Safety of osteoanabolic therapy: A decade of experience


  • Cristiana Capriani,

    1. Metabolic Bone Diseases Unit, Division of Endocrinology, Department of Medicine, College of Physicians and Surgeons, Columbia University, New York, NY, USA
    Search for more papers by this author
  • Dinaz Irani,

    1. Metabolic Bone Diseases Unit, Division of Endocrinology, Department of Medicine, College of Physicians and Surgeons, Columbia University, New York, NY, USA
    Search for more papers by this author
  • John P Bilezikian

    Corresponding author
    1. Metabolic Bone Diseases Unit, Division of Endocrinology, Department of Medicine, College of Physicians and Surgeons, Columbia University, New York, NY, USA
    • Department of Medicine, College of Physicians & Surgeons, 630 W. 168th Street, New York, NY 10032, USA.
    Search for more papers by this author


This article is corrected by:

  1. Errata: Erratum: Safety of osteoanabolic therapy: A decade of experience Volume 28, Issue 2, 431, Article first published online: 15 January 2013


Osteoporosis is a prevalent disease that affects more than 10 million Americans. The lifetime risk of an osteoporosis-related fracture is 50% for Caucasian women and 25% for men. Hip fractures carry a 10% to 20% increase in mortality rate within the first year after fracture.1 The burden of the hip and other fractures can result in chronic pain, disability and other indices of reduced quality of life.

Amidst these sobering statistics is the welcome advent over the past two decades of effective pharmacological therapies that reduce fracture risk. The therapies approved for the management of osteoporosis include bisphosphonates, raloxifene (a selective estrogen-receptor modulator), calcitonin, denosumab, strontium ranelate, and parathyroid hormone [teriparatide (recombinant PTH(1–34) and PTH(1–84)]. The full-length PTH molecule and its foreshortened amino-terminal fragment (teriparatide) are currently the only osteoanabolic agents approved for use. PTH(1–84) is not available in the United States. In November, 2002, teriparatide was approved by the U.S. Food and Drug Administration (FDA) for the treatment of postmenopausal osteoporosis and, later, in men. The approval came with restrictions in duration of therapy (18–24 months) and was limited to individuals with advanced osteoporosis at high risk for fracture. Reasons for these limitations relate, at least in part, to the unexpected occurrence of osteosarcoma and other bone neoplasms in Fischer 344 rat toxicity studies. In fact, the pivotal phase 3 trial of teriparatide was suspended when these rat toxicity data became available. The effects on rats was dose- and duration-dependent, with rats being treated for approximately 80% of their lifetime and at doses three to 58 times the currently approved human dose.2 Similar toxicity was seen for PTH(1–84)3 Given the clear-cut efficacy data when the results of the abbreviated clinical trial results for teriparatide were analyzed, the FDA and equivalent other agencies in Europe and elsewhere granted approval of teriparatide with the stipulations that it should be used for no longer than 2 years. In the United States, the drug carries a “black box warning” and contraindicates its use in patients with existing risk factors for osteosarcoma, including Paget's disease of bone, prior skeletal radiation, and children with open epiphyses.

Teriparatide has now been used in the United States for 10 years. With a decade of experience, it is timely to review the efficacy and safety profile of this medication. PTH(1–84) will also be reviewed. This perspective will be limited to PTH formulations that are available for the treatment of osteoporosis; we will not be discussing other anabolic agents that are currently being studied but are not yet approved, such as the sclerostin antibody ( = sclerostin + antibody).

PTH mechanisms of action

Teriparatide is an amino-terminal fragment of PTH that is biologically identical to the amino-terminal fragment of human PTH. This fragment contains all the classical biological and biochemical actions of the full-length molecule, PTH(1–84). Unlike other osteoporosis agents, PTH is an osteoanabolic that actually promotes bone formation when administered at low doses intermittently. Although the cellular mechanisms by which this occurs are still unclear, it is likely that the Wnt signaling pathway (via suppression of sclerostin),4, 5 Runx2, receptor activator of NF-κB ligand (RANKL), osteoprotegerin, and insulin-like growth factor are all involved.6–8 The stimulation of bone formation that results may first occur on quiescent bone surfaces, akin to bone modeling that occurs in the growing skeleton. Subsequently, bone turnover is stimulated through activation of bone remodeling units. It would appear that most of PTH's osteoanabolic effects are due to the stimulation of bone remodeling, with osteoblast activity exceeding osteoclast activity, and 30% being due to the modeling effect. The initial effect, namely the stimulation of bone modeling, is seen clinically by an increase in bone formation markers such as amino-terminal propeptide of type I procollagen (P1NP) and osteocalcin. The subsequent stimulation of bone remodeling is seen clinically by an increase in bone resorption markers such as serum cross-linked C-telopeptide (CTX) and urinary cross-linked N-telopeptide of type I collagen (NTX). This sequence gives rise to an anabolic window during which the osteoanabolic actions of PTH are optimized (Fig. 1).9 Since the stimulation of bone turnover is a predominant effect of PTH, areas of the skeleton that show an intrinsically high rate of bone turnover, such as the cancellous skeleton (e.g., lumbar spine) will show that greatest accrual of bone density. Skeletal microstructure is improved, with increases in trabecular thickness, connectivity density, and a reduction in trabecular separation.10 An initial increase in cortical porosity is seen; however, this increase does not persist beyond 24 months,11 and one study showed no difference in cortical porosity in iliac crest biopsies between 18 and 24 months.12 Initial declines in bone mineral density (BMD) at cortical sites have also been shown.13–15 However, the net effect on cortical bone is an improvement in both cortical structure and thickness.16–19 These results clearly establish teriparatide and PTH(1–84) as osteoanabolic agents for both trabecular and cortical bone.

