Dr Gagel serves as a consultant for Proctor & Gamble, received speaker's fee from Eli Lilly and Company, Merck, Novartis, Pfizer, and Proctor & Gamble, and received grants and research funding from Novartis and Pfizer.
IN NOVEMBER 2002, the Food and Drug Administration (FDA) approved teriparatide (Forteo) for the treatment of osteoporosis. Teriparatide contains the first 34 amino acids of human PTH. We will use the term teriparatide to describe PTH(1-34) produced by recombinant DNA technology by Eli Lilly & Co. (Indianapolis, IN, USA) and marketed as Forteo. The peptide produced by any other technology will be called PTH(1-34). The approval of teriparatide came after an eventful period of preclinical and clinical studies and was not without controversy. The approval process required the FDA to balance the potential oncologic risks with the clearly established benefits of this unique pharmacologic agent. This situation was created by preclinical data that showed the development of osteosarcomas in a high percentage of rodents treated with large doses of teriparatide for most of their lifespan. Aligned on the side advocating approval of teriparatide was the pharmaceutical maker (Eli Lilly & Co.), some patient groups representing those patients with severe osteoporosis, and a sizable component of basic and clinical investigators studying bone disease. Aligned against approval were several citizen watch groups who vigorously opposed human use of this agent because of concern regarding the potential development of osteosarcoma. Most vociferous was Public Citizen (www.publiccitizen.org), an organization actively opposed to use of this agent because of its potential carcinogenicity. The FDA had been under fire because of recalls of several previously approved pharmaceutical agents; the ensuing criticism created a tense regulatory environment. Perhaps the most interesting role was that of the metabolic bone disease community. After an initial period of surprise regarding the osteosarcoma data, researchers performing studies with teriparatide or other preparations of PTH(1-34) or PTH(1-84) examined their own patient populations treated with the hormone, and no reports of an association of osteosarcoma with PTH therapy have appeared. After a 1- to 2-year period of sorting through their experiences, some extending to the mid-1980s, the reality set in that a novel pharmaceutical agent with shown effectiveness to stimulate new bone formation and little evidence of toxicity in humans might be disapproved. Furthermore, the absence of long-term patent protection for this agent created minimal incentive for the manufacturer to perform the long-term studies that would be necessary to prove safety in a definitive manner. Indeed, as events throughout the pharmaceutical industry have shown, the financial returns from the marketing of this agent (net revenue of $5.6 million in 2002, $56 million in 2003, and $239 million in 2004) would have to be balanced against the potential impact of litigation in the event that there was an adverse effect. More importantly, investigators in the field understood that disapproval of this agent would likely suppress future research using the PTH receptor system and its downstream signaling pathways as a pharmacologic target to stimulate bone formation. The prospect that research on the one signaling system that, at the time, was proven to increase bone formation and had no demonstrable toxicity in humans, was alarming to many basic and clinical researchers in the bone field and led to an advocacy role on behalf of teriparatide from a normally nonpolitical group once the company made a decision to move forward with a request for approval. The FDA chose a middle ground, one that permitted use of this agent for patients with severe forms of osteoporosis, accepting the time-tested tenet that the reduction in morbidity from the use of this agent in high-risk groups far outweighed the hypothetical risks.
The FDA approved teriparatide for treatment of osteoporosis in men and women but placed significant limitations on its application. Use of the drug was limited to a 2-year period and exclusion of patients considered at high risk for osteosarcoma (those with a history of radiation exposure to bone, Paget's disease of bone, and children with open epiphyses) and a recommendation that the pharmacologic agent be used only for those with substantial fracture risk. Despite these limitations, the decision was discussed broadly by experts in the bone community and was strongly supported by many academic and clinical investigators. In addition to the availability of this agent for those patients with severe osteoporosis, implicit in the approval process was a strategy to establish that this agent was indeed as safe as most researchers in this field considered it to be. As a part of the approval process, the manufacturer agreed to conduct a long-term (10-year) surveillance study at several referral centers to assess whether any individuals who developed osteosarcoma had been previously treated with teriparatide. In June 2003, the European Agency for the Evaluation of Medicinal Products also approved teriparatide (Forsteo) for treatment of women with osteoporosis.
We are now >2.5 years down the road. Through August 2005, the number of patients treated with teriparatide worldwide was estimated to be >205,000. Data collected from ∼15,000 retail pharmacies in the United States indicate that ∼90% of the patients are female, and the average age is 68 years (median, 71 years). The age group ≥70 years represent those who have the highest risk of fracture and would be expected to benefit most. Within the limitations of prescription audit data, 82% of patients receiving teriparatide had one or more prior fractures and ∼50% had previously filled a prescription for an antiresorptive drug (bisphosphonate, estrogen or estrogen/progestin combination, or a selective estrogen receptor modulator [SERM]). Among drugs used to treat osteoporosis in the United States, bisphosphonates constitute ∼77%; the remaining drugs (SERMs, calcitonin, teriparatide, or hormone replacement) constitute in aggregate the remaining 23%. Teriparatide prescriptions account for <3% of the total. Each patient in the United States who fills or renews a prescription for teriparatide receives a document called a “Medication Guide” that describes the drug, how it should be taken and used, and possible side effects. It is recommended that all patients treated with teriparatide receive vitamin D and calcium supplementation if dietary sources are likely to be insufficient.
