John W. Strieder, a seasoned thoracic surgeon of the early 20th century, described “the good old days” of thoracic anesthesia colorfully: “the period of operation was, with dismaying frequency, a race between the surgeon and the impending asphyxia of the patient.”¹ Aurelius Cornelius Celsus (25 bc–ad 50), the Roman encyclopedist, knew 2000 years ago that entering the thorax posed unique dangers to the patient. In De Medicina, Celsus describes “for the belly indeed, which is of less importance, can be laid open with the man still breathing; but as soon as the knife really penetrates to the chest…the man loses his life at once.”² This is an early description of the “pneumothorax problem”: opening the chest immediately causes an open pneumothorax. When the lung is exposed directly to the atmosphere, it will rapidly collapse because of the loss of the normally negative intrapleural pressure. In addition, air would be transferred between the two lungs known as “pendulluft,” and the collapsed lung would paradoxically expand during expiration and collapse during inspiration. To further terrify the surgeon, vigorous side-to-side movement of the mediastinum could occur, known as “mediastinal flapping,” that could compress the nonoperative lung. In the lateral decubitus position, it would result in “mediastinal shift” and hypotension. Not surprisingly, respiratory and hemodynamic compromise would ensue as the patient would struggle to breathe spontaneously. Hence most thoracic procedures were limited to the extrathoracic chest wall until the 1930s. Only very brief intrathoracic procedures were possible without patient asphyxiation.

Most areas of surgery flourished after the discovery of inhalational anesthesia in the 1840s, and the delivery of general anesthesia became routine. Until the 1930s, delivery of inhalational anesthesia was typically by mask or open drop administration, using ether or chloroform with or without nitrous oxide. Because patients would typically breathe spontaneously, they could control their own depth of anesthesia with their own respirations. Muscle relaxants were not developed yet, and endotracheal intubation was considered an invasive procedure and only rarely used by a few experts. Most thoracic procedures performed were the same pathology that concerned Celsus 2000 years ago: management of empyema, pulmonary abscess, and tuberculosis. Without antibiotics, patients would frequently present for surgery with copious secretions and formidable coughs. It was common to keep a patient only lightly anesthetized to keep the cough reflex intact to protect the lungs from gastric aspiration and to allow the patient to clear their own copious secretions. Envisioning a harrowing scene of a lightly anesthetized patient choking on their secretions with an unprotected airway, it is hardly surprising that thoracic surgery remained in its infancy well into the 20th century. Better operating conditions and improved anesthesia techniques were needed to allow thoracic surgery to flourish.

Before the discovery of antibiotics, most thoracic procedures were performed to treat infection, opening the pleural cavity did not always result in an open pneumothorax because prolonged infections often resulted in adhesions between the lung and chest wall with a loculated empyema. The utility of these adhesions was known, and repeated aspirations were sometimes attempted to promote adhesion formation before surgery. Alternatively, air or water could be injected into the pleural space as an irritant to promote adhesion formation preoperatively.^3,4 “Muller’s handgrip” was another primitive method used to cope with the pneumothorax problem: while the chest was open, the surgeon would pull the lung into the wound to plug the thoracotomy incision.⁵ Pulmonary resections were frequently performed in a staged manner and had a very high mortality. A snare or tourniquet technique would be used to facilitate a quick resection, and then a reoperation would be needed to remove necrotic tissue later. It is not surprising that sepsis was not uncommon from the remaining necrotic tissue. A review from 1922 reported a mortality rate of 42% for lobectomy, and as high as 70% for cases that involved more than one lobe.⁶ Clearly, surgeons and patients needed safer, less harmful solutions.

Differential Pressure Breathing

The German surgeon Ernst Ferdinand Sauerbruch developed the first promising solution to the “pneumothorax problem.” In 1893, his mentor, Johann von Mikulicz-Radecki, urged him to address the difficulty of operating with an open pneumothorax. His solution, differential pressure breathing, became the principal method for management of ventilation in thoracic surgery until World War II. In Sauerbruch’s experiments on dogs, he found that, during thoracotomy, spontaneous ventilation was maintained and the lung did not collapse if it was exposed to a pressure 10 cm H₂O below atmospheric pressure.⁷ After his experimental thoracotomies on dogs, he applied the technique to humans (Fig. 1.1). To maintain the negative pressure, a large negative pressure chamber was needed that would maintain the normal negative intrapleural pressure. The patient and surgical team were placed within the steel negative pressure chamber while the patient’s head protruded from the chamber and was exposed to atmospheric pressure. With the negative pressure applied directly to the lung, the patient could breathe spontaneously and the lung would remain inflated.

