Correlation between MRI findings of pituitary gland and prolactin level among hyperprolactinemia adult female Saudi patients in rural areas: A retrospective multicentric study

Abstract

Identifying the prolactin threshold that necessitates pituitary magnetic resonance imaging (MRI) in patients with hyperprolactinemia remains challenging. Therefore, developing standards for serum prolactin level criteria to predict prolactinoma is critical. This study aimed to investigate the correlation between hyperprolactinemia and the presence of pituitary adenoma among Saudi female patients with verified prolactin levels. A retrospective multicentric study, including 4 regions from western Saudia Arabia between July 2020 and September 2023, included 168 female patients with abnormal prolactin levels who underwent brain MRI for the pituitary gland evaluation. The prevalence of pituitary adenoma and its associated factors and the relationship between blood prolactin levels and pituitary adenoma on brain MRI, as well as establishing the cutoff value of serum prolactin concentration linked to the existence of pituitary adenoma, were investigated and analyzed. The mean prolactin level was 72.7 ± 63.2 ng/mL. MRI findings were microadenoma in 77 (46.1%), macroadenoma in 17 (10.2%), Empty sella syndrome (ESS) in 7 (4.2%), and normal in 66 (39.5%) patients. In the ROC curve analysis, prolactin levels demonstrated a moderate degree of accuracy in predicting the existence of a pituitary adenoma (AUC = 0.640; 95% CI = 0.563–0.713; P = .0010], and the sensitivity and specificity were 40.59% and 83.33%, respectively. The ideal cutoff prolactin level for diagnosing pituitary adenoma was > 38.71 ng/mL with a sensitivity of 77.23% and specificity of 40.91%. It is prudent to perform pituitary imaging in most cases of hyperprolactinemia without event etiology, even if the condition is minor, due to the high prevalence of pituitary anomalies in female patients with hyperprolactinemia at serial sampling. A multidisciplinary strategy is necessary for a comprehensive diagnosis, treatment, and follow-up approach to improve the clinical outcomes of these individuals.

1. Introduction

Hyperprolactinemia is a frequent pituitary disorder that causes sexual symptoms, such as irregular menstrual periods, galactorrhea, and hirsutism in women, as well as erectile dysfunction and gynecomastia in men.^[1,2] It affects the gonadal axis, causing diminished libido and infertility in both sexes.^[3] Hyperprolactinemia rates vary from 0.4% in the average adult population to 9 to 17% in women with reproductive disorders.^[4,5]

Prolactin release is generally regulated by inhibitory mechanisms, with dopamine playing an important role.^[2] Hyperprolactinemia can be induced by physiological, pathological, or pharmacological causes.^[6] Stress is undoubtedly 1 of the physiological reasons for hyperprolactinemia; it has been established that stress caused by venipuncture can result in an increase in prolactin levels, and in such circumstances, repeated sampling can lead to normalization in up to almost 1-third of cases.^[7]

The most prevalent cause of hyperprolactinemia is prolactin-secreting pituitary adenoma, which accounts for approximately 40% of functioning pituitary adenomas.^[8] Prolactin levels can indicate prolactinoma or macroprolactinoma, with a value over 250 µg/L indicating prolactinoma and > 500 µg/L indicating macroprolactinoma.^[9] However, high prolactin levels in a sellar mass may not always confirm prolactinoma diagnosis because of the “stalk effect” caused by any sellar lesion compressing the pituitary stalk.^[9] The use of a prolactin threshold > 100 µg/L as a cutoff for pituitary imaging was suggested in a previous report.^[10] However, studies advocate magnetic resonance imaging (MRI) for all patients with chronic hyperprolactinemia, even if mild.^[1,3,8,11] In this context, earlier studies examining females referred for reproductive problems with simultaneous hyperprolactinemia failed to discover a prolactin level threshold for pituitary MRI. Therefore, it has always been suggested, regardless of prolactin level.^[1,12] However, hyperprolactinemia was not established by serial sampling in these studies. Thus, venipuncture stress cannot be completely ruled out as a confounding factor.

Few studies have investigated the correlation between hyperprolactinemia and the presence of pituitary adenoma among Saudi patients with verified hyperprolactinemia. For example, Aljabri et al discovered that hyperfunctioning pituitary glands were strongly related to positive MRI, whereas hypo-functioning pituitary glands were associated with normal MRIs.^[13] In the absence of registry data, more cooperative research, including varied population samples from several sites, may be able to provide further information on the fundamental national frequency. Therefore, the goal of our study was to investigate the correlation between hyperprolactinemia and the existence of pituitary adenoma among Saudi female patients with verified abnormal prolactin levels.

