Abstract
The carcinogenic potential of human parathyroid hormone 1–84 (PTH) was assessed by daily subcutaneous injection (0, 10, 50, 150 μg/kg/day) for 2 years in Fischer 344 rats. Histopathological analyses were conducted on the standard set of soft tissues, tissues with macroscopic abnormalities, selected bones, and bones with abnormalities identified radiographically. All PTH doses caused widespread osteosclerosis and significant, dose-dependent increases in femoral and vertebral bone mineral content and density. In the mid- and high-dose groups, proliferative changes in bone increased with dose. Osteosarcoma was the most common change, followed by focal osteoblast hyperplasia, osteoblastoma, osteoma and skeletal fibrosarcoma. The incidence of bone neoplasms was comparable in control and low-dose groups providing a noncarcinogenic dose for PTH of 10 μg/kg/day at a systemic exposure to PTH that is 4.6-fold higher than for a 100 μg dose in humans. The ability of PTH to interact with and balance the effects of both the PTH-1 receptor and the putative C-terminal PTH receptor, may lead to the lower carcinogenic potential observed with PTH than reported previously for teriparatide.
Introduction
Parathyroid hormone 1–84 (PTH) is the principal physiological regulator of systemic calcium and phosphate homeostasis. When injected on a daily basis, PTH and related analogs act through the PTH-1 receptor in bone cells to increase bone formation, mass and strength in animals and humans (Dempster et al., 1993; Fox, 2002). Teriparatide, the N-terminal 1–34 amino acid fragment of PTH, has been shown to prevent additional fractures in postmenopausal women with multiple fractures (Neer et al., 2001) and has been approved for the treatment of postmenopausal women with osteoporosis who are at high risk for fractures. However, the administration of teriparatide to Fischer 344 rats in carcinogenicity studies resulted in time- and dose-dependent increases in the incidence of proliferative lesions in bone, which included focal osteoblast hyperplasia, osteoblastoma, osteoma, and osteosarcoma (Vahle et al., 2002, 2004).
This result prompted the United States Food and Drug Administration to issue a draft guidance stating that the carcinogenic potential of PTH and related peptides should be evaluated in nonclinical studies (U.S. Food and Drug Administration, 2000). Prior to the initiation of such studies, it was hypothesized that chronic administration of PTH would be less likely to induce proliferative changes in bone than teriparatide. PTH, unlike teriparatide, contains both the N-terminal region, which acts at the PTH-1 receptor to stimulate bone growth, and the C-terminal region which binds to a distinct PTH receptor that responds only to the C-terminal region of PTH (Inomata et al., 1995; Murray et al., 2005). This suggests a regulatory role for the C-terminal region of PTH in bone formation and is supported by a number of in vitro and in vivo studies demonstrating the opposing effects of the N- and C-terminal regions of PTH on alkaline phosphatase activity and apoptosis in bone cells, bone resorption and turnover, and plasma calcium levels (Murray et al., 1989; Jilka et al., 1999; Slatopolsky et al., 2000; Divieti et al., 2001, 2002; Nguyen-Yamamoto et al., 2001; Langub et al., 2003). This hypothesis was tested in a carcinogenicity study using Fischer 344 rats that received daily subcutaneous injections of recombinant human PTH for 2 years.
Materials and Methods
Animals and Husbandry
Male and female Fischer 344 rats (F344/NHsd) were obtained from Harlan Sprague–Dawley, Indianapolis, IN. At the initiation of the study the animals were 9–11 weeks old. Based on body weight, the rats were randomly assigned to treatment groups (n = 60/sex/group) and housed individually in stainless steel cages. To ensure that rats of each dose group were exposed for similar periods of time to different areas of the room, the animal cage racks were rotated once per month according to a predetermined schedule. Certified pelleted commercial laboratory diet, Rodent Chow 5002 (PMI Feeds, Inc. St Louis, MO), was provided ad libitum. Municipal tap water, which had been softened, purified by reverse osmosis and exposed to ultraviolet light, was freely available. The work was conducted in accordance with American Association of Laboratory Animal Care and Canadian Association of Laboratory Animal Care guidelines and was approved by the research laboratory Institutional Animal Care and Use Committee. Additionally, this study was conducted in compliance with the Good Laboratory Practice Regulations of the Food and Drug Administration (21 CFR Part 58).
