Slipped capital femoral epiphysis is a rare event in adolescents that, if left untreated, can lead to serious consequences, including osteonecrosis of the femoral head, chondrolysis, and premature degenerative arthritis1-4. The problems are attributable, in part, to a growth plate disruption in which the epiphysis separates from the metaphysis. The etiology and pathogenesis of the condition are poorly understood. Obesity and mechanical factors are found in most, but not all, cases5,6. Endocrine abnormalities such as hypothyroidism, renal osteodystrophy, hypogonadism, and others have also been linked to the condition7-12. Despite these findings, a majority of cases are labeled as idiopathic and occur in otherwise healthy children12,13. It is likely that the etiology of slipped capital femoral epiphysis is multifactorial.
Previous light and electron microscopic studies of the physis of patients with slipped capital femoral epiphysis have demonstrated distortion of the normal architecture and organization as well as chondrocytes that are disordered in their arrangement and frequently clustered2,14,15. Collagen fibrils are reported to be reduced in number, randomly oriented, and widely scattered in their extracellular distribution as well as defective in periodic banding2,14,15. The proliferating and hypertrophic zones appear to be the most severely affected2,14,15. Recent work has demonstrated that the chondrocytes from the zone of hypertrophy undergo earlier apoptosis when compared with normal chondrocytes16. However, we are not aware of any investigations of gene expression of specific messenger RNAs (mRNAs) in the femoral growth plate chondrocytes of individuals with a slip and its relationship to a possible change in protein production that may contribute to weakening of the physis. We examined the hypothesis that slipped capital femoral epiphysis is the result of molecular level decreases in or abnormal expression of genes critical to proper growth plate formation. By applying laser capture microdissection followed by quantitative reverse transcription-polymerase chain reaction analysis of mRNA17-28, we attempted to define gene expression levels of type-II collagen and aggrecan, two principal molecules critical to growth plate development, in patients with slipped capital femoral epiphysis.
Laser capture microdissection allows the isolation of individual cells or clusters of cells from sections of tissues17-21. Once isolated, the cells can be evaluated for gene expression with RNA analysis. The laser capture microdissection technique is not limited by cell type and has been shown to be especially valuable for examining many different tissues, including a number that are pathological in nature22-30. Cartilaginous epiphyseal growth plates have been documented with the technique in only a few studies31-35. However, such tissues are ideal for the utilization of laser capture microdissection because of the defined nature of their constituent chondrocytes and a resultant zonal organization based on the phenotype of these cells. For the evaluation of epiphyseal plates, laser capture microdissection offers an approach with which an investigator can identify a specific cell or region of cells that can be physically isolated from an intact cartilage specimen and studied further in molecular biological detail31-35.
Type-II collagen is the most abundant collagen in growth plate cartilage, where it provides shape and tensile strength to the tissue2,36. Prior studies have revealed that the abnormal collagen fibrils in slipped capital femoral epiphysis2,14,15 could be the source of growth plate weakening, as their structure and organization are related to the biomechanical properties of the tissue. Fibril organization (the number and distribution of fibrils) is directed in part by collagen gene expression levels, so it is possible that an unusual level of type-II-collagen expression in slipped capital femoral epiphysis would lead to the observed weakened state characterizing this condition.
Aggrecan, the most abundant proteoglycan in cartilage, confers a capacity for weight-bearing and compressive resilience by maintaining tissue hydration through osmotic swelling36. Studies of mice have suggested that abnormalities in the aggrecan gene can also lead to growth plate defects6, but a link between aggrecan and the pathogenesis of slipped capital femoral epiphysis has not been established, to our knowledge.
By determining the expression of type-II collagen and aggrecan, or any additional genes that may mediate the development of slipped capital femoral epiphysis, quantitative reverse transcription-polymerase chain reaction analysis allows comparison of samples from affected patients with samples of control tissues. Except under unusual circumstances, normal proximal femoral growth plate cartilage cannot be harvested surgically from patients with slipped capital femoral epiphysis or other patients. As a consequence of this critical limitation, the expression of several genes was measured in the cartilaginous growth plates of the distal part of the femur and proximal parts of the tibia and fibula of patients without slipped capital femoral epiphysis who had undergone a corrective epiphysiodesis. The values were compared with the levels of the same genes found in the proximal femoral growth plates of patients with slipped capital femoral epiphysis.
