By Michael A. Goldfarb, MD, FACS
I had been in the operating room as a medical student at New York University School of Medicine in 1963. I was in the “Old Bellevue.” But I never realized that there was room for an elephant in the operating room until a few months ago.
Since completing my general surgery training, I have performed many operations to remove various cancers from the neck down. Like all surgeons, I hated any postoperative infectious complication. I only recently anecdotally noticed that a patient, who survived a wound infection, did not seem to ultimately live as long as expected. But I did not yet see the elephant.
Then in late 2013, at a multidisciplinary breast conference, the case of a 50-year-old obese, diabetic, woman with a long history of smoking was presented. The patient had a 2-cm invasive breast cancer and insisted on bilateral mastectomy, a sentinel node biopsy and tissue expanders. We all recognized that her risk for infection was “high.” We argued that perhaps she should only have the mastectomies and sentinel lymph node biopsy as a first operation, and the expanders at a later time. Then I remembered my similar patient who did not live a long life after her wound infection, and I vaguely began to see the elephant in the operating room.
Was there a connection between wound infection and long-term survival after infection in breast cancer surgery? I was unaware of any link. I searched the Internet for, “cancer surgery and postoperative infection and survival,” limiting my search to more recent studies (2000 and after). After several searches I only found four relevant articles. Our librarian used those articles to look for other citations. Once I reviewed those articles—and there are not many—I clearly saw the elephant in the operating room.
Patient pain and suffering from any infectious complication after surgery, as well as the massive costs, have been well documented. But the awareness of the reduction in disease-free survival because of postoperative infectious complications has not diffused among surgeons. The evidence for blood transfusion reducing survival, along with colon cancer surgery, has diffused among surgeons. Quoted references below will support the argument that postoperative infection after breast, colon, stomach, liver, lung and head and neck cancer surgery decreases long-term survival.
One of the early articles I noted was “Postoperative Wound Complications and Systemic Recurrence in Breast Cancer,” by Murthy et al (Br J Cancer 2007;97:1211-1217). In this study, “1,065 patients diagnosed with operable primary invasive breast cancer between 1994 and 2001 were assessed for development of systemic recurrence according to whether or not a wound complication occurred after surgery, with a median follow-up of 54 months (range 15-119). There were 93 wound complications (9%). There was a statistically significant greater risk of developing systemic recurrence in patients with wound problems than those without (hazard ratio [HR], 2.87; 95% confidence interval [CI], 1.97-4.18; P<0.0001). This remained in a multivariate analysis after adjustment for case mix variables, including Nottingham Prognostic Index (NPI) and estrogen–progesterone receptor status.”
The article indicated that in the “good prognostic group,” the systemic recurrence-free survival at five years (95% CI) was 93.1% in patients with wound complications and 98.2% in patients with no wound complications. In the moderate prognostic group the systemic recurrence-free survival at five years was 82.2% versus 89.6%. In the poor prognostic group the systemic recurrence-free survival at five years was 61.3% without postoperative infection versus 46.3% with infection. Smokers at the time of diagnosis were more likely to develop metastatic disease than the nonsmokers after adjustment for other factors.
Fischer et al recently published the article “Risk Analysis and Stratification of Surgical Morbidity After Immediate Breast Reconstruction” (J Am Coll Surg 2013;217:780-787). In the article, the authors reported that the following conditions were associated with complications: autologous reconstruction (odds ratio [OR], 1.41; P<0.001), American Society of Anesthesiologists physical status >3 (OR, 1.25; P=0.004), class I obesity (OR, 1.38; P<0.001), class II obesity (OR, 1.91; P<0.001), class III obesity (OR, 1.70; P<0.001) and active smoking (OR, 1.46; P<0.001). The authors concluded that, “surgical complications after IBR [immediate breast reconstruction] are related to preoperatively identifiable factors that can be used to accurately risk stratify patients, which may assist with counseling, selection and perioperative decision making.” The patients were stratified according to risk for infection, with 7.14% of patients falling in the low-risk category and 27.02% in the very-high-risk category.
Turning to cancers of the colon and rectum, the next article I noted was, “Correlation Between Postoperative Infections and Long-term Survival After Colorectal Resection for Cancer,” by Nespoli et al (Tumori 2004;90:485-490). In this study, Dukes’ stage (P=0.048) and occurrence of postoperative infectious complications (P=0.011) were independently associated with outcome. The authors reported: “In patients with infectious complications, the survival rate was more than 50% lower, than in patients without infections. We excluded emergency and palliative operations (intestinal bypass), elderly subjects, previous cancer, immunosuppressive therapy, severe organ dysfunction, and Dukes’ stage D and A because all of these factors may affect the long-term survival.”
