Introduction

Osteoporosis is the most common bone disease, affecting approximately 500 million people worldwide [1]. This systemic skeletal disorder is characterized by microarchitectural deterioration of the trabecular bone tissue, decreased bone mineral density (BMD) and chemical compositional changes, leading to increased skeletal fragility and a higher risk of fractures. The clinical outcome of osteoporosis is often fractures from low-energy trauma, known as fragility fractures, which would not typically occur in healthy bones. Globally, 37 million fragility fractures occur annually, affecting one-third of women and one-fifth of men over the age of fifty [2–4]. These fractures result in significant economic cost and major health consequences for patients such as severe pain, dependency and disability. In this context, it has been shown that the risk of fragility fractures increases 5-fold during the first year following an initial fracture at the same or different skeletal sites [5], then gradually decreases from a factor of 2.7 after one year to 1.4 after ten years [6]. Numerous preclinical and clinical studies indicate that, during the bone regeneration, a significant systemic bone quality loss may occur throughout the skeleton, especially in osteoporotic patients, further aggravating the disease [7–11]. In addition, the bone regeneration process is frequently compromised by the poor osteogenic capability of osteoporosis [7,12], resulting in delayed healing, non-unions, bone deformities, chronic pain and even post-surgical complications such as infections [13–15]. In these challenging bone healing situations, distraction osteogenesis via bone transport is a recognized technique, clinically indicated due to its potential to promote bone healing [16–18]. This gold-standard orthopedic procedure consists of gradually displacing a bone fragment, along an osteotomized gap [16,19]. Despite its medical interest, there is a lack of knowledge about the skeletal bone quality alterations of applying such complex techniques in osteoporotic patients.

In clinical studies, dual energy X-ray absorptiometry (DXA) is the gold standard method for assessing osteoporosis and osteopenia due to its speed, clinical accessibility and low-radiation exposure. Nevertheless, this method is limited by its two-dimensional assessment of trabecular macroscale or apparent BMD and tends to overestimate BMD due to its poor differentiation between trabecular and cortical bone tissues [20,21]. In this context, there are alternative techniques for assessing bone quality that are less common for clinical applications due to their cost, their high levels of radiation or their invasive nature (e.g., bone biopsy extraction). However, these techniques are a great source of knowledge about bone quality in experimental research. In this regard, computed tomography (CT) and micro-computed tomography (μCT) are two widely used three-dimensional and high-resolution techniques for assessing, respectively, the macroscale BMD [22–25], the microscale BMD, and the microstructure of bone samples [26–28], thus offering a multiscale analysis of the bone mineral tissue. In addition, chemical composition analyses of bone samples can provide insights into the bone volumetric composition and stoichiometry of bone tissue as seen in studies on cortical and woven tissues of healthy subjects [29,30]. These techniques could enhance the understanding of bone regeneration in osteoporotic patients, aiding clinicians to develop novel patient-specific treatments and therapies. This could lead to a more effective medical care and follow-up of the disease, reducing recovery time and improving patients’ quality of life, as well as contributing to the development of preventive strategies focused on reducing the risk of refractures or secondary fractures.

Osteoporosis clinical and experimental research primarily focuses on understanding the biochemical and mechano-biological mechanisms underlying the disease, identifying risk factors [31–33], improving early detection techniques [1,34–39], developing better drug treatments or therapies for patients [40–43], or preventing the onset of the disease or fragility fracture through nutrition [44–46] or physical exercise [47,48]. A smaller proportion of this effort is dedicated to enhancing bone healing management. In this context, large animal models faithfully represent human bone conditions due to their size and anatomy, enhancing translational relevance and assessment of treatment efficacy in osteoporosis research. Numerous large animal studies have validated or characterized the effect of various osteoporosis induction protocols on bone quality [49–56], providing the temporal evolution data on BMD or microstructure of cortical or trabecular tissues, commonly by DXA and μCT [57–67]. For example, Zhang et al. [61] studied the influence of ovariectomy in sheep after 12 months from surgery by μCT of the trabecular tissue at the lumbar vertebra, femoral neck, mandibular angle and rib. They reported significant differences in trabecular microstructure degradation depending on the skeletal site location. Similarly, Bisazza et al. [57] compared DXA and CT techniques to assess temporal changes in trabecular and cortical BMD at lumbar vertebrae in the sheep osteoporotic model. They concluded that CT provides greater accuracy in detecting changes in BMD and bone microstructure. Nevertheless, few large animals’ studies have evaluated the skeletal bone quality evolution experienced by individuals adding the complexity of bone regeneration [10,11]. In this challenging context, Lill et al. [11] assessed by CT the macroscale BMD changes in cortical and trabecular tissues of the tibia between healthy and osteoporotic sheep groups after a mid-shaft tibia1 osteotomy. They found a significant apparent BMD downtrend, but without giving information about other bone locations or microstructure and chemical composition changes. In contrast, Bindl et al. [10] studied the influence of osteoporosis and right femoral metaphysis bone gap healing on different skeletal locations, providing insights into the microstructural bone quality loss in tibial cortical and trabecular tissues between healthy and osteoporotic sheep groups. However, there is a lack of knowledge regarding the temporal bone quality loss experienced by the subject-specific not only in terms of microstructure, but also in terms of mineral density and chemical composition, during the onset of osteoporosis and the subsequent bone regeneration process. Furthermore, as these studies present a simple fracture healing model, they do not provide information about how osteoporosis influences on bone regeneration processes of major complexity but frequently used in cases of severe trauma, such as the distraction osteogenesis model. In contrast, there are studies that have assessed different bone regeneration processes in healthy large animals, such as fracture healing [68–72] or distraction osteogenesis [73–77]. The woven bone generated within the callus has been characterized through different approaches like nanoindentation [78,79], biomechanical tests [11,80,81], histology [71,79,81,80], CT [11,73,75], finite element analysis from CT reconstructions [80,82] and gait analysis [74,75] among others. However, these studies lack information on the characterization of this immature tissue in osteoporotic subjects.

