Objective. This study was undertaken to determine whether serial bone age (BA) radiographs were obtained in patients with growth hormone deficiency and to assess whether there were differences in outcome between subjects with and without monitoring of BA radiographs.
Research Design and Methods. Data were collected from the National Cooperative Growth Study database on growth hormone-deficient subjects who were treated for at least 3 years. Comparisons were made among three groups of subjects: 1) those with BAs at entry versus those without; 2) those with BA values in the first year of follow-up if an entry radiograph had not been done versus those with no first-year examination; and 3) those with a BA at entry and yearly for 3 years versus those with no radiographs during the same period. Differences in the change in height standard deviation score (SDS); change in height age, age, pubertal progression, number of visits, growth hormone dosage; and number of growth hormone injections per week were compared.
Results. Of the 6191 subjects assessed, 93% had at least one BA radiograph obtained; there was a mean of 3.6 ± 2.6 total number of BA radiographs per patient during the 5.2 ± 1.9 years of follow-up. Subjects with BA values at entry were older and had slightly higher cumulative height SDS and height age change compared with those without BA values at entry. Subjects with BA assessment during the first year were older and had shorter growth hormone treatment time and slightly better cumulative change in height SDS and height age than did those without BA in the first year. Comparing those with serial BA determination for the first 3 years of treatment versus those with no BA values, those with BA were older, more pubertal, seen more often, had more growth hormone injections per week of a comparable growth hormone dosage, and had slightly larger cumulative change in height SDS and height age than those without x-rays.
Conclusions. These data suggest that National Cooperative Growth Study investigators find it of benefit to obtain baseline and follow-up measurements of BA in most subjects treated with growth hormone. Subjects with BA monitoring do slightly better than do those whose skeletal maturation is not measured. BA assessment should be considered part of the follow-up of patients treated with growth hormone therapy.
The utility of estimating and following skeletal maturation using radiographs of the left hand and wrist and comparing them to the standards of Gruelich and Pyle1 has come into question for subjects treated with growth hormone (GH). It has been suggested that serial assessment of bone age (BA) adds little to clinical decision-making concerning GH dosage and duration of treatment. This is in contrast to the known value of following serial BA radiographs in the research arena. There is large body of literature that has been accumulated for the years during which GH treatment has been available that depends on comparing the change in height age to the change in BA to predict adult height outcome and potential efficacy of GH therapy.2,,3 This is valid in the wide variety of conditions in which GH has been investigated and in the circumstances in which it has been used alone or in combination with other growth-promoting agents.4,,5
The purpose of this study was to determine whether serial BA radiographs were obtained in patients receiving GH treatment and to assess whether there were any differences in monitoring or outcome for subjects with and without these radiograph examinations. Comparisons were made concerning age at initiation of GH therapy, pubertal status, growth velocity, GH dosage, injection schedule, and visit frequency between GH-deficient subjects with and without serial BA determinations from the National Cooperative Growth Study (NCGS) database.
MATERIALS AND METHODS
Analysis concerning the frequency of BA monitoring was made on the NCGS database as of December 31, 1997. Data were included from patients who had GH deficiency (GHD), including patients with idiopathic and organic GHD, who were naive to GH therapy at enrollment in NCGS, had maximal stimulated GH levels <10 mg/mL, and who had at least 3 years of treatment with GH. Comparisons were made for various selected groupings based on the frequency of BA determinations at enrollment and during each of the first 3 years of treatment. The data were analyzed comparing the following: 1) those patients with BA determination at entry versus no BA at entry; 2) subjects who had no BA determination at enrollment but who had BA during the first year of follow-up versus no BA during the first year of follow-up; and 3) the two extreme groupings of patients who had BA at entry and during each of the 3 years of follow-up versus those with no BA at entry or during any of the 3 years of follow-up.
The frequency of BA determinations was compared by age at entry, at pubertal status at entry, and at completion of 3-year follow-up; change in height standard deviation score (SDS); height age over the follow-up period; visit frequency; GH dosage per kilogram of body weight; and number of injections of GH per week. Additionally, the cumulative change in height SDS and height age over the duration of GH treatment were compared.
