PDF
Citation
Print
INTRODUCTION
Pituitary incidentaloma (PI) refers to a lesion around the pituitary gland that is inadvertently detected during an imaging examination for reasons unrelated to pituitary concerns [123]. The name describes the fortuitous nature of its discovery, and PIs histologically correspond to a wide range of diseases [3]. Pituitary adenoma is the most common neoplastic lesion and Rathke’s cleft cyst (RCC) is the most common cystic lesion [1]. With improved access to medical services and rapid advances in imaging technology, the prevalence of PI is increased [2]. However, not all incidental pituitary tumors are operated upon since the exact pathological diagnosis is unknown. Thus, a presumptive diagnosis is generally used [1]. A basal hormone study (BHS) is generally performed for patients diagnosed with PI to check hormonal statuses. Dynamic sella MRI and ophthalmological examinations to confirm visual field disorders are also performed [245]. If the PI is accompanied by tumor-related symptoms, such as headache or visual disturbance, or presents as a functional lesion, it is treated with surgery, bromocriptine, or cabergoline. Except for pituitary apoplexy, the etiological link between headache and a pituitary lesion is hardly proven [3]. In the case of an asymptomatic non-functional lesion, the patient is followed up periodically [4]. The American Society of Endocrinology and the French Society of Endocrinology have proposed the following guidelines. Once a tumor is confirmed as functional, surgery or medication is initiated immediately. If the lesion is non-functional, it is classified as a macro- or micro-incidentaloma, and additional tests or periodic follow-ups are suggested [45]. For asymptomatic non-functional micro-incidentaloma, available guidelines recommend an annual MRI for 3 years [4] or at 6 months, 1 year, and 2 years [5]. While there is a general consensus regarding follow-up periods in existing guidelines, some details remain inconsistent, for example, the follow-up period for non-functional pituitary micro-incidentaloma [3]. Thus, the present study analyzed neuroradiological data from asymptomatic patients with non-functional PI who did not require surgery or medication at the time of diagnosis and assessed to suggest follow-up periods for PI based on MRI characteristics. Presumptive RCC and pituitary adenoma are relatively common and often detected as PIs. Careful radiological analysis can often help to determine the nature of the lesion, although imaging will never replace histological analysis to establish a final diagnosis [3]. However, little is known about changes in PI size based on morphological MRI findings. To address this issue, we also analyzed risk factors for progression of presumptive RCC and pituitary adenoma based on morphological MRI features.
MATERIALS AND METHODS
Participant characteristics
This study included 380 consecutive patients aged 19 years or over who were diagnosed with diseases related to pituitary lesions at our hospital from 2007 to 2023. At the time of diagnosis, dynamic sella MRI, BHS, and/or ophthalmological examination were conducted. Among these, asymptomatic patients with non-functional PI were included whereas patients who underwent surgery or medical treatment at the time of diagnosis were excluded. Seventy-one patients underwent tumor removal surgery using a transsphenoidal approach as indicated following diagnosis. In addition, 44 patients who were diagnosed with prolactinoma received medical treatment. Of the remaining 265 patients, 20 patients were lost to follow-up as they could not be contacted. Finally, 245 patients, who were followed up for more than 12 months, were retrospectively analyzed (Fig. 1). Baseline participant characteristics at the time of diagnosis are summarized in Table 1. Of these 245 patients, 110 and 135 were diagnosed with presumptive RCCs and pituitary adenomas, respectively. In this study, presumptive diagnosis of PI was made.
MR imaging protocol and analysis
A 3.0 T-dynamic sella MRI (Skyra, Siemens, Munich, Germany) with enhancement was performed for all patients diagnosed with PI. All images were interpreted by neuroradiologists. MRI morphological features of PI were determined by two neurosurgeons through extensive analysis of all imaging data. When differences in opinions arose, imaging data were re-examined by both observers. Results were recorded after discussion once an agreement was reached. The MRI protocol used was as follows: 3 mm in thickness (2 mm thickness in dynamic view), matrix size of 448×327, and sagittal+coronal T1/T2 weighted sequences without gadolinium injection; dynamic coronal T1 weighted with gadolinium injection, creating 9 dynamic image frames with temporal resolution of 15 s each; and sagittal T1 weighted sequences with gadolinium injection.