Figure 1.

The anabolic window. This figure demonstrates the window between the rise in bone formation markers and the subsequent rise in bone turnover markers; the effects of PTH actions are optimized during this interval (adapted from Ref. 9).

PTH efficacy

PTH reduces fracture risk. The international pivotal Fracture Prevention Trial assigned 1637 postmenopausal women to daily teriparatide (20 or 40 µg) or placebo injections. After an average of 18 months, subjects who received teriparatide at the daily 20-µg dose showed a 9% increase in lumbar spine BMD, a 65% reduction in risk of new vertebral fractures, and a 35% reduction in risk of nonvertebral fractures; there was also a 53% reduction in nonvertebral fractures that were specifically classified as fragility fractures.20 Several other clinical trials have supported these results since the pivotal trial was originally published. For example, a recent meta-analysis of eight clinical trials (n = 2388) that used teriparatide to treat postmenopausal osteoporosis found that lumbar spine BMD increased by 8.14% (95% confidence interval [CI], 6.72%–9.55%) and by 2.48% in the hip (95% CI, 1.67%–3.29%). In trials that included fracture data, vertebral and nonvertebral fractures were reduced by 70% and 38%, respectively. Limitations of this meta-analysis, according to the authors, included the participation of patients with postmenopausal osteoporosis only, excluding all secondary causes, as well as the clinical heterogeneity present in any broad review, which can confound results.21 In another recent meta-analysis, Murad and colleagues22 compared the effectiveness of the following therapeutic approaches: teriparatide, bisphosphonates, raloxifene, denosumab, and calcium and vitamin D. One hundred and sixteen (116) studies were included in the final analysis, with a total 139,647 patients. Teriparatide had the highest reduction in vertebral fracture incidence (odds ratio [OR] 0.30), hip fracture (OR 0.42), and nonvertebral fracture (OR 0.50), and had the highest probability of being ranked most effective in all three categories (49%, 42%, and 79%, respectively) when compared to the other treatments. Murad and colleagues22 noted limitations of their study that included the small number of fracture events in the trials, which could lead to imprecise estimates and indirect comparisons that are intrinsically not as accurate as head-to-head comparisons. In an update of the 2007 systematic review examining the effectiveness and safety of osteoporosis treatments, Crandall and colleagues23 reported a relative risk of vertebral fractures with teriparatide ranging from 0.31 to 0.36 and a relative risk in nonvertebral fractures ranging from 0.60 to 0.65 when compared to placebo or no treatment after pooling data from two systematic reviews. They also included data from randomized controlled trials (RCTs) not included in the meta-analyses and found an 84% statistically significant reduction in all fractures, reduced risk of vertebral fractures with relative risks (RRs) ranging from 0.34 to 0.44 (with the exception of one RCT that had the fewest vertebral fractures and did not show a statistical difference between PTH and placebo). Nonvertebral fracture risk reduction results were mixed, with several RCTs showing no statistically significant difference when compared with placebo, whereas a pooled analysis of five other RCTs showed a statistically significant risk reduction with teriparatide treatment (RR = 0.62). Their overall conclusion was that teriparatide is associated with a reduced risk of both vertebral and nonvertebral fractures in postmenopausal osteoporotic women,23 a conclusion that was statistically reached by the pivotal clinical trial of teriparatide by Neer and colleagues.20

