Thoracic Anesthesia


Thoracic Anesthesia Practice


Lung Resections for Cancer and Benign Chest Tumors

Mark Stafford-Smith

Key Points

1. Although many of the challenges to the anesthesiologist posed by lung resection surgery are similar to those with other surgeries, acute major hemorrhage is one that is particularly lethal and requires serious preparation for every case.

2. Lung resection surgeries are a highly morbid group of procedures, with mortality rates that are equivalent to or exceed elective coronary artery bypass surgery. Notably, a significant number of the serious complications of lung resection occur beyond the immediate surgical period and are related to postoperative respiratory insufficiency.

3. The anesthesiologist makes many decisions perioperatively that influence respiratory function and can conceivably contribute to postoperative insufficiency. It is imperative in caring for lung resection patients that the anesthesiologist be conscious of these issues and avoid any unwitting contribution to the burden of risk for respiratory impairment and failed tracheal extubation after lung resection surgery.

Clinical Vignette

The patient is a 59-year-old man with a 150 pack-year history of cigarette smoking. After being treated with antibiotics for a persistent productive cough, his sputum has become blood tinged over the past 2 weeks, and a chest x-ray revealed a right upper lobe coin lesion.

Health background includes longstanding hypertension, an anxiety disorder, and peripheral vascular disease, for which he underwent a left femoral-popliteal artery bypass 1 year ago. Current medications include lisinopril, atenolol, aspirin, and alprazolam.

Vital signs: BP 189/88 mm Hg, HR 55, room air SaO2 92%.

Laboratory investigations are notable for white blood cell count 12.1 and prothrombin time 14.0 seconds (normal 12.5-13.8). Pulmonary function tests are notable for a FEV1 of 50% predicted, FEV1/FVC 60%, and DLCO of 45% predicted.

In the past two decades, significant research and innovation has improved both therapy and prognosis for lung cancers and benign tumors. Medical gains in imaging, better timing, prescription, and selectivity of radiation and chemotherapy have complemented surgical advances, including routine tumor staging, port access video-assistance, titanium staplers with scalpel blades, and more targeted operations designed to preserve unaffected lung tissue. Anesthesia advances have kept pace, with better lung isolation methods and a broadened pharmacologic armamentarium providing an enhanced flexibility that combines safe surgery with multiple options for postoperative analgesia and prompt wake-up and extubation, even for patients with limited respiratory reserve or when a procedure is terminated prematurely. Notably, many of these improvements have expanded the candidate pool for lung resection to include patients who would previously have been ineligible due to their marginal lung function.

Despite advances, perioperative morbidity and mortality rates for lung resection still exceed those for many major procedures (eg, aortocoronary bypass surgery), and few dispute the important role of the anesthesiologist’s actions in influencing patient outcome.1-3 The aim of this chapter is to address and integrate numerous elements of thoracic anesthesia, some outlined in more detail in other chapters, which combine to optimize anesthesia provision for lung resection surgery for cancer and benign chest tumors.


Over 170,000 primary lung tumors are diagnosed each year in the United States, with the majority being malignant (>95%). Malignant lung tumors are the largest source of cancer-related deaths in the United States. Cigarettes increase lung cancer risk for the average smoker by approximately tenfold, twentyfold for heavy smokers. Other inhalation exposures act alone or can compound smoking risk, including radiation (eg, radon, uranium), asbestos, nickel, chromate, mustard gas, arsenic, beryllium, iron, and vinyl chloride. Lung cancer is about twice as common in men as women (74 vs 31 per 100,000 per year), presenting most often during the sixth and seventh decades of life. While inhalation exposures are the major risk for lung cancer occurrence, in victims less than 40 years old (<2% patients) genetic vulnerability is also likely important.4

First symptoms of a primary lung cancer may include productive cough, hemoptysis, weight loss, pain or dyspnea, and less commonly clubbing, superior vena cava syndrome, Horner syndrome, muscle weakness, peripheral neuropathy, or ataxia. The most common tumors related to smoking are squamous and small cell and less frequently adenocarcinoma (Table 13–1). At diagnosis, over 80% of small cell tumors are metastatic, whereas less than half of squamous and adenocarcinomas have spread. While most benign tumors are resectable, only 30% of malignant primary tumors are still sufficiently localized to potentially benefit from surgery.

Table 13–1. Classification of Lung Tumors Commonly Presenting for Lung Resection


Surgical resection of lung metastases improves survival in some situations. In the absence of other spread, one or several lung metastases can be removed on one or repeated occasions, often involving a simple wedge resection achieved using VATS surgery. Disease states include melanoma, soft tissue and osteo-sarcomas, germ cell tumors, colorectal, renal cell, uterine and breast cancer, and squamous cell cancers of the head and neck.5


Minimizing loss of healthy tissue is a logical part of surgical planning for any lung resection. Limited procedures such as localized wedge or segmental lung resection are often all that is necessary for benign tumors. For cancerous lesions, evidence of tumor spread from noninvasive (eg, chest x-ray, CT, and PET scan) or invasive procedures such as bronchoscopy and mediastinoscopy provides the most important guide in avoiding unhelpful lung resection surgery. When mediastinoscopy reveals ipsilateral mediastinal lymph-node spread, subsequent response to induction chemotherapy (eg, cis-platinum, paclitaxel), evidenced by a negative re-mediastinoscopy, still indicates eligibility for curative surgery, and long-term outcomes are improved.6

When tumor spread is unlikely, bronchoscopy and mediastinoscopy may be scheduled at the same time as lung resection, but the anesthetic plan must always be able to adapt to early termination if staging samples return positive for cancer. Of practical clinical significance with early termination is not to have used agents with prolonged effects that delay wakeup (eg, muscle relaxants) or requireprolonged observation (eg, neuraxial opioids). Delayed discharge home is particularly distressing for the patient coping with the news of their cancer spread.

