MRI sequences without contrast administration (Set1), focal lesions were already visible in about 85% of cases.
In conventional pituitary MRIs, which included common pre-and post-contrast sequences (Sets2), compared to Sets1 the outcome of pituitary MRI assessment changed only in 2.5% of all patients (mostly benign RCCs) and, importantly, only one patient (0.18%) had MA.
The appearance of both RCC, EPP and MA are almost pathognomonic, thanks to their characteristic location. RCCs are present between the anterior and posterior lobe [21], EPP most often located on the floor of 3rd ventricle, at the median eminence [10] and MAs situated within the anterior lobe. [22] It is well known that RCC is a benign lesion, [23] and if it is small and does not cause a mass effect, it should not be of concern to radiologists or clinicians. Furthermore, its location is so distinctive that the GBCA administration in this case is completely unnecessary.
So called "bright spot" visible on T1-weighted non-contrast MR images in the floor of the third ventricle enables identification of EPP without post-contrast examination. [10] The presence of a focal lesion in the anterior pituitary lobe, which corresponds to MA [22] would seemingly indicate the need for gadolinium administration. However, in light of recent studies incidentalomas should not be considered a contraindication to hormone therapy. As mentioned above, the SP of EPP is very characteristic [10], while SPs of RCC and MA depend on their content. [24], [25] What is more, as our study found, the latter can be very diverse. In The AP dimension is the most accurate factor as a statistically significant predictor in two independent analyzes. Moreover, the ROC analysis itself also indicates the advantage of the AP dimension over other criteria, as it shows the best fitted combination of sensitivity and specificity rates, which indicate the high ability of the predictor to use the GBCA correctly when required, and the proper omission of GBCA administration when unnecessary.
Another argument for the superiority of the AP dimension over others is the fact that children with precocious puberty usually have an enlarged pituitary gland with a convex upper outline, which automatically increases the CC dimension, which obviously is not related to the presence of a focal lesion of this gland. [26], [27] Combining the results of our statistical analysis with the above-mentioned literature data, we believe that in the practical approach only the AP dimension should be taken into account when assessing the size of the pituitary gland.
We hypothesized that administration of GBCA in pituitary MRI is unnecessary in the diagnostics of children with GPD. According to the ROC analysis (Fig. 3), the hypothesis holds true for the small size of the pituitary gland. With the AP dimension above the cut-off value (>7.5 mm), this hypothesis becomes false and GBCA should be administered in such cases.
According to the ROC analysis, 73% of patients with an AP dimension >7.5 mm should receive GBCA, and although the remaining 27% of patients in this group will receive GBCA unnecessarily, statistical analysis showed that the benefit of conventional examination (with pre-and post-contrast sequences) exceeds the potential losses and it is certainly a better solution than using GBCA for all individuals without exceptions.
In children with GPD, a pituitary-oriented MRI is performed prior to hormonal treatment in order to exclude organic causes of hormonal disorders. The conventional MRI examination protocol includes intravenous administration of GBCA, and in younger patients, additionally sedation is used to avoid motor artifacts.
Gadolinium deposition has been a very popular topic in recent years and the current state of knowledge is changing dynamically. The literature is consistent with the occurrence of the accumulation phenomenon of linear GBCA after their (especially) repeated administration [28], e.g.
in the deep structures of the brain. [6] Opinions, however, are divided on the potential accumulation of macrocyclic contrasts, which are commonly considered a safer option and therefore often dedicated to pediatric examinations. [5], [29]- [31] Another aspect is the potential association between gadolinium depositions and possible clinical consequences, which has not been confirmed so far, [31], [32] but further observations of the long-term effects of this matter need to be carried out, especially in children due to the sensitivity of their developing brain and life span. [7], [8] At present, no histological changes have been confirmed that would arise as a result of gadolinium deposition in the brain tissue after its multiple administration, however, larger studies were conducted only on the adult population, [6], [33], [34] and the study of children was performed on a small number of patients. [8] However, the clinical significance and potential long-term consequences of the administration (especially repeated) of gadolinium even in individuals with normal renal function still remain unclear. [5], [32] This is evidenced by the position of the European Medicines Agency, which recommends using the lowest sufficient dose of macrocyclic GBCA and only in cases where native scans are insufficient [33] and the European Society of Urogenital Radiology, which recommends avoiding unnecessary administration of GBCA in children, especially in newborns and infants. [35] The U.S. Food and Drug Administration is also interested in this phenomenon and continuing to assess its effects in the human body. [36] In our opinion, GBCA administration should be omitted in most pituitary MRIs in children with GPD. Therefore we propose the practical approach for pituitary MRI assessment ( Figure 6).
