**3. Discussion**

**Figure 2.** Percentage of chondrocytes without apparent alterations (A), nuclei with pyknosis (B), karyorrhexis (C), and

The mean percentages of normal chondrocytes, pyknotic nuclei, and empty lacunae were significantly affected by lyophilization compared with cryopreservation [normal chondro‐

karyolysis (D) before and after biopreservation of the tracheal segments.

34 Cryopreservation in Eukaryotes

cytes, pyknotic cells, and empty lacunae (p < 0.01), ANOVA + Tukey].

Despite numerous attempts made with synthetic prostheses and tracheal transplants of autologous tissues, none of these alternatives has permitted functional reconstruction due to complications such as devascularization, stenosis, necrosis, dehiscence, infection, immune reactions, and formation of granulation tissue. Reconstruction of long‐segment tracheal defects is an important, unresolved clinical problem. Several attempts at replacing the trachea have been made using lyophilized and cryopreserved tracheal segments. We compared the effect of two preservation methods (lyophilization and cryopreservation) on the canine tracheal cartilage by microscopic evaluation of necrotic cell death.

Cell death by means of histological changes and expression of caspase‐3 in lyophilized and cryopreserved tracheal cartilage was evaluated. Histology was evaluated by light microscopy. Our results showed that the tracheal cartilage treated with lyophilization or cryopreservation induced significant changes in the chondrocyte integrity. These changes were more severe with lyophilization.

The nucleated chondrocytes decreased significantly with preservation (lyophilization or cryopreservation) compared to the tracheal segments from the control group. However, unaltered chondrocytes were significantly affected by lyophilization compared with cryopre‐ servation. Major cell death could lead to a failure in cartilage matrix turnover because chon‐ drocytes are the only source of matrix component synthesis in the cartilage. If chondrocytes are crucial for adequate matrix balance and function, chondrocyte death could be related to tracheal death.

Chondrocytes reside in cavities in the matrix called cartilage lacunae. In the bone, the standard technique for determining osteonecrosis in clinical pathology remains the identification of empty lacunae [10, 11]. Most osteocyte number studies used the number of lacunae as a reference [11]. Empty osteocyte lacunae rarely occur within the first seven days of bone death, and it may take over 16 weeks for a complete loss of osteocytes from the lacunae [10].

We found that the percentage of empty lacunae increased significantly with lyophilization. This increase was 33% higher than in cryopreserved tracheal segments. This increased percentage of empty lacunae might indicate that lyophilized chondrocytes are prone to die sooner than cryopreserved chondrocytes.

A series of characteristic morphological changes occurs when chondrocytes lose their viability. Necrosis, or irreversible cell death, is characterized by nuclear swelling, pyknosis, karyor‐ rhexis, karyolysis, and cytoplasmic eosinophilic staining [10]. Several studies over the past decade have shown convincing evidence that osteocytes die by apoptosis. During the last stages of programmed cell death, osteocytes break up into apoptotic bodies that may be less than 2–3 μm in diameter. Such bodies are likely to remain undetectable to light microscopy and may well be interpreted as empty lacunae [12]. On the other hand, apoptotic bodies are phagocytosed by neighboring cells or macrophages, thereby preventing the retention of cellular debris in the extracellular space. In the bone, some osteocytes die in situ by apoptosis. This process is brief, and the remnants of apoptotic death are recognizable by conventional light microscopy. Apoptotic bodies can remain as pyknotic nuclei for many months. Eventu‐ ally, the products of cell death are removed and become undetectable, leaving apparently empty lacunae. Much later, the empty lacunae become filled with mineralized debris and may no longer be visible by light microscopy [13]. Our results indicate that the percentage of chondrocytes with nuclear pyknosis decreased after both lyophilization and cryopreservation. This decrease was statistically significant only in the lyophilized group of tracheal segments, and pyknosis does not necessarily mean necrosis. Coupled with the decrease in the percentage of pyknotic nuclei after preservation, we also found that caspase‐3 expression was significantly diminished due to lyophilization or cryopreservation of the tracheal segments. If caspase‐3 expression is implicated in tissue damage due to ischemia [5, 6], orotracheal cannulation due to the tidal volume effect [7], and the freezing process inherent in the preservation, it is possible that in lyophilized or cryopreserved tracheal segments, cell death by apoptosis or expression of caspase‐3 occurs before biopreservation. The main cell death pathway after rehydrating or thawing tracheal segments would then be necrosis. According to Hedgecock [12], if large numbers of osteocytes undergo apoptosis at the time of tissue harvest, many other osteocytes likely underwent apoptosis previously, becoming shrunken and fragmented and disappearing from their lacunae. Empty lacunae may represent osteocytes that previously died by apoptosis. We also found that lyophilized or cryopreserved tracheal segments showed a significantly decreased percentage of nuclei with karyorrhexis and an increased percentage of nuclei with karyolysis. However, this increase was significant only after lyophilization. Karyolysis is usually preceded by karyorrhexis and occurs primarily as a result of necrosis. In apoptosis, karyorrhexis usually follows after the core is dissolved in apoptotic bodies. The mean per‐ centage of normal or unaltered chondrocytes, pyknotic nuclei, and empty lacunae was significantly affected by lyophilization compared with cryopreservation. Our results therefore strongly suggest that lyophilization has a deleterious impact on the tracheal cartilage and corroborate the findings reported by Lenot et al. [2] and Villalba‐Caloca [14].