Sauerbruch championed his pneumatic chamber technique as a physiologic method, and differential pressure breathing was widely adapted. However, Sauerbruch’s method was very impractical because of the large, expensive, negative pressure chamber that was needed. Operating conditions were less than ideal. Rudolph Nissen described the limitations of this operating suite: “the surgeon and his assistants had very little room to move; the heat was almost unbearable; and, finally, it was extremely difficult to communicate satisfactorily with the anesthetist outside the chamber.”⁸ An anesthetist would be outside the chamber at the patient’s airway and could only communicate with the surgeon within the chamber by phone over the loud whirring of pumps.

A more practical alternative method for using differential pressure to maintain lung inflation was developed in parallel by a colleague of Sauerbruch’s, Ludolph Brauer. His alternative method for using differential pressure breathing was published alongside Sauerbruch’s. Brauer’s method used a positive pressure chamber to increase the intrapulmonary pressure. Brauer’s chamber was simply a large box and the patient’s head was placed within it after the induction of anesthesia, and anesthesia was maintained with the patient breathing oxygen and chloroform spontaneously. Before the chest was opened, compressed air would be added to the chamber to raise the pressure above atmospheric pressure, and this would prevent the development of an open pneumothorax. The anesthetist would have no access to the head during the procedure.⁷ Brauer’s design resembles specialized helmets developed for delivering continuous positive airway pressure (CPAP) or for noninvasive ventilation that could be used for treating respiratory failure.⁹

Although Brauer’s positive pressure technique was simpler than Sauerbruch’s, Sauerbruch had his devotees in Europe and the United States. In 1909, the American surgeon Willy Meyer created his own “universal differential pressure chamber,” a modified version of Sauerbruch’s negative pressure chamber.¹⁰ Meyer’s chamber was even more complicated than Sauerbruch’s; it included both a positive and negative pressure chamber. The overall chamber was 1000 cubic feet in volume and could contain up to 17 people. The patient, anesthetist, and an assistant could be enclosed in the positive pressure chamber within the negative pressure room. By using both chambers, the normal negative intrapleural pressure gradient could be maintained, either by applying positive pressure to the head, negative pressure to the open chest, or both. Meyer described “if the differential pressure in the universal pressure is composed of part vacuum and part pressure, only the patient is exposed to the full differential, while all others are exposed only to the component…the anesthetizer to the positive fraction and the surgeon…to the negative fraction, which still more reduces any possibility of detrimental effects on the users of the chamber.” This was the only negative pressure chamber built for this purpose in America, and Meyer also used it for improving wound drainage and lung expansion postoperatively.¹¹

Both the positive pressure and negative pressure methods relied on maintaining a pressure gradient between the air outside and within the lungs, otherwise known as differential pressure breathing. Differential pressure breathing was successful at preventing the formerly inevitable open pneumothorax after thoracotomy; however, it was doomed to become a historical relic because it provided dangerously inadequate ventilation. Hypoventilation, hypercarbia, hypoxemia, and impaired venous return were significant problems during prolonged cases and clinical deterioration was not uncommon. Meyer attributed the cause of unexplained shock to hypercarbia, and he recommended applying rhythmic variations in pressure coordinated with spontaneous respirations to assist with ventilation. Remarkably, although this method of preserving respiration with an open chest seems so cumbersome to modern readers, Sauerbruch and his followers felt it was endotracheal intubation that was impractical and unsafe. Meyer felt “combining intubation and masks appears so manifestly inadequate and dangerous for everyday surgery that it cannot deserve preference over apparatus leaving the mouth of the patient unincumbered [sic].”¹⁰

Tracheal Insufflation and Endotracheal Anesthesia

Tracheal insufflation anesthesia, an alternative method for preventing the development of the open pneumothorax, became popular in America in the early 20th century. This new method is the clear precursor to the endotracheal anesthesia we use today. Because of widespread skepticism about the routine use of tracheal intubation, the development did not follow a smooth path. Tracheal intubation and mechanical ventilation were not new discoveries; many pioneers deserve credit in the development of intubation, laryngoscopy, and positive pressure ventilation, especially considering how much skepticism they faced.

