Lymph node dissection along the recurrent laryngeal nerves in patients with oesophageal cancer who had undergone chemoradiotherapy: is it safe?

Lymph node dissection along the recurrent laryngeal nerves in patients with oesophageal cancer... Abstract OBJECTIVES Upper mediastinal lymph node dissection (LND)—especially along the recurrent laryngeal nerve (RN)—is the most challenging part of oesophageal cancer surgery. We investigated whether thoracoscopic RN LND may be safely performed in patients with oesophageal cancer who had undergone chemoradiotherapy (CRT). METHODS Patients with oesophageal cancer who had undergone thoracoscopic RN LND (n = 103) were divided into 2 groups according to whether they had prior treatment with CRT or not [the CRT group (n = 65) vs the upfront surgery group (n = 38), respectively]. All patients were operated on by a single surgeon. Intergroup comparisons were made in terms of (i) the number of dissected nodes, (ii) rates of RN palsy and (iii) rates of perioperative complications. The learning curve for the RN LND procedure was investigated using the cumulative sum method. RESULTS RN LND after CRT was more technically challenging when performed in the left side. Complete skeletonization of the left RN was achieved only in 66.2% of patients in the CRT group (vs 86.8% in the upfront surgery group; P = 0.022). The rate of postoperative left side RN palsy was significantly higher in the CRT group (26.6%) than in the upfront surgery group (7.9%, P = 0.022), albeit resulting in neither higher pneumonia rates nor longer hospital stays. The cumulative sum analysis revealed a steep learning curve for left RN LND in the CRT group. Unfortunately, an acceptable proficiency (left RN palsy rate: 15%) was not achievable even after treatment in 65 cases. CONCLUSIONS Thoracoscopic RN LND is safe but poses significant challenges in CRT-treated patients. Oesophageal cancer, Minimally invasive oesophagectomy, Recurrent laryngeal nerve, Lymph node dissection INTRODUCTION Lymph node metastases are common in patients with oesophageal cancer [1]. In particular, patients with oesophageal squamous cell carcinoma frequently presented with show metastatic spread to upper mediastinum nodes—especially along the bilateral recurrent laryngeal nerve (RN) [1–3]. The reported rates of lymph node metastases over the left- and right-side RN are as high as 15% and 30%, respectively [2, 3]. Based on the Japanese Nationwide Database, the RN nodal stations are those characterized by the highest efficacy index (calculated by multiplying the incidence of metastases to any given station with the 5-year survival rate of patients presented with metastases to that specific station) after removal [3]. However, RN lymph node dissection (LND)—particularly along the left side—is not yet considered as the standard of care mainly because of significant technical challenges that portend a significant risk of morbidity. Accordingly, high rates of RN palsy (unilaterally 10–60%; bilaterally 2–33%) have been reported in the literature [4–7]. The occurrence of RN palsy can in turn result in major postoperative complications (e.g. aspiration pneumonia) and even in postoperative mortality (especially when bilateral palsy occurs) [8, 9]. The implementation of RN LND in patients who had undergone chemoradiotherapy (CRT) is even more technically demanding. Accordingly, radiation therapy-induced fibrosis may obscure mediastinal microanatomy, ultimately making RN LND hazardous to the point that some surgeons consider prior radiation therapy as a relative contraindication to RN LND [10]. Theoretically, an extensive RN LND may be performed under the magnified view of a high-resolution thoracoscope using advanced endoscopes and instruments with multiple degrees of freedom. Unfortunately, published data on the safety of RN nodal dissection in patients with oesophageal cancer who had undergone CRT remain scarce. After the adoption of thoracoscopic oesophagectomy as of 2007, our surgical team has started including the bilateral RN area into its routine LND procedure in 2010. Initially, this approach was limited to patients who did not undergo CRT but—as of 2012—it was extended to all patients, regardless of previous CRT. In this retrospective study, we examined the feasibility and safety of thoracoscopic RN LND in patients with oesophageal cancer who had been previously treated with CRT. The study period spanned through 3 years, and all of the procedures were performed by a single surgeon. The study patients were divided into 2 groups according to previous treatment with CRT or not [the CRT group vs the upfront surgery (US) group, respectively]. Intergroup comparisons were made in terms of (i) the number of dissected nodes, (ii) rates of RN palsy and (iii) rates of perioperative complications. MATERIALS AND METHODS Patients The ethical approval for this retrospective analysis was granted by the local institutional review board. The study sample consisted of consecutive patients with oesophageal cancer who had undergone thoracoscopic oesophagectomy and RN LND between January 2013 and May 2016. All operations were performed by a single surgeon (Y.-K.C.). Patients were divided into 2 groups according to prior use of CRT or not (the CRT group vs the US group, respectively). Pretreatment staging was based on the results of chest and abdomen computed tomography (CT) scans, positron emission tomography (PET) imaging and endoscopic ultrasound. Patients were staged according to the American Joint Committee on Cancer (AJCC) staging criteria, 7th edition (2010). The severity of comorbidities was determined using the Charlson comorbidity index. The total score ranges from 0 to 37, with a score of 0 indicating the absence of comorbidities. Indications for chemoradiotherapy It is our institutional policy to offer CRT before surgery to all patients who are clinically staged as T2N0M0 or higher. CRT consisted of either a combination of cisplatin plus 5-fluorouracil or carboplatin plus paclitaxel administered with concurrent radiation therapy (dose 41.4–45 Gy). Patients underwent surgery at 6–8 weeks after completion of CRT. The standard surgical approach consisted of thoracoscopic oesophagectomy followed by the creation of a gastric tube (through either laparotomy or laparoscopy) and a cervical oesophagogastric anastomosis via left neck incision. A 2-field LND (which included the lower mediastinal/upper abdominal and bilateral RN lymph nodes) was performed in all patients [11]. We limited the use of neck LND only in patients who had an evidence of nodal involvement on preoperative CT or positron emission tomography imaging. Patients were in the intensive care unit for at least 1 night after surgery and subsequently transferred to their appropriate ward. Jejunostomy tube feeding was started 24 h after surgery. We performed a leak test on the 7th postoperative day. Thoracic procedure with the patient lying in the semiprone position The thoracoscopic procedure used in our centre has been previously described in detail [12]. In brief, the patient was placed in the left semiprone position while undergoing a single-lung ventilation (Fig. 1A). We created an artificial pneumothorax through CO2 insufflation (pressure 6–8 mmHg) and then positioned a total of 4 trocars (Fig. 1B). After incision of the mediastinal pleura, the layer between the thoracic duct and the oesophagus was dissected to ensure the maximum mobilization of the posterior part of the oesophagus. The mediastinal pleura was then incised along the course of the right vagus nerve until the lower margin of the subclavian artery with the goal of identifying the right RN. Left RN LND commenced with the exposure of the cartilage portion of the trachea. The oesophagus was then retracted towards the dorsal side either by surgeon’s left hand or a loop; meanwhile, the trachea was pushed aside by an assistant to optimize the operative exposure. The soft tissue between the left side trachea and the oesophagus (which included both lymph nodes and the left RN) was thoroughly dissected (Fig. 2). A careful skeletonization of the left RN was subsequently performed to achieve a complete lymphadenectomy. Figure 1: View largeDownload slide (A) Operating room set-up. (B) Port placement for thoracoscopic oesophagectomy when the patient is placed in the semiprone position. Figure 1: View largeDownload slide (A) Operating room set-up. (B) Port placement for thoracoscopic oesophagectomy when the patient is placed in the semiprone position. Figure 2: View largeDownload slide A representation of the reciprocal positioning of the oesophagus, trachea, left recurrent laryngeal nerve and its ancillary LNs (located ventrally and dorsally) when the patient is placed in the semiprone position. The green dashed line indicates the optimal dissection path for achieving maximum recurrent laryngeal nerve LN exposure. LN: lymph node. Figure 2: View largeDownload slide A representation of the reciprocal positioning of the oesophagus, trachea, left recurrent laryngeal nerve and its ancillary LNs (located ventrally and dorsally) when the patient is placed in the semiprone position. The green dashed line indicates the optimal dissection path for achieving maximum recurrent laryngeal nerve LN exposure. LN: lymph node. Definitions of outcomes The system proposed by the Esophagectomy Complications Consensus Group was used to define perioperative complications [13]. In patients with hoarseness or voice modifications that occurred after surgery, an otolaryngologist assessed vocal cord function using a flexible laryngoscope. Thirty-day mortality and 90-day mortality were defined as any death occurring within 30 or 90 days, respectively, after the date of surgery. Deaths occurring at any time during the postoperative hospital stay were used to calculate in-hospital mortality. Thirty-day readmission was defined as any in-patient admission to our institution within 30 days of discharge from the postoperative stay. Post-therapy surveillance Patients were scheduled for chest X-ray every 3 months and for CT every 6 months postoperatively. Survival was assessed every 6 months through contact with the patient’s physicians or a review of medical records. In case of missing information, data were retrieved from the National Cancer Registry Database of Taiwan. Statistical analysis Continuous data are expressed as both medians and interquartile ranges and means and standard deviations. Skewed variables were compared with the Mann–Whitney U-test, whereas normally distributed data were analysed with the Student’s