This work was conducted under human protocol approval by the institutional review boards of both the participating hospital system and the participating medical school. A core biopsy of the proximal femoral growth plate was performed on nine different patients undergoing surgical repair of a slipped capital femoral epiphysis with use of the bone-graft epiphysiodesis technique5. The core biopsies were part of this routine surgical procedure, and the specimens would normally have been discarded. Five specimens were obtained as controls from four age-matched patients without slipped capital femoral epiphysis who were undergoing distal femoral and proximal tibial and fibular epiphysiodesis procedures to treat limb-length inequality. Following surgical removal of the tissues, samples were immediately placed in RNAlater (Ambion, Austin, Texas) to maintain RNA integrity and then were divided into two segments. One segment was immersed in 10% neutral buffered formalin for further processing into paraffin. The other segment was transported on dry ice to the participating medical school for storage at -20°C and subsequent molecular analysis.
The paraffinized samples were cut into approximately 5-µm-thick sections, mounted on glass slides, and stained with hematoxylin and eosin or toluidine blue for analysis of general tissue morphology, with safranin O to detect the presence of proteoglycan, and with picrosirius red for a qualitative measure of collagen. Sections were examined on an Olympus IX70 light microscope (Olympus America, Melville, New York) and were photographed with use of an Olympus MicroSuite system (Olympus America) for image recording. Montages of image sections were constructed with use of the MicroSuite software.
Initially, one specimen (from the proximal part of the femur) from a patient with slipped capital femoral epiphysis and one specimen (from the distal part of the femur) from a patient without slipped capital femoral epiphysis were decalcified and were separately ground to powder under liquid nitrogen in a freezer/grinder mill (model 6750; Spex, Metuchen, New Jersey). Total RNA was isolated according to the manufacturer's protocol for TRI Reagent (Molecular Research Center, Cincinnati, Ohio). Total RNA was DNase-treated, and its concentration and purity were determined subsequently with ultraviolet spectrophotometric readings at 240, 260, 280, and 320 nm. On the basis of the ultraviolet spectrophotometric readings at 260 nm, 1 µg of each specimen (one from the patient with and one from the patient without slipped capital femoral epiphysis) was reverse-transcribed. The resulting cDNA pools were utilized in a human endogenous control plate (Applied Biosystems, Foster City, California) to identify the most appropriate reference, or housekeeping, gene for this study (more specifically, one that would differ, between the control and pathological samples, by less than twofold). The control plate contained eleven potential candidate genes and one internal plate control. Instructions provided by the manufacturer were followed in the cDNA analysis of the control sample and the sample from the patient with slipped capital femoral epiphysis. Two major genes, type-II collagen and aggrecan, regulating cartilage extracellular matrix formation were also screened with these sample pools to determine if there were significant differences in their levels of expression between the sample from the patient with slipped capital femoral epiphysis and that from the patient without slipped capital femoral epiphysis. A more cell-specific analysis employing the laser capture microdissection technique was performed following the initial screening to verify any key results that had been detected.
The remaining eight frozen specimens (proximal femoral) from eight patients with slipped capital femoral epiphysis and four control specimens (distal femoral and proximal tibial and fibular) from three patients without slipped capital femoral epiphysis were removed from low-temperature storage and embedded in tissue-freezing medium (TFM, Triangle Biomedical Sciences, Durham, North Carolina) inside a cryostat maintained at -25°C. Because of the small amount of individual surgical tissue in certain instances and to enhance subsequent gene expression analysis, some samples from the patients without slipped capital femoral epiphysis were embedded together in single TFM blocks. Sections of 5 µm in thickness were cut and were mounted on glass slides without use of a coverslip. The slides with the sections were fixed for one minute in 70% ethanol, stained with eosin, and subsequently dehydrated through an alcohol gradient. These treatments have been documented to preserve specimen RNA31,34. Sections of tissue on slides without coverslips were observed in a PixCell II laser capture microdissection system (Arcturus Engineering, Mountain View, California), and cartilage chondrocytes were identified for subsequent isolation and analysis. Approximately fifty to 200 chondrocytes were captured with use of a calibrated beam that approximated the diameter of a single cell34. As described previously34, chondrocytes were isolated from tissue sections, bonded to thin polymer-film substrates of laser capture microdissection caps, and treated with lysis buffer.
