Bone histopathology: the effects of pathology on bone microstructure and implications for histological age estimation

Abstract

Age estimation is a vital part of the creation of a biological profile for the purposes of identifying unknown individuals. Numerous macroscopic methods of analysing bone have been developed for this purpose. In some cases, however, fragmentation of the bone, poor preservation, and taphonomic damage will not allow for the implementation of these methodologies. For these cases, microscopic analytical methods have become the standard, and in the past several decades methodologies using different parts of the skeleton have been developed, most notably those using the rib for age estimation. These histological age estimation methods rely upon the measurements for cortical area and osteon population density (OPD) for an accurate age estimation. While these population specific methods have proven accurate and useful for the purposes of age estimation, they do not consider the health of the individual whose age they are attempting to estimate. Chronic pathology can change the microstructure of osseous tissue, which in turn could influence the accuracy of histological age estimation methods. The purpose of this study is twofold. First, to examine the microstructure of chronically ill individuals and provide detailed descriptions of microstructural changes for visual identification of pathology. Second, to quantitatively evaluate pathological samples using both standard measurements used in most age estimation calculations, as well as detailed osteon specific histomorphometrics to measure the specific changes to Haversian systems in cases of chronic pathology. By considering both the qualitative and quantitative characteristics of pathological bone in comparison with healthy controls it will be possible to determine what the effects of pathologically affected bone on methods of age estimation could be. Pathological ribs (N=40) were obtained via autopsy from two populations (Albania and Crete), and from a modern skeletal collection (Crete). These were then compared with healthy ribs (N=40) obtained from autopsy (Crete) and from two modern skeletal collections (Crete and Cyprus). The mid-shafts were then prepared using standard protocols for histological methods. First, samples were assessed qualitatively. Samples were visually evaluated for abnormal microstructural features, and these features were then categorised. Samples were then divided and tested, using these pathological feature categories, by disease class and by age group. Second, samples were assessed quantitatively. Thin sections were first measured using the standard histological metrics of cortical area, OPD, number of complete osteons, number of fragmented osteons, and total number of osteons. Thin sections were then evaluated using a sampling method for ten osteon specific metrics (osteon maximum length, osteon minimum length, osteon circumference, osteon total area, osteon area, osteon circularity, canal area, canal maximum diameter, canal minimum diameter, and canal circularity). Assessment for qualitative features was done by identifying abnormal features of the bone microstructure and dividing them into three defined categories. Samples were then sorted into seven different disease classes and healthy controls, and the presence or absence of these pathological features was noted. The samples were then evaluated using a chi squared test. Inter and intra-observer errors were carried out for both quantitative sections using the Technical error of measurement (TEM), rTEM and the coefficient of reliability (R). Statistical analysis was carried out using RStudio. Next comparative statistics, including correlations and regressions, were run between the complete data set, the pathological set, and the healthy set, specifically testing the relationship between actual age and cortical area, and actual age and OPD. This demonstrated that while the complete set did show a significant relationship between age and the two standard variables (R2 0.24 for cortical area and age, R2 0.21 for OPD and age), this relationship was stronger in the healthy data set (R2 0.32 for cortical area and age, R2 0.46 for OPD and age), and much weaker within the pathological data set (R2 0.09 for cortical area and age, R2 0.01 for OPD and age). Next the measurements for osteon specific metrics were evaluated for their relationship to age as a complete data set, then as a set including the metric means of each sample, and finally as forty age matched pairs. This was done using t-tests, correlations, and regression analysis. Correlation analysis for metric means found significant relationships between osteon area, osteon total area, osteon minimum length, osteon perimeter, osteon maximum length, and age in the healthy group. All these variables had p-values of <0.004. In the pathological group, correlations between the metric variables and age were much weaker, with only three variables, osteon maximum length, osteon area, and osteon minimum length yielding statistically significant relationships, and the p-values for these three relationships were all greater than 0.02, significantly weaker than the healthy group. This research helps define visually identifiable pathological changes to the bone microstructure, and proves that there is a significant difference between the microstructure of pathological and healthy bone. It goes on to show that, while there is a strong relationship between certain microstructural elements and age which would be beneficial for creating population specific age estimation techniques, microstructural elements of pathological samples have much weaker relationships to age, and that their inclusion in data sets used to create population specific formulae would influence the method’s accuracy. Further, if a sample is known to be pathological and shows pathological microstructural features, any histological age estimation technique would be less effective, and that if applied. It is recommended that more research is carried out focusing on specific pathologies, and that more population specific studies should be undertaken.