t-test. Categorical data were compared with the χ2 test or the Fisher’s exact test (for small sample sizes). Cumulative sum (CUSUM) analysis was used to chart the occurrence of RN palsy, the major complication following an extensive RN LND [14–16]. The definition of acceptable and unacceptable failure rates is a key prerequisite to construct a CUSUM chart. Based on the published literature, the RN palsy rates after extensive RN LND are approximately 10–58.4% for the left side and 5–10% for the right side [4–6]. However, these data are referring to CRT naïve patients. We reasoned that such rates in CRT-treated patients would be either similar or slightly higher. Therefore, we set palsy rates of 15% and 50% (for the left side) and 5% and 10% (for the right side) as acceptable and unacceptable, respectively. The probabilities of Type I (α) and Type II (β) error were set at 0.05 and 0.20, respectively. A positive trend of the CUSUM curve indicates failure, whereas a negative trend is suggestive of successful procedures. When the line crossed the upper decision limit while ascending, the actual palsy rate was equal to the unacceptable value (with a Type I error probability of 0.05). When the line crossed the lower decision limit while descending, the actual palsy rate did not differ from the acceptable recurrence rate (with a Type II error probability of 0.20). All calculations were performed using SAS, version 9.3 (SAS Institute Inc., Cary, NC, USA) and SPSS, version 22.0 (IBM Corp., Armonk, NY, USA). Two-tailed P-values <0.05 were considered statistically significant. RESULTS Patient characteristics Between January 2013 and May 2016, we identified 103 patients with oesophageal cancer patients who had undergone thoracoscopic oesophagectomy and bilateral RN LND. Of them, 65 received CRT before surgery (the CRT group), whereas 38 did not (the US group). Table 1 summarizes the demographic and clinical characteristics of the study patients. Squamous cell carcinoma was the most common histological subtype. The 2 study groups did not differ significantly in terms of age, sex, histological subtype, and tumour location. An advanced clinical stage at diagnosis was more common for patients in the CRT group. Table 1: General characteristics of the study patients   Entire cohort   US (n = 38)  CRT (n = 65)  P-value  Age (years)   Mean ± SD  53.3 ± 7.3  54.6 ± 7.9  0.415   Median (IQR)  51.5 (48.8–57.5)  52 (49–60)  0.451  CCI  0.9 ± 1.1  0.8 ± 1.0  0.523  Sex      0.625   Male  36 (94.7)  63 (96.9)     Female  2 (5.3)  2 (3.1)    Histology   SCC  35 (92.1)  62 (95.4)     ADC  3 (7.9)  3 (4.6)    Clinical stage      <0.001   I/II  32 (84.2)  3 (4.6)     III  6 (15.8)  62 (95.4)    Tumour location      0.759   Upper third  8 (21.1)  12 (18.5)     Middle third  17 (44.7)  34 (52.3)     Lower third  13 (34.2)  19 (29.2)    Chemotherapy regimen      NA   5-FU + cisplatin  NA  35     Paclitaxel + carboplatin  NA  30      Entire cohort   US (n = 38)  CRT (n = 65)  P-value  Age (years)   Mean ± SD  53.3 ± 7.3  54.6 ± 7.9  0.415   Median (IQR)  51.5 (48.8–57.5)  52 (49–60)  0.451  CCI  0.9 ± 1.1  0.8 ± 1.0  0.523  Sex      0.625   Male  36 (94.7)  63 (96.9)     Female  2 (5.3)  2 (3.1)    Histology   SCC  35 (92.1)  62 (95.4)     ADC  3 (7.9)  3 (4.6)    Clinical stage      <0.001   I/II  32 (84.2)  3 (4.6)     III  6 (15.8)  62 (95.4)    Tumour location      0.759   Upper third  8 (21.1)  12 (18.5)     Middle third  17 (44.7)  34 (52.3)     Lower third  13 (34.2)  19 (29.2)    Chemotherapy regimen      NA   5-FU + cisplatin  NA  35     Paclitaxel + carboplatin  NA  30    Data are represented as mean ± standard deviation or counts (percentages), as appropriate. ADC: adenocarcinoma; CCI: Charlson comorbidity index; CRT: chemoradiotherapy; 5-FU: 5-fluorouracil; IQR: interquartile range; NA: not applicable; SCC: squamous cell carcinoma; SD: standard deviation; US: upfront surgery. Table 1: General characteristics of the study patients   Entire cohort   US (n = 38)  CRT (n = 65)  P-value  Age (years)   Mean ± SD  53.3 ± 7.3  54.6 ± 7.9  0.415   Median (IQR)  51.5 (48.8–57.5)  52 (49–60)  0.451  CCI  0.9 ± 1.1  0.8 ± 1.0  0.523  Sex      0.625   Male  36 (94.7)  63 (96.9)     Female  2 (5.3)  2 (3.1)    Histology   SCC  35 (92.1)  62 (95.4)     ADC  3 (7.9)  3 (4.6)    Clinical stage      <0.001   I/II  32 (84.2)  3 (4.6)     III  6 (15.8)  62 (95.4)    Tumour location      0.759   Upper third  8 (21.1)  12 (18.5)     Middle third  17 (44.7)  34 (52.3)     Lower third  13 (34.2)  19 (29.2)    Chemotherapy regimen      NA   5-FU + cisplatin  NA  35     Paclitaxel + carboplatin  NA  30      Entire cohort   US (n = 38)  CRT (n = 65)  P-value  Age (years)   Mean ± SD  53.3 ± 7.3  54.6 ± 7.9  0.415   Median (IQR)  51.5 (48.8–57.5)  52 (49–60)  0.451  CCI  0.9 ± 1.1  0.8 ± 1.0  0.523  Sex      0.625   Male  36 (94.7)  63 (96.9)     Female  2 (5.3)  2 (3.1)    Histology   SCC  35 (92.1)  62 (95.4)     ADC  3 (7.9)  3 (4.6)    Clinical stage      <0.001   I/II  32 (84.2)  3 (4.6)     III  6 (15.8)  62 (95.4)    Tumour location      0.759   Upper third  8 (21.1)  12 (18.5)     Middle third  17 (44.7)  34 (52.3)     Lower third  13 (34.2)  19 (29.2)    Chemotherapy regimen      NA   5-FU + cisplatin  NA  35     Paclitaxel + carboplatin  NA  30    Data are represented as mean ± standard deviation or counts (percentages), as appropriate. ADC: adenocarcinoma; CCI: Charlson comorbidity index; CRT: chemoradiotherapy; 5-FU: 5-fluorouracil; IQR: interquartile range; NA: not applicable; SCC: squamous cell carcinoma; SD: standard deviation; US: upfront surgery. Surgical quality and perioperative outcomes Table 2 shows the main data related to surgical quality. No conversion to open thoracotomy was required in either group. There were no significant inter-group differences in terms of thoracic operating time, intraoperative blood loss and the need for blood transfusions. As far as RN LND feasibility is concerned, the ability to achieve a complete LND (i.e. nerve skeletonization) and the number of dissected nodes were significantly lower over the left side in the CRT group; however, such differences were not evident for the right side. Complete skeletonization of the left RN was achieved only in 66.2% of patients in the CRT group, a percentage significantly lower than that obtained in the US group (86.8%; P = 0.022). Table 2: Surgical quality obtained in the study patients   Entire study cohort   US (n = 38)  CRT (n = 65)  P-value  Thoracic operating time (min)   Mean ± SD  166.1 ± 40.1  159.6 ± 50.0  0.500   Median (IQR)  165.5 (134–192.5)  155 (126.5–189)  0.430  EBL (mL)   Mean ± SD  110.0 ± 93.2  139.4 ± 184.3  0.362   Median (IQR)  50 (50–162.5)  100 (50–100)  0.582  Intraoperative BT  1 (2.6)  6 (9.2)  0.199  Successful right RN identification  35 (92.1)  54 (83.1)  0.197  Successful left RN identification  33 (86.8)  43 (66.2)  0.022  R0 surgery  38 (100)  57 (87.7)  0.045  Number of dissected lymph nodes   Total number  41.9 ± 13.1  29.1 ± 12.5  <0.001    Mediastinal nodes  20.8 ± 9.0  14.3 ± 6.1  <0.001    Right RN nodes  3.1 ± 2.5  2.9 ± 2.1  0.556    Left RN nodes  4.9 ± 3.1  3.6 ± 2.6  0.043    Abdominal nodes  19.8 ± 11.8  13.4 ± 7.1  0.001    Entire study cohort   US (n = 38)  CRT (n = 65)  P-value  Thoracic operating time (min)   Mean ± SD  166.1 ± 40.1  159.6 ± 50.0  0.500   Median (IQR)  165.5 (134–192.5)  155 (126.5–189)  0.430  EBL (mL)   Mean ± SD  110.0 ± 93.2  139.4 ± 184.3  0.362   Median (IQR)  50 (50–162.5)  100 (50–100)  0.582  Intraoperative BT  1 (2.6)  6 (9.2)  0.199  Successful right RN identification  35 (92.1)  54 (83.1)  0.197  Successful left RN identification  33 (86.8)  43 (66.2)  0.022  R0 surgery  38 (100)  57 (87.7)  0.045  Number of dissected lymph nodes   Total number  41.9 ± 13.1  29.1 ± 12.5  <0.001    Mediastinal nodes  20.8 ± 9.0  14.3 ± 6.1  <0.001    Right RN nodes  3.1 ± 2.5  2.9 ± 2.1  0.556    Left RN nodes  4.9 ± 3.1  3.6 ± 2.6  0.043    Abdominal nodes  19.8 ± 11.8  13.4 ± 7.1  0.001  Data are represented as mean ± standard deviation or counts (percentages), as appropriate. BT: blood transfusions; CRT: chemoradiotherapy; EBL: estimated blood loss; IQR: interquartile range; RN: recurrent laryngeal nerve; SD: standard deviation; US: upfront surgery. Table 2: Surgical quality obtained in the study patients   Entire study cohort   US (n = 38)  CRT (n = 65)  P-value  Thoracic operating time (min)   Mean ± SD  166.1 ± 40.1  159.6 ± 50.0  0.500   Median (IQR)  165.5 (134–192.5)  155 (126.5–189)  0.430  EBL (mL)   Mean ± SD  110.0 ± 93.2  139.4 ± 184.3  0.362   Median (IQR)  50 (50–162.5)  100 (50–100)  0.582  Intraoperative BT  1 (2.6)  6 (9.2)  0.199  Successful right RN identification  35 (92.1)  54 (83.1)  0.197  Successful left RN identification  33 (86.8)  43 (66.2)  0.022  R0 surgery  38 (100)  57 (87.7)  0.045  Number of dissected lymph nodes   Total number  41.9 ± 13.1  29.1 ± 12.5  <0.001    Mediastinal nodes  20.8 ± 9.0  14.3 ± 6.1  <0.001    Right RN nodes  3.1 ± 2.5  2.9 ± 2.1  0.556    Left RN nodes  4.9 ± 3.1  3.6 ± 2.6  0.043    Abdominal nodes  19.8 ± 11.8  13.4 ± 7.1  0.001    Entire study cohort   US (n = 38)  CRT (n = 65)  P-value  Thoracic operating time (min)   Mean ± SD  166.1 ± 40.1  159.6 ± 50.0  0.500   Median (IQR)  165.5 (134–192.5)  155 (126.5–189)  0.430  EBL (mL)   Mean ± SD  110.0 ± 93.2  139.4 ± 184.3  0.362   Median (IQR)  50 (50–162.5)  100 (50–100)  0.582  Intraoperative BT  1 (2.6)  6 (9.2)  0.199  Successful right RN identification  35 (92.1)  54 (83.1)  0.197  Successful left RN identification  33 (86.8)  43 (66.2)  0.022  R0 surgery  38 (100)  57 (87.7)  0.045  Number of dissected lymph nodes   Total number  41.9 ± 13.1  29.1 ± 12.5  <0.001    Mediastinal nodes  20.8 ± 9.0  14.3 ± 6.1  <0.001    Right RN nodes  3.1 ± 2.5  2.9 ± 2.1  0.556    Left RN nodes  4.9 ± 3.1  3.6 ± 2.6  0.043    Abdominal nodes  19.8 ± 11.8  13.4 ± 7.1  0.001  Data are represented as mean ± standard deviation or counts (percentages), as appropriate. BT: blood transfusions; CRT: chemoradiotherapy; EBL: estimated blood loss; IQR: interquartile range; RN: recurrent laryngeal nerve; SD: standard deviation; US: upfront surgery. Table 3 depicts the perioperative course observed in the 2 study groups. As far as morbidity after RN LND is concerned, a total of 23 patients experienced RN palsy. Of them, 20 and 3 cases had left-sided and right-sided lesions, respectively. No patients had bilateral RN palsy. The rate of postoperative RN palsy in the left side was significantly higher in the CRT group than in the US group (26.6% vs 7.9%, respectively, P = 