In vivo study to assess fat embolism resulting from the Reamer-Irrigator-Aspirator 2 system compared to a novel aspirator-based concept for intramedullary bone graft harvesting

Abstract

Introduction: Fat embolism (FE) following intramedullary (IM) reaming can cause severe pulmonary complications and sudden death. Recently, a new harvesting concept was introduced in which a novel aspirator is used first for bone marrow (BM) aspiration and then for subsequent aspiration of morselized endosteal bone during sequential reaming (A + R + A). In contrast to the established Reamer-Irrigator-Aspirator (RIA) 2 system, the new A + R + A concept allows for the evacuation of fatty BM prior to reaming. In this study, we hypothesized that the risk of FE, associated coagulopathic reactions and pulmonary FE would be comparable between the RIA 2 system and the A + R + A concept.

Materials and methods: Intramedullary bone graft was harvested from intact femora of 16 Merino sheep (age: 1-2 years) with either the RIA 2 system (n = 8) or the A + R + A concept (n = 8). Fat intravasation was monitored with the Gurd test, coagulopathic response with D-dimer blood level concentration and pulmonary FE with histological evaluation of the lungs.

Results: The total number and average size of intravasated fat particles was similar between groups (p = 0.13 and p = 0.98, respectively). D-dimer concentration did not significantly increase within 4 h after completion of surgery (RIA 2: p = 0.82; A + R + A: p = 0.23), with an interaction effect similar between groups (p = 0.65). The average lung area covered with fat globules was similar between groups (p = 0.17).

Conclusions: The use of the RIA 2 system and the novel A + R + A harvesting concept which consists of BM evacuation followed by sequential IM reaming and aspiration of endosteal bone, resulted in only minor fat intravasation, coagulopathic reactions and pulmonary FE, with no significant differences between the groups. Our results, therefore, suggest that both the RIA 2 system and the new A + R + A concept are comparable technologies in terms of FE-related complications.

Keywords: Bone graft; Fat embolism; Harvesting; Reaming.

Fig. 1

Experimental design and protocol. In the RIA 2 group, a one-step procedure with continuous flow of irrigation fluid during reaming, the RIA 2 system, is applied for BG harvesting, and in the A + R + A group, BM is first evacuated from the medullary canal using the aspirator followed by applying sequential IM reaming and aspiration of morselized endosteal bone chips using the aspirator (A). The experimental protocol illustrates the individual time points for the conducted assessments, including the Gurd test, D-dimer concentration, and lung histology (B). A is adapted from Ref. [3], and B is partially created with BioRender.com

Fig. 2

Evaluation of venous fat intravasation assessed with the modified Gurd test. Epifluorescence microscope image (magnification 100x), with a blue rectangle in the inset depicting the scanned area of the filter membrane, showing visible individual fat globules in an exemplary section of a sample that consists of a total of 289 particles ranging in diameter from 1.14 to 18.78 μm, resulting in a slide area covered with fat of 0.048% with a corresponding average median Gurd value of 0.5 (A). Overall, in the RIA 2 (n = 75) and A + R + A (n = 81) samples, no differences in the median number of bone marrow fat particles per sample (p = 0.13) and Gurd rating per sample (p = 0.98) were observed (B, red crossbar = median). The effect seen in the experimental groups on the area of the microscope slide covered with fat globules (C) was not different either for the comparison between the experimental groups per time point or within the respective experimental group in the comparison of the different time points (all p > 0.05). Moreover, there was no difference in the Gurd rating (D) between the two groups at the respective time point or within the groups between the time points (all p > 0.05). NS, nonsignificant

Fig. 3

D-dimer concentrations measured in the blood plasma samples obtained from the experimental groups at multiple time points for up to 4 h after completion of surgery. No significant different interaction effect was observed on D-dimer concentration over time when comparing the RIA 2 with the A + R + A group (p = 0.65). The data are displayed as mean ± SD

Fig. 4

Histological results for sheep lung lobe area covered with fat globules. During fixation with lipids, OsO₄ is reduced to lower oxides that are black and insoluble and are deposited in tissues. The intravascular lipids appeared as single or multiple, well-defined and black (OsO₄) globules (A). The illustrated section in A includes the lung lobe area covered on this section, which was 0.006% of the total stained area with an average particle size of 878.003 μm² and in total eight particles (indicated by black circles) identified as fat globules. No difference between the RIA 2 and A + R + A groups (n = 48) for average lung lobe area covered with fat globules (p = 0.17) was observed (B). For the area covered with fat, no significant differences (all p > 0.05) were observed for the individual lung lobes either within a group or between the experimental groups (C). NS, nonsignificant. The data in panels B and C are displayed as boxplots with individual data points superimposed in panel C