Matter 2 - Biological Plausibility

Matter 2 - Biological plausibility of difference between Teriparatide and comparator

Background to the matter arising

The sponsor raised the following in its requesting letter for review “PBAC claim that the submission provides no evidence to suggest any biologically plausible reasons to explain the claimed difference between Teriparatide and the comparator – paragraph 9 of the PBAC minutes.”

The relevant excerpt from the March 2006 PBAC short minutes is contained within paragraph 5, (not paragraph 9), and reads “Furthermore, no evidence was submitted to suggest any biologically plausible reason that teriparatide is more effective in the high risk sub-group than in the overall trial population. Although a biologically plausible argument could be mounted that alendronate may be less effective in the high risk sub-group because of a lower bone turnover associated with loss of bone architecture in this sub-group, as noted above, the submission did not rely on using the results of a sub-group analysis in place of the overall trial population results for alendronate in its modelled economic evaluation. As noted previously, the submission again sought to base its biological plausibility arguments on the difference mechanisms of action of the two alternative drugs. The PBAC concluded that although this might be relevant to the question of differential treatment effects across the drugs, it is irrelevant to the question here of the biological plausibility of the one drug (teriparatide) having an increasing relative treatment effect as baseline fracture severity worsens. The PBAC thus considered that the re-submission therefore provides no new basis that could change the PBAC’s previous view concerning the invalidity of adopting the results of this sub-group analysis rather than the results of the overall ITT analysis as the basis for deriving an estimate of the effectiveness of teriparatide to compare with alendronate.”

The relevant extract from the sponsor response to the ESC and RWG advice is: “We are somewhat frustrated by the lack of acknowledgement given in the ESC advice relating to this issue. The submission includes reference to a number of clinical, pre-clinical, histomorphological, biological studies and modelling studies presented to support the plausibility that an anabolic agent which stimulates osteoblasts and thus builds new bone would produce superior outcomes to an anticatabolic agent, and that this effect may be moderated by the existing microarchitecture of the bone. We provided independent references relating to the action of anabolic and anti-catabolic drugs on bone, references relating to the structure of bone in mild, moderate and severe patients and evidence demonstrating the differential effects of teriparatide on depleted bone in humans and animals. We have sought independent advice on this issue prior to the submission and believe that this has been addressed “beyond re-stating” that the drugs have a different mechanism of action. We therefore urge the PBAC to discuss this with members of the scientific and medical community. Without further explanation from the evaluator as to why these multiple studies are so easily rejected we cannot further address this issue.”


Top of page

Reviewer’s understanding of the key issues

The key issues involved in this matter appear to relate to the biological plausibility of an increased treatment effect with teriparatide compared with antiresorptive therapy (in particular, aledronate) on the basis of mechanism of action, increasing treatment effect over the spectrum of osteoporosis severity, and the presence of clinical data to support such a claim.

Materials considered

All materials considered by the PBAC relating to the assessment of biological plausibility were reviewed. In addition, key reference studies listed on page 49 were used in forming an opinion.

Reviewer’s opinion

An issue repeatedly identified in the July 2005 and March 2006 submissions was the biological plausibility of an increasing treatment effect with teriparatide over comparator in patients with more severe forms of vertebral osteoporosis. The sponsor claims that vertebral fracture severity is a treatment effect modifier and that an agent such as teriparatide which stimulates osteoblasts as opposed to antiresorptive agents which maintain existing bone is likely to have an enhanced treatment effect in patients with more severe forms of osteoporosis.

Pharmacological therapies for osteoporosis may be classified as either anti-catabolic (antiresorptive) or anabolic depending on their mode of action. Most currently available drug treatments for osteoporosis are anti-catabolic in action and include hormone therapy with oestrogen receptor modulating effects (e.g. raloxifene) and the bisphosphonates (e.g. aledronate and risedronate). Anti-catabolic drugs reverse the negative bone balance seen in osteoporosis by decreasing the associated increase in bone remodelling. They do this by decreasing the number of bone multicellular units (BMU), and hence preferentially decrease bone resorption. The reduction in bone remodelling by anti-catabolic drugs may increase bone mass moderately during the interval in which previously activated BMUs are still completing mineralization. As such, in the period of therapy before bone formation decreases, bone balance becomes positive, but thereafter remains in equilibrium explaining why bone mineral density increases are typically largest during the first two years of therapy with anti-resorptive agents.

