Introduction

As humans age, the rate of bone resorption by osteoclast cells outpaces the rate of bone formation. The mineral content of aged bones declines, eventually resulting in osteoporosis—a condition in which bones become more fragile and prone to fractures [1]. In accordance with World Health Organization (WHO) criteria, 10% of US women older than 50 years had osteoporosis and another 49% had osteopenia at the femur neck in 2005–2006 [2]. In 2010, osteoporosis affected roughly 22 million women and 5.5 million men in the European Union. In view of the variety of fragility fractures, including hip fractures, vertebral fractures, forearm fractures, the estimated economic burden is €37 billion per year [3].

Hence, research on osteoporotic fractures has increased over the past decades. Although bone mineral density (BMD) is considered to be the gold standard for the evaluation of bone strength and fracture risk [4], bone strength is determined by many other factors as bone microstructure, and bone components [4].

Besides methods for bone quality assessment that have been established in the clinical context, there are methods available to directly analyse the mechanical strength of bone tissue, such as micro-indentation, or nano-indentation tests [5, 6].

The aim of this study was to systematically summarize the current techniques commonly used to assess bone quality in human bone specimens, as well as the advantages and limitations of these methods.

Methods

The PRISMA (Preferred Reporting Items for Systematic Review and Meta-Analyses) checklist and algorithm [7] was used to conduct a systematic review of the literature to find all studies concerning the bone quality assessment of human bone specimens. Since data collection has already been completed at the time of PROSPERO registration, this review could not be registered with PROSPERO.

No primary personal data were collected; therefore, no additional ethical approval needed to be obtained.

Information source

A systematic review of the literature was searched by PubMed and Web of Science databases from January 2000 to April 2021. The language of the journal was limited to English, and the searched species was selected to be “human”. To further extend the search, the “similar articles” option of PubMed was employed in each paper.

Search strategy

A search was performed independently by two reviews (F.W. and L.Z.), and terms were used for the systematic search: “(bone quality”. Ti,ab.) AND “(human bone specimens). After removing duplicates, the reviewers scanned the search results by titles and abstracts. After identifying potentially pertinent articles, full-text articles were sourced and checked for suitability according to the inclusion and exclusion criteria (Fig. 1). Any controversy between the two authors was sent and discussed with a third independent author.

Fig. 1
figure 1

Preferred reporting item systematic reviews and meta-analysis (PRISMA) flow diagram of study selection

Full size image

Eligibility criteria

Studies were selected on the basis of the following inclusion criteria: (a) in vitro experiments regarding the bone quality assessment of human bone specimens; (b) research on bone composition (DBM, organic composition); (c) articles published in English. Exclusion criteria were: (a) no access to full text; (b) case reports and review papers; (c) studies on bone implants or screws, bone histological analysis, and clinical patients; (d) non-English language publications; (e) studies on animal bone specimens; (f) studies on finite element analysis (FEA) models.

Data extraction and analysis

Two authors (F.W. and L.Z.) independently performed data extraction and recorded this data using standard spreadsheet software (Excel for Mac 2016, version 16.2.9, Microsoft, Redmond, WA, USA). This included testing methods, authors and year of publication, journal of publication, study design, number of bone specimens, age, the site of specimens, main findings or summaries.

Assessment of study quality

The Newcastle–Ottawa Scale (NOS) [8] which contains three primary components: selection, comparability, and exposure/outcome, is being used to evaluate the quality of non-randomized trials. For this review, the quality of all studies, including bias, was assessed using the adapted Newcastle–Ottawa Quality Assessment Scale (Additional file 1). According to the total quality score, studies were evaluated with the highest score of eleven, as unsatisfactory (0–5), satisfactory (6–8), and good (9–11), which refers to a published article [9]. Two authors (F.W., and L.Z.) assessed all the included articles independently. Disagreements were recorded by discussion.

Statistical analysis

Continuous variables were described by the mean and standard deviation or median and range. Categorical variables were expressed with absolute and relative frequencies. Statistical significance is defined with P < 0.05. The large heterogeneity and lack of randomized controlled trials made it impossible to perform a meta-analysis. Furthermore, since the distributions of some indicators were only ranges, no other statistical analysis was possible.

