No pubertal growth spurt, rapid bone maturation, and menarche post GnRHa treatment in girls with precocious puberty

Introduction

Gonadotropin releasing hormone agonist (GnRHa) treatment is standard of care for central precocious puberty (CPP) [1], [2], [3], [4]. The treatment slows or halts pubertal progression, slows the rate of bone maturation, suppresses gonadotropins, and usually decreases height velocity (HV) to normal pre-pubertal levels. Predicted adult height (PAH) continues to improve while on treatment, and then slowly decreases post treatment as rapid bone maturation resumes.

The decision regarding when to stop treatment needs to be individualized [5] and is based on multiple factors. These include the child reaching an appropriate age for mid-puberty to resume and a PAH reasonable for genetic potential, with caution that PAH may decrease once treatment is discontinued [6]. After discontinuation of GnRHa, the pubertal growth spurt typically does not resume, however growth will continue at a slower velocity until cessation of growth. This emphasizes the importance of not stopping treatment prematurely, as the remaining growth may be minimal. There are limited studies of growth and HV after discontinuation of GnRHa, especially as related to onset of menses [7]. We studied rate of bone maturation and growth after GnRHa treatment was stopped. We hypothesized that HV will not accelerate after treatment is discontinued, and rate of bone maturation will accelerate. These results are helpful to parents anticipating pubertal development, growth and menses post treatment, and provide important information for the physician when deciding how long to continue GnRHa treatment based on anticipated growth and menarche post treatment.

Materials and methods

Results

Twenty girls remained in the study until FH was reached. Baseline characteristics are shown in Tables 1–3. GnRHa treatment was stopped at an average age of 11.8 ± 0.6 years (range 10.6–13.2 years). Girls were treated for an average duration of 4.5 ± 2.4 years (range 1.5–10.4 years). The average reported age at onset of puberty was 5.8 ± 2.1 years (range 1.0–8.4 years). One girl was included in the study who had her onset of puberty estimated by parents at 8.4 years, however she had rapid progression of puberty, her BA was 2 years advanced, and her PAH was below her MPH.

Bone maturation

BA at discontinuation of treatment ranged from 11–14 years (average 12.5 ± 0.8 years). The BA at the onset and end of treatment both had a negative correlation with time to menarche post treatment (0<0.01). Girls with a younger BA at onset or end of treatment developed menarche later than girls with an older BA at either time point during treatment (p<0.01) (Figure 1). BA at end of treatment also had a negative correlation with the amount of growth accrued post treatment, such that the younger the BA at end of treatment, the greater overall growth post-treatment (p<0.01) (Figure 2). No one had menarche prior to a bone age of 12.5 year.

Figure 1:

Time to menarche post GnRHa treatment by bone age at start of treatment (A) and by bone age at end of treatment (B).

Figure 2:

Total growth post GnRHa treatment by BA at end of treatment.

The rate of change of BA advancement averaged 1.3 ± 0.15 prior to treatment, and decreased during treatment to 0.7 ± 0.3 years per chronological year (y/y). Post treatment, rate of BA change accelerated to 1.2 ± 0.8 y/y (range 0–3.0 y/y) (Table 2, Figure 3).

Figure 3:

BA/CA ratio from treatment to final height.

On average, treatment was stopped when CA was within 9 months of BA. Three girls had a BA >2 year beyond CA at end of treatment. They also had the 3 oldest BA at the end of treatment, but 2 girls surpassed their MPH (1 did not have MPH available).

Growth

Average HV on treatment was 4.8 ± 1.2 cm/year (range 2.9–7.6). All girls continued to grow from end of treatment to menarche with an average of 4.7 ± 3.7 cm, with an average height velocity of 3.2 ± 2.0 cm/year. On average, they grew an additional 4.6 ± 2.1 cm from menarche until FH, with a HV of 2.4 ± 1.9 cm/year during that time (Figure 4). Only 2/20 girls, had an increased height velocity (>6 cm/year) after treatment was discontinued, similar to a pubertal height velocity. Growth after discontinuation of treatment ranged from 4.6 to 12.4 cm in girls who stopped GnRHa with a BA≥13 years, and 7.9–16.7 cm in girls who stopped GnRHa with a BA<13 years.

