The chasm: what can be done, and what should be done.

There is certainly no lack of contentious issues in health care. The high price tag of technology, financial support for medical research, and cost effectiveness are familiar topics that have been endlessly debated in the context of different technologies. However, occasionally an issue emerges that seems to crystallize many of the most contentious issues simultaneously. In the early 1990s, autologous bone marrow transplantation served as a focal point for discussing the high cost of new technologies and payer coverage policies, which tended to exclude therapies considered "investigational." Now, the use of growth hormone in short stature children is focusing debate on another set of issues, including the high cost of bioengineered therapies (one can assume cost will be an element of virtually all health policy discussions), off label uses of drugs, diffusion of medical practice that outpaces the underlying science, and the associated ethical issues.

Prior to the commercial availability of biosynthetic growth hormone (GH), GH was only available as extracts of pituitary gland harvested at autopsy. Due to this very limited supply, GH therapy was rationed to patients with a severe or "classic" GH deficiency, resulting in pituitary dwarfism. Biosynthetic GH has been available since 1985, and this new unlimited supply has naturally created interest in expanding its use to short-statured children without classic GH deficiency.

However, these broadened patient selection criteria have remained controversial due to uncertainties in almost every step in the diagnosis and treatment process: selecting patients to be tested, limitations in the laboratory testing for GH, establishing diagnostic cut-offs for normal versus abnormal GH levels, ability of the laboratory tests to predict response to GH therapy, changes in growth velocity due to GH therapy, whether resulting final height is significantly improved, and whether this improvement is clinically or emotionally significant for the patient. Despite these uncertainties, it is estimated that the annual U.S. expenditure for growth hormone, which costs about $20,000 per year for a 30 kg child, exceeds $375 million dollars.

At the very core of this issue is the lack of clarity about what exactly is being treated. A diagnosis of growth hormone "deficiency" suggests that a disease is present, and with the ready availability of biosynthetic GH, GH "deficiency" becomes a treatable disease. This rationale is consistent with the standard practice of endocrinology: deficient hormones are replaced, and excessive hormones are suppressed. Patients with short stature frequently undergo extensive testing to establish whether their GH levels are normal or abnormal. However, aside from those patients with classic growth hormone deficiency, there is uncertainty about what levels of GH would actually constitute a disease.

It is apparent that GH levels lie along a continuum, and that defining a cut-off as to what is normal or abnormal is arbitrary. For example, if a GH hormone level of 10 ng/ml is considered normal, would a patient with a GH level of 9.9 ng/ml be treated, while a child of the same height with a level of 10 ng/ml would not? Due to these uncertainties, some have abandoned laboratory testing for growth hormone levels entirely, instead basing treatment decisions on the child's height, growth velocity, and estimated final height based on skeletal maturity and the parents' height. Normal height and growth velocity are usually defined as above the fifth percentile based on standardized growth charts. Obviously, by definition, this approach would automatically define 5 percent of the population as "abnormal."

If short stature cannot be attributed to a disease, what then is the rationale for treatment? Is short stature below the fifth percentile considered a deformity, a disability, or a functional impairment warranting treatment? Or alternatively, is the rationale for treatment the anticipated psychosocial distress associated with short stature, which is in turn related to our cultural bias toward heightism? If the psychosocial distress, i.e., low self esteem, social isolation, etc., is the real underlying rationale for treatment, then either this disability should be present at the time treatment is initiated in childhood, or if not, it should be predictable when the child reaches his or her final height.

Unfortunately, there are no data to show that psychosocial distress can be predicted. Various studies have pointed out that psychosocial morbidity is variably related to short stature. In one large study of 258 people, there was no correlation between short stature and clinically significant psychosocial morbidity in either sex. (1) In addition, severity of the height deficit did not correlate with behavioral adaptation. Furthermore, GH therapy is typically initiated by parents or physicians and given to children, who, at the start of therapy, are not experiencing any behavioral problems.

This discussion is only a theoretical rationale for GH therapy. The situation becomes even more complicated when one considers whether GH therapy really works, either in terms of increasing final height, or in diminishing the presumed morbidity of short stature. Unfortunately, the data are inadequate on this score, as well. While GH therapy does seem to initially increase the growth velocity in most patients, it is unclear whether this observation translates into a significantly increased final height, which, of course, requires long-term follow-up. Furthermore, the use of GH therapy is not without risks. Although the medical complications appear to be minimal, GH therapy itself is burdensome, requiring regular injections. Therefore, there is some concern that GH therapy itself may create psychosocial problems by stigmatizing the child as "sick."

When distilled down to its core, the broadened use of GH in short-stature children without classic deficiency can be characterized as a prophylactic chronic injection therapy of uncertain efficacy that addresses a set of potential psychosocial issues that the grown child may or may not experience. When presented in this stark light, the $378 million expenditure for GH is astonishing.

As pointed out by Lantos in a 1989 article, there is no precedent for this type of therapy in pediatrics. (2) Aside from attention deficit disorder, drug therapy is generally not given for psychosocial morbidities, which are more typically addressed by counseling. Additionally, in this setting, drug therapy is used to alter behavior. In contrast, GH therapy is used to change height with only an uncertain effect on behavior. GH therapy is also unusual in that it requires regular injections. There are certainly pediatric precedents for regular injections, for example treating diabetes, or chronic transfusion therapy for sickle cell disease. However, this therapy is easily justified by the significant morbidity and mortality of these diseases if left untreated. This scenario does not apply to GH therapy.

Given the lack of firm diagnostic criteria for GH deficiency and long-term outcomes data, endocrinologists have developed individual prescribing practices that have been investigated in several surveys. Most recently, Cuttler and colleagues reported that the use of GH was, in part, related to parents' wishes and endocrinologists' perceptions regarding treatment effectiveness. (3) Given eight case scenarios, 64 percent would recommend GH treatment in those with the shortest stature and lowest growth velocity. However, only 31 percent responded that they thought GH therapy was either likely or very likely to have a positive impact on the emotional well-being of the patient. Ultimately, at this point in its evolution, the use of GH therapy illustrates the dilemma commonly created by new medical technology--the chasm between what can be done and what should be done. Unfortunately, the knowledge of what can be done precedes the understanding of what should be done.

References

(1.) Sandberg D. E., Brook A.E., Campos S.P. Short stature: A psychosocial burden requiring growth hormone therapy? Pediatr 1994; 94:832-40.

(2.) Lantos J, Siegler M, Cuttler L. Ethical issues in growth hormone therapy. J Amer Med Assoc 1989;461:1020-24.

(3.) Cuttler L, Silvers B, Singh J et al. Short stature and growth hormone therapy. A national study of physician recommendation patterns. J Amer Med Assoc 1996;276:531-37.

Elizabeth Brown, MD, is Director of Technology Assessment and Clinical Guidelines at Aetna, Inc., in Chicago, Illinois. The opinions expressed in this column are those of the author and not necessarily those of Aetna U.S. Healthcare. Dr. Brown can be reached at 312/441-3292.