There is further support published in the Annals of Surgical Oncology (2007;14:2559-2566) by Law et al in the study “The Impact of Postoperative Complications on Long-term Outcomes Following Curative Resection for Colorectal Cancer.” In this series of patients, “With exclusion of patients who died within 30 days, the median follow-up of the surviving patients was 45.3 months. The 5-year overall survival was 64.9%, and the overall recurrence rate was 29.1%. The presence of postoperative complications was an independent factor associated with a worse overall survival (P=0.023; HR, 1.26; 95% CI, 1.03-1.52) and a higher overall recurrence rate (P=0.04, HR, 1.26; 95% CI, 1.01-1.57). The presence of postoperative complication not only affects the short-term results of resection of colorectal cancer, but the long-term oncologic outcomes are also adversely affected.”
In 2011, Mirnezami et al published a very convincing study in the Annals of Surgery: “Increased Local Recurrence and Reduced Survival from Colorectal Cancer Following Anastomotic Leak” (2011;253:890-899). The authors examined the long-term oncologic effects of anastomotic leakage (AL) after restorative surgery for colorectal cancer by performing a search of studies published between 1965 and 2009. They evaluated local recurrence, distant recurrence and cancer-specific survival. In all, the analysis included 21,902 patients from 21 studies, including 13 prospective nonrandomized studies, one prospective randomized study and seven retrospective studies. The authors reported the following results: “For rectal anastomoses, the odd ratios (OR) of developing a local recurrence when there was AL was 2.05 (95% CI, 1.51-2.8; P=0.0001). For studies describing both colon and rectal anastomoses, the OR of local recurrence when there was an AL was 2.9 (95% CI, 1.78-4.71; P<0.001). The OR of developing a distant recurrence after AL was 1.38 (95% CI, 0.96-1.99; P=0.083). Long term cancer specific mortality was significantly higher after AL with an OR of 1.75 (95% CI, 1.47-2.1; P=0.0001).” The authors concluded “AL has a negative prognostic impact on local recurrence after restorative resection of rectal cancer. A significant association between colorectal AL and reduced long-term cancer specific survival was also noted.” No association between AL and distant recurrence was found in this particular study.
Turning to another cancer, Tsujimoto et al published the following article in Annals of Surgical Oncology: “Impact of Postoperative Infection on Long-term Survival After Potentially Curative Resection for Gastric Cancer” (2009;16:311-318). They concluded: “Patients with complications due to postoperative infection had significantly more unfavorable outcomes compared with those patients without postoperative infection. Multivariate analysis demonstrated that age, preoperative comorbidity, blood transfusion, tumor depth, nodal involvement, and postoperative infection correlated with overall survival.”
In 2011, Okamura et al published the paper “Prognostic Significance of Postoperative Complications After Hepatectomy for Hepatocellular Carcinoma” [HCC] (J Surg Oncol 2011;104:814-821). In this study, the authors compared results of patients with complications (177) and those without complications (199) after hepatectomy for liver cancer. Clinical data, surgical outcome, and long-term survival of patients were analyzed retrospectively. The authors concluded that “postoperative complications (major or minor) are one of prognostic indicators in hepatectomy for HCC, and further stressed that “high quality surgical technique to minimize complications will improve the prognosis of HCC.”
The theme continues with the paper “Prognostic Significance of Postoperative Wound Infections After Total Laryngectomy” (Head Neck 2012;34:1023-1027) by Gonzalez-Marquez et al. The authors analyzed 129 patients with previously untreated laryngeal or hypopharyngeal squamous cell carcinoma who underwent a total laryngectomy with a minimum follow-up of 24 months. Fifty-seven patients (44%) developed a wound infection. The study showed that the only parameters significantly associated with a worse disease-specific survival in these cases were nodal extracapsular invasion (P<0.001) and surgical wound infection (P=0.02).
Further support of the role of infection on survival was detailed in “Influence of Postoperative Infectious Complications on Long-term Survival of Lung Cancer Patients: A Population-based Cohort Study” in the Journal of Thoracic Oncology (2013;8:554-561). In this article, Andalib found an overall 30-day postoperative complication rate of 58.2% in a group of 4,033 patients. They noted a major infectious complication (pneumonia, empyema, or mediastinitis) in 378 patients, and concluded that “postoperative complications, particularly of a major infectious type, are strong negative predictors of long-term survival in lung cancer patients.”