Data are presented as means ± standard deviation (SD) for continuous variables and as frequencies or percentages for categoric variables.
There were a total of 6191 subjects in the NCGS database who had GH deficiency, were naive to GH treatment, and who had at least 3 years of follow-up. Of these subjects, 5778 (93%) had a least one BA radiograph obtained. The average age at first BA was 9.2 ± 3.9 years, whereas the average BA was 7.3 ± 3.8 years. At entry, 3882 (62.7%) of the subjects had a BA. During the subsequent 3 years of follow-up, 39.8% (year 1), 53.5% (year 2), and 52.3% (year 3) of the patients had BA determinations each year. For all subjects, the average GH treatment time was 5.2 ± 1.9 years, and there was a mean 3.6 ± 2.6 total number of BA x-rays (including entry and follow-up) per patient.
Comparison was made between subjects with BA at entry (n = 3882) and those with no BA (n = 2309) at entry (Table 1). As shown, more patients with entry BA x-rays compared with those without entry BA had subsequent BA determinations during the 3-year follow-up. In addition, subjects with BA were older at entry and had slightly higher cumulative height SDS and height age change over the duration of GH treatment than those without BA at entry.
Similar analyses were performed for subjects who did not have a BA at entry but who had a BA determination during the first year of follow-up compared with those with no BA during that first-year period. There were 1409 patients who did not have a BA determination, and 900 patients who did during the first year. Patients with a BA during the first year were older at enrollment (8.7 ± 3.9 years) compared with those without a BA (7.5 ± 4.2 years), and GH treatment time was shorter for the group with BA (5.4 ± 2.1 vs 5.1 ± 1.8 years, respectively, for the group without BA vs with BA during year 1). For patients who did not have a BA determination during the first year, the percentages of patients with BA during years 2 and 3 were 38.1% and 34.3%, respectively, compared with the percentages of subjects with BA during the first year, with 63.4% having a repeat BA in year 2, and 64% having a repeat BA in year 3. The group without BA in the first year had a similar height SDS and height age advancement over the treatment period (cumulative change in height SDS was 1.6 ± 1.3, and cumulative change in height age was 6.0 ± 2.3 years over 5.4 years of follow-up). Because the treatment period was shorter for those with BA (cumulative change in height SDS was 1.6 ± 1.2, and cumulative change in height was 5.9 ± 2.2 years over 5.1 ± 1.8 years of follow-up), the data suggest that there was a slightly better outcome in the group who had a BA obtained during follow-up year 1.
Finally, comparisons were made between the two extreme groups that had either no BA x-rays done or had BA determinations done yearly during the first 3 years of follow-up. Group 1 had no BA at entry nor during the 3-year follow-up; there were 662 subjects with an average calendar age of 7.0 ± 4.1 years at enrollment and a mean height age of 5.0 ± 3.4 years. Group 2 had BA at entry and during each year of the 3-year follow-up; there were 772 patients with an average calendar age of 9.7 ± 4.0 years at enrollment and a mean height age of 7.0 ± 3.3 years. As shown in Table 2, a slightly larger cumulative change in height SDS and height age and a shorter treatment period indicate that the rate of cumulative change in height SDS and height age was greater for group 2 than for group 1. The cumulative changes over the 3-year period were also greater for group 2. With regard to BA advancement in group 2 subjects, there was an average of 1.3 years BA advancement per chronologic year for each of the 3 years, and a cumulative BA advancement of 3.8 ± 1.3 years for the 3-year follow-up period. During the same period, there was a mean cumulative height age advancement of 4.0 ± 0.8 years for the subjects in group 2.
As presented in Table 2, there were more subjects in group 2 who entered the study in puberty; in addition, more group 2 patients entered puberty during the observation period than did group 1 patients. Group 2 subjects received slightly more injections than those in group 1, but the dosage of GH (mg/kg/wk) was not different between the groups. Group 2 patients had more regular annual visits than did group 1 subjects.