Test Article and Study Design
Recombinant human PTH (NPS Pharmaceuticals Inc. Salt Lake City, UT) was administered as a solution in a citrate buffer containing mannitol. PTH doses of 10, 50 or 150 μg/kg body weight were administered daily by subcutaneous injection into the dorsal region for the 2-year period (Table 1). The high-dose (150 μg/kg/day) was expected to be the maximum tolerated dose based on a 6-month rat toxicology study and the low dose (10 μg/kg/day) was expected to provide a systemic exposure to PTH that was at least 2-fold greater than the systemic exposure in humans following a 100 μg dose. The study contained 2 groups of control rats that only received vehicle injections. A 6th group of rats (delayed start high-dose group; DS) was included to assess the skeletal response in the absence of PTH treatment during the period of active bone growth. This group began dosing (150 μg/kg/day) on Week 24 of the study and continued for the next 80 weeks. Dose volume was adjusted weekly for the first 16 weeks and monthly thereafter to account for changes in body weight.
All rats were observed twice daily for mortality and signs of ill health or reaction to treatment. In addition, a detailed physical examination was performed weekly for all rats. From Week 26 onwards, all rats were also examined weekly for the presence of palpable masses. The site, size, and appearance of these masses were recorded when first detected and, following this initial description, their progression was monitored.
Toxicokinetic Analysis
Blood samples were collected from satellite animals in Groups 2, 3, 4, and 5 at 8 time points on 4 separate days throughout the first 12 months of treatment. Blood was obtained by jugular venipuncture without anesthesia into tubes containing ethylenediaminetetraacetic acid. Plasma was stored at −80°C until PTH analysis (Scantibodies Whole PTH Immunoradiometric Assay, Scantibodies Laboratory, Santee, CA). The animals in the satellite groups were euthanized after the 12-month blood sample was obtained.
Noncompartmental toxicokinetic analysis was performed on plasma PTH concentration data. Toxicokinetic analysis included assessment of the tmax, Cmax and area under the curve, i.e., total systemic exposure. Area under the curve was calculated by the trapezoidal method (Gibaldi and Perrier, 1982) using WinNonlin, v3.2 (Pharsight Corporation, Mountain View, CA).
Radiographs
To enhance the detection of hidden or “occult” tumors that might otherwise have gone undetected at necropsy, radiographs of the whole skeleton (dorso/ventral and lateral planes) were performed under isoflurane anesthesia within 2 weeks of study end for all surviving carcinogenicity study rats. A Clinix-R (Picker International, Cleveland, OH) radiography machine was used with a Trimax 6 screen (3M, St. Paul, MN) and TMS/RA (Kodak, Rochester, NY) radiographic film. Radiographs were not taken for unscheduled necropsies. Each radiograph was examined by the study pathologist or veterinarian prior to necropsy in order to determine whether additional bones needed to be collected besides those listed in the protocol.
Necropsy
Animals euthanized on completion of treatment, as well as those euthanized for humane reasons, were anesthetized with isoflurane and exsanguinated from the abdominal aorta. All rats, including those found dead during the study, were necropsied and tissue samples were preserved.
A comprehensive collection of soft tissues was made, together with several routinely sampled bones: femur (distal left), tibia (proximal left), lumbar vertebrae (L5 and L6), sternum, and calvarium. Additionally, representative samples from any tissue with an external or internal macroscopic abnormality and bones with neoplasm-suspect radiological changes were collected. Collected tissues were fixed and preserved in 10% neutral-buffered formalin with the exception of ocular structures and male gonads, which were fixed in Zenker’s fluid.
Bone Densitometry
The right femur and first four lumbar vertebrae (L1–L4) were also collected at necropsy, preserved overnight in 10% neutral-buffered formalin and transferred into 70% alcohol until densitometric scanning. These bones were collected from all main study animals euthanized at the end of treatment, with the exception of those bones with tumor-suspect macroscopic or radiological abnormalities, which were preferentially processed for histological evaluation. Bone mineral content (BMC) and bone mineral area (BMA) were measured by dual energy X-ray absorptiometry with a QDR-2000 plus bone densitometer (Hologic, Bedford, MA) and were used to derive bone mineral density (BMD). Measurements were made on segmented regions of the central and distal femur and the L1–L4 segments.
