Assessment of Anastomotic Viability With Spectroscopic Real-time Oxygen Saturation Measurement in a Porcine Study

Objective Anastomotic leakage (AL) is a severe complication following intestinal procedures. Intra.Ox™ by ViOptix Inc (Newark, CA, USA) is a novel, FDA-approved spectroscopic device which enables real-time measurement of mixed tissue oxygen saturation (StO2). Using a porcine model, this study explores the correlation between StO2 measurements and AL formation as well as investigates the applicability of Intra.Ox™ in the clinical setting. Methods Eleven female swine were divided into 3 groups to explore AL formation in different ischemic conditions. Group 1: 100% mesenteric-vascular ligation, n = 3; Group 2: 50% ligation, n = 5; Group 3: No mesenteric ligation, n = 3. StO2 at the anastomotic line was measured before and after vessel ligation and anastomosis. Measurements were taken at 6 distinct locations along afferent and efferent loops. AL was evaluated on postoperative day 5 by re-laparotomy. Results AL rate was 100%, 60% and 0% in groups 1, 2 and 3, respectively. Post-anastomotic StO2 in group 1 (22.9 ± 18.5%) and 2 (39.2 ± 20.1%) were significantly lower than in group 3 (53.1 ± 8.3%, p<.0001). Post-anastomotic StO2 readings ≤40% indicated AL potential with 100% sensitivity,+ 80% specificity, positive predictive value of 85.7% and negative predictive value of 100%. Conclusion This study demonstrates the value of Intra.Ox™ in assessing local perfusion and indicate the association between low StO2 and AL by providing accurate, real-time, noninvasive tissue oxygenation measurements at anastomotic sites. Further studies are required to investigate the clinical application of this novel device in intestinal surgery. Graphical Abstract


Introduction
Anastomotic leak (AL) is considered one of the most severe postoperative complications following intestinal procedures. Despite appropriate techniques and advances in surgical technology, AL occurs in 6.4% to 11% of patients undergoing colorectal surgery 1 and is associated with a mortality risk of up to 22%. [2][3][4][5] While surgical dogma attributes the majority of AL to tissue hypoperfusion and poor surgical technique, its exact causes remain unknown. Today, AL continues to pose a significant postoperative challenge with prolonged hospitalization, increased health care costs, possible stoma application, reduced overall survival, as well as increased overall morbidity. Since tissue oxygenation is a key determinant in tissue healing, adequate perfusion is a wellrecognized prerequisite for complete healing of colorectal anastomosis. [6][7][8] The extent of available local tissue oxygen inversely correlates with the risk of intestinal AL formation and tissue ischemia at the anastomotic site. 9,10 Currently, intestinal microcirculation and bowel viability are assessed macroscopically by evaluating the color of the serosal surface of the intestine, presence of bowel peristalsis as well as pulsation and bleeding from mesenteric arteries. 10 However, this type of intraoperative evaluation is subjective and based on a surgeon's individual experience. Thus, accessibility to objective and quantitative intraoperative methods to assess bowel microcirculation is paramount. Measurement of intestinal tissue blood perfusion at the intended site of anastomosis intraoperatively has been evaluated via multiple methods including electromyography, 11 ultrasound Doppler system, 12,13 laser doppler flowmetry 14,15 and fluorescence imaging. [16][17][18][19] However, these technologies do not meet the clinical needs of reliability and objectivity, and are often invasive and inconvenient.
Recently, a novel near-infrared spectroscopy (NIRS) based tissue oximeter called Intra.OxÔ has been developed by ViOptix (ViOptix, Inc Newark, CA) which provides fast, real-time measurement of mixed tissue oxygen saturation (StO 2 ). This FDA approved device measures the ratio of oxygenated to deoxygenated hemoglobin within a target tissue and can evaluate renal and skin flap ischemia. 20,21 Intra.OxÔ is a wireless and battery-operated device, and allows for rapid, serial measurements throughout the operative procedure. A recent study has presented encouraging first data supporting the use of Intra.OxÔ for intraoperative tissue oxygenation measurements following large bowel resection and vessel ligation in a porcine model. 22 However, only 2 animals were included, and intestinal anastomoses were not conducted in that study. In a more comprehensive approach, we implemented a large-scale porcine model mimicking critical ischemia in the small bowel to investigate both the applicability and efficacy of Intra.-OxÔ in assessing intestinal tissue perfusion and exploring the association between low StO 2 and AL following intestinal anastomosis.