From all of this, it is essential to deepen understanding the skeletal bone quality changes in osteoporotic patients adding the complexity of bone regeneration. As a novelty, the present work will explore this combined condition through in vivo experiments using a large osteoporotic animal model undergoing a bone regeneration process treated by distraction osteogenesis via bone transport.

We hypothesize that the osteoporotic bone regeneration process significantly aggravates the systemic bone quality deterioration beyond the baseline effects of osteoporosis, which could be tested by measuring the mineral density, microstructure and chemical composition. In addition, the bone impairment may be intensified due to the presence of the disease. Thus, the study aims to elucidate and compare the bone quality changes in different bone locations of individuals during osteoporosis and osteoporotic bone regeneration. This will be achieved through multiscale imaging techniques (CT and μCT) and chemical composition analyses to explore temporal evolution of specific subject’s skeletons. In this sense, the macroscale and microscale BMD, microstructure, and chemical composition of different bone tissues (cortical, trabecular and woven bones) across different skeletal sites (the operated bone, its contralateral counterpart and a bone far from the operated bone) will be evaluated temporally and individually.

Materials and methods

This section is organized as follows: firstly, the osteoporotic animal model and the distraction procedure using bone transport (BT) are described. Next, the study design is outlined to quantify the individual temporal impact of osteoporosis and the regeneration of osteoporotic bone defect across different bone locations and tissues. Finally, the methodology followed in the different analyses is described: the apparent or macroscale BMD (BMD_CT) data measured by CT, the microscale BMD (BMD_μCT) and microstructure data measured by μCT, and chemical composition analysis (ash fraction, volumetric composition, and elemental mass content).

Osteoporotic animal model and distraction procedure

Osteoporosis was induced in a total of fifteen female merino sheep (weight 60.2 ± 5.6 kg), 2–4 years old using the protocol provided by Zarrinkalam et al. [54]. The animals were selected from a research farm and marked on the wool to avoid confounders. The randomly selection criteria ensures that the specimens are healthy and have proper vaccination and deworming protocols. The sample size was calculated to reduce the number of animals required to a relevant minimum, obtaining significant differences. For this purpose, the insights reported by Bindl et al. [10] was used as a reference, in which similar parameters were analyzed in the same osteoporotic animal model. The sheep were transported to the Clinical Veterinary Hospital of the University of Cordoba, where all in vivo experiments were conducted. This research center has spacious, fenced and partially roofed outdoor facilities, where the animals were housed and cared for. The induction period began with a bilateral ovariectomy. The ovariectomized sheep received periodic intramuscular injections of glucocorticoid (500 mg Solu-Moderín® + 7.8 ml injectable water) every 3 weeks for 33 weeks. Additionally, they were exclusively fed with a calcium-free diet till sacrifice, comprising 12% crude protein, 9% crude fiber, 6.5% crude ash, 2% crude fat, 0% calcium, 0.1% phosphorus and 0.1% sodium.

The BT experiments began with a surgical procedure performed on the right hind metatarsus of the ovariectomized animal at week 33 ± 2.5. The sheep was under general anesthesia and intubated during both surgical procedures (BT surgery and ovariectomy), while the body temperature, blood pressure, oxygen levels, exhaled carbon dioxide levels, and electrocardiograms were continuously monitored. As illustrated in Fig 1, an Ilizarov-type external fixator [73–75,83] was initially implanted using Ø4 mm Schanz screws. Once the metatarsus was stabilized, three cross-sectional osteotomies were made using a guided oscillating saw (1.2 mm thickness). These osteotomies created two diaphyseal bone fragments: a 25 mm proximal transportable bone fragment (previously attached to the distractor through two Ø2.5 mm Steinmann pins) and a distal 15 mm bone segment which was removed to generate a critical-size bone defect (Fig 1). Following the bone surgery, the calcium free diet was continued throughout the experimental period, but steroid injections were ceased to preserve postsurgical animal welfare. After one-week latency period, a 15-days distraction phase was carried out, applying a 1 mm/day distal displacement of the transportable bone fragment per day along the 15 mm bone gap. In this way, the naïve tissue formed within the proximal osteotomy was elongated (distraction callus) while the tissue of the critical-size bone defect was compressed (docking site callus). Finally, the animals were randomly sacrificed by an overdose of sodium pentobarbital IV Euthasol® at two different time points during the osteoporotic bone regeneration, at day 40 (n = 5) or at day 100 (n = 5) after the BT surgery.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Bone transport (BT) surgery procedure in the right hind metatarsus of an ovariectomized sheep.

Fixator implantation and bone defect performance: (A) proximal cross-sectional osteotomy; (B) 25 mm transport fragment; (C) 15 mm critical-size bone defect; (D) Ø4 mm Schanz-screws; (E) Ø2.5 mm Steinmann pins; (F) external fixator frames; (G) non-instrumented fixed bars.

https://doi.org/10.1371/journal.pone.0319910.g001

Animal welfare was guaranteed throughout the induction period, surgical interventions and experimental phases following the ARRIVE guidelines, European (63/2010/EU) and national (RD 53/2013) regulations on animal research. The animal ethics of this study was approved by University of Córdoba (Protocol Number: 2021PI/21).

Study design at different bone locations

As shown in Fig 2, various bone tissue types from three bone locations of the sheep were analyzed: the right hind metatarsus (M_O, cortical and woven tissues) as the operated bone, the distal left hind metatarsus metaphysis (M_NO, trabecular tissue) as the contralateral counterpart, and the iliac crest (IC, trabecular tissue) as a bone distant from the operated bone. The bone samples were evaluated at different time points: on the day of ovariectomy as a non-osteoporotic time point (Healthy), on the day of the BT surgery (week 33 after the ovariectomy) representing the osteoporotic time point (OP), and on the day of sacrifice, 40 or 100 days after surgery, representing the osteoporotic bone regeneration time point (OP + R40 or OP + R100, respectively). Table 1 shows the time points at which the samples were measured by the different analyses.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Study design overview.

https://doi.org/10.1371/journal.pone.0319910.t001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Study design scheme.

