Guidance for removal of fetal bovine serum from cryopreserved heart valve processing

Abstract

Bovine serum is commonly used in cryopreservation of allogeneic heart valves; however, bovine serum carries a risk of product adulteration by contamination with bovine-derived infectious agents. In this study, we compared fresh and cryopreserved porcine valves that were processed by 1 of 4 cryopreservation formulations, 3 of which were serum-free and 1 that utilized bovine serum with 1.4 M dimethylsulfoxide. In the first serum-free group, bovine serum was simply removed from the cryopreservation formulation. The second serum-free formulation had a higher cryoprotectant concentration, i.e. 2 M dimethylsulfoxide, in combination with a serum-free solution. A colloid, dextran 40, was added to the third serum-free group with 2 M dimethylsulfoxide due to theoretical concerns that removal of serum might increase the incidence of tissue cracking. Upon rewarming, the valves were inspected and subjected to a battery of tests. Gross pathology revealed conduit cracking in 1 of 98 frozen heart valves. Viability data for the cryopreserved groups versus the fresh group demonstrated a loss of viability in half of the comparisons (p < 0.05). No significant differences were observed between any of the cryopreserved groups, with or without bovine serum. Neither routine histology, autofluorescence-based multiphoton imaging nor semiquantitative second-harmonic generation microscopy of extracellular matrix components revealed any statistically significant differences. Biomechanics analyses also revealed no significant differences. Our results demonstrate that bovine serum can be safely removed from heart valve processing and that a colloid to prevent cracking was not required. This study provides guidance for the assessment of changes in cryopreservation procedures for tissues.

Fig. 1

Leaflet biomechanical testing methods. A Preconditioning of the leaflet sample by cycling through several physiological loading cycles. Low load modulus is measured at this step. B The leaflet section is then elongated to result in a stress of 400 kPa and held at constant deformation for 1,000 s while recording the decay in stress. C Measurement of ultimate properties by pulling the leaflet at a constant strain rate until failure.

Fig. 2

Summary of viability data obtained using the Alamar Blue assay. All data are displayed as the mean ± SE (n > 4–6). RFU = Relative fluorescent units. ∗ p < 0.05 versus fresh control samples.

Fig. 3

A Cross-section of a fresh intact heart valve leaflet, stained with Movat's pentachrome stain. B–F Multiphoton-induced autofluorescence images of the leaflet inflow side of a fresh specimen (B) and specimens cryopreserved in serum (control; C), serum-free DMEM (D), serum-free Unisol (E) and serum-free Unisol + dextran 40 (F). No significant differences were observed between the samples with regard to the ECM components such as elastic fibers (green, 760 nm) and collagen bundles (red, 840 nm). Scale bars = 50 μm.

Fig. 4

A–C Histology of the worst areas seen in fresh tissue, demonstrating white spaces (arrows, A), bovine serum-containing cryopreserved controls with tissue breaks (arrows, B) and serum-free cryopreserved tissue in DMEM with a single focal area of tissue damage (arrows, C). D Fresh tissue without a white space. E Serum-free cryopreserved tissue in Unisol with white spaces (arrows). F Serum-free cryopreserved tissue in Unisol + dextran 40 with a white space (arrow). The damage seen in B and C may be due to induced artifacts or, more likely, ice damage. Hematoxylin and eosin stain. Scale bar = 100 μm.

Fig. 5

Leaflet and conduit ultimate load. All data are displayed as means ± SE (n = 7–10). No significant differences were observed.