Histopathology
All specified tissues from all animals were prepared for histopathological examination by embedding in paraffin wax, sectioning and staining with hematoxylin and eosin. Bones were decalcified in dilute formic acid prior to processing. Preparation and evaluation of soft and bony tissues was done as follows:
– All tissues from all animals in both control groups and both groups dosed with 150 μg/kg/day
– All tissues from all animals in the 10 and 50 μg/kg/day groups which died or were euthanized prior to scheduled necropsy
– All gross and radiological abnormalities in all groups
– rior to study start, bones were specified as a target organ and were examined in all groups
– The spleen was identified as a target organ in the 150 μg/kg/day female group and, thus, was examined in 10 and 50 μg/kg/day female groups
The terminology and diagnostic criteria employed to interpret the various lesions seen in this study are broadly consistent with those contained within the Standardized System of Nomenclature and Diagnostic Criteria Guides for Toxicologic Pathology published by the Society of Toxicologic Pathologists (Long et al., 1993). The diagnostic criteria used to classify proliferative changes in bones are based on those used in the teriparatide carcinogenicity study (Vahle et al., 2002). Comparisons were also made with the human literature for some unusual bone lesions (Schwamm and Millward, 1995; Adler, 2000).
A single pathologist evaluated all tissues. Subsequently, a pathology peer review was performed to confirm the microscopic observations. The reported results reflect the mutually agreed-upon diagnoses by the primary and peer review pathologists, both certified by the American College of Veterinary Pathologists.
Statistical Analyses
Analyses were carried out to compare data from the main study test animals with data from each of the control groups and the combined control groups. Mortality data were analyzed using a Peto’s 1-tailed trend test for most pairwise comparisons between groups (Peto et al., 1980); the analysis was performed using SAS Release 8.1 (SAS Institute, Cary, NC). Peto’s 2-sided test was used to compare mortality for both control groups and both groups treated with 150 μg/kg/day. The significance of a linear dose-related increase in tumor occurrence rates was evaluated using Peto’s survival adjusted 1-sided trend test. The analyses were performed in 3 phases; in turn, each control group and the combined control groups were considered. The overall trend test was performed excluding the 150 μg/kg/day DS group and using the arithmetic dose level scores (0, 10, 50, and 150). Peto’s 1-sided trend test was also used to test if the tumor rate in each treated group used for the trend test was significantly higher than the one in the combined control group (pairwise comparisons). Additional pairwise comparisons between both 150 μg/kg/day groups were performed using Peto’s 2-sided trend test.
Each trend test (the overall trend and the pairwise comparison tests) of the tumor data was conducted at the 5% significance level and was interpreted according to the recommendations of Lin and Rahman (1998).
The 1-sided Cochran-Armitage trend test followed by Fisher’s exact test for comparison between groups was used to analyze nonneoplastic lesion data. The dual energy X-ray absorptiometry data were analyzed with Levene’s test and, if the outcome was significant (p ≤0.05), data were subjected to a Kruskal-Wallis test followed by a Wilcoxon-rank-sum test. Analysis of variance and t-tests on least-square means were carried out if the outcome of Levene’s test was not significant. All references to statistical significance refer to comparisons to the combined control groups, unless otherwise noted.
Results
Systemic Plasma Drug Levels
Total systemic exposure to PTH at steady-state was reflected by the 6- and 12-month area under the curve measurements, the mean of which was used for comparison with human exposure (Table 2). The low dose provided a systemic exposure to PTH that was 4.6-fold greater than the systemic exposure in humans following a 100 μg dose. Although not evaluated in this study, no anti-PTH antibody formation was observed in a separate 6-month rat toxicology study employing PTH doses similar to and greater than the doses used in this study (unpublished observations). This finding and the absence of a decline in systemic exposure over the 12 months of toxicokinetic sampling in the current study, suggest that there is no immune response by the rat to human PTH.