Typical MRI findings of a pituitary adenoma include a slow enhancement compared to a pituitary gland [6] and a typical location of an intrasellar mass [7]. Pituitary adenomas show various signal intensities on T1- and T2-weighted images depending on cells of origin or the type of secretion [89]. RCC is observed on an MRI as a well-defined non-enhancing midline cyst within the sella arising between anterior and intermediate lobes of the pituitary gland. RCCs are usually unicameral and homogeneous in signal intensity and show various signal intensities on both T1- and T2-weighted images [6]. No contrast enhancement of the cyst was seen. However, a thin enhancing rim of surrounding compressed pituitary tissue might be apparent [1011]. The presence of a fluid-fluid level, septation, and off-midline location strongly favors diagnosis of a cystic pituitary adenoma over RCC [6].
Changes in PI size were compared between the most recent MRI and that taken at the time of initial diagnosis. The presence or absence of pituitary stalk deviation was determined using the MRI performed at the time of initial diagnosis. Progression or regression of PI was defined as a tumor height change of more than 1 mm.
Basal hormone study
BHS was performed for patients diagnosed with PI. Blood sampling times were considered. Central hypocortisolemia was defined as a low serum cortisol in the morning or an inappropriate response to a rapid adrenocorticotropic hormone (ACTH) test. Serum cortisol and ACTH levels were determined using an Elecsys Cortisol II (Roche Diagnostics GmbH, Mannheim, Germany). An electrochemiluminescence immunoassay based on the competition principle and Elecsys ACTH (Roche Diagnostics GmbH), an electrochemiluminescence immunoassay based on the Sandwich principle were performed. Central hypothyroidism was determined based on low free plasma thyroxine (fT4). Normal TSH, free T4 (Elecsys FT4 IV, Roche Diagnostics GmbH), T3 (Elecsys T3, Roche Diagnostics GmbH), and TSH (Elecsys TSH, Roche Diagnostics GmbH) tests were performed. Hyperprolactinemia was diagnosed if serum prolactin level was above relevant gender-specific cutoff (Elecsys Prolactin II, Roche Diagnostics GmbH) test. Insulin-like growth factor I (IgF-1) was tested using a sandwich chemiluminescence immunoassay, LIAISON IGF-I (DiaSorin S.P.A, Saluggia, Italy). If the level of IgF-1 was higher than the age-specific cutoff, 75-g oral glucose tolerance test (OGTT) was additionally performed to definitively diagnose acromegaly. Blood glucose was measured using an enzymatic reference method with hexokinase (GLUC3, Roche Diagnostics GmbH). Hypogonadism was defined as LH and FSH levels lower than cutoffs, with reference to menopausal status in women. LH and FSH concentrations were measured using Elecsys LH and Elecsys FSH (Roche Diagnostics GmbH), respectively. In men, low testosterone was defined as below the cutoff of normal LH and FSH levels. Testosterone was measured using a VIDAS Testosterone II (bioMérieux S.A., Marcy l’Étoile, France), an enzyme-linked fluorescent assay based on the competition principle.
Ophthalmologic examination
To check for optic nerve disorders in patients diagnosed with PI, visual acuity tests were performed using a 4-m visual acuity chart (developed by Jin [12]). Humphrey Visual Field 30-2/60-2 tests were performed using a Humphrey analyzer (Carl Zeiss Meditec, Inc., Dublin, CA, USA).
Follow-up evaluation
Patients with asymptomatic non-functional PI were followed periodically with dynamic sella MRI at 6-month or 1-year intervals per existing guidelines to monitor for change in lesion size. A presumptive diagnosis based on MRI was made for patients who did not satisfy criteria for surgery. Patients were instructed to present to the hospital immediately if they showed any of the symptoms of PI, such as headache, nausea, diminished visual acuity, visual field cut, and ophthalmoparesis [13]. If patients demonstrated hormone changes during follow-up, they were seen by an endocrinologist. Additional hormone testing was performed if necessary. If the PI enlarged and led to symptoms or hormonal fluctuations requiring treatment, patients were scheduled for surgery or drug therapy. To evaluate the PI enlargement, the study population (n=245) was divided into two groups as follows: 1) a progression group consisting of patients whose PI size increased; and 2) a reduction/no change group consisting of patients whose PI size remained unchanged or decreased. The study population was also divided into presumptive RCC (n=110) and pituitary adenoma (n=135) groups to analyze the radiological features.