Although precise estimates of compliance with the recommended treatment regimen and persistence of use are not available, the duration of teriparatide use averages 12 months, a number that compares favorably with the average duration of use for bisphosphonates of 8-9 months.(1,2) This may be because the patients using teriparatide have more severe osteoporosis and are therefore more compliant. In the following sections, we will address a number of issues, both scientific and pragmatic, associated with the use of this agent.
EXPERIENCE WITH SELF-ADMINISTRATION
Teriparatide is self-administered by daily subcutaneous injection of 20 μg in the thigh or abdomen using a pen-like device. Recognizing that many of the patients would be elderly and some would be visually impaired, Lilly created an educational program for patients who plan to use this agent. This instruction, by a trained registered nurse or other health care professional, includes hands-on use of the pen with self-administration in the presence of the educator. Approximately 8000 group sessions involving >22,000 patients have been held during the first 2 years since launch. This experience has proven important. As might be anticipated for this elderly population with little or no experience with parenteral administration of medications, difficulty in the use of the pen device is the most frequent complaint of patients taking teriparatide and the most common reason for patients to contact the manufacturer's call center. Improvements in pen design were introduced in mid-2004.
Absence of osteosarcoma in treated patients
As described in a previous perspective article,(3) near-lifetime studies in Fischer 344 rats showed that teriparatide induced the formation of osteosarcomas at exposure multiples over the dose approved for human use (20 μg/day) of 3-58 times. In the first rat study, bone tumors were observed in all treatment groups.(4) In a second rat study to evaluate dose, treatment duration, and age at initiation, no bone tumors were observed at 3-fold the human exposure in rats treated for 20 months, ∼80% of their lifetime.(5)
That this is a class effect of activation of the PTH receptor and not unique to the first 34 amino acids of PTH is shown by the occurrence of osteosarcomas in rats treated for 24 months with human PTH(1-84) (PREOS; NPS/Allelix, Ontario, Canada) at doses of 50 and 150 μg/kg/day, representing clinical exposure multiples of 27 and 66 times that of the dosage currently being studied in humans (Table 1). According to the report from NPS, the effects were dose-related, with no statistically significant effect seen in rats treated with 10 μg/kg/day, a clinical exposure multiple of 4.6. There have been no reported carcinogenesis studies directly comparing teriparatide to recombinant PTH(1-84) (PREOS). Collectively, however, these findings are consistent with the hypothesis that dose and duration of treatment with PTH are relevant to the formation of osteosarcomas in rats.
Table Table 1.. Incidence of Osteosarcomas in Rats Treated With Either Teriparatide or PTH(1-84) for 24 Months
What has happened in humans? In Lilly-sponsored clinical trials of teriparatide studying 1943 patients, no patient treated with either 20 (the FDA-approved dose) or 40 μg/day has developed osteosarcoma. All ongoing human teriparatide studies were stopped in 1998 at the time that osteosarcomas were identified in rodents; these patients have been followed for >5 years since their last exposure to teriparatide. In addition, there have been no reports of osteosarcoma in any patient previously treated with teriparatide, PTH(1-34), or PTH(1-84). Some of these studies date to the 1980s,(3) providing ample time for osteosarcoma development (if there were a latency between treatment and development of an osteosarcoma), although it is unclear how thoroughly the investigators in these prior studies examined the records of patients who participated in those studies. The recent failure by physicians in the bone and oncology communities to identify the development of osteonecrosis associated with high-dose intravenous bisphosphonate use (because most patients complained to their dentists rather than to their bone or oncology physicians) provides a cautionary note regarding the inadequacy of ascertainment when the side effect is unexpected.(7-9)
Finally, there have been no reports of osteosarcoma in any of the 205,000 patients treated through August 2005 since teriparatide approval in the United States or Europe. It can be estimated that >30-40% of these patients have been on teriparatide for >1 year, and an increasing percentage of patients are now completing their second year of therapy. It is important to recognize that the first osteosarcomas in the rats treated with 20- to 58-fold clinical exposure multiples of teriparatide did not develop until 1 year of exposure had been completed. Thus, the absence of osteosarcomas, while reassuring, should be considered an initial progress report rather than conclusive proof of the long-term clinical safety of this pharmacologic agent.
This issue has also been examined from a different perspective. There have been reports of the coincidence of primary hyperparathyroidism and osteosarcoma in <10 individuals.(10-12) Reasoning that the long-term exposure of bone to PTH seen in primary hyperparathyroidism would approximate some aspects of the exposure to an exogenous preparation of PTH, albeit with the important difference of continuous versus once-a-day exposure, Jimenez et al.(13) examined a cohort of 1234 patients with osteosarcoma who presented to a tertiary referral center over a 56-year period to determine whether the prevalence of hyperparathyroidism in this population was greater than expected in the normal population. Approximately 80% of these patients had serum calcium measurements during their evaluation. Three patients in this cohort had both osteosarcoma and primary hyperparathyroidism. However, a comparison of the prevalence of hyperparathyroidism in the osteosarcoma and normal populations showed no significant difference between the two groups. Whereas these findings do not exclude a relationship between hyperparathyroidism and osteosarcoma with certainty, they do suggest that a relationship, if it exists, is not robust. Again, it needs to be recognized that continuous exposure to endogenous PTH in hyperparathyroidism does not mimic the pharmacologic profile obtained with daily exogenous administration of the hormone. Finally, addressing this question prospectively, a 10-year case-finding surveillance study is currently underway by Eli Lilly & Co. to determine whether a sample of men or women, ≥40 years of age with newly diagnosed osteosarcoma, have ever received treatment with teriparatide or PTH(1-34).
