Questions and facts regarding denosumab discontinuation among postmenopausal women

KEYWORDS:

• Alendronate
• denosumab
• discontinuation
• osteoporosis
• zoledronate

1. Introduction

In contrast to other chronic diseases where treatment is lifelong, in osteoporosis treatment discontinuation may be considered under certain conditions. Bisphosphonates (BPs) remain the mainstay of osteoporosis treatment for the last three decades and both their long biological half-life in bone and their retained anti-fracture effect, even after their discontinuation, have inevitably influenced our perception regarding the therapeutic approach of this disease [1]. The only easily accessible and valuable surrogate marker to guide our decisions regarding discontinuing treatment or not is bone mineral density (BMD), due to its strong correlation with fracture prediction [2]. However, although BMD can be efficiently used as a target while on treatment [1], it can be revisited not sooner than 1 year from discontinuation, which is probably a relatively long time for a chronic condition to remain unmonitored. For example, no one would consider repeating HbA1c or lipid measurements 1 year after discontinuation of diabetes or hyperlipidemia treatment, respectively, while probably the majority of physicians would not have advised cessation of treatment even if the patient was perfectly controlled under therapy. However, given the lack of reliable predictive information from bone markers following osteoporosis treatment discontinuation [3], BMD measurement is the only tool to monitor the course of the disease post-treatment.

2. Discussion

Denosumab (Dmab) discontinuation is currently a ‘hot’ topic in the literature. The reversal of its favorable skeletal effects after its cessation was known from its launch in the therapeutic field [4], but it is only until recently that the increase in bone turnover above pre-treatment values, described as ‘rebound phenomenon,’ was associated in a few patients with multiple vertebral fractures [5]. The mechanism underlying the rebound phenomenon is largely unknown although upregulation of osteoclastogenesis has been implicated [6]. Given the above, there exist some arguments stating that the onset of Dmab might involves the potential for lifelong treatment and thus it should be at least reserved as an optimal regimen for older individuals. Based on the pharmacodynamics of both Dmab and BPs, the latter are currently recommended to succeed any Dmab therapy in an effort to prevent the expected bone loss and hopefully the rebound-associated fractures [7]. Although this recommendation sounds quite simple, it appears that there is still a lot to learn in order to effectively tackle any challenge that might arise from Dmab discontinuation at the patient level.

Probably the major question regarding Dmab treatment concerns the timing of its discontinuation. Data from a recent analysis suggest that the incidence of non-vertebral fractures under Dmab treatment gets lower as total hip T-score increases and reaches a plateau at a range between −2.0 and −1.5 independently of age and prevalent fractures, while this is not the case for vertebral fractures which might exhibit the same inverse relationship with BMD however without an upper T-score threshold [8]. Therefore, it appears that BMD monitoring might determine a time point of Dmab treatment when the risk, at least for non-vertebral fractures, has been minimized and thus discontinuation could be considered as it seems to provide no further antifracture benefits. This notion inevitably leads to the next question which concerns the ideal therapeutic management following Dmab discontinuation when the patient has reached an osteopenic (namely with a Tscore > −2.0) total hip range. As already discussed, it is currently recommended to switch to a BP in order to consolidate the Dmab treatment gains. However, a number of queries raise again from the rationale and the implementation of such an approach in everyday clinical practice both among physicians and patients: Does Dmab discontinuation inevitably lead to a substantial bone loss in every patient? Does previous treatment with BPs protect against the rebound phenomenon? Is there a specifically increased fracture risk or it is just an expected wane of treatment’s protection? What is the real incidence of multiple vertebral fractures and is there a specific patient’s profile which is prone to this adverse event? Is there a specific BP regimen that could be used in order to minimize the losses from Dmab discontinuation? When and for how long should this BP administered? Does the duration of Dmab treatment play a significant role?

Regarding the identical response of all patients following an untreated cessation of Dmab, data from all studies undoubtly support a fast bone loss. However, at an individual basis, responses may significantly vary, as for example in our recent randomized controlled trial (RCT) where 13 out of the 30 patients, who remained untreated for 1 year after reaching osteopenia with Dmab, exhibited changes smaller than the least significant change in lumbar spine BMD [9].

Conventional thinking and the known retention of BPs in the skeleton support the notion of a possible protective effect against both bone loss and increased fracture risk from previous BP use as this could potentially blunt the rebound phenomenon. This, however, was not proved by a recent RCT, which included patients with an up to 3 years previous alendronate treatment or even longer treatment with any other BP [10], and a real-world observational study [11].