Results

Study description and quality assessment

After scrutinizing the titles and abstracts, as well as examining the full texts, the remaining sixty-eight studies were included in the systematic review. Among them, forty-seven articles investigated several imaging modalities, including radiography, dual-energy X-ray absorptiometry (DEXA), CT-based techniques, and MRI-based modalities. Nineteen articles dealt with mechanical testing approaches, including traditional testing methods and novel indentation techniques. Nine articles reported the correlation between bone quality and bone composition, such as DBM, organic composition (Figs. 2 and 3).

Fig. 2
figure 2

Different testing methods of bone quality, as well as the intrinsic bone composition that affects bone quality

Full size image
Fig. 3
figure 3

The percentage of different modalities in assessing bone quality

Full size image

A total of 2898 human cadaveric bone specimens were included (Table 1). The number of specimens ranged from 4 to 189 with a mean of 42.62 ± 42.39. Each bone quality assessment method has its advantages and limitations, the details see Table 2. The individual scores of each study are recorded in Table 3. Overall, 4, 52, and 12 of the included studies were rated as “unsatisfactory”, “satisfactory”, and “good”, respectively.

Table 1  Summary of methods of bone quality with important outcomes, advantages, and limitations
Full size table
Table 2  Summary of studies characteristics, patient or specimen demographic details and main findings or summaries
Full size table
Table 3 Quality Assessment of the Studies by the Newcastle–Ottawa Scale
Full size table

Imaging modalities

Imaging modalities for assessing bone quality have various advantages, including non-invasiveness, multiple measurements. Especially, the development of advanced imaging techniques allows the assessment of bone quality at the three-dimensional (3D) microstructure level, such as QCT, micro-CT, high-resolution magnetic resonance imaging (HR-MRI).

X-ray-based modalities

Radiography

Traditional radiography is a cost-effective, widely available method for examining bone geometry, structure, and fracture risk that has been utilized in a wide range of studies [10,11,12,13,14,15,16]. The fracture toughness of bone tissue is highly connected to the bone shape defined by parameters based on plane X-ray radiogrammetry [11]. Furthermore, cortical thickness (CTI) and cortical-medullar index (CMI), as predictors of bone quality, can be obtained from anteroposterior radiographs [10, 15]. Tingart et al. [15] used anteroposterior radiographs to measure the CTI of 19 human cadaver humeri. The results indicated that the CTI of the proximal diaphysis can be a reliable indicator of the bone quality of the proximal humerus. Moreover, the CTI measured by radiographs has a significant positive correlation with BMD evaluated by DEXA. Clavert et al. [10] tested the CMI of 21 cadaveric distal humeri by plain radiographs showed that it is a predictor of the bone quality of the distal humerus and has a significant positive correlation with BMD measured by DEXA and CT-scan (pQCT). Aside from CTI and CMI, the trabecular homogeneity index (THI) was also used to assess bone quality using a plain radiograph, and it shows a strong connection with DEXA and CT-derived data [12]. However, radiography has its limitations, such as low sensitivity, unable to further visualize the microstructure of bone specimens, and that only 2D images are available.

Dual-energy X-ray absorptiometry (DEXA)

DEXA can provide an integrated examination of cortical and trabecular bone, which is frequently practiced in routine practice [10, 17,18,19,20,21]. Choel et al. [17] used 63 mandibular bone specimens to investigate the potential utilization of DEXA for the assessment of bone mineral content (BMC) and BMD prior to implant placement. Furthermore, Hua et al. [18] used 19 mandibular bone samples to evaluate the accuracy of fractal analysis and morphometry measured by DEXA. However, the limitations of the DEXA technique also need to be considered. Yang et al. [19] suggested that BMD measured by DEXA is only one aspect of the complex understanding of bone quality. Tan et al. [20] used 189 human lumbar specimens to verify and quantify the difference in DEXA-BMD between unexplained (in situ) and explanted (in vitro) scans. They found that the in vitro BMDs of the specimens were lower than those of in situ scans. This implied that several factors can affect the accuracy of the DEXA technique, such as the process of preparation, the surrounding soft tissue and their composition, and the scanning conditions. Additionally, Johannesdottir F [21] found that the aBMD of the femoral neck by DEXA (R2 = 0.69) was significantly lower than bone strength measured by QCT-based FEA in predicting femoral failure load.