Statistical analyses
SPSS for Windows (Version 20.0; IBM Corp., Armonk, NY, USA) was used for all statistical analyses. Associations between categorical variables and tumor enlargement were analyzed using Pearson’s χ2 test (or Fisher’s exact test as appropriate). Continuous variables were compared using Student’s t-tests, Mann–Whitney test, and Kruskal–Wallis test. Prognostic factors considered for univariate analysis included patient demographics, tumor size at diagnosis (<5 mm, 5–10 mm, or ≥10 mm), pituitary stalk deviation, and characteristics of MR images (presence of enhancement after contrast injection, T1, or T2 signals). Multivariate logistic regression analysis was used to assess associations of variables with tumor enlargement. Factors associated with progression and reduction of pituitary adenomas were analyzed using multinomial logistic regression. Results are presented as adjusted hazard ratios (HR) or odds ratios (OR) with 95% confidence interval (CI). The level of statistical significance was set at p≤0.05.
Ethics statement
The Institutional Review Board (IRB) of Dongguk University Ilsan Hospital approved all aspects of this study (IRB FILE No: 2023-10-013). The requirement for informed consent was waived by the IRB as this study only depended on information obtained as a part of routine clinical care from patient medical records.
RESULTS
Median clinical and neuroradiological follow-up periods after diagnosis were 74 months (range, 12–195 months) and 27 months (range, 12–123 months), respectively. During follow-up, 33 (13.5%) of the 245 patients demonstrated PI enlargement. PI remained unchanged in 180 (73.4%) patients but decreased in size in 32 (13.1%) patients. Their radiological features are shown in Table 2. Of the 33 patients with PI size increased, 10 patients underwent tumor removal surgery at a median of 28 months (range, 6–44 months) after diagnosis because of newly developed neurological symptoms, such as visual disturbances caused by enlarged tumor (Table 3).
PI size changes during neuroradiological follow-up period
Baseline characteristics at the time of diagnosis for the progression group (n=33, 13.5%) and the reduction/no change group (n=212, 86.5%) are shown in Table 4. Factors significantly associated with PI progression were pituitary stalk deviation (hazard ratio, 12.31; 95% CI, 4.24–35.80; p<0.001) (Fig. 2) and presence of PI enhancement (hazard ratio, 2.69; 95% CI, 1.12–6.44; p=0.026). The larger the PI at diagnosis, the greater the tendency for PI progression (hazard ratio, 1.86; 95% CI, 0.64–5.43; p=0.064); however, the multivariate analysis result was not statistically significant.
The mean age of 110 patients diagnosed with presumptive RCC was 45.5 years. Median clinical and neuroradiological follow-up periods were 32 months (range, 12–133 months) and 23 months (range, 12–123 months), respectively. Participant characteristics at diagnosis are shown in Table 5. At the last neuroradiological follow-up, RCCs had progressed in 6 (5.5%) patients but spontaneously reduced in 15 (13.6%) patients. In patients with RCCs that reduced in size, these RCCs resolved and became undetectable in the most recent MRI in 7 (6.4%) patients. The remaining 89 (80.9%) patients had RCCs that remained unchanged throughout the follow-up period. Representative MR images of RCCs that reduced in size are shown in Fig. 3. MRI morphological factors were not related to a change in size. Notably, the no change group was older than the reduction group, and the progression group was older than the no change group (p=0.226).
The mean age of 135 patients diagnosed with presumptive pituitary adenoma was 51.0 years. Median clinical and neuroradiological follow-up periods were 44 months (range, 12–195 months) and 24 months (range, 12–114 months), respectively. Patient characteristics at diagnosis are presented in Table 6. Of 135 patients, 14 were diagnosed with cystic pituitary adenoma. At the last neuroradiological follow-up, pituitary adenomas had progressed in 27 (20.0%) patients and decreased in 17 (12.6%) patients. In the remaining 91 (67.4%) patients, pituitary adenomas were unchanged during the follow-up period. Absence of tumor enhancement was significantly associated with tumor reduction (odds ratio, 4.54; 95% CI, 2.26–9.11; p<0.001) while pituitary stalk deviation was significantly associated with tumor progression (odds ratio, 0.04; 95% CI, 0.01–0.14; p<0.001). In univariate analysis, non-cystic pituitary adenoma and isointensity in T1-weighted images were significant factors associated with tumor enlargement (p<0.001 and p=0.030, respectively). However, they were not significant factors in multivariate analysis.