0.022); however, no significant inter-group differences were observed in terms of right-side RN palsy rates. Despite a significantly higher left RN palsy rate in the CRT group, no significant differences were observed between the 2 groups with regard to pulmonary complications, anastomotic leaks, 30-day mortality/90-day mortality, and 30-day readmission rates. With regard to 90-day mortality in the CRT group, we registered a total of 3 cases. Of them, 2 died because of pulmonary complications while the patients were still hospitalized after surgery. Consequently, these 2 patients were also considered as cases of in-hospital mortality. The remaining patient died at 58 days after surgery because of brain haemorrhage caused by tumour metastases. Table 3: Short-term perioperative outcomes in the study patients   US (n = 38)  CRT (n = 65)  P-value  MV time (h)   Mean ± SD  10.1 ± 10.0  26.0 ± 101.1  0.337   Median (IQR)  5.6 (3.5–14.9)  7.2 (4.1–17.2)  0.340  ICU stay (h)   Mean ± SD  38.9 ± 46.2  57.9 ± 109.9  0.315   Median (IQR)  20.1 (17.6–43.9)  34.3 (18.5–45.4)  0.367  Pneumonia  3 (7.9)  11 (16.9)  0.197  Chylothorax  4 (10.5)  2 (3.1)  0.190  MV >72 h  0 (0)  3 (4.6)  0.295  RN palsy   Right sided  2 (5.3)  1 (1.6)  0.554   Left sided  3 (7.9)  17 (26.6)  0.022  Anastomotic leak  2 (5.3)  6 (9.2)  0.468  30-Day mortality  0 (0)  0 (0)  NA  In-hospital mortality  0 (0)  2 (3.1)  0.275  90-Day mortality  0 (0)  3 (4.6)  0.179  30-Day readmission  4 (10.5)  13/63 (20.6)  0.188    US (n = 38)  CRT (n = 65)  P-value  MV time (h)   Mean ± SD  10.1 ± 10.0  26.0 ± 101.1  0.337   Median (IQR)  5.6 (3.5–14.9)  7.2 (4.1–17.2)  0.340  ICU stay (h)   Mean ± SD  38.9 ± 46.2  57.9 ± 109.9  0.315   Median (IQR)  20.1 (17.6–43.9)  34.3 (18.5–45.4)  0.367  Pneumonia  3 (7.9)  11 (16.9)  0.197  Chylothorax  4 (10.5)  2 (3.1)  0.190  MV >72 h  0 (0)  3 (4.6)  0.295  RN palsy   Right sided  2 (5.3)  1 (1.6)  0.554   Left sided  3 (7.9)  17 (26.6)  0.022  Anastomotic leak  2 (5.3)  6 (9.2)  0.468  30-Day mortality  0 (0)  0 (0)  NA  In-hospital mortality  0 (0)  2 (3.1)  0.275  90-Day mortality  0 (0)  3 (4.6)  0.179  30-Day readmission  4 (10.5)  13/63 (20.6)  0.188  Data are represented as mean ± standard deviation or counts (percentages), as appropriate. CRT: chemoradiotherapy; ICU: intensive care unit; IQR: interquartile range; MV: mechanical ventilator; NA: not applicable; RN: recurrent laryngeal nerve; SD: standard deviation; US: upfront surgery. Table 3: Short-term perioperative outcomes in the study patients   US (n = 38)  CRT (n = 65)  P-value  MV time (h)   Mean ± SD  10.1 ± 10.0  26.0 ± 101.1  0.337   Median (IQR)  5.6 (3.5–14.9)  7.2 (4.1–17.2)  0.340  ICU stay (h)   Mean ± SD  38.9 ± 46.2  57.9 ± 109.9  0.315   Median (IQR)  20.1 (17.6–43.9)  34.3 (18.5–45.4)  0.367  Pneumonia  3 (7.9)  11 (16.9)  0.197  Chylothorax  4 (10.5)  2 (3.1)  0.190  MV >72 h  0 (0)  3 (4.6)  0.295  RN palsy   Right sided  2 (5.3)  1 (1.6)  0.554   Left sided  3 (7.9)  17 (26.6)  0.022  Anastomotic leak  2 (5.3)  6 (9.2)  0.468  30-Day mortality  0 (0)  0 (0)  NA  In-hospital mortality  0 (0)  2 (3.1)  0.275  90-Day mortality  0 (0)  3 (4.6)  0.179  30-Day readmission  4 (10.5)  13/63 (20.6)  0.188    US (n = 38)  CRT (n = 65)  P-value  MV time (h)   Mean ± SD  10.1 ± 10.0  26.0 ± 101.1  0.337   Median (IQR)  5.6 (3.5–14.9)  7.2 (4.1–17.2)  0.340  ICU stay (h)   Mean ± SD  38.9 ± 46.2  57.9 ± 109.9  0.315   Median (IQR)  20.1 (17.6–43.9)  34.3 (18.5–45.4)  0.367  Pneumonia  3 (7.9)  11 (16.9)  0.197  Chylothorax  4 (10.5)  2 (3.1)  0.190  MV >72 h  0 (0)  3 (4.6)  0.295  RN palsy   Right sided  2 (5.3)  1 (1.6)  0.554   Left sided  3 (7.9)  17 (26.6)  0.022  Anastomotic leak  2 (5.3)  6 (9.2)  0.468  30-Day mortality  0 (0)  0 (0)  NA  In-hospital mortality  0 (0)  2 (3.1)  0.275  90-Day mortality  0 (0)  3 (4.6)  0.179  30-Day readmission  4 (10.5)  13/63 (20.6)  0.188  Data are represented as mean ± standard deviation or counts (percentages), as appropriate. CRT: chemoradiotherapy; ICU: intensive care unit; IQR: interquartile range; MV: mechanical ventilator; NA: not applicable; RN: recurrent laryngeal nerve; SD: standard deviation; US: upfront surgery. As far as pathology results are concerned, 9 (13.8%) patients in the CRT group showed nodal involvement in the RLN area. The positive rates over the right and left RN were 10.8% and 4.6%, respectively. Notably, 1 patient had bilateral RN metastases. Seven (18.4%) patients in the US group had RLN metastases. The positive rates over the right and left RN were 15.8% and 7.9%, respectively. Of note, bilateral RN metastases were observed in 2 cases. Learning curve of recurrent laryngeal nerve lymph node dissection after chemoradiotherapy Because the rates of right RN palsy (in both groups) and left RN palsy (in the US group) were low, we specifically focused on the learning curves of left RN LND in the CRT group—which was characterized by the highest palsy rate. For this reason, only the learning curves for complete RN LND in the subgroup of 43 patients in which this procedure was successfully completed were presented. Although the CUSUM curve never breached the unacceptable threshold (50% palsy) throughout the study, we were also unable to observe a point at which the line descended to the boundary set as acceptable (i.e. 15% palsy; Fig. 3). Figure 3: View largeDownload slide The cumulative sum chart for the left recurrent laryngeal nerve (RN) lymph node dissection in patients who had undergone chemoradiotherapy. Only patients who received a complete left RN lymph node dissection (n = 43) were included. A positive trend of the cumulative sum curve indicates failure, whereas a negative trend is suggestive of successful procedures. An RN palsy rate of 15% was deemed as acceptable, whereas a 50% rate was regarded as unacceptable. When the learning curve reached the boundary for acceptability (while descending) or unacceptability (while ascending), the surgical proficiency of the preceding series was classified as acceptable or unacceptable, respectively. Figure 3: View largeDownload slide The cumulative sum chart for the left recurrent laryngeal nerve (RN) lymph node dissection in patients who had undergone chemoradiotherapy. Only patients who received a complete left RN lymph node dissection (n = 43) were included. A positive trend of the cumulative sum curve indicates failure, whereas a negative trend is suggestive of successful procedures. An RN palsy rate of 15% was deemed as acceptable, whereas a 50% rate was regarded as unacceptable. When the learning curve reached the boundary for acceptability (while descending) or unacceptability (while ascending), the surgical proficiency of the preceding series was classified as acceptable or unacceptable, respectively. DISCUSSION To our knowledge, this is the first study to specifically investigate the feasibility, safety and learning curve of thoracoscopic RN LND in patients with oesophageal cancer who had undergone CRT. Our data indicate that RN LND is feasible even after CRT, although the technical challenges to be faced are greater than in the CRT-naïve patients. Despite the RN palsy rate being 3.4-fold greater in the CRT group than in the US group (7.9% vs 26.6%, respectively), such higher occurrence did not result in increased rates of pneumonia and/or death. Such positive results are likely due to (i) prompt diagnosis and aggressive clinical management of RN palsy and (ii) the absence of bilateral palsy (0%). All our patients underwent a vocal cord assessment by an otolaryngologist even in the presence of small and/or subtle voice modifications. In cases showing unilateral palsy and a wide glottal gap, intracordal hyaluronic acid injections were performed as early as possible (in general within 3 months) to reduce the likelihood of subsequent choking and aspiration pneumonia [17, 18]. Despite such recent therapeutic advances, unilateral RN palsy remains a significant source of morbidity and impaired quality of life. More remarkably, major efforts should be made to avoid bilateral RN palsy—a devastating complication for which no effective treatment exists [19]. In an era during which preoperative CRT is becoming the standard of care, the prevention of nerve injury during surgery is of utmost importance. Complete left RN LND requires a thorough removal of soft tissues and lymph nodes located along the nerve course (i.e. from the aortic arch level to the inferior thyroid artery). Because of the limited exposure of the left paratracheal space, the dissection of nodes located ventrally to the left RN generally represents the most challenging part of the surgical procedure. Two main options have been proposed to increase surgical exposure, including (i) the suspension of the oesophagus with loops of umbilical tapes (with the goal of maximizing the exposure of the left RN) and (ii) the replacement of a double-lumen tube with a single-lumen tube (either with or without a blocker) coupled with CO2 insufflation (with the goal of reducing the tracheal rigidity) [20, 21]. Previous studies conducted in the CRT-naïve patients by experienced surgeons supported the utility of these manoeuvres to reduce both RN palsy and deaths [20, 21]. In accordance with previous reports, the implementation of these surgical options allowed us to obtain low rates of left RN nerve palsy (7.9%) and mortality (0%) in our US patients [22]. However, left RN LND remains generally difficult to be achieved in patients who had undergone CRT—even when repeatedly performed by the same surgeon. As shown in Table 2, CRT-induced mediastinal fibrosis hampered a successful RN identification in approximately one-third of patients who had been treated with this approach (when compared with 10% only in the US group). When the nerve is not clearly identifiable, contusions, excessive stretching and thermal injury that may occur during surgical manipulation can ultimately result in nervous damage. This offers a compelling explanation for the RN palsy rate being higher in the CRT group than in the US group. Another important observation from our study is that RN LND is characterized by a steep learning curve (as shown by the CUSUM analysis), which suggests that an adequate proficiency was not achieved even after treatment of 65 cases. There are at least 3 strategies that may be helpful to reduce the occurrence of RN palsy. First, left RN LND can be selectively applied only to patients at high risk of having left ypN-RN disease (while invariably maintaining a radical approach for right RN LND). It should be noted that the clinical utility of