Mortality
There was an increased rate of mortality at Week 94 in Group 5 males receiving 150 μg/kg/day compared with the combined vehicle control groups (67% versus 30%, respectively) leading to discontinuation of treatment for Group 5 males at the end of Week 94. For females, there was no significant effect of PTH treatment on mortality (Table 3). The mortality rate for males was notably higher than in females, irrespective of treatment status and was attributed to the greater severity of chronic progressive nephropathy, which was the most common fatal, nonneoplastic condition in males. An increased incidence of fatal osteosarcoma was noted in males treated at 50 μg/kg/day, and males and females treated at 150 μg/kg/day in Groups 5 and 6. The first osteosarcoma fatalities were seen at Weeks 58, 55 and 65 in a control female, a 150 μg/kg/day DS male and a 150 μg/kg/day male, respectively.
Bone Densitometry
There were no significant differences in variances or in means between the 2 vehicle control groups in either gender at any skeletal site, so both vehicle groups were combined prior to statistical analysis. Treatment with PTH resulted in a dose-dependent increase in BMC in both genders at the lumbar spine and central femur, skeletal sites that contain primarily trabecular and cortical bone, respectively (Figure 1).
The increases in BMC were significantly greater with all doses. The increase in BMC was similar in Groups 5 and 6, which received PTH for approximately 2 years and for 18 months, respectively. Similar increases in BMC also occurred at the distal femur (data not shown), which contains both trabecular and cortical bone. Similar changes were observed in BMD, although the magnitude of the increase in BMD was smaller than in BMC because BMA also increased slightly, but significantly, in PTH-treated animals. The increase in femoral BMA was consistent with a significant elongation of femoral length (~5% and ~3% in Group 5 males and females, respectively).
Radiology
Bone sclerosis, defined by an increased diffuse radiopacity, was commonly observed in all PTH groups (Figure 2) with an incidence of approximately 100% in the ≥50μg/kg/day dose groups (Table 4). In bones of the appendicular skeleton, the sclerosis was generally bilateral, affecting most commonly the femurs and tibiae. In the femurs, the typical presentation was a thickening of the cortices with a concurrent reduction of the medullary space. Both the spine and skull were often affected in the axial skeleton. Bone sclerosis was occasionally observed in the control groups. However, the skeletal distribution of sclerosis in control animals was generally not as widespread as in PTH-treated rats.
Localized and generally mono-ostotic radiological changes, either characterized by bone production, bone loss or a mixed reaction (both bone production and loss), were common observations in PTH-treated groups (Table 4). Overall, radiological changes were more frequently seen in the 50 or 150 μg/kg/day groups than in the 10 μg/kg/day and control groups. These changes were used to identify additional bones for histopathological evaluation.
Radiological assessments made possible the identification of additional small bone tumors, approximately 10% overall and approximately 40% in the 150 μg/kg/day female DS group, at sites not routinely sampled at necropsy and that would otherwise have gone undetected. These tumors were referred to as “occult” neoplasms (Table 4). Radiographs also proved a useful complement to support the histological diagnosis of some neoplasms.
Neoplastic Histological Observations
Several types of bone neoplasms were observed in both males and females treated with PTH at ≥50 μg/kg/day. Most neoplasms were regarded as malignant and classified as osteosarcoma. Other skeletal neoplasms included malignant fibrosarcoma, benign osteoblastoma, and osteoma. The incidence of all bone neoplasms was significantly increased in the whole skeleton of both males and females treated at ≥50 μg/kg/day (Groups 4 and 5). In groups treated with vehicle or PTH at 10 μg/kg/day, primary bone tumors were rare and limited to the occurrence of osteosarcomas in 2 female control animals and 1 male in the 10 μg/kg/day group (Table 5). No statistically significant differences in the incidence of bone neoplasms were found between males and females treated with PTH at 10 μg/kg/day and their respective control groups, consistent with the 10 μg/kg/day dose having no tumorigenic effect on bones.