2. Material and method

2.1. Study design and inclusion criteria

This was a retrospective multicenter study involving 4 regions (Alkarj, Afif, Najran, and Riyadh) in western Saudi Arabia between July 2020 and September 2023. This study included 168 female patients with abnormal prolactin levels who underwent brain MRI for pituitary gland evaluation. All female individuals with relevant clinical symptoms and disturbed serum prolactin levels were included in the study.

2.2. Exclusion criteria

Patients with claustrophobia with a pacemaker, aneurysm clip, orbital metallic foreign body, prior pituitary disease, severe renal insufficiency, liver cirrhosis, uncompensated primary hypothyroidism, physiological causes of hyperprolactinemia (i.e., pregnancy and breastfeeding), and patients on hyperprolactinemic drugs.

2.3. Blood sampling

Blood samples were collected in response to symptoms, such as infertility, galactorrhea, amenorrhea, oligomenorrhea, headache, or visual field defects. The normal serum prolactin level in females ranged from 4.79 23.3 ng/mL. The blood prolactin levels were determined using chemiluminescence. All blood samples were collected between 8:00 and 11:00 am in a quiet room at the endocrine investigation day unit while the patients fasted. Initially, stable venous access was achieved by inserting an intravenous cannula into the antecubital vein and keeping it patent with a steady infusion of 0.9% standard saline solution. After withdrawing the first prolactin sample from the indwelling cannula, the patient remained reclined for 30 minutes. After another 30 minutes, a second prolactin sample was collected, and the lowest prolactin value was used for analysis.^[1] The serum prolactin concentration level was classified as low (≤90 ng/mL), mild hyperprolactinemia (border zone) (between 90.1 and 200 ng/mL), and high (>200 ng/mL). Mild hyperprolactinemia, a non-definitive concentration between prolactinoma and nonfunctioning pituitary adenoma, was defined as 90 to 200 ng/mL, and patients in this range are considered borderline zone cases.^[14]

2.4. Magnetic resonance imaging protocol

MRI was performed on a SIEMENS AVANTO (1.5 Tesla) MRI scanner unit with the following sequences: T1-weighted spin-echo (T1-SE; TR 420; TE 15 ms) with 3 mm contiguous coronal slices; T2-weighted spin-echo (T2-SE; TR 3610; TE 111 ms) with 3 mm contiguous coronal slices; and a sagittal T1 acquisition. The field of vision was 21 cm and the acquisition matrix was 272 × 320 pixels. All individuals were investigated without gadolinium (T1-SE), followed by 0.1 mg/kg of gadolinium diethylenetriaminepentaacetic acid (DTPA) immediately before the acquisition began. Dynamic research has also been conducted. The result was regarded as positive if the pituitary gland included a focal point of low signal on unenhanced images and/or less augmentation than the neighboring gland on contrast-enhanced images.^[15] The tumor size was determined using hard-copy sagittal and coronal photographs. Maximum sagittal, transverse, and coronal diameters were measured.^[16] An adenoma was classified as a microadenoma (tumor <10 mm in diameter) or macroadenoma (tumor > 10 mm in diameter).^[1] Empty sella syndrome (ESS) is a radiologic disease in which the sella turcica appears empty because cerebral spinal fluid herniates into the area, squeezes, and flattens the pituitary gland.

2.5. Collected data

The collected data included the patient’s age, primary symptoms such as visual field defects, dizziness and headache, amenorrhea, infertility, symptoms of hypothyroidism, symptoms of hypogonadism, breast problems, hirsutism, and androgenic alopecia; comorbidities such as (hypertension, hypothyroidism, diabetes, lung disease, hypercholesterolemia, and adrenal disease); laboratory data such as [thyroid stimulating hormone (TSH) and prolactin levels]; and radiologic brain MRI findings.

2.6. Main outcome

The primary outcome was the prevalence of pituitary adenomas among female patients with hyperprolactinemia and its associated factors. The secondary outcome involved examining the relationship between blood prolactin levels and brain MRI findings of pituitary adenoma as well as establishing the cutoff value of serum prolactin concentration linked to the existence of pituitary adenoma on MRI.