Early stage cancerous tumors are treated by complete resection of the involved lung lobe. However, some early stage tumors, by reason of more extensive local spread or their relationship to major airways, are ineligible for lobectomy and have traditionally been candidates for pneumonectomy. Examples of lung-sparing alternatives to pneumonectomy for selected individuals include bi-lobectomy and upper lobe/sleeve bronchus resection with reattachment of the lower lung lobes. In high-risk patients with limited pulmonary reserve, localized wedge or segmental lung resection may be all that is possible even for malignant disease.

Video-assisted thoracic surgery (VATS) is taking an ever expanding role in the approach to lung tumor surgery. The introduction of VATS surgery for lung resection has been associated with good results and a reduction in complication rates (Table 13–2),7,8 although some institutions have not seen improvements in outcome.9 VATS procedures are particularly reliant on perfect lung isolation and place added responsibility on the anesthesiologist in this regard. Lung resection more extensive than lobectomy is generally not eligible for VATS, in part due to the disproportionately small size of the port incision relative to the resected tissue that must be extracted through it. However, even small tumors are sometimes so placed as to be ineligible for resection by a VATS approach (eg, hilar). Studies indicate that appropriate use of VATS procedures capably achieve surgical goals with lower dehiscence, bleeding and infection rates, reduced pain, and faster recovery. Compared to open thoracotomy, VATS also often reduces the need for rib spreading and the possibility of rib fractures. Other differences between VATS and open thoracotomy for equivalent procedures is the tendency for less pain and blood loss and quicker recovery from VATS surgery. Sources of morbidity and mortality following lung resection surgery vary by procedure and surgical approach, increasing with more extensive lung resection and open (vs VATS) surgical approaches (see Table 13–2).

Table 13–2. Complication Rates (%) in a Population of 1079 Patients Undergoing Lobectomy Lung Resection by Conventional Thoracotomy (n = 382) and video-assisted thoracoscopic surgery (VATS), n = 697 Approaches



Beyond standard preoperative assessment, patients scheduled for lung tumor procedures commonly have considerations specific to their presenting condition and planned surgery. Perioperative risk may be increased by issues related to the origin of their cancer (eg, smoking), but tumor-derived concerns can also contribute to risk (eg, paraneoplastic syndromes). For lung tumor surgeries, a detailed assessment of airway, bleeding risk, and eligibility for neuraxial procedures is particularly relevant.

Smoking increases the risk of lung cancer, and also chronic bronchitis, reactive airway disease, and obstructive lung disease. Some patients present for surgery so affected by these accompanying conditions as to require supplemental home oxygen therapy. For these patients, preparation should include pulmonary function data to quantify their respiratory impairment and a plan for preoperativeoptimization. In patients with poor respiratory function, preoperative arterial blood gas assessment can also inform postoperative management (Figure 13–1). Risk for atherosclerotic heart and vascular disease is also increased in smokers, as is cor pulmonale, supraventricular tachyarrhythmias, and atrial fibrillation. Assessment of cardiac risk follows standardized protocols as for any nonscardiac surgery patient.10


Figure 13–1. Evidence of an elevated arterial partial pressure of carbon dioxide (PaCO2) while breathing spontaneously is strongly suggestive that a patient has very impaired pulmonary function, as represented in this data by severely impaired forced expiratory volume in 1 second with maximal effort (FEV1). The shaded area represents normal PCO2 values.

Paraneoplastic syndromes sometimes are a presenting symptom of lung cancer. Tumor-mediated autoantibodies to calcium channels and Purkinje cells can cause Lambert-Eaton myesthenic syndrome and subacute cerebellar degeneration, respectively.11 The former involves muscle weakness that may be symptomatic, but whose diagnosis is sometimes missed until an exaggerated muscle relaxant response requires postoperative ventilator management; the latter is a degenerative neurologic disorder characterized by broad ataxic gait and nystagmus without other gross neurologic deficits. Lambert-Eaton syndrome contrasts with myesthenia gravis by its involvement of proximal more than distal limb muscles and improved strength with repetitive movements but not in response to acetylcholinesterase inhibitor therapy (eg, neostigmine). Other lung tumor-related hormone effects include hyponatremia due to inappropriate anti-diuretic hormone secretion, Cushing syndrome from excess adrenocorticotrophic hormone, and hypercalcemia from parathyroid hormone release. Among patients with benign lesions, those with neurofibromatosis-1 are notable for the frequent coexistence of neuroendocrine tumors including pheochromocytoma (up to 6% patients, and 20%-50% of those with associated hypertension) and carcinoids (up to 10% patients).12-14 Although neurofibromatosis-1 is associated with an increased lung cancer risk, VATS in these patients is usually performed for neurofibroma resection with the offending tumor most often protruding from an intervertebral foramen. Presumably due to tumor friability and location, epidural hematoma with paraplegia can complicate neurofibroma resection and must be considered in the risk/benefit analysis for neuraxial analgesia.

Identification of patients with increased bleeding risk and those who may be ineligible for spinal/epidural analgesia (see Chapters 6924) requires careful assessment of bleeding history (eg, with tooth extraction), chronic drug therapies (eg, clopidogrel), and review of coagulation tests. Neuraxial procedure guidelines are available regarding acceptable coagulation parameters and timing of anticoagulation cessation.15 The plan for coordinated epidural catheter removal and postoperative thromboprophylaxis must also be formulated preoperatively to minimize spinal hematoma risk. In some patients presenting with conditions requiring chronic warfarin anticoagulation (eg, atrial fibrillation), heparin “bridging” therapy may be required until just prior to surgery. Concurrent infection and anatomic spinal abnormalities are also factors in determining suitability for spinal/epidural procedures.

