Growth hormone rising: did we quit too quickly?

Abstract

TWELVE YEARS AGO Daniel Rudman proposed that daily administration of recombinant human growth hormone (rhGH) was safe and efficacious for the treatment of reduced bone and muscle mass in elderly men.(1) Although the study design and the subsequent interpretation of his original paper were greeted with tremendous skepticism, a decade's worth of small but important randomized trials, funded federally or privately, were undertaken. The results, as critics repeatedly have pointed out, were equivocal at best. However, at the same time, several companies initiated clinical trials of rhGH for the growth hormone deficiency (GHD) syndrome. Most were performed in Europe, led by a group of dedicated investigators who were unwilling to give up on rhGH and felt there was a subset of patients with primary osteoporosis that exhibited low circulating levels of insulin-like growth factor I (IGF-I) even if this was not a result of decreased endogenous GH secretion.(2-6) Late in the 1990s, rhGH was approved by the U.S. Food and Drug Administration (FDA) as “replacement” therapy for GHD principally because of data that unequivocally demonstrated improvement in muscle and bone mass, as well as quality of life in patients without pituitary function. These changes are reflected by increases in both serum IGF-I and insulin-like growth factor binding protein-3 (IGFBP-3). However, coincidental epidemiologic evidence emerged that circulating IGF-I concentrations in the high-normal range were associated with a greater risk of prostate, breast, and colon cancer.(7-9) Because rhGH increased IGF-I expression and secretion from the liver in a dose-dependent manner, support, both publicly and privately, for rhGH or rhIGF-I, in common disorders such as osteoporosis, quickly faded. Clearly there had to be other drugs or hormones that could carry the anabolic banner with similar efficacy but better safety profiles. Parathyroid hormone (PTH) rapidly filled that vacuum and is now approved for the treatment of postmenopausal osteoporosis. With regards to cancer risk, new data have recently emerged that low serum levels of IGFBP-3 in combination with high normal IGF-I levels were more powerful indicators of prostate cancer risk than IGF-I alone.(7) Interestingly, in a recent animal model, Cohen et al.(10) showed that tumor incidence in diet-restricted TRAMP mice was markedly increased by the administration of rhIGF-1 but not by rhGH. Although rhGH was relegated to a forgotten corner of skeletal anabolics, every now and then, data, such as was published last year about the positive myotrophic effects of rhGH in the frail elderly, have emerged that both tantalize and beg a very important question: did we give up too quickly on rhGH?(11) The paper by Landin-Wilhelmsen et al.(12) in this month's Journal is one such example. These investigators performed a relatively small (i.e., by osteoporosis standards) randomized placebo controlled trial of 80 postmenopausal women on hormone replacement therapy to determine the effects of two doses of daily rhGH on bone mass. After 18 months, the placebo group stopped injections, but the two treatment arms were followed for an additional 18 months. Thirty-six months after randomization, rhGH was discontinued, and the subjects were followed for an additional 2 years. Women in the two rhGH groups showed remarkable increases in bone mineral content (BMC) and bone mineral density (BMD) at several skeletal sites after 60 months. In addition, rhGH was extremely well tolerated, and serum IGF-I, as well as IGFBP-3, concentrations returned to baseline values after cessation of rhGH. Several features of this paper are worth discussing in respect to trial design as well as outcome. First, the double-blind portion of the randomized trial lasted only 18 months. Thus, rhGH treated women knew they were receiving growth hormone for the last 1.5 years of the trial. Second, there was a 2-year period of observation, posttreatment, in all rhGH-treated women, when other lifestyle interventions could have occurred to enhance the positive effects of the previous GH treatment. Third, it is not clear whether the subjects were blinded to their BMD results after cessation of the double-blind portion of the trial. Knowledge about positive changes in BMD can reinforce behaviors that could further enhance BMD. This is clearly demonstrated by the increase in BMD in the placebo group, which might also be a function of calcium/vitamin D supplementation in a population at risk for low