Ovariectomized sheep analyses of the right hind metatarsus cortical and woven bone tissue (M_O), the distal metaphysis trabecular tissue of the left hind metatarsus (M_NO), and the iliac crest trabecular tissue (IC).

https://doi.org/10.1371/journal.pone.0319910.g002

The influence of osteoporosis and osteoporotic bone regeneration on the bone type and location was evaluated with BMD_CT measurements in woven, cortical and trabecular bones by CT. And the analysis through scales and locations were assessed with BMD_μCT, microstructure and chemical composition measurements of trabecular bone (see Table 1). For this purpose, μCT and chemical composition analyses were performed in trabecular biopsies extracted from M_NO and IC. M_NO biopsies were only extracted after sacrificing (OP + R40 or OP + R100), as performing this procedure in vivo could compromise animal welfare. Meanwhile, the CT scans and the IC biopsy extractions were conducted in vivo for Healthy and OP time points during the surgical procedures (under general anesthesia and intubated), and ex vivo for OP + R40 and OP + R100 time points. Unlike the M_NO, the IC as distant bone is suitable for in vivo biopsy extractions as it is easily accessible and without major harvesting risk, with abundant trabecular tissue. The biopsies were extracted using a biopsy punch and were preserved at -80ºC in PBS-soaked gauzes.

For the different analysis performed on all bone samples, each sheep data reported at the Healthy time point is used as its individual control data. In this way, each sheep’s data at subsequent time points were also normalized by its respective control data, thus providing an individualized temporal evolution of the animal.

Macroscale data: BMD_CT using CT

As illustrated in Fig 3A, CT scans were performed in vivo on the animal under general anesthesia and intubated at the Clinical Veterinary Hospital of the University of Cordoba. All CT images were acquired using Revolution ACT (General Electric, Pekin, China) CT scanner (XYZ voxel size 460–570 x 460–570 x 625 μm/px). A CT phantom QRM-BDC/6–200® (PTW, Freiburg, Germany) was also included to linearly correlate BMD_CT (0–800 mg HA/cm³) with the stack’s Hounsfield Units (Fig 3B). All the bone samples (M_O, M_NO and IC) and the phantom inserts (6 x Ø = 18 mm, h = 200 mm) were individually segmented using the open-source image processing tool ImageJ-Fiji. From each axial cross-sectional slice, a mean and standard deviation value of the BMD_CT was estimated at each time point. Intermediate CT scans were taken at weeks 10 and 20 after ovariectomy (10w and 20w time points, respectively) to assess the temporal evolution of BMD_CT in M_NO during the osteoporosis induction phase. The trabecular tissue was segmented without including the cortical tissue in the M_NO (13% of total length of the left hind metatarsus) and IC (complete bone).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Macroscale BMD characterization of the osteoporotic bone.

CT assessment of the BMD_CT in cortical (operated metatarsus, M_O: proximal, transport and distal fragments), in woven (M_O: distraction callus, docking site callus) and in trabecular (non-operated metatarsus, M_NO). (A) CT measurement of an ovariectomized sheep. (B) CT phantom included in the CT measurement. C) CT XZ (top) and XY (down) measurement projections. (D) Cross-sectional BMD_CT throughout the bone distal percentage of M_O length (0% proximal, 100% distal) in an ovariectomized sheep at Healthy, OP and OP + R100 time points.

https://doi.org/10.1371/journal.pone.0319910.g003

As shown in Fig 3D, different representative regions within the operated metatarsus (M_O) were monitored to quantify the local and temporal evolution of the BMD_CT in the cortical and woven bone tissues according to each animal’s final CT scan (OP + R40 or OP + R100). Mean and standard deviation values of BMD_CT were calculated from three cortical fragments and two bone calluses. The cortical BMD_CT was quantified in the proximal, transport and distal fragments (13%, 9% and 12% of the total length of M_O, respectively). The woven bone BMD_CT was measured in the distraction and docking site calluses (5%, and 0.5% of the total length of M_O, respectively).

Microscale data: BMD_μCT and microstructure using μCT

The μCT measurements were taken with the μCT scanner (XYZ voxel size 22.17–34.61 μm/px) model Y. COUGAR SMT (YXLON, Hamburg, Germany) at the University of Seville Research, Technology and Innovation Center (CITIUS). As shown in Figs 4A and 4B, the μCT measurement includes harvested trabecular biopsies from the iliac crest (IC) or contralateral metatarsus (M_NO), along with a cylindrical μCT phantom, both fixed to a 3D printed support and placed within the μCT scanner. The μCT phantom (Ø = 4.5 mm, h = 5 mm) model MicroCT-HAD4.5® (QRM, Moehrendorf, Germany) was included to establish a linear correlation between the BMD_μCT of the phantom inserts (0–1200 mg HA/cm³) and the Hounsfield Units of the μCT stack. The μCT measurements were carried out using an open μCT multifocus tube of 25–160 kV voltage and current intensity of 0.01–1 mA.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Microscale BMD and microstructure characterization of the osteoporotic trabecular tissue samples.

