Surgical Quality: More Than Just Margins and Lymph Nodes

For the majority of gastrointestinal malignancies, surgical resection is the mainstay of treatment. Resection involves the complete removal of the neoplasm, with the intent of achieving negative microscopic margins. In cancers involving the pancreas, stomach, bile duct, and esophagus, data have shown that leaving residual tumor at the time of operations (R2 resection) yields no survival advantage when compared to patients not undergoing any resection at all. Such resections are best reserved for palliative intent in highly selected patients. Following complete resection with removal of all gross disease, there is a significant survival difference between negative and positive microscopic margins. Results from numerous studies have affirmed the negative effect of positive resection margins on oncologic outcome. This includes resections for pancreatic cancer, rectal cancer, gastric cancer, esophageal cancer, and hepatectomy for metastatic colorectal cancer.

In addition to resection margins, the adequacy of resection is often judged by the number of lymph nodes in the specimen. Data regarding the benefit of regional lymph node dissection are more difficult to obtain; however, removal of the surrounding lymph nodes does provide for adequate staging information. The extent of lymphadenectomy required is debated; however, for colon cancer it is estimated that a minimum of 12 lymph nodes need to be removed and examined.¹ The extent of lymphadenectomy for gastric cancer is more controversial.

A large cooperative group trial examining adjuvant therapy following resection for gastric cancer reported that nearly 50% of the patients underwent what would be considered an inadequate lymph node dissection. ² Although there is ongoing debate regarding standard vs. extended lymph node dissections, the effect of inadequate lymph node dissection on oncologic outcome is difficult to quantify. In select patients with colon cancer, the removal and examination of fewer than 10 lymph nodes can be considered an indication for postoperative adjuvant therapy.³ Determining the detrimental effect of lymph node dissection on disease recurrence is very difficult and requires further study.

In this issue of GCR, Attili and colleagues reviewed data regarding patients undergoing surgical resection with curative intent for pancreatic, esophageal, gastric, and colorectal cancers.⁴ They identified anastomotic leak, postoperative weight loss, and prolonged recovery as significant factors for development of recurrent disease. It is known that the morbidity associated with these operations can result in a delay or inability to deliver appropriate adjuvant therapy. In this study, however, the authors report that the negative effect of postoperative complications was independent of the use of postoperative adjuvant therapy.

The authors studied resections in a wide variety of gastrointestinal malignancies of substantially disparate biologic aggressiveness. Despite this, their analysis did not identify the primary tumor site as an independent risk factor, though this may be explained by the small sample size of patients with pancreas and esophageal cancers. In addition, anastomotic leak, postoperative weight loss, and delayed recovery were significant factors when stage of disease was controlled.

This study raises several points that need to be addressed. There is no clear description of how postoperative complications were defined. Was anastomotic leak based on clinical suspicion or radiologic confirmation? And how were these leaks graded? It is likely that a grade 1 pancreatic leak would have a different postoperative course than a grade 3 leak, as defined by the International Study Group for Pancreatic Fistula (ISGPF) classification.⁵ Similarly, what parameters were used to define delayed wound healing? In a study that analyzes postoperative morbidity on outcome, an established and objective method to determine morbidity should be employed.⁶ A more detailed description of the surgical approaches and inclusion in analysis of surgical variables would add weight to the study.

It is interesting to note that the average age of patients in this study was in the 30s. In contrast, most studies from the West that have analyzed the influence of postoperative morbidity on long-term outcome have consisted of older patients. Similarly, it is commendable that there was no difference in the number of patients that received adjuvant chemotherapy regardless of the presence or absence of complications. This is also different than in the West, where a good number of patients with postoperative complications do not recover sufficiently to receive chemotherapy.

Despite the limitations of a retrospective study, Attili and his colleagues have brought to light another measure of surgical quality that is often overlooked. It has long been known that postoperative complications have been associated with an increased risk of disease recurrence; the exact magnitude, though, is unknown.

A recent review of over 1,500 patients undergoing colorectal surgery identified a 5-year cancer-specific survival of 76% in patients without a postoperative anastomotic leak as opposed to 57% in patients with a postoperative anastomotic leak. This study would suggest a significant negative influence of postoperative complications. To put this in perspective, such a difference in cancer specific survival would rival any benefit of postoperative adjuvant therapy.⁷

A recent case control study from the Cleveland Clinic suggests that the consequences of anastomotic leak and intra-abdominal abscess are more significant following resections for rectal cancer than for colon cancer.⁸

The negative effect of postoperative complications on outcomes for other gastrointestinal sites has not been well studied. However, a recent trial from Belgium did demonstrate such a relationship in an analysis of complications following esophagectomy. ⁹ Similar relationships have been reported for hepatic resection for metastatic colorectal cancer.^10,11 A study by Mynster et al suggests that the risk of disease recurrence due to postoperative complications may be multifactorial. These authors reported that individually, perioperative blood transfusions and postoperative infections were not associated with an increase in cancer recurrence. However, when both factors were noted in the same patient, the combination was associated with a significant increase in cancer recurrence.¹²

The mechanism by which postoperative complications results in an increase in cancer recurrence is unknown. Whether it is immune suppression as a result of infectious complications, perioperative blood transfusions, or release of humoral factors, the process needs to be identified. Mynster suggested that the combination of perioperative blood transfusions and postoperative infection was significantly associated with a poor prognosis (as opposed to individual occurrence).¹² In addition, not all complications lead to poor outcomes. A recent study by Rizk et al¹³ suggested that surgical complications may reduce survival whereas postoperative medical complications do not.

Attili and his colleagues have reminded us of the importance of good surgical technique and minimizing postoperative surgical complications. Although not easy to quantify, it appears that the magnitude of postoperative anastomotic leak may have a significant negative influence on cancer specific survival for several gastrointestinal malignancies. Further studies are needed in this area, as there has been a greater interest in improving surgical quality. Understanding the pathophysiology of how postoperative complications may result in poor clinical outcome would go a long way toward improving our understanding of tumor biology.