Both in a clinical trial and in a routine clinical setting Dmab discontinuation [5,11], or even delay by more than 4 months [12], could be associated by a 3 to 5-fold higher risk for vertebral, major osteoporotic, and hip fractures. However, in the placebo-controlled trials, the off-treatment fracture risk among patients who had received Dmab was not different than that of the placebo group [5,13], and one could argue that this is simply a relapse of the given fracture risk as the protective effect of treatment has been lost. It was only when the multiple vertebral fractures (MVFs) were censored as an adverse event that a significant increase was found amongst those discontinuing Dmab [5]. It seems that the percentage of patients experiencing this adverse event is low and ranges between 3.4% and 7.7% according to the small number of prospectively studied patients [5,9,14]. Although there are a lot yet to understand regarding the risk factors predisposing to this phenomenon, it is the prevalent vertebral fractures that are associated with an approximately 4-fold increase of the possibility of the rebound-associated MVFs [5]. Therefore, patients with established osteoporosis should be closely monitored in everyday clinical practice in order to ensure adherence to the subsequent antiresorptive treatment in case Dmab discontinuation is decided, while any other patient profile cannot be currently excluded from this recommended therapeutic approach.

Zoledronate (ZOL) and alendronate (ALN) are the two most studied BPs in this context. The difference in their way of administration, namely once i.v. versus weekly per os, probably raises another query regarding whether a single large BP dose or a frequent administration of small BP doses would more efficiently control the rebound phenomenon. Transition to ALN seems to maintain or even increase the BMD attained after a 1 year of Dmab treatment, while a BMD loss can be roughly expected in up to 1/5 of patients [15]. However, this finding should be cautiously interpreted as it corresponds to a quite short, and relatively unusual in common clinical practice, period of Dmab treatment. It seems that the duration of Dmab treatment might influence the efficacy of the subsequent BP treatment to preserve bone gains as suggested from the two currently published RCTs with ZOL. In specific, a single i.v. infusion of ZOL 5 mg, given 6 months after the last injection of denosumab therapy maintained for 3 years the BMD gains in the majority (around 80%) of patients previously treated with Dmab for an approximate period of 2.5 years [16]. On the contrary, treatment with ZOL, irrespective of the timing of administration, did not fully prevent bone loss among patients with an approximate 4.5 years of Dmab treatment [10]. Therefore, besides other yet unknown factors, Dmab treatment duration probably plays a role in the diversity of results among different studies. The timing of BP administration is also a matter of controversy as it was hypothesized that BP could be more efficacious if given with an already increased rate of bone turnover. However, this hypothesis was not verified [9]. Finally, a rational and probably optimal approach has been recently suggested involving the administration of ZOL 6 months following the last Dmab injection and then close monitoring of the BTMs 3 and 6 months after ZOL infusion rather than waiting for the annual BMD changes; with this approach in case of increased BTMs (i.e. above the mean found in age and gender-matched cohorts) an early repeated infusion of ZOL could be considered [17]. Nevertheless, this approach, which is up to now based on an opinion of experts, might confront difficulties with the current country-specific guidelines regarding the use of ZOL in postmenopausal osteoporosis.

3. Expert opinion

The value of Dmab as a regimen for the treatment of osteoporosis is undoubtful. However, its positive effects on the skeleton are rapidly reversed upon discontinuation. Furthermore, the risk for this unique entity of MVFs following its discontinuation is a fact, affecting somewhat less than 10% of the patients.

A potent antiresorptive is recommended to follow Dmab at all instances in order to restrain bone losses and advert the risk for MVFs. Among them, ZOL is the most studied up to date and current evidence suggests that it can largely prevent bone loss. Although not proven yet, we believe that the performance of BPs in maintaining bone mass following Dmab discontinuation is probably affected by the duration of Dmab treatment. Dmab administration for 3 or more years may render subsequent BPs less efficacious and some level of bone loss should be expected despite the BP administration.

We recommend ZOL as the BP of choice 6 months after the last Dmab injection, because of the larger amount of existing data supporting its use on this setting and the convenience of its administration. Alternatively, weekly ALN should be preferred. The optimal duration of BP treatment is currently unknown but should be at least 1 year. Close monitoring is recommended thereafter.

Declaration of interests

P Makras has received lecture fees and research grants from Amgen; lecture fees from Glaxo, Lilly, Pfizer, Leo, Genesis, ELPEN, VIANEX, Rafarm, UCB, Takeda. ADA has received lecture fees from Amgen, Eli-Lilly, ELPEN, ITF Hellas, VIANEX. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Additional information

Funding

This paper was not funded.