(A) μCT set up. (B) XY μCT stack: IC biopsy and μCT phantom fixed to a 3D printed support (top). (C) μCT measurement procedure.

https://doi.org/10.1371/journal.pone.0319910.g004

The complete post-processing methodology for the μCT measurements is illustrated in Fig 4C, according to the guidelines for assessment of bone microstructure provided by Bouxsein et al. [84]. First, the BMD_μCT of the phantom inserts (5 x Ø = 0.8 mm, 4.5 mm height) was measured using ImageJ-Fiji® by means of the plugging BoneJ. Using the Amira-Avizo® software, all trabecular samples were segmented with a lower threshold of 400 mg HA/cm³ for normalized segmentation value of Hounsfield Units. The threshold was confirmed by visual inspection after checking its capability to adequately discriminate between trabeculae and other components presented in the different bone samples (pores, blood, air, etc.). From the 3D reconstruction of the trabecular volume, the trabecular microstructural parameters were calculated: the trabecular thickness (Tb.Th), the trabecular separation (Tb.Sp), the bone volume per total volume (BV/TV), the trabecular number (Tb.N), and the structure model index (SMI). Finally, the degree of anisotropy (DA), the connectivity density (Conn.D), the fractal dimension (FD), and the mean cross-sectional microscale BMD_μCT value of the sample was measured using ImageJ-Fiji® by means of the plugging BoneJ.

Chemical composition analyses

Chemical composition analyses were performed on the trabecular biopsies from the iliac crest (IC) and the contralateral metatarsus (M_NO) after the μCT measurements at the University of Seville Research, Technology and Innovation Center (CITIUS). Firstly, a manual grinding procedure was carried out using a sterilized pestle and mortar to reduce the samples to a particle size of 1 mm. Subsequently, the total mass m_t of the sample was measured. This mass comprises the mass of water m_w, the organic mass or organic phase m_o and the mineral mass or mineral phase m_m, as defined in Eq. 1.

(1)

Afterwards, a drying procedure was applied to the sample by heating it at 105ºC for 1 hour in a BINDER VD 23 vacuum drying chamber (BINDER GmbH, Tuttlingen, Germany). The samples were weighed and heated again at 105ºC every 15 minutes until a constant dry mass m_d was achieved.

(2)

Then, the samples underwent an ashing process in a Nabertherm Muffle Furnace (Nabertherm GmbH, Lilienthal Germany), following the protocol provided by Martínez-Reina et al. [29]: (A) a lineal increase to 250ºC for 30 min; (B) a constant temperature of 250ºC for 1 hour; (C) a lineal increase from 250ºC to 650ºC for 30 min; (D) a constant temperature of 650ºC for 2 hours; (E) sample weighing and 650ºC constant temperature for 30 min until a constant ash mass m_a is achieved. The protocol burns the organic phase of the sample, so m_a corresponds with the mineral phase, as shown in Eq. 3.

(3)

To assess the mineral content, the ash fraction, α, was calculated using Eq. 4.

(4)

The volumetric fractions v_x of the mass x is obtained by:

(5)

where the density of the mass is ρ_w = 1 g/cm³, ρ_o = 1.43 g/cm³ [85] and ρ_m = 3.12 g/cm³ [29].

Later, an elemental analysis was carried out using a TruSpec Chns Micro analyzer (LECO Corporation, St Joseph, MI, USA) to determine the mass percentage of carbon. After previously dissolving the samples in hydroalcoholic acid, the mass percentage of calcium, phosphorus, magnesium, potassium, sodium and strontium were finally acquired through inductively coupled plasma atomic emission spectrometer Ultima 2 (HORIBA Jobin Yvon, Edison, NJ, USA).

Results

Impact of osteoporosis and osteoporotic bone regeneration on the bone tissue type

The temporal evolution of the BMD_CT in the cortical, woven, and trabecular bone tissues are presented in Tables 2 and 3. These data normalized to the individual Healthy time point value are also represented in Figs 5A-5C. The underlying data are presented in S1 and S2 Tables, while the respective statistical analysis is shown in S5 and S6 Tables.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Macroscale BMD characterization in the osteoporotic operated bone.

https://doi.org/10.1371/journal.pone.0319910.t002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Macroscale BMD characterization in the osteoporotic contralateral bone.

https://doi.org/10.1371/journal.pone.0319910.t003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Individual temporal evolution of macroscale BMD in osteoporotic hind metatarsi.

(A) Cortical BMD_CT in M_O: proximal, transport and distal fragments. (B) Woven BMD_CT in M_O: distraction callus and docking site callus. (C) Trabecular BMD_CT in M_NO. Data normalized to the Healthy time point (compiled in S1 and S2 Table) and presented as mean ± standard deviation values. Significant differences among time groups: * means p < 0.05, ** means p < 0.01 and *** means p < 0.001. Significant differences of one time point versus Healthy time point: # means p < 0.05, ## means p < 0.01 and ### means p < 0.001. Significance evaluated by Kruskal-Wallis test followed by Dunn–Sidak post-hoc and correction for cortical and trabecular tissue and Mann-Whitney U test for woven tissue. Statistical analyses compiled in S5 and S6 Tables.

https://doi.org/10.1371/journal.pone.0319910.g005

Fig 5A shows the mean and standard deviation of the individual temporal evolution of the BMD_CT in the different cortical bone regions within the operated metatarsus (M_O): proximal, transport and distal fragments. No significant differences were observed between the Healthy and OP time points. However, a significant decrease in BMD_CT was observed during the osteoporotic bone regeneration process when comparing the bone regeneration time points (OP + R40 and OP + R100) with the BT surgery time point (OP). This reduction increases over time in the distal direction of the metatarsus, with BMD_CT loss of 20% of in distal fragment and 8% in proximal one at OP + R100 relative to Healthy. This reduction of cortical BMD_CT is more significant on day 100 post-BT surgery than on day 40.

Regarding the woven bone tissue_, Fig 5B shows the individual temporal evolution of BMD_CT in the distraction and docking site bone calluses, relative to the mean cortical BMD_CT measured in the corresponding cortical region of the metatarsus at the Healthy time point. During the osteoporotic bone regeneration, a BMD_CT increase was observed in the distraction callus, with mean values of 37% at OP + R40 and 46% at OP + R100 relative to Healthy measures. The docking site callus BMD_CT was also increased, with mean values of 45% at OP + R40 and 48% at OP + R100, and a reduction in the standard deviation over time.