Management of chronic drug therapies must involve attention to respiratory depressant effects, particularly in patients with marginal pulmonary reserve. Delayed tracheal extubation is a strong predictor of poor outcome, and agents such as extended release opioids can sometimes take partial blame for this complication after lung resection. Even depressant effects from well-intentioned but ill-considered “modest” preanesthetic intravenous sedative (eg, midazolam for anxiety) can complicate lung resection in high-risk patients; by compounding the respiratory depression due to agents that are arguably of more importance at the end of surgery, such as opioids for analgesia. In these circumstances, patient reassurance for anxiolysis, including an explanation of the need to limit preoperative sedation, is often sufficient, but if pharmacologic intervention is deemed essential (eg, during complicated epidural placement), then in the author’s experience, for an average-sized adult, small doses of a short acting intravenous sedative (eg,10-20 mg propofol) alone or potentiated by a very small dose of longer acting agent (eg, 0.25-0.5 mg midazolam) are generally very effective.



Preparation for even the most “straightforward” lung surgery demands sufficient monitoring and vascular access to appropriately respond to complications that can occur, most notably including significant hemorrhage, an uncommon but ever-present risk for any intrathoracic procedure. Beyond standard monitoring requirements and two large bore peripheral intravenous catheters, invasive lines generally include a peripheral intra-arterial catheter for continuous blood pressure recording and repeated blood gas assessment. A right-sided radial arterial line is convenient if staging procedures are also planned since this location also facilitates recognition of innominate artery compression during mediastinoscopy (see Chapter 10).

Neuraxial injection or catheter placement prior to anesthesia induction (eg, epidural catheter insertion) most conveniently occurs following placement of peripheral intravenous and intra-arterial catheters but prior to central venous access. Such a sequence provides intra-arterial monitoring for recognition of intravascular or intrathecal injection of an epidural catheter test-dose and avoids the cumbersome movement required to have a patient transfer into the sitting or lateral position with a central venous line in situ.

While central venous access is not essential for lung resection surgery, patient and procedural factors such as comorbidities (eg, cardiac history), bleeding risk (eg, redo thoracotomy), and the likelihood of postoperative pulmonary edema (eg, pneumonectomy) may warrant such monitoring. If placement of a central venous catheter is deemed necessary, selection of the side ipsilateral to the operative lung is highly preferable for subclavian or internal jugular central venous puncture, due to possibilites such as unrecognized bullous lung disease and the potential for tension pneumothorax in the nonoperative lung. “Down-side” tension pneumothorax can be lethal when it manifests intraoperatively and is difficult to detect. One way is to attach a stethoscope earpiece to a suitable esophageal temperature probe and use it as an esophageal stethoscope. In the case of a “down-side” tension pneumothorax, the chest is completely silent on manual inflation of the reservoir bag.

Bladder catheter placement is not essential for all lung resection procedures and can occur after anesthesia induction. Urinary output monitoring provides some information regarding intravascular volume status in the absence of central venous pressure data and should always be used to avoid the possibility of urinary retention in patients with postoperative epidural analgesia.

Standard preinduction considerations include administration of intravenous antibiotic prophylaxis within 1 hour prior to surgical incision and planning for postoperative disposition (PACU vs stepdown vs ICU observation). Postoperative analgesia strategy often influences disposition when continuous epidural infusions are used, since the care team must be equipped and trained to recognize and treat potentially serious complications associated with their use such as hypotension from local anesthetic-mediated reductions in sympathetic tone and delayed respiratory depression due to cephalad spread of neuraxial opioids (see Chapter 6).

The potential for major hemorrhage with lung surgery is partially explained by the thin walled and high flow characteristics of the pulmonary arterial tree, making these vessels both vulnerable to injury and difficult to repair, with the same capacity for rapid blood loss as major systemic arteries. The uncommon but serious bleeding complication with lung resection mandates preparatory steps in addition to good intravenous access and monitoring. Confirmation that blood products are available and/or in the operating room is essential immediately prior to surgery. Routine availability of colloid volume expanders, use of fluid and patient warming technology, and a rapid transfusion device immediately available or nearby should also be considered.

Nonetheless, the liberal availability of blood products required in preparation for lung resection surgery must be accompanied by thoughtful application of transfusion “triggers” and strict avoidance of unjustifiable blood product administration. Poorer outcomes with “unnecessary” transfusion including pulmonary complications are particularly relevant to lung surgery. Also, a note of caution is warranted regarding the potential for resuscitation “overshoot” that can occur in any response to acute hemorrhage. An effort to keep scale in the resuscitation of hemorrhage, assisted by good communication with the surgeon, will help avoid the lung edema that can occur from fluid overload.

Intraoperative Management


Anesthesia induction should be immediately preceded by meticulous preoxygenation; note that compared to the standard instruction to “take a deep breath in,” prolonged exhalation to expel room air priorto each inhalation of enriched oxygen is actually more efficient, particularly in patients with obstructive lung disease. Immediately prior to anesthesia, induction is also a convenient moment to test and/or supplement any local anesthetic block that may be developing. Standard anesthesia induction agent selection to achieve hypnosis, paralysis, and blunting of the hemodynamic intubation response should be matched to the characteristics of the lung resection procedure and patient, as highlighted above. Becoming experienced with drug selections to minimize lingering respiratory depressant effects, even for patients at low risk for delayed tracheal extubation, pays dividends when a patient is unexpectedly sick or at high risk. For example, midazolam is not essential for anesthesia induction and, while not routinely recommended, even intravenous opioids to blunt the intubation response (eg, fentanyl) can be replaced by a bolus of intravenous lidocaine (1.0-1.5 mg/kg) for the most impaired patient, particularly if postoperative continuous epidural analgesia is planned.