PTH in men

Also prevalent in men, osteoporosis carries an even higher mortality rate after hip fracture than in women.24 Clinical trials in osteoporotic men are typically smaller in scope and are not powered to see reductions in fracture incidence. Surrogate markers such as BMD are monitored. The first randomized trial examining the efficacy of teriparatide in men was conducted by Kurland and colleagues25 in 2000. The teriparatide group showed BMD increases of 13.5% at the lumbar spine and 2.9% at the femoral neck after 18 months. A subsequent, larger but shorter trial involving 437 men with osteoporosis showed a 5.9% increase in lumbar spine BMD and 1.5% increase at the femoral neck BMD.26 The shorter duration of this larger clinical trial, only 11 months, was due to the timing of the results of the rat toxicity studies. The increase in BMD over this 11-month period matched virtually identically the gains in the pivotal clinical trial in postmenopausal women, after the same period of time. A majority of these study subjects (n = 279) were followed after the treatment was completed. There was a significant (83%, p = 0.01) decrease in new moderate or severe vertebral fractures at 30 months after the discontinuation of treatment. It should be noted that some of these patients did receive other osteoporosis therapies during the follow-up period; however, the majority of the patients that received other therapies were in the placebo group and, therefore, the results are likely still to be valid.27

PTH in glucocorticoid-induced osteoporosis

Teriparatide was approved for the treatment of glucocorticoid-induced osteoporosis in 2009 for both men and women. Glucocorticoids are the most common secondary cause of osteoporosis, increasing fracture risk even when bone density is not significantly decreased. Teriparatide's osteoanabolic activity is ideally suited to counter the deleterious effects of glucocorticoids such as reduced bone formation and increased osteoblastic apoptosis. In 2007, Saag and colleagues28 conducted a head-to-head study comparing daily parenteral teriparatide (20 µg) with daily oral alendronate (10 mg) in 428 osteoporotic patients who had received glucocorticoids for at least 3 months (≥5 mg prednisone daily or equivalent). Significant densitometric differences between the groups were evident by 6 months, and by 18 months the teriparatide group showed lumbar spine BMD increases of 7.2% versus the alendronate group, which showed increases of only 3.4% (p < 0.001). Total hip BMD also showed significantly greater increases with teriparatide than with the alendronate group by 12 months. There were significantly fewer new vertebral fractures in the teriparatide group than in the alendronate group (0.6% versus 6.1%, respectively, p = 0.004).28 In the follow-up 18-month extension trial, significant differences in BMD and fracture risk between the two groups were maintained.29


Full-length, native PTH [PTH(1–84)] has also shown efficacy in postmenopausal osteoporosis and is approved for use in Europe. Hodsman and colleagues30 conducted the first randomized, placebo-controlled trial that compared 50, 75, or 100 µg PTH(1–84) daily with placebo over 12 months. Efficacy was found to be dose-dependent, with the highest dose cohort showing 7.8% increases in BMD in the lumbar spine and a 1.6% increase in the total hip at 12 months.30

Greenspan and colleagues31 conducted a much larger randomized, placebo-controlled trial comparing PTH(1–84) 100 µg versus placebo in 2532 postmenopausal women for 18 months (the TOP trial). Increases in BMD were greater in the treatment group by 6.9% in the lumbar spine and 2.5% in the femoral neck. PTH treatment also prevented first vertebral fractures (RR 0.32%; 95% CI, 0.14–0.75) and decreased incidence of additional vertebral fractures in women with baseline fractures (RR 0.47; 95% CI, 0.23–0.98).31 An open-label extension study was conducted by Zanchetta and colleagues,32 monitoring the effect of PTH(1–84) after 36 months in 91 patients who were originally in the placebo arm of the TOP trial. Compared to baseline, lumbar spine BMD increased by 8.5%, total hip by 3.2%, and femoral neck by 3.4% (p < 0.001 for all three values). Lumbar spine and femoral neck data showed initial increases and then plateaued between 24 and 36 months, whereas the total hip showed continuous improvement throughout the duration of the trial. Although the trial was not powered to assess vertebral fracture prevention, the rate did seem to decrease, because only one patient experienced a vertebral fracture after starting open-label treatment (this patient had a previous vertebral fracture at the onset of the trial).32 Various subgroups from these pivotal trials have been studied, showing that these two PTH forms are effective in subjects who have or have not previously fractured or have experienced a single fracture or multiple fractures, and that complaints of back pain are reduced.20, 31–35