Total RNA from captured chondrocytes was extracted, DNase-treated, and reverse-transcribed according to published procedures31,34. Human-specific TaqMan primer and probe sets for type-II collagen and aggrecan were purchased from Applied Biosystems. 18S rRNA (ribosomal RNA) was used as an endogenous control, on the basis of the results of the human endogenous control plate analysis (Applied Biosystems). Primers for 18S rRNA were designed with use of Primer Express software (Applied Biosystems). The following 18S rRNA primer sequences were used: forward: AACGAGACTCTGGCATGCTAACTA, reverse: CCACTTGTCCCTCTAAGA.
Gene expression was analyzed with an ABI PRISM 7700 Sequence Detection System (Applied Biosystems). Each sample for quantitative reverse transcription-polymerase chain reaction analysis for 18S rRNA contained SYBR Green Master Mix (Applied Biosystems), 0.3 µM of each primer, and cDNA. Appropriate negative controls containing no reverse transcriptase and buffer blanks were added to confirm the absence of genomic DNA or DNA contamination. Dissociation curves were generated with use of the software program v1.Ob.1 (Applied Biosystems) to verify the correct melting temperatures for the amplicon products34. Instructions from the manufacturer were followed for the TaqMan primer and probe sets for type-II collagen and aggrecan.
Standard curves were constructed in this study from serial dilutions of stock RNA isolated from human articular chondrocytes grown to confluence in vitro. Triplicate threshold cycle values were produced for each serial dilution and sample. With use of the methodology for relative standard curves37, plots of log input versus threshold cycle for each gene were obtained. Expression levels of type-II collagen and aggrecan were normalized to 18S rRNA, which was utilized as the reference gene. The averaged normalized amount of target gene expression was then divided by the averaged control expression for each gene. The control samples as the calibrator were set equal to the value of 1, and the fold changes in the samples from the patients with slipped capital femoral epiphysis as compared with the control samples were plotted with their respective standard errors of the mean34. Gene expression results were analyzed statistically with an independent-samples t test (SPSS, Chicago, Illinois), and p values were recorded. Analysis was based on a sample size of eight for the pathological specimens and four for the counterpart controls. A p value of <0.05 was considered significant.
Source of Funding
The funding for this study was used to pay for technician time and effort in performing preparation of specimens as well as laser capture microdissection and quantitative reverse transcription-polymerase chain reaction analysis. Funding was also utilized for the purchase of reagents, TaqMan primer and probe sets, endogenous control plates, chemicals, and related disposable supplies and consumable products.
Figure 1, A, presents a histological reference as an image montage of sections of a portion of a normal femoral head obtained during surgery from a patient without slipped capital femoral epiphysis. For comparison, Figure 1, B, shows a montage of sections of a region of the femoral growth plate from a patient with slipped capital femoral epiphysis. The two patients were age-matched, and the figures illustrate the location and specific regions and cells examined in this study. Figure 1, A, demonstrates the site of articular cartilage far removed from its underlying growth plate (circled) where, in the core-biopsy specimens from the patients with slipped capital femoral epiphysis, constituent proliferating and hypertrophic chondrocytes were isolated by laser capture microdissection for gene expression analysis. Thus, only growth plate cartilage from the patients with or without slipped capital femoral epiphysis was investigated in this study. The growth plate from a patient with slipped capital femoral epiphysis can be readily distinguished by the dark blue staining separating the secondary center of ossification and the subchondral bone of the specimen (Fig. 1, B).
Figure 2 shows various images representative of the cartilaginous epiphyseal growth plates from a patient who underwent an epiphysiodesis (in the proximal part of the tibia) and from individuals with slipped capital femoral epiphysis. The two patients represented in Figure 2, A through D, were age-matched, and the tissues shown in Figure 2, A and C, served as controls for those shown in Fig. 2, B and D. The images in Figure 2 show the results of selective section staining with safranin O (Fig. 2, A and B), picrosirius red (Fig. 2, C and D), and hematoxylin and eosin (Fig. 2, E and F). The deep red staining for proteoglycan across the entire growth plate of a typical control section (Fig. 2, A) is in direct contrast to the paucity and nonuniformity in proteoglycan staining intensity observed in a section from a patient with slipped capital femoral epiphysis (Fig. 2, B). Figure 2, C and D, shows sections (corresponding to those shown in Figure 2, A and B) stained with picrosirius red, which specifically identifies collagen. These sections show the tissue from the patient with slipped capital femoral epiphysis (Fig. 2, D) to be clearly different from the control tissue (Fig. 2, C) with regard to the intensity and uniformity of the picrosirius-red staining, results indicative of a reduced number or disordered extracellular arrangement of collagen fibrils in slipped capital femoral epiphysis.