The FDA and the manufacturer have adopted a cautious stance regarding use of teriparatide in patients who might have a higher risk of osteosarcoma. Teriparatide use is contraindicated in patients with Paget's disease of bone or unexplained elevations of alkaline phosphatase, patients with a prior history of radiation to the skeleton, children or young adults with open epiphyses, or in patients with bone metastasis. Importantly, both the manufacturer and FDA have reaffirmed the importance of adhering to these guidelines. One of us (RFG) works in a cancer center and inquired whether incidental radiation to ribs during external beam radiotherapy for treatment of breast cancer constituted skeletal radiation. After uncovering several examples of presumed radiation-induced osteosarcomas in the rib cages of patients exposed to external beam radiotherapy,(14) both the manufacturer and the FDA concluded that patients with any skeletal radiation should be excluded from treatment with teriparatide. That it will be important to continue to educate physicians regarding these prohibitions is highlighted by two examples. The first is a patient treated by one of us (RFG) who had breast carcinoma treatment >25 years ago at another institution and had forgotten she had received external beam radiotherapy. Her memory was triggered by the educational materials and package insert provided by the manufacturer. In a second case, seen by a colleague, teriparatide had been initiated shortly before by a primary care physician, despite a clear history of prior radiation to the upper chest wall.
As long-term, real world experience regarding the safety of teriparatide accrues, physicians should keep the prohibitions regarding its use firmly in mind. Having convinced the regulatory agencies that this is a safe agent for most patients, it will be important that we in the bone field not jeopardize the appropriate use of this pharmacologic agent by applying this therapy indiscriminately in patients at greater risk for osteosarcoma. The development of several cases of osteosarcoma in patient populations at higher risk for osteosarcoma, whether or not caused by teriparatide, could force the FDA to take regulatory action that would jeopardize its use in the larger groups of patients with severe osteoporosis.
Teriparatide is contraindicated in patients with cancer and known bone metastasis. It has been established that increased osteoclastic bone resorption plays an important role in the localization and growth of bone metastases. There is abundant evidence that inhibition of osteoclast function by treatment with bisphosphonates retards the development of bone metastasis in breast cancer.(15) Intermittent treatment with teriparatide, in addition to stimulating bone formation, also enhances bone resorption. Current hypotheses regarding the mechanisms of osteolytic bone resorption associated with breast cancer invokes a pathophysiologic role for PTH-related peptide (PTHrP), a peptide that binds to and activates the same receptor as PTH, in the pathogenesis of breast cancer metastasis.(16) PTHrP is thought to activate osteoclastogenesis, leading to increased bone resorption. It has been further postulated that release of TGF-β, or other matrix proteins, by osteoclast-mediated bone resorption further stimulates breast cancer growth and PTHrP production.(17) While an attractive hypothesis, these findings can't be reconciled with the results of long-term clinical trials where there is clearly defined evidence that patients with PTHrP-producing breast cancers survive longer and with fewer bone metastases than those whose tumors do not produce PTHrP.(18) These seemingly contradictory observations, which have now been extended to 10 years of follow-up (MA Henderson, JA Danks, JL Slavin, GB Byrnes, PFM Choong, JB Spillane, JL Hopper, TJ Martin, unpublished data, 2006) suggest the need for additional prospective clinical study.
A similar logic can be applied to myeloma and other malignancies associated with increased bone resorption. There is very compelling evidence in myeloma that bisphosphonate treatment has reduced the incidence of skeletal complications related to bone loss. Again it is unclear whether treatment with a substance that enhances osteoclast formation would be helpful or detrimental. Until these issues are addressed in controlled trials, teriparatide should not be used in patients with documented or suspected bone metastases or in patients with lytic changes associated with myeloma or other hematologic malignancies.
A more difficult dilemma is the patient with a past history of breast carcinoma who has developed severe osteoporosis. One of us (RFG) has used teriparatide in several patients with severe osteoporosis (vertebral fractures and very low BMD) treated for breast cancer more than a decade earlier. In each of these cases, care has been taken to exclude a prior history of skeletal radiation. The absence of any prospective clinical trials in such patients suggests that these cases should be evaluated on a case-by-case basis; because the percentage of patients with a past history of malignancy will grow substantially over the next several decades, prospective clinical investigation should be pursued in this group.
Other adverse effects
The original cohort of 1943 patients treated with teriparatide has been followed for a mean period of 5 years to determine whether cardiovascular disease, diminished renal function or urinary calculi, or vertebral fractures occur with greater frequency after cessation of teriparatide. A review of these data by Eli Lilly & Co. has identified no discernible increase in any of these pathologies.(19)
Based on postlaunch surveillance, Eli Lilly & Co. has updated the teriparatide package insert label (http://www.fda.gov/cder/foi/label/2004/21318s004lbl.pdf) to include a section on “postmarketing reports.” This section includes the following listing of adverse events with associated spontaneous reporting rates—“possible allergic events soon after injection: acute dyspnea, facial edema, generalized urticaria, chest pain (<1 in 1000 patients treated); hypercalcemia >2.76 mM (11 mg/dl; <1 in 100 patients treated; hypercalcemia >3.25 mM (13 mg/dl; <1 in 1000 patients treated).” (Postmarketing surveillance for hypercalcemia is based on spontaneous safety reports to the company and does not represent an incidence rate.)