Although the primary analyses were performed for the first 3 years of follow-up in group 1 and 2 subjects, a subset of subjects had been followed for up to 9 to 10 years. The cumulative mean number of BA determinations performed during this period for group 1 was 2.21 ± 1.8 (26 subjects), and the mean number performed during this period for group 2 was 10.4 ± 3.0 (24 subjects). Data on mean number of BA x-rays obtained and total GH treatment time indicate that the longer the treatment period, the greater the number of BA x-rays obtained.
These data show that nearly two thirds of subjects in the NCGS database had a BA radiograph at entry and that ∼40% to 50% of subjects each year had a follow-up BA determination. Subjects with BA x-rays obtained at entry or during the first year were more likely to have subsequent BA measurements during the 3-year follow-up period. They were older and appeared to have slightly improved height outcomes during treatment. By comparing the two extreme groups, those with no BA determinations and those with yearly BA assessments, it again appeared that those with BA measurements were older and grew better. In addition, this analysis suggested that the subjects with BA readings were more likely to begin GH treatment after the initiation of puberty and more likely to enter puberty during the study period. The subjects with BA x-rays received more GH injections and had more frequent visits documented in the NCGS database than did the cohort with no BA measurements. This would suggest that NCGS investigators find it of value to obtain BA measurements in most subjects. Although the slightly improved height SDS and height age outcomes during treatment of the subjects with BA readings might be attributed to their older age and pubertal status, there still remains the suggestion that somehow radiographs are used by the NCGS investigators to maximize GH therapy. Because final height data were not yet available, it is possible that by assessing skeletal maturation at frequent intervals, GH treatment might be extended longer, other agents (such as sex steroids or luteinizing releasing hormone analogues) might be added more often and in a more efficacious manner, or, because of more frequent visits, subjects might be more compliant with treatment regimens, all which would ultimately improve the outcome of the cohort followed with BA radiographs.
The conclusion of this study, that it is important to follow BA x-rays for the pediatric endocrinologist, is supported by the survey of GH treatment practices that was conducted by Wyatt et al6in 1993 and reported in 1995. Of the 251 pediatric endocrinologists surveyed, BA determinations were used by 60% of them to determine who should start GH treatment, and BA delay was ranked 5 of 14 on the scale of auxologic and laboratory criteria used to initiate GH. These findings were similar to the percentages of those who ranked obtaining a BA determination as an important criterion to discontinue GH. Again, 61% of those surveyed stated that BA was important, and they ranked it 5 of 14 in criteria used to stop GH therapy.
There is no doubt that assessment of skeletal maturation is at the foundation of research performed evaluating GH treatment.7–9 Although BA radiographs are used to distinguish benefit from harm during studies of GH efficacy in novel clinical situations and in innovative therapeutic regimens, what role BA radiographs play during routine clinical practice has been questioned. These data from the NCGS support the notion that not only do pediatric endocrinologists find it of value to obtain BA determinations at enrollment, but that they also find it beneficial to assess serially skeletal development. BA assessment should be considered an important component of the follow-up of patients treated with GH.
Supported by an educational grant from Genentech, Inc, South San Francisco.
We thank Kevin Connelly for analysis and programming support, and Sandra Blethen, MD, PhD, and Jim Frane, PhD, for helpful suggestions.
- Received May 13, 1999.
- Accepted June 22, 1999.
Reprint requests to (F.R.K.) 4650 Sunset Blvd, Box 61, Los Angeles, CA 90027. E-mail:
Presented in part at the National Cooperative Growth Study Twelfth Annual Investigators Meeting; October 8–11, 1998; New Orleans, LA.
- GH =
- growth hormone •
- BA =
- bone age •
- NCGS =
- National Cooperative Growth Study •
- GHD =
- growth hormone deficiency •
- SDS =
- standard deviation score(s) •
- SD =
- standard deviation
- ↵Gruelich WW, Pyle SI. Radiographic Atlas of Skeletal Development of the Hand and Wrist. Stanford, CA: Stanford University Press; 1993
- Plotnick L,
- Attie KM,
- Blethen SL,
- Sy JP
- August GP,
- Julius JR,
- Blethen SL
- Copyright © 1999 American Academy of Pediatrics