Andreas Vesalius used tracheal intubation for positive pressure ventilation of a pig in 1543. He performed a tracheotomy and passed a reed into the trachea of a pig and blew into the tube to provide artificial ventilation during a thoracotomy and thus prevented a potentially fatal open pneumothorax. His findings went unnoticed and were only later rediscovered. In 1788, Charles Kite resuscitated victims of drowning from the River Thames using curved metal cannulas that he placed blindly in the trachea. Soon after the development of inhalational anesthesia, there were early enthusiasts trying to apply these resuscitation techniques to anesthesia delivery. In 1869, Friedrich Trendelenburg used a tracheostomy tube with an inflatable cuff to administer chloroform during head and neck surgery. William MacEwan, a Scottish surgeon, is credited with the first use of oral endotracheal intubation for an anesthetic. On July 5, 1878, MacEwan placed a flexible metal tube in the larynx of an awake patient who was to have an oral tumor removed at the Glasgow Royal Infirmary.¹² In 1885, Joseph O’Dwyer, a pediatrician unaware of earlier uses of intubation, performed blind oral tracheal intubations on children suffering from diphtheria.¹³ O’Dwyer designed a rigid tube with a conical tip that could occlude the larynx to facilitate positive-pressure ventilation. In 1893, George Fell attached O’Dwyer’s metal tube to a bellows and T-piece, creating the Fell-O’Dwyer apparatus. Fell used the apparatus to provide ventilatory support for opiate-induced respiratory depression (Fig. 1.2).

By the 1890s, there was interest in applying endotracheal anesthesia technique to thoracic surgery in an attempt to prevent the pneumothorax problem. In 1896, the French surgeons Tuffier and Hallion reported on their use of tracheal intubation with artificial ventilation to perform thoracotomies on animals.¹² They used a device with a bellows for the rhythmic inflation of the lungs, and a water valve that could control the degree of resistance to expiration, a precursor to the modern use of positive end-expiratory pressure (PEEP). Inspired by Tuffier and Hallion, Rudolph Matas made modifications to the Fell-O’Dwyer apparatus to make it appropriate for use during surgery. Matas was convinced that such a device would be ideal for thoracic cases. His modifications included adding a graduated cylinder for delivery of precise volumes of gases and a mercurial manometer for the measurement of intrapulmonary pressures. He also modified it to be a simple anesthesia machine by adding an intralaryngeal cannula connected by a stopcock to a rubber tube and funnel that could be used for administering chloroform.¹⁴

These early pioneers of endotracheal techniques were using endotracheal tubes that were similar in size to the trachea, through which inspiration and exhalation occurred. In 1907, Barthélemy and Dufour used a new method called “tracheal insufflation.”¹² A thin tube was placed in the trachea and gases were continuously insufflated under positive pressure into the lower portion of the trachea. Expired gases exited between the tracheal tube and the tracheal wall. Meltzer and Auer, American physiologists, used this technique extensively in animal studies and showed that curarized dogs could be anesthetized and kept alive by blowing air and ether continuously into a tube inserted into the trachea. Gas exchange would still occur “without any normal or artificial rhythmical respiratory movements whatever” because expired gases could escape around the tracheal tube.¹⁵ This was essentially an improvement of Brauer’s method of continuously applying positive pressure; however, because dead space was decreased significantly by the placement of the cannula in the trachea, gas exchange was improved although still not optimized.