Choice and dose of muscle relaxant must be made to optimize conditions for tracheal intubation but also to respect any potential for a shortened procedure. When bronchoscopy and mediastinoscopy precede lung resection, a single-lumen endotracheal tube with minimum internal diameter of 7.5 to 8.0 mm is used to allow passage of an adult bronchoscope. Otherwise, tracheal intubation requires lung separation to allow for operative lung deflation and nonoperative lung ventilation.

Double-lumen endotracheal tubes provide excellent lung collapse and versatile lung isolation in most circumstances, with bronchoscopic confirmation of positioning. The course of the left mainstem bronchus under the aortic arch provides the most predictable and uninterrupted airway “landing zone” for endobronchial cuff placement. A left-sided tube is the preferred method unless tube proximity to the surgical field obliges an alternate strategy (eg, left pneumonectomy). Adult left- and right-sided double lumen tubes are identified by their endobronchial tube caliber; 32, 35, and 37 are common female adult sizes, while 37, 39, and 41 French are available for men. Both undersized and oversized double lumen tubes may complicate lung isolation and even cause difficulties with tracheal intubation.16Correct endotracheal tube size selection is therefore critical (see Chapter 5).

Bronchial blocker options for lung isolation are sometimes preferable. However, particular effort must be directed to assuring good lung deflation for surgery with this technique, including, (1) the earliest possible blocker inflation time to maximize absorption, and (2) prior to balloon inflation, an extended period of apneic exhalation (>60 s) following ventilation with 100% oxygen, to minimize residual lung volume and hasten gas absorption, respectively. Proximal balloon migration in the airway sufficient to lose lung isolation is common with bronchial blockers. One approach to reduce the frequency of this problem is to trap the inflated balloon at the junction of the upper lobe and mainstem bronchus. While effective, this approach should not be used for upper lobectomy surgery due to the risk of including a piece of the balloon in the resected specimen.

Secure positioning of the patient in the lateral decubitus position after anesthesia induction with the operative side up is typical for most lung resection procedures. Additional reverse-Trendelenberg and reflex positioning of the operating table is common to spread the rib spaces and level the operative chest wall. Safe positioning must include inspection of pressure points, particularly of the down-side arm, including the brachial plexus in the axilla, radial nerve at the humerus, and ulnar nerve at the elbow. In addition, reconfirmation of lung isolation in the lateral position and support of the weight of the ventilator circuit to prevent drag on the endotracheal tube is prudent.

Mild hypotension with anesthesia induction is ubiquitous but can be more common and marked in lung resection patients, often due to an evolving sympathetic epidural block, but more of concern if caused by the interaction of positive pressure ventilation with severe obstructive lung disease. Recognition of so called “dynamic hyperinflation” from inefficient exhalation (also known as breath “stacking” or “auto-PEEP”) is critical both for its hemodynamic implications, but also its potential for serious complications during a procedure (eg, barotrauma, “down-side” tension pneumothorax, cardiac arrest).17While low levels of hyperinflation are common in patients with severe obstructive lung disease or asthma, hemodynamically significant dynamic hyperinflation is uncommon. In the author’s experience, this type of dynamic hyperinflation is best diagnosed during a period of complete cessation of manual or mechanical positive pressure ventilation—a gradual rise of 10 to 20 mm Hg systolic blood pressure over a 1 to 2 minute apneic period in the absence of other stimuli is characteristic. In patients with dynamic hyperinflation, ventilator settings should be titrated to minimize breath stacking; suggested changes include eliminating positive end expiratory pressure (PEEP), adjusting ventilator settings to extend expiratory and reduce inspiratory breath times (eg, 1:4 or 1:5 ratio), and even modestly reduce breath depth and frequency (mild “permissive hypercapnia” is generally well-tolerated). If bronchoconstriction is contributing to dynamic hyperinflation, interventions such as bronchodilator therapy and inhaled volatile agent may also be helpful. Rarely, vasoactive infusions and/or fluid bolus interventions are required to support blood pressure. Periodic arterial blood gas assessment and episodic reinflation of the operative lung for ventilation, coordinated with the surgical team, may also be required for safe management of these patients. Intraoperative transesophageal echocardiography may even occasionally be considered to assess ventricular filling.


Anesthesia for lung resection requires provision of optimal surgical conditions including the need for prolonged one-lung ventilation, while maintaining an extended period of hypnosis, muscle relaxation, and analgesia (in continuum with the postoperative period). Monitoring to minimize any chance of recall is also important (eg, bispectral index or end-tidal volatile agent monitoring), particularly with drug selection being biased away from amnestic agents. Ventilator settings must account for transitions from two to one-lung ventilation, including a goal to minimize barotrauma (eg, peak inspiratory pressures below 30 cm H2O). Finally, strategies that keep infused perioperative fluids and inspired oxygen levels to a minimum are embraced by many thoracic surgery teams.

Major decisions in selecting agents for anesthesia maintenance include the relative roles of regional, intravenous, and inhalational anesthesia. So-called “balanced anesthesia” refers to either inhalational or infused intravenous agents to achieve “light” general anesthesia, combined with local anesthetic epidural block for surgical chest wall anesthesia. Potential advantages of balanced anesthesia include reduced need for hypnotic agents and pre-emergence establishment of analgesia. For patients with severely limited respiratory reserve who are considered at extremely high risk for failure to achieve tracheal extubation, a balanced anesthesia strategy combining intraoperative and postoperative local anesthetic epidural infusion and intraoperative light general anesthesia without the use of benzodiazepines and opioids may be useful methods to consider.

