Combined Aggressive Distal Gastrectomy and Double-Tract Reconstruction for Palliation of Incurable Locally Invasive Distal Gastric Cancer With Gastric Outlet Obstruction

Advanced cancer of the distal stomach invades directly to adjacent structures (e.g., pylorus, duodenum, pancreas) and occasionally metastasizes to distant organs. Thus, patients with distal gastric cancer typically experience unpleasant disease manifestations, such as gastric outlet obstruction (GOO), gastrointestinal hemorrhage, and a palpable mass. Because this malignant disease is frequently incurable at the time of diagnosis, many palliative treatments, including nonresection gastric bypass (NRGB) and endoscopic stent placement, have been developed to address these problems.^1–3 However, the procedures often fail to eliminate unpleasant symptoms even after reestablishing a route for food passage. Thus, even today, patients with GOO secondary to incurable, locally invasive (T4) gastric cancer still require a clinically effective method for palliation. Therefore, I designed a new surgical palliative procedure that combines aggressive distal gastrectomy (ADG) and double-tract (DT) reconstruction (ADGDTR). ADG serves to eliminate unpleasant symptoms caused by the tumor mass, whereas DT reconstruction reestablishes a food passage. My objective in this study was to evaluate the efficacy and safety of ADGDTR in patients with GOO secondary to T4 gastric cancer.

Patients and Methods

Patients

I retrospectively evaluated patients presenting to my clinic for palliative surgical treatment of GOO secondary to T4 distal gastric cancer between March 1996 and March 2011. All patients provided written informed consent to undergo ADGDTR. The local Institutional Review Board approved this study. Baseline patient characteristics were recorded from the clinical records. Preoperative cancer staging included physical examination, gastroduodenal endoscopy, endoscopic biopsy, and barium study. Barium examinations showed the degree of GOO and the extent of gastric cancer invasion (Fig. 1A). Abdominal ultrasound and whole-body computed tomography studies revealed the presence of metastatic disease and the degree of gastric cancer invasion to adjacent structures.

View Figure

• Download as Powerpoint
• Download Figure

Operative procedure

ADG included the tumor even in cases where part of the lesion could not be removed (Fig. 1C). I began the DT reconstruction by developing the jejunal limb 20 cm distal to the ligament of Treitz. After that, I performed an end-to-side gastrojejunostomy and then created a side-to-end jejunoduodenostomy 20 cm distal to the gastrojejunostomy. I then developed an end-to-side jejunojejunostomy 30 cm distal to the jejunoduodenostomy (Fig. 2A and 2B). Several weeks after surgery, the reconstructed gastrointestinal tract was examined by barium study (Fig. 2C).

View Figure

• Download as Powerpoint
• Download Figure

Results

During the study period, 7 patients underwent ADGDTR (5 male; mean age, 71 years [range, 60–83 years]). The clinical, surgical, and disease characteristics of the patients are shown in Table 1. The mean operation time was 207 minutes (range, 150–295 minutes), and the average blood loss was 290 g (range, 110–480 g). The mean length of hospitalization was 43 days (range, 28–73 days) and mean survival was 8.3 months (range, 2–22 months). There was 1 death due to sepsis, and no other complications occurred. All surviving patients tolerated a low-residue or regular diet postoperatively. The only patient who required emergent surgery (No. 2) was an 83-year-old man. He was hospitalized with GOO secondary to advanced gastric cancer with liver metastasis. He also had ischemic heart disease and severe anemia. He developed acute myocardial ischemia preoperatively and was moved to the cardiovascular department for medical treatment. After he became hemodynamically stable, he was transferred back to me for gastric cancer treatment. Soon thereafter, the patient experienced massive hematemesis and shock caused by gastric hemorrhage, and ADGDTR was performed emergently. He had an uneventful clinical course, tolerated oral intake satisfactorily, and was discharged on the 21st postoperative day.