Abnormal plasma calcium concentrations, when detected, have generally been treated by adjustment of supplementary dietary calcium and/or vitamin D intake. Alternatively, stopping drug administration has also been reported.
“Injection site and injection technique events including pain, swelling, erythema, localized bruising, pruritus, and minor bleeding at the injection site (<1 in 30 patients treated). These usually have been mild and transient.”
In addition, a section on “overdosage” was added to the teriparatide label.
“In postmarketing spontaneous reports, there have been cases of medication error in which the entire contents (up to 800 μg) of the teriparatide pen have been administered as a single dose. Transient events reported (in such patients) have included nausea, weakness/lethargy, and hypotension. In some cases, no adverse events occurred as a result of the overdose. No fatalities associated with overdose have been reported.”
CLINICAL USE OF TERIPARATIDE
Two and a half years have passed since teriparatide was introduced into general use in the United States and Europe. Over the next year, increasing numbers of patients will have completed 2 years of therapy. In most cases, bone mass of normal quality will have increased substantially. Physicians are now addressing the question of what to do after cessation of teriparatide. In addition, there are questions regarding combination of teriparatide with other osteoporosis therapies. Many of these topics have been addressed in recent reviews by Mitlak,(20) Miller et al.(21) and, more recently, by Hodsman et al.(22)
We wish to comment on several frequently addressed areas, including efficacy and optimizing therapy, although definitive data are still lacking on a number of the important clinical questions.
In which conditions is PTH effective?
Teriparatide has been approved for treatment of osteoporosis in postmenopausal women with osteoporosis and men with primary or hypogonadal osteoporosis. The evidence for efficacy has been reviewed extensively,(20-22) and only the highlights of this topic will be addressed in this Perspective. In addition, we will review the evidence for efficacy in glucocorticoid-induced osteoporosis.
The highest level of teriparatide efficacy has been observed in postmenopausal women with osteoporosis. The Fracture Prevention Trial studied 1637 postmenopausal women with a prevalent fracture who were not receiving hormone replacement therapy (HRT) or other antiresorptive treatments. These patients were randomly assigned to self-administer teriparatide 20 μg, teriparatide 40 μg, or placebo injection. This study was terminated after a median time of treatment of 19 months because of the unexpected finding of osteosarcomas in a 2-year rat carcinogenesis study.(4) Analysis of the data set showed a 65% reduction in the risk of vertebral fractures in women taking 20 μg of teriparatide and a 69% reduction in the group treated with 40 μg. Total nonvertebral fractures were reduced 53% and 54%, respectively, in the 20- and 40-μg groups. Because the number of nonvertebral fractures was small, the effect of teriparatide at any single skeletal site was not statistically significant. An analysis of this data set showed no difference in efficacy at the 20- and 40-μg doses and less hypercalcemia with the 20-μg dose. A 20-μg/day dose was approved by the FDA.
The data in men were equally compelling, but the premature cessation of the trial designed to establish fracture efficacy precluded establishment of fracture prevention efficacy. In this trial, 437 men were randomized to teriparatide 20 μg, teriparatide 40 μg, or placebo. The study was terminated because of the finding of osteosarcomas in rats as described earlier.(4) After 11 months of therapy, spine BMD was increased by 5.9% above baseline in the 20-μg group, and in the femoral neck, BMD increased by 1.5%.(23) Treatment duration was too short to detect significant effects on fracture risk. BMD responses to teriparatide were similar regardless of gonadal status (49% had low testosterone) or baseline BMD. Based on these and other smaller studies that also showed efficacy of PTH(1-34),(24,25) the FDA approved teriparatide for use in men. In a follow-up of the original study,(23) men who received teriparatide and who may have received subsequent antiresorptive therapy had a decreased risk of moderate and severe vertebral fractures.(26)
There is also considerable evidence for efficacy of PTH(1-34) in glucocorticoid-induced osteoporosis.(27,28) A 1-year randomized study comparing BMD in women on glucocorticoids (5-20 mg/day prednisone equivalent) taking estrogen alone or estrogen and PTH(1-34) showed efficacy.(27) Those on estrogen and PTH had a 35% (QCT) or 11% (DXA) increase in lumbar spine BMD compared with 1.7% (QCT) or 0% (DXA) in the estrogen-only group. In this study, there were no significant hip or forearm BMD differences between the two treatment groups. There were no fracture data. This study is also interesting because it was perhaps the first to show a continued effect of PTH when combined with another agent after cessation of the drug. The PTH was discontinued after 12 months of therapy, and both groups were continued on estrogen and glucocorticoid therapies.(28) In the group treated with PTH/estrogen/glucocorticoid during the first year, BMD rose from 35% to 46% (QCT) and from 11% to 13% (DXA) increment. This contrasts with a 1.7% to 1.3% (QCT) or 0% to 2.6% (DXA) change in the estrogen/glucocorticoid-only group. There were also small positive changes noted at the hip in the PTH-treated group during the second year. Not only was the PTH effective during therapy, but effects continue for some period of time after cessation.
Should PTH be used in combination with other bone-active pharmacologic agents?
