If Gonadotropin-Releasing Hormone Plus Growth Hormone (GH) Really Improves Growth Outcomes in Short Non-GH-Deficient Children, Then What?

Paul B. Kaplowitz, M.D., Ph.D.

Department of Pediatrics Virginia Commonwealth University School of Medicine Richmond, Virginia 23298

Address correspondence and requests for reprints to: Paul B. Kaplowitz, M.D., Ph.D., Department of Pediatrics, Virginia Commonwealth University School of Medicine, Richmond, Virginia 23298.

Since biosynthetic GH became widely available in 1985, pediatric endocrinologists around the world have been working to extend the use of GH beyond the less than 10% of short children with classic GH deficiency to as many categories of short children as possible. In the cases of Turner syndrome and chronic renal insufficiency, these efforts have met with sufficient success to have received approval from the FDA for those indications. The biggest challenge (certainly in terms of the number of patients who could potentially be helped) remains the group of short children who have no identifiable disorder of the GH/insulin-like growth factor axis, and no other endocrine, genetic, or organ system disorder. Many have called this group of patients "idiopathic short stature," although it is very heterogeneous and includes a large number of children with either constitutional growth delay, familial short stature, or both. Over the past 10–15 yr, a large number of groups have investigated the effects of various doses and duration of treatment with biosynthetic GH on adult height. While all of the three possible outcomes have been reported (increase in adult height, no significant effect, and even loss of predicted height), the recent report by Hintz et al. (1) of the final results of the large Genentech Collaborative Group study provides reason for both optimism and pessimism. The mean age of starting therapy was 10 yr, and after a mean duration of therapy of 6.0 yr for boys and 5.5 yr, both boys and girls had mean heights of 5–6 cm above the pretreatment heights predicted from the bone age. However, whereas some children achieved heights of 10 cm above predicted, about 20% of both boys and girls failed to improve on their predicted height. Furthermore, there did not seem to be any auxological parameter or blood test that predicted which children were most likely to obtain benefit. Another recent report, from Children’s Hospital of Buffalo, found that in 36 patients with idiopathic short stature treated with GH for a mean of 41 months, height gain was 6.8 cm for females (n = 6) but only 3.0 cm (n = 30) for males (2). The smaller benefit in males in this study may be due to the fact that the mean age at starting treatment in the latter study was 12.7 yr for boys and 12.2 yr for girls. Although much additional data will surely be published over the next few years, the final answer to the question "Does GH in non-GH-deficient short stature increase adult height?" will very likely be "Yes, in most cases, but not by very much."

In reviewing the literature on GH in idiopathic short stature, one gets the impression that no matter how long one continues treatment, most of the benefit on predicted height is achieved in the first 2 yr, or at least before the onset of the pubertal growth spurt. In one study from Austria, the entire improvement in predicted height was achieved during yr 1 and 2, with no further improvement for yr 3 through 5 (3). This is most likely due to a number of factors. First, the growth response to GH invariably wanes after 1 or 2 yr of therapy. Second, once puberty is well underway, GH-sufficient children produce high amplitude spikes of GH and high levels of insulin-like growth factor 1, making the effects of exogenous GH on growth difficult to separate from the effects of increased endogenous GH. Third, pubertal levels of sex steroids cause a rapid advancement of bone maturation, foreshortening the time left for exogenous GH to have any effect on growth, and some but not all studies indicate that giving GH to non-GHD children may accelerate the tempo of puberty and the rate of bone maturation. Thus, it is not surprising that the effects of GH alone in short children who are about to start puberty or are in early puberty have not been very impressive. To circumvent the problem of earlier than desirable epiphyseal fusion in such children, a number of investigators have tried to slow the advancement of bone maturation by using GnRH analogs to suppress sex steroid production. It was reasoned that if GnRH analogs could improve adult height in children with central precocious puberty, they might also work in short adolescents with normally timed puberty. GnRH analogs do, in fact, prolong the period of growth in short adolescents, but improvement in adult vs. predicted height has been disappointing (4), most likely because the benefit of slower bone age advancement is offset by slower than normal linear growth once sex steroids are suppressed.

Over the past 8 yr, several groups of investigators have reported attempts to improve growth outcomes in short normal children by combining GnRH analogs with GH, to reverse the slowing of growth observed in children treated with GnRH analogs alone, and many of these reports have previously appeared in this journal. Some have reported disappointing results, but others have reported improvements in predicted and/or completed height in the range of 5–10 cm. These studies are reviewed in detail in the study of Kamp et al. (5), in this issue of the journal, which reports encouraging news from The Netherlands for those hoping that this form of combined therapy will prove to be effective. During a 3-yr treatment period, for girls with either idiopathic short stature or intrauterine growth retardation started on treatment at a mean age of 11.4 yr, there was a gain in predicted height of 8.0 cm in treated vs. control subjects; for boys, who were started on treatment at a mean age of 12.2 yr, the gain was reported to be 10.4 cm. We must ask then if this new study is methodologically superior to many of those that have preceded it and, if so, in what ways?