serum 25OH vitamin D levels. Fourth, all women were taking hormone replacement therapy before study initiation and throughout the trial. As such, changes in BMC and BMD must be viewed in the context of combination therapy, rather than as a result of rhGH treatment alone. Finally, the number of subjects in this trial was small, leaving considerable room to doubt the overall impact of the treatment in respect to BMD and providing no evidence whatsoever concerning fracture efficacy. All these design limitations are likely to be familiar to investigators who have worked with rhGH. Indeed, almost all rhGH trials “post-Rudman” suffer from limitations such as open labeling after 18 months, absence of blinding, and underpowering for fracture efficacy. Notwithstanding those concerns, this paper contains some interesting elements that might resurrect the debate about the use of rhGH in the management of osteoporosis. The increase in BMD at both the spine and femur is considerable and clinically significant. A nearly 15% rise in spine and femoral BMC after 3 years of rhGH and 2 years of observation is consistent with a strong anabolic response in the skeleton of women who entered with very low BMD, and in some cases, with prevalent vertebral fractures. Moreover, bone area increased significantly at both sites with rhGH. Such changes are reminiscent of the skeletal response to PTH and raise some interesting questions. Moreover, the authors demonstrated an increase in lean body mass, which in itself, may have significant implications in respect to falls and fracture susceptibility. Several questions immediately arise from this study. First, does GH treatment preferentially increase bone mineral? We know GH induces hepatic IGF-I expression in a dose-dependent manner, and this leads directly to a rise in serum IGF-I. Less certain is whether rhGH also induces a similar increase in skeletal IGF-I expression. Of considerably more interest, however, is the putative connection between mineral accretion and IGF-I. Very recently, Zhang et al.(13) demonstrated by targeted deletion of the IGF type I receptor in bone, that mineralization of the skeleton is dramatically impaired. Furthermore, IGF-I can enhance mineralization of stromal cells and osteoblasts in vitro (CJ Rosen, personal communication, 2002). So it is conceivable that part of the effect of rhGH on BMD occurs through an increase in mineral apposition. Unfortunately, we can only guess whether skeletal mineralization was enhanced by GH because the investigators did not perform paired bone biopsy specimens; they elected only to label and biopsy women at the beginning of the trial.(12) Second, does bone size increase in women treated with a combination of rhGH and estrogen? In this study, bone area increased by as much as 10% in the femoral neck at year 5 in those women treated for 3 years with the highest dose of rhGH (2.5 U/day). This must mean that growth hormone increases periosteal apposition, a feature also noted with PTH treatment.(14) The combination of increased mineral deposition and enhanced periosteal growth would imply that bone strength could be markedly improved by growth hormone treatment, especially in combination with an antiresorptive. This is reflected by animal data from Andreassen et al.(14) who have shown an significant increase of the ultimate stiffness in ovariectomized rats treated with PTH(1-34), rhGH, or the combination (Fig. 1). This effect also explains why there is a huge discrepancy between BMD data in GH-treated patients as measured by DXA and the possibility that rhGH, like PTH, could have a greater effect on fracture risk reduction than what can be explained by DXA. This is illustrated in Fig. 2. Mechanical properties of femora from sham-operated rats (Sham) vs. ovariectomized (OVX) rats treated with parathyroid hormone 1-34 (OVX-PTH), recombinant human growth hormone (OVX-rhGH), and the combination of PTH and rhGH (OVX-PTH-rhGH). *p < 0.05, **p < 0.01. Adapted from Andreassen TT, Oxlund H 2000 The influence of combined parathyroid hormone and growth hormone treatment on cortical bone in aged ovariectomized rats J Bone Miner Res 15:2266-2275.(14) Percentage of change to baseline of hip bone measurements as performed by DXA (Hologic QDR 1000) in patients with pituitary insufficiency treated with rhGH. BMC, bone mineral content; BMD, bone mineral density. Adapted from Wüster C, Härle U, Rehn U, Müller C, Knauf K, Köppler D, Schwabe C, Ziegler R 1998 Benefits of grouch hormone treatment on bone metabolism, bone density, and bone strength in growth hormone deficiency and osteoporosis. Growth Horm IGF Res 8:87-94.