extensive LND after CRT remains highly controversial—especially for ycN-negative patients [23]. We have previously shown that the pre-CRT cN-RN status is the most important predictor for positive ypN-RN [24]. Notably, routine RN nodal dissection still yielded positive results [ypN-RN(+)] in 30.8% of patients with cN-RN(+) who were judged to be ycN-RN(−) following CRT. Conversely, RN LND resulted in a low (10%) positive LN discovery rate in the subgroup of patients who did not have radiological evidence of RN LN involvement prior to CRT [24]. We speculate that RN LND might be omitted in this specific subgroup, especially when potential surgical risks are expected to outweigh clinical benefits. Second, we recommend a careful monitoring of RN function during left RN LND. This approach may reduce the incidence of nerve injury and has been shown to be clinically useful in other surgical procedures (e.g. thyroid operation) [25, 26]. However, data in the field of oesophageal cancer surgery are controversial. Zhong et al. [27] have previously shown that RN monitoring during oesophagectomy results in (i) a higher number of harvested positive mediastinal lymph nodes and (ii) a lower incidence of postoperative RN paralysis and pneumonia (when compared with patients who did not undergo nerve monitoring). However, other reports failed to confirm such a protective effect [5]. Recently, promising results have been obtained by replacing classical intermittent monitoring with continuous intraoperative nerve monitoring. Continuous intraoperative nerve monitoring may allow detecting an impending RN injury in an earlier phase when compared with traditional intermittent monitoring. Consequently, a surgical manoeuvre that causes stretching or pressure on the RN may be promptly interrupted, resulting in a superior nerve function preservation [28]. Third, the use of EndoWrist instruments may be clinically useful during fine dissections, ultimately reducing the risk of RN palsy. There is ample evidence that robotic surgical systems incorporating 3-dimensional stereoscopic vision and arms with wrist-like joints can allow obtaining meticulous dissections in limited anatomical space. In addition, they may be helpful to reduce the steep learning curve typically associated with conventional laparoscopic surgery (e.g. as encountered for radical prostatectomy). We hypothesize that similar trends may exist for interventions performed in the narrow left paratracheal space, although further investigations are required to support this possibility. Limitations Some limitations of our study merit comment. First, our study design was retrospective, and the sample size was small, ultimately resulting in a reduced statistical power. Second, generalization from this study should be made with caution, as data were based on a single-centre experience obtained by a single surgeon. In this context, randomized longitudinal studies are warranted to confirm and expand our findings. CONCLUSION Thoracoscopic RN LND is safe but poses significant challenges in patients with oesophageal cancer who had undergone CRT. A significant learning curve exists for finalizing this procedure in a complication-free manner. Funding This work was supported by the Chang Gung Memorial Hospital, Taiwan [CORPG3G0771 and CORPG3G0781]. Conflict of interest: none declared. REFERENCES 1 Tachimori Y. Pattern of lymph node metastases of squamous cell esophageal cancer based on the anatomical lymphatic drainage system: efficacy of lymph node dissection according to tumor location. J Thorac Dis  2017; 9: S724. Google Scholar CrossRef Search ADS PubMed  2 Udagawa H, Ueno M, Shinohara H, Haruta S, Kaida S, Nakagawa M et al.   The importance of grouping of lymph node stations and rationale of three-field lymphoadenectomy for thoracic esophageal cancer. J Surg Oncol  2012; 106: 742– 7. Google Scholar CrossRef Search ADS PubMed  3 Tachimori Y, Ozawa S, Numasaki H, Matsubara H, Shinoda M, Toh Y et al.   Efficacy of lymph node dissection for each station based on esophageal tumor location. Esophagus  2016; 13: 1– 45. Google Scholar CrossRef Search ADS PubMed  4 Sato Y, Kosugi S-I, Aizawa N, Ishikawa T, Kano Y, Ichikawa H et al.   Risk factors and clinical outcomes of recurrent laryngeal nerve paralysis after esophagectomy for thoracic esophageal carcinoma. World J Surg  2016; 40: 129– 36. Google Scholar CrossRef Search ADS PubMed  5 Hikage M, Kamei T, Nakano T, Abe S, Katsura K, Taniyama Y et al.   Impact of routine recurrent laryngeal nerve monitoring in prone esophagectomy with mediastinal lymph node dissection. Surg Endosc  2017; 31: 2986– 96. Google Scholar CrossRef Search ADS PubMed  6 Taniyama Y, Miyata G, Kamei T, Nakano T, Abe S, Katsura K et al.   Complications following recurrent laryngeal nerve lymph node dissection in oesophageal cancer surgery. Interact CardioVasc Thorac Surg  2015; 20: 41– 6. Google Scholar CrossRef Search ADS PubMed  7 Ninomiya I, Okamoto K, Fujimura T, Fushida S, Osugi H, Ohta T. Oncologic outcomes of thoracoscopic esophagectomy with extended lymph node dissection: 10-year experience from a single center. World J Surg  2014; 38: 120– 30. Google Scholar CrossRef Search ADS PubMed  8 Gockel I, Kneist W, Keilmann A, Junginger T. Recurrent laryngeal nerve paralysis (RLNP) following esophagectomy for carcinoma. Eur J Surg Oncol  2005; 31: 277– 81. Google Scholar CrossRef Search ADS PubMed  9 Berry MF, Atkins BZ, Tong BC, Harpole DH, D’Amico TA, Onaitis MW. A comprehensive evaluation for aspiration after esophagectomy reduces the incidence of postoperative pneumonia. J Thorac Cardiovasc Surg  2010; 140: 1266– 71. Google Scholar CrossRef Search ADS PubMed  10 Osugi H, Narumiya K, Kudou K. Supracarinal dissection of the oesophagus and lymphadenectomy by MIE. J Thorac Dis  2017; 9: S741. Google Scholar CrossRef Search ADS PubMed  11 Wong B. More or less surgery for esophageal cancer: extent of lymphadenectomy in esophagectomy for squamous cell esophageal carcinoma How much is necessary. Dis Esophagus  1994; 7: 151– 5. Google Scholar CrossRef Search ADS   12 Chao Y-K, Hsieh M-J, Liu Y-H, Liu H-P. Lymph node evaluation in robot-assisted versus video-assisted thoracoscopic esophagectomy for esophageal squamous cell carcinoma: a propensity-matched analysis. World J Surg  2018; 42: 590– 8. Google Scholar CrossRef Search ADS PubMed  13 Low DE, Alderson D, Cecconello I, Chang AC, Darling GE, D'Journo XB et al.   International consensus on standardization of data collection for complications associated with esophagectomy: Esophagectomy Complications Consensus Group (ECCG). Ann Surg  2015; 262: 286– 94. Google Scholar CrossRef Search ADS PubMed  14 Lim T, Soraya A, Ding L, Morad Z. Assessing doctors’ competence: application of CUSUM technique in monitoring doctors’ performance. Int J Qual Health Care  2002; 14: 251– 8. Google Scholar CrossRef Search ADS PubMed  15 Noyez L. Control charts, Cusum techniques and funnel plots. A review of methods for monitoring performance in healthcare. Interact CardioVasc Thorac Surg  2009; 9: 494– 9. Google Scholar CrossRef Search ADS PubMed  16 Williams SM, Parry BR, Schlup M. Quality control: an application of the cusum. BMJ  1992; 304: 1359. Google Scholar CrossRef Search ADS PubMed  17 Fang T-J, Hsin L-J, Chung H-F, Chiang H-C, Li H-Y, Wong AM et al.   Office-based intracordal hyaluronate injections improve quality of life in thoracic-surgery-related unilateral vocal fold paralysis. Medicine  2015; 94: e1787. Google Scholar CrossRef Search ADS PubMed  18 Fang TJ, Pei YC, Li HY, Wong A, Chiang HC. Glottal gap as an early predictor for permanent laryngoplasty in unilateral vocal fold paralysis. Laryngoscope  2014; 124: 2125– 30. Google Scholar CrossRef Search ADS PubMed  19 Baba M, Aikou T, Natsugoe S, Kusano C, Shimada M, Nakano S et al.   Quality of life following esophagectomy with three-field lymphadenectomy for carcinoma, focusing on its relationship to vocal cord palsy. Dis Esophagus  1998; 11: 28– 34. Google Scholar CrossRef Search ADS PubMed  20 Xi Y, Ma Z, Shen Y, Wang H, Feng M, Tan L et al.   A novel method for lymphadenectomy along the left laryngeal recurrent nerve during thoracoscopic esophagectomy for esophageal carcinoma. J Thorac Dis  2016; 8: 24. Google Scholar PubMed  21 Zhang R, Liu S, Sun H, Liu X, Wang Z, Qin J et al.   The application of single-lumen endotracheal tube anaesthesia with artificial pneumothorax in thoracolaparoscopic oesophagectomy. Interact CardioVasc Thorac Surg  2014; 19: 308– 10. Google Scholar CrossRef Search ADS PubMed  22 Shen Y, Zhang Y, Tan L, Feng M, Wang H, Khan MA et al.   Extensive mediastinal lymphadenectomy during minimally invasive esophagectomy: optimal results from a single center. J Gastrointest Surg  2012; 16: 715– 21. Google Scholar CrossRef Search ADS PubMed  23 Shridhar R, Hoffe SE, Almhanna K, Weber JM, Chuong MD, Karl RC et al.   Lymph node harvest in esophageal cancer after neoadjuvant chemoradiotherapy. Ann Surg Oncol  2013; 20: 3038– 43. Google Scholar CrossRef Search ADS PubMed  24 Li Z-G, Zhang X-B, Wen Y-W, Liu Y-H, Chao Y-K. Incidence and predictors of unsuspected recurrent laryngeal nerve lymph node metastases after neoadjuvant chemoradiotherapy in patients with esophageal squamous cell carcinoma. World J Surg  2018; doi: 10.1007/s00268-018-4516-y [Epub ahead of print]. 25 Randolph GW, Kamani D. Intraoperative electrophysiologic monitoring of the recurrent laryngeal nerve during thyroid and parathyroid surgery: experience with 1, 381 nerves at risk. Laryngoscope  2017; 127: 280– 6. Google Scholar CrossRef Search ADS PubMed  26 Barczyński M, Konturek A, Pragacz K, Papier A, Stopa M, Nowak W. Intraoperative nerve monitoring can reduce prevalence of recurrent laryngeal nerve injury in thyroid reoperations: results of a retrospective cohort study. World J Surg  2014; 38: 599– 606. Google Scholar CrossRef Search ADS PubMed  27 Zhong D, Zhou Y, Li Y, Wang Y, Zhou W, Cheng Q et al.   Intraoperative recurrent laryngeal nerve monitoring: a useful method for patients with esophageal cancer. Dis Esophagus  2014; 27: 444– 51. Google Scholar CrossRef Search ADS PubMed  28 Tsang RK, Law S. Adaptation of continuous intraoperative vagus nerve stimulation for monitoring of recurrent laryngeal nerve during minimally invasive esophagectomy. World J Surg  2016; 40: 137– 41. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png European Journal of Cardio-Thoracic Surgery Oxford University Press

Lymph node dissection along the recurrent laryngeal nerves in patients with oesophageal cancer who had undergone chemoradiotherapy: is it safe?

