Thermal effects of glenoid reaming during shoulder arthroplasty in vivo.

Background: Glenoid component loosening is a common cause of failure of total shoulder arthroplasty. It has been proposed that the heat generated during glenoid preparation may reach temperatures capable of producing osteonecrosis at the bone-implant interface. We hypothesized that temperatures sufficient to induce thermal necrosis can be produced with routine drilling and reaming during glenoid preparation for shoulder arthroplasty in vivo. Furthermore, we hypothesized that irrigation of the glenoid during reaming can reduce this temperature increase.

Methods: Real-time, high-definition, infrared thermal video imaging was used to determine the temperatures produced by drilling and reaming during glenoid preparation in ten consecutive patients undergoing total shoulder arthroplasty. The maximum temperature and the duration of temperatures greater than the established thresholds for thermal necrosis were documented. The first five arthroplasties were performed without irrigation and were compared with the second five arthroplasties, in which continuous bulb irrigation was used during drilling and reaming. A one-dimensional finite element model was developed to estimate the depth of penetration of critical temperatures into the bone of the glenoid on the basis of recorded surface temperatures.

Results: Our first hypothesis was supported by the recording of maximum surface temperatures above the 56°C threshold during reaming in four of the five arthroplasties done without irrigation and during drilling in two of the five arthroplasties without irrigation. The estimated depth of penetration of the critical temperature (56°C) to produce instantaneous osteonecrosis was beyond 1 mm (range, 1.97 to 5.12 mm) in four of these patients during reaming and one of these patients during drilling, and two had estimated temperatures above 56°C at 3 mm. Our second hypothesis was supported by the observation that, in the group receiving irrigation, the temperature exceeded the critical threshold in only one specimen during reaming and in two during drilling. The estimated depth of penetration for the critical temperature (56°C) did not reach a depth of 1 mm in any of these patients (range, 0.07 to 0.19 mm).

Conclusions: Temperatures sufficient to induce thermal necrosis of glenoid bone can be generated by glenoid preparation in shoulder arthroplasty in vivo. Frequent irrigation may be effective in preventing temperatures from reaching the threshold for bone necrosis during glenoid preparation.