Figure 4:

Height velocity from treatment to final height – individual patients shown. Solid bold line is average.

Discussion

Time to menarche is a great concern for many children and their families. The range of onset of menarche post GnRHa treatment is consistent across studies [12], [13], [14], [15] from 3 months to 3 years. The present study confirms that range but shows for the first time the relationship between BA and menarche. We found that the younger the BA was prior to treatment or the younger the BA was at the end of treatment, predicted a longer time to menarche. No girls had menarche prior to BA of 12.5 years. When girls were compared by onset of CPP <6 years vs. >6 yo, the younger girls had menarche on average one year later than the older girls, but at a similar BA. This supports the importance of BA in predicting menarche. There were no other differences between these 2 groups, supporting the robustness of this population with truly rapidly progressing CPP, rather than a mixed population of early normal puberty and CPP. This is one of the strengths of the population, since many publications likely include some early normal puberty when girls close to 8 years old are studied.

Rate of BA advancement prior to treatment in girls with CPP is accelerated. On GnRHa treatment, rate of advancement slows to less than 1 year per year when treatment is optimal, and therefore leads to increases in growth potential and FH [16]. Once treatment is stopped, we suspected that rate of BA advancement accelerates, and this is the first study with data to support that pathophysiology. Prior to GnRHa treatment rate of BA advancement in this population on average was 1.3 years per chronological year [5]. This decreases to an average of 0.7 years per year on treatment. Post treatment, rate of BA advancement accelerates back to pretreatment rapid maturation averaging 1.2 years per year. This is expected since whatever process causes rapid bone maturation in CPP, is not cured by GnRHa treatment, but rather slowed down during treatment. Lazar et al. [17] reported data from 115 girls post-GnRHa treatment. They describe greater height gain post treatment in girls who had onset of puberty prior to 6 years old compared to older girls. Final height in the older groups was less than PAH at end of treatment. They measured BA every 12 months and report that BA at end of treatment significantly contributed to prediction of height gain post treatment. However, they do not report change in BA post treatment. They raised the question of whether post-treatment growth would have been greater if BA was closer to 11 at end of treatment. The present data support the opposite hypothesis, since rate of BA advance accelerates post treatment; stopping treatment at an older BA may lead to greater final height, achieved by more growth on treatment rather than more growth post-treatment. Shim et al. also looked at growth post-GnRHa treatment in 85 boys but compared only PAH at end of treatment to final height, and did not delineate BA or growth rate changes over the years post treatment [18].

On average, treatment was stopped when CA was close to BA. Adequate pubertal suppression with GnRHa theoretically corrects the discrepancy between BA and CA seen prior to treatment. The girls in this study who did not reach a CA within 2 years of BA prior to end of treatment, were treated until an older BA than the rest of the girls, which was likely necessary to achieve improved FH outcomes. This supports the likely the importance of duration of treatment for optimal height outcome. Of course, we cannot know what height the girls would have achieved if treatment had been discontinued earlier, but we have shown the rapid acceleration of rate of bone maturation post treatment so it would be unlikely they would have reached the same FH.

Previous studies showed that earlier onset of treatment, less BA advancement, and less delay from onset of CPP to onset of treatment have taller height outcomes as compared to MPH [6, 19], [20], [21], [22]. In the present study, girls with less BA advancement had greater growth post treatment. This is the first study to quantify growth and growth rate between end of treatment and menarche and between menarche and FH. This is very important for clinicians and families when making decisions about the timing for cessation of GnRHa treatment. Those decisions cannot be based on expectations for resumption of a pubertal growth spurt post treatment. Only 2 girls in this population had increased HV to >6 cm/year post treatment. All others grew at ≤5 cm/y after GnRHa was stopped. An interrupted pubertal growth spurt does not resume post GnRHa treatment for CPP.

