- Email: [email protected]
Pulmonary injury in patients undergoing complex spine surgery
The Spine Journal 5 (2005) 269–276
Pulmonary injury in patients undergoing complex spine surgery Michael K. Urban, MD, PhDa,*, Kethy M. Jules-Elysee, MDa, James B. Beckman, MDa, Khillil Sivjee, MDb, Thomas King, MDb, Webster Kelsey, BAc, Oheneba Boachie-Adjei, MDd a Department of Anesthesiology, Hospital for Special Surgery, 535 East 70th Street, New York, NY 10021, USA Department of Pulmonary and Critical Care Medicine, New York-Presbyterian Hospital, Weill Medical College of Cornell University, 525 E. 68th Street, New York, NY 10021, USA c Department of Anesthesia, Hospital for Special Surgery, 535 East 70th Street, New York, NY 10021, USA d Department of Orthopedic Surgery, Weill Medical College of Cornell University, and Scoliosis Service, Hospital for Special Surgery, 535 East 70th Street, New York, NY 10021, USA
b
Received 19 March 2004; accepted 21 October 2004
Abstract
BACKGROUND CONTEXT: Previous reports have shown that 15% of patients who undergo sequential anterior, then posterior, surgical corrections for spinal deformities demonstrate evidence of acute lung injury. By analyzing the bronchoalveolar lavage (BAL) fluid from these patients for evidence of acute inflammation, we might gain some insight into the etiology of this acute lung injury. PURPOSE: To elucidate the etiology of acute lung injury after corrective surgery for adult spinal deformities. STUDY DESIGN/SETTING: Fifteen adult patients with scoliosis scheduled for elective sequential anterior then posterior corrective (A/P) spinal deformity surgery. PATIENT SAMPLE: Consecutive adult patients with scoliosis scheduled for elective corrective surgery with the author (OBA). OUTCOME MEASURES: Patients were assessed for postoperative respiratory complications by oxygen requirements, continued mechanical ventilation, and radiological evidence of diffuse bilateral interstitial or alveolar infiltrates. An acute pulmonary inflammatory response included the presence of inflammatory cells and elevated cytokines in BAL fluid. METHODS: BAL were performed after induction of anesthesia but before surgery, at the completion of surgery, and on the morning after surgery with the patient still intubated. BAL fluid was analyzed for inflammatory cells and cytokine interleukin-6 (IL-6) and tumor necrosis factor alpha (TNF-α) levels. Patients were assessed postoperatively for increased pulmonary vascular resistance, radiological evidence of diffuse bilateral alveolar infiltrates, and the requirement for ventilatory support beyond the first postoperative day (POD1). RESULTS: The cell counts of BAL fluid demonstrated significant increases in neutrophils, lymphocytes, and lipid laden macrophages (LLMAC) with surgery. The concentration of the cytokines IL-6 and TNF-α also increased with surgery. The elevations in BAL inflammatory cells and cytokine levels correlated positively with increased pulmonary vascular resistance and the requirement for mechanical ventilation. CONCLUSIONS: After A/P spine fusions, patients have evidence of an acute inflammatory pulmonary injury. Several etiologies exist for this finding, including blood and fluid infusions, direct trauma to the lung, a systemic inflammatory response, and the embolization of fat and bone-marrow debris. The presence of LLMAC in the lungs of these patients and the finding that the patient with the requirement for the longest ventilatory support also had the highest BAL LLMAC count, suggest that the embolization of fat and bone debris released from the spine during surgery may be at least
FDA device/drug status: not applicable. Nothing of value received from a commercial entity related to this research.
1529-9430/05/$ – see front matter doi:10.1016/j.spinee.2004.10.049
쑖 2005 Elsevier Inc. All rights reserved.
* Correspondence. Department of Anesthesiology, Hospital for Special Surgery, 535 East 70th Street, New York, NY 10021. Tel.: (212) 606-1792; fax: (212) 517-4481. E-mail address: [email protected] (M.K. Urban)
270
M.K. Urban et al. / The Spine Journal 5 (2005) 269–276
partially responsible for the lung injury. Further studies on the mechanism of lung injury during this procedure are warranted. 쑖 2005 Elsevier Inc. All rights reserved. Keywords:
Deformity surgery; Pulmonary injury; Pulmonary inflammation; Bronchoalveolar lavage; Lipid laden macrophages; Lung cytokine levels
Introduction There is significant morbidity in adult patients who undergo complex corrective procedures for spinal deformities [1]. These procedures are associated with large blood losses which may exceed the patient’s total estimated blood volume. This dramatic blood loss and the associated release of endogenous mediators of the stress response can have significant physiological effects including coagulopathies, myocardial depression, hypotension, and acute lung injury. It was previously reported that pulmonary artery catheter monitoring of patients who undergo complex spinal fusions facilitates the identification of patients who have sustained pulmonary injury [2]. These patients exhibited elevated pulmonary vascular resistance and hypoxemia, which is clinically similar to the fat embolism syndrome described in long bone fractures and hip arthroplasty [3]. The surgical interventions, including spinal instrumentation, thoracoplasties, and osteotomies may be responsible for the embolization of fat and bone marrow debris to the pulmonary microvasculature. However, the pulmonary effects could also be caused by transfusion-related lung injury or the release of a variety of cytokines that are associated with major trauma and sepsis. This study represents an initial attempt to characterize the pulmonary inflammatory response in lung injury during complex deformity spinal surgery.