Charles Elsberg, a thoracic surgeon in New York City, was familiar with Meltzer and Auer’s research and applied this method to thoracic surgery. He first used tracheal insufflation to resuscitate a myasthenic patient who had become cyanotic and pulseless. The technique was successful in that she regained spontaneous circulation; however, she did not regain consciousness so the resuscitation was eventually discontinued. Elsberg modified Meltzer and Auer’s apparatus by replacing the bellows with an electric motor. He also placed the tracheal cannula under visualization after topicalization of the larynx with cocaine by using either a Killian bronchoscope or a Chevalier Jackson laryngoscope.¹⁶ In February 1910, Elsberg presided over the historical first use of tracheal insufflation anesthesia for thoracotomy.¹⁷ The thoracic surgeon Howard Lilienthal recruited Elsberg to help him treat a butcher with a 13-month history of productive cough. The presumptive diagnosis was lung abscess, and Lilienthal wanted to attempt an operative cure. When the pleura was opened, 15 mm Hg was applied intratracheally, and the lung was noted to be “two-thirds of its capacity, mottled, and rosy pink in color.” Different pressures were applied and the lung collapsed and swelled. Elsberg periodically interrupted the insufflation every 2 to 3 minutes, to allow the lungs to collapse and facilitate carbon dioxide elimination, thus resembling modern positive-pressure ventilation. After his success in this landmark surgery, Elsberg promoted tracheal insufflation for all surgeries requiring general anesthesia. Only 1 year later, he published on his experiences using this technique to anesthetize over 200 patients.¹⁸ Elsberg’s method of tracheal insufflation is very similar to the modern practice of oxygen insufflation during rigid bronchoscopy that was first introduced by Sanders in 1968.¹⁹

Endotracheal Intubation and Laryngoscopy

After Elsberg’s triumph, tracheal insufflation anesthesia became the most popular anesthetic method for thoracic surgery in the United Sates in the 1920s and 1930s. Differential pressure anesthesia was still preferred in Europe for thoracic procedures. Tracheal insufflations remain popular in Europe only for head and neck procedures where mask or hand drop techniques could interfere with the surgical field. A major reason for the reluctance to widely adopt the tracheal insufflation technique in Europe was the dominance of Sauerbruch and his unwillingness to adopt any other method. Sauerbruch’s own assistant, Giertz, performed experiments on animals that showed that rhythmic inflation of the lungs was superior to differential pressure breathing. He also showed that differential pressure anesthesia resulted in inadequate ventilation, hypercarbia, impaired venous return, and circulatory collapse.⁷ Although better than the alternative, tracheal insufflation was far from perfect. Carbon dioxide accumulation would occur if gas flow was interrupted. This was addressed with modifications to Elsberg’s apparatus that periodically stopped airflow to allow the lungs to collapse. Also, barotrauma was possible when dangerously high intrapulmonary pressure occurred when the return of gas was impeded. Alveolar rupture and surgical emphysema could occur and were called “wind-tumor,” likely caused by an interruption in the exit of expired gases when laryngospasm occurred around thin tracheal insufflation catheters.¹²

Another impediment to the routine use of endotracheal techniques was that blind placement of endotracheal tubes was the norm. Instruments for direct laryngoscopy existed by the 1920s, but were infrequently used. Blind placement required considerable skill and could be traumatic and cause airway laceration from the rigid tube. Alfred Kirstein, a physician in Berlin, is credited with inventing the first direct laryngoscope in 1895; before 1895, direct visualization of the larynx was considered impossible. Kirstein’s “autoscope” was not used for anesthesia, but it was the prototype for many laryngoscopes to follow.²⁰ In 1913, Chevalier Jackson developed his own laryngoscope and described proper positioning and technique for laryngoscopy in a landmark paper.²¹ In 1941, Robert Miller created the still familiar Miller blade, its origins clearly rooted in the laryngoscopes of Kirstein and Jackson. Sir Robert Macintosh released his curved blade in 1943, that remains until today the most popular laryngoscope blade in the world because of its ease of use.

Improvements in endotracheal tubes occurred alongside these developments in direct laryngoscopy. World War I produced many wounded warriors requiring reconstructive surgery for head and neck injuries. In 1919, the British anesthetists Ivan Magill and Stanley Rowbotham were assigned to work with the British army plastics unit. Under pressure to provide unhindered access to the face and airway, they became experts in blind nasal intubations. They rejected the popular insufflation technique and used larger tubes that permitted inhalation and exhalation to occur through the tube. Magill’s wide-bore red rubber tubes resisted kinking and adjusted to the contours of the upper airway. They remained the standard endotracheal tube until plastic tubes were introduced.