Fig 5C illustrates the relative individual temporal BMD_CT evolution of the contralateral distal trabecular tissue. A progressively significant BMD_CT loss is observed throughout the osteoporotic induction period, reaching 6.4% relative to Healthy value at the BT surgery (OP). Then, this reduction is stabilized and remains constant during the bone regeneration process, reaching 8.6% and 7.8% BMD_CT loss at OP + R40 and at OP + R100, respectively.

Impact of osteoporosis and osteoporotic bone regeneration through scales and locations

To analyze the impact of osteoporosis and the osteoporotic bone regeneration through scales (macroscale and microscale) and locations, firstly the microscale data are described. The mean and standard deviation values of the temporal evolution of the BMD_μCT and the microstructure parameters in the trabecular iliac crest (IC) biopsies and contralateral metatarsus (M_NO) are presented in Table 4. The IC data are normalized with respect to the Healthy time point value in Fig 6B. The underlying data are presented in S3 Table, while the respective statistical analysis is shown in S7 Table.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Microscale BMD and microstructure characterization in osteoporotic non-fractured trabecular bones.

https://doi.org/10.1371/journal.pone.0319910.t004

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. Microscale characterization of osteoporotic trabecular iliac crest (IC) biopsies.

(A) 3D reconstruction comparison based on μCT, rendered using Amira-Avizo® software. (B) Individual temporal evolution of microscale BMD and microstructure. Data normalized to the Healthy time point (compiled in S3 Table) and presented as mean ± standard deviation values. Significant differences among time groups: * means p < 0.05, ** means p < 0.01 and *** means p < 0.001. Significant differences of one time point versus Healthy time point: # means p < 0.05, ## means p < 0.01 and ### means p < 0.001. Significance evaluated by Kruskal-Wallis test followed by Dunn–Sidak post-hoc and correction. Statistical analyses compiled in S7 Table.

https://doi.org/10.1371/journal.pone.0319910.g006

In the IC, no significant differences were observed in any parameter measured during the osteoporotic induction period (Healthy to OP). In contrast, significant differences were reported when comparing the trabecular separation (Tb.Sp), the bone volume per total volume (BV/TV), the trabecular number (Tb.N) and the connectivity density (Conn.D) between the bone regeneration time point (OP + R40 or OP + R100) with the OP time point. As shown in Figs 6A and 6B, there was a significant increase in the Tb.Sp (68.1%), and a significant decrease in the BV/TV (30.3%), in the Tb.N (28.3%), and in the Conn.D (53.9%) from the OP to the bone regeneration time point (OP + R40 or OP + R100). Regarding the remaining measured parameters, no significant differences between time points were obtained for the BMD_μCT, the trabecular thickness, the structure model index, the degree of anisotropy and the fractal dimension. In addition, no significant differences were found between OP + R40 and OP + R100 time points in any of the parameters measured.

With respect to the previously mentioned impact through scales and locations, microscale BMD (BMD_μCT) and macroscale BMD (BMD_CT) are shown for the trabecular tissue from the IC and M_NO in Figs 7A and 7B, respectively (statistical analysis in S6 and S7 Tables, respectively). No significant BMD differences were found between both bone regeneration time points (OP + R40 and OP + R100) for the trabecular bone locations assessed at both scales.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 7. Multiscale BMD evaluation in osteoporotic non-fractured trabecular bones.

Assessment at OP + R40 and OP + R100 time points, 40 and 100 days after the BT surgery, respectively. (A) Iliac crest (IC) assessment. (B) Non-operated metatarsus (M_NO) assessment. Data presented as mean ± standard deviation values. * means p < 0.05, ** means p < 0.01 and *** means p < 0.001. Significance evaluated by Kruskal-Wallis test followed by Dunn–Sidak post-hoc and correction for IC and Mann-Whitney U test for M_NO. Statistical analyses compiled in S6 and S7 Tables.

https://doi.org/10.1371/journal.pone.0319910.g007

Chemical composition characterization

The temporal evolution of the chemical composition analyses (ash fraction and volumetric composition, and the percentage of the elemental mass content) in the trabecular iliac crest (IC) biopsies and contralateral metatarsus (M_NO) are shown in Table 5 as mean and standard deviation values. Both IC analyses are also presented relative to the individual Healthy time point values in Figs 8A and 8B. The underlying data are presented in S4 Table, while the respective statistical analysis is shown in S8 Table.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Chemical composition characterization in osteoporotic non-fractured trabecular bones.

https://doi.org/10.1371/journal.pone.0319910.t005

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 8. Individual temporal evolution of chemical composition in osteoporotic trabecular IC biopsies.

(A) Individual temporal evolution of ash fraction (α), and volumetric composition of water (ν_w), organic phase (ν_o), and mineral phase (ν_m). (B) Individual temporal evolution of the percentage of elemental mass content of C, Ca, P, Mg, K, Na, and Sr. Data normalized to the Healthy time point (compiled in S4 Table) and presented as mean ± standard deviation values. Significant differences among time groups: * means p < 0.05, ** means p < 0.01 and *** means p < 0.001. Significant differences of one time point versus Healthy time point: # means p < 0.05, ## means p < 0.01 and ### means p < 0.001. Significance evaluated by Kruskal-Wallis test followed by Dunn–Sidak post-hoc. Statistical analyses compiled in S8 Table.

https://doi.org/10.1371/journal.pone.0319910.g008

For the IC evolution presented in Fig 8A, no significant differences were observed in the individual ash fraction and volumetric composition values during the osteoporotic induction period (between Healthy and OP time points). However, these parameters experienced different changes during the osteoporotic bone regeneration process. A significant reduction of up to 31.6% relative to Healthy value was reported for the ash fraction (α) with no significant differences between both bone regeneration time points (OP + R40 and OP + R100). The water and organic phase volumetric fractions followed opposite trends, reaching a significant reduction of up to 35.7% in the volumetric fraction of water (ν_w) and a significant increase of up to 48.8% in the volumetric fraction of organic phase (ν_o) at the OP + R100 time point. Regarding the volumetric fraction of the mineral phase (ν_m), it showed a significant reduction of 45.4% relative to Healthy between OP and OP + R40 time points, which recovered Healthy values at OP + R100.

