Assessing the cost-effectiveness of withdrawing growth hormone treatment after mid-puberty in adolescents with idiopathic isolated growth hormone deficiency

If children who are diagnosed as idiopathic isolated growth hormone deficiency
are retested for growth secretion after adult height has been reached, a normal
test result is often observed. It appears plausible that if a normal GH
secretion is observed in mid-puberty, GH treatment may only have a minor effect
on adult height. We hypothesize that withdrawing GH treatment in mid-puberty
has no negative effect on attained adult height and on patients' satisfaction
with adult height.

The aim of this study is to assess whether withdrawing GH treatment after
mid-puberty in adolescents with idiopathic isolated GH deficiency, who showed a
normal result in a GH stimulation test at retesting, is as effective as
continuing GH until adult height.

Prospective patient preference design with additional historic control group,
studied up to adult height. All children with IIGHD will be retested in
mid-puberty, according to the current treatment protocol. If GH secretion is
normal, patients will be asked if they prefer to continue GH treatment until
near-adult height is reached (traditional approach) or discontinue GH
treatment. We expect that the preference of each choice will be approximately
50%. It is expected that groups will differ in baseline characteristics (e.g.
those who choose discontinuing GH may be older and taller). Because the number
of included patients will be too low (and the between-group differences too
large) to show statistically significant *non-inferiority* of discontinuing GH
at mid-puberty, a retrospective analysis will be performed of growth, pubertal
stages and bone age of a historic control group (anonimized) with IIGHD, in
whom a normal GH provocation test was found after stopping GH treatment at
final height. Based on these data, a model will be constructed of expected
height gain on GH treatment as a function of sex, age, bone age, Tanner stage,
GH peak in childhood, GH peak at retesting, and GH dosage. For both
prospectively followed groups the expected height gain at inclusion will be
calculated based on the model. At the end of the observation period, the
effectively attained height gain in both groups will be compared with the
predicted one. We hypothesize that the difference in attained minus predicted
height gain in both groups will not be significantly different from zero, and
that the 95% CI will exclude a difference >0.5 SD to the detriment of the group
who discontinued GH in mid-puberty.

The intervention is withdrawing growth hormone treatment in mid-pubertal
adolescents with adequate growth hormone secretion at retesting. At the start
of growth hormone treatment growth hormone secretion was insufficient. Patients
were retested in mid-puberty and growth hormone secretion appeared normal.

The burden is minimal, and approximately 50% of the participants will benefit
by stopping daily injections for 2-3 years, less venepunctures and visits to
the clinic. A theoretical, but very low risk is that the group who stops GH at
mid-puberty may achieve a slightly shorter adult height. Besides a benefit for
50% of participants, there is a substantial benefit for health budget. The
average GH dose for a GH deficient adolescent is 1.6 mg/day. The cost of 1 mg
GH is ¤ 25. GH treatment costs ¤ 14.600 per year per adolescent. Resource
utilization is estimated as ¤ 365 per year per adolescent. If our study shows
that GH treatment can be withdrawn 2 years earlier with a similar and equally
satisfying adult height, the reduction in costs are substantial (¤ 2.394.400).
This study can only be done in this patient group.