of the patient was appropriate for spontaneous ventilation. Patients with radiographic evidence of diffuse bilateral interstitial or alveolar infiltrates and hypoxemia (PaO2⬍60 mm Hg on 3 L/m FIO2) were considered to have pulmonary complications. All patients were monitored with a pulmonary artery catheter which was inserted through a 9F cannula in the right internal jugular vein after the induction of general anesthesia but before surgery. The pulmonary artery catheter was used to maintain filling pressures (pulmonary artery occlusive pressure [PAOP]) during the procedure [2]. The mean PAOP for all 15 patients was 10.3⫾1.4 mm Hg at baseline and 10.9⫾1.9 mm Hg at the completion of surgery. An increase in the pulmonary artery diastolic pressure to PAOP gradient is indicative of increased pulmonary vascular resistance. Pain management consisted of subdural morphine injected during the posterior fusion and intravenous patient controlled analgesia hydromorphone. Bronchoalveolar lavages (BAL) were performed while the patients were ventilated. Three separate 60-cc volumes of sterile 0.9% saline were injected through a flexible fiberoptic bronchoscope wedged in the right middle lobe and lingula of the lung and then aspirated back with a syringe. The three lavages were combined for each time sample. A sample was spread on slides and then centrifuged; the cell-free supernatant was stored at ⫺80⬚C for cytokine analysis. The BAL were obtained at baseline (after induction of general anesthesia but before surgery), upon closure (at the completion
Materials and methods With institutional review board approval, 15 adult patients with scoliosis scheduled for elective sequential anterior then posterior corrective spinal deformity (A/P) surgery were enrolled. All of the patients received a preoperative pulmonary consultation which consisted of pulmonary function studies. The surgery consisted of sequential anterior thoracolumbar discectomies with insertion of bone (5–9 levels) and then posterior thoraco-lumbar fusions with DePuy AcroMed instrumentation (9–15 levels). All of the patients required a chest tube after closure of the anterior incision (9/ 15 right thoracotomies), and osteotomies were performed in 9 of the 15 patients (4–6 levels). All patients received the same general endotracheal anesthetic, which consisted predominantly of nitrous oxide and fentanyl, and a fraction of inspired oxygen (FIO2) ⬍30%. Patients were monitored in an intensive care unit for at least the first 24 hours, and sedated and ventilated for at least the first 12 postoperative hours, FIO2ⱕ60%. Positive pressure ventilation was terminated when both the anesthesiologist and pulmonologist determined that the respiratory condition
Table 1 Perioperative patient characteristics Patients (n) Age (yr) Weight (kg) Deformity Scoliosis Kyphoscoliosis Curve⬚ Preoperative RLD (n) LOS (h) Levels fused Anterior Posterior Osteotomies (n) Thoracotomy: Right/Left EBL (mL) Ventilated ⬎1 day (n) Mean pulmonary pressures (mm Hg) Baseline PAD-PCWP Closure PAD-PCWP
15 46⫾12 60⫾9 13 2 68⫾16 9 10⫾2 7⫾2 12⫾3 10 9/6 2437⫾1261 5 2 6
EBL⫽estimated intraoperative blood loss; LOS⫽length of surgery; PAD⫽pulmonary artery diastolic pressure; PCWP⫽pulmonary capillary wedge pressure; RLD⫽restrictive lung disease.
M.K. Urban et al. / The Spine Journal 5 (2005) 269–276
of surgery), and on the first postoperative day (POD1, the morning after surgery). Total cell counts of the BAL fluid were made using a hemocytometer and the differential cell counts on cytocentrifuge preparations of the BAL fluid for neutrophils, lymphocytes, and macrophages for each of the time points. In addition, the macrophages that contained fat-stained globules were counted as lipid laden macrophages (LLMACs). For cytokine levels, the BAL fluid was first concentrated using a centrifugal filter device (Millipore, Amicon Centriplus, model YM-3, Millipore Corp., Billerica, MA). The concentrated BAL samples were assayed for tumor necrosis factor alpha (TNF-α) and interleukin-6 (IL-6) via a commercially available sandwich-type enzyme immunoassay (PeliKine; Research Diagnostics Inc., Flanders, NJ). Because there is variability in the recovery of endothelial lining fluid (ELF) in the BAL samples, we corrected for the dilution of ELF with the saline washes via the urea method of Rennard et al. [4]. The urea concentration in the lavage fluid and plasma was quantified using a colorimetric assay (SigmaAldrich Urea Nitrogen Kit). The ELF cytokine level was calculated using the formula: ELF · Cytokine⫽BAL · Cytokine (Urea · plasma/Urea · BAL). Chest X-rays were taken on each postoperative day for 5 days or until the patient was extubated and the chest tube was removed. Hemodynamic data were collected for all patients during their hospital stay. Data were analyzed by analysis of variance with p⬍.05 considered significant and coefficient of correlation within 95% confidence levels, using the STAT View program.
