Report on a large animal study with Göttingen Minipigs where regenerates and controls for articular cartilage were created in a large number. Focus on the conditions of the operated stifle joints and suggestions for standardized procedures.

The characterization of regenerated articular cartilage (AC) can be based on various methods, as there is an unambiguous accepted criterion neither for the natural cartilage tissue nor for regenerates. Biomechanical aspects should be considered as well, leading to the need for more equivalent samples. The aim of the study was to describe a large animal model where 8 specimens of regenerated AC can be created in one animal plus the impact of two surgeries on the welfare of the animals. The usefulness of the inclusion of a group of untreated animals (NAT) was to analyzed. Based on the histological results the conditions of the regenerates were to be described and the impact on knee joints were to be explored in terms of degenerative changes of the cartilage. The usefulness of the statistical term "effect size" (ES) will be explained with histological results. We analyzed an animal model where 8 AC regenerates were obtained from one Göttingen Minipig, on both sides of the trochleae. 60 animals were divided into 6 groups of 10 each, where the partial thickness defects in the trochlea were filled with matrices made of Collagen I with or without autologous chondrocytes or left empty over the healing periods of 24 and 48 weeks. One additional control group consisting of 10 untreated animals was used to provide untouched "external" cartilage. We harvested 560 samples of regenerated tissue and "external" controls, besides that, twice the number of further samples from other parts of the joints referred to as "internal" controls were also harvested. The animals recovered faster after the 1st operation when the defects were set compared to the 2nd operation when the defects were treated. 9% of all animals were lost. Other complications were for example superficial infections, seroma, diarrhea, febrile state and an injury of a claw. The histological results of the treatments proved the robustness of the study design where we included an "external" control group (NAT) in which the animals were not operated. Comparable significant differences between treated groups and the NAT group were detected both after ½ year and after 1 year. Spontaneous regenerated AC as control revealed differences after an observation time of nearly 1 year. The impact of the treatment on cartilage adjacent to the defect as well as the remaining knee joint was low. The ES was helpful for planning the study as it is shown that the power of a statistical comparison seems to be more influenced by the ES than by the sample size. The ranking of the ES was done exemplarily, listing the results according to their magnitude, thus making the results comparable. We were able to follow the 3 R requirements also in terms of a numerical reduction of animals due to the introduction of a group of untreated animals. This makes the model cost effective. The presented study may contribute as an improvement of the standardization of large animal models for research and regulatory requirements for regenerative therapies of AC.