Session V - Femur

Pulmonary Gas Exchange during Intramedullary Fixation of Femoral Shaft Fractures

Matthew J. Weresh, MD; Kyle J. Jeray, MD; Michael D. Stover, MD; Michael J. Bosse, MD; James F. Kellam, MD; Stephen H. Sims, MD; Michael Gillette, MD; Jeffrey A. Kline, MD, Carolinas Medical Center, Charlotte, NC

Introduction: Experimental and clinical evidence has indicated that the process of intramedullary femur fracture fixation can extrude bone medullary contents into the venous circulation, leading to embolization of bone marrow to the lung. Whether or not this marrow embolization causes clinically significant pulmonary compromise remains controversial. By examining pulmonary gas exchange using the alveolar dead space fraction (Vd/Vt) and the alveolar-arterial oxygen gradient (A-a DO₂) measurements, the effects , if any, of the embolic burden on the lungs during intramedullary femoral nailing can be determined. Evidence from animal models and human studies of pulmonary embolism (PE) suggests that Vd/Vt, and to a lesser extent, A-a DO₂ will increase if a significant embolic burden is exerted on the lungs. The physiological basis for Vd/Vt rests on the principle that alveoli that are poorly perfused will not participate in gas exchange. Thus, gas expired from embolized lung units will be relatively dilute in CO₂ and, with exhalation, gas from these lung units will mix with gas from well perfused lung units, thereby depressing the peak end-tidal CO₂, increasing the Vd/Vt. This study was undertaken to determine if intramedullary femoral nailing produces immediate worsening of pulmonary gas exchange in patients with femoral shaft fractures.

Methods: Fifty hemodynamically stable patients with femur fractures that were treated with intramedullary nailing were enrolled prospectively: 40 males and 10 females with an average age of 30 years. Three serial measurements were obtained: immediately prior to femoral nailing, 30 minutes after insertion of the nail, and 120 minutes after nailing. The Vd/Vt was calculated from the equation (PaCO₂-PetCO₂)/PaCO₂ , where PaCO₂ was the arterial partial pressure of carbon dioxide and PetCO₂ was the steady-state end-tidal carbon dioxide. The A-a DO₂ was calculated from the equation (715-47)FiO₂ - (PaCO₂/0.8 + PaO₂), where PaO₂ was the arterial partial pressure of oxygen and FiO₂ was the inspired fraction of oxygen. The patients' pulmonary clinical courses were assessed pre- and post-surgery and correlated to observed gas-exchange values.

Results: Forty-eight patients were enrolled as recent trauma (< 24 hrs), two for nonunions. The median interval between injury and nailing was 4.0 hours. No patients had a pre-injury history of pulmonary problems. Seven were smokers. Six had evidence of pulmonary contusions on pre-operative chest radiographs. Three patients had chest tubes placed pre-operatively for pneumothoraces. Two patients had bilateral femur fractures, and 43 patients had reamed nails. One patient developed fat embolism syndrome (FES) prior to femoral nailing. Another developed a PE on postoperative day 4.

Vd/Vt prior to nailing was 0.09±0.09,at 30 min 0.10±0.06 and at 120 min 0.08±0.07 (P > 0.2 using paired t-test at all time points). A-a DO₂ prior to nailing was 84±85 mm Hg, at 30 min 89±69 and at 120 min 51±45. No significant difference was found between times. No difference in data was found with analysis of fracture grade, number of reamer passes or time to nailing. Of significant difference (P < 0.05) was the Vd/Vt and A-a DO₂ in patients with lung contusion prior to nailing (Vd/Vt - 0.16+0.07, A-a DO₂ - 162+86) when compared to the uninjured lung group (Vd/Vt - 0.07+0.06, A-a DO₂ - 73+71). However, in this group there was no increase in the values after nailing. The patient with FES had a statistically significant difference from all the other patients at each time interval but did not have a significant change pre- to post-nailing (pre Vd/Vt - 0.35, 30 minutes post Vd/Vt - 0.33 and 120 minutes post Vd/Vt - 0.31). The patient with a PE had the third largest relative change in Vd/Vt (pre Vd/Vt - 0.06 to 120 minutes post Vd/Vt - 0.17). The only clinical manifestation exhibited by the patient was tachypnea (RR>22) at 72 hours postoperative.

Discussion: This study showed no significant change in Vd/Vt or A-a DO₂ with intramedullary fixation of femur fractures. A change would be expected if a significant degree of pulmonary vascular occlusion had occurred. In experimental glass bead embolization, the Vd/Vt increased proportionally to the degree of pulmonary vascular obstruction in mechanically ventilated goats. In humans, Vd/Vt has been examined as a method of screening for acute PE with excellent sensitivity. Carroll et al. have shown that a one-ml embolism, produced with a balloon- tipped catheter, produced a noticeable decrease in end-tidal CO₂ in mechanically ventilated humans (24 patients - sensitivity was 85%, and specificity was 94%). Thus, ample preliminary evidence has been set to justify the argument that pulmonary gas exchange should have changed if a significant, persistent pulmonary marrow element embolic burden were routinely induced by femoral nailing. The absence of any significant effect with reaming agrees with the work of Wolinsky et al., who found that, although the lungs are showered with marrow contents, as demonstrated with transthoracic echocardiography, no clinical change in pulmonary gas exchange was noted.

The development of FES and PE was too rare an occurrence in our study to draw any conclusions with regard to the sensitivity and specificity. However , the fact that the Vd/Vt was elevated prior to intramedullary fixation in the patient with FES and the fact that it did not change with surgery suggest that the insult to the lungs had already occurred, and the further embolic burden did not change the gas exchange. The patient with a PE did not become symptomatic until 72 hours after surgery. There was an increase in Vd/Vt over the 120 minutes. This may have represented an early indication of potential pulmonary worsening, but more studies with larger numbers need to be done before any conclusions can be reached.

Conclusion: The process of femoral nailing with and without reaming does not appear to produce significant pulmonary marrow element emboli to alter pulmonary gas exchange, as indicated by the absence of any change in either the alveolar dead space fraction or the alveolar-arterial oxygen gradient.