34 Image Analysis Histomorphometry Stereology
Jargon Simplified: Stereology
Stereology is a method to get unbiased quantitative estimation of the first-order stereological parameters in a three-dimensional sample from measurements made on two-dimensional planar sections.
Many systems and techniques have been developed for quantitative analyses in orthopedic research. Bioquant systems (Nashville, TN) are widely used in the field of orthopedic research. Bioquant′s Osteo System is ideal for analyzing bone structure, osteoblast activity, and osteoclast activity. Detailed instruction about the Bioquant systems can be found on the company′s website at http://www.bioquant.com/. For researchers who are interested in fracture repair and bone regeneration, recently developed stereology technique may have advantages over the Bioquant systems. Stereology is a method to get unbiased quantitative estimation of the first-order stereological parameters in a three-dimensional sample from measurements made on two-dimensional planar sections. These parameters include volume, surface area, length, and number. Stereology allows us to quantify not only the amount of tissue formation, but also many of the regulating mechanisms that underlie bone repair. For example, tissue vascularization and infiltration of inflammatory cells can be accurately measured using stereology technique. In this chapter, we describe our experience of using stereology to quantify fracture healing, tissue vascularization, and inflammation.
34.1 Basic Rules of Stereology
To achieve unbiased estimation in stereology, certain rules have to be followed through the whole process of the assay. The number of samples, the time points of tissue collection, the parameters to be analyzed, and appropriate sampling need to be decided before the experiment even starts. A pilot study will provide information about the variation of the parameters, which can be used to determine the sample size (power analysis) and the adequacy of sampling (coefficient of error). The following are some basic rules of stereology.
34.1.1 Comparable Reference Space
Reference space is the anatomical region in the tissue that contains all the objects of interest. Reference space must be unambiguously defined, and 100% of it should be available for analysis in each sample. In addition, the reference spaces of the control group and the treated group in one experiment need to be comparable. For later stages of fracture healing (i.e., 7 days or more after injury in a murine tibia fracture model 1, 2), the callus usually has a clear margin and it is reasonable to use the whole callus as the reference space. For early fracture healing (i.e., before 5 days after injury in a murine tibia fracture model1,2), it may be difficult to outline the callus because it has no clear margin yet. For samples collected early after fracture, the reference space needs to be carefully defined. One method is to use the whole tibia with all the surrounding soft tissues as the reference space. An alternative approach is to add permanent markers to the tissue before creating bone fracture. The markers will be used to define the reference space. 3 This second approach will be discussed in more detail later.
34.1.2 Isotropic Uniform Random and Vertical Uniform Random Sections
Two types of sections can be used for stereology (Fig. 34.1). Isotropic uniform random sections are sections with complete randomness. On the contrary, vertical uniform random (VUR) sections are prepared by randomly rotating the sample around an axis before embedding. In fracture studies, this axis can be the long bone that is fractured. VUR sections are less random compared to isotropic uniform random sections and are good for tissues with defined orientations. Fracture callus, bone, and cartilage in the callus do not have defined orientation, and the estimation of their volume does not require random sectioning. Blood vessels, however, may have defined orientation, and isotropic uniform random or VUR sections are required for their quantification.
34.1.3 Random Systemic Sampling
This is a central role of stereology. All tissues of interest in the reference space should have the same probability to get analyzed. To achieve this, random systemic sampling is performed, which means a starting point is chosen at random from the sampling frame and selections thereafter are at regular intervals (Fig. 34.1). The intervals between selected sections in one sample have to be consistent; however, a different sample can have different intervals between its selected sections.
34.1.4 Adequate Sampling
Adequate and efficient sampling is required to capture most of the biological variation in a parameter and to use your time and effort productively. Adequate sampling depends on the heterogeneity of the tissue and the amount of the object of interest in the reference space. It is easy to appreciate that quantification of tissues of high heterogeneity or counting rare events will need a higher fraction of sampling. The adequacy of sampling is measured by the coefficient of error, which is defined as the standard deviation divided by the mean.
As a rule of thumb, analysis of 6 to 10 levels (sections) and counting at least 200 intersections or points of the object of interest are usually enough. To guarantee adequate sampling, researchers may rely on a reasonable degree of oversampling by analyzing at least eight sections or more than 500 counts for each sample.
