Elderly patients are at risk of developing cardiopulmonary and cognitive impairment following major orthopaedic surgery. One of the mechanisms believed to be responsible for such complications after total knee arthroplasty (TKA) is the release of embolic debris that may travel from the surgical site, through the lungs, and into the brain following tourniquet release. Removal of fat globules and marrow particulates from bone surfaces prior to pressurization and cementation of prosthetic components may reduce the number and size of embolic particles. We conducted a prospective, randomized clinical trial to compare the effect of carbon dioxide (CO2) gas versus saline lavage on the number and size of embolic particles observed during cemented TKA.

Twenty patients undergoing elective TKA were randomly assigned to one of two groups. In group A, standard high-pressure pulsatile saline lavage was used to clean the resected bone surfaces. In group B, the femoral canal was cleaned using CO2 lavage techniques and the resected bone surfaces were cleaned with a manual saline wash followed by CO2 lavage. All patients received the same TKA implant design. The presence of embolic particles in the heart and brain was intraoperatively monitored using transesophageal echocardiography (TEE) and transcranial Doppler (TCD) techniques, respectively. For each patient, TEE images were analyzed at tourniquet release and during the final range of motion (ROM) assessment prior to wound closure using the following five point cardiac echogenic scoring system: Grade 0: no emboli; Grade I: a few fine emboli; Grade II: a cascade of many fine emboli; Grade III: a cascade of fine emboli mixed with at least one embolus > 1 cm in diameter; and Grade IV: large embolic masses > 3 cm in diameter. The highest grade observed during either tourniquet release or ROM assessment was assigned to each patient. Cardiac emboli were then categorized according to embolic grade as follows: Grade 0 or I = Low; Grade II, III, or IV = High. For analysis of cerebral emboli, the total number of positive counts measured using TCD was recorded for each patient. TEE data were available for nine patients in group A and eight patients in group B. Comparative TCD data were available for seven patients in group A and six patients in group B. Fischer’s Exact Test was used to check for differences between groups.

For cardiac emboli, nine of nine (100%) patients in group A were in the High category based on their TEE grade, with eight patients being Grade II and one Grade III. In contrast, three of eight (37.5%) patients in group B were in the Low category, leaving only five (62.5%) in the High category (p = 0.08). All five group B patients in the High category were Grade II. No patients in group A had cerebral emboli detected using TCD. In group B, three of six patients had one cerebral embolus and the remaining three had none. Three patients in group B were excluded from the comparative TCD analysis due to the presence of a patent foramen ovale (PFO). These three patients with a PFO had one, three, and four cerebral emboli, respectively. No patients in group A had a PFO.

This study examines the effect of pulsatile saline versus CO2 gas lavage on intraoperative embolic events during TKA. Thirty-seven percent of patients in the CO2 lavage group had a Low cardiac echogenic score compared with 0% of patients in the standard pulsatile saline lavage group. A single cerebral embolus was detected in three of six patients in the CO2 lavage group compared with none in the seven patients in the standard pulsatile saline lavage group. Compared to published studies on cerebral emboli in TKA, the overall incidence of cerebral emboli in the current study was very low across both groups. The results of this study suggest that CO2 gas, as compared to pulsatile saline, lavage reduces the number of intraoperative cardiac emboli during total knee arthroplasty.