In the past, six histological grading systems for classical chondrosarcoma have been published. Due to the inhomogeneity and complexity of these classifications, the comparison of clinical data, survival rates and local failures has to be considered critically. In 1996, the author published a grading system that was simple to use and easily reproduced. This system was based on a few nuclear features. The main intention of the current study was to verify whether the histological grade, which was defined by the author's classification, correlates with the recurrence rate. In a retrospective study, clinical data, X-rays and histological material from 35 patients with classical chondrosarcoma and 16 patients with enchondroma were analysed. Statistical analysis was done using the chi-squared test and the Fisher exact test. Local recurrence occurred in 25.7% of all patients. The difference in recurrence rate among grades 1-3 was statistically significant ( P=0.002). The frequency of grades 1-3 varied up to 54%, when published grading systems were compared. No significant difference between the histological grade and features such as double nuclei and mitosis were observed. The frequency of cellularity, double nuclei and mitoses was similar between enchondromas and low-grade chondrosarcomas. Of chondrosarcoma patients, 90.6% of total patients and 87.5% of those with grade-1 lesions reported pain, whereas only 43.8% of the enchondroma patients did. Even in patients with grade-1 chondrosarcomas, radiological findings were much more aggressive in comparison with enchondromas. The histological grade, defined on the basis of the author's simple and reproducible grading system, indicates the risk of local recurrence, especially in cases that are inadequately treated. Grade-3 chondrosarcomas and lesions located in regions where the removal of the tumour would be difficult have to be given special attention.

The combined histological and microcomputed analysis of human iliac crest biopsies leads to major advances in our understanding of three-dimensional bone architecture. Microcomputed tomography avoids the time-consuming reconstruction and artifacts of serial sections. Furthermore, its high resolution allows the recording of structural differences as low as 10 microns. Thus, three-dimensional analysis in combination with histological evaluation of cellular dynamics facilitates earlier and easier recording of changes of cancellous bone.