Loading next page...
 
/lp/ou_press/lymph-node-dissection-along-the-recurrent-laryngeal-nerves-in-patients-CcmmPFy0kY
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.
ISSN
1010-7940
eISSN
1873-734X
D.O.I.
10.1093/ejcts/ezy127
Publisher site
See Article on Publisher Site

Abstract

Abstract OBJECTIVES Upper mediastinal lymph node dissection (LND)—especially along the recurrent laryngeal nerve (RN)—is the most challenging part of oesophageal cancer surgery. We investigated whether thoracoscopic RN LND may be safely performed in patients with oesophageal cancer who had undergone chemoradiotherapy (CRT). METHODS Patients with oesophageal cancer who had undergone thoracoscopic RN LND (n = 103) were divided into 2 groups according to whether they had prior treatment with CRT or not [the CRT group (n = 65) vs the upfront surgery group (n = 38), respectively]. All patients were operated on by a single surgeon. Intergroup comparisons were made in terms of (i) the number of dissected nodes, (ii) rates of RN palsy and (iii) rates of perioperative complications. The learning curve for the RN LND procedure was investigated using the cumulative sum method. RESULTS RN LND after CRT was more technically challenging when performed in the left side. Complete skeletonization of the left RN was achieved only in 66.2% of patients in the CRT group (vs 86.8% in the upfront surgery group; P = 0.022). The rate of postoperative left side RN palsy was significantly higher in the CRT group (26.6%) than in the upfront surgery group (7.9%, P = 0.022), albeit resulting in neither higher pneumonia rates nor longer hospital stays. The cumulative sum analysis revealed a steep learning curve for left RN LND in the CRT group. Unfortunately, an acceptable proficiency (left RN palsy rate: 15%) was not achievable even after treatment in 65 cases. CONCLUSIONS Thoracoscopic RN LND is safe but poses significant challenges in CRT-treated patients. Oesophageal cancer, Minimally invasive oesophagectomy, Recurrent laryngeal nerve, Lymph node dissection INTRODUCTION Lymph node metastases are common in patients with oesophageal cancer [1]. In particular, patients with oesophageal squamous cell carcinoma frequently presented with show metastatic spread to upper mediastinum nodes—especially along the bilateral recurrent laryngeal nerve (RN) [1–3]. The reported rates of lymph node metastases over the left- and right-side RN are as high as 15% and 30%, respectively [2, 3]. Based on the Japanese Nationwide Database, the RN nodal stations are those characterized by the highest efficacy index (calculated by multiplying the incidence of metastases to any given station with the 5-year survival rate of patients presented with metastases to that specific station) after removal [3]. However, RN lymph node dissection (LND)—particularly along the left side—is not yet considered as the standard of care mainly because of significant technical challenges that portend a significant risk of morbidity. Accordingly, high rates of RN palsy (unilaterally 10–60%; bilaterally 2–33%) have been reported in the literature [4–7]. The occurrence of RN palsy can in turn result in major postoperative complications (e.g. aspiration pneumonia) and even in postoperative mortality (especially when bilateral palsy occurs) [8, 9]. The implementation of RN LND in patients who had undergone chemoradiotherapy (CRT) is even more technically demanding. Accordingly, radiation therapy-induced fibrosis may obscure mediastinal microanatomy, ultimately making RN LND hazardous to the point that some surgeons consider prior radiation therapy as a relative contraindication to RN LND [10]. Theoretically, an extensive RN LND may be performed under the magnified view of a high-resolution thoracoscope using advanced endoscopes and instruments with multiple degrees of freedom. Unfortunately, published data on the safety of RN nodal dissection in patients with oesophageal cancer who had undergone CRT remain scarce. After the adoption of thoracoscopic oesophagectomy as of 2007, our surgical team has started including the bilateral RN area into its routine LND procedure in 2010. Initially, this approach was limited to patients who did not undergo CRT but—as of 2012—it was extended to all patients, regardless of previous CRT. In this retrospective study, we examined the feasibility and safety of thoracoscopic RN LND in patients with oesophageal cancer who had been previously treated with CRT. The study period spanned through 3 years, and all of the procedures were performed by a single surgeon. The study patients were divided into 2 groups according to previous treatment with CRT or not [the CRT group vs the upfront surgery (US) group, respectively]. Intergroup comparisons were made in terms of (i) the number of dissected nodes, (ii) rates of RN palsy and (iii) rates of perioperative complications. MATERIALS AND METHODS Patients The ethical approval for this retrospective analysis was granted by the local institutional review board. The study sample consisted of consecutive patients with oesophageal cancer who had undergone thoracoscopic oesophagectomy and RN LND between January 2013 and May 2016. All operations were performed by a single surgeon (Y.-K.C.). Patients were divided into 2 groups according to prior use of CRT or not (the CRT group vs the US group, respectively). Pretreatment staging was based on the results of chest and abdomen computed tomography (CT) scans, positron emission tomography (PET) imaging and endoscopic ultrasound. Patients were staged according to the American Joint Committee on Cancer (AJCC) staging criteria, 7th edition (2010). The severity of comorbidities was determined using the Charlson comorbidity index. The total score ranges from 0 to 37, with a score of 0 indicating the absence of comorbidities. Indications for chemoradiotherapy It is our institutional policy to offer CRT before surgery to all patients who are clinically staged as T2N0M0 or higher. CRT consisted of either a combination of cisplatin plus 5-fluorouracil or carboplatin plus paclitaxel administered with concurrent radiation therapy (dose 41.4–45 Gy). Patients underwent surgery at 6–8 weeks after completion of CRT. The standard surgical approach consisted of thoracoscopic oesophagectomy followed by the creation of a gastric tube (through either laparotomy or laparoscopy) and a cervical oesophagogastric anastomosis via left neck incision. A 2-field LND (which included the lower mediastinal/upper abdominal and bilateral RN lymph nodes) was performed in all patients [11]. We limited the use of neck LND only in patients who had an evidence of nodal involvement on preoperative CT or positron emission tomography imaging. Patients were in the intensive care unit for at least 1 night after surgery and subsequently transferred to their appropriate ward. Jejunostomy tube feeding was started 24 h after surgery. We performed a leak test on the 7th postoperative day. Thoracic procedure with the patient lying in the semiprone position The thoracoscopic procedure used in our centre has been previously described in detail [12]. In brief, the patient was placed in the left semiprone position while undergoing a single-lung ventilation (Fig. 1A). We created an artificial pneumothorax through CO2 insufflation (pressure 6–8 mmHg) and then positioned a total of 4 trocars (Fig. 1B). After incision of the mediastinal pleura, the layer between the thoracic duct and the oesophagus was dissected to ensure the maximum mobilization of the posterior part of the oesophagus. The mediastinal pleura was then incised along the course of the right vagus nerve until the lower margin of the subclavian artery with the goal of identifying the right RN. Left RN LND commenced with the exposure of the cartilage portion of the trachea. The oesophagus was then retracted towards the dorsal side either by surgeon’s left hand or a loop; meanwhile, the trachea was pushed aside by an assistant to optimize the operative exposure. The soft tissue between the left side trachea and the oesophagus (which included both lymph nodes and the left RN) was thoroughly dissected (Fig. 2). A careful skeletonization of the left RN was subsequently performed to achieve a complete lymphadenectomy. Figure 1: View largeDownload slide (A) Operating room set-up. (B) Port placement for thoracoscopic oesophagectomy when the patient is placed in the semiprone position. Figure 1: View largeDownload slide (A) Operating room set-up. (B) Port placement for thoracoscopic oesophagectomy when the patient is placed in the semiprone position. Figure 2: View largeDownload slide A representation of the reciprocal positioning of the oesophagus, trachea, left recurrent laryngeal nerve and its ancillary LNs (located ventrally and dorsally) when the patient is placed in the semiprone position. The green dashed line indicates the optimal dissection path for achieving maximum recurrent laryngeal nerve LN exposure. LN: lymph node. Figure 2: View largeDownload slide A representation of the reciprocal positioning of the oesophagus, trachea, left recurrent laryngeal nerve and its ancillary LNs (located ventrally and dorsally) when the patient is placed in the semiprone position. The green dashed line indicates the optimal dissection path for achieving maximum recurrent laryngeal nerve LN exposure. LN: lymph node. Definitions of outcomes The system proposed by the Esophagectomy Complications Consensus Group was used to define perioperative complications [13]. In patients with hoarseness or voice modifications that occurred after surgery, an otolaryngologist assessed vocal cord function using a flexible laryngoscope. Thirty-day mortality and 90-day mortality were defined as any death occurring within 30 or 90 days, respectively, after the date of surgery. Deaths occurring at any time during the postoperative hospital stay were used to calculate in-hospital mortality. Thirty-day readmission was defined as any in-patient admission to our institution within 30 days of discharge from the postoperative stay. Post-therapy surveillance Patients were scheduled for chest X-ray every 3 months and for CT every 6 months postoperatively. Survival was assessed every 6 months through contact with the patient’s physicians or a review of medical records. In case of missing information, data were retrieved from the National Cancer Registry Database of Taiwan. Statistical analysis Continuous data are expressed as both medians and interquartile ranges and means and standard deviations. Skewed variables were compared with the Mann–Whitney U-test, whereas normally