Inhalational and total intravenous anesthetic (TIVA)-based maintenance strategies without local anesthetic blockade comfortably achieve general anesthesia for most thoracic surgery patients but compared to a balanced anesthesia approach require more specific planning for transition to postoperative analgesia. Agents selected for a primarily inhalational anesthetic, such as desflurane and sevoflurane, have a low blood/gas partition coefficient that facilitates rapid emergence. Notably, the recurring requirement for 100% oxygen and the potential for air emboli surgery are clear reasons for the rare use of nitrous oxide during lung resection, although associated immune suppression also runs counter to the goals of cancer surgery. Intravenous infusion agents for TIVA are also those with “effervescent” properties that facilitate prompt emergence, such as propofol and remifentanil. Both inhalational and TIVA techniques can be combined with postoperative analgesia approaches involving longer acting opioids by intravenous bolus or epidural infusion, and/or direct local anesthetic infiltration of intercostal nerves, port wounds, and chest tube exit sites.

Purported advantages of TIVA over inhalational strategies include avoidance of volatile agent-mediated inhibition of hypoxic pulmonary vasoconstriction, thereby reducing V/Q mismatch and oxygen desaturation episodes during one-lung ventilation (OLV).18 However, more recent clinical studies have questioned these observations.19 One recent study comparing oxygenation and desaturation episodes with equipotent TIVA or volatile anesthesia (as judged by a bispectral monitoring target 40-60) in combination with epidural local/opioid analgesia found no difference between these approaches (Figure 13–2). In contrast, purported advantages of inhalational over TIVA include fewer episodes of hypotension and/or vasopressor use and a smoother transition to postoperative analgesia, particularly when TIVA is combined with epidural local anesthesia. In clinical terms, the impression of this author is that the anesthesiologist is the most important ingredient regardless of their selected technique.


Figure 13–2. Changes in arterial oxygen levels over time (mins) with the transition from two-lung (TLV) to one-lung ventilation (OLV) during 65 thoracic surgery procedures randomized to equipotent levels of inhalational sevoflurane or intravenous propofol anesthesia (A). The lowest arterial oxygenation value observed for each patient is also depicted (B). (Used with permission from Pruszkowski O, Dalibon N, Moutafis M, et al.31 Copyright Oxford University Press.)

Arterial oxygen desaturation (<90%) is common during one-lung ventilation, occuring in up to 10% of patients20 but is rarely due to shunting alone, which amounts to only about 30% after normal adaptation to one-lung physiology. Other sources of hypoxemia with OLV are numerous, including insufficient inspired oxygen concentration, endotracheal tube obstruction or malposition, inadequate tidal volume, and low oxygen delivery from hemodilution or depressed cardiac output due to deep anesthesia, hypovolemia, or right ventricular dysfunction related to hypervolemia or myocardial dysfunction.

A logical approach to the management of arterial desaturation during OLV is an essential skill for the thoracic anesthesiologist (Figure 13–3; see also Chapter 5). The first response is to acutely rectify the clinical situation by returning inspired oxygen concentration to 100% and asking the surgeon to temporarily discontinue surgery while reinflating the operative lung by manual re-expansion. A brief period of two-lung ventilation generally returns the oxygen saturation to acceptable levels, providing a period for troubleshooting to identify addressable factors to prevent or delay its recurrence. The two main causes of hypoxemia, sources of inadequate ventilation and/or perfusion, must be promptly identified and optimized. Simple steps include suctioning of tube secretions and auscultatory and fiberoptic visual confirmation of correct tube positioning (including lung isolation and identifying any occult upper lobe obstruction from endobronchial tube malposition) and adjusting ventilator settings to 3 to 4 mL/kg tidal volume (ideal body weight) or using pressure limited ventilation (maximum peak inspiratory pressure <30 cm H2O). The physiology of OLV increases normal demands on the right ventricle, requiring higher filling pressures to maintain cardiac output—a modest colloid or crystalloid bolus (eg, 250 mL) may expose a state of volume-responsive hypovolemia. Excessively deep anesthesia should also be easily recognized and corrected.


Figure 13–3. Suggested algorithm for response to hypoxemia during one-lung ventilation (OLV). CPAP - continuous positive airway pressure; DLT - double lumen tube; FiO2 - inspired fraction of oxygen; PEEP - positive end-expiratory pressure; PGI2 - prostacyclin; ST-ST segment. (Used with permission from Roze H, Lafargue M, Ouattara A.20 Copyright Wolters Kluwer Health.)

If hypoxemia recurs during one-lung anesthesia despite the simple steps outlined above, then a sequential approach to improving oxygenation with ventilatory maneuvers may help the problem. The first simple step is to introduce continuous oxygen insufflation to the operative lung (nondependent, upside). This can be easily achieved by placing an endotracheal suction catheter into the open airway lumen attached to standard tubing with oxygen flows at 2 to 3 L/min (remember to occlude the thumb port). A more elaborate setup can be added to further recruit perfused upside lung alveoli with continuous positive airway pressure (CPAP) 5 to 10 cm H2O; this rarely causes sufficient lung inflation to distract the surgical team.

If hypoxemia persists despite adding oxygen and CPAP to the upside lung as outlined above, then changes to improve ventilation of the nonoperative (dependent, downside) lung may be helpful. First, recruitment maneuvers for a few seconds with a peak airway pressure of 40 cm H2O will assure that atelectasis in the dependent lung is not contributing to hypoxemia. Second, an experiment with the addition of PEEP is warranted. For some patients, the addition of 10 cm H2O PEEP will recruit alveoli and improve oxygenation; whereas, for others, it will paradoxically shunt blood towards the upside lung, while in some individuals added PEEP will aggravate “auto PEEP” and precipitate hypotension without improving oxygenation (see also One-lung ventilation in patients with severe chronic obstructive pulmonary disease, in Chapter 15). Converting from volume-controlled to pressure-limited ventilation may also help for some patients.