Osteosarcoma
At necropsy, a pale and firm mass was the most common macroscopic presentation of osteosarcomas in PTH-treated rats. In some animals, these tumors had markedly infiltrated adjacent soft tissues or resulted in pathological fractures. These observations accounted for the increased incidence of hind limb disuse in males and females given 150 μg/kg/day, including the DS group. However, most osteosarcomas had not breached the periosteal envelope and required radiological and/or histological examination for diagnosis.
Histologically, osteosarcomas were highly variable with the osteoplastic subtype predominating (Figure 3). This subtype was characterized by abundant production of tumor bone and, typically, resulted in expansive bone growth beyond the periosteal margins where anaplastic osteoblasts were most evident along the tumoral outer rim. In decreasing order of incidence, the fibroblastic, telangiectatic and compound subtypes were also seen in PTH-treated rats. The production of tumor bone or matrix by anaplastic mesenchymal cells was used to differentiate between the telangiectatic osteosarcoma subtype and a hemangiosarcoma, and between the fibroblastic osteosarcoma subtype and a fibrosarcoma. The compound subtype, combining a bone and cartilaginous tumoral matrix, was observed infrequently. The marked variation in amount of tumor bone between subtypes explains the marked variability of radiological density noted with these neoplasms.
Multicentric osteosarcomas, originating on different bones, were found in 25% of PTH-treated rats diagnosed with this neoplasm; 4 primary sites were noted in 1 male treated with 150 μg/kg/day. In animals with primary osteosarcoma, metastases were noted in 45% of PTH-treated rats and in 1 control female (Table 5). Soft tissues presenting most frequently with metastases were the lungs, spleen, liver, and adrenal glands.
Osteosarcomas were found in a wide variety of bones in both the appendicular and axial skeleton. In the appendicular skeleton, the tibia and femur were most commonly affected, while the forelimbs were less frequently involved (Figure 4). The two osteosarcomas noted in control females were located in the femur and pelvis. In the axial skeleton, osteosarcomas were distributed all along the spine, ribcage and, occasionally, on the skull. Notably, the bones of the tail, feet, and face were unaffected.
There was a significant, dose-related trend in the occurrence of osteosarcomas. Osteosarcomas, combined for all sites, were significantly increased in males treated with PTH at 50 μg/kg/day and males and females treated with 150 μg/kg/day. At individual bone sites, a significant increase in osteosarcomas was observed in the lumbar vertebrae, tibia, and femur of males in the 150 μg/kg/day group and in the tibia of males treated with 50 μg/kg/day. High-dose PTH was also associated with a significant increase of osteosarcomas in lumbar vertebrae of female rats. Delaying the start of PTH treatment for 6 months resulted in a significantly lower bone tumor incidence in males, but not in females (Group 6 versus 5).
Fibrosarcoma
This rare, malignant mesenchymal neoplasm of the skeleton was only seen in 4 animals treated with PTH at ≥50 μg/kg/day (Table 5). The distal femur was the primary site affected with a single involvement of the lumbar vertebrae. These fibrosarcomas were small, confined within the bone, and were detected during radiological examination as sharply localized focal osteolysis. A dense proliferation of anaplastic, usually fusiform, mesenchymal cells that lacked the production of any tumor osteoid matrix and that replaced preexisting bone, were the criteria used for histological diagnosis (Figure 5). None of the statistical analyses conducted with this tumor reached significance.
Osteoblastoma
Osteoblastoma is a benign osseous neoplasm which was first reported in rats treated with teriparatide (Vahle et al., 2002). This tumor typically contains a moderate number of large and active osteoblasts along bone trabeculae. Diagnostic hallmarks used to identify an osteoblastoma included an intramedullary expansive growth pattern with minimal invasion at the margins, minimal cytological atypia, and scattered normal-looking mitotic figures (Figure 3). Only animals treated with PTH at ≥ 50 μg/kg/day were diagnosed with osteoblastoma (Table 5), which were mainly found in the tibia, femur, lumbar vertebra and sternum. When analyzed across all examined osseous sites, significant increases in incidence were noted for females treated at 50 μg/kg/day and males and females at 150 μg/kg/day (Group 5). In several animals, some tumors in the same section consisted of a portion typical of an osteoblastoma that was adjacent to cells with a malignant phenotype. These dual occurrences were diagnosed solely as osteosarcomas, suggesting neoplastic progression.