A section of growth plate cartilage from a second patient with slipped capital femoral epiphysis is shown in Figure 2, E, and enlarged in Figure 2, F. Typical of these specimens, this tissue region was markedly different from the control and normal growth-plate cartilage in terms of the number and arrangement of its composite proliferating and hypertrophic chondrocytes. Notably, these cells were not well organized; in particular, they were not in long, parallel columns, as they are in normally developing growth plates.
Figure 3, A, B, and C, documents the isolation, by laser capture microdissection, of chondrocytes from representative fragments of control tissue (from the distal part of the femur and the proximal parts of the tibia and fibula) removed during an epiphysiodesis in a patient without slipped capital femoral epiphysis. From a section from this sample on a glass slide (Fig. 3, A), areas of proliferating and hypertrophic chondrocytes were located (Fig. 3, B), and these cells were then captured (Fig. 3, C) for subsequent total-RNA isolation and quantitative reverse transcription-polymerase chain reaction analysis. Figure 3, D, E, and F, shows isolation by laser capture microdissection of a group of chondrocytes from the proximal growth plate in the normal femoral head depicted in Figure 1, A. To make this series of images, a portion of the femoral head was embedded and frozen on the chuck in a cryostat (Fig. 3, D1). Section enlargement (Fig. 3, D2) identifies the region of interest within the growth plate designated for laser capture microdissection, and such areas served as reference sites for the capture of chondrocytes from samples from the patients with slipped capital femoral epiphysis. Two histological images (Fig. 3, E and F) document stages in the laser microdissection of a frozen and eosin-stained section of a specimen from a patient without slipped capital femoral epiphysis. Here, a region of growth plate cartilage is shown before and after laser microdissection of numerous chondrocytes (Figs. 3, E, and 3, F, respectively). The identical approach was used to capture growth plate chondrocytes from samples from the patients with slipped capital femoral epiphysis (not illustrated because these fragile sections were typically disrupted following the performance of the technique).
Results of initial screening with a human endogenous control plate are shown in Figure 4. In general, there was a less than two-threshold-cycle difference (a fourfold difference in gene expression) in any individual candidate gene, except for huTfR (transferrin receptor), between the proximal femoral tissue from the patient with slipped capital femoral epiphysis and the distal femoral tissue from the patient without slipped capital femoral epiphysis (Fig. 4, A). The plate also showed that the housekeeping genes of interest varied with regard to their threshold cycles, and three genes (huCYC [cyclophilin], huHPRT [hypoxanthine ribosyl transferase], and huTBP [transcription factor IID]) were detected at threshold cycles of >30. The latter result is indicative of the low abundance of each particular gene in the whole ground-to-powder samples, a consequence of which is that the same gene would be even less abundant and therefore inappropriate for use in cells isolated by laser capture microdissection. The specimen variability measured in this analysis may be the result of inaccuracies inherent in the ultraviolet spectrophotometric readings at 260 nm and/or natural differences in the expression of the genes. Regardless, the choice of a reference gene for application to any quantitative reverse transcription-polymerase chain reaction analysis is ideally one that varies minimally (by less than one threshold cycle) between the pathological state of the tissue and the counterpart control. On the basis of this criterion, 18S rRNA was selected as the normalizer for the genes in this study (Fig. 4, A), and its effect is evident in Figure 4, B, where concentration differences between the different endogenous genes have been normalized. In this figure, most of these constitutively expressed genes depict very limited differences (less than one threshold cycle) or small differences (one to two threshold cycles), which represent a two to fourfold statistical difference in expression between the pathological and control samples (Fig. 4, B). The threshold-cycle difference exceeds two for only two of ten normalized genes (huB2m [ß2-microglobulin] and huTfR).