The therapeutic effects of teriparatide or PTH(1-84) on bone are complex. There are not only different effects on trabecular and cortical bone, but the temporal sequence of change in these two different types of bone also differs. Daily treatment with PTH causes an increase of lumbar and hip trabecular bone that is continuous over a 2-year period of treatment. Cessation of PTH after a year of therapy is associated with stability of bone mass in some studies(29) and loss in others.(30) Daily treatment with teriparatide or PTH(1-84) causes loss of cortical bone at the hip and distal third of the radius, measured by either DXA or QCT.(31,32) Continued treatment results in resorption of bone on the endosteal surface and deposition of bone on the periosteal surface, combined with increased endocortical porosity. Cessation of PTH after 1 or 2 years of therapy results in filling in of the resorbed cortical bone and stability or small increases of cortical bone.
The aforementioned changes in cortical bone occur over an extended period of time during continuous daily treatment (2-3 years), and a single snapshot of changes at different points in this temporal sequence may yield very different results. The addition of a second therapeutic agent (all those studied have been antiresorptive agents) must be analyzed in the context of the underlying temporal changes.
Several studies have addressed the question of concomitant use of teriparatide or PTH(1-84) in combination with a potent oral bisphosphonate, alendronate. In the most thorough study of this issue, Black et al.(30) examined the effect of PTH(1-84), PTH(1-84) and alendronate, or alendronate during a 1-year trial. Evaluation of the QCT data set was informative. PTH(1-84) caused a significantly greater increase of trabecular bone (spine or hip) than either PTH(1-84) and alendronate or alendronate alone. The effects on hip cortical bone (total or femoral neck), measured by QCT, were exactly the opposite. There was a small increase with alendronate alone, no change with PTH(1-84) combined with alendronate, and loss of bone with PTH(1-84). A similar pattern was observed for these three groups in the assessment of bone by DXA at the distal third of the radius, an area composed largely of cortical bone. No fracture data were available.
A second study by Finkelstein et al.(31) showed similarly complex effects of combining PTH(1-34) and alendronate. This study differed from the one described above in several important ways. A higher dose of PTH(1-34) (37 μg) was used than the current approved dose of 20 μg/day, and PTH was started at 6 months in the PTH alone and PTH-alendronate groups. BMD was assessed by DXA or QCT (lumbar spine only). At the conclusion of 30 months, BMD was highest in the PTH(1-34) alone group (after the initial 6 months of alendronate) in the lumbar spine, femoral neck, and total hip, whereas BMD at the distal radius was highest in either the alendronate or combination therapy group compared with PTH(1-34) alone. Total body BMD was increased with all groups, but there were no statistically significant intergroup differences. Alendronate or alendronate combined with PTH(1-34) produced linear increases in BMD over the 30-month period of study. In contrast, PTH(1-34) alone stimulated early increases of BMD in the spine; increases at the femoral neck and total hip were delayed until a rapid rise during the last 12 months of the study. These results provide additional confirmation of the complex effects of PTH at different skeletal sites and the even more complex interpretation of results when PTH and a bisphosphonate are combined.
These studies raise a number of interesting questions regarding the combination of PTH and a potent bisphosphonate. The protective effect of alendronate on BMD in cortical bone, particularly early in the course of PTH treatment, suggests that combining PTH with a low bisphosphonate dose might provide greater protection for cortical bone without interfering with trabecular bone formation. However, in the absence of specific studies including fracture data showing such benefit, most in the osteoporosis field have chosen to use bisphosphonates after PTH as will be described below.
Another unresolved issue is whether bisphosphonate therapy should be stopped for a period of time before initiation of teriparatide therapy. In contrast to the inhibitory effect of bisphosphonate therapy on bone formation and resorption, there is evidence that the efficacy of PTH therapy on bone structure is tightly correlated with its ability to increase biochemical markers of bone formation.(33) Increases in procollagen I C-terminal propeptide (PICP) at 1 month and procollagen I N-telopeptide (PINP) at 3 months were the most sensitive and accurate predictors of the response to teriparatide in lumbar spine BMD.(34) Furthermore, a number of the more potent bisphosphonates have a prolonged and profound suppressive effect on markers of bone resorption, leading some to recommend discontinuance of bisphosphonate therapy for a period before the initiation of teriparatide therapy.
This issue was also addressed in a recent report from Cosman et al.,(35) who pretreated all patients entering a study for 1 year with alendronate. One-third of this group continued on weekly alendronate; one-third were treated with weekly alendronate and PTH(1-34) (25 μg/day); and one-third received weekly alendronate with a 3-month period of PTH(1-34) (25 μg/day) followed by a 3-month period without PTH, cycling for 15 months or three cycles of PTH(1-34). Two findings were of significance. First, PTH(1-34) stimulated an increase of bone formation markers (procollagen, osteocalcin, and alkaline phosphatase) during each of the intermittent PTH cycles. The magnitude of increase in the cyclic PTH group was much greater than observed with alendronate alone, but less than in the weekly alendronate/daily PTH group. Second, there were roughly comparable increases in BMD measured by DXA at the lumbar spine and total hip with either the continuous or cyclic PTH, and both were significantly greater than with alendronate alone. Unfortunately, this study did not examine any site by QCT, making it impossible to determine whether there were differential effects on trabecular and cortical bone. In addition, the overall increase in lumbar spine and total hip BMD was less in either of these groups than in many other studies in which daily PTH alone was examined.