The greatest advantage of the study by Kamp et al. (5), as the authors point out, is that it is the only trial of GnRH analog plus GH reported so far that used a randomized control group. Without a control group, the only measure of efficacy of treatment is the comparison between either adult height or predicted height at the end of therapy and predicted height at the start of therapy. Because height predictions based on bone age are less reliable in individuals than in large groups of patients, studies that have a small number of patients and rely only on height prediction are subject to error in estimating the treatment effect. The study by Kamp et al. (5) found that treatment for 3 yr was effective in improving predicted height whether one uses untreated controls or pretreatment height prediction as the basis for comparison. In the case of girls, the mean predicted height after 3 yr in controls actually improved 2.4 cm, so that the gain in predicted height in treated subjects was less than the 10.4 cm gain based on pretreatment height prediction. For boys, the opposite effect was seen, as height prediction in controls declined over 3 yr by 3.9 cm, increasing the gain in height prediction from 6.5 cm (compared with pretreatment height prediction) to 10.4 cm (compared with controls). Thus, although the magnitude of the treatment effect varies according to which comparison is used, the effect is large enough that we can conclude a real improvement has taken place. Why some similar studies have shown no such improvement is still not clear. For example, the study by Lanes and Gunczler (6) from Venezuela was similar in design, did use a control group (apparently not a randomized one), and found, after a mean of 21/2 yr of GnRH plus GH therapy in seven girls and three boys, no real increase in adult height compared with either pretreatment predicted height or the untreated controls.

Where the study by Kamp et al. (5) is lacking, as the authors readily admit, is that the subjects have not been followed through to completion of growth. Although in some studies gains in predicted height during treatment are sustained when adult height is achieved, in others the effect of treatment on adult height is considerably reduced when compared with the effect based on height prediction at the end of therapy. One variable that may explain some of the disparity from study to study is the length of the treatment period. For example, the study by Balducci et al. (7), in which seven of nine girls were treated for only 2 yr, showed a gain in predicted height after 2 yr of 4.4 cm (from 143.2 to 147.6 cm), whereas the mean completed height was a mere 1.4 cm better than the pretreatment predicted height. In the report by Pasquino et al. (8), also from Italy, the treatment period with GnRH plus GH was 4.6 ± 1.7 yr, and of the 10.5 cm gain in predicted height (from 146.3 to 156.8 cm), only 0.5 cm was lost when the girls were followed to completion of growth. The treatment period in the study by Kamp et al. (5) was intermediate between the above studies, and treated subjects at the time the study ended still had mean bone ages of 12–13 yr depending on the subgroup. Thus, one might guess that if there is rapid advancement of bone age, loss of improvement in adult height between discontinuation of therapy and completion of growth would result in an intermediate gain in height when compared with the two studies above. If it turns out that duration of treatment is a major predictor of outcome in combined GnRH plus GH therapy, the estimated expense of this form of therapy per centimeter of increased height will be extremely high as discussed below.

Whereas the adult height issue will hopefully be resolved in a follow-up report, there are other questions that should be raised about the study by Kamp et al. (5). The authors, in criticizing the controlled study of Municchi et al. (9) using GnRH vs. placebo, point out that "the study population was very heterogeneous, and therefore the results are difficult to interpret." To try to increase the number of subjects, however, the authors have not only used both girls and boys (who are analyzed separately in Table 1 but together in most of the figures), but also both ISS and IUGR subjects, who are analyzed mostly together. The only subgroup that is of a reasonable size is the ISS girls (n = 23). It is probably premature to draw any conclusions about outcomes for the boys based on treatment and control groups of only 2–4 subjects for ISS and IUGR boys or even 11 total subjects (5 controls and 6 treated) for IUGR and ISS boys combined. From a brief look at Table 1, one might well conclude that treatment of ISS boys resulted in a loss of over 4 cm of predicted height compared with controls. This is likely due to the table reporting only the pretreatment predicted height of the entire group of eight ISS boys entered into the study, not the separate means of the treated and untreated subgroups, which must differ substantially. This confusion would not exist if the groups were larger and/or the treated and control groups more comparable at baseline.

While we are on the subject of boys, one might well question the inclusion of subjects in the protocol whose mean predicted adult heights were 168.4 cm (for ISS boys) or 165.9 cm (for IUGR boys). Inclusion criteria were stated to be an actual height or a predicted adult height of less than -2 SD, which for Dutch children seems to be about 168.5 cm. Clearly, any child with a height of -2 SD and a delayed bone age will have a predicted height within the normal range. Although there is no universally accepted definition of "short" or "too short," a boy with a height prediction of 168 cm, while perhaps short by the Dutch standards used, may not be short enough in the minds of many of our readers to justify 3 yr of expensive therapy with two hormonal agents. Two other European studies of GnRH plus GH cited above have used girls whose initial mean height predictions were extremely short (143.2 cm and 146.3 cm as noted above) (7, 8). The fact that GnRH plus GH may increase adult height in boys who would attain a normal or near-normal height if they were just left alone should not imply that offering this treatment is appropriate.

