In this study, we found that the peak LH, peak FSH, LH/FSH, bone age, and Δage were significantly higher in the ICPP group when comparing to the PT group. The pituitary length, height, and PV of the ICPP group were also significantly higher than those of the PT group, as demonstrated by the MRI. The PV might be a predictive marker for ICPP, with a sensitivity of 54.10% and a specificity of 72.20% at the cutoff value of 196.01 mm3. According to univariate analysis, PV was significantly and positively associated with peak LH, LH/FSH, age, bone age, and BMI. However, after stepwise multivariate regression analysis, the bone age and peak LH were the only significant predictors for PV. Taken together, pituitary MRI should be combined with clinical and laboratory tests to maximize the diagnostic value of PV for PP.
It is difficult to distinguish ICPP and PT in the clinical diagnosis, but their differences in prognosis are extremely obvious. PT does not require treatment while ICPP has lasting adverse effects, such as short adult stature, and needs to be treated with long-acting luteinizing hormone releasing hormone (LHRH) agonists[9, 11, 19]. Therefore, it is of great significance to separate the two types of PP.
Due to the disadvantages of the GnRH stimulation test, MRI has become an alternative method to evaluate the pituitary gland, and to be performed in many tertiary care centers to exclude brain abnormalities in girls diagnosed with CPP. In this study, to examine the effectiveness of MRI, we compared the MRI results among the PT, ICPP, pubertal, and control groups. The results showed significant differences in the PV and pituitary height. Patients in the ICPP and pubertal groups had the largest PV and pituitary height, followed by patients with PT. In contrast, the control group had the smallest PV and pituitary height. Moreover, patients with larger pituitary glands had higher levels of LH and FSH.
CPP is accompanied by significant changes in the shape and size of the pituitary gland; patients with ICPP have higher pituitary grade, height, and Sagittal cross-sectional area compared to age-matched normal subjects[14, 15, 20]. Besides, it has been showed that the PV of CPP children were higher, and the upper pituitary surface in CPP patients appeared convex in a higher proportion[20]. However, no significant difference in the pituitary length, width, and volume was observed between the control, PT, and ICPP groups. In our study, the results demonstrated that PV and pituitary height in ICPP group were higher than those in control group. In addition, we first reported that PV and pituitary height of ICPP group were higher than those of PT group, and similar to those of pubertal group. The ratio of convexity in the pituitary gland increased in the ICPP group when comparing to the control and PT groups, even though the difference was not obvious.
The stepwise multivariate regression analysis in this study built a model (R2 = 0.159) that explained only 15.9% of the variability in PV, even though PV showed a meaningful difference between the four groups. This phenomenon might be resulted from the following reasons. First, the lack of variables for generating a proper regression model, since many factors such as age, nutrition, race, sex, and pubertal stage would influence PV[21-23]. Second, cells secreting gonadotrophs (LH and FSH) account for only 10% of the anterior pituitary cells, and are distributed diffusely throughout the anterior lobe without much effects on PV. Finally, previous study showed that pituitary hypertrophy arose from the stimulation of growth-hormone-producing cells (somatotrophs) in the pituitary gland. The pituitary enlargement in puberty was also possibly correlated with the serum levels of somatomedin C[14].
Besides, our data indicate that a larger PV is correlated with a higher peak LH value, but not peak FSH value. It has been showed that a larger PV is associated with a higher FSH production but is independent of pubertal development in normal subjects[16]. However, in this study, the positive association between LH and PV is linked to the pubertal development. Previous study has demonstrated that the gradually elevated FSH had been already ongoing for several years prior to the onset of puberty[24]. We therefore propose that increased LH levels are associated with a larger pituitary gland during early puberty, whereas the association between peak FSH and PV is not obvious in the early adolescence.
We also conducted the univariate logistic regression analysis with each MRI parameter as an independent variable to evaluate the diagnostic value of pituitary MRI in CPP girls. The PV might be a predictive marker for ICPP, with a sensitivity of 54.10% and a specificity of 72.20% at the cutoff value of 196.01 mm3. However, these results did show a reliable predictor. In general, the cutoff values obtained by ROC curves showed low sensitivity and specificity, even for the most potential predictor, PV.
This study had several limitations. First, even though this was a retrospective study, the research goal was not accurately set at the stage of data collection. Second, we did not collect basal LH, basal FSH, E2, and Tanner stage. Third, only one experienced radiologist performed all imaging studies. Finally, although the sample size of this study was larger than that of previous study, the sample size was still relatively small considering the incidence of precocious puberty.
In conclusion, the PV in the ICPP group was significantly higher compared with the PT and control groups, but there was no reliable cutoff value to distinguish ICPP from PT. Pituitary MRI should be combined with clinical and laboratory tests to maximize the diagnostic value of PV for PP.