Use in Hypoparathyroidism

Hypoparathyroidism is characterized by low or absent PTH levels in individuals whose serum calcium is low. It is most often caused by surgical removal of, or damage to, the parathyroid glands during thyroid, parathyroid, or other neck surgery. An autoimmune disorder can be responsible, along with rare genetic etiologies. The current treatment regimen for hypoparathyroidism consists of high doses of calcium and vitamin D supplementation, including the active form, 1,25-dihydroxyvitamin D. However, these high doses can cause hypercalciuria, as well as renal, basal ganglia, and other soft tissue calcifications. Hypoparathyroidism is currently the only endocrine-deficiency disease for which the missing hormone, namely PTH, is not available.36 Several small teriparatide studies ranging from 20 weeks to 3 years have shown better control of serum calcium levels along with reduced needs for calcium and vitamin D supplementation when compared to calcium and vitamin D supplementation alone.37–40

PTH(1–84) has also shown promising results for the treatment of hypoparathyroidism. Several recent clinical trials have shown a reduced need for vitamin D and/or calcium supplementation as well as improvement in abnormal bone remodeling and metabolism on daily or every other day dosing. Some patients in the PTH(1–84) treatment arms were able to discontinue their 1,25 supplementation altogether.41–45 Cusano and colleagues46 recently presented results from a 4-year study examining 18 hypoparathyroid patients treated with PTH(1–84) at a dose of 100 µg every other day, every third day, or every day. Data showed that serum and urine calcium levels improved and were maintained for the duration of the study. Supplement requirements for calcium were reduced by 36%; calcitriol requirements were also reduced, with three subjects stopping calcitriol altogether.46 PTH(1–84) was found to be effective with once-daily dosing while teriparatide required multiple daily dosing. Positive results were also recently reported in an infusion pump administration trial using teriparatide in hypoparathyroidism.47

Accelerated Fracture Healing

In animals, PTH has been shown to be beneficial in fracture healing by stimulating and accelerating hard-callus formation and increasing the strength of the fracture site.48. A study examining the ability of PTH to accelerate distal radial fracture healing in human subjects, involving 102 postmenopausal women, showed a reduction in the median time to radiographic healing.49 The effect was seen at the 20-µg dose of teriparatide, but not the 40-µg dose. Similar results were reported by CT assessment in a study with PTH(1–84) in 65 osteoporotic women aged more than 70 years who experienced a pelvic fracture.50 A study of teriparatide to accelerate healing after hip fracture is currently being conducted.51 More evidence from randomized controlled trials is needed to substantiate many anecdotal reports that seem to demonstrate remarkable effects in this regard.

Other reports include the successful use of the current approved teriparatide regimen in one case of bisphosphonate-associated atypical femoral fracture and two cases of bisphosphonate-associated osteonecrosis of the jaw (ONJ).52–54 In the first case, the authors described radiographic improvement with the closure of the femur fracture one month after starting teriparatide.52 Harper and Fung53 and Lau and Adachi54 reported bone regeneration at extraction sites and the absence of ulcerations in two distinct patients with ONJ after 10 months of anabolic therapy.


Most of the attention related to the safety of PTH administration in human subjects relates to the rat toxicity data, clearly indicating that osteosarcoma occurs regularly in a dose- and time-dependent manner.2, 3, 55, 56 In the United States, the FDA identifies major safety concerns with a “black box” warning, a designation given to teriparatide with specific reference to osteosarcoma. The labeling instructions contraindicate the use of teriparatide in any situation in which the risk of osteosarcoma is increased. Both teriparatide and PTH(1–84) are contraindicated, therefore, in anyone: (1) with a history of osteosarcoma or any bone cancer (primary or metastatic); (2) at risk for osteosarcoma (Paget's disease of bone, previous radiation therapy to the skeleton or to other organs if skeletal tissue is exposed); or (3) children and young adults with open epiphyses.


In the general population, osteosarcoma is the most common nonhematologic primary bone malignancy. Its incidence shows a biphasic trend by age, with a first peak around puberty and a smaller peak in subjects over 60 years of age.57, 58 Nevertheless, the overall incidence of osteosarcoma is low, representing less than 1% of all cancers diagnosed in the United States.59 The incidence ranges worldwide between 3 and 4.5 cases/million/year in childhood and adolescence, 2 cases/million/year in individuals 25 to 59 years old, and 1.5 to 4.5 cases/million/years in subjects over the age of 60 years.57, 58