2.7. Statistical analysis

IBM SPSS software was used for all statistical analyses (IBM SPSS, version 18, Armonk, IBM Corp, Armonk, NY). Continuous variables were reported as mean and standard deviation (SD); categorical variables were reported as absolute numbers or percentages and compared using χ² or Fisher exact tests. Univariate and multivariate analyses were performed using odds ratios (ORs), and their corresponding 95% confidence intervals (CIs) were calculated from β coefficients and standard errors. The diagnostic sensitivity, specificity, positive and negative predictive values, and diagnostic accuracy of the prolactin levels were recorded. Receiver operating characteristic (ROC) curve analysis established a threshold value for serum prolactin levels that might detect pituitary adenomas in hyperprolactinemic individuals. Additionally, it was constructed to determine the optimal measurable cutoff for diagnosis (corresponding to the maximum Youden index). The area under the curve (AUC) was compared using the Student t test. Statistical significance was defined as a 2-tailed P-value < .05.

2.8. Ethical approval

The study was approved by the Ethics Research Committees of Al Maarefa University, Riyadh, Saudi Arabia, in compliance with the ethical standards outlined in the Declaration of Helsinki. All participants provided informed consent, emphasizing voluntary participation, anonymity, and confidentiality.

3. Result

The mean age was 33.6 ± 9.5 years (17.0–67.0 years). Most patients were aged between 30 and 39 years (n = 63, 37.5%), followed by those age–20 to 29 years (n = 59, 35.1%). The main comorbidities were hypothyroidism, diabetes, and hypertension in 17 (10.2%), 12 (7.2%), and 6 (3.6%) patients, respectively. The main symptom was visual field defects in 50 (29.8%) cases, followed by dizziness and headache in 30 (17.9%), amenorrhea in 37 (22.0%), and infertility in 16 (9.5%) cases. Regarding laboratory data, the mean TSH was 2.9 ± 2.1 mIU/L (range: 0.03–13.0 mIU/L) and was within normal ranges in most cases (n = 134, 80.2%). The mean prolactin level was 72.7 ± 63.2 ng/mL (24.3–470.0 ng/mL), and prolactin level > 100 ng/mL was found in 32 (19.2%) cases. The MRI findings were microadenoma in 77 patients (46.1%) (Fig. 1), macroadenoma in 17 (10.2%) (Fig. 2), ESS with pituitary microadenoma in 7 (4.2%) (Fig. 3), and normal brain MRI findings without pituitary adenoma in 66 (39.5%) cases (Table 1).

Figure 1.

Coronal view of contrast-enhanced MRI revealing the pituitary gland with a focal hypointense lesion along the left side of the pituitary gland (White arrow) (A) and contrast-enhanced T1 weighted MRI of the pituitary gland shows a delayed enhancement of a small microadenoma (White arrow) (B). MRI = magnetic resonance imaging.

Figure 2.

Coronal view of contrast-enhanced MRI revealing a bulky appearance of the pituitary gland concerning macroadenoma (A) and macroadenoma with a hypo-enhancing lesion about the remainder of the pituitary gland (B) (White arrow). MRI = magnetic resonance imaging.

Figure 3.

Coronal view of contrast-enhanced MRI revealing a partially empty sella (A) and normal, homogeneously enhancing pituitary gland (B) (White arrows). MRI = magnetic resonance imaging.

Table 1.

Clinicopathological characteristics of female patients with hyperprolactinemia (n = 168).

Variables	N (%)
Age (yr), Mean ± SD	33.6 ± 9.5 (Range: 17.0–67.0)
Age groups
<20 yr	9 (5.4%)
Between 20 and 29 yr	59 (35.1%)
Between 30 and 39 yr	63 (37.5%)
Between 40 and 49 yr	30 (17.9%)
Between 50 and 59 yr	4 (2.4%)
More 60 yr	3 (1.8%)
Comorbidities
Hypertension	6 (3.6%)
Diabetes	12 (7.2%)
Hypothyroidism	17 (10.2%)
Asthma	2 (1.2%)
Hypercholesterolemia	2 (1.2%)
Main symptoms
Visual field defects	50 (29.8%)
Dizziness and headache	30 (17.9%)
Amenorrhea	37 (22.0%)
Infertility	16 (9.5%)
Symptoms of hypothyroidism	14 (8.3%)
Symptoms of hypogonadism	12 (7.1%)
Breast problem	4 (2.4%)
Hirsutism	3 (1.8%)
Androgenic alopecia	2 (1.2%)
Laboratory findings
TSH (mIU/L), Mean ± SD	2.9 ± 2.1 (Range: 0.03–13.0)
TSH subgroup
Normal	134 (80.2%)
High	33 (19.8%)
Prolactin (ng/mL), Mean ± SD	72.7 ± 63.2 (Range: 24.3–470.0)
MRI findings
Microadenoma	77 (46.1%)
Macroadenoma	17 (10.2%)
Normal	66 (39.5%)
Empty Sella Syndrome	7 (4.2%)