CT-based modalities

Quantitative computed tomography (QCT)

As a reliable and accurate technique, QCT and HR-pQCT scans have been used in the laboratory and provide us with valuable and comprehensive insight into bone quality [10, 22,23,24,25,26,27,28,29,30,31,32]. A study by Wachter et al. [30] concluded that QCT is a better predictor for the mechanical strength of the intertrochanteric region with objectivity and high precision. Liu et al. [28] reported that microstructural measurements and mechanical characteristics of the distal tibia can be efficiently derived from HR-pQCT images. Also, HR-pQCT is a promising tool for assessing the fracture healing process at the microscale [23].

Briefly, pQCT has emerged as an accurate technique for measuring bone quality with multiple advantages, including measured density-independent of overlying tissue, less susceptible to interference from bone size, relatively safe, higher accuracy, and 3D visualization [26, 31, 32]. Compared to the first-generation HR-pQCT with a nominal isotropic voxel size of 82 μm, the second-generation HR-pQCT has been improved to 61 μm, which allows a more accurate assessment of trabecular thickness (Tb.Th) [24].

Micro-computed tomography (Micro-CT, μCT)

Micro-CT is an advanced imaging modality for quantifying bone quality with high resolution. Currently, it has been gradually applied to assess the bone quality of human bone specimens, with a range of isotropic voxel resolution from 9 to 37 μm [14, 25, 28, 33,34,35,36,37,38,39,40,41,42,43,44,45,46,47]. Micro-CT imaging technique has higher accuracy compared to HR-pQCT, and it is considered as the “gold standard” in bone quality assessment, which allows objective and quantitative evaluation of trabecular bone structure [28, 38, 46]. Moreover, the combination of micro-CT images and mechanical tests can provide valuable and comprehensive information about the microarchitecture and microdamage of human cancellous bone specimens [41, 43]. To explore the influences of osteoporosis and gender on the microstructure of bone grafts, Xie et al. [35] used micro-CT to measure several important microstructure parameters, including bone volume fraction (BV/TV), bone surface density (BS/TV), specific bone surface (BS/BV), trabecular thickness (Tb.Th), trabecular number (Tb.N), trabecular separation (Tb.Sp), structure model index (SMI), and connectivity density (Conn.D.). This non-invasive technique may be sufficient to enhance the evaluation of the bone quality of human bone tissue.

MR-based modalities

The MR-based modalities are a promising tool for evaluating bone morphometry due to their non-invasiveness, and high innate contrast between bone and soft tissue [46, 48,49,50]. Applications of MR-based scanning include nuclear magnetic resonance (NMR), high-resolution MRI (HR-MRI), and micro-MRI (μMRI).

Link et al. [49] compared parameters of the trabecular bone structure obtained from HR-MRI and multi-slice computed tomography (MSCT) with 39 distal radius bone specimens. Their data indicated that structure parameters derived from HR-MRI performed better in the prediction of trabecular bone structure, although this technology is more susceptible to image post-processing. In addition, the distribution and changes of water within bone tissue in relation to bone quality (i.e. bone strength and toughness) were also investigated by nuclear magnetic resonance (NMR) [48]. The findings demonstrated that quantification of mobile and bound water by MR imaging techniques could potentially serve as an indicator of bone quality. Correspondingly, Ni et al. [48] reported that the distribution of bound and free water measured by NMR could be considered as an important factor to determine bone quality. For microstructure analysis of micro-MRI images, Liu et al. [46] validated the 3D model-independent microstructure measurements by micro-MRI and micro-CT. They concluded that the microstructural and mechanical properties of most bone specimens could be efficiently derived from micro-MRI, as well as provide additional information on bone quality.

Mechanical testing methods

Traditional testing

In addition to indirect imaging modalities, traditional mechanical methods can provide an accurate and direct assessment of bone quality at the tissue level, such as structural stiffness, bone strength, elastic modulus, and ultimate stress [51,52,53,54,55]. Several studies quantitatively investigate the changes in microstructure and morphology of human bone specimens using a combination of mechanical testing methods and micro-CT [56,57,58,59,60].