Results We assessed lung injury in 15 adult patients (46⫾12 years) with spinal deformities after elective sequential anterior then posterior reconstructive spinal fusions (Table 1). All of the procedures consisted of anterior thoracotomies, multiple decompressions and removal of discs, and posterior segmental instrumentation (DePuy AcroMed, DePuy Spine, Raynham, MA). All of the procedures were associated with significant blood loss. Eight of the patients had preoperative pulmonary function tests suggestive of moderate restrictive lung disease. Baseline pulmonary artery pressures were normal in all patients but the pulmonary artery diastolic pressures increased with surgery whereas the PAOP was kept stable, indicative of increased pulmonary vascular resistance. Our BAL analysis indicates that surgery induced an acute inflammatory response in the lungs of all the patients. Baseline counts for neutrophils and LLMACs were very low (1–2 cells per 100) before surgery. The BAL cell counts (per 100 cells) of neutrophils, lymphocytes, and LLMACs
271
increased significantly with surgery (Fig. 1). The concentration of the cytokines TNF-α and IL-6 in the BAL also increased with surgery for all patients (Fig. 2). The inflammatory response with regard to cytokines was consistent within patients, because elevations in one cytokine (IL-6) correlated positively with elevations in the other cytokine, TNF (r2⫽.85, p⬍.001). Also, the complete inflammatory response was consistent within patients, because cytokine levels correlated positively with the number of BAL neutrophils (R2⫽.324, p⫽.03) and LLMACs (R2⫽.356, p⫽.03) on POD1. The inflammatory response, however, did not correlate positively with length of surgery, estimated intraoperative blood loss (EBL), or preoperative demographic characteristics, including age, weight, and the presence of preoperative restrictive lung disease. The number of LLMACs in the BAL on the day after surgery correlated positively (r2⫽.34; p⫽.02) with the number of fused vertebral segments (Fig. 3), but not with the EBL or volume of reinfused scavenged blood. However, the correlation between elevated cytokine levels and fused segments was not as strong (p⫽.12). Patient 8 had significantly higher LLMAC counts than the other patients and required 6 days of postoperative ventilatory support for respiratory failure. Five patients required ventilatory support for longer than 1 day. Three of these five patients (Patients 8, 10, and 12) had TNF-α and IL-6 levels greater than 1 SD above the mean on POD1 (Fig. 4). These three patients had radiographic evidence of acute lung injury, diffuse infiltrates over both lung fields and not limited to the side of the lavage or the surgery. With regard to the other two of the five patients, in Patient 11 the cytokine data were incomplete and in Patient 3 the IL-6 and TNF levels were higher intraoperatively (closure) than on POD-1. Patient 3 also had the highest BAL neutrophil count. In addition, there was a statistically significant association between elevated pulmonary vascular resistance at the end of the procedure (PVR⬎200 dynes/s/ cm⫺5), as well as the requirement for ventilatory support for longer than 1 day and elevated TNF and IL-6 levels (p⫽.01). Patient 8, who required 6 days of postoperative ventilatory support for respiratory failure, had the highest LLMAC counts and IL-6 levels. Patient 8 underwent the largest number of combined anterior and posterior discectomies and fusions.
Discussion After sequential anterior then posterior thoraco-lumbar fusions for spinal deformities, patients demonstrate evidence of acute lung injury [1,2]. This evidence includes radiological changes, elevated pulmonary vascular resistance, and the requirement for ventilatory support. This report supports the hypothesis that the lung injury may be the result of an acute pulmonary inflammatory response. All patients had increased BAL inflammatory cells and cytokine levels with
272
M.K. Urban et al. / The Spine Journal 5 (2005) 269–276
Fig. 1. Cell counts (per 100 cells) of (A) bronchoalveolar lavage neutrophils, (B) lymphocytes, and (C) LLMACs for 15 patients undergoing A/P spinal fusions. Baseline (after intubation, before surgery); closure (after both anterior and posterior spinal surgery), and POD1 (morning of the first postoperative day, patient still intubated). Gray bars⫽baseline; black bars⫽closure; white bars⫽POD1.
surgery. Patients with postoperative respiratory complications, who required ventilatory support beyond POD1, had the highest BAL inflammatory cell counts and cytokine levels. However, our sample size was small and differences between
patients for BAL inflammatory cell counts were not significant, except for LLMACs in Patient 8. Respiratory failure is a common complication of trauma [5]. This respiratory failure can be the result of direct injury
M.K. Urban et al. / The Spine Journal 5 (2005) 269–276
273
Fig. 1. Continued
to the lung or indirect damage from trauma-induced systemic reactions. The indirect lung injury occurs as a result of a generalized increase in pulmonary microvascular permeability to fluids and proteins [6]. These changes are possibly the result of the activation of humoral and cellular mediators of inflammation triggered by the systemic consequences of trauma. A/P surgery involves direct trauma to the lung on the side of the thoracotomy along with injury to the contralateral lung as manifested by diffuse infiltrates in that lung. In our study, 11 of the 15 patients demonstrated postoperative bilateral radiological pulmonary changes. Several studies have demonstrated that the acute lung injury associated with sepsis or trauma is the result of the damage produced by inflammatory cells and cytokines [7]. The BAL fluid of patients with adult respiratory distress syndrome after sepsis syndrome had elevated levels of neutrophils and proinflammatory cytokines (IL-6, IL-8), produced by alveolar macrophages [8]. In addition, the number of neutrophils and the concentration of IL-6 and IL-8 increases in BAL fluid after cardiopulmonary bypass, and there is a positive correlation between some cytokine levels and a decrease in arterial oxygenation [9]. Similarly, other studies support the hypothesis that TNF-α plays an important role in the inflammatory processes involved in aspiration pneumonitis [10].
In this study, level of BAL cytokines and inflammatory cells correlated positively with postoperative respiratory complications. However, patient demographics, the degree of spinal curvature, the number of patients with moderate restrictive lung disease, and the extent of surgery (time and number of osteotomies) did not correlate positively with the pulmonary inflammatory process. Large blood losses and the requirement for multiple transfusions have been associated with pulmonary insufficiency [11,12]. In this study, however, there was not a direct positive correlation between EBL and cytokine levels. Diffuse alveolar permeability changes will occur in patients exposed to barotraumas and oxygen toxicity during mechanical ventilation. However, none of our patients experienced inspired oxygen levels above 60% and the inflammatory process was initiated during surgery, not after days of ventilatory support. Similar pulmonary inflammatory changes have been noted during systemic circulatory collapse (sepsis). However, all of the patients in this study had a stable hemodynamic profile perioperatively. In this study, patients with pulmonary inflammatory changes had significantly higher LLMAC counts, which correlated positively with the number of vertebral segments instrumented. Patient 8 was a 57-year-old female with kyphoscoliosis and moderate restrictive lung disease, who required 11 hours of surgery with 10 anterior discectomies with
274
M.K. Urban et al. / The Spine Journal 5 (2005) 269–276
Fig. 2. The ELF concentration of (A) IL-6 and (B) TNF-α (pg/mL) in BAL fluid from 15 patients after A/P spinal fusions. Gray bars⫽baseline; black bars⫽closure; white bars⫽POD1.