Abbreviations: MRI = magnetic resonance imaging, SD = standard deviation, TSH = thyroid stimulating hormone.

3.1. Factors associated with the presence of pituitary adenoma in MRI

The only factor associated with pituitary adenoma was higher prolactin levels, which was statistically significant in univariate (OR = 1.01; 95% CI = 1.01–1.02; P = .003) and multivariate analyses (OR: 1.02; 95% CI: 1.01–1.03, P = .001) (Table 2). In contrast, higher TSH levels were not associated with pituitary adenoma, were associated with idiopathic hyperprolactinemia, and were statistically significant in multivariate analysis (OR, 0.82; 95% CI: 0.67–0.99, P = .043).

Table 2.

Association between pituitary adenoma and other factors in univariate and multivariate analysis.

Variables	Subgroup	No pituitary adenoma (N = 66)	Pituitary adenoma (N = 101)	Univariate analysis		Multivariate analysis
				OR (95% CI)	P-value	OR (95% CI)	P-value
Age (yr)	Mean ± SD	32.0 ± 9.5	34.7 ± 9.4	1.03 (1.00–1.07)	.079	1.01 (0.97–1.05)	.584
Prolactin (ng/mL)	Mean ± SD	53.0 ± 32.4	85.5 ± 74.4	1.01 (1.01–1.02)	.003	1.02 (1.01–1.03)	.001
TSH (mIU/L)	Mean ± SD	3.2 ± 2.3	2.7 ± 1.9	0.89 (0.75–1.03)	.118	0.82 (0.67–0.99)	.043
Main symptoms	Visual problem	19 (38.0)	31 (62.0)	−		−
	Hypothyroidism	2 (14.3)	12 (85.7)	3.68 (0.88–25.31)	.111	4.55 (0.96–34.33)	.084
	Hypogonadism	2 (16.7)	10 (83.3)	3.06 (0.71–21.36)	.176	4.30 (0.92–31.36)	.090
	Dizziness	15 (51.7)	14 (48.3)	0.57 (0.22–1.44)	.237	0.52 (0.18–1.46)	.217
	Amenorrhea	18 (48.6)	19 (51.4)	0.65 (0.27–1.53)	.322	0.64 (0.25–1.67)	.367
	Androgenic alopecia	1 (50.0)	1 (50.0)	0.61 (0.02–16.11)	.735	0.68 (0.03–18.21)	.791
	Breast problem	1 (25.0)	3 (75.0)	1.84 (0.22–38.66)	.609	0.61 (0.04–15.64)	.721
	Hirsutism	0 (0.0)	3 (100.0)	3528982.08 (0.00–NA)	.986	9319283.88 (0.00-NA)	.990
	Infertility	8 (50.0)	8 (50.0)	0.61 (0.19–1.93)	.398	0.63 (0.18–2.19)	.471
Hypertension	No	65 (40.4)	96 (59.6)	−	.271	−	.821
	Yes	1 (16.7)	5 (83.3)	3.39 (0.53–65.66)		0.72 (0.04–19.75)
Diabetes	No	65 (41.9)	90 (58.1)	−	.050	−	.161
	Yes	1 (8.3)	11 (91.7)	7.94 (1.49–147.02)		5.85 (0.71–148.76)

Note: Boldface indicates a statistically significant result (P < .05).

Abbreviations: CI = confidence interval, OR = odds ratio, SD = standard deviation, TSH = thyroid stimulating hormone.

3.2. The cutoff value of prolactin for detecting pituitary adenoma

Most of the patients with normal pituitary MRI (90.9%), microadenoma (81.6%), and ESS (85.7%) had low prolactin levels ≤ 90 ng/mL. While 17.1% of microadenomas and 61.1% of macroadenomas had mild hyperprolactinemia (between 90.1 and 200 ng/mL) or were in the border zone (Fig. 4). Only 6 (33.3%) macroadenoma patients had high hyperprolactinemia (>200 ng/mL) and were statistically significant in univariate and multivariate analyses (OR: 170.00; 95% CI: 27.98–3353.44, P < .001) (Table 3).