The next step was the development of the cuffed tracheal tube. Without this, controlled positive-pressure would not be effective. In the 19th century, there were sporadic attempts at using cuffed tubes. In 1871, Trendelenburg used a cuffed tracheotomy tube, as did Eisenmenger in 1893, and Dorrance in 1910.²² None of these attempts sparked much interest in cuffed endotracheal tubes. In 1928, Guedel and Ralph Waters introduced their endotracheal tube with a detachable inflatable cuff, and became strong advocates for the routine use of cuffed endotracheal tubes (Fig. 1.3).²³ Guedel performed his famous “dunked dog” demonstrations to show the effectiveness of the tube’s seal. He submerged his intubated and sedated dog in an aquarium, from which he emerged unscathed.²⁴ Not only would this tube facilitate the use of controlled positive-pressure ventilation, it could prevent aspiration of gastric contents, no longer making it necessary for patients to be kept lightly anesthetized to preserve the cough reflex. With deeper planes of anesthesia, the trachea could be suctioned and operating conditions improved. Through hyperventilation, it was often possible to suppress respiratory efforts even without muscle relaxation. Control of ventilation and protection from aspiration of gastric contents represent an historic milestone in patient ventilation strategy.

Even though all of the components of airway management necessary to conquer the “pneumothorax problem” existed by 1930, unfortunately, these methods did not immediately gain widespread use. Sauerbruch’s differential pressure breathing was still commonly used in Europe until World War II. Cuffed endotracheal tubes were not initially deemed necessary and took many years to gain widespread approval. In 1948, a review of 309 anesthetics for thoracic cases still advocated routine use of steep Trendelenburg to promote drainage of secretions around uncuffed endotracheal tubes and still did not recommend routine use of controlled positive-pressure ventilation.²⁵

Milestones in Thoracic Surgery

Thoracic surgery progressed at a snail’s pace in the 1920s. Improvements in anesthetic techniques in the 1930s made several advances in thoracic surgery possible. In 1929, Harold Brunn used the individual-structure ligation technique to replace the two-stage snare or tourniquet technique for lung resection. This new technique reduced complications, such as air leak, tension pneumothorax, hemorrhage, and infection from necrotic residual tissue.²⁶ Rudolph Nissen performed the first two-stage pneumonectomy in 1931, soon followed by Evarts Graham’s one-stage total pneumonectomy for lung cancer in 1933.^27,28 The trajectory of thoracic surgery was changing; opening the chest had been so risky that it had been reserved only for refractory infections, but now the role of thoracic surgery for treating malignancy could flourish and overshadow its use for the treatment of infection. The addition of routine postoperative pleural drainage in the 1930s by closed chest thoracostomy also aided surgical progress. Advances in esophageal surgery also occurred in the 1930s. The first transthoracic esophagectomy with an intrathoracic esophagogastric anastomosis was performed successfully in Japan in 1933.²⁹ Thoracic surgery was starting to flourish as thoracic anesthesia improved.

Thoracic Surgery Under Regional Anesthesia

Regional anesthesia for thoracic surgery had its advocates before the 1940s. Proponents of regional anesthesia claimed its safety because it kept the cough reflex intact and maintained spontaneous ventilation. These are still valuable attributes of regional anesthesia. In a 1936 review of thoracic anesthesia, Magill describes spinal anesthesia as an excellent technique for a wide range of thoracic procedures, even pneumonectomy! He recognized that regional anesthesia is best for cooperative patients, as it still is today. The awake patient could assist more easily with breath-holding because controlled ventilation was not routine during general anesthesia.³⁰ Not everyone was so enamored with spinal anesthesia for thoracic surgery. Nosworthy declared, “I like my anesthetic technique to be such that I have the whole situation under control. I do not feel that I am in a position to cope with any emergency when chest surgery is performed under spinal anesthesia.”³¹ Nosworthy went on to describe an inadequate cough reflex and frequent dyspnea during open chest procedures under spinal anesthesia. It is interesting that tubeless thoracic procedures are presently gaining widespread popularity because of concerns that positive pressure ventilation has the potential to injure the lung parenchyma.