Osteoma
This very rare benign tumor in rats was only seen in animals treated with PTH at ≥50 μg/kg/day, where it was localized to the skull, lumbar vertebrae and ilium. When the osteomas were analyzed across all examined skeletal sites, a significant increase was found for males treated at 150 μg/kg/day (Group 5; Table 5).
Miscellaneous Neoplastic Findings
PTH treatment did not increase the incidence of other neoplasms in any of the soft or skeletal tissues examined.
Nonneoplastic Histologic Observations
Osteosclerosis
Osteosclerosis was commonly seen in males and females given 10 μg/kg/day PTH and it affected almost all routinely examined bones from animals given ≥50 μg/kg/day (Table 6) confirming the widespread distribution seen with radiography. This change was primarily characterized by trabecular hypertrophy, which increased the volume of bone occupying the medullary cavity (Figure 6). The distribution was usually diffuse within the endosteal compartment and involved the epiphyseal, metaphyseal and diaphyseal regions in long bones. Often, bone accumulation in the medullary cavity was sufficiently severe to almost entirely obliterate the space occupied by the bone marrow. Osteosclerosis was regarded as an exaggerated pharmacological response to treatment with PTH.
Osteosclerosis was also observed in control rats, especially in females (Table 6). The low incidence of osteosclerosis was consistent with spontaneous hyperostosis or osteopetrosis; a condition reported predominantly in aging female Fischer 344 rats (Leininger and Riley, 1990; Thurman and Bucci, 1994). This observation also correlated with the radiological bone sclerosis noted in a small number of control rats. Nonetheless, the incidence of osteosclerosis was significantly lower in both male and female control groups compared with the 10 μg/kg/day dose group at each of the routinely sampled bone sites.
Fibrous Osteodystrophy:
Fibrous osteodystrophy, predominantly noted in the lumbar vertebrae and sternum, was a finding seen predominantly in males and females treated at ≥50 μg/kg/day (Table 6). These microscopic, focal or multifocal lesions in the spongiosa were generally admixed with osteosclerosis and consisted of increased numbers of bone cells lining a woven bone matrix and of stromal cells and fibrous connective tissue in the adjacent marrow cavity (Figure 7). Such lesions were differentiated from those resulting from renal osteodystrophy, a complication of the advanced chronic progressive nephropathy seen in many control and PTH-treated males. Prominent osteoclastic resorption, a feature of renal osteodystrophy, was lacking in animals diagnosed with fibrous osteodystrophy. Overall, the sporadic occurrence of fibrous osteodystrophy in males and females was comparable between the control and 10 μg/kg/day PTH groups.
Focal Osteoblast Hyperplasia:
Focal osteoblast hyperplasia is characterized by microscopic focal proliferation of well-differentiated osteoblasts filling enlarged medullary spaces without excessive bone lysis or formation (Figure 7). PTH treatment was associated with an increased incidence of focal osteoblast hyperplasia at all sites in males and females treated with 150 μg/kg/day and at the lumbar vertebrae in females treated with 50 μg/kg/day (Table 6). In male rats, the most commonly affected bone was the tibia, while the femur and lumbar vertebrae were most commonly affected in females. Focal osteoblast hyperplasia was a rare observation in animals dosed with 10 μg/kg/day and, when all bones were evaluated together, its incidence in this dose group was comparable to or slightly less than in control animals.
Osteofibrous Dysplasia
Osteofibrous dysplasia, localized primarily in the femur, was another microscopic change seen in several female rats dosed at ≥ 50 μg/kg/day, but not in the control or 10 μg/kg/day groups or in any males (Table 6). The histology of these lesions resembled its human counterpart (Schwamm and Millward, 1995) and was characterized by replacement of marrow and trabecular bone by well-differentiated spindled cellular elements in a collagenous matrix admixed with predominantly small woven bone trabeculae lined with an osteoblastic rim. In contrast to fibrous osteodystrophy, these lesions presented with substantial bone resorption, which was often seen on radiographs as focal bone loss. The fibroblastic and osteoblastic cells of fibrous osteodystrophy and osteofibrous dysplasia lacked anaplastic features, which were used to differentiate them from bone malignancies.