Clearly, clamping of the arterial supply to lung being resected should reduce shunt and improve oxygen saturation, but active steps to otherwise improve hypoxemia through redistribution of blood flow during one-lung anesthesia are otherwise rarely employed. Simple retraction of the upside lung has been noted to reduce hypoxemia. As mentioned above, the importance of anesthetic agent selection is controversial, with evidence for the inhibition of hypoxic pulmonary vasoconstriction by volatile anesthetics relative to TIVA being restricted to comparisons with isoflurane; whereas, more recently, sevoflurane has even been suggested to reduce inflammatory response and improve outcomes compared to TIVA.21 Infusions of the hypoxic pulmonary vasoconstriction potentiator almitrine have shown benefit in some studies, alone and in combination with nitric oxide. Notably, almitrine has an important toxicity profile and is not currently available in the United States.

Attentiveness to inspired air:oxygen ratios and tidal volume management during lung resection may affect outcomes, with the bulk of evidence supporting a strategy that keeps O2 concentrations and tidal volumes at lowest tolerable levels.22,23 For patients receiving induction therapy prior to their surgery, this has particular importance since adjuvant chemotherapy (eg, cisplatin, paclitaxel) can inflict subclinical acute lung injury making parenchyma more vulnerable to effects such as barotrauma, oxygen toxicity, and the formation of reactive oxygen species. Unacceptable oxygen desaturation clearly warrants steps to prevent hypoxemia, including recruitment maneuvers (peak airway pressure <40 cm H2O) and inspired oxygen concentration up to 100%, but, when other options exist, avoiding high airway pressures (eg, keep peak airway pressure <30 cm H2O) and oxygen concentrations (eg, <50%) may be beneficial. A common mistake with volume-controlled ventilation is to overlook the risk of barotrauma at the onset of one-lung anesthesia; to avoid inadvertent large breaths, minute ventilation during one-lung anesthesia can be safely achieved by more frequent smaller breaths or by changing to pressure limited ventilation and modest permissive hypercapnia.

Acute lung injury and adult respiratory distress syndrome (ALI/ARDS) is a very serious complication that arises within the first 3 to 4 days following lung resection surgery and is associated with a 40% mortality rate.1 ALI/ARDS prevalence after pneumonectomy, lobectomy/bi-lobectomy, and sublobar resections are 8, 3, and 0.9%, respectively. Numerous patient and procedural characteristics have been associated with increased risk of ALI/ARDS, notably including high tidal volume and airway pressure during OLV.24 Other reported factors include advanced age, male gender, chronic suppurative disease, concurrent cardiac disease, low diffusion capacity for carbon monoxide, and resection of more than 45% of lung vasculature.2 Additional perioperative ALI/ARDS risk factors include hypervolemia, greater extent of tissue resection, extended surgery time, increased blood loss, and reoperation. While no singular cause of ALI/ARDS after lung resection has been identified, evidence supports multiple contributing mechanisms, causes and aggravating factors, including hyperoxia, reactive oxygen species and barotrauma, lymphatic disruption, microembolization, elevated pulmonary vascular pressures, and ischemia-reperfusion and inflammatory lung injury.

Concern over perioperative fluid therapy for lung resection comes from reports of ALI/ARDS associated with large volumes of intravenous fluid and hypervolemia.2 While “excessive” fluid administration is avoidable and may be contributory, it does not always precede the onset of ALI/ARDS. Nonetheless, conservative fluid management seems prudent in most cases, although fluid restriction sufficient to precipitate acute kidney injury following lung resection also correlates with adverse outcome (see also Chapter 23).25 Recent advances in fluid management include better understanding of the importance of the endothelial glycocalyx in the formation of edema, and factors that affect the redistribution of colloid and crystalloid from the intravascular space.26 Administration algorithms have moved away from the traditionally calculated “third space” fluid loss for a procedure and instead advocate for measured replacement of insensible fluid deficits with crystalloid (eg, lactated Ringer solution), and, in the absence of major bleeding or anemia, the use of colloid infusion for subsequent euvolemia maintenance. This strategy rarely translates intraoperatively into more than 1 liter of crystalloid and 1 L of colloid total for most patients.


As outlined above, in formulating an anesthetic plan for thoracic surgery, considerable respect must be paid to a serious complication of emergence whose occurrence is partly under the influence of the anesthesiologist—failure to achieve tracheal extubation. This is particularly important since major pulmonary complications of lung resection surgery are more than twice as likely in the setting of postoperative respiratory failure27 and highly associated with other markers of adverse outcome, including postoperative mortality (see Table 13–2). Contributors to the generally high risk of respiratory failure after lung resection include “variable” factors amenable to optimization such as inadequate respiratory mechanics (from residual paralysis, suboptimal positioning, and pain-related chest wall splinting), post-extubation upper airway obstruction, and respiratory depression due to residual anesthetic agents, and “fixed” factors such as pre-existing disease, infection, and surgery-related loss of parenchyma, lung contusion, and airway soiling. An effective tracheal extubation strategy must focus on optimizing “variable” factors, including preemergence interventions (airway suctioning, positioning, analgesia, etc), appropriate agent selection throughout surgery (as outlined above), and a logical brief sequence of anesthetic withdrawal and tracheal extubation.28 Subsequent to tracheal extubation but equally important activities include the continuous maintenance of a patent airway and precise monitoring to identify those who will require reintubation.