A recent presentation by Cosman et al.(36) also suggests that initiation of PTH immediately after cessation of alendronate is associated with an adequate bone formation response. They rechallenged a group of patients previously treated with PTH(1-34) followed by alendronate for a year with a second course of PTH(1-34). Bone formation and resorption markers during the second exposure to PTH were similar to those observed during the first course of therapy.
It is possible that suppression of osteoclast function, by whatever means, diminishes bone formation caused by PTH(1-34). However, it has been reported that blockade of the osteoclast chloride channel leads to inhibition of bone resorption without changing bone formation, suggesting that selective pharmacologic blockade of the resorptive, but not other, functions of the osteoclast may be possible.(37) Another chance to explore this observation will be provided by osteoprotegerin (OPG) or the monoclonal antibody, AMG 162, that binds to RANKL and prevents activation of the RANK receptor. In mice, disruption of this pathway by overexpression of osteoprotegerin inhibits bone resorption but does not affect bone formation, leading to increased bone mass.(38) Curiously, treatment of mice with OPG leads to periosteal deposition of bone.(39) Whether there will be similar effects in humans is the subject of phase II studies; if there are similar effects in humans as in mice, it would be logical to combine one of these agents with teriparatide.
There have been no prospective studies that address the combined effect of PTH and estrogen combinations. It may be inferred from the robust increase in BMD in patients on estrogen therapy in whom PTH(1-34) was added that estrogen does not seem to inhibit the bone formation-enhancing effect of PTH(1-34).(40,41)
Combined use of teriparatide or PTH(1-34) and raloxifene has been examined in rodents and humans. Studies in rodents showed that pretreatment with raloxifene does not prevent PTH(1-34)-mediated increases in bone mass.(42) In other studies, it has been shown that a SERM enhances PTH-stimulated bone formation in rodents.(43) Results in humans have shown that prior treatment with raloxifene does not blunt the effect of teriparatide.(44) The findings of Cosman et al.(45) and Deal et al.(46) also indicate no attenuation of the beneficial effect of teriparatide when used together with raloxifene. Clearly long-term prospective studies are needed.
There are a number of potential bone-active compounds that have not been studied in conjunction with teriparatide or PTH(1-84). Notable among them are vitamin D analogs. Vitamin D metabolites or analogs have been shown in a variety of studies in rodents and humans to increase bone mass, and 1,25 dihydroxyvitamin D3 also contributes to osteoclast differentiation. Clinical studies of calcitriol for treatment of osteoporosis have shown an increase in vertebral BMD, but no fracture data have been reported, and there is a significant incidence of hypercalcemia and hypercalciuria.(47) Combining vitamin D analogs with PTH in humans has been limited or precluded by the potential for development of hypercalcemia in patients treated with both. The introduction of vitamin D analogs with reduced calcemic effects and positive effects on bone formation provide an opportunity to study such analogs in combination with PTH.(48)
Other agents that affect bone formation include fluoride and strontium. It is not clear whether combining these agents with PTH would be of benefit.
What to do after completion of 2 years of therapy with teriparatide?
A question that is now confronting clinicians using this agent is whether to initiate (or reinitiate) an antiresorptive drug after 2 years of teriparatide treatment. In the previously reported Fracture Prevention Trial using teriparatide, 1637 postmenopausal women with a prevalent fracture, who were not receiving HRT or other antiresorptive treatments, were randomly assigned to self-administer teriparatide 20 μg, teriparatide 40 μg, or a placebo injection.(49)
In a subsequent analysis to address long-term efficacy, a group of 1262 women from the Fracture Prevention Trial was followed for 28 months after stopping teriparatide. They were treated according to standard medical practice, with elective osteoporosis drugs other than teriparatide being used in 47% of the women in the follow-up period. A 41-45% reduction in new vertebral fracture risk was observed in women who had been treated with teriparatide in the original trial in comparison with women treated with placebo in the original trial, suggesting a persistent beneficial effect of teriparatide for at least 18 months after stopping the drug.(29) This study has limitations in that ∼45% (44% in the 20 μg and 46% in the 40 μg teriparatide dose groups) also received some other form of osteoporosis treatment during the follow-up period, raising the possibility that it was some combination of teriparatide and another osteoporosis agent (most commonly a bisphosphonate) that contributed to persistent fracture reduction effect. Although the authors indicate that there did not seem to be any effect of bisphosphonate use, no specific data are presented. More importantly splitting the sample size of each group by roughly one-half reduces the power of this analysis. Indeed, the authors hint at the fact that bisphosphonate therapy may further enhance the effect of teriparatide by presenting data showing that bone mass increments caused by teriparatide are further enhanced by the subsequent use of bisphosphonate, providing a preview of what will be shown more definitively in studies discussed below.
More recently Prince et al.(19) examined the fracture outcome in the cohort of patients included in the Fracture Intervention Trial 30 months after cessation of teriparatide therapy. Approximately 50% of each group (teriparatide or control) were treated with a bisphosphonate for some period of time during the 30-month period, and a higher percentage was on a bisphosphonate at the conclusion. A comparison of hip bone mass during this follow-up period documents that patients treated with bisphosphonates after discontinuation of teriparatide shows continued increments in BMD, whereas those who received no treatment had declines. These studies are also limited in the same sense as the Fracture Prevention Trial described above—it was difficult to know whether to attribute the effect to the PTH or the combination of PTH and alendronate because of the lack of control over administration of bisphosphonates.
