Association between enrollment in an enhanced recovery program for colorectal cancer surgery and long‐term recurrence and survival

Enhanced Recovery After Surgery (ERAS) programs have been shown to minimize the surgical inflammatory response in colorectal cancer. Our objective was to determine the association between an ERAS program for colorectal cancer surgery and oncologic recurrence and survival.


| INTRODUCTION
Cancer recurrence increases patient morbidity and represents a significant economic burden for health care systems. 1 Although significant advancements have been made in surgical care and associated therapy, recurrence rates are still high depending on the quality of care provided during cancer surgery. 2 Literature has shown that multidisciplinary perioperative care models facilitate early recovery and hasten the return to indicated medical oncologic therapy, 3 which may reduce cancer recurrence following surgery and improve long-term survival. 4 Nevertheless, there is a lack of knowledge regarding the association between specific perioperative interventions (e.g., anesthetic type, pain regimen, nutrition, rehabilitation) and long-term oncologic outcomes. 5 Enhanced Recovery After Surgery (ERAS) protocols involve the bundled application of evidence-based perioperative care interventions that are primarily aimed at hastening patient recovery and reducing the surgical stress response. 6 In recent years, the ERAS ® Society has compiled a set of guidelines across multiple surgical specialties, 7 all of which support the safety and efficacy of this multidisciplinary model as a standard of care to achieve better patient satisfaction, superior pain control, and lower postoperative morbidity. 8 The current ERAS guideline for elective colorectal surgery includes a total of 24 interventions that have been shown to improve postoperative clinical outcomes. 7 As new evidence accumulates in favor of ERAS, more research is now targeting longer term outcomes, such as survival in oncologic patients. 9 Prior work involving patients at high risk for postoperative complications or advanced cancer stage has suggested that they benefit from care within the context of an ERAS program. [10][11][12] Investigators have hypothesized that interventions associated with ERAS mediate this benefit by mitigating inflammation and surgical insult or through direct prevention of hospital-acquired conditions. 9,13 Unfortunately, the evidence regarding management associated with an ERAS program and cancer recurrence is still inconclusive, with a handful of studies yielding inconsistent results about the impact on survival. 11,[14][15][16] Our group aimed to determine the association between an ERAS program for colorectal surgery and long-term cancer recurrence and survival.

| Design
This is a retrospective, cohort study conducted in a single institution, community-based academic hospital with the same group of anesthesiologists and surgeons during the study period. The protocol for this study was approved by our institutional review board and all data were deidentified to protect confidentiality. Informed consent was waived due to the retrospective nature of this study. Two groups of patients were identified based on the period of ERAS initiation in our institution (March 2013

| Inclusion and exclusion criteria
This study included adult patients undergoing elective colorectal surgery. Emergent procedures were excluded, along with any patient who was unable or unwilling to participate in the ERAS program. In this cohort, none of the patients denied receiving perioperative care under our ERAS program.

| Outcomes
Our primary outcome was overall 5-year survival, which was defined as the living or deceased status after surgical treatment. Time to death was also recorded for the purpose of the analysis. This was assessed through hospital and primary care medical records. It was classified into three categories, including postoperative causes, defined as death secondary to postoperative complications occurring in the first 30 days after surgery; oncologic causes, defined as secondary to tumor progression despite planned curative surgical treatment; and other causes, defined as those not due to tumor progression or postoperative, not related to disease (i.e., accident or other illness). Two researchers reviewed all medical records to ensure adequate follow-up.
Oncologic recurrence was defined as the identification of a tumor mass consistent with primary cancer in any part of the body after surgery or any treatment modality with curative intent. This was typically performed through active surveillance based on the cancer stage (I-annual colonoscopy; II/III/IV-clinical review at least every 6 months during the first 3 years and annually until the fifth year with carcinoembryonic antigen, chest X-ray and abdominal ultrasound, and annual colonoscopy were performed). The time to the first recurrence was also recorded.

| Statistical analysis
An initial exploratory analysis was performed using descriptive statistics. The univariate analysis compared demographics and clinical variables between pre-ERAS and ERAS periods. Kaplan-Meier curves were plotted along with log-rank p values to identify potential differences in time to mortality events or recurrence events between both periods. Cox hazard regression analysis was performed at each year of follow-up to evaluate the association between ERAS program enrollment and overall survival, cancer-related mortality, and oncologic recurrence. Certain clinically relevant confounders (i.e., age, American Society of Anesthesiologists, cancer stage, and comorbidities) were included in the multivariable survival analysis. Hazard ratios (HRs) were reported along with their corresponding 95% confidence intervals (CIs). Kaplan-Meier curves were constructed.
We performed parametric survival analysis using Weibull distribution in those cases where the Cox assumptions were violated. Survival

| Patient characteristics
In total, 646 patients were included, of which 339 were pre-ERAS and 307 ERAS. Our overall median compliance rate with the ERAS protocol was 90% (interquartile range [IQR] 85%-95%). There was a greater proportion of female patients in the ERAS period (42% vs. 32%; p = 0.03). The remainder of baseline demographic and clinical characteristics were comparable between periods (Table 1)  Univariate analysis of patients who died, recurred, and were disease-free is presented in Table 2. Disease-free 5-year survival did not differ between groups (66% ERAS vs. 60% pre-ERAS; p = 0.14), but it was significantly higher in the ERAS group compared to conventional care among patients with advanced cancer stage (58% ERAS vs. 39% pre-ERAS; p < 0.01). While there was no difference in disease-free 5-year survival between groups among low cancer stage   Table 3.

| Subgroup analysis
Mediation analysis showed that a lower surgical duration (<2 h) contributed to 32.6% of the effect of ERAS on mortality at 2 years, while total intravenous anesthesia (propofol) and restrictive fluid therapy (<2 L) contributed 24.6% and 14.9%, respectively.

| Oncologic recurrence
Overall recurrence rates (17% ERAS vs. 21% pre-ERAS; p = 0.31) as   11,16 but not at 3 years of study follow-up. 15 Quiram et al. 14 found a statistically significant association between ERAS and overall survival, but no relationship was detected for disease-free survival.
As shown, our study demonstrated a significant improvement in both overall and disease-free survival rates among patients with advanced initial cancer stage.

| CONCLUSION
Although enrollment was not associated with a difference in survival at 5 years after surgery, patients who received perioperative care within an ERAS program with advanced colorectal cancer did have improved survival and a lower likelihood of oncologic recurrence compared to conventional care. These findings should be considered hypothesis-generating and large, prospective trials designed to assess for cancer recurrence and long-term survival are necessary to confirm these results.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.