Serum hyperprolactinemia in asymptomatic patients with PI
Nineteen patients had high serum prolactin (mean±SD, 65.0±31.8 ng/mL). However, none of these patients had an increase in serum prolactin level during follow-up. Of these patients, seven patients were medicated with typical antipsychotics or antiemetic drugs. Their serum prolactin levels decreased to normal levels or decreased after changing or stopping medications without any remarkable symptoms. At the time of tests, two patients were breastfeeding. In one of these patients, her serum prolactin level returned to normal (81.9→16.4 ng/mL) after stopping breastfeeding. Furthermore, two patients showed stalk section effect suspected of non-functioning pituitary adenoma. In one of these patients, serum prolactin level changed from high to normal (56.0→16.0 ng/mL) after 6 months. However, follow-up MR images showed no changes. In another patient, serum prolactin level did not elevate during follow-up at 12 months. One patient had subclinical hypothyroidism. During endocrinology follow-up at 6 months, the patient’s hormone levels decreased to normal levels. Of patients who had higher-than-normal serum prolactin levels without any drug history or specific cause, three returned to normal serum prolactin levels during follow-up at 6 months, while four continued follow-up at 6 months without any changes in serum prolactin levels.
DISCUSSION
In our study, stalk deviation and contrast enhancement of the PI at the time of diagnosis were significantly different between progression and reduction/no change groups during follow-up. In the reduction/no change group, there was no increase in PI size or abnormal symptoms after a mean of 19.3 months of neuroradiological follow-up. Thus, imaging follow-up was discontinued. Clinical follow-up was continued for a mean of 36.4 months after discontinuing neuroradiological follow-up. However, abnormal symptoms caused by an increase in the size of the pituitary incidentaloma were not observed [13]. Although it was not significant, 76 (35.9%) patients from the reduction/no change group had tumor sizes ≥10 mm at the time of diagnosis. However, 21 (63.7%) patients from the progression group had tumor sizes ≥10 mm at the time of diagnosis. Of these 21 patients, 10 underwent tumor removal surgery, and all demonstrated stalk deviations in MRI.
The US Endocrine Society recommends long-term follow-up with serial MRI for asymptomatic non-functioning micro-PIs, regardless of the lesion size at the time of diagnosis [4]. The French Endocrine Society recommends surveillance for lesions measuring 5–10 mm for up to 2 years, but no follow-up for lesions smaller than 5 mm [5]. Constantinescu and Maiter [3] have divided PIs into microadenomas, macroadenomas, and RCC and recommended MRI at 6 months and 1 year after diagnosis. Chong et al. [14] have reported that a longer follow-up period and pituitary stalk displacement were associated with lesion progression. Galland et al. [5] have reported that while PIs might grow, it is rare (10%–13%). Even with a long-term follow-up, less than 5% of patients showed size changes exceeding 10 mm. In light of our results and findings from a previous study, we recommend that for patients with PI measuring less than or equal to 10 mm without stalk deviation and contrast enhancement, short-term radiological follow-up suffices as long as the patient is instructed to seek outpatient or emergency care if they develop symptoms that suggest potential PI growth, such as headache, nausea, diminished visual acuity, visual field cut, and ophthalmoparesis [13].
In our study, the pathological diagnosis of patients who underwent surgery was confirmed to be pituitary adenoma in seven patients and RCC in three patients. The presumptive diagnosis of the three patients with RCC was cystic pituitary adenoma. Their initial radiological images revealed a cystic pituitary macroadenoma with hemorrhagic degeneration. Clinical manifestations at the time of diagnosis included incidental findings, dizziness, and visual disturbances. Intratumoral hemorrhage in pituitary adenomas is common even without clinical symptoms [15]. RCCs are sometimes indistinguishable from cystic pituitary adenomas in an MRI [6]. Wen et al. [16] have reported that 50% of surgically proven RCCs are preoperatively misdiagnosed as pituitary adenomas. Various signal intensities of an RCC can mimic a cystic pituitary adenoma with hemorrhage, which makes imaging diagnosis of a cystic pituitary adenoma or RCC challenging [6].
