Spine surgery is one of the fastest growing branches of orthopedic surgery. Patients often present with a relatively high acuity and, depending on surgical approach, morbidity and mortality can be comparatively high. Among the most prevalent and most frequently fatality-bound perioperative complications are those affecting the pulmonary system; evidence of clinical or subclinical lung injury triggered by spine surgical procedures is emerging. Increasing burden of comorbidity among the patient population further increases the likelihood of adverse outcome. This review is intended to give an overview over some of the most important causes of pulmonary complications after spine surgery, their pathophysiology and possible ways to reduce harm associated with those conditions. We discuss factors surrounding surgical trauma, timing of surgery, bone marrow and debris embolization, transfusion associated lung injury, and ventilator associated lung injury.

The number of spine surgical procedures has been increasing dramatically over the last decade[1]. This trend has made this population a prominent group among hospitalized patients, especially since they are accompanied by relatively high medical acuity. Depending on surgical approach, perioperative complications may occur in over 20% of patients, with those requiring thoracic approaches suffering the highest morbidity[2]. Amongst complications, those affecting the pulmonary system are of special concern as they have been linked to high rates of mortality[2]. Indeed, almost half (43.7%) of patients dying after lumbar spine fusion do so with a diagnosis of pulmonary compromise. Adult respiratory distress syndrome (ARDS) occurs in up to 3% of patients and will increase the risk of in-hospital death by more then 6-fold[2]. This is especially of concern as pulmonary complications have been on the rise amongst an increasingly comorbidity ridden spine surgical population[1]. Further, studies suggest that a large number of spine surgical patients show evidence of lung injury, albeit often subclinical[3,4].

A better understanding of the associated pathophysiology could allow for improvements in outcome. While the mechanisms leading to the development of pulmonary complications are likely multifactorial, they do include direct mechanical trauma due to parenchymal contusion, embolization of marrow material into the lung, ventilator associated lung injury and the entity of transfusion related lung injury[3-8]. In this article, we will review potential mechanisms contributing to lung injury and point to possible ways to reduce it.

The most important indications for spine surgery include degenerative disc disease and associated pain and disability, followed by stenosis, scoliosis and spondylolisthesis. Other indications include trauma and neoplastic disease. In the time period between 1988 and 2001, the number of spine fusions in the United States increased from some 24 000 to more than 120 000 procedures per year[9,10]; degenerative disc disease as a primary diagnosis was present in almost two thirds of cases (65.3%) by 2001[11]. Taking a closer look at the available data, the rate of surgery increased by approximately 100% during the 1980s and more than 220% during the 1990s. After 1996, when intervertebral fusion cages received approval by the food and drug administration, an exponential increase in case burden was seen. In 2001, depending on age group, as many as 60 to 90 per 100 000 patients underwent cervical, thoracic or lumbar spine fusion[11]. The most dramatic rise was evident in the group of patients aged 60 or older in all procedures except cervical spinal fusion, which peaked in the group aged 40 to 59 years[9]. Rates of spinal fusion surgery have been maintaining their steady increase during the first decade of the 21st century, surpassing 110 per 100 000 patients in 2003, while at the same time exhibiting considerable regional variability[12]. In contrast, the volume of discectomy and laminectomy procedures stagnated, but the fraction of outpatient procedures among these increased[13]. Given the increasing age of patients and preferred treatment patterns for degenerative disease[14], the upwards trend of spinal fusion procedures incidence is likely to perpetuate in the next years[1].

The aforementioned developments, which essentially represent a shift towards a rising number of procedures with increased invasiveness being performed among an older, comorbidity ridden population, necessitate a close analysis of outcomes and associated risk factors. Procedure-related immediate complications occur in up to 20% and in-hospital mortality rates of 0.2%-0.5% have been recorded[1]. These numbers are contingent upon a number of independent risk factors, including male gender, advanced age, surgical approach (anterior vs posterior vs anterior/posterior approach, the latter being associated with the highest odds for morbidity and mortality), and preexisting comorbidities[15]. Most significantly associated with a higher mortality are known diagnoses of congestive heart failure, liver disease, coagulopathy, neurologic disorders, renal disease, electrolyte imbalances and pulmonary circulatory disease. Aside from device-related adverse events, respiratory complications rank among the most common and most serious procedure-related complications after spine surgery. They occur in up to 3.82% of subjects, especially after anterior and anterior/posterior surgical approaches and are associated with high mortality. Patients are fundamentally more prone to suffer in-hospital death after having developed ARDS [odds ratio (OR) 5.85], pulmonary embolism (OR 8.17) or other lung-associated conditions (OR 1.45)[2]. These ORs represent the relative adjusted risk for in-hospital mortality in patients who suffered from the complication, compared to those who did not (1 = reference).

