Denosumab discontinuation in the clinic: implications of rebound bone turnover and emerging strategies to prevent bone loss and fractures

Abstract

Denosumab is a highly effective treatment for osteoporosis, and its long-term use is associated with incremental gains in BMD and sustained fracture risk reduction. Stopping denosumab, however, results in a rebound increase in bone turnover, loss of treatment-associated BMD gains, and in the worst case, spontaneous vertebral fractures (VFs). Insights into the risk factors, the underlying mechanisms for rebound-associated bone loss, and true incidence of rebound VFs are emerging. Conventional strategies using bisphosphonates to mitigate post-denosumab rebound display mixed success. Bisphosphonates may preserve bone density following short-term denosumab but the optimal sequential approach after longer-term denosumab remains elusive. Patients at particular risk of are those with prevalent VFs or greater on-treatment BMD gains. To greater understand these risks and strategies to preserve bone after denosumab, an emerging body of translational and preclinical work is shedding new light on the biology of RANKL inhibition and withdrawal. Discovering an effective "exit strategy" to control rebound bone turnover and avoid bone loss after denosumab will improve confidence among patients and clinicians in this highly effective and otherwise safe treatment for osteoporosis. This perspective characterizes the clinical problem of post-denosumab rebound, provides a comprehensive update on human studies examining the use of bisphosphonates following denosumab and explores mechanistic insights from preclinical studies that will be critical in the design of definitive human trials.

Keywords: BMD; bisphosphonates; bone resorption; bone turnover; denosumab; discontinuation; fracture; osteoporosis; rebound; sequential therapy.

Figure 1

Changes in bone mineral density (BMD), c-terminal telopeptide of type 1 collagen (CTx) and vertebral fracture (VF) risk at baseline, during and post-denosumab cessation. CTx falls rapidly with initiation of denosumab treatment and remains suppressed for the duration of treatment. Upon denosumab cessation (defined as 6-mo after the last denosumab dose) there is a rapid rebound and overshoot in CTx before slowly returning to baseline levels. BMD rises gradually during denosumab treatment. After cessation and along with the rebound in CTx, there is a rapid decline in BMD back towards baseline levels, earlier than the return of CTx to baseline. Fracture risk, including VF risk, falls with ongoing denosumab treatment. In the setting of exaggerated bone turnover and declining BMD, there is a rapid increase in VF risk, particularly multiple VFs, with a return towards baseline risk. Graphical concepts adapted from Bone et al. and Miller et al.

Figure 2

Changes in bone homeostasis during denosumab treatment and discontinuation and targets for sequential therapy. (A) Bone homeostasis is maintained by interaction between bone formation by osteoblasts and bone resorption by osteoclasts which is regulated by RANKL/RANK/osteoprotegerin (OPG) signaling. Osteoclasts may undergo “recycling” where, under RANKL stimulation, osteoclasts undergo fission to form osteomorphs which can circulate and fuse to re-form osteoclasts. (B) Denosumab inhibits RANKL/RANK binding, preventing osteoclast differentiation, and promoting accumulation of osteoclast precursors. Osteoclasts may undergo fission to form osteomorphs but fusion is also inhibited by denosumab, leading to an accumulation of osteomorphs. (C) upon denosumab discontinuation and offset of RANKL inhibition, increased RANKL signaling promote differentiation and activation of accumulated osteoclast precursors. Accumulated osteomorphs may also fuse to form active osteoclasts. Bone turnover resumes and osteoblastic bone formation and OPG secretion also restarts but delayed relative to osteoclastogenesis and bone resorption, leading to net bone loss. (D) Targets for sequential therapy following denosumab discontinuation: Bisphosphonates primarily target and inhibit active osteoclasts resorbing bone without impacting osteoclast differentiation and activation. Romosozumab inhibits sclerostin signaling, promoting osteoblast formation and resultant OPG production and bone gains, as well as inhibiting osteoclast formation. Rescue therapy with denosumab reverses the pathophysiology of the rebound phenomenon, targeting RANKL/RANK binding, inhibiting osteoclast formation as well as potentially inhibiting re-fusion of accumulated osteomorphs. Abbreviations: OPG, osteoprotegerin; RANKL, receptor activator of nuclear factor kappa beta ligand; RANK, receptor activator of nuclear factor kappa beta; TRAP, tartrate resistant acid phosphatase. Created with BioRender.com.

Figure 3

Mechanistic representation of effects of sequential bisphosphonates on targeting rebound bone loss based on concepts extrapolated from preclinical studies. (A) Standard “on time” intervention and repeat dosing. (i) As the effect of denosumab wanes before the next dose, accumulated osteoclast precursors and osteomorphs form osteoclasts and bone resorption resumes. Serum RANKL and TRAP may become elevated above on-treatment suppressed levels. CTX, however, would likely remain suppressed at this time. (ii) 6-mo following the last denosumab dose, bone loss is occurring. Rising TRAP and CTX levels may be seen at this time due to increased bone resorption. “On time” sequential bisphosphonate (sBP) at this time targets bone resorbing osteoclasts. Standard timing of BPs may mitigate some bone loss, however osteoclasts are more abundant at this stage compared to an earlier intervention window. (iii) rebound bone loss continues as accumulated osteoclast precursors and osteomorphs not targeted by the initial bisphosphonate doses are purported to continue to form osteoclasts. Repeat doses of bisphosphonates (rBP) may be required at this time to target these newly formed osteoclasts. Serum TRAP and CTX levels may remain elevated as increased osteoclastic bone resorption continues, or return to baseline levels if resorption is suppressed with repeated BP dosing. (B) “early” intervention and repeat dosing. (i) Early bisphosphonate administration (eBP) within 6 mo post last denosumab dose may inhibit bone resorption by the initial wave of osteoclasts differentiating from the accumulated pool of precursors. (ii) Osteoclasts continue to form and repeat bisphosphonate doses (rBP) may be needed to target these newly formed osteoclasts. Abbreviations: BP, bisphosphonates; CTX, cross-linked C-telopeptide of type I collagen; Dmab, denosumab; OC, osteoclasts; RANK, receptor activator of nuclear factor kappa beta; RANKL, receptor activator of nuclear factor kappa beta ligand; TRAP, tartrate resistant acid phosphatase. Created with BioRender.com.

Figure 4

Clinical considerations for denosumab initiation, discontinuation and re-initiation. ^*Prior to initiating denosumab therapy, consider fracture risk and appropriateness for long-term antiresorptive therapy. Stopping denosumab and transitioning to other osteoporosis medications may be more appropriate in cases where long-term antiresorptive therapy is unlikely to be required. Abbreviations: ALN, alendronate; BTM, bone turnover markers; BMD, bone mineral density; BP, bisphosphonates; Dmab, denosumab; T/L spine XR, thoracolumbar spine X-ray; VF, vertebral fracture; ZOL, zoledronate. Created with BioRender.com.