At present, video-assisted thoracic surgery (VATS) is relatively well-developed in China and has been widely used for the treatment of lung bullae, pulmonary cysts, and isolated mediastinal tumors and for sleeve resection, wedge resection, segmentectomy, lobectomy, and thymectomy (1–5). The collapse of the ipsilateral lung, which allows surgeons to operate in a limited space, is essential for VATS. A double-lumen endotracheal tube can be used in adults for one-lung ventilation (OLV). However, double-lumen endotracheal tubes suitable for children, especially for infants, are not available in the Chinese market. Endobronchial blocker (BB) placement is the preferred and most frequently used lung separation technique in children (6–9). Due to the immaturity of organ development and the small body size of infants, the risk of lung separation in these children is higher than that in preschool-age or older children. In recent years, to avoid the injuries and risks related to intubated anesthesia, some medical institutions have reported their experience with non-intubated general anesthesia with spontaneous breathing in thoracic surgery for adults. These thoracic surgeries have all been performed with regional nerve blocks [epidural nerve block, paravertebral nerve blocks (PVNB), or intercostal nerve block] when patients are awake or under sedation (2, 10, 11). For preschool children, even simple operations cannot be performed under local anesthesia only if with deep sedation or general anesthesia. Wei et al. had combined non-intubated anesthesia and PVNB to enhance recovery in children aged 3–8 years undergoing VATS (12).

We tried an anesthetic technique for VATS to achieve a safer and faster postoperative recovery with fewer postoperative complications in infants and young children. The technique of non-intubated general anesthesia with spontaneous breathing combined with PVNB was performed in 25 infants from May 2019 to June 2020 in our department. In this study, we introduce the experience from the retrospective data of these patients.

During the study period, a total of 46 ASA I–II children underwent VATS. Of these patients, 25 underwent non-intubated general anesthesia with preserved spontaneous breathing combined with PVNB (non-intubation group). In the non-intubation group, there were 12 males and 13 females, with an average age of 1.0 ± 0.5 years and a weight of 9.3 ± 2.0 kg; there were 14 cases of pulmonary sequestration (PS) and 11 cases of bronchogenic cyst (BC). Conventional intubated general anesthesia was performed in 21 patients (intubation group). In the intubation group, there were 9 males and 12 females, with an average age of 1.1 ± 0.3 years and weight of 9.6 ± 1.2 kg; there were 3 cases of PS and 18 cases of BC. The children in both groups had no history of surgical trauma and no comorbid diseases. All children underwent chest x-ray or chest computed tomography examination before surgery to determine the location and nature of the lesion. Preoperative cardiac color ultrasound and electrocardiogram showed no obvious surgical contraindications. There was no significant difference in general data, such as sex, age, weight, duration of surgery, and intraoperative blood loss, between the two groups. The diagnosis was PS or BC, and all patients underwent lung segmentectomy or lobectomy (Table 1).

In both groups, sufficient depth of anesthesia was achieved, the operation went smoothly, and the perioperative vital signs were stable. Compared with the intubation group, the non-intubation group had a lower HR at T_1,2,4, a lower MAP at T_1–4, a higher SpO₂ at T_0–2, a higher P_ETCO₂ at T₃, and a higher BIS score at T₃. Compared with T₁, the end-tidal carbon dioxide (P_ETCO₂) at T₃ decreased in the intubation group, while it increased at T_2–3 in the non-intubation group. There was no significant difference in the highest P_ETCO₂ between the two groups, but the lowest SpO₂ in the non-intubation group was higher than that in the intubation group (93% vs. 88%, P < 0.001). The two groups had similar duration of surgery. Compared with the intubation group, the non-intubation group had a shorter duration of anesthesia (Table 2).

In the non-intubation group, temporary manual assisted ventilation was performed for six patients (24%) because of SpO₂ <90%. SpO₂ improved after transient manual assisted ventilation. One patient (4%) had coughing during the operation, which caused significant mediastinal flutter and affected the operation. Therefore, anesthesia machine-assisted ventilation was performed after anesthesia was deepened to ensure a smooth operation. In the intubation group, SpO₂< 90% during the operation was observed in 15 patients (71.4%), and breathing parameters returned to normal after sputum suction and adjustment of ventilation parameters. Displacement of the BB after positioning or during surgery occurred in six patients (28.6%), which caused the failure of one-lung ventilation, and the lung on the operation side was expanded. These operations were completed under TLV. There were no perioperative deaths in the two groups.