Osteosarcoma in rats

The FDA and the European Medicines Agency (EMA) warning regarding osteosarcoma is related to the appearance of the disease in Fisher 344 rats during the 2-year carcinogenicity study for teriparatide.2 This study reported a 26% incidence of osteosarcoma in 360 Fisher 344 rats that received teriparatide at doses of 5, 30, or 75 µg/kg/day, beginning at 6 to 8 weeks of age.2 The occurrence of osteosarcoma was dose-dependent, with the first histopathological and clinical detection occurring after 13 and 17 months, respectively.2 Other bone neoplasms, including osteomas, osteoblastomas, and focal osteoblast hyperplasia, were found. It was noted that the duration of exposure to rats, up to 2 years, represented most of their lifespan. Moreover, the doses employed were large multiples (threefold to 58-fold) of the human dose.2 The second rat carcinogenicity study by Vahle and colleagues55 confirmed that the development of bone neoplasm in rats was strongly dependent on the dose and the duration of treatment, whereas the age at initiation had only a minor effect. Consistent findings were reported in the PTH(1–84) study, which examined Fisher 344 rats given doses of 10, 50, or 150 µg/kg/d for 24 months.3 A 150 µg/kg/d group with a delayed start of 6 months was also included. Radiological evaluations mainly showed widespread bone sclerosis in treated animals and small bone tumors, particularly in the highest dose group.3 Histological findings revealed a significant and dose-dependent increase in bone neoplasm in the 50 and 150 µg/kg/d groups and in the delayed-start groups, compared to controls.3 Osteosarcoma was first detected at week 55 in the highest-dose group.3

Data on other rat strains

A 2001 guidance document from the FDA on the development of PTH for the prevention and treatment of osteoporosis reported preclinical data on the rare occurrence of osteosarcoma in two strains of rat and one strain of mice treated with PTH and related peptides from weaning to 18 months of age, whereas no bone malignancy was observed in the control animals.60 A recent study by Watanabe and colleagues61 investigated the 2-year carcinogenicity of teriparatide in male and female Sprague-Dawley rats. The authors reported data from five individual studies testing the effect of different teriparatide doses and treatment regimens and substantially confirmed the results in Fisher 344 rats.61 A significant increase in the incidence of osteosarcoma was found in the highest-dose group, compared to controls.61 Hence, as for the Fisher 344 studies, a dose- and duration-dependent development of bone tumors was observed.61 No influence of age at initiation of treatment was found.61 The study also established the noncarcinogenic and the carcinogenic dose level of teriparatide as 4.5 µg/kg/d and 13.6 µg/kg/d, respectively.61 The area under the receiver operating characteristic curve (AUC) values of teriparatide per week at these doses were 3.9- to 11.6-fold and 17.8- to 37.1-fold; the corresponding AUC values in humans receiving the weekly injection would be the dose of 56.5 µg.61

Studies in humans and nonhuman primates

The osteoanabolic effect of PTH has also been extensively studied in cynomolgus macaque monkeys. Osteosarcoma has not developed in this non-human primate with exposure range of 1 to 25 µg/kg/d of teriparatide or PTH(1–84) for up to 18 months of treatment and 36 months of observation.62–66 For reference, the 5 µg/kg/d dose is approximately eight times the human dose.64 The lack of evidence for PTH-induced osteosarcoma under these conditions extends to clinical, radiological, and histological examination. Although studies in monkeys are not considered in the assessment of the carcinogenic effect of drugs, these data seem to indicate much less sensitivity to a oncogenic effect of high-dose PTH in non-human primates than in lower animal forms such as rats. The monkey studies may also have more direct relevance to human subjects.

In human subjects, results from clinical trials with both PTH(1–34) and PTH(1–84) have not reported any skeletal malignancies.20, 26, 29, 31 The cumulative number of patients in these trials and observational studies amounts to more than 16,000 and up to 3 years of continuous treatment.67

The postapproval experience with teriparatide and PTH(1–84) has given us additional information. The cumulative experience has reached 10 years for teriparatide and 6 years for PTH(1–84). The worldwide experience is well over 1 million subjects for teriparatide and 57,000 patient-years of exposure for PTH(1–84), respectively. Three cases of osteosarcoma in patients treated with teriparatide have been reported.68–70 A few other reports of osteosarcoma in patients receiving teriparatide have been identified and are currently under review. To date, no cases of osteosarcoma have been reported with the use of PTH(1–84).71

The first case of osteosarcoma in a patient treated with teriparatide was reported in 2007. The postmenopausal woman was a smoker in her 70s who, during her second year of teriparatide therapy, was found to have a metastatic pulmonary lesion that was ultimately diagnosed as osteogenic sarcoma.68 No primary lesion was ever discovered in bone and no autopsy was performed. Despite some ambiguity with the histological examination, the pulmonary metastasis was considered to be an osteosarcoma.68 The second case was a 67-year-old man with a past medical history of prostatectomy and subsequent pelvic radiation therapy for prostate cancer 7 years previously, who was diagnosed with chondroblastic osteosarcoma of the left pubic ramus 2 months after starting teriparatide.69 The third patient was an elderly woman who experienced the growth of an enlarging subcutaneous left posterior calf mass after starting teriparatide treatment.70 The timing of treatment and development of the mass are not clear. Despite the complexity in the definition of the pathology, the mass was diagnosed as an extraskeletal osteosarcoma. These three cases, correctly and cautiously included among the reports of osteosarcoma in patients receiving teriparatide, do not appear to provide strong evidence confirming a direct and unique effect of PTH therapy in the development of osteosarcoma.