The skeletal effects of parathyroid hormone (PTH) depend upon the pattern of systemic exposure. When used in an intermittent or cyclical manner, PTH can be classified as an anabolic therapy for osteoporosis. In contrast, a continual excess of endogenous PTH, as seen in hyperparathyroidism, may be detrimental to the skeleton as bone resorption may be stimulated more than bone formation. Teriparatide is the active fragment (1-34) of endogenous human PTH manufactured using recombinant DNA technology. It is the most extensively studied form of PTH in patients with osteoporosis. The intermittent administration of PTH acts in two principal ways to induce a positive bone balance that persists in the medium term. Firstly, activation of the BMU results in a preferential increase in bone formation over bone resorption (i.e. augmentation of osteoblast function). Secondly, there is also increased conversion of bone lining cells to osteoblasts. This response has been demonstrated in cell culture systems, in intact animals, and in humans. It is this property of the proposed administration of teriparatide that increases bone formation and results in a positive bone balance. In addition, teriparatide induces renewed modelling which is an efficient way to increase bone mass, because the formation of the BMU is not preceded by a resorption phase. Other effects of teriparatide include improvements in trabecular micro-architecture, and cortical bone geometry and thickness. In contrast, anti-catabolic agents do not increase bone tissue mass or trabecular bone volume. Instead, these agents increase the tissue mineral content of density present in the existing bone tissue mass.

Top of page

The different mechanisms of action between the two drug classes raise the theoretical possibility of an enhanced treatment effect in favour of anabolic agents in patients with more severe forms of osteoporosis, particularly involving skeletal sites comprised of predominately high-turnover trabecular bone. In healthy individuals, bone is constantly remodelled through the process of active formation and resorption which occurs in a close temporal sequence. The hallmark of osteoporosis is a reduction in skeletal mass with an associated micro-architectural disturbance caused by an imbalance between bone resorption and bone formation. Loss of gonadal function and ageing are the two most important elements contributing to this imbalance. Bone is made up of two types: cortical and trabecular. Cortical bone comprises 80% of all bone and is found mainly in long bones. Bone loss that occurs primarily in cortical bone results in fractures of the proximal femur and humerus. Trabecular bone is predominately found in flat bones (vertebrae, sternum, and pelvis) and the metaphyses of long bones. Thus, when trabecular bone is most affected by bone loss, vertebral, pelvic and Colles’ fractures of the forearm are manifestations of this pattern of osteoporosis. In healthy adults, 25% of trabecular bone is resorbed and replaced every year, compared with 3% of cortical bone, which indicates the rate of remodeling and hence susceptibility to the pathophysiological changes of osteoporosis is different between the different types of bone. Vertebrae are principally composed of trabecular bone encased in a thin shell of cortical bone. Some underlying medical conditions or risk factors predispose certain individuals to one particular pattern of osteoporosis. For example, the use of corticosteroids may be associated with rapid bone loss, especially in high-turnover trabecular bone, and may lead to fractures in predominately trabecular bone sites such as the vertebrae and ribs. However, for most individuals with osteoporosis there is no differential loss of bone between the two types of bone, i.e. osteoporosis is a systemic skeletal disease and the process affects all bony sites.

The pathophysiology of osteoporosis does appear to alter with increasing severity of osteoporosis with bone density (quantity) and bone architecture (quality) becoming disproportionately disturbed. Whereas it has been shown that bone mass is the major determinant of bone strength, the mass-based concept does not fully account for the contribution of bone microarchitecture to mechanical efficiency and ability to withstand fracture at normal physiological stress levels. The pivotal paper by Parkinson and Fazzalari on this topic was included by the sponsor in the fourth submission. This study showed that changes to cancellous bone structure are bone volume-dependent in a nonlinear manner. Specifically, at low bone volume (<15%), structural parameters of cancellous bone such as trabecular thickness and trabecular separation change at a much greater rate than at higher bone volume. This indicates that the structural integrity of cancellous bone may become rapidly compromised when bone volume falls below a critical value.