bone insertion and 16 posterior levels fused with segmental instrumentation. This patient with the highest LLMAC count required ventilatory support for 6 postoperative days. Pulmonary microvascular fat has been considered a source of lung
injury in long bone trauma and during hip and knee arthroplasty [3,13,14]. The pulmonary changes which follow the embolization of fat and bone marrow debris during these procedures, result in acute inflammatory lung injury. In
M.K. Urban et al. / The Spine Journal 5 (2005) 269–276
275
Fig. 3. Regression analysis comparing number of vertebral levels fused to number of LLMAC in BAL on POD1. Correlation p⫽.02. LLMAC⫽⫺25.972⫹ 2.417 * Total levels; R2⫽.339.
our previous report we demonstrated that A/P surgical patients sustained increases in pulmonary vascular resistance with surgery, similar to the increases seen during revision total hip arthroplasty [14]. The multiple bone invasive procedures that occur during A/P procedures could serve as the embolic source of fat and bone marrow debris. This process would lead to the recruitment of inflammatory cells and the release of proinflammatory cytokines, which ultimately produce diffuse pulmonary injury. The presence of LLMACs in the BALs of these patients after surgery, however, is only
suggestive of fat embolism. Furthermore, we were unable to determine why only some patients exhibited increased alveolar LLMACs and subsequently developed an inflammatory response. Further studies are required to validate this hypothesis, delineate the patients at risk, and develop preventative treatments. Transesophageal echocardiography has been used to detect and quantify pulmonary emboli (fat; bone marrow debris) during total hip arthroplasty [14]. Future analysis will involve transesophageal echocardiography monitoring during these procedures.
Fig. 4. Scattergram of IL-6 levels in BAL on POD1 in 15 patients after A/P spinal fusions.
276
M.K. Urban et al. / The Spine Journal 5 (2005) 269–276
References [1] Rizzi PE, Winter RB, Lonstein JE, Denis F, Perra JH. Adult spinal deformity and respiratory failure. Spine 1997;22:2517–30. [2] Urban MK, Urquhart B, Boachie-Adjei O. Evidence of lung injury during reconstructive surgery for adult spinal deformities with pulmonary artery pressure monitoring. Spine 2001;26:387–90. [3] Gossling HR, Donohue TA. The fat embolism syndrome. JAMA 1979;241:2740–2. [4] Rennard SL, Basset G, Lecossier D, et al. Estimation of volume of epithelial lining fluid recovered by lavage using urea as marker of dilution. J Appl Physiol 1986;60:532–8. [5] Pepe PE. Acute post-traumatic respiratory physiology and insufficiency. Surg Clin North Am 1989;69:157–73. [6] Putensen C, Waibel U, Koller W, Putensen-Himmer G, Hormann C. Assessment of changes in lung microvascular permeability in posttraumatic acute lung failure after direct and indirect injuries to lungs. Anesth Analg 1992;74:793–9. [7] Warren JS. The role of cytokines in experimental lung injury. In: Kunkel SL, Remick DF, eds. Cytokines in health and disease. New York: Marcel Dekker, 1992:257–69.
One Hundred Years Ago in Spine
[8] Kobayashi A, Hashimoto S, Koogchi K, et al. Expression of inducible nitric oxide synthetase and inflammatory cytokines in alveolar macrophages of ARDS following sepsis. Chest 1998;113:1632–9. [9] Kotani N, Hashimoto H, Sessler D, et al. Neutrophil number and interleukin-8 and elastase concentrations in bronchoalveolar lavage fluid correlate with decreased arterial oxygenation after cardiopulmonary bypass. Anesth Analg 2000;90:1046–51. [10] Davidson BA, Knight PR, Helinski JD, Nader ND, Shanley TP, Johnson KJ. The role of tumor necrosis factor-α in the pathogenesis of aspiration pneumonitis in rats. Anesthesiology 1999;91:486–99. [11] Jacobs RR, Asher MA, Gilbert JL. Correlation of coagulopathy and pulmonary insufficiency with blood transfusion in spinal fusions. Spine 1980;5:1–3. [12] Silliman CC, Paterson AJ, Dickey WO. The association of biologically active lipids with the development of transfusion related acute lung injury; a retrospective study. Transfusion 1997;37:719–26. [13] Masson RG, Ruggieri J. Pulmonary microvascular cytology: a new diagnostic application of the pulmonary artery catheter. Chest 1985; 88:908–14. [14] Urban MK, Sheppard R, Gordon MA, Urquhart BL. Right ventricular function during revision total hip arthroplasty. Anesth Analg 1996; 82:1225–9.
nally called it Spirochaeta pallida; Schaudinn subsequently changed the name to Treponema pallidum. By then, the hypothesis that syphilis was the cause of tabes dorsalis, as proposed by Fournier in 1876 [2], was generally accepted. References
In 1905, Schaudinn and Hoffmann discovered the organism that caused syphilis [1]. They origi-
[1] Schaudinn FR, Hoffmann E. Vorla¨ufiger Bericht u¨ber das Vorkommen von Spirochaeten in syphilitischen Krankeitsprodukten und bei Papillomen. Arb. k. GesundhAmte 1905;22:527–34. [2] Fournier JA. De l’ataxie locomotrice d’origine syphilitique. Paris: G. Masson, 1876.