Chest wall closure and skin suturing identify the final phase of all lung resection surgeries. Routine interventions poorly tolerated at lighter anesthetic depths should occur early during chest closure, such as gastric, throat and endotracheal tube suctioning, and oral and/or nasal airway insertion. Reducing anesthetic depth and neuromuscular block over the 10 to 15 minute closure period is critical to preparation for emergence. Until just prior to extubation, spontaneous respiration is discouraged to avoid hypoventilation and unrecognized CO2 narcosis, the somnolence from carbon dioxide accumulation that can develop at levels as low as 70 mm Hg.29 An alternate approach during chest closure is to allow spontaneous respiration assisted by pressure-support mode ventilation; this strategy can prevent gas rebreathing and hypercarbia (from the considerable dead space of the double-lumen tube), while allowing respiratory drive to develop.

Since pain at emergence is extremely difficult to treat without adding acute respiratory depression and interfering with efforts to extubate the patient, the anesthesiologist must be confident that analgesia is established preemergence. If a thoracic epidural catheter has been placed, a common practice is to supplement existing analgesia with an additional 2 mL bolus of 2% preservative-free lidocaine (for an average adult male) 10 to 15 minutes prior to emergence; this represents a modest “insurance policy” against emergence pain that is rarely associated with block-mediated hypotension but allows the patient to be awake and extubated before more pain management decisions are needed.

Prior to emergence, the patient should be moved into a sitting or deck chair position; the flex at the waist and supine head up position raises the chest and displaces the abdominal contents caudad. Practically speaking, such positioning establishes or increases functional residual capacity, which reduces the likelihood of oxygen desaturation through the emergence phase.28 If the patient is otherwise stable, and tracheal extubation is expected to be a major challenge, transfer from the operating table to the bed prior to awakening (while maintaining adequate monitoring) eliminates this interruption to coordinated breathing. When dealing with a patient with extremely impaired pulmonary function, a near vertical sitting position supported by pillows is essential.

As wound dressing is completed, steps can be coordinated to eliminate residual anesthetic and muscle relaxant drugs while simultaneously positioning the patient. Only a brief period of responsiveness, coordinated strength, and regular breathing pattern with acceptable tidal volume and/or maximum inspiratory pressure should precede prompt tracheal extubation and 100% O2 by face mask. Among post-extubation options for maintaining a patent airway, the nasopharyngeal trumpet is best tolerated for prolonged periods but must be placed very carefully to minimize any risk of epistaxis.

Patients with severely impaired pulmonary function commonly cannot meet “ideal” tidal volume or maximum inspiratory pressure extubation criteria, and experience and judgement from the anesthesiologist is particularly neccessary in timing endotracheal tube removal for these patients, appreciating that some will require reintubation.28 For patients with severe disease, appropriate responsiveness, strength, and breathing pattern may warrant a trial of tracheal extubation without evidence of deep breaths, since delay increases the chances of CO2 narcosis in this population.29 Yang and colleagues found an f/Vt ratio (frequency of breaths per minute divided by the tidal volume in liters) <100 to be highly predictive of successful tracheal extubation in critically ill patients with limited respiratory reserve, but the value of this metric in perioperative patients has not been assessed.30 Fortunately, such difficult emergence occurrences are infrequent, but the patient with limited respiratory reserve likely has the most to gain from an experienced anesthesia team and avoidance of a prolonged episode of postoperative mechanical ventilation.

To distinguish the stable extubated patient with limited respiratory reserve from one who needs reintubation, beyond clinical appearance and pulse oximetry monitoring, an extremely useful tool is arterial blood gas CO2 trends. Sequential repeated blood gas determinations immediately following tracheal extubation (eg, every 3 min) identify two main patterns; the first, a steady decrease of CO2 levels toward the patient’s baseline, even if starting from extremely high values (eg, 100 to 120 mm Hg), favorably predicts successful extubation and generally indicates simple ongoing conservative care will be sufficient; the second, where CO2 levels are stable or rising, is much more concerning and requires further prompt intervention and optimization to avert respiratory failure and tracheal reintubation. Should tracheal reintubation be required, it is important to remember that bag-mask ventilation and oxygenation should avert major problems and allow this to be a controlled event.


1. Dulu A, Pastores SM, Park B, Riedel E, Rusch V, Halpern NA. Prevalence and mortality of acute lung injury and ards after lung resection. Chest. 2006;130(1):73-78.

2. Grichnik KP, D’Amico TA. Acute lung injury and acute respiratory distress syndrome after pulmonary resection. Semin Cardiothorac Vasc Anesth. 2004;8(4):317-334.

3. Harpole DH Jr, DeCamp MM Jr, Daley J, et al. Prognostic models of thirty-day mortality and morbidity after major pulmonary resection. J Thorac Cardiovasc Surg. 1999;117(5):969-979.

4. Lindstrom I, Nordling S, Nissen AM, Tammilehto L, Mattson K, Knuutila S. DNA copy number changes in lung adenocarcinoma in younger patients. Mod Pathol. 2002;15(4):372-378.

5. Kaifi JT, Gusani NJ, Deshaies I, et al. Indications and approach to surgical resection of lung metastases. J Surg Oncol. 2010;102(2):187-195.

6. Call S, Rami-Porta R, Obiols C, et al. Repeat mediastinoscopy in all its indications: experience with 96 patients and 101 procedures. Eur J Cardiothorac Surg. 2011;39(6):1022-1027.

7. Villamizar NR, Darrabie MD, Burfeind WR, et al. Thoracoscopic lobectomy is associated with lower morbidity compared with thoracotomy. J Thorac Cardiovasc Surg. 2009;138(2):419-425.