Anti-catabolic therapies such as bisphosphonates may be less effective in the high risk subgroup of patients with severe bone loss, particularly in predominately higher-turnover trabecular bone sites such as vertebral body osteoporosis. This hypothesis appears to be supported by in vivo histomorphometric data favoring the differential effect of teriparatide compared to aledronate in postmenopausal women. This theory (supported by scientific data) is relevant in considering a differential treatment effect across the two drug classes used in osteoporosis. Anti-catabolic drugs work by inhibiting bone resorption and reducing the chance that trabecular plates will perforate. In patients with more severe grades of osteoporosis, bone structure is lost, thus abnormal stress is placed on the remaining structures which increases the possibility of damage. In response, bone turnover is stimulated, but osteoblastic bone formation increases at an insufficient rate, so overall bone mass decreases and trabecular architecture diminishes. Anti-catabolic drugs can slow this cycle and try to maintain bone strength. However, in patients with severe vertebral osteoporosis with fractures, the bone structure is damaged to the threshold where maintenance of the existing bone is insufficient and bone formation is required.

Because of the mechanism of action of teriparatide, it is conceivable that this drug may have an increasing relative treatment effect as baseline severity risk of fracture worsens. There are several studies (supplied by the sponsor) that show that teriparatide induces beneficial changes in bone quality, structural architecture and geometry. These attributes are particularly advantageous in more severe forms of vertebral osteoporosis. As such, it is theoretically conceivable that teriparatide has an increasing relative treatment effect as baseline severity risk of fracture worsens because of the beneficial inter-relationship between the anabolic mechanism of action of the drug and the pathophysiology of severe vertebral osteoporosis. Consequently, on the basis of biological plausibility, I would disagree with the PBAC statement that the different mechanisms of action of the two alternative drug classes is irrelevant to whether or not a single drug (i.e. teriparatide) has an increasing relative treatment effect as baseline severity risk of fracture worsens.

Top of page

Despite the theoretical plausibility, the submissions did not contain convincing clinical outcome data to support such a hypothesis. Data from the pivotal GHAC study did not conclusively demonstrate an enhanced treatment effect in the severe fracture (SQ3) subgroup compared to the overall Intention-to-Treat (ITT) analysis. Using the overall ITT data (Table 1- page 15), new vertebral fractures occurred in 64 of 544 (14.3%) of women in the placebo group compared to 22 of 541 (5.0%) of women treated with teriparatide 20ug/day (relative fracture risk 0.35, 95% CI 0.22 to 0.55; p<0.001). Using the SQ3 subgroup data from the GHAC study, new vertebral fractures occurred in 27 of 95 (28.4%) of women in the placebo group compared to 5 of 86 (5.8%) of women treated with teriparatide 20ug/day (relative fracture risk 0.205, 95% CI 0.08 to 0.51). Because the 95% confidence intervals for each type of population analysis lie within each other, the treatment effect by population should be interpreted with caution. Moreover, the PBAC has conducted a test for interaction of treatment effect modification across the two baseline fracture subgroups: - the nominated (“severe”) subgroup and its comparator (“non-severe”, i.e. nil + mild + moderate fracture subgroups at baseline) and shown a non-significant result (p=0.1433) for relative risk reduction. This information is contained within the evaluator’s commentary on the July 2005 re-submission (refer to section 7.6-page 13). These results are also shown in this review in Tables 3 and 4 (pages 16-17). Finally, the interpretation of these results should be viewed cautiously. According to Altman and Bland, results of tests for interaction are likely to be convincing only if they were specified at the start of the study. As already discussed in matter 1B, such analysis occurred post-hoc.

Reviewer’s summary

There is sufficient evidence to support the sponsor claim of biological plausibility that teriparatide may have an improved treatment effect over comparator therapy (antiresorptive drugs) in a population of patients with severe vertebral osteoporosis. This is based on the drug’s anabolic mechanism of action and how such an action may be disproportionately effective in severe osteoporosis where there is significant alteration in the pathophysiology of osteoporosis favouring such a therapeutic action.

However, there is insufficient clinical outcome data in the submissions to conclusively support the sponsor claim that teriparatide is more effective in the high risk vertebral fracture subgroup than in the overall treatment population. In addition, It remains unclear whether or not teriparatide has a superior treatment effect compared to alternative anti-osteoporosis agents for patients with more severe forms of osteoporosis as direct comparative (head-to-head) clinical outcome data in this sub-population does not currently exist.