Figure 4.

Frequency distribution of pituitary MRI findings according to serum prolactin concentration. MRI = magnetic resonance imaging.

Table 3.

Association between prolactin level and pituitary adenoma.

Variables	Subgroup	Prolactin level (ng/mL)			Univariate analysis		Multivariate analysis
		Low (≤90)	Border* (90.1–200)	High (>200)	OR (95% CI)	P-value	OR (95% CI)	P-value
Pituitary Magnetic resonance imaging	Normal	60 (90.9)	5 (7.6)	1 (1.5)	−		−
	Microadenoma	62 (81.6)	13 (17.1)	1 (1.3)	2.26 (0.85–6.73)	.118	2.26 (0.85–6.73)	.118
	Macroadenoma	1 (5.6)	11 (61.1)	6 (33.3)	170.00 (27.98–3353.44)	<.001	170.00 (27.98–3353.44)	<.001
	Empty Sella Syndrome	6 (85.7)	1 (14.3)	0 (0.0)	1.67 (0.08–12.34)	.660	1.67 (0.08–12.34)	.660

Note: Boldface indicates a statistically significant result (P < .05).

Abbreviations: CI = confidence interval, OR = odds ratio.

^*Mild hyperprolactinemia, a non-definitive concentration between prolactinoma and nonfunctioning pituitary adenoma, is defined as 90 to 200 ng/mL, and patients with this range are considered border zone cases.

The area under the ROC curve for predicting pituitary adenoma was 0.640 (95% CI = 0.563–0.713; P = .0010), with a sensitivity and specificity of 40.59% and 83.33%, respectively (Fig. 5). Prolactin levels in ROC curve analysis indicated considerable accuracy in predicting the presence of pituitary adenoma. The best criteria cutoff value was more than 38.71 ng/mL, with a sensitivity of 77.23% (CI: 67.8% to 85.0%) and specificity of 40.91% (CI: 29.0% to 53.7%).

Figure 5.

The prolactin threshold level’s receiver operating characteristic curve analysis demonstrated a moderate degree of accuracy in predicting the presence of a pituitary adenoma; the area under the ROC curve was 0.640 (95% CI = 0.563–0.713; P = .0010), with sensitivity and specificity of 40.59% and 83.33%, respectively. CI = confidence intervals, ROC = receiver operator characteristic.

4. Discussion

This study evaluated the relationship between prolactin levels and pituitary adenoma on MRI in Sudi females with abnormally verified prolactin levels. To the best of our knowledge, this is the first study in Saudi Arabia to examine whether prolactin levels can predict the presence of pituitary adenoma in female patients with hyperprolactinemia after excluding the known etiology. Our findings indicate that a prolactin level more than 38.71 ng/mL can predict the presence of a pituitary lesion with moderate accuracy.

Hyperprolactinemia is frequently caused by excessive prolactin production in prolactinomas. However, parasellar or intrasellar masses can also impair dopamine flow, resulting in increased blood prolactin levels. Elevated prolactin levels may suggest a hypothalamic-pituitary lesion, trauma, surgery, radiation, skull fracture, or internal carotid artery aneurysm.^[17] Furthermore, estrogens can cause hyperprolactinemia, although the effect of oral contraceptives on prolactinoma development is still debatable with varied results.^[18] Furthermore, primary hypothyroidism, intracranial hypotension, stress, and physiological variables such as exercise, high-protein meals, and alcohol intake can also result in hyperprolactinemia.^[17]

The mean age in our research was 33.6 ± 9.5 years, consistent with the result of Al-Futaisi et al, which indicated an overall mean age of 32 ± 12 years in patients with hyperprolactinemia.^[19] Additionally, in this study, hyperprolactinemia was more common in patients aged 30 to 39 years (37.5%), followed by those aged 20 to 29 years (35.1%). Our findings are consistent with prior research from Saudi Arabia, such as Mahzari et al, who reported that hyperprolactinemia was detected more commonly in individuals aged 21 to 30 years (42.6%) and 31 to 40 years (24.1%).^[20] Moreover, our findings showed no age differences between patients with and without pituitary anomalies, consistent with the findings of Varaldo et al.^[1] It is important to remember that pituitary adenomas predominantly affect young, economically active individuals, for whom diagnostic delays lead to a loss of productivity. These findings underscore the importance of increasing knowledge about these curable illnesses to reduce the negative consequences of late diagnosis.^[8] Since our study included a limited number of patients, the results should be interpreted cautiously, as patient phenotypic variation may impact prolactin levels across different age groups and etiologies.^[21]

