Effect of combination versus sequential teriparatide initiation on vertebral bone marrow microenvironment in postmenopausal women with prior denosumab treatment: A propensity score-matched study.

Hao, Jie; Yan, Fayao; Yao, Yu; Zhang, Tongrui; Zhang, Yongqi; Li, Zheng; Xu, Hao; Jiang, Xingjie · Bone · 2026

retrospective_cohort · Level III

Where this comes from

Abstract

The optimal strategy for transitioning from long-term denosumab (DMAb) to teriparatide (TPTD) in postmenopausal osteoporosis remains a clinical dilemma, as DMAb withdrawal can trigger a deleterious rebound in bone resorption. While bone mineral density (BMD) is the traditional benchmark, it fails to capture critical aspects of bone quality, such as the bone marrow microenvironment. This study investigated whether the method of TPTD initiation-adding it to ongoing DMAb (combination) versus switching after DMAb withdrawal (sequential)-differentially affects the vertebral bone marrow microenvironment, assessed by the MRI-derived vertebral bone quality (VBQ) score, a novel marker of marrow adiposity. In this retrospective cohort study, postmenopausal women with osteoporosis who received TPTD for ≥12 months after at least 24 months of DMAb were included. Propensity score matching (1:1) yielded 60 patients in the combination group and 60 in the sequential group. The primary outcome was the percentage change in VBQ (ΔVBQ%) from baseline to 12 months. Combination therapy resulted in a significant improvement in the bone marrow microenvironment (ΔVBQ%: -7.1% ± 3.0%), whereas sequential therapy did not (-0.5% ± 5.4%), with an adjusted between-group difference of -6.5% (95%CI: -8.1 to -4.9, P < 0.001). This was accompanied by superior gains in lumbar spine BMD (+9.4% vs. +5.6%, P < 0.001) and trabecular bone score (+4.8% vs. +1.5%, P < 0.001). Notably, the combination group maintained complete suppression of bone resorption (CTX), while the sequential group experienced a marked and sustained CTX rebound. Furthermore, in the combination group, the improvement in VBQ was significantly correlated with early bone formation activation (ΔPINP at M + 3, r = -0.50) and with gains in lumbar spine BMD (r = -0.40) and TBS (r = -0.77), suggesting a mechanistic link between modeling-based bone formation and microenvironmental restoration. In postmenopausal women with long-term prior denosumab treatment, initiating teriparatide while continuing denosumab was associated with significantly greater improvement in the vertebral bone marrow microenvironment (greater VBQ reduction), more consistent gains in lumbar spine BMD and TBS, and better control of bone resorption markers over 12 months compared with a sequential switch after denosumab withdrawal. These findings suggest that the combination strategy may be more favorable than a sequential switch for improving vertebral bone quality. While our data provide mechanistic support for current guideline recommendations, confirmation in prospective randomized studies is required before clinical translation.