Table 1 Baseline characteristics in patients with distal gastric cancer and gastric outlet obstruction

View Fullscreen

The patient who died before postoperative hospital discharge (No. 4) was an 80-year-old man who developed abdominal sepsis after surgery. Interoperative findings included necrotic tissue adherent to the entire mucosal surface of his gastric lesion and tumor invasion to both the pancreas and liver (lateral segment). A small volume of necrotic tissue was not completely resected macroscopically. After surgery, intraperitoneal abscess formation was detected at the site of the residual infected necrotic tissue, although this area was electrocauterized as much as possible during the procedure (Fig. 1C). Radiographic examinations, including computed tomography, showed neither fistula formation between the intraperitoneal abscess and gastrointestinal tract nor suture line leakage. He received a diagnosis of systemic sepsis and died 87 days after surgery in spite of our efforts.

Discussion

Surgical palliation for patients with GOO secondary to incurable and locally advanced malignant disease (e.g., gastric, pancreatic, and duodenal cancer) has been controversial among surgeons for many years. NRGB is a widely accepted palliative surgery and has some advantages, such as low cost, acceptable morbidity, and simple technique.^7,8 However, this procedure is not always effective, and not all patients recover the ability to tolerate oral intake after undergoing painful surgery. Thus, NRGB has not been recognized as a standard surgical palliation.^9–13

In patients with GOO secondary to pancreatic cancer, Lucas et al¹⁴ found that NRGB led to delayed gastric emptying in 95%, as evidenced by nausea and vomiting. Therefore, they evaluated combined simple distal gastrectomy (excluding the tumor) and gastrojejunostomy in 19 patients with unresectable pancreatic cancer. They reported that all patients tolerated solid food until death and concluded that their combination surgery was the most effective palliative procedure. This combination surgery eliminates the GOO and avoids postoperative delayed gastric emptying. However, their method does not involve surgical removal of the pancreatic lesion and does not address unpleasant tumor-related symptoms (e.g., palpable mass and abdominal pain).

In patients with gastric cancer, DG results in the surgical removal of the primary tumor and relieves unpleasant symptoms, including GOO, delayed gastric emptying, gastrointestinal hemorrhage, and the palpable mass. Many gastric surgeons insist that DG is more effective than NRGB for patients with gastric cancer.^15–20 Stupart et al²¹ stated that although NRGB offered worthwhile palliation, DG provided more meaningful palliation and longer survival compared with NRGB.²¹ Furthermore, several other surgeons concluded that DG is beneficial for patients with incurable disease even in the presence of peritoneal dissemination, provided the degree is slight.^22,23 Kim et al²⁴ recommended DG as a standard method for patients with locally advanced gastric cancer, regardless of the surgical curability, and Geoghegan et al²⁵ stated that DG is the policy in their surgery department whenever feasible.

However, Hartgrink et al²⁶ suggested that the efficacy of DG is conditional. In their evaluation of prognostic factors, including age, liver metastasis, peritoneal dissemination, distant lymph node metastasis, and combinations of these factors, they found that patients with 1 metastatic site only benefitted from DG if they were 70 years or younger.²⁶ In the present series, my cohort included 3 patients who were 70 years or younger, 2 patients in their seventies, and 2 octogenarians. The oldest patient was age 83 years. All patients had T4 lesions, and 5 had 2 or more incurability factors. I found that ADGDTR was clinically beneficial even in patients who did not meet the criteria proposed by Hartgrink et al.

Surgeons need an accurate preoperative assessment of tumor resectability to perform gastric cancer surgery successfully. However, advanced gastric lesions incite a local inflammatory or fibrous reaction, making it difficult to characterize mass resectability preoperatively.^27–29 These reactions are easily misidentified as tumor invasion, and the clinical stage of gastric cancer is often overestimated compared with the histologic stage.^24,30 Some surgeons, relying on the results of the preoperative examination, readily consider an advanced gastric tumor to be unresectable and offer NRGB alone. Thus, patients lose the opportunity to undergo DG and may lose the chance to obtain effective palliation. Hallissey et al³¹ insisted that only an experienced surgeon can accurately assess tumor resectability at the time of laparotomy and determine the best surgical treatment. Further, they proposed that, even in patients with locally invasive or metastatic gastric cancer, DG should be considered the best choice whenever resection is feasible.³¹ In this series, I chose the procedure at the time of laparotomy.

