Outcomes of Transhiatal and Intercostal Pleural Drain After Ivor Lewis Esophagectomy: Comparative Analysis of Two Consecutive Patient Cohorts

Introduction

Intercostal pleural drainage is the standard of care after non-cardiac thoracic surgery. However, no conclusive data are available regarding the effect of continuous suction, the optimal drain size, and the threshold for removal. ¹ Small bore tubes appear to be safe and effective, and have also been shown to cause less pain after lung resection. ² Also, early removal can reduce postoperative chest drainage time without compromising patient safety. ³ However, despite this emerging evidence, intercostal drainage is still considered the gold standard even in the era of enhanced recovery for surgery protocols. ⁴ Pain at the exit site of the intercostal drain is a major source of patient's discomfort and may be inadequately controlled by thoracic epidural analgesia. ⁵ Use of parenteral nonopioid analgesics is preferable and is common practice in these patients to improve pain coverage, ⁶ but it may be associated with renal adverse effects depending on dosage and plasma concentration. ⁷ The aim of this study was to compare the short-term outcomes of transthoracic esophagectomy after transhiatal or conventional intercostal pleural drain.

Patients and Methods

We designed an observational, retrospective, single-center cohort study. Primary outcomes were postoperative complication rates and analgesia requirements. The Internal Review Board approved the study protocol and all patients gave informed consent. We reviewed the hospital charts and the prospective research database of consecutive patients who underwent a hybrid Ivor Lewis esophagectomy through laparoscopy and right thoracotomy for carcinoma of the esophagus or the esophagogastric junction between January 2014 and December 2016. Patients were treated with a conventional intercostal pleural drainage until May 2015, and with a transhiatal drain thereafter. A 19Fr intercostal drain (Blake; Ethicon, Somerville, NJ) connected to an underwater seal and continuous suction was used in the first patient cohort and a 15Fr transhiatal drain connected to a portable vacuum drainage system (J-Vac; Ethicon) was used in the second patient cohort. Patients in whom a double drain (intercostal and transhiatal) was used were excluded from the study. Routine antibiotic prophylaxis with Cefazolin was given, and a thoracic epidural catheter was inserted before operation. Pneumoperitoneum was induced with a Veress needle and a 5-port laparoscopic access was used to perform gastric mobilization, supramesocolic lymphadenectomy, and lower mediastinal dissection. A 4 cm large gastric tube was formed using a linear stapler (Endo-GIA; Medtronic, Minneapolis, MA), starting with a 45 mm cartridge above the third branch of the right gastric vessels. The upper part of the stomach was left undivided. In the most recent patient cohort, the transhiatal drain was advanced through the 5 mm subxiphoid port into the right pleural cavity.

The second step of the operation consisted of a right posterior-lateral thoracotomy in the fifth interspace. The anterior serratus muscle was routinely spared. The arch of azygos vein was divided, and the esophagus was mobilized en bloc with the mediastinal pleura and the periaortic lymphofatty tissue up to the diaphragm. The thoracic duct was routinely ligated above the diaphragm. Lymphadenectomy of the low paratracheal and tracheobronchial nodes was performed, and the vagal branches supplying the airways were preserved whenever possible. The gastroesophageal junction was retrieved in the chest along with the pre-formed gastric tube, and an end-to-side esophagogastric anastomosis was performed at the apex of the chest using a 25 mm circular stapler (Medtronic). The gastric tubulization was completed at this point using the endo-GIA stapler. In the first patient cohort, a 19Fr Blake drain was introduced through the eighth intercostal space in the mid-axillary line, while in the most recent patients, the active part of the 15Fr Blake drain was retrieved through the hiatus.^8,9 Postoperatively, the intercostal drain was connected to underwater seal and external aspiration (−10 cm water) and the transhiatal drain to the portable vacuum bag. The vacuum portable system was checked every 3 hours for the first 12–24 hours to ensure that the bellow was fully charged.

Preoperative and postoperative data, including serial albumin levels, were recorded in an electronic database. Postoperative complications were scored according to the Dindo-Clavien classification. ¹⁰ The efficiency of the drainage systems was assessed by recording the daily and the total volume of fluid drainage, the presence of subcutaneous emphysema on the chest wall and/or neck, the presence air leakage in the underwater bottle or the rapid loss of vacuum in the J-Vac, the presence of residual pneumothorax on a standard chest radiography or CT-scan, and the duration of drainage. A chest X-ray was performed immediately after operation and on day 1, 3, and 6. In eligible patients, the nasogastric tube was removed on postoperative day 3 and clear liquids per os were allowed according to the institutional fast-track protocol. The epidural catheter was removed on postoperative day 4. A gastrografin swallow study was performed on postoperative day 5 and patients were then allowed a soft diet. Paracetamol and ketorolac were given for postoperative analgesia, and the required dose of each drug until discharge was recorded. Patients were scheduled for a follow-up visit at 1 and 3 months after the operation.

Continuous data are presented as mean and standard deviation or median and interquartile range (IQR). Categorical variables are shown as frequencies and percentages. Two-sided Chi-square test, Wilcoxon signed-rank, or t-test for independent values were performed as appropriate. P-values were considered significant when <0.05. Bonferroni correction was used as appropriate. All analyses were carried out using R version 3.2.2 software. ¹¹

Results

During the study period, 165 patients underwent esophagectomy for cancer. Of these, 65 did not meet the inclusion criteria. The final patient sample consists of 100 patients who underwent a hybrid Ivor Lewis esophagectomy and in whom a single intercostal (n = 50) or transhiatal (n = 50) Blake drain was used.