This brings us to the next and most important point: Let us assume that subjects in the study of Kamp et al. (5) are followed until growth is completed and that the mean improvement in final height is in the range of 7–8 cm. Let us also assume that additional studies confirm that GnRH plus GH given for at least 3 yr in short children with normally timed puberty show treatment effects of similar magnitude. What then should we do about it? Let us first estimate the cost of adding an extra 8 cm for a short 12-yr-old boy living in the United States. I will use for this calculation the recommended maximum GH dose of 0.3 mg/kg·week. Looking at a growth chart, a short (-2.5 SD) boy treated from age 12–15 whose weight is average for height will have an average weight during the 3 yr of therapy in the vicinity of 32 kg. At about 32 x 0.3 {approx}10 mg GH per week x 52 weeks, the yearly cost of the 520 mg GH (at a cost of about $50/mg when provided by a home health care agency) comes close to $26,000/year. We then add to this the cost of the GnRH analog. At a local pharmacy, the cost of Depot Lupron at 7.5 mg, which is the recommended starting dose, is $632/dose. Given every 4 weeks, 13 doses per year are needed, adding up to $8200/year. If we combine the two and multiply it by 3 yr, the drug cost alone comes to just over $100,000. We should then add the cost of about nine office visits over 3 yr, yearly GnRH testing to confirm hormonal suppression, at least yearly bone age determinations and additional visits after treatment is stopped until growth is complete, which could add another $5,000. If one generously assumes an average 8 cm height gain, the estimated cost per extra centimeter of adult height is about $14,000. Let us now look at this scenario from the perspective of the child and family. One of the pleasures of treating a short child with GH is to monitor height increments every 3–6 months and communicate to the child and family the improvement in growth rate during therapy. However, with GnRH plus GH, growth in the study by Kamp et al. (5) was actually slower during the 3 yr of the study (7.0 + 5.4 + 4.9 cm = 17.4 cm) than in the untreated controls (7.4 + 7.0 + 4.7 cm = 19.1 cm). Of course, the point of the combined therapy is in the slower advancement in bone age that, while important, is invisible to the family. Thus, we should explain to the family at the outset that the benefits of therapy on adult height will only become apparent in the follow-up period, after treatment is stopped and the extra growth is added on while the epiphyses remain open. This is truly delayed gratification, and one can speculate that many children and their parents will be unwilling to wait so long to reap the benefits of combined therapy. We should also consider that many short 12-yr-old boys and girls will have reservations about a treatment which puts progression of puberty on hold for at least 3 yr, since for many such children, having the physical changes of puberty at the expected time may be as important as the additional growth they hope to achieve at least 5 yr down the road.

How will we as endocrinologists be able to justify this expensive form of therapy to those who will be asked to pay for it? There is already much resistance among insurers in covering the cost of GH for idiopathic short stature, which is understandable given the uncertain and, on the average, not very impressive increase in adult height achieved. If we request coverage for combination therapy that is even more expensive, and we are not very careful in selecting our patients, can we blame the insurers for balking? Ultimately, if there is to be any hope of getting this treatment covered by most insurers, guidelines will need to be established (hopefully by endocrinologists and not by insurance companies) as to how short the adult height prediction should be in boys and girls before we undertake treatment. Should it be 2 SD below the population mean, as the Dutch group decided when they started their study? Should we be more conservative and use a -2.5 SD or even a -3 SD cut-off? Should we factor in the target height of the parents? (This raises the knotty problem of whether short children of average or tall parents are more deserving of this therapy than are equally short children of short parents.) Should some measure of the emotional impact of short stature on the child be factored in? Obviously, there are no easy answers to these questions, but they deserve serious thought. Perhaps if organizations like the Lawson Wilkins Pediatric Endocrine Society can reach a consensus on where to draw the line so that the number of patients eligible for this therapy is not too great, there may be a chance of getting combined GnRH plus GH therapy accepted as treatment for severe idiopathic short stature.

