Glutathione content but not gamma glutamyl cysteine synthetase mRNA expression predicts cisplatin resistance in head and neck cancer cell lines.

PURPOSE: To correlate cellular glutathione content and gamma-glutamyl cysteine synthetase (gamma GCS) mRNA expression with cisplatin sensitivity in a panel of seven head and neck squamous cancer cell lines. METHODS: Cisplatin IC50 was determined for each cell line using a sodium tetreazolium (XTT) assay. Cellular glutathione content was measured by using a previously reported enzymic method. gamma GCS mRNA expression was measured using an RNase protection assay. RESULTS: Total cellular glutathione was an excellent predictor of cisplatin sensitivity in this series of cell lines. The IC50 for cisplatin in the cell line with the highest glutathione concentration was approximately 90 times higher than in the cell line with the lowest glutathione concentration. Regression analysis showed a highly statistically significant positive correlation between cisplatin IC50 and cellular glutathione (coefficient of determination R2 = 0.81, P = 0.0012). Some-what surprisingly, in contrast to previous studies in ovarian cancer, gamma GCS mRNA expression in these cell lines was not significantly predictive of either total cellular glutathione or cisplatin sensitivity (R2 = 0.005, P = 0.84). As expected, treatment of resistant cell lines with buthionine sulfoximine resulted in decreased cellular glutathione and enhanced cisplatin sensitivity. CONCLUSIONS: Our results suggest that glutathione may be an important determinant of cisplatin sensitivity in clinical head and neck cancer. Since cisplatin is the most active chemotherapy drug for the treatment of this disease, this correlation may have important clinical relevance. The lack of correlation between glutathione level and gamma GCS expression suggests that salvage or alternate synthetic pathways may be critical in these cells.

Rat model for a vascularized laryngeal allograft.

A new rat model was developed to reexamine the potential for laryngeal transplantation. The final anatomic derivation evolved from two earlier developmental phases. The first model had only a single arterial anastomosis; the second had an end-to-end arterial anastomosis with an end-to-end arteriovenous shunt. The final product employed an end-to-side arterial shunt and an end-to-side arteriovenous shunt for revascularization. The allografts were sited in tandem with the intact recipient larynges and were not innervated. A total of 16 animals were studied in phase 3; 2 died and the remaining 14 had a 64% arterial patency at intervals of 1 to 14 days. Our purpose is to detail the relevant technical considerations of this new model and compare it with historical controls.

The use of absorbable sutures in laser-assisted microvascular anastomoses.

The efficacy of the laser in performing microvascular anastomoses has been well established in the laboratory using a number of wavelengths. These studies have concluded that laser-assisted microvascular anastomoses are at least comparable to, if not superior to, ordinary suture techniques. The advantages have been the diminished foreign body reaction that occurs as a consequence of using only a few stay sutures to hold the vessels in approximation while the laser bonding is performed, as well as the rapidity of the surgery. We have already shown that absorbable sutures (polyglactin 910) are as efficacious as standard nonabsorbable sutures (nylon) in both arterial and venous microanastomoses. The current study was undertaken to see if the foreign body reaction could be diminished even further by the use of absorbable 10.0 sutures and compare these findings to laser-assisted microvascular anastomoses performed with nonabsorbable sutures. The carbon dioxide milliwatt laser was used to perform laser-assisted microvascular anastomoses in rat femoral arteries and veins. Patency rates and histological response were compared at intervals of 3 days and 1, 2, 4, 8, and 12 weeks postoperatively. Both arterial and venous patency rates were comparable (overall absorbable, 91.2%-52/57; overall nonabsorbable, 87.7%-50/57), as was the degree of inflammation and fibrosis. We conclude that absorbable sutures can be used for laser-assisted microvascular anastomoses and have the potential of allowing healing to occur without any foreign material within the surgical site.

The efficacy of absorbable suture for microvascular anastomoses.

Although microvascular anastomoses are routinely performed with nonabsorbable sutures, a foreign body reaction can be stimulated that acts as a nidus for inflammation, infection, and possible thrombosis. Absorbable sutures should be able to diminish this reaction. There are sparse data describing the use of absorbable sutures for both arterial and venous anastomoses. This investigation compares standard microvascular anastomotic technique using nonabsorbable 10-0 sutures (nylon) with absorbable 10-0 sutures (polyglactin 910), using a previously reported tubed superficial epigastric flap model in rats. Patency rates and histologic responses are compared at intervals of 3 days and 1, 2, 8 and 12 weeks postoperatively. Arterial and venous patency rates were similar for both materials (overall nonabsorbable, 85.9%; overall absorbable, 84.4%). A somewhat increased inflammatory response was noted in the arterial absorbable group at 1 week and in both the venous and arterial nonabsorbable groups at 8 and 12 weeks postoperatively. The later finding most probably represents the continued presence of the nylon sutures. The incidence of intraluminal thrombosis was greater for nonabsorbable suture, occurring in two arterial and four venous anastomoses, compared with only one absorbable suture venous anastomosis. We conclude that nonabsorbable suture is as efficacious as standard absorbable material in both arterial and venous microanastomoses with the potential benefit of diminished foreign body reaction.