Intuitively, surgical extent and invasiveness are closely correlated with the risk for adverse events. It has been shown that more extensive procedures, especially those necessitating invasion of the thoracic cavity, are burdened with higher rates and risk of complications[2]. Indeed, approximately 3% of patients will develop ARDS after anterior/posterior spine surgery compared to approximately 1% after a posterior or 1.6% after an anterior approach only[2].

Possible explanations include contusion and direct mechanical trauma due to the invasion of the thoracic cavity and displacement of the lung to gain access to the spine. However, it remains unknown if simple displacement of the lung offers any advantage to lung isolation and collapse during surgery as both may be burdened with disadvantages. In the first case, direct mechanical trauma may lead to tissue damage and its sequelae, while in the latter the phenomenon of re-expansion pulmonary edema may become a source of concern[16]. In addition, the effect on the contralateral lung has to be kept in mind in the context of pathologic events summarized under the term “down lung syndrome”[17].

Similarly to events described during joint arthroplasty procedures, evidence is emerging that the intravasation of marrow and bone debris during the instrumentation process may contribute to perioperative lung injury[18]. Here, the embolization of material into the pulmonary vasculature has been associated both with parenchymal damage and increases in pulmonary vascular resistance in a dose dependent manner[19,20]. In the spine surgical setting, Urban et al[6] was able to demonstrate an adverse pulmonary effect of perioperative events in the form of an increase in pulmonary vascular resistance in 15% (8/55) of patients, usually during or after posterior instrumentation. In a follow-up study, the same author analyzed bronchoalveolar specimens and linked the presence of lipid-laden macrophages to possible embolization of fat and debris entering the bloodstream during the surgical procedure[5]. This mechanism of lung injury is supported by echocardiographic studies, in which 80% of spine surgery patients experienced moderate to severe embolic events during instrumentation of the spine[7]. Markers of lung catabolism showed a significant increase in the postoperative period compared to baseline[4].

While the load of embolic material may be a determinant of the extent of lung injury, the patient’s ability to compensate for the pathophysiologic changes is almost certainly a major factor in the ability of patients to compensate for the insult. In this context, we previously reported a dramatically increased risk for mortality in those patients with pre-existing pulmonary hypertension {OR 8.37 [confidence interval (CI) 5.95-11.78]}[2].

It has been speculated that an increase in pulmonary vascular resistance may lead to increased right ventricular and atrial pressures and thus decreased venous return. This in turn may promote arrhythmias, secondary to right heart dilatation and the formation of venous thrombosis and embolic events[21].

Significant blood loss is an important issue in major spine surgery and frequently necessitates replacement of blood products. Although the risk of infectious disease transmission through transfusion has declined substantially, other complications must be kept in mind. These complications include immediate and delayed hemolysis, febrile and allergic reactions, systemic inflammatory response syndrome, disruptions of coagulation, electrolyte and acid-base household as well as other immune-system mediated conditions like graft-versus-host disease and, most significantly, transfusion-associated acute lung injury (TRALI). TRALI has emerged as the leading cause for transfusion-associated mortality[22]; it is defined as new acute lung injury occurring during or within 6 h after a transfusion. All available blood products have been reported to cause TRALI; those with high plasma content seem to be involved more frequently. While initial reports stated a per component incidence between as high as 1:432 for whole blood platelets and 1:557 000 for red blood cells[22], recent efforts to screen for and exclude high risk cases both on the sides of donor and recipient seemed to succeed in reducing the incidence of TRALI. A reduction in TRALI incidence rate for plasma transfusion from 1:4000 in 2006 to 1:12 000 in 2009 has been reported[23]. Clinical and diagnostic features of TRALI include dyspnea, hypoxia, fever, tachycardia, bilateral pulmonary edema without evidence of congestive heart failure or volume overload and bilateral fluffy infiltrates on chest radiograph[24]. In contrast to the non-immunologic transfusion-associated circulatory overload-an important and very similarly presenting differential diagnosis-TRALI is probably caused by antibodies adhering to neutrophils in the recipients’ pulmonary epithelium. These activated neutrophils release oxidases and other reactive substances provoking damage to the capillary membrane and, subsequently, capillary leak and pulmonary edema[25]. For about 15% of cases in which no antibodies could be found, other sources of capillary damage have been suggested, including involvement of lymphocytes and monocytes, cytokines, lipid priming molecules and endotoxins. Treatment of TRALI is primarily supportive; ventilatory support and supplemental oxygen is required in most cases, while a beneficial influence of steroid administration is not proven. Diuresis is not recommended[26]. Unlike in ARDS caused by other pathologies, patients usually recover quickly after TRALI, most within 96 h of the transfusion, and mortality is relatively low, ranging between 5% and 10%[24].