(23) As such, there is a compelling case being built for a prospective trial on fractures in GH-treated individuals, especially because there are observational data that rhGH treatment for pituitary patients reduces fracture risk.(4) Once again, however, the absence of biopsy data in the study by Landin-Wilhelmsen et al.(12) limits confidence in this conclusion, as does the small number of subjects. However, it is tantalizing to speculate that daily rhGH acts very much like PTH, through induction of skeletal IGF-I and other growth factors, to increase bone mass, especially in the periosteal compartment, and thereby enhance bone strength. Third, is discontinuation of rhGH associated with an increase in BMD? In the current study, cessation of rhGH was associated with a continuous increase in BMC and bone area for 2 years after treatment. This phenomenon of “catch-up” bone gain after cessation of treatment was also noted recently by Biller et al.(15) in younger individuals treated with rhGH. Moreover, Finkelstein et al.(16) observed continual bone accretion after discontinuation of PTH. The mechanism of this “catch-up” is not readily apparent, although there are some interesting possibilities. First, it has long been recognized that GH, IGF-I, and PTH all stimulate bone resorption while activating bone formation. In respect to GH and IGF-I, however, it has never been clear precisely how IGF-I could increase bone resorption. However, Rubin et al.(17) recently demonstrated that IGF-I in a dose-dependent manner stimulated RANKL expression and downregulated OPG in ST-2 cells. These findings imply that bone turnover with anabolics is accelerated by targeting stromal cells, which in turn generate not only growth factors that increase bone formation, but also cytokines that are necessary for osteoclast recruitment and differentiation. Thus, discontinuation of rhGH might lead to an immediate cessation in IGF-I-mediated RANKL expression, while at the same time continuing the stimulatory actions on osteoblasts initiated earlier during treatment. Finally, an obvious question is whether combination therapy is the appropriate therapeutic paradigm for anabolic drugs for disorders such as osteoporosis. The current study(12) is remarkable because it truly represents a combination trial using estrogen as an antiresorptive and rhGH as an anabolic. Although at least one other study used rhGH + an antiresorptive (calcitonin) to treat postmenopausal women, the current trial is bigger, has longer follow-up, and employs a stronger antiresorptive agent.(18) The combination of rhGH with pamidronate resulted in absence of rhGH efficacy on bone.(19) Indeed, the current trial(12) is very reminiscent of an earlier PTH + estrogen study conducted by Lindsay et al.(20) That study produced the first solid evidence for a marked increase in bone mass at both lumbar and femoral sites with PTH(1-34) and estrogen. Combination therapy has a strong rationale in clinical medicine (i.e., to enhance trabecular and cortical bone mass by directly stimulating osteoblasts while simultaneously blocking endosteal resorption). Currently, there is a large National Institutes of Health sponsored randomized trial of PTH with or without alendronate, which will determine whether a potent antiresorptive impacts positively or negatively on bone accretion with PTH.(21) Notwithstanding, evidence from the current paper strongly suggests that the right combination of an antiresorptive that blocks IGF-I-induced osteoclast recruitment, with an anabolic, could produce a substantial increase in bone mass and strength.(22) In summary, there may be room to reconsider our notion about rhGH therapy for the management of osteoporosis. In some ways, the progression of rhGH in clinical trials is very similar to that of PTH both in respect to understanding the mechanism of action, and in regards to defining potentially serious side effects. Of course we are far from determining whether 1-3 years of rhGH could have long-term neoplastic sequelae. However, data on tumorigenesis and IGFBP-3, which is stimulated by rhGH and is cancer-protective, might also be promising. However, potential benefits of rhGH on muscle strength and coordination that might lead to a reduction in falls and might reduce fracture incidence are more than expected from BMC alone. However, the paper by Landin-Wilhelmsen et al.(12) is another fresh reminder that what goes around comes around. In science, that is often a good thing.