As I mentioned earlier, there are not many other articles that I have found that document evidence of postoperative infection decreasing survival in cancer surgery patients. There are a few articles that attempt to explain the mechanism. One is “Perioperative Host-Tumor Inflammatory Interactions: A Potential Trigger for Disease Recurrence Following a Curative Resection for Colorectal Cancer,” by Mike et al (Surgery Today 2008;38:579-584).
The authors of this paper “investigated whether host-tumor interactions are causal or consecutive clinical factors associated with surgical stress that influences the long-term survival after a curative resection of colorectal cancer. A Medline/PubMed search was conducted to identify the relevant articles investigating the factors related to surgical stress and their effects on the long-term survival after a curative resection of colorectal cancer. The intraoperative state is defined as a cytokine storm in which ongoing local cytokine production occurs at the site of the tumor, thus further enhancing the autocrine cytokine loop for angiogenic factor production. The postoperative state is defined as tissue regeneration in which surgery-related clinical events enhance the systemic induction of inflammatory cytokines, which in turn synergistically exaggerate the local activation of tumor growth factors.
Host-tumor interactions under surgical stress may act synergistically as potent tumor growth factors, and may thus influence long-term survival. Controlling surgical insults and/or regulating perioperative inflammatory responses may therefore lead to new therapeutic approaches for controlling disease recurrence.”
More granular explanations of potential mechanisms were elucidated in the following articles:
Surgical Decision Making and Informing Patients
The article “Development and Validation of a Novel Stratification Tool for Identifying Cancer Patients at Increased Risk of Surgical Site Infection,” by Anaya et al describes a prospective cohort study of 503 patients undergoing elective operations at a tertiary cancer center (Ann Surg 2012;255:134-139) to assess postoperative surgical site infections (SSIs). The authors used multivariate logistic regression analyses to identify predictors of SSI from which they created a scoring system called Risk of Surgical Site Infection in Cancer (RSSIC). SSI incidence after 30 days was 24%. Significant predictors of SSI included preoperative chemotherapy (OR, 1.94; 95% CI, 1.16-3.25), clean-contaminated wounds (OR, 2.1; 95% CI, 1.24-3.55), operative time more than two hours (OR, 1.75; 95% CI, 1.01-3.04) and equal to four hours (OR, 2.24; 95% CI, 1.22-4.1). The authors also determined predictors according to surgical site, including the groin (OR, 4.65; 95% CI, 1.69-12.83) and head/neck (OR, 0.12; 95% CI, 0.02-0.89). The authors concluded: “SSIs are common following cancer surgery. Preoperative chemotherapy, in addition to other common risk factors, was identified as a significant predictor for SSI in cancer patients. The RSSIC [scoring system] improves risk stratification of cancer patients and identifies those that may benefit from more aggressive or novel preventive strategies.”
This report from the Annals of Surgery noted a 30-day SSI rate of 24%. Even if the wound infection rate is 5%, the life span of those patients may be compromised. And survival is shortened more than any benefit from chemotherapy or radiation. With regard to simultaneous breast surgery and reconstruction, surgeons must inform their patients about the risk for infection and the resulting chance of decreased long-term survival. Delayed reconstruction is more reasonable in patients at high risk for postoperative infection. If the chance of a wound infection in a high-risk patient were 25%, which might decrease long-term survival by 25%, would you advise using expanders at the same time as mastectomy? What percent risk for infection would you accept and still insert expanders? Now do you see the elephant?
As far as colon surgery, surgeons should inform patients of the risk for infectious complications and the consequences. The first step is to get surgeons on board just as they have accepted the risk for transfusion in colon cancer surgery. If surgeons accept the notion that postoperative infection reduces survival, then certain algorithms in surgery should be modified. Sometimes a surgeon does a low-anterior resection and considers adding a loop ileostomy, because he or she is not “sure” about the anastomosis. If the survival of that patient is reduced because of a leak, the decision for a loop ileostomy should be easier.
Perhaps earlier recognition of infection will promote earlier treatment and higher long-term survival. In the future, modulators of the immune response will also help treat infectious processes. The evidence that postoperative infectious complications decrease survival in cancer surgery patients is compelling, and should resonate with all physicians caring for these patients. Will patients realize that if they have an infection, their life span will be shortened? Should that be part of informed consent? These are uncomfortable questions for us, but they will be addressed, especially because some of us will deal with these questions on a personal level. Hopefully this article will help alert surgeons that there is an elephant in the operating room.
Dr. Goldfarb is program director, Department of Surgery, Monmouth Medical Center, Long Branch, New Jersey. He is an advisory board member for General Surgery News.