In this retrospective study, we found that nearly 20% (238/1182) of acute OVCF diagnosed by MR imaging simultaneously involved two or more vertebra. The incidence of multiple OVCF was quite near to the incidence of 1/5 of new fractures post-vertebroplasty [4, 14, 15]. Moreover, we for the first time showed that simultaneous two OVCF (67.6%) was the most common form of MSVF, and more than half (53.3%) of MSVF had at least two OVCF in adjacent. The propensity of adjacent new fractures was also reported in the vertebral fracture cascade after cement augmentation, and the risk of new OVCF would increase 2.5 fold once the number of pre-existing fractured vertebra increased by one level [16, 17]. Probably due to the tendency of paired fractures in adjacent, MSVF demonstrated the same distribution predominance as SSVF in the thoracolumbar segment, but involved more vertebra in the nearby lumbar and thoracic spine. Therefore, based on the similarities in occurrence rate and distribution pattern between MSVF and new OVCF post-vertebroplasty [4, 14,15,16,17], we presume there might be two models of vertebral fracture cascade. One might start from a thoracolumbar vertebral fracture and extend gradually beyond to the nearby lumbar and thoracic spine, and the other one might occur much promptly as the form of MSVF simultaneously involving two or more vertebra in adjacent.

While multiple risk factors have been revealed to be associated with post-vertebroplasty re-fracture [3,4,5,6,7,8, 11, 14,15,16,17], it remains elusive what potentiates acute MSVF. Here, we found that females had 3.4-fold higher risk of vertebral fractures and experienced SSVF 3.1 years earlier than males. However, females rather than males in MSVF-2 were comparatively older than in SSVF, suggesting a risk factor of increasing age for acute MSVF in females. This is in line with previous report that older females with lower baseline BMD were at higher risk of repeated and multiple vertebral fractures [9, 16]. As multiple vertebral fractures were reported to be often caused by high-energy trauma [9], it is possible acute MSVF might experience higher incidence of spine trauma than SSVF. However, just on the contrary, we found that fewer MSVF reported apparent trauma such as fall on ground or crush injury to the spine, but more often denied evident spine trauma prior to the onset of back pain. Evidence accumulates that deteriorating osteoporosis promotes new fractures post-vertebroplasty [3, 8, 9, 16,17,18,19], and it is possible acute MSVF might have lower baseline BMD than SSVF. Here, BMD index was collected from 387 females (32.7% of study population), and only in the age group of 70–80 years old, MSVF-3/m showed lower T-score values than MSVF-2 and SSVF. Besides, comparison of comorbidities also identified no significant difference in the comorbidity type and number between MSVF and SSVF. Our study demonstrated that MSVF tended to occur in older females without significant spine trauma, lower baseline BMD, or multiple geriatric comorbidities. Vertebral fracture cascade in the form of acute MSVF might be driven by gender- and age-specific synergistic interactions of diverse risk factors.

OVCF are most often hospitalized with a complain of acute or chronic back pain [1, 2]. In this study of 1182 OVCF, the duration of pre-hospital back pain varied significantly from 1 day to longer than 6 months. While more than half of SSVF (58.9%) were hospitalized early with back pain ≤ 1 week, both MSVF-2 (19.2%) and MSVF-3/m (38.9%) reported higher ratio of pre-hospital back pain > 1 month and delayed hospitalization. One reason for the delay might be the fact that MSVF tended to occur in older females without significant spine trauma. In much aged patients with multiple comorbidities, prolonged observation of the back pain or conservative therapy would probably lead to delayed diagnosis and vertebroplasty in MSVF. As delayed vertebroplasty was reported to increase the risk of less maintained vertebral height and re-fracture [20, 21], it is also possible new vertebral fracture might occur during the observation or conservative therapy of back pain initiated by previous OVCF. In support of this hypothesis, although the ratio of vertebral bone marrow edema was not significantly different between SSVF and MSVF, MSVF-3/m caused more bone marrow edema restrained in the cranial or caudal half of vertebrae. Besides, the thoracolumbar vertebra in MSVF showed less compression and loss of anterior vertebral height than in SSVF. Collectively, our findings indicated a feature of inconspicuous but accelerated vertebral fracture cascade in the form of acute MSVF, highlighting a challenge of early diagnosis and vertebroplasty to alleviate back pain.

Our study was limited in its retrospective nature and relatively small number of included patients from a single spine center. The duration of pre-hospital back pain and type of spine trauma were based on patients’ chief complains after hospitalization, which unavoidably yielded lead time bias and recall bias [22]. Other risk factors of vertebral fracture cascade such as regional or global kyphosis [23], anti-osteoporosis medication [24,25,26,27], body mass index and life styles [6,7,8] were not evaluated to fully understand the mechanism of OVCF involving multiple vertebra. BMD was unavailable in 2/3 of the study population and in most of the males, which might underestimate the role of advanced osteoporosis in promoting MSVF. Recent evidence shows that biochemical marker of bone turnover is useful for evaluating bone fragility and therapy monitoring in postmenopausal osteoporosis [28, 29]. Further prospective studies with stratified participants and long-term follow-up are warranted for a better understanding of vertebral fracture cascade in the form of acute MSVF.