8. Whitson BA, Groth SS, Duval SJ, Swanson SJ, Maddaus MA. Surgery for early-stage non-small cell lung cancer: a systematic review of the video-assisted thoracoscopic surgery versus thoracotomy approaches to lobectomy. Ann Thorac Surg. 2008;86(6):2008-2016; discussion 2016-2008.

9. Gopaldas RR, Bakaeen FG, Dao TK, Walsh GL, Swisher SG, Chu D. Video-assisted thoracoscopic versus open thoracotomy lobectomy in a cohort of 13,619 patients. Ann Thorac Surg. 2010;89(5):1563-1570.

10. Fleisher LA, Beckman JA, Brown KA, et al. ACC/AHA 2007 guidelines on perioperative cardiovascular evaluation and care for noncardiac surgery: a report of the American College of Cardiology/American Heart Association Task Force on practice guidelines (writing committee to revise the 2002 guidelines on perioperative cardiovascular evaluation for noncardiac surgery): developed in collaboration with the American Society of Echocardiography, American Society of Nuclear Cardiology, Heart Rhythm Society, Society of Cardiovascular Anesthesiologists, Society for Cardiovascular Angiography and Interventions, Society for Vascular Medicine and Biology, and Society for Vascular Surgery. Circulation. 2007;116(17):e418-e499.

11. Leonovicz BM, Gordon EA, Wass CT. Paraneoplastic syndromes associated with lung cancer: a unique case of concomitant subacute cerebellar degeneration and lambert-eaton myasthenic syndrome.Anesth Analg. 2001;93(6):1557-1559, table of contents.

12. Griffiths DF, Williams GT, Williams ED. Duodenal carcinoid tumours, phaeochromocytoma and neurofibromatosis: islet cell tumour, phaeochromocytoma and the von Hippel-Lindau complex: two distinctive neuroendocrine syndromes. Q J Med. 1987;64(245):769-782.

13. Zografos GN, Vasiliadis GK, Zagouri F, et al. Pheochromocytoma associated with neurofibromatosis type 1: concepts and current trends. World J Surg Oncol. 2010;8:14.

14. Jensen RT, Berna MJ, Bingham DB, Norton JA. Inherited pancreatic endocrine tumor syndromes: advances in molecular pathogenesis, diagnosis, management, and controversies. Cancer. 2008;113(7 Suppl):1807-1843.

15. Horlocker TT, Wedel DJ, Benzon H, et al. Regional anesthesia in the anticoagulated patient: defining the risks (the Second Asra Consensus Conference on Neuraxial Anesthesia and Anticoagulation).Reg Anesth Pain Med. 2003;28(3):172-197.

16. Chow MY, Liam BL, Thng CH, Chong BK. Predicting the size of a double-lumen endobronchial tube using computed tomographic scan measurements of the left main bronchus diameter. Anesth Analg. 1999;88(2):302-305.

17. Vanden Hoek TL, Morrison LJ, Shuster M, et al. Part 12: cardiac arrest in special situations: 2010 American Heart Association guidelines for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation. 2010;122(8 Suppl 3):S829-S861.

18. Abe K, Shimizu T, Takashina M, Shiozaki H, Yoshiya I. The effects of propofol, isoflurane, and sevoflurane on oxygenation and shunt fraction during one-lung ventilation. Anesth Analg. 1998;87(5):1164-1169.

19. Ishikawa S. Oxygenation may improve with time during one-lung ventilation. Anesth Analg. 1999;89(1):258-259.

20. Roze H, Lafargue M, Ouattara A. Case scenario: management of intraoperative hypoxemia during one-lung ventilation. Anesthesiology. 2011;114(1):167-174.

21. De Conno E, Steurer MP, Wittlinger M, et al. Anesthetic-induced improvement of the inflammatory response to one-lung ventilation. Anesthesiology. 2009;110(6):1316-1326.

22. Grocott HP. Oxygen toxicity during one-lung ventilation: is it time to re-evaluate our practice? Anesthesiol Clin. 2008;26:273-280, v.

23. Walker MG, Yao LJ, Patterson EK, et al. The effect of tidal volume on systemic inflammation in acid-induced lung injury. Respiration. 2011;81(4):333-42.

24. Jeon K, Yoon JW, Suh GY, et al. Risk factors for post-pneumonectomy acute lung injury/acute respiratory distress syndrome in primary lung cancer patients. Anaesth Intensive Care. 2009;37(1):14-19.

25. Martin D, Kushins S, Phillips-Bute B, Onaitis M, Stafford-Smith M. Comparison of post-pneumonectomy outcomes and AKI by AKIN creatinine rise and AKIN oliguria criteria. Anesthesiology. 2009;111:A176.

26. Chappell D, Jacob M, Hofmann-Kiefer K, Conzen P, Rehm M. A rational approach to perioperative fluid management. Anesthesiology. 2008;109(4):723-740.

27. Stephan F, Boucheseiche S, Hollande J, et al. Pulmonary complications following lung resection: a comprehensive analysis of incidence and possible risk factors. Chest. 2000;118(5):1263-1270.

28. Miller KA, Harkin CP, Bailey PL. Postoperative tracheal extubation. Anesth Analg. 1995;80(1):149-172.

29. Langford NJ. Carbon dioxide poisoning. Toxicol Rev. 2005;24(4):229-235.

30. Yang KL, Tobin MJ. A prospective study of indexes predicting the outcome of trials of weaning from mechanical ventilation. N Engl J Med. 1991;324(21):1445-1450.

31. Pruszkowski O, Dalibon N, Moutafis M, et al. Effects of propofol vs sevoflurane on arterial oxygenation during one-lung ventilation. Bri J Anaesth. 2007;98(4):539-544.