34.1.5 Choose the Right Probe
Many probes have been developed to analyze different parameters in stereology. One rule of probe selection is that the number of dimensions in the parameter of interest and the probe must have a sum of 3. For example, to estimate the number of cells, which is a zero-dimensional parameter, three-dimensional probes like a physical dissector or optical dissector should be used. To estimate surface area, a two-dimensional parameter, a one-dimensional line probe is used. For length of blood vessels, a two-dimensional planar probe is used as length is a one-dimensional parameter. For total volume, which is a three-dimensional parameter itself, a zero-dimensional point grid will be appropriate.
34.1.6 Follow the Same Protocol of Tissue Processing
Tissue processing has significant effects on the tissues. For example, dehydration steps in paraffin section preparation significantly shrink the tissue. Compared to frozen sectioning, which does not require dehydration, paraffin sectioning gives a lower reading of tissue volume. Tissue processing also affects the quality of histological staining, immunohistochemistry, and molecular analyses. For one analysis in one study, the same protocol of tissue processing, including harvesting, fixation, decalcification, embedding, sectioning, and staining, should be followed. For fracture studying, paraffin sections can be used to quantify the volume of callus, bone, or cartilage. To analyze fluorescent cells or perform immunostaining, frozen sections are usually better.
34.2 Stereology in Research on Bone Regeneration
Stereology is a valuable tool in orthopedic research. It is used to estimate volume. More importantly, stereology allows the unbiased quantification of cells including inflammatory cells and stem cells and tissue vascularization. Based on an Olympus CAST system (Center Valley, PA) and Visiopharm software (Hoersholm, Denmark), we have developed protocols to estimate the volume of tissue, number of cells, and length and surface area of blood vessels during fracture healing. The animal model we used is a murine tibia fracture model.
34.2.1 Creation of Tibia Fractures
Mice are anesthetized and a transverse fracture is created by three-point bending at the midshaft of tibia. Fractures are either left unstabilized or stabilized with a custom-made external fixator.2 To apply the external fixator, the proximal and distal metaphyses of the tibia are transfixed using two insect pins each, which are oriented perpendicular to the long axis of the tibia. Two rings are secured to the pins and then bridged with three threaded bars (Fig. 34.2). This external fixator provides rigid stability to the fracture. Animals are allowed to ambulate freely after recovery.
34.2.2 Tissue Processing
Animals with tibia fracture are euthanized at day 3 to 28 after injury. The whole tibia from the knee joint to the ankle is collected, skinned, and fixed in 4% paraformaldehyde for 24 hours at 4°C. Muscles are kept intact. Tissues are then decalcified in 19% ethylenediaminetetraacetic acid for 2 weeks at 4°C. Complete decalcification is confirmed by X-ray.
34.2.3 Estimation of Volume
Collect Systemic Random Sections
To estimate the volume of a tissue of interest in fracture callus, systemic random sections are collected. Either frozen or paraffin sections can be used. If immunohisto-chemistry is not part of the analysis, we routinely choose paraffin sections for their easy handling and better quality of histology. Longitudinal sections are used to estimate the volume of callus, bone, and cartilage within the callus. Section thickness is 10 µm or less. A nonstabilized tibia fracture normally has a large callus and 200 to 400 serial sections can be collected for each callus.1,3, 4, 5 Therefore, we choose an interval of 30 sections (equals to 300 µm if each section is 10 µm thick) between two systemic randomly selected sections for histomorphometry. This setting allows us to analyze at least six sections for each sample. Fig. 34.1 illustrates the strategy of random systemic sampling. Before the start of sectioning, a random number between 1 and 30 is assigned to each sample. For example, sample X is assigned a random number of 15. The tissue block of sample X is mounted on the microtome. The section thickness is set to 10 µm, and cutting is started. Sections are counted starting at the first cut, even if there is no tissue. The researcher then checks whether there is callus tissue on the 15th section. If there is callus tissue, this section will be collected as the first section for histomorphometry, the 45th section will be second, the 75th the third section, and so on. If there is no callus tissue on the 15th section, the researcher will need to continue cutting to the 45th section. The presence of callus tissue on the 45th section is then checked again, and this will be the first section to be collected if there is callus tissue. If the 45th section still has no callus tissue, the researcher will need to move on to the 75th section. After the first section with callus tissue is collected, every 30th section will be collected throughout the whole callus tissue.
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