Miscellaneous Nonneoplastic Histological Observations
Extraskeletal effects associated with PTH treatment were noted in a few soft tissues. The incidence of increased extramedullary hematopoiesis was higher in females treated at ≥50 μg/kg/day. This dose-related change was considered secondary to the marked osteosclerosis that displaced hematopoietic production from the obliterated bone marrow to the spleen. Dilatation and/or inflammation of the urinary bladder was more common in both male groups treated with 150 μg/kg/day. These findings were regarded as secondary, at least in some treated rats, to a malignant bone tumor localized in the vertebrae or pelvis.
Discussion
This study characterized the effects of chronic, near lifetime exposure to recombinant human PTH on the rat skeleton in a cancer bioassay. A PTH dose of 10 μg/kg/day for 2 years significantly increased bone mass, but did not increase the incidence of osteosarcoma or induce other neoplastic and non-neoplastic proliferative changes. This dose provided a systemic exposure to PTH that was 4.6-fold greater than the systemic exposure in humans following a 100 μg dose. A dose-related increase in osteosarcoma was observed with the 50 and 150 μg/kg/day doses of PTH.
Fibrosarcoma, another malignant skeletal tumor, was observed in four of 360 rats exposed to ≥50 μg/kg/day. While the teriparatide oncogenicity studies did not report such tumors (Vahle et al., 2002, 2004), the introduction of radiography in this study allowed detection and sampling of these small lesions. This tumor was assumed to be associated with PTH treatment.
The incidence, skeletal distribution and histological appearance of osteoblastoma and osteoma resulting from mid-and high-dose PTH administration were similar to teriparatide (Vahle et al., 2002, 2004). However, these benign bone tumors were not observed in the low-dose PTH group.
Focal osteoblast hyperplasia has been described as part of the morphologic continuum of hyperplasia leading to neoplasia associated with near-lifetime administration of teriparatide to the rat (Vahle et al., 2002). It was observed in this study with comparable incidence in control and low-dose PTH animals and with higher incidence in animals dosed with PTH at the mid-and high-dose.
Fibrous osteodystrophy and osteofibrous dysplasia were observed in this study, but were not previously observed following chronic teriparatide exposure (Vahle et al., 2004). However, focal stromal and stromal-vascular proliferations reported following long-term exposure to teriparatide (Vahle et al., 2004) were not observed in this study. The incidence and histological appearance of focal stromal and stromal-vascular proliferations following teriparatide administration and of fibrous osteodystrophy and osteofibrous dysplasia in this study appeared to be similar, suggesting that they represent the same entities. Our selection of a different diagnostic terminology reflects our opinion that these lesions are more than just stromal proliferation because they include some bone loss and abnormal woven matrix deposition.
The DS group treated with 150 μg/kg/day PTH from 8 months to 2 years of age, was included to assess the skeletal response in absence of PTH treatment during the period of active bone growth. In both genders the delayed start of high-dose PTH was also associated with a high incidence of osteosarcomas suggesting that rapid longitudinal bone growth is not necessary for the induction of neoplasms at this high dose.
While this study did not directly compare PTH with teriparatide, it was designed to allow comparisons between the effects of both peptides. As in the teriparatide study (Vahle et al., 2002), this bioassay used the same inbred strain of Fischer 344 rats that were fed ad libitum, kept in a similarly controlled environment and exposed for the same time period and by the same route of administration at similar molar dose levels. Furthermore, the diagnostic criteria used for the classification of most proliferative lesions associated with teriparatide were applied to the PTH study. However, unlike the teriparatide study, radiographs were used to detect small tumors at skeletal sites not routinely sampled at necropsy and that otherwise would have gone undetected. This increased the number of skeletal neoplasms identified by 10% and supports previous recommendations that radiography be used to detect bone tumors in oncogenicity studies (Stanton, 1979; Zwicker and Eyster, 1996).