Following determination of 18S rRNA as the most appropriate housekeeping gene, cDNA pools were utilized for quantitative reverse transcription-polymerase chain reaction with aggrecan and type-II-collagen primer sets for the single specimen from the patient with slipped capital femoral epiphysis and the single control specimen described above. The results listed in Table I show marked downregulation of these genes on analysis with the comparative threshold-cycle methodology37. The fold change in expression of type-II collagen compared with the control value was 0.006, and the fold change in the expression of aggrecan compared with the control value was 0.009.
Type-II-collagen and aggrecan gene expression levels measured, after laser capture microdissection, with quantitative reverse transcription-polymerase chain reaction analysis of the remaining eight specimens from eight patients with slipped capital femoral epiphysis and the remaining four specimens from three patients without slipped capital femoral epiphysis are shown in Figure 5. Figure 5 provides a fold-difference plot of the levels of expression of type-II collagen and aggrecan in the samples from the patients with slipped capital femoral epiphysis relative to the control samples. This figure shows the normalized average values and standard errors of the mean for each of the four groups of specimens calibrated to the respective control samples, set to the value of 1.0, according to standard curve methodology37 for comparing gene expression measures. In Figure 5, the fold differences between the specimens from the patients with slipped capital femoral epiphysis and those from the patients without the disorder were 13.7% ± 0.2% for type-II collagen (p < 0.05) and 26% ± 0.2% for aggrecan (p < 0.05). Thus, on the basis of selected chondrocyte isolation by laser capture microdissection and subsequent quantitation, it appears that the gene expression levels of type-II collagen and aggrecan are significantly downregulated in proximal femoral samples from patients with slipped capital femoral epiphysis as compared with samples of combined material from the distal part of the femur and proximal parts of the tibia and fibula from patients without slipped capital femoral epiphysis. This result is consistent with that derived from the analysis, performed with the comparative threshold-cycle approach, of the whole ground-to-powder specimen from the proximal part of the femur of the patient with slipped capital femoral epiphysis and from the distal part of the femur of the patient without slipped capital femoral epiphysis described above (Table I).
Little is known about the expression of genes that may mediate and impact growth plate formation and development in the pathogenesis of slipped capital femoral epiphysis. Indeed, the cause or causes of this serious problem are basically only conjectural at this time. Utilizing laser capture microdissection followed by mRNA analysis with quantitative reverse transcription-polymerase chain reaction, we measured type-II-collagen and aggrecan gene expression in physeal chondrocytes from human subjects with slipped capital femoral epiphysis in an attempt to understand the etiology of this pathological condition.
This study has several important limitations, including the small number of samples from patients with slipped capital femoral epiphysis. The sample size was small because of the relatively few patients with slipped capital femoral epiphysis who are treated with bone-graft epiphysiodesis. Despite the few specimens, p values were calculated by means that are appropriate for these sample sizes.
There is also some concern about our use of specimens obtained during distal femoral and proximal tibial and fibular epiphysiodeses as control tissues in this study. The validity of using physes from different donors and from different anatomical locations for comparison with samples from the sites of slipped capital femoral epiphyses may be questioned. To be sure, aside from the differences between individual patients, there are clear differences between the loading characteristics of the proximal part of the femur and those of the distal femoral and proximal tibial and fibular physes. Fishkin et al. recently used finite element analysis to demonstrate that increasing shear strain across the proximal femoral physis can result in slipping as a result of increased weight, femoral retroversion, and varus load38. However, distal femoral physes and proximal tibial and fibular physes obtained in another study were not found to be exposed to such shear forces, and it is not known what effect these loading factors have on gene expression in the proximal part of the femur as compared with that in the distal part of the femur or the proximal part of the tibia (or fibula)39. Furthermore, it has been shown that increasing strain on cartilage can induce the apoptotic cascade40. In addition, baseline differences in gene expression between the physes have not yet been defined. Finally, histological studies have shown that the different growth plates appear identical but molecular level parameters characterizing the respective plates are poorly understood2,15. Thus, some variability in mechanical parameters and structural features has been documented and there may be other such changes, as well as those of a molecular, biochemical, or biological nature, between the proximal and distal growth plate cartilage from either normal femora or those obtained from patients with slipped capital femoral epiphysis. Likewise, there may be such variability between these tissues and proximal tibial and fibular growth plate cartilage from patients without slipped capital femoral epiphysis.