In this study, the most common symptom was visual field abnormalities (29.8%), followed by dizziness and headache (17.9%), amenorrhea (22.0%), and infertility (9.5%). These findings are consistent with those of prior research from Oman. Futaisi et al found that the most common symptoms were headache (59.8%), visual field abnormalities (35.7%), galactorrhea (23.2%), and exhaustion (17.0%). In addition, most of the women (57.0%) had atypical menstrual cycles.^[3] Identifying the pathological state in symptomatic patients is crucial, regardless of the presence of macroadenoma, as increased monomeric prolactin levels are a primary concern.^[22] Furthermore, hypothyroidism was more common in non-pituitary adenoma cases in this study. This finding was similar to those reported by Bayrak et al and Leca et al, who showed that concomitant thyroid disease was unusual among hyperprolactinemia cases.^[23,24]

The probability of encountering a pituitary adenoma increases with elevated serum prolactin levels.^[17] Generally, hyperprolactinemia correlates positively with the tumor dimension, whereby microadenomas demonstrate serum prolactin levels ranging from 94 to 188 ng/mL. In contrast, macroprolactinomas frequently present prolactin concentrations exceeding 235 ng/mL. However, the phenomenon known as the “hook effect” necessitates consideration in patients diagnosed with substantial pituitary adenomas, particularly when utilizing outdated assay techniques.^[14] Hyperprolactinemia resulting from hypothalamic dysfunction or compression of the pituitary stalk is generally observed at < 94 ng/mL.^[24] In this study, brain MRI identified microadenomas in 77 (46.1%) patients, macroadenomas in 17 (10.2%), ESS accompanied by pituitary microadenoma in 7 (4.2%), and normal findings without adenoma in 66 (39.5%). Our analysis showed that prolactin levels were higher among patients with pituitary adenoma than those with normal MRI (Mean ± SD: 53.0 ± 32.4 vs 85.5 ± 74.4 ng/mL) and were a predictor for pituitary adenoma. Furthermore, regression analysis showed a statistically significant association between hyperprolactinemia and the presence of macroadenoma in pituitary images; however, no correlation was observed between prolactin levels and microadenoma and ESS. In a parallel study by Rand et al, involving 74 premenopausal women with heightened prolactin levels, the association between prolactin concentrations and microadenomas was evaluated. Their findings indicated microadenomas in 51.3% of subjects, macroadenomas in 8.1%, and infundibular gliomas in 39.2%. The size of adenomas correlated with prolactin levels, suggesting that increased prolactin concentrations were linked to the detection of MRI pituitary abnormalities, as mentioned in our results.^[10] Varaldo et al examined prolactin levels as a prognostic indicator of pituitary lesions in 139 subjects. They discovered that 76.3% of the patients demonstrated pituitary pathology, with microlesions present in 69.8% and macrolesions in 25.5%. Additionally, prolactin measurements exhibited modest predictive accuracy for identifying pituitary abnormalities.^[1] However, longitudinal data are lacking in the present study, and an analysis of prolactin fluctuations before, during, and after treatment could elucidate whether prolactin correlates with MRI-detected adenoma growth and therapeutic response, rather than merely reflecting adenoma status. Additional clinical studies may reveal whether prolactin levels change throughout therapy and whether such variations are related to the clinical treatment results.

Moderate hyperprolactinemia is often recommended as a prerequisite for pituitary imaging; however, constraints related to accessibility and financial implications can render this approach unfeasible.^[12,25] The absence of a definitive diagnosis does not necessarily indicate a critical issue; however, initiating adjunctive therapy with dopaminergic agonists may adversely influence long-term outcomes by exposing patients to potential side effects and postponing the identification of pituitary macroadenoma lesions.^[1,26] Conversely, Souter et al reported a notably reduced prevalence of pituitary abnormalities, with 60.9% of MRIs yielding negative results.^[12] This study was distinctive in that it exclusively examined individuals with low or mild-to-moderate hyperprolactinemia based on a single measurement, which was subsequently replicated on the second occasion. Thus, it is conceivable that some of the participants assessed in such studies may not have been “truly” hyperprolactinemic, and the reported levels may have normalized upon serial evaluation.^[27] Our findings are consistent with those of other studies, including Varaldo et al.^[1] In this context, even when considering the subset of patients with low or mild-to-moderate hyperprolactinemia within our cohort, the proportion of patients with negative MRI results was still restricted to 39.5% of the female population.