Pulmonary injury in patients undergoing complex spine surgery Michael K. Urban, MD, PhDa,*, Kethy M. Jules-Elysee, MDa, James B. Beckman, MDa, Khillil Sivjee, MDb, Thomas King, MDb, Webster Kelsey, BAc, Oheneba Boachie-Adjei, MDd a Department of Anesthesiology, Hospital for Special Surgery, 535 East 70th Street, New York, NY 10021, USA Department of Pulmonary and Critical Care Medicine, New York-Presbyterian Hospital, Weill Medical College of Cornell University, 525 E. 68th Street, New York, NY 10021, USA c Department of Anesthesia, Hospital for Special Surgery, 535 East 70th Street, New York, NY 10021, USA d Department of Orthopedic Surgery, Weill Medical College of Cornell University, and Scoliosis Service, Hospital for Special Surgery, 535 East 70th Street, New York, NY 10021, USA
b
Received 19 March 2004; accepted 21 October 2004
Abstract
BACKGROUND CONTEXT: Previous reports have shown that 15% of patients who undergo sequential anterior, then posterior, surgical corrections for spinal deformities demonstrate evidence of acute lung injury. By analyzing the bronchoalveolar lavage (BAL) fluid from these patients for evidence of acute inflammation, we might gain some insight into the etiology of this acute lung injury. PURPOSE: To elucidate the etiology of acute lung injury after corrective surgery for adult spinal deformities. STUDY DESIGN/SETTING: Fifteen adult patients with scoliosis scheduled for elective sequential anterior then posterior corrective (A/P) spinal deformity surgery. PATIENT SAMPLE: Consecutive adult patients with scoliosis scheduled for elective corrective surgery with the author (OBA). OUTCOME MEASURES: Patients were assessed for postoperative respiratory complications by oxygen requirements, continued mechanical ventilation, and radiological evidence of diffuse bilateral interstitial or alveolar infiltrates. An acute pulmonary inflammatory response included the presence of inflammatory cells and elevated cytokines in BAL fluid. METHODS: BAL were performed after induction of anesthesia but before surgery, at the completion of surgery, and on the morning after surgery with the patient still intubated. BAL fluid was analyzed for inflammatory cells and cytokine interleukin-6 (IL-6) and tumor necrosis factor alpha (TNF-α) levels. Patients were assessed postoperatively for increased pulmonary vascular resistance, radiological evidence of diffuse bilateral alveolar infiltrates, and the requirement for ventilatory support beyond the first postoperative day (POD1). RESULTS: The cell counts of BAL fluid demonstrated significant increases in neutrophils, lymphocytes, and lipid laden macrophages (LLMAC) with surgery. The concentration of the cytokines IL-6 and TNF-α also increased with surgery. The elevations in BAL inflammatory cells and cytokine levels correlated positively with increased pulmonary vascular resistance and the requirement for mechanical ventilation. CONCLUSIONS: After A/P spine fusions, patients have evidence of an acute inflammatory pulmonary injury. Several etiologies exist for this finding, including blood and fluid infusions, direct trauma to the lung, a systemic inflammatory response, and the embolization of fat and bone-marrow debris. The presence of LLMAC in the lungs of these patients and the finding that the patient with the requirement for the longest ventilatory support also had the highest BAL LLMAC count, suggest that the embolization of fat and bone debris released from the spine during surgery may be at least
FDA device/drug status: not applicable. Nothing of value received from a commercial entity related to this research.
1529-9430/05/$ – see front matter doi:10.1016/j.spinee.2004.10.049
쑖 2005 Elsevier Inc. All rights reserved.
* Correspondence. Department of Anesthesiology, Hospital for Special Surgery, 535 East 70th Street, New York, NY 10021. Tel.: (212) 606-1792; fax: (212) 517-4481. E-mail address: [email protected] (M.K. Urban)
270
M.K. Urban et al. / The Spine Journal 5 (2005) 269–276
partially responsible for the lung injury. Further studies on the mechanism of lung injury during this procedure are warranted. 쑖 2005 Elsevier Inc. All rights reserved. Keywords:
Deformity surgery; Pulmonary injury; Pulmonary inflammation; Bronchoalveolar lavage; Lipid laden macrophages; Lung cytokine levels
Introduction There is significant morbidity in adult patients who undergo complex corrective procedures for spinal deformities [1]. These procedures are associated with large blood losses which may exceed the patient’s total estimated blood volume. This dramatic blood loss and the associated release of endogenous mediators of the stress response can have significant physiological effects including coagulopathies, myocardial depression, hypotension, and acute lung injury. It was previously reported that pulmonary artery catheter monitoring of patients who undergo complex spinal fusions facilitates the identification of patients who have sustained pulmonary injury [2]. These patients exhibited elevated pulmonary vascular resistance and hypoxemia, which is clinically similar to the fat embolism syndrome described in long bone fractures and hip arthroplasty [3]. The surgical interventions, including spinal instrumentation, thoracoplasties, and osteotomies may be responsible for the embolization of fat and bone marrow debris to the pulmonary microvasculature. However, the pulmonary effects could also be caused by transfusion-related lung injury or the release of a variety of cytokines that are associated with major trauma and sepsis. This study represents an initial attempt to characterize the pulmonary inflammatory response in lung injury during complex deformity spinal surgery.