These relatively rare cases of osteosarcoma would appear to be consistent with the epidemiology of osteosarcoma in adults. The incidence is similar to unselected populations of adult humans who develop osteosarcoma and are not receiving PTH.

In this issue of JBMR, Andrews and colleagues67 report the first 7-year data from the postmarketing surveillance study of osteosarcoma and teriparatide in the United States. This study was conducted by the FDA in 2003 with the aim to create a 15-year surveillance program on the exposure to teriparatide in patients with osteosarcoma. Between 2003 and 2009 from 15 U.S. registries, 549 of 1448 patients diagnosed with osteosarcoma were interviewed. It was expected that by chance alone, the number of patients who developed osteosarcoma and were given teriparatide would amount to one or two. Numbers higher would indicate an association that was greater than expected on epidemiological grounds alone. The report by Andrews and colleagues67 does not reveal any cases of teriparatide exposure prior to the diagnosis of osteosarcoma. The three published cases of osteosarcoma were not included in the series, because they were not identified in the 15 cancer and medical registries participating in the study. This postmarketing study, which provides important and further reassuring data, will continue for another 7 to 8 years.67

There are other databases that one could access to determine whether there is an increased risk of osteosarcoma in subjects who are exposed to PTH. An obvious source is primary hyperparathyroidism, a chronic disorder in which patients can be exposed to elevated levels of PTH for years. The Swedish Cancer Registry database reported no increase in the incidence of primary bone neoplasia among 12,644 patients with primary hyperparathyroidism who did not undergo parathyroidectomy.72 Similarly, in patients with disorders associated with secondary hyperparathyroidism, such as chronic kidney disease, no increase in osteosarcoma has been reported.72 Jimenez and colleagues73 conducted a retrospective analysis of all cases of osteosarcoma at MD Anderson Hospital between 1948 and 2005. Of the 1234 cases identified, three patients also carried the diagnosis of primary hyperparathyroidism. In two subjects, primary hyperparathyroidism was a preexisting diagnosis, whereas in the third subject both were diagnosed simultaneously. The prevalence of primary hyperparathyroidism in these patients with osteosarcoma did not differ from the expected prevalence in a normal population without osteosarcoma.

Rat bone metabolism versus human bone metabolism

The predictable increase in osteosarcoma when rats are exposed to PTH raises questions as to why the rat and the mouse, and it would appear not the human or the monkey, are at risk. A facile explanation could be the extended exposure time, virtually a rodent's lifetime, and the high doses employed. If human subjects could tolerate such high doses of PTH over their lifetime, would they develop osteosarcoma? This is a question that will never be answered. A way to approach this question was addressed by Tashjian and Chabner.72 They plotted bone mineral content (BMC) data from humans, rats, and monkeys exposed to comparable amounts of PTH and reported a higher pharmacological effect of teriparatide in increasing both trabecular and cortical bone mass in rats.72 This raises the likelihood that there are other explanations, besides dose and duration of exposure, for the lack of evidence for a relationship between PTH and osteosarcoma in monkeys and in humans. The rat is a species that models its skeleton for virtually its entire life. The modeling is associated with lifetime skeletal growth. When exposed to an osteoanabolic agent like PTH, the rat responds exuberantly with a profound increase in endocortical and trabecular bone mass and a significant reduction in marrow space.2, 55 Older studies from the 1930s showed that the entire marrow space can be obliterated by chronic exposure of rats to PTH.74 The ever-growing rat skeleton, therefore, is a developing organ harboring immature and potentially tumorigenic cells that might respond to PTH with uncontrolled behavior, losing its capacity for ordered modeling or growth. In contrast to rat skeletal physiology, the mature adult human skeleton remodels itself on a continual basis. Modeling essentially ceases as a normal occurrence after skeletal maturity has been reached. Osteoblasts are rarely active on quiescent surfaces but are rather exclusively found in the bone remodeling unit, reacting to the excavating work of the osteoclast. An exception to this rule, at least in part, is the human skeleton exposed to PTH. Initially, modeling is stimulated but this phase is limited in scope and in time. Remodeling is influenced by PTH to a much greater extent than the initiating PTH-associated modeling events. This difference in modeling/remodeling characteristics between rats and humans is an important distinction to be made and may have relevance to the issue of PTH and osteosarcoma.