A uniaxial compression test was employed by Kalouche et al. [55] to determine the mechanical characteristics of glenoid cancellous bone in the three planes (axial, coronal, and sagittal). Bayraktar et al. [51] compared the mechanical properties at the tissue level by compression and tensile tests using human trabecular and cortical bone specimens. Charlebois et al. [56] studied 148 bone specimens from different anatomical regions using unconfined and confined compression methods. The data on the behaviour of human trabecular bone at large strain under compression indicated that the influence of tissue fabric would decrease with strain and plays a significant role in the softening behaviour of bone tissue. Furthermore, the application of mechanical loading also allows the microdamage of bone specimens to be studied, which is an aspect of bone quality [59, 60]. Lambers et al. [57] suggested that microdamage has a greater impact on the bone quality of human cancellous bone. Hence, the application of these traditional testing methods can provide more direct data on bone quality at both the tissue and micro-levels when combined with micro-CT.

Indentation testing

Currently, micro-indentation testing can measure bone properties at the millimeter level, and nano-indentation testing has the potential to measure the mechanical properties of bone at the level of trabeculae or osteons. These novel techniques are being used in vitro to evaluate bone quality at various anatomical sites [5, 6, 61,62,63,64,65,66,67].

A study by Dall'Ara et al. [5] concluded that micro-indentation has the ability to distinguish between severely damaged and intact tissue for human vertebral bone tissue. Jenkins et al. [63] claimed that reference point micro-indentation (RPI) can be used as a useful tool for evaluating the mechanical properties of bone in the laboratory. Two studies directly compared RPI with traditional mechanical tests (bending test) [66, 67]. Granke et al. [66] claimed that RPI properties are likely to be influenced by both elastic and plastic behaviour of bone tissue. However, Krege et al. [67] reported that the RPI technique alone is not sufficient to evaluate the mechanical properties of bone.

For nanoindentation, it provides a novel perspective that has been applied to the research on bone materials, especially for volumes as small as lamellae [6, 65]. Albert et al. [6] used nanoindentation to investigate the effects of disease severity (osteogenesis imperfecta) on the local elastic modulus and hardness of bone tissue. The nanoindentation technique makes it possible to investigate the characterization of bone material properties and evaluate modulus and hardness at a smaller scale.

Compositional characterization

As mentioned above, water within bone tissue has a certain effect on bone quality (i.e. bone strength and toughness), which can be measured by NMR. But additionally, compositional characterization (i.e. DBM and organic compositions) is generally acknowledged as being important.

Degree of bone mineralization

The mineralization process consists of a primary deposition of mineral substance on the calcification front, followed by a slow and progressive increase in mineral deposition named secondary mineralization [68]. According to Follet et al. [68], the more mineralized the cancellous bone, the greater the stiffness and compressive strength. Although the increase in DBM may make bone stiffer and more resistant to mechanical loading, too high a mineral to matrix ratio would result in increased brittleness (i.e. higher tendency to crack propagation), and decreased toughness (i.e. the ability to deform without fracturing). Contrastly, this ratio that is too low can lead to bone softening, reduced stiffness and strength. Saito et al. [69] reported that DBM is related to distinct patterns of enzymatic and non-enzymatic cross-links in human bones and is an important element in assessing bone quality.

Organic composition

The ability of bone strength is not only determined by DBM, but also by organic composition (i.e. collagen glycation, collagen cross-links), which has been explored in various studies [47, 70,71,72,73,74,75]. As the major organic interagent, type I collagen is vulnerable to enzymatic and non-enzymatic biomechanical alterations that impact bone quality in numerous ways [47, 73]. The testing data by Poundarik et al. [72] indicated that advanced glycation end products (advanced glycation end products, AGEs) created by non-enzymatic glycation could be used for diagnostic applications in fracture risk assessment. Ural et al. [70] measured total fluorescent AGEs from 96 human cortical bone specimens indicated that AEGs may contribute to bone fragility by altering bone matrix properties. Furthermore, the extent of non-enzymatic glycation (NEG) is linked to alterations in the microarchitecture and microdamage of cancellous bone [73]. Willett et al. [71] used hydrothermal isometric tension (HIT) to measure the collagen’s thermal stability and network connectivity in order to observe the correlation between bone collagen integrity and fracture toughness of cortical bone. They found that the integrity of bone collagen is a critical factor for the fracture toughness of cortical bone. Therefore, the investigation of the bone matrix at the microstructure, and in particular collagen, plays a fundamental role in the mechanical properties of bone tissue at the macroscopic level.