When the non-intubation group was compared with the intubation group, the time from the completion of the surgery to extubation (P < 0.001), the time to first feeding after surgery (P < 0.001), and the length of postoperative in-hospital stay (P < 0.001) were significantly reduced. The FLACC pain score of the two groups was below 3 on the first postoperative day, and the analgesia effect was satisfactory. In the non-intubation group, there were two cases (8%) of nausea and eight cases of pneumothorax (32%), and the complication rate was 40%. In the intubation group, there were seven cases of pneumothorax (33.3%), six cases of pleural effusion (28.6%), and one case of atelectasis (4.8%), and the complication rate was 66.7%. The patient with atelectasis also had a severe lung infection. There was no statistically significant difference in the overall incidence of complications between the two groups (P = 0.071) (Table 3).

In recent years, VATS has been performed widely in children with diseases of the esophagus, lung, mediastinum, and so on. Lung isolated technique in children, especially young children, is a challenge for anesthesiologists. The smallest available double-lumen tube (DLT) (26 Fr) has an external diameter (ED) of 8.7 mm and is equivalent to single-lumen tube (SLT) of 6 mm (ED: 8.2 mm) or 6.5 mm (ED: 8.9 mm). For this reason, DLT is not suitable for children <8 years of age. Additionally, performing an endobronchial intubation on a dependent lung can achieve an OLV for children <2 years of age. There is an anatomical variation in the take-off of the right upper lobe, which remains very close to the carina (≤1 cm) in children <8 years of age (15). Therefore, a right endobronchial intubation may not get efficient the right upper robe ventilation. A BB also can be performed to isolate an independent lung. In young children, especially children <2 years of age, an extraluminal placement of a BB is used commonly. In these cases, blocker compression and traction may damage vocal cords. Over distension of the cuff may injure bronchial walls. BB may not block the right upper robe completely for its higher opening location resulting in an incomplete collapse. Moreover, changes in body position and/or intraoperative traction may cause the displacement of SLT or BB, and result in the failure of OLV (15, 16). Compared with conventional mechanical ventilation, preserved spontaneous breathing can make lung collapse quickly after thoracic cavity opened and negative pressure disappeared. The likelihood of postoperative hypoxemia is greatly reduced due to avoiding residual muscle relaxants. Because of separation anxiety, preoperative sedation is essential in young children. Dexmedetomidine has been shown to produce stable sedation without respiratory depression and can provide analgesia with stable blood flow dynamics (17, 18). Children have large tongues and relatively narrow oropharyngeal cavities surrounded by soft structures. After sedation, the tongue is likely to retract and obstruct the airway. Using a laryngeal mask rather than a face mask can effectively open the airway and avoid airway obstruction caused by deep sedation. Furthermore, it can reduce sore throat, pharyngeal discomfort, hoarseness, laryngeal edema, vocal cord injury, and recurrent laryngeal nerve paralysis as Janík et al. describe (19). Ultrasound-guided PVNB technique is well-accepted and has associated advantages, including high safety, high success rate, and less related complications. In recent years, many studies have shown that the analgesic effects of PVNB and thoracic epidural block are equivalent (20, 21). The former only blocks the sympathetic nerve on the surgical side; has little effect on the normal physiological functions; does not inhibit the cardiac sympathetic nerve, blood vessels and myocardium contractile function; and is less likely to affect hemodynamics; consequently, it has achieved better results in anesthesia for thoracic surgery (22). This study showed that general anesthesia with PVNB combined with a laryngeal mask could be performed in pediatric thoracic surgery. This technique could produce complete analgesia with fewer stress reactions, stable vital signs, reduced perioperative opioid dosages, and no residual muscle relaxation. Gastrointestinal reactions after the operation were minimal. It was conducive to intestinal peristalsis and accelerates the recovery of digestive system, thus allowing patients to start eating earlier. Additionally, this method reduced the length of hospital stay after surgery and the total hospitalization expenses. Spontaneous breathing combined with PVNB was safe and feasible in infants and young children for VATS. This result was consistent with Wei's research in 3–8 year old children (12).