The lack of rapid pubertal growth post treatment is likely related to growth plate senescence. Our understanding of growth plate senescence includes the effect of estrogen to first stimulate growth plate chondrogenesis and therefore linear growth, while simultaneously decreasing the number of resting zone chondrocytes, accelerating the rate of chondrocyte proliferation leading to earlier proliferative exhaustion with eventual growth plate fusion [23, 24]. The early exposure of estrogen in precocious puberty causes early growth plate senescence. Perhaps GnRHa treatment is usually timed to interrupt this process at the point when estrogen has already diminished enough chondrocytes, such that as bone aging continues during treatment and no further increase is possible post treatment. The older the BA was at the beginning of treatment, the less growth is accrued post treatment. We can interrupt the process of CPP causing rapid maturation of BA, but the process of senescence, although slowed during treatment, is not stopped, so we cannot regain all the growth potential that was lost. This is further support of continuation of GnRHa longer in those girls with late onset of treatment who have short height potential compared to MPH.

It cannot be determined from this study whether growth would have been different if treatment was stopped earlier in the girls with more advanced BA. However, in all girls HV was good throughout treatment, there was similar amount of growth post treatment, and height outcomes compared to MPH were similar across all girls. This suggests that the individual treatment decisions of when to stop GnRHa treatment led to the best expected outcomes.

The girls in this study were all treated until almost 11 yo or more in order to maximize height benefit as well has had their physical findings of puberty coincide with their peers. When the primary goal of treatment is prevention of early menarche, treatment is often discontinued between 10 and 11 yo, anticipating onset of menarche on average 18 months post treatment. While not the focus of this study, the cost-implications of continuing treatment should always be kept in mind [25]. Aside from cost-implications, however, the range of onset of menarche is 3 months to 2.5 years post treatment, similar to other published studies. Therefore, the older age at discontinuation in this cohort is unlikely to have affected the results of time to menarche or growth post treatment. It is always clinically relevant to inform families that a 10 yo girl who stops treatment may have menarche by 10 years 3 months, which would still be quite early after years of treatment to delay menarche”.

The major strengths of this study include following girls to FH with measures of growth, bone maturation and menses every 6 months for as long as 5–6 years post GnRHa treatment. This study also has girls who continued GnRHa until 13–14 years with good growth during the treatment interval, as well as post treatment. We found that 18/19 had FH within 3 inches of MPH, with more than half of the girls reaching or surpassing MPH. The wide range of age at onset of puberty allowed assessment of age as a factor contributing to timing of menarche.

The major limitations include the small number of participants, although this does show a real-world example from one center. All girls were offered participation, but not all girls wanted to continue to be followed every 6 months once treatment was discontinued, so there may have been some bias in those who continued coming. For example, those who did not continue follow up may have been obviously growing less and therefore not interested in continued bone age to assess potential growth. It is unlikely that those who preferred not to continue were those growing more rapidly, but that cannot be determined. Another limitation is that bone age at cessation of treatment was not randomized, so physician decisions to stop treatment could bias outcomes. However, it would not be ethical to randomize treatment cessation if physician assessment indicated possible benefit to continued treatment.

In conclusion, the present study quantitates growth after treatment with GnRHa in girls with CPP. In most girls, there is no resumption of pubertal growth spurt post GnRHa treatment. Time to onset of menarche post-GnRHa treatment is proportionate to BA at the start as well as the end of treatment, with no girls in the present study having menarche prior to BA of 12.5 years. Since time to menarche post treatment is multifactorial and not yet able to be predicted, this study adds some guidelines and suggests continuation of treatment beyond a bone age of 12.5 years in younger girls, where earlier menarche continues to be a concern. Rate of bone maturation accelerates post treatment. These factors are important in assessing potential height outcome and the decision regarding timing for cessation of treatment. This study will help clinicians give patients and families better estimates of growth and onset of menarche post treatment.