One more point should be made on this subject: There may be another less expensive form of treatment on the horizon that can accomplish the goal of slowing bone age advancement and improving adult height in short normal children. It is now known from the study of rare cases with deficient estrogen production or action that estradiol and not testosterone is the critical factor in determining the rate of bone age advancement and the timing of epiphyseal fusion in males (10). Mauras et al. (11) have recently reported on the metabolic effects of the aromatase inhibitor anastrozole in teenage boys, and although no significant adverse effects were seen over a treatment period of 10 weeks, questions about the long-term safety of estrogen blockade on pubertal boys remain unanswered until larger and longer studies are completed (12). One recent case of a boy with precocious puberty due to a hamartoma who received anastrozole for 3 yr at a time at which his epiphyses were nearly fused, and achieved a higher than predicted adult height (164.4 vs. 158.4 cm), provides some hope for this approach (13). Whether these drugs are efficacious and safe in females is another critical question, but in another 3–5 yr we may have evidence that aromatase inhibitors are as good as or superior to the combination of GnRH plus GH. In the meantime, we can expect to see additional (hopefully larger and better controlled) reports of studies using GnRH plus GH in idiopathic short stature, and we can anticipate that hormonal manipulation of growth and adult height in short but endocrinologically normal children will continue to be pursued vigorously in pediatric endocrine clinics around the world.

Received May 9, 2001.

Accepted May 9, 2001.

References

  1. Hintz RL, Attie KM, Baptista J, et al. 1999 Effect of growth hormone treatment on adult height in children with idiopathic short stature. N Engl J Med. 340:502–507.[Abstract/Free Full Text]
  2. Buchlis JG, Irizarry L, Crotzer BC, et al. 1998 Comarison of final heights of growth hormone-treated versus untreated children with idiopathic growth failure. J Clin Endocrinol Metab. 83:1075–1079.[Abstract/Free Full Text]
  3. Schmitt K, Blumel P, Waldhor T, et al. 1997 Short- and long-term (final height. data in children with normal variant short stature treated with growth hormone. Eur J Pediatr. 156:680–683.[CrossRef][Medline]
  4. Carel J-C, Hay F, Coutant R, et al. 1996 Gonadotropin-releasing hormone agonist treatment of girls with constitutional short stature and normal pubertal development. J Clin Endocrinol Metab. 81:3318–3322.[Abstract]
  5. Kamp GA, Mul D, Waelkens JJJ, et al. 2001 A randomized controlled trial of three years growth hormone and GnRH agonist treatment in children with idiopathic short stature and intrauterine growth retardation. J Clin Endocrinol Metab. 86:2969–2975.[Abstract/Free Full Text]
  6. Lanes R, Gunczler P. 1998 Final height after combined growth hormone and gonadotropin-releasing hormone analogue therapy in short healthy children entering into normally timed puberty. Clin Endocrinol. 49:197–202.[CrossRef][Medline]
  7. Balducci R, Toscano V, Mangiantini A, et al. 1995 Adult height in short normal adolescent girls treated with gonadotropin-releasing hormone analog and growth hormone. J Clin Endocrinol Metab. 80:3596–3600.[Abstract]
  8. Pasquino AM, Pucarelli I, Roggini M, Segni M. 2000 Adult height in short normal girls treated with gonadotropin-releasing hormone analogs and growth hormone. J Clin Endocrinol Metab. 85:619–622.[Abstract/Free Full Text]
  9. Municchi G, Rose SR, Pescovitz OH, et al. 1993 Effect of deslorelin-induced pubertal delay on the growth of adolescents with short stature and normally timed puberty: preliminary results. J Clin Endocrinol Metab. 77:1334–1339.[Abstract]
  10. Grumbach MM, Auchus RJ. 1999 Estrogen: consequences and implications of human mutations in synthesis and action. J Clin Endocrinol Metab. 84:4677–4694.[Abstract/Free Full Text]
  11. Mauras N, O’Brien KO, Klein KO, Hayes V. 2000 Estrogen suppression in males: metabolic effects. J Clin Endocrinol Metab. 85:2370–2377.[Abstract/Free Full Text]
  12. Rochira V. 2001 Aromatase inhibitors in pubertal boys: clinical implications (Letter). J Clin Endocrinol Metab. 86:1836–1837.[Free Full Text]
  13. Faglia G, Arosia M, Porretti S. 2000 Delayed closure of epiphyseal cartilages induced by the aromatase inhibitor anastrozole: would it help short children grow up? J Endocrinol Invest. 23:721–723.[Medline]




This Article
Full Text (PDF)
Submit a related Letter to the Editor
Purchase Article
View Shopping Cart
Alert me when this article is cited
Alert me when eLetters are posted
Alert me if a correction is posted
Citation Map
Services
Email this article to a friend
Similar articles in this journal
Similar articles in PubMed
Alert me to new issues of the journal
Download to citation manager
Request Copyright Permission
Google Scholar
Articles by Kaplowitz, P. B.
Articles citing this Article
PubMed
PubMed Citation
Articles by Kaplowitz, P. B.


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Endocrinology Endocrine Reviews J. Clin. End. & Metab.
Molecular Endocrinology Recent Prog. Horm. Res. All Endocrine Journals