Comparison of dosimetric variability in the cervical spine between carbon fiber and titanium instrumentation.

Objective: Carbon fiber instrumentation has gained popularity in spine oncology for its radiographic advantage of reduced artifact on imaging. With its increased use in patients who undergo radiation therapy, the dosimetric accuracy of postoperative radiation with carbon fiber constructs compared to classic titanium instrumentation becomes an important question. The purpose of this study was to compare the dosimetric accuracy of postoperative radiation in carbon fiber-instrumented versus titanium-instrumented cadaveric cervical spines after corpectomy.

Methods: Two cadaveric specimens underwent two-level corpectomy using either titanium or carbon fiber instrumentation. Dosimeter chips were placed circumferentially around the constructs to calculate the dose of radiation to surrounding areas. The cadavers underwent one round of radiation with their respective constructs, and the dose of radiation was calculated and compared to the measured dose in each chip. After the first round of radiation, the instrumentation was switched between cadavers and the radiation therapy was repeated. The difference between the calculated and measured dose in carbon fiber versus that in titanium instrumentation in each cadaveric model was subsequently compared.

Results: There was a significant difference in the dosimetry calculated at the area of the spinal cord dorsal to the corpectomy cage, with 68% less variability between the calculated and measured dose in the carbon fiber construct compared to that of the titanium construct. The mean variation of the measured dose at the spinal cord was 7.73% in titanium versus 4.6% in carbon fiber (p = 0.024, 1-tail; p = 0.048, 2-tail). There was also 30% less variability between the measured and calculated dose in the carbon fiber construct at the dosimeter chips lateral to the spinal cord, with a mean variation in the carbon fiber-instrumented cadaver of 4.94% compared to 6.45% with titanium (p = 0.01, 1-tail; p = 0.02, 2-tail). When all the dosimeters were combined without the control group, there was a statistically significant 14.5% difference in the mean variation of the calculated versus measured doses between the carbon fiber-instrumented cadavers versus titanium-instrumented cadavers (4.8% vs 5.5%, respectively; p = 0.015, 1-tail; p = 0.03, 2-tail).

Conclusions: There is less variation in the calculated versus measured doses of radiation therapy in carbon fiber instrumentation compared to titanium instrumentation. This suggests improved radiation therapy delivering accuracy and complication avoidance in carbon fiber-instrumented spines.