of the patient was appropriate for spontaneous ventilation. Patients with radiographic evidence of diffuse bilateral interstitial or alveolar infiltrates and hypoxemia (PaO2⬍60 mm Hg on 3 L/m FIO2) were considered to have pulmonary complications. All patients were monitored with a pulmonary artery catheter which was inserted through a 9F cannula in the right internal jugular vein after the induction of general anesthesia but before surgery. The pulmonary artery catheter was used to maintain filling pressures (pulmonary artery occlusive pressure [PAOP]) during the procedure [2]. The mean PAOP for all 15 patients was 10.3⫾1.4 mm Hg at baseline and 10.9⫾1.9 mm Hg at the completion of surgery. An increase in the pulmonary artery diastolic pressure to PAOP gradient is indicative of increased pulmonary vascular resistance. Pain management consisted of subdural morphine injected during the posterior fusion and intravenous patient controlled analgesia hydromorphone. Bronchoalveolar lavages (BAL) were performed while the patients were ventilated. Three separate 60-cc volumes of sterile 0.9% saline were injected through a flexible fiberoptic bronchoscope wedged in the right middle lobe and lingula of the lung and then aspirated back with a syringe. The three lavages were combined for each time sample. A sample was spread on slides and then centrifuged; the cell-free supernatant was stored at ⫺80⬚C for cytokine analysis. The BAL were obtained at baseline (after induction of general anesthesia but before surgery), upon closure (at the completion
Materials and methods With institutional review board approval, 15 adult patients with scoliosis scheduled for elective sequential anterior then posterior corrective spinal deformity (A/P) surgery were enrolled. All of the patients received a preoperative pulmonary consultation which consisted of pulmonary function studies. The surgery consisted of sequential anterior thoracolumbar discectomies with insertion of bone (5–9 levels) and then posterior thoraco-lumbar fusions with DePuy AcroMed instrumentation (9–15 levels). All of the patients required a chest tube after closure of the anterior incision (9/ 15 right thoracotomies), and osteotomies were performed in 9 of the 15 patients (4–6 levels). All patients received the same general endotracheal anesthetic, which consisted predominantly of nitrous oxide and fentanyl, and a fraction of inspired oxygen (FIO2) ⬍30%. Patients were monitored in an intensive care unit for at least the first 24 hours, and sedated and ventilated for at least the first 12 postoperative hours, FIO2ⱕ60%. Positive pressure ventilation was terminated when both the anesthesiologist and pulmonologist determined that the respiratory condition
Table 1 Perioperative patient characteristics Patients (n) Age (yr) Weight (kg) Deformity Scoliosis Kyphoscoliosis Curve⬚ Preoperative RLD (n) LOS (h) Levels fused Anterior Posterior Osteotomies (n) Thoracotomy: Right/Left EBL (mL) Ventilated ⬎1 day (n) Mean pulmonary pressures (mm Hg) Baseline PAD-PCWP Closure PAD-PCWP
15 46⫾12 60⫾9 13 2 68⫾16 9 10⫾2 7⫾2 12⫾3 10 9/6 2437⫾1261 5 2 6
EBL⫽estimated intraoperative blood loss; LOS⫽length of surgery; PAD⫽pulmonary artery diastolic pressure; PCWP⫽pulmonary capillary wedge pressure; RLD⫽restrictive lung disease.
M.K. Urban et al. / The Spine Journal 5 (2005) 269–276
of surgery), and on the first postoperative day (POD1, the morning after surgery). Total cell counts of the BAL fluid were made using a hemocytometer and the differential cell counts on cytocentrifuge preparations of the BAL fluid for neutrophils, lymphocytes, and macrophages for each of the time points. In addition, the macrophages that contained fat-stained globules were counted as lipid laden macrophages (LLMACs). For cytokine levels, the BAL fluid was first concentrated using a centrifugal filter device (Millipore, Amicon Centriplus, model YM-3, Millipore Corp., Billerica, MA). The concentrated BAL samples were assayed for tumor necrosis factor alpha (TNF-α) and interleukin-6 (IL-6) via a commercially available sandwich-type enzyme immunoassay (PeliKine; Research Diagnostics Inc., Flanders, NJ). Because there is variability in the recovery of endothelial lining fluid (ELF) in the BAL samples, we corrected for the dilution of ELF with the saline washes via the urea method of Rennard et al. [4]. The urea concentration in the lavage fluid and plasma was quantified using a colorimetric assay (SigmaAldrich Urea Nitrogen Kit). The ELF cytokine level was calculated using the formula: ELF · Cytokine⫽BAL · Cytokine (Urea · plasma/Urea · BAL). Chest X-rays were taken on each postoperative day for 5 days or until the patient was extubated and the chest tube was removed. Hemodynamic data were collected for all patients during their hospital stay. Data were analyzed by analysis of variance with p⬍.05 considered significant and coefficient of correlation within 95% confidence levels, using the STAT View program.