In our study, among the 110 patients diagnosed with presumptive RCC, some patients showed small RCCs located between the anterior and intermediate lobes of the pituitary [6] and immediately below the pituitary stalk without evidence of stalk deviation. In 15 (13.6%) of these patients, the lesion reduced in size during follow-up. Kim et al. [17] and Truong et al. [18] have reported spontaneous shrinkage or complete resolution of RCCs. However, the underlying mechanism is poorly understood [14]. According to Kinoshita et al. [19], 10% of patients diagnosed with RCC experience increased lesion size. Moreover, patients were older than those with stable lesions. Although not significant, our results of RCC showed that patients with reduced sizes during follow-up tended to be younger than those with unchanged lesions and that patients with unchanged lesions tended to be younger than those with lesion enlargement, similar to Kinoshita et al.’s findings [19]. Moreover, the odds of lesion enlargement increased with advancing age in our study (with 245 patients). In light of our results and previous findings [171819], termination of neuroradiological follow-up at the time of diagnosis and clinical follow-up after 6 months might be feasible for patients with micro-PIs located in anterior and intermediate lobes of the pituitary without stalk deviation.
In our study, among 135 patients diagnosed with presumptive pituitary adenoma, non-enhancing lesions were significantly associated with lesion reduction while stalk deviation was significantly associated with lesion progression. Fernández-Balsells et al. [20] have found a higher risk of growth in macroadenomas and solid lesions than in microadenomas and cystic lesions. In the present study, non-cystic pituitary adenoma and isointensity in T1-weighted images were associated with lesion progression in the univariate analysis. However, such associations were not statistically significant in the multivariate analysis. Considering our results and previous findings, follow-up according to existing guidelines is necessary for patients diagnosed with presumptive pituitary adenoma with stalk deviation.
Physiological causes of hyperprolactinemia without prolactinoma include pregnancy, lactation, exercise, stress, seizures, and sexual intercourse. It might also be caused by the use of antipsychotic drugs and antiemetics or by underlying medical conditions such as chronic renal failure or liver cirrhosis [21]. In our study, some patients diagnosed with asymptomatic PI had high serum prolactin levels. However, these levels were below 200 ng/mL [22]. Patients were followed up without pharmacotherapy after excluding the above causes. After considering patient’s medications or medical history, if a tumor is unlikely, current medications are stopped or changed, if possible. Periodic hormone testing is recommended during follow-up [23]. If there is no change in lesion size and serum prolactin level remains unchanged or reduced during follow-up while the underlying cause is being managed, early termination of follow-up might be considered after educating patients on when to re-present to the hospital, considering both clinical and radiological characteristics.
Our study had some limitations. For example, it was retrospective in nature and had a small sample size with varied PI sizes. Numbers of patients in studies by Imran et al. [24] and Tresoldi et al. [25] about the natural history of PI were 265 and 203 patients, respectively. Additionally, there was a risk of examination error due to MRI thickness, as the MRI protocol included a 3 mm in thickness (2 mm thickness in dynamic view). Following PI without performing a surgery, accurate diagnosis without histological confirmation is always a problem [26]. In this study, we used presumptive diagnoses of PI based on MRI findings. Since there has been no study analyzing changes in PI size over the follow-up period based on neuroradiological characteristics, our study has strengths in this respect. Considering the lack of consensus regarding the management of micro-PIs from available guidelines [134527] and that repetitive and long-term clinical and neuroradiological follow-ups can increase patient time and cost burdens and exhaust healthcare resources [2], our findings may offer an alternative to address this.
In conclusion, we recommend an additional guideline for patients diagnosed with asymptomatic non-functioning micro-PI without stalk deviation and contrast enhancement on MRI. In particular, early discontinuation of clinical and neuroradiological follow-up should be supported, as long as patients are educated regarding the possible neurological symptom indicative of tumor progression.
XML Download