Fig 8B shows the individual temporal evolution of the elemental mass content percentage of C, Ca, P, Mg, K, Na and Sr from the previously mentioned mineral phase of the IC. Table 5 shows the mass percentage of these elements, with Ca representing around 40%, P around 20%, and the sum of the other elements analyzed representing less than 6%. During the osteoporotic induction period, the mass percentage of C and Mg showed an increase of 20.7% and 22.8% relative to Healthy measurements, respectively. No changes were found in the remaining elements analyzed. During the osteoporotic bone regeneration process, the elements exhibited different trends, although the mass percentage of Na and Sr remained unchanged. The mass percentage of C significantly increased again by 29.0% at OP + R40 (49.7% relative to Healthy value), but it returned to Healthy values at OP + R100. The mass percentage of Ca and P reported an increase of 10.14% and 14.23% at OP + R100, respectively, without differences in the early period of the osteoporotic bone regeneration (OP + R40). With respect to the mass percentage of Mg, it showed also a new significant increase of 33.34% during the osteoporotic bone regeneration (56.12% relative to Healthy value), without significant differences between both bone regeneration time points. Finally, the mass percentage of K reported a reduction trend of up to 35.02% at OP + R100, although without significant differences.

In addition, the osteoporotic trabecular bone comparison presented in Table 5 shows no significant differences in the ash fraction, volumetric composition, or elemental mass content between both osteoporotic bone regeneration time points (OP + R40 and OP + R100) at both trabecular bone locations.

Discussion

This in vivo study quantitatively assesses the individual temporal bone quality alterations due to osteoporosis and subsequent bone defect regeneration via bone transport in a large animal model. Different bone tissues (cortical, trabecular and woven) at different bone locations were evaluated by CT, μCT and chemical composition analyses. Specifically, three different osteoporotic bones were analyzed: the operated metatarsus with a critical-size bone defect under the regeneration process, its contralateral counterpart, and a bone far from operated bone, the iliac crest. Thus, this work covers, for the first time in the literature, the mineral, microstructural and chemical composition alterations suffered by the skeletal system of individual subjects with osteoporosis undergoing a complex bone regeneration process treated by distraction osteogenesis via bone transport.

Beginning with the macroscale analysis of the operated bone by CT, the results showed that the cortical BMD_CT of the operated metatarsus is not affected by 33 weeks of osteoporosis induction (Healthy to OP in Table 2 and Fig 5A). In contrast, the trabecular BMD_CT suffered a significant reduction at the distal contralateral counterpart (almost 7% relative to Healthy value) after the osteoporosis induction process (Healthy to OP in Table 3 and Fig 5C). These results are in line with previous findings in several studies in the same osteoporotic animal model [52,55,57]. Osteoporotic trabecular BMD_CT loss is associated with increased porosity, a hallmark of osteoporosis. Note that the apparent trabecular CT segmentation does not differentiate between trabeculae and pore phase. After the bone surgical intervention (OP), the distal contralateral trabecular tissue appears to be affected by the osteoporotic bone regeneration, reporting a non-significant reduction of BMD_CT (Fig 5C). During this period (OP to OP + R40 or OP + R100), the woven tissue of the operated bone experienced an increased level of ossification (Table 2 and Fig 5B) while all cortical tissue samples (located at the proximal, transport and distal fragments) reported a reduction of the BMD_CT. This decrease over time is more significant in the distal direction of the bone (Table 2 and Fig 5A). Particularly, the apparent cortical fragments BMD_CT decreased by 9.2% in proximal, 14.3% in transport, and 20% in distal fragments with respect to Healthy values after 100 days from BT surgery (OP + R100). Ren et al. [87] observed this early-stage cortical remodeling after osteotomy (63 and 84 days of bone regeneration) in healthy sheep tibia fracture healing model, with BMD_CT reductions of 1.3% and 2.2% at the proximal and distal bone fragments ends, respectively. The greater reductions in the current study are likely due to the higher size of our defect and the influence of induced osteoporosis. Augat et al. [68] proposed that this phenomenon might result from reduced load-bearing capacity in the operated limb following surgery, leading to bone tissue readaptation and resorption [74,83]. This condition induces a bone tissue readaptation in the intervened limb resulting in bone resorption. Another possible explanation that also could lead to an operated bone decreased osteoblastic activity, is the reduction of vascularization in the distal direction due to the osteotomies and the size of the defect [88]. This hypothesis can be supported by the more pronounced degradation of the cortical tissue in the distal direction of the metatarsus (Fig 5A). Augat et al. [68] also observed BMD_CT differences although not significant, probably due to the simplicity of the defect generated (2 mm osteotomy). One last hypothesis could be the need for bone mineral to repair the bone defects, which is in line to the mentioned increased level of ossification reported in both calluses at both bone regeneration, especially early in the docking site callus regeneration, as mechanically characterized by Mora-Macías et al. [78] in BT healthy sheep model. In the present study, the mentioned mineral need becomes even more critical due to the poor osteogenic capability caused by the pathology [7,12]. In this context, bone defect regeneration processes could produce a systemic skeletal disorder, making the areas closest to the defect a potential source of mineral resources to contribute to the bone regeneration [68]. However, the transport fragment does not seem to have a significant mineral contribution during the first 40 days after the bone intervention (OP to OP + R40 in Table 2 and Fig 5A). This outcome is probably due to the vascular disconnection of this transport fragment which prevents its minerals from being used as a resource for both adjacent woven mineralization. In contrast, there was a significant loss of BMD_CT at 100 days after surgery (OP to OP + R100), suggesting the need for prior angiogenesis and woven bridging to contribute to both callus osteogenesis. Concluding with the macroscale characterization, in the operated bone, the significant cortical BMD_CT loss over time and woven mineralization indicates that the mineral has been mobilized for the bone defects regeneration treated by distraction osteogenesis. This need appears to be even more critical due to osteoporosis. In contrast, the influence of this disease alone in the BMD_CT seems to be limited to the trabecular tissue of the contralateral bone (Fig 5C).