distributed data were analysed with the Student’s t-test. Categorical data were compared with the χ2 test or the Fisher’s exact test (for small sample sizes). Cumulative sum (CUSUM) analysis was used to chart the occurrence of RN palsy, the major complication following an extensive RN LND [14–16]. The definition of acceptable and unacceptable failure rates is a key prerequisite to construct a CUSUM chart. Based on the published literature, the RN palsy rates after extensive RN LND are approximately 10–58.4% for the left side and 5–10% for the right side [4–6]. However, these data are referring to CRT naïve patients. We reasoned that such rates in CRT-treated patients would be either similar or slightly higher. Therefore, we set palsy rates of 15% and 50% (for the left side) and 5% and 10% (for the right side) as acceptable and unacceptable, respectively. The probabilities of Type I (α) and Type II (β) error were set at 0.05 and 0.20, respectively. A positive trend of the CUSUM curve indicates failure, whereas a negative trend is suggestive of successful procedures. When the line crossed the upper decision limit while ascending, the actual palsy rate was equal to the unacceptable value (with a Type I error probability of 0.05). When the line crossed the lower decision limit while descending, the actual palsy rate did not differ from the acceptable recurrence rate (with a Type II error probability of 0.20). All calculations were performed using SAS, version 9.3 (SAS Institute Inc., Cary, NC, USA) and SPSS, version 22.0 (IBM Corp., Armonk, NY, USA). Two-tailed P-values <0.05 were considered statistically significant. RESULTS Patient characteristics Between January 2013 and May 2016, we identified 103 patients with oesophageal cancer patients who had undergone thoracoscopic oesophagectomy and bilateral RN LND. Of them, 65 received CRT before surgery (the CRT group), whereas 38 did not (the US group). Table 1 summarizes the demographic and clinical characteristics of the study patients. Squamous cell carcinoma was the most common histological subtype. The 2 study groups did not differ significantly in terms of age, sex, histological subtype, and tumour location. An advanced clinical stage at diagnosis was more common for patients in the CRT group. Table 1: General characteristics of the study patients   Entire cohort   US (n = 38)  CRT (n = 65)  P-value  Age (years)   Mean ± SD  53.3 ± 7.3  54.6 ± 7.9  0.415   Median (IQR)  51.5 (48.8–57.5)  52 (49–60)  0.451  CCI  0.9 ± 1.1  0.8 ± 1.0  0.523  Sex      0.625   Male  36 (94.7)  63 (96.9)     Female  2 (5.3)  2 (3.1)    Histology   SCC  35 (92.1)  62 (95.4)     ADC  3 (7.9)  3 (4.6)    Clinical stage      <0.001   I/II  32 (84.2)  3 (4.6)     III  6 (15.8)  62 (95.4)    Tumour location      0.759   Upper third  8 (21.1)  12 (18.5)     Middle third  17 (44.7)  34 (52.3)     Lower third  13 (34.2)  19 (29.2)    Chemotherapy regimen      NA   5-FU + cisplatin  NA  35     Paclitaxel + carboplatin  NA  30      Entire cohort   US (n = 38)  CRT (n = 65)  P-value  Age (years)   Mean ± SD  53.3 ± 7.3  54.6 ± 7.9  0.415   Median (IQR)  51.5 (48.8–57.5)  52 (49–60)  0.451  CCI  0.9 ± 1.1  0.8 ± 1.0  0.523  Sex      0.625   Male  36 (94.7)  63 (96.9)     Female  2 (5.3)  2 (3.1)    Histology   SCC  35 (92.1)  62 (95.4)     ADC  3 (7.9)  3 (4.6)    Clinical stage      <0.001   I/II  32 (84.2)  3 (4.6)     III  6 (15.8)  62 (95.4)    Tumour location      0.759   Upper third  8 (21.1)  12 (18.5)     Middle third  17 (44.7)  34 (52.3)     Lower third  13 (34.2)  19 (29.2)    Chemotherapy regimen      NA   5-FU + cisplatin  NA  35     Paclitaxel + carboplatin  NA  30    Data are represented as mean ± standard deviation or counts (percentages), as appropriate. ADC: adenocarcinoma; CCI: Charlson comorbidity index; CRT: chemoradiotherapy; 5-FU: 5-fluorouracil; IQR: interquartile range; NA: not applicable; SCC: squamous cell carcinoma; SD: standard deviation; US: upfront surgery. Table 1: General characteristics of the study patients   Entire cohort   US (n = 38)  CRT (n = 65)  P-value  Age (years)   Mean ± SD  53.3 ± 7.3  54.6 ± 7.9  0.415   Median (IQR)  51.5 (48.8–57.5)  52 (49–60)  0.451  CCI  0.9 ± 1.1  0.8 ± 1.0  0.523  Sex      0.625   Male  36 (94.7)  63 (96.9)     Female  2 (5.3)  2 (3.1)    Histology   SCC  35 (92.1)  62 (95.4)     ADC  3 (7.9)  3 (4.6)    Clinical stage      <0.001   I/II  32 (84.2)  3 (4.6)     III  6 (15.8)  62 (95.4)    Tumour location      0.759   Upper third  8 (21.1)  12 (18.5)     Middle third  17 (44.7)  34 (52.3)     Lower third  13 (34.2)  19 (29.2)    Chemotherapy regimen      NA   5-FU + cisplatin  NA  35     Paclitaxel + carboplatin  NA  30      Entire cohort   US (n = 38)  CRT (n = 65)  P-value  Age (years)   Mean ± SD  53.3 ± 7.3  54.6 ± 7.9  0.415   Median (IQR)  51.5 (48.8–57.5)  52 (49–60)  0.451  CCI  0.9 ± 1.1  0.8 ± 1.0  0.523  Sex      0.625   Male  36 (94.7)  63 (96.9)     Female  2 (5.3)  2 (3.1)    Histology   SCC  35 (92.1)  62 (95.4)     ADC  3 (7.9)  3 (4.6)    Clinical stage      <0.001   I/II  32 (84.2)  3 (4.6)     III  6 (15.8)  62 (95.4)    Tumour location      0.759   Upper third  8 (21.1)  12 (18.5)     Middle third  17 (44.7)  34 (52.3)     Lower third  13 (34.2)  19 (29.2)    Chemotherapy regimen      NA   5-FU + cisplatin  NA  35     Paclitaxel + carboplatin  NA  30    Data are represented as mean ± standard deviation or counts (percentages), as appropriate. ADC: adenocarcinoma; CCI: Charlson comorbidity index; CRT: chemoradiotherapy; 5-FU: 5-fluorouracil; IQR: interquartile range; NA: not applicable; SCC: squamous cell carcinoma; SD: standard deviation; US: upfront surgery. Surgical quality and perioperative outcomes Table 2 shows the main data related to surgical quality. No conversion to open thoracotomy was required in either group. There were no significant inter-group differences in terms of thoracic operating time, intraoperative blood loss and the need for blood transfusions. As far as RN LND feasibility is concerned, the ability to achieve a complete LND (i.e. nerve skeletonization) and the number of dissected nodes were significantly lower over the left side in the CRT group; however, such differences were not evident for the right side. Complete skeletonization of the left RN was achieved only in 66.2% of patients in the CRT group, a percentage significantly lower than that obtained in the US group (86.8%; P = 0.022). Table 2: Surgical quality obtained in the study patients   Entire study cohort   US (n = 38)  CRT (n = 65)  P-value  Thoracic operating time (min)   Mean ± SD  166.1 ± 40.1  159.6 ± 50.0  0.500   Median (IQR)  165.5 (134–192.5)  155 (126.5–189)  0.430  EBL (mL)   Mean ± SD  110.0 ± 93.2  139.4 ± 184.3  0.362   Median (IQR)  50 (50–162.5)  100 (50–100)  0.582  Intraoperative BT  1 (2.6)  6 (9.2)  0.199  Successful right RN identification  35 (92.1)  54 (83.1)  0.197  Successful left RN identification  33 (86.8)  43 (66.2)  0.022  R0 surgery  38 (100)  57 (87.7)  0.045  Number of dissected lymph nodes   Total number  41.9 ± 13.1  29.1 ± 12.5  <0.001    Mediastinal nodes  20.8 ± 9.0  14.3 ± 6.1  <0.001    Right RN nodes  3.1 ± 2.5  2.9 ± 2.1  0.556    Left RN nodes  4.9 ± 3.1  3.6 ± 2.6  0.043    Abdominal nodes  19.8 ± 11.8  13.4 ± 7.1  0.001    Entire study cohort   US (n = 38)  CRT (n = 65)  P-value  Thoracic operating time (min)   Mean ± SD  166.1 ± 40.1  159.6 ± 50.0  0.500   Median (IQR)  165.5 (134–192.5)  155 (126.5–189)  0.430  EBL (mL)   Mean ± SD  110.0 ± 93.2  139.4 ± 184.3  0.362   Median (IQR)  50 (50–162.5)  100 (50–100)  0.582  Intraoperative BT  1 (2.6)  6 (9.2)  0.199  Successful right RN identification  35 (92.1)  54 (83.1)  0.197  Successful left RN identification  33 (86.8)  43 (66.2)  0.022  R0 surgery  38 (100)  57 (87.7)  0.045  Number of dissected lymph nodes   Total number  41.9 ± 13.1  29.1 ± 12.5  <0.001    Mediastinal nodes  20.8 ± 9.0  14.3 ± 6.1  <0.001    Right RN nodes  3.1 ± 2.5  2.9 ± 2.1  0.556    Left RN nodes  4.9 ± 3.1  3.6 ± 2.6  0.043    Abdominal nodes  19.8 ± 11.8  13.4 ± 7.1  0.001  Data are represented as mean ± standard deviation or counts (percentages), as appropriate. BT: blood transfusions; CRT: chemoradiotherapy; EBL: estimated blood loss; IQR: interquartile range; RN: recurrent laryngeal nerve; SD: standard deviation; US: upfront surgery. Table 2: Surgical quality obtained in the study patients   Entire study cohort   US (n = 38)  CRT (n = 65)  P-value  Thoracic operating time (min)   Mean ± SD  166.1 ± 40.1  159.6 ± 50.0  0.500   Median (IQR)  165.5 (134–192.5)  155 (126.5–189)  0.430  EBL (mL)   Mean ± SD  110.0 ± 93.2  139.4 ± 184.3  0.362   Median (IQR)  50 (50–162.5)  100 (50–100)  0.582  Intraoperative BT  1 (2.6)  6 (9.2)  0.199  Successful right RN identification  35 (92.1)  54 (83.1)  0.197  Successful left RN identification  33 (86.8)  43 (66.2)  0.022  R0 surgery  38 (100)  57 (87.7)  0.045  Number of dissected lymph nodes   Total number  41.9 ± 13.1  29.1 ± 12.5  <0.001    Mediastinal nodes  20.8 ± 9.0  14.3 ± 6.1  <0.001    Right RN nodes  3.1 ± 2.5  2.9 ± 2.1  0.556    Left RN nodes  4.9 ± 3.1  3.6 ± 2.6  0.043    Abdominal nodes  19.8 ± 11.8  13.4 ± 7.1  0.001    Entire study cohort   US (n = 38)  CRT (n = 65)  P-value  Thoracic operating time (min)   Mean ± SD  166.1 ± 40.1  159.6 ± 50.0  0.500   Median (IQR)  165.5 (134–192.5)  155 (126.5–189)  0.430  EBL (mL)   Mean ± SD  110.0 ± 93.2  139.4 ± 184.3  0.362   Median (IQR)  50 (50–162.5)  100 (50–100)  0.582  Intraoperative BT  1 (2.6)  6 (9.2)  0.199  Successful right RN identification  35 (92.1)  54 (83.1)  0.197  Successful left RN identification  33 (86.8)  43 (66.2)  0.022  R0 surgery  38 (100)  57 (87.7)  0.045  Number of dissected lymph nodes   Total number  41.9 ± 13.1  29.1 ± 12.5  <0.001    Mediastinal nodes  20.8 ± 9.0  14.3 ± 6.1  <0.001    Right RN nodes  3.1 ± 2.5  2.9 ± 2.1  0.556    Left RN nodes  4.9 ± 3.1  3.6 ± 2.6  0.043    Abdominal nodes  19.8 ± 11.8  13.4 ± 7.1  0.001  Data are represented as mean ± standard deviation or counts (percentages), as appropriate. BT: blood transfusions; CRT: chemoradiotherapy; EBL: estimated blood loss; IQR: interquartile range; RN: recurrent laryngeal nerve; SD: standard deviation; US: upfront surgery. Table 3 depicts the perioperative course observed in the 2 study groups. As far as morbidity after RN LND is concerned, a total of 23 patients experienced RN palsy. Of them, 20 and 3 cases had left-sided and right-sided lesions, respectively. No patients had bilateral RN palsy. The rate of postoperative RN palsy in the left side was significantly higher in the CRT