A human endogenous control plate analysis was undertaken in the present study to determine whether a reference gene could be identified and used to compare distal femoral growth plate cartilage from a patient without slipped capital femoral epiphysis with proximal femoral growth plate cartilage from a patient with slipped capital femoral epiphysis. The endogenous control plate results for these two specimens were similar. For eight of the ten different housekeeping genes common to the growth plate tissues obtained from both the patient with and the patient without slipped capital femoral epiphysis, there was no more than a two-threshold-cycle difference in expression between the pathological and control tissues. There was an approximately three-threshold-cycle difference in the remaining two genes of interest. Thus, these paired variations in the genes, all of which are known to be expressed constitutively for maintenance of cellular activities41, appear to be relatively minor, particularly in view of the facts that tissue from the site of a slipped capital femoral epiphysis is pathological and that it is known that there are mechanical and structural differences between the proximal and distal parts of the femur38-40. The data from the two whole ground-to-powder samples suggest that it is acceptable to use distal femoral growth plate cartilage at least for constitutively expressed gene comparisons with proximal femoral growth plate cartilage, regardless of whether the proximal femoral growth plate cartilage is affected or not by the condition of slipped capital femoral epiphysis.
Thus, on the basis of the conclusion drawn from the endogenous control plate analysis that the sampling and the control tissues were appropriate and could be normalized to 18S rRNA, subsequent quantitative data were derived from other ground-to-powder samples from the proximal part of the femur of the patient with slipped capital femoral epiphysis and from the distal part of the femur of the patient without slipped capital femoral epiphysis. Levels of expression of type-II collagen and aggrecan in the tissues from the patient with slipped capital femoral epiphysis were found to be remarkably downregulated, with fold changes of 0.006 and 0.009, respectively, compared with the control levels.
The findings from the powdered specimens supported those of the analysis of type-II-collagen and aggrecan gene expression in chondrocytes isolated by laser capture microdissection from additional individuals with and without slipped capital femoral epiphysis. In the latter analysis, since only small amounts of individual surgical tissue were available in certain instances, and to enhance the gene expression analysis following the laser capture microdissection, some samples from the patients without slipped capital femoral epiphysis were pooled and consisted of material principally from the distal part of the femur with minor contributions from the proximal parts of the tibia and fibula. Statistical analysis of such mixed control tissues yielded small errors of the mean for both type-II-collagen and aggrecan gene expression, normalized to 18S rRNA. Interestingly and importantly, laser capture microdissection and quantitative reverse transcription-polymerase chain reaction analysis of the specimens from the patients with slipped capital femoral epiphysis demonstrated expression levels of type-II collagen and aggrecan mRNA that were 13.7% ± 0.2% and 26% ± 0.2%, respectively, of those of the pooled tissue from their counterpart age-matched controls.
Thus, analyses of both whole ground-to-powder specimens and the sections of tissue providing laser-captured chondrocytes revealed consistent results, demonstrating that both type-II-collagen and aggrecan gene expression, each normalized to 18S rRNA, were significantly lower in growth plate samples from patients with slipped capital femoral epiphysis compared with samples from patients without slipped capital femoral epiphysis. These downregulated expression levels in the specimens from the patients with slipped capital femoral epiphysis, together with the data showing that the pathological and control specimens had similar levels of ten of eleven housekeeping genes, suggest that these changes in type-II-collagen and aggrecan expression are real and not the consequence of variations between samples, differences in biopsy sites, or other factors.
From this study, it is impossible to determine whether these abnormalities in type-II-collagen and aggrecan gene expression levels are the cause or result of slipped capital femoral epiphysis. In other words, it is not clear whether gene expression abnormalities of the chondrocytes result in abnormal production of structural proteins and thus a weakened growth plate that is prone to slip or whether a fractured growth plate causes chondrocyte destruction and thus decreased gene expression and protein production. Adamczyk et al. showed increased apoptosis throughout the proximal femoral growth plates of patients with slipped capital femoral epiphysis, a result suggesting that the growth plate is abnormal with respect to the programmed death of its constituent cells16. However, those data, like those reported here, do not resolve the question of whether changes in apoptotic proteins, or their genes and other genes, are the cause or effect of slipped capital femoral epiphysis.
This study demonstrated gene expression and histological changes of the proximal femoral growth plate in patients with slipped capital femoral epiphysis that could adversely affect the intrinsic strength of the growth plate and its ability to resist displacement. The biological implications and relevance of these changes have yet to be determined. 