Most women experience no or modest decreases in aBMD during pregnancy and larger decreases during lactation with spontaneous recovery postweaning [7]. Additionally, they may experience a decrease in volumetric BMD and changes in bone microarchitecture at the distal tibia and proximal radius [8, 9]. In PLO, bone loss is aggravated, and impaired volumetric BMD and bone microarchitecture have been reported compared to healthy controls [10] and healthy lactating women [11]. The woman in this report also had a severely impaired aBMD, volumetric BMD, and bone microarchitecture compared to healthy controls, which even after treatment persisted until 40 months postpartum.

In line with previous studies [7, 8, 10, 11], the trabecular bone was more affected than the cortical bone but recovered earlier. At 7 months postpartum, aBMD T-score was worse at the lumbar spine (predominantly trabecular) than at the total hip (cortical and trabecular), and trabecular BMD was relatively lower than cortical BMD at the distal radius and tibia. Until 40 months postpartum, lumbar spine aBMD showed larger increases than total hip aBMD, and trabecular BMD increased more at the radius and tibia than cortical BMD. The increase in trabecular BMD seemed the result of thickening of existing trabeculae and possibly the fusion of neighboring trabeculae. Nevertheless, a large increase was found in cortical BMD at the right tibia, possibly due to the filling of cortical pores and an increase in tissue mineralization. Despite the improvements, BMD and microarchitecture remained severely impaired at 40 months postpartum compared to healthy controls. This could have contributed to compensatory periosteal bone apposition in order to improve bone strength, as reflected by the changes in cortical and trabecular area and previously also suggested by Kovacs [7].

A possible factor in the pathological bone loss in PLO is heredity. Recent studies showed that multiple genetic mutations, in LRP5 and WNT1, can aggravate bone loss in PLO [2, 10, 12, 13]. Although no monogenic cause was found in this woman, osteoporosis and osteopenia occur in her immediate family. Additionally, HR-pQCT scans of her mother’s distal radius and tibia showed similar cortical discrepancies (Online Resource 4). It suggests a polygenic origin in this patient with unknown contribution of both her parents’ side.

Currently, there is no clinical guideline for the management of patients with PLO, and treatment strategy remains challenging and controversial due to the lack of intervention-controlled research. Termination of breastfeeding is important but may not be sufficient, especially in case of multiple vertebral fractures [14]. Prolonged bisphosphonate treatment is suggested to be successful [15], but bisphosphonates accumulate in the skeleton and can be released in a subsequent pregnancy, possibly harming the fetus. TPTD does not accumulate in the body and has been found to increase aBMD, relieve pain, and prevent new vertebral fractures in PLO [16,17,18]. It also results in greater increases in aBMD compared to no TPTD treatment in PLO [19]. Interestingly, Lee et al. showed no significant difference in aBMD 3 years after discontinuation of TPTD treatment between PLO-patients with and without successive antiresorptive therapy [20]. It suggests that it may be advantageous not to start antiresorptive treatment after TPTD discontinuation when a PLO-patient desires to have children in the future. As this finding was not available when the woman presented at our hospital, we treated her with one dose of ZA after TPTD.

The largest improvements were generally seen during TPTD treatment and until 1 year after the ZA-dose, albeit site-dependent. At the distal radius, lumbar spine, and hip, the largest changes occurred during treatment. At the distal tibiae in contrast, they occurred before treatment. The latter may be due to spontaneous recovery and possibly also due to the physiotherapy that the woman followed after pregnancy because of right leg pain after stumbling during pregnancy. The stumbling and consequent disuse of the right leg for 5 months may also explain the worse HR-pQCT parameters at 7 months at the right tibia compared to the left tibia.