As shown in Table 1, there were no statistically significant differences between the 2 patient cohorts with respect to demographics and clinicopathological data. No intraoperative complications occurred. No conversions from the portable vacuum system to underwater seal and suction occurred. The incidence of grade ≥3 complications was similar in both patient groups (Table 2). Overall, 4 patients had persistent pneumothorax and/or air leakage; of these, 2 patients in the transhiatal group and 1 in the intercostal group required percutaneous catheter drainage on postoperative day 2. Three patients required revisional surgery due to hemothorax, anastomotic leakage, and chylothorax, respectively. One patient in the transhiatal group had persistent gastric outlet obstruction, possibly related to kinking of the Blake drain at the hiatus: drain removal at the time of the gastrografin swallow study resulted in immediate relief of obstruction.

The median volume of fluid drained was 1580 mL (IQR = 880) in patients with a transhiatal drain and 1610 mL (IQR = 850) in those with an intercostal drain (P = .505). Figure 1 shows the daily amount of pleural fluid drained in both patient cohorts. Patients with transhiatal drain had a significantly lower median drainage output only on postoperative day 1 (P < .001). All patients required supplemental analgesic medications during the postoperative period. However, patients with the transhiatal drain required a significantly lower dose of ketorolac (P < .001). The length of hospital stay was similar in the two groups. No complications related to the method of pleural drain occurred up to 3 months after hospital discharge. There was only one hospital readmission in the transhiatal group due to severe nutritional impairment (Table 2).

OPEN IN VIEWER

FIG. 1.

Comparison of daily volume of pleural fluid in patients with transhiatal (white box) and intercostal drain (gray box). There was a statistically significant difference between the two groups only on postoperative day 1 (P < .001). No significant differences were noted from POD 2 to POD 8 (P = .823, P = .067, P = .471, P = .841, P = .473, P = .502, and P = .039, respectively; Bonferroni correction significance level = 0.00625). POD, postoperative day.

Discussion

This comparative study shows that the transhiatal pleural drain seems as safe and efficacious as the intercostal drain, but is associated with a significantly reduced use of postoperative ketorolac. More than 10 years ago, we started to replace the conventional 28–32Fr Argyle tube with a 24Fr Blake drain. Later on, we switched to a 19Fr Blake drain, still connected to underwater seal and suction. ⁹ More recently, we hypothesized that a 15Fr transhiatal drainage of the pleural cavity may reduce postoperative pain and enhance patient's recovery after surgery. ¹²

Even in the era of minimally invasive and fast-track surgery it is not unusual to see patients with significant postoperative pain, at rest and/or during cough, and persistent discomfort at the chest drain site lasting months after drain removal. It is plausible that use of large-bore, semirigid chest tubes greatly reduce the benefits of a minimally invasive surgical procedure by causing pain, restricting mobility, and increasing the risk of pneumonia and thromboembolic complications. In addition, currently used drainage systems connected to underwater seal are cumbersome and often inconvenient to both patients and hospital staff.

The use of a small-size Blake drain independent from underwater seal and connected to a small portable vacuum bag provides multiple benefits. The bag is easily carried by the patient and fits into a jacket pocket, allowing early ambulation and physiotherapy. In addition, nursing care is greatly facilitated, and no purse-string suture or Valsalva maneuver is required at the time of drain removal. Although the one-way flutter valve prevents intrapleural reflux of air or fluid, we recommend to check the efficiency of aspiration and the level of fluid drained at regular intervals, especially during the first 12–24 hours. Repeat charging of the bellow and frequent emptying of the J-Vac bag ensure early and optimal lung reexpansion.

The role of the chest drainage after esophagectomy has been discussed in our previous work. ¹² In this study, the total amount of pleural fluid drained was similar in the 2 patient groups, confirming that routine underwater seal drain may be safely avoided after transthoracic esophagectomy. The only difference in the fluid volume was found on postoperative day 1 (Fig. 1), probably indicating that care of the J-Vac bag is critical in the early postoperative period.

Interestingly, patients with transhiatal drain received roughly 50% less of the mean ketorolac dose, and the difference between the two groups was statistically significant (P < .001). This finding, although inconclusive based on this observational cohort study, is anyhow important because it indicates that the intercostal drain exit site may be a neglected and independent source of pain in patients with thoracotomy.

We limited our method of transhiatal pleural drainage to patients without significant coagulation disorders, intraoperative technical difficulties, pleural adhesions, and/or air leakage at the time of thoracotomy closure, ¹² and this may explain the low rate of residual pneumothorax in our series.

Among the strengths of this study are the fairly good sample size, the comparison with an historical cohort, and the consistency of the surgical procedures. However, treatment allocation was not randomized, and the study was not appropriately powered to assess statistical significance. No visual analogue scales were used in the assessment of postoperative pain, but patients were treated on demand with paracetamol and ketorolac. Also, the site of pain was not consistently recorded. Selection bias cannot be excluded and one can argue that the intercostal drain was connected to a water-seal bottle and the transhiatal drain to a portable vacuum system. In addition, being a single-center study, the results cannot be generalized to other hospitals and to other thoracic surgical procedures due to variability in nursing protocols and management.

Conclusions

A pleural drain connected to a portable vacuum drainage system placed through the upper abdominal port site after hybrid Ivor Lewis esophagectomy could safely replace the traditional intercostal drain in selected patients. It also enables early ambulation and quick recovery time by reducing chest pain and eliminating the need of cumbersome underwater seal. A randomized trial could be performed to compare intercostal versus transhiatal drainage in patients undergoing transthoracic esophagectomy to provide more evidence about safety and effectiveness.