This study identifies a value of 5 IU/L as a possible new LH threshold after GnRH analogous assay with triptorelin, in the diagnosis of CPP in girls with thelarche onset before 8 years old. Only a few studies have analyzed the LH level after a triptorelin test for the diagnosis of PP and we believe that our data may be useful for a better definition of this test. As a matter of fact, the proposed cut-off value of 5 IU/L corresponds to the value suggested by the Italian Society of Pediatric Endocrinology and Diabetology guidelines on the basis of the 2009 consensus statement on the use of gonadotropin-releasing hormone analogues in children [14], confirming the validity of our result.

It is well-known that the gold standard for the diagnosis of central PP is represented by LH levels measurements after a stimulation test. However, different cut-offs to set the diagnosis have been proposed according to the test used, the time of assessment and the population analyzed [14].

So far, few data aiming to a standardized interpretation of the subcutaneous triptorelin assay have been presented in literature, and a great variability has emerged among assays. A retrospective study proposed an LH post stimulus cut-off of 6 IU/L collecting serial blood samples at baseline and at 30, 60, 90, 120 minutes after the triptorelin stimulus [15]. On the contrary, a prospective case-control study tested 46 girls with premature thelarche using both intravenous GnRH test and subcutaneous triptorelin assay. Gonadotropin levels were measured at 0, 3 and 24 h after the stimulus, proposing a cut-off of 7 IU/L (with immunofluorometric assay - IFMA) or 8 IU/L (with electrochemiluminescence immunoassay - ECLIA) to differentiate premature thelarche from central PP [7]. This study also highlighted the usefulness of the 24 h estradiol dosage in identifying pubertal evolution in borderline cases. A third prospective study has recently proposed a cut-off of LH post-stimulus measured at 180 minutes of 3.4 IU/L [16]. Finally, a Korean study suggested a peak of stimulated LH level ≥4.5 IU/L at 120 min after triptorelin injection as indicative of CPP [17].

Although all these studies have adequately demonstrated the diagnostic value of triptorelin, there is no consensus regarding the appropriate blood sampling intervals and the time for LH peak determination. This is due to different causes, and first of all, to pharmacokinetic differences in the LH response after intravenous gonadorelin administration, compared to the triptorelin test following subcutaneous injection. Consequently, the time required for the gonadotropin to reach its peak concentration depends on the type and on the modality of administration of the stimulus used. Secondarily, the drug levels following the subcutaneous injections may be influenced by the injection volume, the tissue distribution of proteolytic enzymes, and the regional variation in blood flow. Moreover, triptorelin acetate is known to have a longer half-life and a stronger affinity with the GnRH receptor compared to gonadorelin. As a final point, triptorelin manages to increase the plasma concentration of GnRH for a longer period compared to gonadorelin, and therefore the LH level should be measured over a longer period of time in girls with suspected CPP. Our cut-off results are similar to those proposed by other studies but our method involves a 4-hour control, hence an exact comparison is not possible, given that previous studies proposed cut-off values measured 2 or 3 h after the stimulus. Our Centre was the first to use the triptorelin test in humans, using 0.1 mg/m2 of triptorelin subcutaneously and analysing blood samples for LH, FSH, and estradiol 4 h after the injection, based on the data of a previous study involving five healthy adult volunteers, which demonstrated that gonadotropins show its peaks 4 h after triptorelin administration [18]. An LH value of 15 IU/L, dosed 4 h after the subcutaneous triptorelin stimulus, has been routinely used as cut-off for over two decades to diagnose CPP, without any reported missed diagnoses. The value of 15 IU/L derived from a comparison of post-stimulus LH levels after LHRH and triptorelin stimulus in a cohort of girls (n = 108) with CPP compared with other girls with isolated premature thelarche without activation of the HPGA (unpublished data). Lately, it has been noticed that some patients with clinically progressive CPP showed an LH value below the diagnostic threshold and needed a second assay after a few months to confirm the diagnostic hypothesis. The method used for the assay in our Centre was a chemiluminescence immunoassay, which has remained unchanged for the last twenty years; the variation in the progression of the PP was therefore independent from the method used. Hence the decision to review the cut-off for this test. In fact, these girls underwent a watchful wait approach through clinical follow-up, which in some cases led to a delay in the beginning of the replacement therapy.