The bronchial blocker used for OLV is often displaced due to body position change and intraoperative traction, which affects the lung collapse on the operation side. In the intubation group in this study, displacement of tracheal blocker caused the failure of OLV in six patients, since the adjusting the blocker position was difficult. Thus, TLV had to be used for the operation. In the non-intubation group, the success rate of OLV established through iatrogenic pneumothorax under spontaneous breathing was high. During OLV, in addition to preventing hypoxemia and ventilator-related lung injury, attention must be paid to preventing CO₂ retention (23). Compared with the intubation group, the non-intubation group had a higher intraoperative minimum SpO₂ and a lower incidence of hypoxia. In the lateral decubitus position, patients in the non-intubation group can preserve spontaneous breathing and diaphragmatic movement. Thus, the incidence of atelectasis is low, and there is no significant imbalance in the ratio of ventilation to blood flow (5, 15). To avoid mediastinal flutter that may result in heart and respiration disfunction, we applied an iatrogenic pneumothorax with a low pressure of 6–12 mmHg. In addition, the absence of stimulation from the tracheal tube reduces airway stress and secretion production. There was no significant difference in the maximum P_ETCO₂ during surgery between the two groups, and there was no accumulation of CO₂. This is mainly because the operation time was short, and breathing could return to normal soon after the operation (24)). However, due to prolonged one-lung spontaneous breathing or mediastinal flutter during the operation in the non-intubation group, six children (24%) had a short-term failure to effectively maintain blood oxygen saturation. For patients with spontaneous breathing, mediastinal flutter may occur due to the pressure differences between the left and right sides of the chest cavity after the chest is opened (24). In this study, one patient (4%) had severe coughing and mediastinal flutter. The operation continued after anesthesia was deepened, and the patient was switched to anesthetic machine-assisted ventilation. Respiratory management is most critical in this technique. The maintenance of oxygenation depends not only on whether the airway is open but also on the compression of the contralateral lung, the underlying lung disease, and airway secretions when patients are in the lateral decubitus position. Since the thoracic rib support ability of infants is low, the healthy lung is compressed in the lateral decubitus position, and its function can be greatly impacted. Due to the small diameter of the airway, a small volume of secretions or blood can cause a sharp increase in airway pressure, which can dramatically increase the breathing workload and affect anesthesia management. Therefore, it is necessary to strictly control indications and perform a comprehensive evaluation before surgery. The use of non-intubated anesthesia should be determined based on the technical capabilities of the anesthesia and surgical teams, monitoring and treatment conditions, surgical methods, disease types, individual differences in patients, and the actual conditions of various medical institutions (19). It is not applicable for patients who are undergoing complicated surgery that may involve a prolonged operation, massive blood loss (19), and the occurrence of mediastinal flutter. Appropriate sedation and analgesia during the operation should be performed with strict monitoring and enhanced management, and endotracheal intubation should be always prepared and ready for use if necessary.

Several limitations should be considered. First, this is a retrospective study so patients’ characteristics may result in severe selection bias. There are many respiratory parameters of spontaneous respiration, and mechanical ventilation cannot be collected completely. For a retrospective study, we collected all information from medical records. Regretfully, most of the time, the medical records cannot offer detailed information of the operation. It only mentions which side or lobe of lung was resected but not specific segments. Second, the duration of lung segmentectomy is short so we cannot draw a conclusion that non-intubated general anesthesia with preserved spontaneous breathing combined with PVNB is feasible for other thoracic surgeries. Third, there is not a reliable method to monitor the depth of anesthesia for young children. The BIS monitor is recommended for children over 2 years old commonly (25). Hence, the value of BIS to be monitored needs more studies. Furthermore, only patients classified as ASA I or II were included in the present study, and there are insufficient data about patients with perioperative lung dysfunction. According to Janík et al., not all patients are suitable for such a procedure, especially those with severe emphysema, obese patients, and those with a problematic oropharyngeal configuration—Mallampati score (19). The indication for non-intubated anesthesia techniques needs more precise study, especially about perioperative lung function (26).

Non-intubated general anesthesia with preserved spontaneous breathing combined with PVNB is a safe and feasible lung separation technique in infants and young children for VATS. It can provide good exposure during thoracoscopic surgery, can avoid the complications associated with mechanical ventilation, and is prone to quick postoperative recovery and fewer postoperative complications. It can promote rapid rehabilitation and shorten the length of hospital stay, thus reducing medical costs. However, strict control of indications and anesthesia management by experienced anesthesiologists are the keys to success. For some limits in this study, prospective studies should be encouraged to verify the results by compatible surgery methods and refined airway management.