In this animal model, the responses of the skeleton to PTH and teriparatide were similar in many ways (Vahle et al., 2004). Like teriparatide, PTH induced a significant dose-dependent increase in bone mass with all doses at all skeletal sites evaluated. This pharmacological effect resulted in a generalized osteosclerosis observed by radiology and histology. At the higher doses, these changes were often so marked that the marrow was almost completely obliterated. Also like teriparatide, the administration of PTH for 2 years resulted in dose-related proliferative changes in bone where osteosarcoma predominated. However, in contrast to teriparatide, the lowest dose of PTH was not carcinogenic with the incidence of osteosarcoma equal to that observed in control animals. The histological features of the osteosarcomas in the control and low-dose PTH groups were similar.
Collectively, the data support the concept that induction of skeletal neoplasia associated with PTH is dose-dependent, but that the threshold for bone tumor development occurs at a higher dose than that required to obtain the desired pharmacological effect. Our data also suggest that at exposures similar to those observed clinically, the oncogenic potential of PTH is lower than teriparatide in comparable 2-year rat carcinogenicity studies. When compared to 24 months of treatment with teriparatide (Vahle et al., 2002), treatment with PTH resulted in a notable right-shift in the dose response curve for osteosarcomas (Figure 8). This affords a safety margin for PTH of at least 4.6-fold between exposures following the clinically recommended dose for humans (100 μg/day) and the demonstrated noncarcinogenic dose administered to rats (10 μg/kg/day). In contrast, the lowest dose of teriparatide evaluated in rats (5 μg/kg/day) resulted in an exposure only 3 times that observed with the 20 μg/day clinical dose and was associated with an increased incidence of osteosarcoma (Eli Lilly and Company, 2004).
The reason long-term exposure to PTH is associated with a lower incidence of osteosarcoma than teriparatide is unknown, but it is possible that the C-terminal region of PTH, which is absent in teriparatide, may be responsible. C-terminal fragments of PTH are secreted by the parathyroid glands and produced following peripheral PTH metabolism, primarily in the liver (Nguyen-Yamamoto et al., 2002). It is becoming increasingly recognized that a protein that binds PTH and C-terminal PTH fragments, but not teriparatide, is present in bone cells (Murray et al., 2005).
Activation of this putative C-terminal PTH receptor elicits biological responses that are often in opposition to those that occur following PTH-1 receptor activation. For example, C-terminal PTH fragments blunt the increase in serum calcium levels induced by teriparatide and PTH infusions in thyroparathyroidectomized rats (Slatopolsky et al., 2000; Nguyen-Yamamoto et al., 2001; Langub et al., 2003). Additionally, teriparatide stimulates bone resorption while C-terminal PTH fragments inhibit bone resorption. Finally, teriparatide inhibits apoptosis in bone cells while C-terminal PTH fragments stimulate apoptosis in bone cells (Bringhurst, 2002).
It is this latter phenomenon that provides a possible explanation for the difference in oncogenicity between the 2 peptides; the N-terminal and C-terminal regions of PTH may act in concert to control and maintain a normal rate of osteoblast turnover. PTH-1 receptor activation in the absence of such feedback regulation could lead to osteoblast hyperplasia and neoplastic transformation. A similar regulatory mechanism has been described in the prolactin/growth hormone endocrine system where the full-length hormones stimulate proliferation of capillary endothelial cells in vitro and in vivo, but N-terminal peptides derived from these hormones are inhibitory and act through a distinct receptor (Struman et al., 1999).
In conclusion, recombinant human PTH represents a potentially new treatment for individuals with osteoporosis who are at a risk of fracture. In the rat carcinogenicity study reported here, treatment with PTH resulted in significant new bone growth at all the evaluated skeletal sites, with no increase in the incidence of proliferative, benign or malignant neoplastic bone changes in the 10 μg/kg/day dose group. Although the application of these findings to humans with osteoporosis needs to be carefully considered, the noncarcinogenic dose in rats occurred at systemic exposure 4.6-times that in humans at the proposed clinical dose. Recombinant human PTH is currently the only anabolic agent being investigated for the treatment of osteoporosis with a demonstrated noncarcinogenic dose in a 2-year rat bioassay.