Discussion

The application of each method is closely related to the study design and the outcomes of interest. The X-ray-based imaging methods, including radiography and DEXA, have the advantages of low-cost, low-radiation. However, their limitations make it impossible to provide comprehensive and accurate information on bone quality. For CT-based techniques, including QCT, HR-pQCT, and micro-CT, they can perform 3D image reconstruction and microstructure analysis of human bone specimens, which enables more accurate and comprehensive bone quality information. Many studies have taken micro-CT imaging analysis as the “gold standard” of bone quality assessment [28, 38, 46]. The micro-CT technique can be considered as a comprehensive, high-resolution, three-dimensional, non-invasive technique for the assessment of bone microstructure. Moreover, the development of advanced MRI-based techniques, such as NMR, HR-MRI, micro-MRI, has shown promising results in the assessment of bone structure and water composition, providing additional information [46, 50]. However, most of the advanced imaging techniques described in this review are limited to a minority of laboratories due to expensive equipment and professional operation. Related research and technological breakthroughs need to be explored in order to make these novel imaging techniques to in vivo research and eventually to the clinic.

Compared to indirect imaging techniques, conventional mechanical methods can directly provide the performance of whole bone or bulk tissue specimens. Nevertheless, they are used only for ex vivo bone specimens due to the nature of destruction. However, with the development of indentation techniques, it is possible to directly test bone quality in a minimally invasive manner [5, 62, 66]. This technique has the advantages of being direct, simple, minimally invasive, as well as allowing in vivo testing. The shortcomings are that its results are relatively sole (only tissue hardness and brittleness) [66] and are limited to superficial sites, such as the tibial midshaft. Also, the reliability and significance of the parameters need to be validated further [76]. From our perspective, the combination of imaging modalities and mechanical testing methods would be a good choice for the assessment of bone quality ideally and comprehensively at both micro- and tissue levels.

Furthermore, compositional characterization, both DBM and organic composition, plays an essential role in assessing the mechanical properties of bone tissue and can provide more fundamental information that yields mechanistic insights into affecting bone quality [68, 69, 71]. Especially for bone collagen, there is a significant association with clinically relevant bone diseases, such as osteoporosis, osteogenesis imperfecta, and diabetes-related diseases [77, 78].

Previous studies have reported that collagen content in human bones reaches a maximum during adolescence and gradually decreases thereafter with aging [77, 79]. Compared with age-matched healthy subjects, osteoporotic bone indicated that reductions in the enzymatic cross-links and an increase in AGEs cross-links in bone [69, 80]. In diabetic bone tissue, BMD may be normal, but bone strength has decreased, which correlates with the increased formation of AGEs [77, 81]. For osteogenesis imperfecta, it is also a disease closely associated with collagen, and studies have shown that the orientation of collagen is highly disordered and that the collagen-mineral particle network is profoundly altered [78]. Therefore, the alteration of bone quality and biomechanical performances is the macroscopic result of a sequence of composition and microstructural events.

There are several limitations to our systematic review. First, not all test methods and studies were summarized in our review, which is a limitation of all systematic reviews. In this article, “bone quality” and “human bone specimens” were used as search terms. Actually, “bone quality” is not universally defined, and there are several other interchangeable phrases used, including “bone material quality”, “bone matrix quality”, etc. Similarly, the search term “human bone specimens” is interchangeable with “human bone samples”. Due to the limitation of content, it is difficult, or even almost impossible, to use all possible search terms in a review.

In order to further expand the search, the “similar articles” option of PubMed and the references for main articles were used in this review. Second, there is the risk of selection bias since the presence of heterogeneous. Finally, this review focuses only on imaging techniques, mechanical testing methods, and the effects of compositional characterization, computational techniques such as FEA are not included.

Conclusions

Advanced techniques are playing an increasingly important role due to their multiple advantages, focusing on the assessment of bone morphology and microarchitecture. Non-invasive imaging modalities and mechanical testing techniques, as well as the assessment of bone composition, need to complement each other in order to provide comprehensive and ideal information on the bone quality of human bone specimens.