Some surgeons believe that ADG is a high-risk surgery and inappropriate for palliation, and, indeed, 1 patient in my cohort died of abdominal sepsis. However, the intraperitoneal abscess that caused this mortality could have been prevented by complete removal of the necrotic mucosa. Thus, ADG is not intrinsically a high-risk surgery. The most severe, refractory, and fatal complication of gastric surgery is abdominal sepsis, often caused by suture line leakage. Especially in patients with locally advanced T4 distal gastric cancer, the duodenal wall is often edematous or friable after resection. This condition, known as the difficult duodenal stump, is the most significant risk factor for suture line leakage. Moreover, a difficult duodenal stump increases the risk of tumor invasion. Invasive cancer may also be associated with suture line leakage, although this has not been proven.³²

The risk of difficult duodenal stump suture line leakage may depend on the type of gastrointestinal reconstruction after ADG. In patients who undergo Billroth I reconstruction, the difficult duodenal stump is not closed and is anastomosed with the stomach remnant. As a result, the duodenal contents are drained into the stomach through the gastroduodenostomy, and the internal duodenal pressure remains low, as does the risk of gastroduodenostomy suture line leakage. In Billroth II and Roux-en-Y reconstructions, the difficult duodenal stump is closed, forming a blind duodenal stump. Consequently, the duodenum does not drain easily, and internal duodenal pressure is high. Therefore, the risk of difficult duodenal stump suture line leakage is higher after these procedures than after Billroth I reconstruction.

Barnett³³ and Barnett and Tucker³⁴ conducted laboratory and clinical studies of surgical blind duodenal stump management. They developed a drainage anastomosis between the duodenum and jejunal limb of the Roux-en-Y segment to decrease the risk of difficult duodenal stump suture line leakage, and obtained good results. They concluded that the duodenal drainage procedure helps prevent difficult duodenal stump suture line leakage. A blind duodenal stump is not formed in patients who undergo DT reconstruction. Instead, a side-to-end anastomosis between the jejunal limb and duodenal stump is performed, and the duodenum is drained through the resulting jejunoduodenostomy, allowing internal duodenal pressure to remain low. Consequently, the risk of jejunoduodenostomy suture line leakage is negligible. Both DT and Billroth I procedures are superior to Billroth II and Roux-en-Y reconstructions in that the former minimize the risk of difficult duodenal stump suture line leakage.

All patients who undergo palliative surgery wish to maintain oral intake for as long as possible. However, sooner or later after surgery, tumor invasion will recur at the site of retained cancer tissue. As this tissue gradually grows and invades into the wall of the adjacent upper gastrointestinal tract, it will cause malignant restenosis that clinically manifests as GOO. Some gastrointestinal tract reconstruction methods can delay the onset of malignant restenosis. After the Billroth I reconstruction, food passes through a single route around the site of the primary lesion. Because this site is at high risk for cancer recurrence, malignant gastrointestinal tract restenosis is likely to occur early in patients who undergo Billroth I reconstruction. On the other hand, after Billroth II or Roux-en-Y reconstruction, the route through which food passes is not near the primary cancer site, and early malignant restenosis is unlikely.

After DT reconstruction, the jejunoduodenostomy and jejunojejunostomy form 2 routes for food passage, and neither is near the site at risk for cancer recurrence. Thus, DT reconstruction also delays the onset of malignant restenosis of the gastrointestinal tract. DT, Billroth II, and Roux-en-Y reconstructions are superior to Billroth I reconstruction in that they lower the risk of gastrointestinal tract malignant restenosis.

DT reconstruction has other benefits for patients in addition to surgical palliation of GOO. Noh et al³⁵ proposed a new procedure similar to DT reconstruction that consists of a gastrojejunostomy, jejunoduodenostomy, and jejunojejunostomy, and stated that the incidence of postoperative delayed gastric emptying was lower after their reconstruction than following Roux-en-Y reconstruction. Namikawa et al³⁶ compared the clinical outcomes of DT, Roux-en-Y, and Billroth II reconstructions, and found that only DT reconstruction decreased the severity of reflux esophagitis and remnant gastritis.

In conclusion, ADGDTR is an effective palliative treatment with low surgical and malignant restenosis risks in patients with GOO secondary to incurable, locally invasive distal gastric cancer.