In a report by Rittmaster et al.,(50) using PTH(1-84), subsequent treatment with alendronate resulted in greater BMD increments at each dose of PTH(1-84) (Fig. 1). In addition, the finding in this study that bone turnover in the PTH(1-84) group followed by alendronate (as measured by serum osteocalcin, alkaline phosphatase, and N-telopeptide) was elevated above the control group followed by alendronate at the conclusion of the study (a full year after cessation of PTH), argues that effects of PTH to increase bone turnover persist long after it is discontinued.
Kurland et al.(51) have reported similar findings for teriparatide. Men treated with alendronate, risedronate, or cyclic etidronate after PTH(1-34) therapy gained an additional 5.1 ± 1.0% BMD, whereas those placed on no therapy after teriparatide lost 3.7 ± 1.7%. Furthermore, the addition of alendronate or risedronate, initiated 1 year after discontinuing PTH(1-34) therapy, caused a subsequent increase in bone mass comparable with that observed in patients treated with a bisphosphonate at the time of cessation of PTH(1-34). This latter study suffers from small numbers of patients and the lack of a prospective design.
The most compelling evidence for a beneficial effect of a potent bisphosphonate is provided by the recent report from Black et al.(30) Patients were randomized to one of four groups: PTH(1-84) during the first year with no therapy in the second year; PTH(1-84) during the first year followed by alendronate in the second year; PTH and alendronate during the first year followed by alendronate in the second year; and alendronate for 2 years. The results from this trial showed very clearly that PTH(1-84) followed by alendronate was associated with the greatest increment in bone mass, whether measured by DXA (Fig. 2) or QCT (data not shown).
These studies provide compelling evidence that bisphosphonate therapy after PTH(1-34) or PTH(1-84) has value, although there are currently no fracture data. There is less information about the use of other antiresorptive agents, such as raloxifene, after teriparatide treatment, although there are data available that concomitant use of estrogen or raloxifene with PTH does not adversely impact the effect of PTH.(42,52)
Informal discussions with physicians have pointed out another logic for considering antiresorptive therapy after teriparatide therapy. Although teriparatide is the most effective therapy available for restoring bone quality(20,22) and reducing vertebral fractures, it is also the most expensive, and its use is currently limited to a total of 2 years of therapy. Many physicians want to preserve and protect “hard-earned” increments in bone mass, particularly because there is currently no option for reinitiating therapy if bone mass gained during the initial 2 years of teriparatide therapy is subsequently lost. Combining this logic with the aforementioned data showing efficacy of antiresorptive agents after cessation of teriparatide therapy provides a compelling case for initiation or reinitiation of antiresorptive therapy.
Is there an increase of trabecular bone at the expense of cortical bone?
There was early concern that teriparatide therapy might preferentially increase trabecular bone mass at the expense of cortical bone.(53,54) Subsequent reports addressing this question have provided reassurance on this point. The findings of Zanchetta et al.(55) indicated that teriparatide induced beneficial architectural changes in the distal radial diaphysis with no detrimental effects on cortical mineral content. Histomorphometric and μCT studies of 51 iliac crest biopsy specimens showed that PTH increased cancellous bone volume by 14%, increased cancellous connectivity by 19%, and increased cortical thickness by 22% in women with osteoporosis. Similar effects have been observed in the femoral neck, an anatomic location where thinning of the cortex might be expected to have a detrimental effect. A broad spectrum of evidence(19) indicates that periosteal bone formation combined with endosteal remodeling leads to bone with a greater diameter and a cortex with greater thickness. The net effect of these architectural changes is a bone that is biomechanically stronger. Long-term studies that include fracture endpoints will be needed to provide complete reassurance on this effect.
How does teriparatide stimulate increased bone formation?
There is a substantial literature indicating that intermittent exposure to teriparatide is associated with net bone formation, whereas continuous exposure to teriparatide preferentially activates osteoclast-mediated resorption. In a report by Locklin et al.,(56) intermittent exposure to teriparatide enhances transcription of osteoblast differentiation markers Runx2, alkaline phosphatase (TNALP), and type 1 collagen, whereas continuous exposure was associated with increases in RANKL and decreased expression of osteoprotegerin (OPG). This constellation of effects may explain the differences between intermittent and continuous administration teriparatide. A recent report from Murshed et al.(57) argues that mineralization of a particular tissue is dependent on the co-expression of TNALP, through its effect to cleave pyrophosphate, an inhibitor of mineralization, and type 1 collagen. In this report the authors were able to induce bone formation in the dermis by co-expression of these two genes. It is therefore possible that the new bone formation observed during PTH treatment is caused by the enhanced co-expression of these two genes. It will be important to pursue these observations in other experimental systems. The potential to activate bone formation by increasing the expression of only two genes by other strategies, without concomitant activation of bone resorption, is an attractive possibility.
Is increased bone resorption necessary for the beneficial effect of teriparatide?