It should be recognized that the cumulative experience with teriparatide and PTH(1–84) is still rather limited in numbers and in time. It is possible, but considered remote, that after a much longer period of surveillance post-PTH therapy and with even greater numbers of subjects treated, a signal associating PTH with osteosarcoma in human subjects will be recognized. The bulk of data would argue that this scenario is an unlikely one.

Safety of PTH in patients with history of malignancy

No clinical recommendations exist on the use of PTH in patients with a history of malignancy in tissues other than bone. The studies in Fisher 344 rats did not reveal any nonosseous tumors. The tumorigenic effect of PTH in rats appears to be specific to bone.2, 3, 55 In the clinical trials with teriparatide and PTH(1–84), no increase in nonosseous cancers was found.20, 26, 29, 31 However, because many tissues and solid tumors could express the PTH/PTHrP receptor, it has been suggested that patients with a history of cancers in the past 5 years avoid PTH therapy.75 There may be wisdom in this recommendation for those whose malignancies harbor osteoblastic potential, such as breast and prostate cancer. However, there is no information about changes in tumor behavior in response to PTH. Of some interest are epidemiological data that suggest mortality rates from cancer in primary hyperparathyroidism are actually lower than control populations.76

Pregnancy and breastfeeding

Pregnancy and breastfeeding are relative contraindications to PTH use. This recommendation is classified by FDA as class C, indicating that no data are available.

Hypercalcemia and hypercalciuria

The physiological and homeostatic actions of PTH on bone, kidney, and 1,25(OH)2D synthesis are all in the direction of conserving calcium and maintaining serum calcium levels. When used as a pharmacological agent, however, PTH has the potential to raise the serum calcium level and, by virtue of a greater filtered glomerular calcium load, to increase urinary calcium excretion. When PTH is used to treat osteoporosis, the serum calcium will rise transiently but is virtually always maintained within the normal range. The peak increase in serum calcium after PTH administration occurs earlier (4–6 hours) after teriparatide administration and a bit later with PTH(1–84) (8–10 hours).77, 78 By the next dose, 24 hours later, the administered PTH is no longer detectable in the circulation and the small rise in serum calcium has usually returned to baseline.

In the pivotal fracture trial of teriparatide, an 11% incidence of mild hypercalcemia in postmenopausal women was observed 4 to 6 hours after injection.20 The recurrence rate was much less, being confirmed in only one-third of those women.20 It was rare to discontinue the PTH because of hypercalcemia.20 Similar results were reported in other clinical trials, which also showed little or no incidence of hypercalcemia within 24 hours after teriparatide injection.26, 28, 29, 79, 80 In a post hoc analysis, Miller and colleagues81 showed a significantly higher incidence of hypercalcemia in patients with moderate renal impairment (glomerular filtration rate [GFR] 30–49 mL/min).

The incidence of hypercalciuria in women was 4.8% to 11% within the first 12 months of teriparatide administration, compared to 3.2% at baseline. However, there was a significant difference between the treatment and the placebo group only when calcium excretion was expressed by baseline body weight (>4 mg/kg/d) at 6 months.82 High baseline urinary calcium excretion was a predictive factor.82 Nevertheless, recurrence was low.

Concurrent hypercalciuria and hypercalcemia were found in less than 1% of women on teriparatide, with a significant correlation between serum and urinary calcium values.82 Postapproval data have confirmed that hypercalcemia and hypercalciuria are rarely seen in subjects treated with teriparatide. Gold and colleagues83 reported an incidence of hypercalcemia (serum calcium above 11 mg/dL) of less than 1% of patients treated with teriparatide in the United States in a 2004 postmarketing surveillance update.

The pivotal trial for PTH(1–84) in osteoporosis, however, showed different results. Hypercalcemia and hypercalciuria were found in 27.8% and 46% of women.31 These relatively high rates are hard to interpret because a significant proportion of the study population was allowed to enroll even if their baseline serum calcium and/or urinary calcium excretion values were elevated. With an 8.9% and 5.5%31 baseline incidence of hypercalcemia and hypercalciuria, it is not surprising that hypercalcemia and hypercalciuria would be seen rather commonly. Data from other clinical trials with PTH(1–84) in which baseline hypercalcemia and hypercalciuria were exclusionary criteria did not show these relative high rates of hypercalcemia and hypercalciuria.71, 84 They were not significantly different from the reports of the teriparatide trials.

In both the teriparatide and the PTH(1–84) trials, discontinuation of calcium supplements and, wherever necessary, reduction in PTH dose frequency were reported as effective in correcting the incidence of hypercalcemia. In practice, many experts limit the amount of supplemental calcium when teriparatide is being prescribed to no more than 500 mg per day.