The decision whether to treat or not a girl for CPP must be taken individually after assessing both the age of onset of secondary sexual characters and the evidence of pubertal progression during follow-up. The degree of bone age advancement and uterine dimensions on pelvic ultrasound may also be helpful in the diagnostic process and in the decision to start the treatment. The most notable long-term complications of an untreated CPP are represented by growth impairment and by a final height below the predicted familiar target [19], as well as by several psychological consequences of premature menarche, which, despite being often difficult to analyse, can persist over time, modifying self-perception, even in adulthood. Since puberty and adolescence are particularly vulnerable periods for the emotional chances they bring on, another reason for a rapid treatment of CPP lies in a psychosocial discomfort, which might lead to precocious sexual activity and to the risk of early pregnancies [10, 19]. In fact, a prompt treatment allows to preserve the growth potential and to reach the familiar height target avoiding psychological issues.

Our data show that other useful parameters to assess the diagnosis of CPP, such as the uterine longitudinal diameter, the bone age progression and the estradiol levels, provide useful indicators from a clinical viewpoint but are not decisive for establishing a diagnosis. Actually, our data confirm that pelvic ultrasound alone is not indicative of CPP and should be combined with clinical and laboratory tests to maximise its diagnostic value in the diagnosis of CPP [20]. In fact, although uterine and ovarian measurements are significantly higher in girls with CPP, there is a significant overlap of normal prepubertal and early pubertal values [21]. Moreover, none of the girls enrolled in this study shows evidence of HPGA activation at the first assay, consequently the role of pelvic ultrasound for them is even more limited. Finally, we did not measure the 24-hour estradiol value because the double sampling appeared to be too expensive, long and above all not easy for our patients.

Interestingly, our population presents an advancement, although not significant, of bone age in comparison with the chronological age. This data may be of difficult interpretation. It may be related to an activation of the hypothalamic-pituitary-gonadal axis, or to be a consequence of being overweight. In fact, it is well-known that obesity could lead to a bone age progression due to the aromatization of androgens into estrogens in the adipose tissue [22, 23].

In our cohort we enrolled some overweight and obese girls, and we noticed that these patients developed thelarche earlier compared with the normal weight girls, with an average advancement of 0.45 years (7.26 ± 0.80 years in normal weight girls vs 6.81 ± 0.92 in obese girls, p < 0.05). These data confirm what previous literature has already shown: obese patients tend to begin the pubertal developmental process earlier than normal weight girls do. Unfortunately, data concerning the relationship between central PP and patients’ nutritional status are lacking and it is difficult to establish whether a higher BMI influences the earlier onset of puberty or, on the contrary, whether CPP could be partially responsible for the excessive weight gain [10].

Beyond post-stimulus LH values, other variables resulted significantly different amongst girls with a precocious activation of the HPGA, such as LH-to-FSH ratio and growth velocity. For what concerns LH-to-FSH ratio, in literature values between 0.6 and 1 have been proposed as cut-offs to distinguish a progressive CPP from a non-evolutive one [24]. Our study demonstrates that this ratio could be useful to support the diagnosis, even though the main biochemical diagnostic value remains the LH levels. Remarkably, even growth velocity represents a significant parameter to detect progressive CPP. Yet since this variable must be evaluated at least in a 6-months period, it becomes useful only during the auxological follow-up and not at the first evaluation of the patient. Therefore, growth velocity might be something to take into consideration during clinical follow-up without necessarily repeating a stimulus test.

This study employs a retrospective design and includes patients who did not exhibit HPGA activation in the initial assay. This distinction might account for some discrepancies between our findings and those of other studies, as we hypothesise that patients with clear pubertal activation would demonstrate significant bone age advancement and larger uterine longitudinal diameters. Despite these limitations, our objective is not to compare patients with CPP and isolated premature thelarche but rather to identify the most appropriate LH value that could obviate the need for repeated GnRH analogous tests during follow-up in girls previously exhibiting only partial HPGA activation.

Even though further studies are needed to better define a universal cut-off for LH post-stimulus with GnRH analogues, the present article allows to identify a potential new threshold for LH levels measured 4 h after the subcutaneous triptorelin administration of 5 IU/L in a cohort of Italian girls with thelarche onset before 8 years old. Other parameters that appeared to be statistically significant and that might support the activation of the HPGA are the LH/FSH ratio and the growth velocity; yet, when considered together with the LH value, they did not significantly enhance the diagnostic capability. The other variables we hypothesised as useful in supporting the diagnosis of PP, such as the uterine longitudinal diameter, the bone age advancement and estradiol levels, are not statistically significant to set the diagnosis.