group than in the US group (26.6% vs 7.9%, respectively, P = 0.022); however, no significant inter-group differences were observed in terms of right-side RN palsy rates. Despite a significantly higher left RN palsy rate in the CRT group, no significant differences were observed between the 2 groups with regard to pulmonary complications, anastomotic leaks, 30-day mortality/90-day mortality, and 30-day readmission rates. With regard to 90-day mortality in the CRT group, we registered a total of 3 cases. Of them, 2 died because of pulmonary complications while the patients were still hospitalized after surgery. Consequently, these 2 patients were also considered as cases of in-hospital mortality. The remaining patient died at 58 days after surgery because of brain haemorrhage caused by tumour metastases. Table 3: Short-term perioperative outcomes in the study patients   US (n = 38)  CRT (n = 65)  P-value  MV time (h)   Mean ± SD  10.1 ± 10.0  26.0 ± 101.1  0.337   Median (IQR)  5.6 (3.5–14.9)  7.2 (4.1–17.2)  0.340  ICU stay (h)   Mean ± SD  38.9 ± 46.2  57.9 ± 109.9  0.315   Median (IQR)  20.1 (17.6–43.9)  34.3 (18.5–45.4)  0.367  Pneumonia  3 (7.9)  11 (16.9)  0.197  Chylothorax  4 (10.5)  2 (3.1)  0.190  MV >72 h  0 (0)  3 (4.6)  0.295  RN palsy   Right sided  2 (5.3)  1 (1.6)  0.554   Left sided  3 (7.9)  17 (26.6)  0.022  Anastomotic leak  2 (5.3)  6 (9.2)  0.468  30-Day mortality  0 (0)  0 (0)  NA  In-hospital mortality  0 (0)  2 (3.1)  0.275  90-Day mortality  0 (0)  3 (4.6)  0.179  30-Day readmission  4 (10.5)  13/63 (20.6)  0.188    US (n = 38)  CRT (n = 65)  P-value  MV time (h)   Mean ± SD  10.1 ± 10.0  26.0 ± 101.1  0.337   Median (IQR)  5.6 (3.5–14.9)  7.2 (4.1–17.2)  0.340  ICU stay (h)   Mean ± SD  38.9 ± 46.2  57.9 ± 109.9  0.315   Median (IQR)  20.1 (17.6–43.9)  34.3 (18.5–45.4)  0.367  Pneumonia  3 (7.9)  11 (16.9)  0.197  Chylothorax  4 (10.5)  2 (3.1)  0.190  MV >72 h  0 (0)  3 (4.6)  0.295  RN palsy   Right sided  2 (5.3)  1 (1.6)  0.554   Left sided  3 (7.9)  17 (26.6)  0.022  Anastomotic leak  2 (5.3)  6 (9.2)  0.468  30-Day mortality  0 (0)  0 (0)  NA  In-hospital mortality  0 (0)  2 (3.1)  0.275  90-Day mortality  0 (0)  3 (4.6)  0.179  30-Day readmission  4 (10.5)  13/63 (20.6)  0.188  Data are represented as mean ± standard deviation or counts (percentages), as appropriate. CRT: chemoradiotherapy; ICU: intensive care unit; IQR: interquartile range; MV: mechanical ventilator; NA: not applicable; RN: recurrent laryngeal nerve; SD: standard deviation; US: upfront surgery. Table 3: Short-term perioperative outcomes in the study patients   US (n = 38)  CRT (n = 65)  P-value  MV time (h)   Mean ± SD  10.1 ± 10.0  26.0 ± 101.1  0.337   Median (IQR)  5.6 (3.5–14.9)  7.2 (4.1–17.2)  0.340  ICU stay (h)   Mean ± SD  38.9 ± 46.2  57.9 ± 109.9  0.315   Median (IQR)  20.1 (17.6–43.9)  34.3 (18.5–45.4)  0.367  Pneumonia  3 (7.9)  11 (16.9)  0.197  Chylothorax  4 (10.5)  2 (3.1)  0.190  MV >72 h  0 (0)  3 (4.6)  0.295  RN palsy   Right sided  2 (5.3)  1 (1.6)  0.554   Left sided  3 (7.9)  17 (26.6)  0.022  Anastomotic leak  2 (5.3)  6 (9.2)  0.468  30-Day mortality  0 (0)  0 (0)  NA  In-hospital mortality  0 (0)  2 (3.1)  0.275  90-Day mortality  0 (0)  3 (4.6)  0.179  30-Day readmission  4 (10.5)  13/63 (20.6)  0.188    US (n = 38)  CRT (n = 65)  P-value  MV time (h)   Mean ± SD  10.1 ± 10.0  26.0 ± 101.1  0.337   Median (IQR)  5.6 (3.5–14.9)  7.2 (4.1–17.2)  0.340  ICU stay (h)   Mean ± SD  38.9 ± 46.2  57.9 ± 109.9  0.315   Median (IQR)  20.1 (17.6–43.9)  34.3 (18.5–45.4)  0.367  Pneumonia  3 (7.9)  11 (16.9)  0.197  Chylothorax  4 (10.5)  2 (3.1)  0.190  MV >72 h  0 (0)  3 (4.6)  0.295  RN palsy   Right sided  2 (5.3)  1 (1.6)  0.554   Left sided  3 (7.9)  17 (26.6)  0.022  Anastomotic leak  2 (5.3)  6 (9.2)  0.468  30-Day mortality  0 (0)  0 (0)  NA  In-hospital mortality  0 (0)  2 (3.1)  0.275  90-Day mortality  0 (0)  3 (4.6)  0.179  30-Day readmission  4 (10.5)  13/63 (20.6)  0.188  Data are represented as mean ± standard deviation or counts (percentages), as appropriate. CRT: chemoradiotherapy; ICU: intensive care unit; IQR: interquartile range; MV: mechanical ventilator; NA: not applicable; RN: recurrent laryngeal nerve; SD: standard deviation; US: upfront surgery. As far as pathology results are concerned, 9 (13.8%) patients in the CRT group showed nodal involvement in the RLN area. The positive rates over the right and left RN were 10.8% and 4.6%, respectively. Notably, 1 patient had bilateral RN metastases. Seven (18.4%) patients in the US group had RLN metastases. The positive rates over the right and left RN were 15.8% and 7.9%, respectively. Of note, bilateral RN metastases were observed in 2 cases. Learning curve of recurrent laryngeal nerve lymph node dissection after chemoradiotherapy Because the rates of right RN palsy (in both groups) and left RN palsy (in the US group) were low, we specifically focused on the learning curves of left RN LND in the CRT group—which was characterized by the highest palsy rate. For this reason, only the learning curves for complete RN LND in the subgroup of 43 patients in which this procedure was successfully completed were presented. Although the CUSUM curve never breached the unacceptable threshold (50% palsy) throughout the study, we were also unable to observe a point at which the line descended to the boundary set as acceptable (i.e. 15% palsy; Fig. 3). Figure 3: View largeDownload slide The cumulative sum chart for the left recurrent laryngeal nerve (RN) lymph node dissection in patients who had undergone chemoradiotherapy. Only patients who received a complete left RN lymph node dissection (n = 43) were included. A positive trend of the cumulative sum curve indicates failure, whereas a negative trend is suggestive of successful procedures. An RN palsy rate of 15% was deemed as acceptable, whereas a 50% rate was regarded as unacceptable. When the learning curve reached the boundary for acceptability (while descending) or unacceptability (while ascending), the surgical proficiency of the preceding series was classified as acceptable or unacceptable, respectively. Figure 3: View largeDownload slide The cumulative sum chart for the left recurrent laryngeal nerve (RN) lymph node dissection in patients who had undergone chemoradiotherapy. Only patients who received a complete left RN lymph node dissection (n = 43) were included. A positive trend of the cumulative sum curve indicates failure, whereas a negative trend is suggestive of successful procedures. An RN palsy rate of 15% was deemed as acceptable, whereas a 50% rate was regarded as unacceptable. When the learning curve reached the boundary for acceptability (while descending) or unacceptability (while ascending), the surgical proficiency of the preceding series was classified as acceptable or unacceptable, respectively. DISCUSSION To our knowledge, this is the first study to specifically investigate the feasibility, safety and learning curve of thoracoscopic RN LND in patients with oesophageal cancer who had undergone CRT. Our data indicate that RN LND is feasible even after CRT, although the technical challenges to be faced are greater than in the CRT-naïve patients. Despite the RN palsy rate being 3.4-fold greater in the CRT group than in the US group (7.9% vs 26.6%, respectively), such higher occurrence did not result in increased rates of pneumonia and/or death. Such positive results are likely due to (i) prompt diagnosis and aggressive clinical management of RN palsy and (ii) the absence of bilateral palsy (0%). All our patients underwent a vocal cord assessment by an otolaryngologist even in the presence of small and/or subtle voice modifications. In cases showing unilateral palsy and a wide glottal gap, intracordal hyaluronic acid injections were performed as early as possible (in general within 3 months) to reduce the likelihood of subsequent choking and aspiration pneumonia [17, 18]. Despite such recent therapeutic advances, unilateral RN palsy remains a significant source of morbidity and impaired quality of life. More remarkably, major efforts should be made to avoid bilateral RN palsy—a devastating complication for which no effective treatment exists [19]. In an era during which preoperative CRT is becoming the standard of care, the prevention of nerve injury during surgery is of utmost importance. Complete left RN LND requires a thorough removal of soft tissues and lymph nodes located along the nerve course (i.e. from the aortic arch level to the inferior thyroid artery). Because of the limited exposure of the left paratracheal space, the dissection of nodes located ventrally to the left RN generally represents the most challenging part of the surgical procedure. Two main options have been proposed to increase surgical exposure, including (i) the suspension of the oesophagus with loops of umbilical tapes (with the goal of maximizing the exposure of the left RN) and (ii) the replacement of a double-lumen tube with a single-lumen tube (either with or without a blocker) coupled with CO2 insufflation (with the goal of reducing the tracheal rigidity) [20, 21]. Previous studies conducted in the CRT-naïve patients by experienced surgeons supported the utility of these manoeuvres to reduce both RN palsy and deaths [20, 21]. In accordance with previous reports, the implementation of these surgical options allowed us to obtain low rates of left RN nerve palsy (7.9%) and mortality (0%) in our US patients [22]. However, left RN LND remains generally difficult to be achieved in patients who had undergone CRT—even when repeatedly performed by the same surgeon. As shown in Table 2, CRT-induced mediastinal fibrosis hampered a successful RN identification in approximately one-third of patients who had been treated with this approach (when compared with 10% only in the US group). When the nerve is not clearly identifiable, contusions, excessive stretching and thermal injury that may occur during surgical manipulation can ultimately result in nervous damage. This offers a compelling explanation for the RN palsy rate being higher in the CRT group than in the US group. Another important observation from our study is that RN LND is characterized by a steep learning curve (as shown by the CUSUM analysis), which suggests that an adequate proficiency was not achieved even after treatment of 65 cases. There are at least 3 strategies that may be helpful to reduce the occurrence of RN palsy. First, left RN LND can be selectively applied only to patients at high risk of having left ypN-RN disease (while invariably maintaining a radical approach