Worldwide, similar trends are also emerging in other fields of medicine where, especially in studies on the pediatric population scientists are trying to abandon GBCA administration [4], [37], [38] or at least reduce its dose. [9], [10] There are also few studies in the literature indicating a need to reconsider the use of GBCA in pituitary MRI, which is not always necessary for the correct interpretation of the sellar region. [38], [39] As with the harmfulness of gadolinium deposits, scientists are also divided over the toxicity of anesthetics in children [11]- [14] Animal studies have confirmed the neurotoxicity of anesthetics and sedatives on the developing brain tissue, [15], [16] and while these observations cannot be directly extrapolated to the pediatric population, they are of concern. Due to the specificity of the studied age group, most research are retrospective, and therefore not ideal. Some of these studies have found an association between exposure to anesthetics and neurobehavioral problems or cognitive impairment in the later stages of development, which is especially noticeable with repeated or prolonged exposure. [12], [13] Although several new studies have shown no noticeable neurocognitive impairment in children after short-term anesthetics exposure, [12], [14] however, as with the gadolinium deposition, the long-term side effects of such exposure are not yet known.
Due to constant concerns about the potential harmfulness of GBCA and the potential risk of general anesthesia in younger children, and in accordance with the results presented above, in our opinion, GBCA administration in unquestionable cases is not only unnecessary, but even could be considered an impropriety in the medical decision-making process.
Out of concern for the welfare of the youngest patients, we designed the procedure algorithm ( Figure 6) for pituitary MRI assessment in children with GPD before hormonal therapy.
The suggested MR protocol includes only T1 and T2-WI in the coronal and sagittal planes without contrast administration. Post-contrast examination may be omitted if the native MR examination meets the following criteria: 1. The pituitary gland must not present any focal lesions on T1 and/or T2 image or the lesion must be in the typical location for RCC, EPP or MA.
2. The AP dimension of the pituitary gland must be smaller than the cut-off value (7.5 mm).
Obviously, if a suspicious or potentially malignant lesion outside the sellar region is detected, not only should GBCA be administered, but the scope of the examination should also be extended to include a complete MRI protocol of the brain.
The present study has several limitations. First, it has a retrospective design; second, none of our patients underwent surgery, therefore histopathological confirmation of the diagnoses made by MRI was not possible. Additionally, the subgroups of children aged <2 and >12 years were small. Despite its limitations, to the best of our knowledge this is the first study involving an attempt of complete GBCA omission in pituitary MRI in children.

CONCLUSION
The results of our research have relevant implications for clinical practice or health policy.
For most children, it will be possible to shorten the examination time by omitting GBCA administration, which in turn has a number of advantages. Briefer examination means reduced time of general anesthesia in younger children.
Furthermore, reducing the procedure time would increase the chances of obtaining good-quality MRI of the pituitary gland in a conscious young patient who is, however, old enough not to move during the shortened examination without GBCA administration.
Shortening the examination time is also important for health policy due to its economic aspects.
Reducing the number of post-contrast examinations will save time, during which more children with GPD waiting in line might be examined. This aspect is particularly important in countries where access to MRI examinations is limited. This will increase the availability of MRI examinations for waiting children and accelerate their further clinical management.     analysis (AUC lower CI 95% >50%), but only AP dimension is statistically significant according to the logistic regression analysis (P= .002) and also has the best the best fitted combination of sensitivity (69%) and specificity (74%) rates.

Compliance with Ethical Standards:
Conflict of Interest: The author (Marta Michali-Stolarska, Andrzej Tukiendorf, Jagoda Jacków-Nowicka, Anna Zacharzewska-Gondek, Joanna Chrzanowska, Joanna Bladowska) declares that he/ she has no conflict of interest Ethical approval: All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Approval for this research was given by the Commission of Ethics at the Wrocław Medical University (number of permission: 117/2019). Informed consent was obtained from each patient to participate in this study. ∼ ∼