HOUSTON – Surgical-site infections occurring in patients who underwent curative resection for localized colon cancer were associated with worse overall survival in a large retrospective study.
Among nearly 10,000 patients with nonmetastatic colon cancer who underwent surgery with curative intent, surgical-site infections (SSIs) were associated with both worse overall survival and a reduced likelihood of receiving adjuvant chemotherapy, reported Dr. Gala Markia Barden, a surgical resident at Baylor College of Medicine, Houston.
Both SSIs and failure to receive adjuvant chemotherapy are independently associated with worse overall survival, she said at the annual Society of Surgical Oncology Cancer Symposium.
“Future studies and practice guidelines should focus on target areas for improving these potentially preventable problems, including active surveillance for and early recognition of surgical-site infections, as well as vigilant follow-up to ensure treatment completion and to improve the transition between the surgical and medical oncology teams to mitigate losses to follow-up,” she said.
Tapping into the merged Veterans Affairs Surgical Quality Improvement Program and VA Central Cancer Registry (VASQIP-VA) databases, the authors identified 9,946 patients aged 18 years and older who underwent radical resection for colon cancer from 1999 through 2009. Patients with rectal cancers or early postoperative deaths (within 90 days of surgery) were excluded.
The investigators examined the relationships between SSIs and both 5-year overall survival and receipt of adjuvant chemotherapy, which has been documented to improve survival in patients with stage III colon cancer. Delivery of adjuvant chemotherapy in these patients is considered to be a measure of the quality of cancer care, Dr. Barden noted.
Of the 9,946 patients included in the study, 1,340 (13.5%) developed SSIs. These patients were slightly but significantly younger (P < .001), had worse functional status (P = .002), and had higher American Society of Anesthesiologists (ASA) physical status scores (P < .001).
In univariate analysis, the investigators found that, in the entire cohort, SSIs were associated with worse overall survival (OS); in multivariate analysis controlling for sex, nutrition, functional status, ASA score, and number of lymph nodes resected, they saw that SSI was associated with a hazard ratio (HR) for worse overall survival of 1.35 (P < .0001).
When they looked at the association of SSI and OS stratified by cancer stage, however, they found that it was significant only for stage III disease. Patients with stage III who developed an SSI had a median OS of 29 months, compared with 33 months for those with no site infections (P< .001).
Dr. Barden and her associates also found that 42% of patients with infections did not receive adjuvant chemotherapy, compared with 34% of patients without SSIs (P = .002).
To see whether the worse survival among patients with SSI was primarily driven by the failure to deliver chemotherapy, they created a model adjusted for cancer risk factors, which showed that patients with stage III disease who developed an SSI and did not undergo adjuvant chemotherapy had an HR of worse overall survival of 1.59 (P < .0001).
They then added into the model those patients with SSIs who did receive adjuvant therapy but, contrary to their expectations, saw that the HR was only slightly reduced (1.56) and remained significant (P < .0001). The model also confirmed that failure to deliver chemotherapy was associated with worse survival (HR 1.52, P <.0001)
Dr. Barden acknowledged that the study was limited by the retrospective design, predominantly male VA cohort, and the lack of information in the databases about why patients did not receive adjuvant therapy.
The study was internally supported. Dr. Barden reported having no conflicts of interest.