PTH [teriparatide, PTH(1-34), and PTH(1-84)] causes both an increase in bone formation and bone resorption. In this respect, PTH differs from all prior approved osteoporotic agents including estrogen, SERMs, calcitonin, and bisphosphonates, whose major effects are to inhibit bone resorption. Treatment with teriparatide, PTH(1-34), or PTH(1-84) causes increases of bone formation (increases of serum bone-specific alkaline phosphatase, bone collagen propeptides, osteocalcin, and bone formation rates) and bone resorption (increases of deoxypyridinoline or N-telopeptide and histologic evidence of increased osteoclast-mediated bone resorption). The increase in bone formation markers tends to occur before the increase of resorption markers. Furthermore, in human studies of patients with glucocorticoid-induced osteoporosis treated with PTH(1-34), there is evidence that increments in BMD correlate with increases in bone turnover.(58) Attempts to uncouple the PTH-induced effects on bone formation and resorption have to date been largely unsuccessful. Combined treatment with alendronate and PTH results in a lower rate of bone accretion than treatment with PTH(1-34) or PTH(1-84) alone, suggesting that either increased resorption is needed for bone formation or that the loss of mature osteoclasts caused by bisphosphonates eliminates some signaling process initiated by the osteoclast. These collective findings and evidence obtained from a variety of transgenic mouse models, reviewed by Martin and Sims,(59) argue persuasively for the presence of an essential cross-talk from osteoclasts to osteoblasts, necessary for PTH-mediated bone formation. Several hypothetical mechanisms have been suggested. The first postulates that activation of the RANK receptor system by PTH causes osteoclasts to release a soluble factor that enhances bone formation. A second model postulates osteoclast-mediated cell-cell interaction within the resorption lacuna. A third variant is that osteoclasts cause release of growth factors such as TGF-β from bone, thereby enhancing bone formation.(60) The recent demonstration that teriparatide treatment causes expansion of hematopoietic stem cell precursors has further increased the complexity of cellular interactions within the bone marrow compartment.(61) It is possible that the products of osteoclasts or osteoclast-mediated resorption or osteoclasts themselves interact with a third marrow cell type to facilitate bone formation. A better understanding of the relationship between bone resorption and formation would undoubtedly lead to strategies to uncouple formation and resorption, raising the possibility that bone formation by PTH could be activated independently of resorption.
FUTURE USES OF TERIPARATIDE
Hematopoietic stem cells
The unexpected observation that PTH(1-34) stimulates hematopoietic stem cell proliferation and the finding that irradiated mice subsequently transplanted and treated with PTH engrafted more rapidly than controls had a major impact on the world of bone marrow transplantation.(61) Within months after the appearance of this report, investigators at two institutions had initiated phase I/II treatment protocols to determine whether there are similar effects in humans. Results are awaited.
Short-term treatment of hypoparathyroidism
Before the observation that PTH caused osteosarcomas, PTH(1-34) was used successfully to treat chronic hypoparathyroidism in investigational studies.(62) Until the issues related to oncogenesis are sorted out, long-term treatment of hypoparathyroidism is not an acceptable therapeutic option. Short-term management of postoperative hypoparathyroidism, however, can be safely studied. Patients who have extensive head and neck surgical procedures for invasive thyroid, pharyngeal, or laryngeal carcinoma commonly develop permanent hypoparathyroidism. Management of these patients during the first week after surgery can be challenging. When a head and neck surgical procedure is combined with a bypass procedure to recreate a functioning esophagus (and short-term placement of a jejunal feeding tube), very large doses of oral calcium (6-12 g of elemental calcium/day or continuous intravenous calcium infusion), and vitamin D (50,000-100,000 IU/day) are commonly required to prevent symptomatic hypocalcemia. Hospital stays are routinely prolonged because of the lag between initiation of vitamin D and calcium therapy and normalization of serum calcium; in ∼1-2% of patients this can lead to postdischarge hypercalcemia necessitating readmission. In this situation, study of the potential for short-term treatment with teriparatide in the postoperative period, thereby permitting a more gradual and safe escalation of calcium and vitamin D therapy, seems appropriate.
Teriparatide use in fracture healing and implant fixation
Although there are as yet no clinical trial data in patients on the potential use of teriparatide use in fracture healing or implant fixation, there are some animal data. In rats, it has been reported that daily subcutaneous injections of teriparatide increase callus formation, speed of fracture repair, and mechanical strength.(63-66) Additional studies in rats have indicated that intermittent treatment with teriparatide can enhance metal implant fixation and local BMD and increase implant-bone contact.(67-69) These findings suggest that limited duration clinical trials using teriparatide are justified in selected clinical circumstances to enhance fracture healing or improve the outcome of orthopedic or dental implant surgery.
In the 2.5 years since launch, there has been a steady increase in the use of teriparatide, reaching >205,000 patients worldwide. Expected inconvenience with daily self-injection is noted but is modest. Patient education and improvements in pen design seem useful in decreasing injection-related events. The spectrum of adverse events in the wider community of clinical practice is similar to those observed in the phase III clinical trial,(49) with the most frequent being light-headedness and leg cramps. Elevations of plasma calcium have been reported less frequently than noted in the clinical trial setting. These events have tended to be transient, not symptomatic, and have been managed by adjusting either calcium or vitamin D intake. For rare persistent or marked elevations of plasma calcium, cessation of use is indicated. There have been no reported cases of osteosarcoma in patients taking teriparatide or any other preparation of PTH. A number of unanswered questions remain about how to use this bone anabolic agent to achieve optimum benefit for women and men with osteoporosis, including the prior, concurrent, and post-treatment use of drugs that act principally to inhibit bone resorption. Academic investigator-initiated clinical investigations are underway to examine the potential short-term use of teriparatide to stimulate hematopoietic stem cells and postoperative hypoparathyroidism and to accelerate healing of bone in specific clinical situations such as spinal fusion and dental implants.
The authors thank Julie Larrabee for expert editorial assistance.