It should be self-evident that serum and urinary calcium should be measured prior to prescribing PTH as per FDA and EMA guidelines. It is also self-evident that PTH should not be used in patients with hypercalcemia of any origin. Because the incidence of hypercalcemia was higher in patients with renal impairment, baseline creatinine clearance is also a helpful baseline measurement. Finally, as the strategies adopted by PTH trials varied, clinical recommendations on monitoring and management of hypercalcemia in PTH-treated patients are currently lacking. It is common to measure the serum calcium within the first month after starting therapy. If the calcium is elevated, it should be repeated. If the serum calcium continues to be elevated, the calcium supplement should be reduced or eliminated. As useful algorithm to follow comes from the PaTH trial.84


A dose-dependent incidence of hyperuricemia of about 3% was found in the teriparatide pivotal fracture trial in patients with normal renal function,82 with the highest occurrence in patients with moderately impaired renal function. In the pivotal fracture trial of PTH(1–84), serum uric acid remained within the normal range.31 In neither trial was there any increased risk of clinical manifestations, such as gout, nephrolithiasis, or arthralgias.20, 31

Only in patients with a history of hyperuricemia or renal, arthritic, or soft tissue manifestations of altered uric metabolism would it seem helpful to measure the serum uric acid prior to embarking upon therapy.

Other safety issues

The most common and pervasive, but not troublesome, adverse events that have been confirmed in clinical practice over the past 10 years are headache and upper gastrointestinal (GI) symptoms such as nausea and vomiting. Although the incidence of such events ranged between 8% and 28.5% in the two pivotal trials,20, 31 they do not appear to be as common with wider clinical experience. Because PTH is a vasodilator, postural hypotension can sometimes be heralded by dizziness, but dizziness can also occur without any change in blood pressure. There are no substantial changes in heart rate with use of PTH. No other manifestations of cardiovascular involvement (e.g., change in electrocardiogram) or impaired renal function or nephrolithiasis have been noteworthy.

Safety in hypoparathyroidism trials

Clinical trials of both teriparatide and PTH(1–84) in patients with hypoparathyroidism have not shown, after a treatment period of up to 4 years, and in one instance of a 6-year-old child treated for 14 years, any report of osteosarcoma or other malignancies.38–42, 47, 85 Winer and colleagues39 reported no cases of high serum or urinary calcium levels in hypoparathyroid patients treated with teriparatide twice daily over a 3-year period. Data from PTH(1–84) trials showed transient episodes of mild hypercalcemia, accounting for only 4% of all measurements and a significant but minimal and isolated 3-month increase in urinary calcium excretion among 30 subjects receiving the hormone at the dose of 100 µg every other day for 24 months.42 The 4-year data reported three transient episodes of mild hypercalcemia in three of 18 subjects and no episodes of hypercalciuria.47

Compliance and Persistence of PTH Use

Both teriparatide and PTH(1–84)are generally well tolerated, demonstrating high levels of persistence and compliance. The landmark trial conducted by Neer and colleagues20 found that only 6% of patients taking teriparatide withdrew before the end of the study; high rates of persistence have also been found in subsequent studies in the United States and Canada.28, 86–88

Trials that have reported compliance rates of PTH(1–84) demonstrate a range of 75% to 97.6% compliance30, 32, 84, 89; in the TOP study, 64% of the PTH-treated cohort and 70% of the placebo arm completed the study.31 A British observational study showed a persistence rate of 87% at 12 months of treatment; this was likely influenced by the comprehensive education and follow-up program provided to each patient on PTH.90 Similar results were seen in a French study looking at persistence in postmenopausal patients enrolled in an education and follow-up program and found that 81.5% were still on treatment at 15 months.91 Recently, Black and colleagues92 have shown that those who comply with therapy of PTH(1–84) show better densitometric results.


With a decade of experience with teriparatide and more than 6 years of experience with PTH 1–84, these two medications have been shown to be effective for the treatment of osteoporosis for both men and women at high risk for fracture and for glucocorticoid-induced osteoporosis. The cumulative experience has not revealed safety issues that were not recognized during the clinical trials. The significantly decreased fracture risk and low side-effect profile make these medications attractive options for situations in which indications are met. Newer potential indications such as in hypoparathyroidism and to accelerate fracture healing are attractive. Whereas osteosarcoma continues to be a theoretical issue, the clinical experience with both forms of PTH has not substantiated such concerns.


JPB is a consultant for Eli Lilly, NPS Pharmaceuticals, Merck, GSK, and Amgen, and receives research support from NPS Pharmaceuticals and Amgen. CC and DI state that they have no conflicts of interest.


Authors' roles: CC, DI, and JPB were all involved in the drafting and revision of this manuscript.