Individuals with hyperprolactinemia must undergo pituitary imaging.^[24] Younger individuals with prolactin levels between 47 and 189 μg/L may have nonfunctional pituitary adenomas, whereas those with prolactin levels above 204 μg/L are more likely to have macroadenomas.^[24,28] However, the literature suggests variable thresholds for diagnosing prolactinomas, especially in the ambiguous zone (between 25 and 200 µg/L), which may easily lead to diagnostic errors.^[24,28,29] This emphasizes the complexity of distinguishing between functional and nonfunctional adenomas, or macroadenomas and microadenomas, based on prolactin levels. While Herlihy et al found no clear link between prolactin hormone levels and prolactinoma prevalence,^[30] other reports have found a strong correlation. For example, Leca et al discovered a substantial association between baseline serum prolactin levels and tumor growth, indicating good diagnostic accuracy. ROC curve investigation demonstrated an accurate distinction between micro- and macro-adenomas at a cutoff value of 204 ng/L with a sensitivity and specificity of 93.2% and 89.1%, respectively.^[24] In another report by Kawaguchi et al, the recommended prolactin level for diagnosing prolactinoma and nonfunctioning adenomas was 38.6 ng/L.^[14] According to Varaldo et al, prolactin concentrations exhibit a moderate level of accuracy in forecasting the presence of pituitary disorders. The optimal thresholds identified were >25 ng/L and >44.2 ng/L for males and females, respectively.^[1] Kyristi et al suggested a prolactin threshold of 85.2 ng/mL for screening pituitary adenoma in patients with polycystic ovarian syndrome,^[31] while Kim et al recommended a cutoff value of > 52.9 ng/mL.^[11] Osorio et al discovered a substantial association between tumor volume and blood prolactin levels in 219 prolactinoma patients, with a significant increase in prolactin levels every 1-cm3 tumor volume, especially in female patients.^[32] In the current study, ROC curve analysis yielded an AUC of 0.640 (moderate accuracy), with a sensitivity of 40.59% and specificity of 83.33%, establishing an optimal cutoff of >38.71 ng/mL for pituitary adenoma diagnosis, which was lower than that in previous reports. Notably, this lower threshold may result from the inclusion of cases of normal pituitary glands in our MRI evaluations and lower prolactin levels (>90 ng/L) in several patients. Overall, these thresholds allow providers to use an additional data point when determining the likelihood of hyperprolactinemia and to increase confidence when optimizing preoperative planning for patients whose prolactin level or volume exceeds the threshold, indicating a higher risk of pituitary adenoma. On the other hand, low prolactin levels do not rule out the possibility of a pituitary tumor or adenoma, even if it is rather large, due to the aforementioned stalk effect. In this context, prolactin levels appear to be more beneficial as a rule-in rather than a rule out test, and given this, it seems to be safer to perform an MRI of the sellar region in all patients with confirmed hyperprolactinemia.

4.1. Study strengths

The strength of this study is the serial sampling-based exclusion of patients with hyperprolactinemia, which eliminates cases of venipuncture stress. The patient population comprised people with a variety of conditions, not simply infertility. Therefore, the sample was representative of most endocrinologists’ clinical practices.

4.2. Study limitations

The main limitation of this study is its reliance on secondary data, the quality of which may vary due to variances in documentation, data integrity, and record-keeping practices. Furthermore, the study’s retrospective nature and limited sample size may have resulted in inherent bias. Excluding records with insufficient data may have resulted in selection bias in the analysis. Other limitations include the fact that it was only performed on females, and not all brain MRIs were conducted at the same centers. Other factors, such as tumor size, treatments, and pathological findings, which may influence prolactinoma diagnosis, were not investigated in this report. Furthermore, the study did not examine macroprolactinemia, and individuals without clinical symptoms were excluded from therapy even if their serum prolactin levels were high. Future research should involve a larger sample size and extended follow-up to address these limitations and to yield more robust findings for patient treatment.