Regarding the microscale analysis of the trabecular tissue performed in the iliac crest by μCT (Table 4 and Figs 6A and 6B), no significant differences were reported after 33 weeks of osteoporosis induction in any of the computed parameters (Healthy to OP). This finding supports previous studies indicating that osteoporosis does not uniformly affect trabecular tissues at different locations [89–91]. In contrast, the iliac crest trabecular microstructure was shown to remodel during the first 40 days after the BT surgery (OP to OP + R40), with a significant increase in the Tb.Sp (67%), and a decrease in the BV/TV (31%), in the Tb.N (28%), and in the Conn.D (54%) in relation to Healthy value. Consequently, the iliac crest trabeculae after the intervention are notably less dense and interconnected, with fewer trabeculae. As far as the authors are concerned, there are no studies assessing the individual microstructural changes of the trabecular tissue throughout a bone defect regeneration process in an osteoporotic sheep model to compare. However, our findings are in line with Bindl et al. [10], who compared osteoporotic and healthy sheep groups, the osteoporotic right femoral fracture healing also induces a decrease in the BV/TV of 32% and the Tb.N of 31%, and a 57% increase in the Tb.Sp at the right tibial trabecular tissue at 56 days post-fracture. In addition, the μCT analysis conducted by Fischer et al. [4] in osteoporotic mice revealed a significant degradation of trabecular microstructure (25% reduction in BV/TV and Tb.N) in the lumbar vertebrae of the femur fractured group compared to the non-fractured group. Regarding the BMD_μCT measured 40 days after BT surgery (OP to OP + R40), no significant differences were reported. Unlike CT, which computes an apparent BMD (BMD_CT) with trabeculae and pores, μCT has the resolution to analyze the trabeculae focusing on the mineral phase, which appears to maintain the mean BMD_μCT while impairing the trabecular iliac crest microstructure. By 100 days post-surgery, the absence of significant differences in BMD_μCT and microstructure parameters of the iliac crest with respect to 40 days after surgery (OP + R40 to OP + R100, 60 days difference) suggests that trabecular tissue far enough from the regeneration focus is not affected by the bone regeneration in an osteoporotic subject (Table 4 and Figs 6A and 6B). This is in line with the multiscale BMD results of both trabecular bones (Figs 7A and 7B) and the trabecular microstructure results of the contralateral metatarsus (Table 4). This microstructure was observed to be weaker (higher Tb.Sp, lower Tb.N and Conn.D) than that reported by Blázquez-Carmona et al. [77] in the operated metatarsus trabecular tissue of healthy sheep undergoing a bone lengthening process using distraction osteogenesis. According to this, the osteoclastic activity seems to increase in the osteoporotic trabecular tissue of far non-fractured bones at the early stage, during the first 40 days after the bone surgical intervention (OP to OP + R40). This fact is probably a systemic response of the osteoporotic skeleton to further contribute to the supply of minerals for the bone repair of the operated metatarsus. In addition, the main microscale changes in the trabecular tissue in the non-fractured bones are due to the combination of osteoporosis with the bone callus regeneration process, rather than solely osteoporosis.

In terms of chemical composition analyses of the iliac crest trabecular tissue, the ash and volumetric composition temporal evolution results are illustrated in Table 5 and Fig 8A. As far as the authors know, there are no studies in which the volumetric composition is assessed during osteoporosis and osteoporotic bone regeneration (the same for the ash fraction) at the non-fractured trabecular tissue to compare. Considering the differences, the results could be compared with those of Martínez-Reina et al. [29], who analyzed the volumetric composition and elemental mass content percentage of the lamellar (cortical) and woven tissue from the metatarsus of healthy sheep undergoing a BT regeneration process. According to Table 5 and Fig 8A, the parameters measured did not report significant changes at the end of the osteoporosis induction (Healthy to OP), consistent with Bloebaum et al. [92], who found no differences in the ash fraction between young and osteoporotic women groups at different trabecular sites. During the first 40 days after the bone surgical intervention (OP to OP + R40), a significant loss of the ash fraction took place (31.6% with respect to Healthy value). With respect to the volumetric composition of the iliac crest during the bone regeneration (OP to OP + R40 or OP + R100), the significant opposite trends reported in water volumetric fraction (downtrend) and organic phase volumetric fraction (uptrend) seems to indicate a volumetric change comparable to that reported by Martínez-Reina et al. [29] in healthy woven bone. In this immature bone tissue, a high volumetric fraction of organic phase means that it is undergoing a process of mineral change, either absorption or resorption. In addition, a mineral phase loss of volumetric fraction occurred in the early period of osteoporotic bone regeneration, which appears to recover over time. All this information seems to indicate that the osteoporotic bone regeneration stimulates the production of organic matrix in distant non-fractured bones, which suggest the displacement of blood volume to contribute with the inflammation and the transport of minerals. This early loss of mineral phase mass in relation to the mineral phase plus organic phase mass (ash fraction) and water volumetric fraction reflects, once again, a mineral mass resorption and mobilization from the trabecular tissue of the iliac crest to contribute to the osteoporotic regeneration of the operated metatarsus. This hypothesis is consistent with the trabecular iliac crest microstructure parameters degradation discussed above (Table 4 and Figs 6A and 6B).