Massive transfusion has been defined as administration of more than 10 units of packed red blood cells in a 24 h period[27]. One of the most important problems arising with this entity is the alteration of intravascular blood homeostasis, resulting in coagulopathy, electrolyte imbalance, acidosis and hypothermia, which can all be difficult to counteract and give rise to a number of associated conditions on their part. Particularly imbalances of coagulation as well as pre-existing coagulopathy have been associated with a high predictive risk for mortality after spine fusion [OR 5.46 (CI 4.34-6.86)][2]. Further, as the risk for development of complications is proportionate to the number of units transfused, a cautious transfusion regimen can be advantageous wherever possible[28]. In terms of survival and improved surgical outcome, a liberal transfusion regimen has recently not proven beneficial even in high risk patients after hip surgery, another form of orthopedic surgery where high blood loss is frequently an issue[29].

Mechanical ventilation is necessary during the perioperative period in order to provide sufficient respiration and oxygenation to the anesthetized patient. Historically, tidal volumes as high as 15 mL/kg ideal body weight (IBW) have been utilized, generating mechanical stress on the lung parenchyma. These high tidal volumes were believed to ensure bronchiolar patency and avoid atelectasis[30]. Only recently studies suggested that the usage of tidal volumes in a physiologic range (6 mL/kg IBW) could prevent pulmonary adverse outcomes including acute lung injury and ARDS[31]. In principle, mechanical ventilation is thought to generate physical damage through overdistension (volutrauma) and excessive transpulmonary pressure (barotrauma) as well as pulmonary inflammatory response and disruption of structural element metabolism[32]. During intrathoracic procedures this adds up to surgically induced trauma and the extent of pulmonary injury is likely contingent upon the accumulated total tissue interference. A number of recent studies suggest that a protective ventilation strategy can contribute to beneficial outcome of spine surgery[8]. Efforts to quantify the trauma by measuring markers of inflammation and elastin catabolism, including plasma cytokines, myeloperoxidase and elastin in bronchoalveolar lavage fluid and desmosin in the urine, showed conflicting results. However, a substantial rise in inflammatory markers after surgery could be noted irrespective of which ventilation approach (i.e., low or traditional tidal volumes) was chosen[33].

One-lung ventilation (OLV), achieved using bronchial blockers or double-lumen endobronchial tubes, improves intrathoracic visualization for the surgeon while reducing the risk of trauma inflicted through mechanical interference with the lung. Although the cardiac output shunt fraction increases drastically as soon as one lung is excluded from ventilation and gas exchange, satisfactory oxygen saturation (SpO₂) levels above 90% can be maintained in up to 80%-90% of cases even with low to medium sized tidal volumes[34]. Other complications of OLV include tracheal or bronchial injury during placement of the double lumen tube, re-expansion pulmonary edema[16] and adverse effect on the contralateral lung through ventilation-, surgery- or position-associated additive stress[17].

It has been well established that more extensive surgery harbors increased risk of pulmonary complications, with combined anterior and posterior procedures being associated with highest risk. To date it remains controversial if performing both portions of surgery in one or in separate sessions can modify this risk. Studying this question poses extensive limitations as sufficient sample sizes are difficult to achieve in order to provide sufficient statistical power.

In a study utilizing nationally representative data from 113 991 cases, we found increased rates of pulmonary complications in those undergoing staged procedures performed during the same hospitalization (5.0% vs 3.3%, P < 0.0001). While controlling for surgical indication, patient demographics including comorbidity burden, many clinical factors could not be accounted for in this analysis, thus limiting the interpretation of data. Although we cannot exclude that staged cases may have been more invasive, the findings suggest that the performance of a second surgery during the time of increased perioperative systemic inflammation may contribute to higher rates of complications. This is a finding also observed in staged bilateral total knee arthroplasties, which share similar pathophysiologic set ups[35].

A myriad of factors that individually, but also cumulatively contribute to the development of pulmonary complications after spine surgery has been described. Therefore, a multimodal approach is required to effectively counteract these factors and thereby improve outcomes.

Spine pathologies impart a large socioeconomic impact on our health care system. Given an increasingly ageing population with comorbid conditions, the number of spine surgery procedures must be expected to rise. Pulmonary compromise is seen in almost half of patients who die following spine surgery. Thus, pulmonary complications and associated risk factors deserve special consideration in preoperative planning and patient selection. Strategies to proactively improve patient outcome include careful preoperative identification of patients with pulmonary and/or systemic comorbidity, stratification of patients by risk, and, subsequently, selection of appropriate surgical and anesthesiologic management. A multimodal approach is required to effectively counteract factors that contribute to the development of pulmonary complications after spine surgery.