Results We assessed lung injury in 15 adult patients (46⫾12 years) with spinal deformities after elective sequential anterior then posterior reconstructive spinal fusions (Table 1). All of the procedures consisted of anterior thoracotomies, multiple decompressions and removal of discs, and posterior segmental instrumentation (DePuy AcroMed, DePuy Spine, Raynham, MA). All of the procedures were associated with significant blood loss. Eight of the patients had preoperative pulmonary function tests suggestive of moderate restrictive lung disease. Baseline pulmonary artery pressures were normal in all patients but the pulmonary artery diastolic pressures increased with surgery whereas the PAOP was kept stable, indicative of increased pulmonary vascular resistance. Our BAL analysis indicates that surgery induced an acute inflammatory response in the lungs of all the patients. Baseline counts for neutrophils and LLMACs were very low (1–2 cells per 100) before surgery. The BAL cell counts (per 100 cells) of neutrophils, lymphocytes, and LLMACs
271
increased significantly with surgery (Fig. 1). The concentration of the cytokines TNF-α and IL-6 in the BAL also increased with surgery for all patients (Fig. 2). The inflammatory response with regard to cytokines was consistent within patients, because elevations in one cytokine (IL-6) correlated positively with elevations in the other cytokine, TNF (r2⫽.85, p⬍.001). Also, the complete inflammatory response was consistent within patients, because cytokine levels correlated positively with the number of BAL neutrophils (R2⫽.324, p⫽.03) and LLMACs (R2⫽.356, p⫽.03) on POD1. The inflammatory response, however, did not correlate positively with length of surgery, estimated intraoperative blood loss (EBL), or preoperative demographic characteristics, including age, weight, and the presence of preoperative restrictive lung disease. The number of LLMACs in the BAL on the day after surgery correlated positively (r2⫽.34; p⫽.02) with the number of fused vertebral segments (Fig. 3), but not with the EBL or volume of reinfused scavenged blood. However, the correlation between elevated cytokine levels and fused segments was not as strong (p⫽.12). Patient 8 had significantly higher LLMAC counts than the other patients and required 6 days of postoperative ventilatory support for respiratory failure. Five patients required ventilatory support for longer than 1 day. Three of these five patients (Patients 8, 10, and 12) had TNF-α and IL-6 levels greater than 1 SD above the mean on POD1 (Fig. 4). These three patients had radiographic evidence of acute lung injury, diffuse infiltrates over both lung fields and not limited to the side of the lavage or the surgery. With regard to the other two of the five patients, in Patient 11 the cytokine data were incomplete and in Patient 3 the IL-6 and TNF levels were higher intraoperatively (closure) than on POD-1. Patient 3 also had the highest BAL neutrophil count. In addition, there was a statistically significant association between elevated pulmonary vascular resistance at the end of the procedure (PVR⬎200 dynes/s/ cm⫺5), as well as the requirement for ventilatory support for longer than 1 day and elevated TNF and IL-6 levels (p⫽.01). Patient 8, who required 6 days of postoperative ventilatory support for respiratory failure, had the highest LLMAC counts and IL-6 levels. Patient 8 underwent the largest number of combined anterior and posterior discectomies and fusions.
Discussion After sequential anterior then posterior thoraco-lumbar fusions for spinal deformities, patients demonstrate evidence of acute lung injury [1,2]. This evidence includes radiological changes, elevated pulmonary vascular resistance, and the requirement for ventilatory support. This report supports the hypothesis that the lung injury may be the result of an acute pulmonary inflammatory response. All patients had increased BAL inflammatory cells and cytokine levels with
272
M.K. Urban et al. / The Spine Journal 5 (2005) 269–276
Fig. 1. Cell counts (per 100 cells) of (A) bronchoalveolar lavage neutrophils, (B) lymphocytes, and (C) LLMACs for 15 patients undergoing A/P spinal fusions. Baseline (after intubation, before surgery); closure (after both anterior and posterior spinal surgery), and POD1 (morning of the first postoperative day, patient still intubated). Gray bars⫽baseline; black bars⫽closure; white bars⫽POD1.
surgery. Patients with postoperative respiratory complications, who required ventilatory support beyond POD1, had the highest BAL inflammatory cell counts and cytokine levels. However, our sample size was small and differences between
patients for BAL inflammatory cell counts were not significant, except for LLMACs in Patient 8. Respiratory failure is a common complication of trauma [5]. This respiratory failure can be the result of direct injury
M.K. Urban et al. / The Spine Journal 5 (2005) 269–276
273
Fig. 1. Continued
to the lung or indirect damage from trauma-induced systemic reactions. The indirect lung injury occurs as a result of a generalized increase in pulmonary microvascular permeability to fluids and proteins [6]. These changes are possibly the result of the activation of humoral and cellular mediators of inflammation triggered by the systemic consequences of trauma. A/P surgery involves direct trauma to the lung on the side of the thoracotomy along with injury to the contralateral lung as manifested by diffuse infiltrates in that lung. In our study, 11 of the 15 patients demonstrated postoperative bilateral radiological pulmonary changes. Several studies have demonstrated that the acute lung injury associated with sepsis or trauma is the result of the damage produced by inflammatory cells and cytokines [7]. The BAL fluid of patients with adult respiratory distress syndrome after sepsis syndrome had elevated levels of neutrophils and proinflammatory cytokines (IL-6, IL-8), produced by alveolar macrophages [8]. In addition, the number of neutrophils and the concentration of IL-6 and IL-8 increases in BAL fluid after cardiopulmonary bypass, and there is a positive correlation between some cytokine levels and a decrease in arterial oxygenation [9]. Similarly, other studies support the hypothesis that TNF-α plays an important role in the inflammatory processes involved in aspiration pneumonitis [10].
In this study, level of BAL cytokines and inflammatory cells correlated positively with postoperative respiratory complications. However, patient demographics, the degree of spinal curvature, the number of patients with moderate restrictive lung disease, and the extent of surgery (time and number of osteotomies) did not correlate positively with the pulmonary inflammatory process. Large blood losses and the requirement for multiple transfusions have been associated with pulmonary insufficiency [11,12]. In this study, however, there was not a direct positive correlation between EBL and cytokine levels. Diffuse alveolar permeability changes will occur in patients exposed to barotraumas and oxygen toxicity during mechanical ventilation. However, none of our patients experienced inspired oxygen levels above 60% and the inflammatory process was initiated during surgery, not after days of ventilatory support. Similar pulmonary inflammatory changes have been noted during systemic circulatory collapse (sepsis). However, all of the patients in this study had a stable hemodynamic profile perioperatively. In this study, patients with pulmonary inflammatory changes had significantly higher LLMAC counts, which correlated positively with the number of vertebral segments instrumented. Patient 8 was a 57-year-old female with kyphoscoliosis and moderate restrictive lung disease, who required 11 hours of surgery with 10 anterior discectomies with
274
M.K. Urban et al. / The Spine Journal 5 (2005) 269–276
Fig. 2. The ELF concentration of (A) IL-6 and (B) TNF-α (pg/mL) in BAL fluid from 15 patients after A/P spinal fusions. Gray bars⫽baseline; black bars⫽closure; white bars⫽POD1.