The percentage of the elemental mass content from the mineral phase of the iliac crest trabecular tissue is shown in Table 5 and Fig 8B. While no direct comparisons are available in the literature, Martínez-Reina et al. [29] measured the mass percentages of C, P, and Ca in non-osteoporotic cortical and woven tissue. The comparisons between this study and the current work are limited due to the osteoporosis influence and the differences between these bone tissues with the trabecular tissue, without also analyzing other elements such as Mg, K, Na and Sr. As shown in Fig 8B, the percentage of elemental mass content of C and Mg experienced an increase during the osteoporosis induction (Healthy to OP). An increase in the mass percentage of C (probably from a degradation of the organic matrix) and Mg, indicates an increase in the carbonate content of the hydroxyapatite, which may make it less stable and more fragile and susceptible to an osteoporotic increased reabsorption. However, the elemental mass percentages of Ca, P, K, Na and Sr remained constant, showing that in this trabecular bone the percentage of hydroxyapatite has not changed significantly due to osteoporosis. The most abundant metallic elements in the mass percentage of hydroxyapatite (Ca and P) appear not to change significantly during the osteoporotic bone regeneration (OP to OP + R40 or OP + R100) of the operated metatarsus. However, it cannot be assumed that the mineral mass is not being lost, since the results are described as mass percentage. The same applies to three of the less abundant elements in the bone mineral, C, Na and Sr. In contrast, the mass percentage of K and Mg reported a significant decreasing and increasing trend, respectively, during the osteoporotic bone regeneration process, not reported during the osteoporotic induction period. This fact suggests a metabolic and homeostatic adaptation through the redistribution of secondary elements. In this regard, the increase in Mg could improve the formation of new bone tissue and cellular activity [93], while the decrease in K could indicate its less direct involvement in mineralization, thus focusing on the needs of osteoporotic bone regeneration of the operated bone. With respect to the comparison of the osteoporotic bone regeneration points for both trabecular tissues of non-fractured bones (iliac crest and contralateral metatarsus) in all the analyses (CT, μCT and chemical composition) performed in this work (Tables 4 and 5), no significant differences were found. This fact may indicate that both trabecular tissues have behaved in a similar way between days 40 and 100 after the bone surgery (OP + R40 to OP + R100), prioritizing the osteogenesis of operated bone woven tissues over the recovery of non-fractured trabecular osteoporotic tissue.

As for the limitations of the work, it should be noted that not all the analyses presented have been carried out in all the bone tissues and locations considered throughout all the time points (Healthy, OP, OP + R40 or OP + R100) of measurement of the sheep. In the case of μCT scanning and chemical composition analysis, the individual temporal evolution is only completed at the iliac crest, as any in vivo biopsy extractions at the operated or contralateral metatarsus would have compromised the welfare of the animals. In addition, two sheep suffered health difficulties and had to be slaughtered before the BT surgery. Meanwhile, three animals experienced post BT surgery complications that made it necessary to slaughter them before the bone regeneration time point. Thus, these specimens did not provide the complete individual temporal evolution data. The CT scans were limited to a single time point during bone regeneration due to the presence of metal elements in the implanted external fixator, which required post-sacrifice CT scans after fixator removal. Consequently, the comparison between the OP + R40 and OP + R100 time points refer to animals from different sacrifice groups, therefore it is not possible to talk about an individual temporal evolution when comparing both time points. Finally, the influence of osteoporosis during the bone regeneration process could not be determined explicitly since the study does not consider a healthy control group undergoing the same bone regeneration process.

In conclusion, this study provides novel quantitative insights into the alterations in bone quality experienced by individual subjects during the onset of osteoporosis combined with bone regeneration processes treated by distraction osteogenesis. Specifically, multiscale BMD, microstructure and chemical composition were assessed in different bone tissues and locations in osteoporotic large animal subjects undergoing a critical-size bone regeneration process. It is determined firstly that osteoporosis alone only seems to significantly impair the bone quality of trabecular tissue heterogeneously depending on the bone localization. Secondly, the osteoporotic bone regeneration process following bone transport interventions results in a systemic skeletal disorder, which is greater than osteoporosis itself, and appears to be intensified by this disease. At the operated bone, this combined condition caused a significant loss of cortical bone quality that was aggravated over time in the distal direction. In addition, the trabecular tissues of far bones were also especially affected during the early stage of bone regeneration, deteriorating their microstructure and altering their chemical composition. This generalized skeletal bone quality impairment seems to be a quick response of the organism to regulate the blood mobilization of bone mineral sources that can contribute to enhance the woven tissue mineralization of the operated bone. This mineral requirement is even more critical in osteoporotic individuals in whom the osteogenic capacity is severely diminished by estrogen deficiency, accelerating bone resorption and decreasing osteoblastic activity.

In this sense, the present work reflects the need to study the bone quality by different measurement approaches (multiscale BMD, microstructure, and chemical composition) at different bone locations to evidence the aggravation of the osteoporotic bone regeneration process. In osteoporotic patients who need distraction osteogenesis procedure, it is important to prioritize the continuous monitoring of the bones statistically more prone to a new fracture, especially the already threated bone under regeneration, whose fixation is at high risk of being mechanically and fatally compromised. Therefore, it is crucial the continuous bone quality care and clinical follow-up, especially during the early stage of bone regeneration, to ensure adequate healing, considering the several risks of refractures or secondary fractures.

Supporting information

Acknowledgments

Grant PID 2020–113790RB-I00 funded by MICIU/AEI/10.13039/501100011033.

The authors would like to thank veterinary technicians Juan Ramón Ríos Pérez, Ana Soto Fernández and Álvaro Rincón Panadero for their commitment, effort and involvement in this work.