for right RN LND). It should be noted that the clinical utility of extensive LND after CRT remains highly controversial—especially for ycN-negative patients [23]. We have previously shown that the pre-CRT cN-RN status is the most important predictor for positive ypN-RN [24]. Notably, routine RN nodal dissection still yielded positive results [ypN-RN(+)] in 30.8% of patients with cN-RN(+) who were judged to be ycN-RN(−) following CRT. Conversely, RN LND resulted in a low (10%) positive LN discovery rate in the subgroup of patients who did not have radiological evidence of RN LN involvement prior to CRT [24]. We speculate that RN LND might be omitted in this specific subgroup, especially when potential surgical risks are expected to outweigh clinical benefits. Second, we recommend a careful monitoring of RN function during left RN LND. This approach may reduce the incidence of nerve injury and has been shown to be clinically useful in other surgical procedures (e.g. thyroid operation) [25, 26]. However, data in the field of oesophageal cancer surgery are controversial. Zhong et al. [27] have previously shown that RN monitoring during oesophagectomy results in (i) a higher number of harvested positive mediastinal lymph nodes and (ii) a lower incidence of postoperative RN paralysis and pneumonia (when compared with patients who did not undergo nerve monitoring). However, other reports failed to confirm such a protective effect [5]. Recently, promising results have been obtained by replacing classical intermittent monitoring with continuous intraoperative nerve monitoring. Continuous intraoperative nerve monitoring may allow detecting an impending RN injury in an earlier phase when compared with traditional intermittent monitoring. Consequently, a surgical manoeuvre that causes stretching or pressure on the RN may be promptly interrupted, resulting in a superior nerve function preservation [28]. Third, the use of EndoWrist instruments may be clinically useful during fine dissections, ultimately reducing the risk of RN palsy. There is ample evidence that robotic surgical systems incorporating 3-dimensional stereoscopic vision and arms with wrist-like joints can allow obtaining meticulous dissections in limited anatomical space. In addition, they may be helpful to reduce the steep learning curve typically associated with conventional laparoscopic surgery (e.g. as encountered for radical prostatectomy). We hypothesize that similar trends may exist for interventions performed in the narrow left paratracheal space, although further investigations are required to support this possibility. Limitations Some limitations of our study merit comment. First, our study design was retrospective, and the sample size was small, ultimately resulting in a reduced statistical power. Second, generalization from this study should be made with caution, as data were based on a single-centre experience obtained by a single surgeon. In this context, randomized longitudinal studies are warranted to confirm and expand our findings. CONCLUSION Thoracoscopic RN LND is safe but poses significant challenges in patients with oesophageal cancer who had undergone CRT. A significant learning curve exists for finalizing this procedure in a complication-free manner. Funding This work was supported by the Chang Gung Memorial Hospital, Taiwan [CORPG3G0771 and CORPG3G0781]. Conflict of interest: none declared. REFERENCES 1 Tachimori Y. Pattern of lymph node metastases of squamous cell esophageal cancer based on the anatomical lymphatic drainage system: efficacy of lymph node dissection according to tumor location. J Thorac Dis  2017; 9: S724. Google Scholar CrossRef Search ADS PubMed  2 Udagawa H, Ueno M, Shinohara H, Haruta S, Kaida S, Nakagawa M et al.   The importance of grouping of lymph node stations and rationale of three-field lymphoadenectomy for thoracic esophageal cancer. J Surg Oncol  2012; 106: 742– 7. Google Scholar CrossRef Search ADS PubMed  3 Tachimori Y, Ozawa S, Numasaki H, Matsubara H, Shinoda M, Toh Y et al.   Efficacy of lymph node dissection for each station based on esophageal tumor location. Esophagus  2016; 13: 1– 45. Google Scholar CrossRef Search ADS PubMed  4 Sato Y, Kosugi S-I, Aizawa N, Ishikawa T, Kano Y, Ichikawa H et al.   Risk factors and clinical outcomes of recurrent laryngeal nerve paralysis after esophagectomy for thoracic esophageal carcinoma. World J Surg  2016; 40: 129– 36. Google Scholar CrossRef Search ADS PubMed  5 Hikage M, Kamei T, Nakano T, Abe S, Katsura K, Taniyama Y et al.   Impact of routine recurrent laryngeal nerve monitoring in prone esophagectomy with mediastinal lymph node dissection. Surg Endosc  2017; 31: 2986– 96. Google Scholar CrossRef Search ADS PubMed  6 Taniyama Y, Miyata G, Kamei T, Nakano T, Abe S, Katsura K et al.   Complications following recurrent laryngeal nerve lymph node dissection in oesophageal cancer surgery. Interact CardioVasc Thorac Surg  2015; 20: 41– 6. Google Scholar CrossRef Search ADS PubMed  7 Ninomiya I, Okamoto K, Fujimura T, Fushida S, Osugi H, Ohta T. Oncologic outcomes of thoracoscopic esophagectomy with extended lymph node dissection: 10-year experience from a single center. World J Surg  2014; 38: 120– 30. Google Scholar CrossRef Search ADS PubMed  8 Gockel I, Kneist W, Keilmann A, Junginger T. Recurrent laryngeal nerve paralysis (RLNP) following esophagectomy for carcinoma. Eur J Surg Oncol  2005; 31: 277– 81. Google Scholar CrossRef Search ADS PubMed  9 Berry MF, Atkins BZ, Tong BC, Harpole DH, D’Amico TA, Onaitis MW. A comprehensive evaluation for aspiration after esophagectomy reduces the incidence of postoperative pneumonia. J Thorac Cardiovasc Surg  2010; 140: 1266– 71. Google Scholar CrossRef Search ADS PubMed  10 Osugi H, Narumiya K, Kudou K. Supracarinal dissection of the oesophagus and lymphadenectomy by MIE. J Thorac Dis  2017; 9: S741. Google Scholar CrossRef Search ADS PubMed  11 Wong B. More or less surgery for esophageal cancer: extent of lymphadenectomy in esophagectomy for squamous cell esophageal carcinoma How much is necessary. Dis Esophagus  1994; 7: 151– 5. Google Scholar CrossRef Search ADS   12 Chao Y-K, Hsieh M-J, Liu Y-H, Liu H-P. Lymph node evaluation in robot-assisted versus video-assisted thoracoscopic esophagectomy for esophageal squamous cell carcinoma: a propensity-matched analysis. World J Surg  2018; 42: 590– 8. Google Scholar CrossRef Search ADS PubMed  13 Low DE, Alderson D, Cecconello I, Chang AC, Darling GE, D'Journo XB et al.   International consensus on standardization of data collection for complications associated with esophagectomy: Esophagectomy Complications Consensus Group (ECCG). Ann Surg  2015; 262: 286– 94. Google Scholar CrossRef Search ADS PubMed  14 Lim T, Soraya A, Ding L, Morad Z. Assessing doctors’ competence: application of CUSUM technique in monitoring doctors’ performance. Int J Qual Health Care  2002; 14: 251– 8. Google Scholar CrossRef Search ADS PubMed  15 Noyez L. Control charts, Cusum techniques and funnel plots. A review of methods for monitoring performance in healthcare. Interact CardioVasc Thorac Surg  2009; 9: 494– 9. Google Scholar CrossRef Search ADS PubMed  16 Williams SM, Parry BR, Schlup M. Quality control: an application of the cusum. BMJ  1992; 304: 1359. Google Scholar CrossRef Search ADS PubMed  17 Fang T-J, Hsin L-J, Chung H-F, Chiang H-C, Li H-Y, Wong AM et al.   Office-based intracordal hyaluronate injections improve quality of life in thoracic-surgery-related unilateral vocal fold paralysis. Medicine  2015; 94: e1787. Google Scholar CrossRef Search ADS PubMed  18 Fang TJ, Pei YC, Li HY, Wong A, Chiang HC. Glottal gap as an early predictor for permanent laryngoplasty in unilateral vocal fold paralysis. Laryngoscope  2014; 124: 2125– 30. Google Scholar CrossRef Search ADS PubMed  19 Baba M, Aikou T, Natsugoe S, Kusano C, Shimada M, Nakano S et al.   Quality of life following esophagectomy with three-field lymphadenectomy for carcinoma, focusing on its relationship to vocal cord palsy. Dis Esophagus  1998; 11: 28– 34. Google Scholar CrossRef Search ADS PubMed  20 Xi Y, Ma Z, Shen Y, Wang H, Feng M, Tan L et al.   A novel method for lymphadenectomy along the left laryngeal recurrent nerve during thoracoscopic esophagectomy for esophageal carcinoma. J Thorac Dis  2016; 8: 24. Google Scholar PubMed  21 Zhang R, Liu S, Sun H, Liu X, Wang Z, Qin J et al.   The application of single-lumen endotracheal tube anaesthesia with artificial pneumothorax in thoracolaparoscopic oesophagectomy. Interact CardioVasc Thorac Surg  2014; 19: 308– 10. Google Scholar CrossRef Search ADS PubMed  22 Shen Y, Zhang Y, Tan L, Feng M, Wang H, Khan MA et al.   Extensive mediastinal lymphadenectomy during minimally invasive esophagectomy: optimal results from a single center. J Gastrointest Surg  2012; 16: 715– 21. Google Scholar CrossRef Search ADS PubMed  23 Shridhar R, Hoffe SE, Almhanna K, Weber JM, Chuong MD, Karl RC et al.   Lymph node harvest in esophageal cancer after neoadjuvant chemoradiotherapy. Ann Surg Oncol  2013; 20: 3038– 43. Google Scholar CrossRef Search ADS PubMed  24 Li Z-G, Zhang X-B, Wen Y-W, Liu Y-H, Chao Y-K. Incidence and predictors of unsuspected recurrent laryngeal nerve lymph node metastases after neoadjuvant chemoradiotherapy in patients with esophageal squamous cell carcinoma. World J Surg  2018; doi: 10.1007/s00268-018-4516-y [Epub ahead of print]. 25 Randolph GW, Kamani D. Intraoperative electrophysiologic monitoring of the recurrent laryngeal nerve during thyroid and parathyroid surgery: experience with 1, 381 nerves at risk. Laryngoscope  2017; 127: 280– 6. Google Scholar CrossRef Search ADS PubMed  26 Barczyński M, Konturek A, Pragacz K, Papier A, Stopa M, Nowak W. Intraoperative nerve monitoring can reduce prevalence of recurrent laryngeal nerve injury in thyroid reoperations: results of a retrospective cohort study. World J Surg  2014; 38: 599– 606. Google Scholar CrossRef Search ADS PubMed  27 Zhong D, Zhou Y, Li Y, Wang Y, Zhou W, Cheng Q et al.   Intraoperative recurrent laryngeal nerve monitoring: a useful method for patients with esophageal cancer. Dis Esophagus  2014; 27: 444– 51. Google Scholar CrossRef Search ADS PubMed  28 Tsang RK, Law S. Adaptation of continuous intraoperative vagus nerve stimulation for monitoring of recurrent laryngeal nerve during minimally invasive esophagectomy. World J Surg  2016; 40: 137– 41. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

European Journal of Cardio-Thoracic SurgeryOxford University Press

Published: Mar 28, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off