5. Conclusion

Pituitary imaging should be performed in all cases of hyperprolactinemia without an event etiology, even if the condition is mild, due to the high prevalence of pituitary anomalies found in female patients with hyperprolactinemia at serial sampling. Our findings indicate that female patients with prolactin levels more than 38.71 ng/mL may require brain imaging to detect pituitary adenomas. To improve the clinical outcomes of these patients, a multidisciplinary approach to diagnosis, treatment, and follow-up is required.

Author contributions

Conceptualization: Nasher Alyami, Ghazlan Alhenaki, Salem Al Atwah, Nawras Alhenaki, Fatema Smaisem, Asmaa Alotaibi, Joud Abu Risheh, Mustafa Smaisem, Abdulmalik Alhenaki, Sultan Alanazi, Maram Alshammeri, Dana Alsayed, Sarah Musallam, Faisal Ahmed.

Data curation: Nasher Alyami, Ghazlan Alhenaki, Salem Al Atwah, Nawras Alhenaki, Fatema Smaisem, Dana Alsayed, Arwa Wadaan, Faisal Ahmed.

Formal analysis: Nasher Alyami, Ghazlan Alhenaki, Nawras Alhenaki, Fatema Smaisem, Asmaa Alotaibi, Joud Abu Risheh, Abdulmalik Alhenaki, Sultan Alanazi, Maram Alshammeri, Sarah Musallam.

Funding acquisition: Nasher Alyami, Ghazlan Alhenaki, Salem Al Atwah, Fatema Smaisem, Asmaa Alotaibi, Joud Abu Risheh, Mustafa Smaisem, Sultan Alanazi, Arwa Wadaan, Sarah Musallam, Faisal Ahmed.

Investigation: Nasher Alyami, Ghazlan Alhenaki, Salem Al Atwah, Nawras Alhenaki, Joud Abu Risheh, Mustafa Smaisem, Abdulmalik Alhenaki, Sultan Alanazi, Maram Alshammeri, Dana Alsayed, Arwa Wadaan, Sarah Musallam, Faisal Ahmed.

Methodology: Nasher Alyami, Ghazlan Alhenaki, Salem Al Atwah, Nawras Alhenaki, Fatema Smaisem, Asmaa Alotaibi, Joud Abu Risheh, Mustafa Smaisem, Sultan Alanazi, Maram Alshammeri, Dana Alsayed, Arwa Wadaan, Sarah Musallam, Faisal Ahmed.

Project administration: Nasher Alyami, Nawras Alhenaki, Asmaa Alotaibi, Mustafa Smaisem, Abdulmalik Alhenaki, Maram Alshammeri.

Resources: Nasher Alyami, Ghazlan Alhenaki, Salem Al Atwah, Joud Abu Risheh, Mustafa Smaisem, Maram Alshammeri, Dana Alsayed, Arwa Wadaan, Sarah Musallam, Faisal Ahmed.

Software: Nasher Alyami, Salem Al Atwah, Fatema Smaisem, Asmaa Alotaibi, Joud Abu Risheh, Mustafa Smaisem, Abdulmalik Alhenaki, Dana Alsayed, Faisal Ahmed.

Supervision: Nasher Alyami, Ghazlan Alhenaki, Nawras Alhenaki, Fatema Smaisem, Asmaa Alotaibi, Joud Abu Risheh, Mustafa Smaisem, Sultan Alanazi, Maram Alshammeri, Arwa Wadaan.

Validation: Nasher Alyami, Ghazlan Alhenaki, Nawras Alhenaki, Asmaa Alotaibi, Joud Abu Risheh, Abdulmalik Alhenaki, Faisal Ahmed.

Visualization: Nasher Alyami, Salem Al Atwah, Fatema Smaisem, Joud Abu Risheh, Mustafa Smaisem, Maram Alshammeri, Dana Alsayed, Arwa Wadaan, Faisal Ahmed.

Writing – original draft: Nasher Alyami, Ghazlan Alhenaki, Salem Al Atwah, Fatema Smaisem, Asmaa Alotaibi, Joud Abu Risheh, Mustafa Smaisem, Abdulmalik Alhenaki, Sultan Alanazi, Faisal Ahmed.

Writing – review & editing: Nasher Alyami, Nawras Alhenaki, Fatema Smaisem, Asmaa Alotaibi, Abdulmalik Alhenaki, Sultan Alanazi, Maram Alshammeri, Dana Alsayed, Arwa Wadaan, Sarah Musallam, Faisal Ahmed.