Cartilage-flow phenomenon and evidence for it in perichondrial grafting

The cartilage-flow phenomenon has been frequently described but not in connection with transplantation procedures for treatment of cartilage lesions. Consequently, we examined this particular phenomenon in an experimental sheep model originally set up to study the use of perichondrial grafts for repair of full-thickness cartilage lesions. Osteochondral lesions were made in non-weight-bearing and weight-bearing areas of knee joints in 36 sheep. The defects were filled with autologous rib perichondrial grafts and secured by either collagen sponges or fibrin glue (n = 24 animals). Defects without perichondrial grafts served as controls (n = 12 animals). Following 1 week of immobilisation of the operated leg, the plaster was removed, and the animals were allowed to move freely. Animals were killed after 4, 8, 12 and 16 weeks. Grafts including rims of original surrounding cartilage and bone were removed and investigated by means of macroscopy, histology and micromorphology, including scanning electron microscopy and analysis under polarized light. Cartilage flow was observed in all specimens by 4 weeks after drilling the defects, independent of the weight-bearing condition. These flow formations exhibited a bending of the collagen fibres centrally into the defects, reduction of metachromasia, cell cluster formation and areas of reduced cell density. Time-dependent flow formations were observed related to the weight-bearing condition and whether or not the defects had been grafted. In grafted, non-weight-bearing defects further cartilage flow was stopped 8 weeks after transplantation by the growing perichondrial transplants, which had filled the defects completely. In contrast, control defects exhibited further flaw formations in both areas. The same was observed in grafted defects in the weight-bearing area. In defects without complete filling as not enough spontaneously growing fibrous tissue had arisen from the bottom of the defects, two different observations were made: either the defects exhibited a fungiform mass of fibrous tissue that had overgrown the lateral flow formations, or the central mass of fibrous tissue was overgrown by the lateral flow formations. In conclusion, cartilage flow seems to be a mechanically induced phenomenon at the rims of cartilage lesions that contributes little to the reduction of size of large osteochondral defects. There was no evidence for new cartilage proliferation or production of cartilaginous matrix at the rims of the lesions.