bone insertion and 16 posterior levels fused with segmental instrumentation. This patient with the highest LLMAC count required ventilatory support for 6 postoperative days. Pulmonary microvascular fat has been considered a source of lung
injury in long bone trauma and during hip and knee arthroplasty [3,13,14]. The pulmonary changes which follow the embolization of fat and bone marrow debris during these procedures, result in acute inflammatory lung injury. In
M.K. Urban et al. / The Spine Journal 5 (2005) 269–276
275
Fig. 3. Regression analysis comparing number of vertebral levels fused to number of LLMAC in BAL on POD1. Correlation p⫽.02. LLMAC⫽⫺25.972⫹ 2.417 * Total levels; R2⫽.339.
our previous report we demonstrated that A/P surgical patients sustained increases in pulmonary vascular resistance with surgery, similar to the increases seen during revision total hip arthroplasty [14]. The multiple bone invasive procedures that occur during A/P procedures could serve as the embolic source of fat and bone marrow debris. This process would lead to the recruitment of inflammatory cells and the release of proinflammatory cytokines, which ultimately produce diffuse pulmonary injury. The presence of LLMACs in the BALs of these patients after surgery, however, is only
suggestive of fat embolism. Furthermore, we were unable to determine why only some patients exhibited increased alveolar LLMACs and subsequently developed an inflammatory response. Further studies are required to validate this hypothesis, delineate the patients at risk, and develop preventative treatments. Transesophageal echocardiography has been used to detect and quantify pulmonary emboli (fat; bone marrow debris) during total hip arthroplasty [14]. Future analysis will involve transesophageal echocardiography monitoring during these procedures.
Fig. 4. Scattergram of IL-6 levels in BAL on POD1 in 15 patients after A/P spinal fusions.
276
M.K. Urban et al. / The Spine Journal 5 (2005) 269–276
References [1] Rizzi PE, Winter RB, Lonstein JE, Denis F, Perra JH. Adult spinal deformity and respiratory failure. Spine 1997;22:2517–30. [2] Urban MK, Urquhart B, Boachie-Adjei O. Evidence of lung injury during reconstructive surgery for adult spinal deformities with pulmonary artery pressure monitoring. Spine 2001;26:387–90. [3] Gossling HR, Donohue TA. The fat embolism syndrome. JAMA 1979;241:2740–2. [4] Rennard SL, Basset G, Lecossier D, et al. Estimation of volume of epithelial lining fluid recovered by lavage using urea as marker of dilution. J Appl Physiol 1986;60:532–8. [5] Pepe PE. Acute post-traumatic respiratory physiology and insufficiency. Surg Clin North Am 1989;69:157–73. [6] Putensen C, Waibel U, Koller W, Putensen-Himmer G, Hormann C. Assessment of changes in lung microvascular permeability in posttraumatic acute lung failure after direct and indirect injuries to lungs. Anesth Analg 1992;74:793–9. [7] Warren JS. The role of cytokines in experimental lung injury. In: Kunkel SL, Remick DF, eds. Cytokines in health and disease. New York: Marcel Dekker, 1992:257–69.
One Hundred Years Ago in Spine
[8] Kobayashi A, Hashimoto S, Koogchi K, et al. Expression of inducible nitric oxide synthetase and inflammatory cytokines in alveolar macrophages of ARDS following sepsis. Chest 1998;113:1632–9. [9] Kotani N, Hashimoto H, Sessler D, et al. Neutrophil number and interleukin-8 and elastase concentrations in bronchoalveolar lavage fluid correlate with decreased arterial oxygenation after cardiopulmonary bypass. Anesth Analg 2000;90:1046–51. [10] Davidson BA, Knight PR, Helinski JD, Nader ND, Shanley TP, Johnson KJ. The role of tumor necrosis factor-α in the pathogenesis of aspiration pneumonitis in rats. Anesthesiology 1999;91:486–99. [11] Jacobs RR, Asher MA, Gilbert JL. Correlation of coagulopathy and pulmonary insufficiency with blood transfusion in spinal fusions. Spine 1980;5:1–3. [12] Silliman CC, Paterson AJ, Dickey WO. The association of biologically active lipids with the development of transfusion related acute lung injury; a retrospective study. Transfusion 1997;37:719–26. [13] Masson RG, Ruggieri J. Pulmonary microvascular cytology: a new diagnostic application of the pulmonary artery catheter. Chest 1985; 88:908–14. [14] Urban MK, Sheppard R, Gordon MA, Urquhart BL. Right ventricular function during revision total hip arthroplasty. Anesth Analg 1996; 82:1225–9.
nally called it Spirochaeta pallida; Schaudinn subsequently changed the name to Treponema pallidum. By then, the hypothesis that syphilis was the cause of tabes dorsalis, as proposed by Fournier in 1876 [2], was generally accepted. References
In 1905, Schaudinn and Hoffmann discovered the organism that caused syphilis [1]. They origi-
[1] Schaudinn FR, Hoffmann E. Vorla¨ufiger Bericht u¨ber das Vorkommen von Spirochaeten in syphilitischen Krankeitsprodukten und bei Papillomen. Arb. k. GesundhAmte 1905;22:527–34. [2] Fournier JA. De l’ataxie locomotrice d’origine syphilitique. Paris: G. Masson, 1876.