Fluorescent viability stains overestimate chondrocyte viability in osteoarticular allografts

Background: Allografts from many tissue banks are carefully processed and stored with the goal of preserving chondrocyte viability. However, the importance of living chondrocytes for graft stability is unclear, in part because actual viabilities of individual allografts at the time of placement are seldom known. Hypotheses: Cell yields from allograft and fresh cartilage differ significantly if chondrocyte viability in allografts is lower than indicated by fluorescence staining with conventional viability probes. In addition, transmission electron microscopy will show significant differences in the percentage of morphologically abnormal chondrocytes in allograft and fresh cartilage. Study Design: Controlled laboratory study. Methods: Fluorescence viability staining, chondrocyte yield, and chondrocyte characteristics were studied in 8 commercial osteochondral allografts (7 hemicondyles, 1 talus) and 4 freshly harvested cartilage samples from an adult distal femur (age, 46 years), from an adult talus (age, 51 years), and from an adult tibial plateau (age, 29 years) and from a juvenile distal tibia (age, 9 years). Selected fresh and allograft specimens. were repeatedly frozen and thawed to deliberately kill chondrocytes by membrane disruption. The findings were analyzed to determine if allograft and fresh cartilage were significantly different with respect to each of the 3 different outcome measures. Results: Although fluorescent staining indicated that approximately 75% of chondrocytes were viable (calcein AM-labeled) in allograft cartilage, counterstaining with 4,' 6-diamidino-2-phenylindole showed that fewer than 30% contained identifiable nuclei. In contrast, 100% of cells labeled as viable contained nuclei in fresh cartilage. Killing chondrocytes by freeze-thawing before staining did not diminish calcein AM staining in allograft cartilage but caused a significant reduction in fresh cartilage. The average yield of chondrocytes from allograft cartilage was less than 200 000/100 mg tissue, significantly lower than in fresh cartilage, which averaged more than 1.5 million/100 mg tissue. The yield from freeze-thawed controls was less than 24 000/100 mg. Cell numbers increased after 7 days of culture in all cases except for chondrocytes from freeze-thawed cartilage, an indication that the isolated cells were viable. Morphologic analysis by transmission electron microscopy revealed significant increases in the numbers of chondrocytes with pyknotic or absent nuclei or with disintegrated plasma membranes in allograft versus fresh cartilage. Conclusion: Conventional fluorescence probes are unreliable for analyzing chondrocyte viability in osteoarticular allografts. Alternative methods for assessment of viability, such as cell culture and ultrastructural imaging, may provide more accurate assessment of viability in allografts. Clinical Relevance: Conventional staining methods that overestimate chondrocyte viability in osteoarticular allografts may mislead investigators attempting to assess the effects of chondrocyte viability on graft stability following implantation. A more reliable means to measure chondrocyte viability will be required to accurately assess these effects.