Animals and StO 2 Measurements
A total of 11 age-matched female domestic pigs (13)(14)(15) week old) weighing between 40 and 50 kg were purchased from S&S Farms (Ramona, CA) for this study. Approval of the local board on animal welfare was obtained from the local Institutional Animal Care and Use Committee (IACUC) at the University of California, Irvine. Swine were fed a balanced diet and had unlimited access to water. All swine received humane care throughout the study in compliance with the eighth edition of the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health. 23 Intra.-Ox TM was provided by ViOptix, Inc to measure intraoperative small bowel anastomotic tissue oximetry. Intra.Ox TM tissue oximeter was used to measure StO 2 via tissue light absorption and scattering at several different wavelengths of near-infrared light to compute StO 2 in real time. The tip of the device consists of an optical head containing light-emitting diodes that generate light and contact the target tissue. By generating different wavelengths of light in the red to near infrared, the tip can acquire data multiple times per second to compute oxygenation levels. Approximately 10 seconds of data were obtained during each measurement. The head position was adjusted during measurements to ensure uniform contact and pressure.

Experimental Design and Surgical Procedure
Food was restricted for 16h before operation and water was allowed until anesthesia. On the day of the procedure, a pre-anesthetic mixture consisting of xylazine (2 mg/kg) and ketamine (25 mg/kg) was injected intramuscularly. Following the onset of sedation, an endotracheal tube and an intravenous line were placed and secured. Animals were monitored for electrocardiogram, pulse oximetry, body temperature, heart rate, respiration rate, blood pressure and tidal volume. Survival surgery animals were given enrofloxacin (7.5 mg/kg) for 3 days after surgery. After placement into supine position, a midline laparotomy was performed. At 40 cm from the Treitz ligament, a 12 cm segment of jejunum was identified for enterotomy and anastomosis. 11 animals were randomly divided into 3 groups which were distinguished based on the percentage of applied degree of mesenteric vessel ligation: Group 1: 100% ligation group (n = 3): small bowel anastomosis with ligation of the entire peri-anastomotic mesentery and its vasculature; Group 2: 50% ligation group (n = 5): small bowel anastomosis with 50% intact small bowel mesentery; Group 3, CTL (n = 3): small bowel anastomosis with completely intact small bowel mesentery (Figure 1). At both afferent and efferent sites, three 2 cm intervals oximetry readings by Intra.Ox TM were taken along the antimesenteric edge of the serosa before ligation. After ligation, GIA TM stapler (Medtronic, Norwalk, CT) loads were used to transect the small intestine proximally and distally. A side-to-side intestinal anastomosis was created by the same GIA TM stapler. To eliminate potential pre-anastomotic ischemia caused by the GIA staplers as well as to allow a small reperfusion effect, we delayed the first tissue measurements for 20 minutes after completion of side-to-side anastomosis. Twenty minutes after completing the side-to-side anastomosis, StO 2 measurements were taken at both afferent and efferent sites at three 2 cm intervals (Figure 2(a)). We were able to measure well defined areas adjacent to the staple line, including serosal surfaces in less than 1 cm proximity to the staple line. After completion of all StO 2 measurements, the abdominal fascia and skin were closed with sutures. On postop day 5, repeat exploratory laparotomy was performed to evaluate AL formation and areas of tissue necrosis. Following this assessment, animals were euthanized by lethal injection of euthasol.

Statistical Analysis
The data are respectively presented as mean ± standard deviation. Statistical analysis was carried out in GraphPad Prism

Impact of Mesentery Ligation on Tissue Oxygenation
Prior to mesenteric ligation, baseline StO 2 was measured using Intra.Ox TM . Before ligation, the average oximetry readings between the 3 groups were similar, and no significant difference in StO 2 levels was detected at the anastomotic site (Figure 3). Side-to-side anastomosis was performed following different degrees of ischemia by ligating the mesenteric vessels. After anastomosis, StO 2 at the anastomotic site was significantly decreased in both the 50% ligation group and 100% ligation group compared to control group; moreover, there was a significant difference between the 50% ligation group and 100% ligation group (Figure 3, post-anastomotic StO 2 in group 1 (22.9 ± 18.5%) and 2 (39.2 ± 20.1%) were significantly lower than in group 3 (53.1 ± 8.3%) (p<.0001 for group 1 vs 3; p = .02 for group 1 vs 2; P-values were corrected using Tukey adjustment).

The Correlation Between StO 2 at Anastomotic Site and AL Formation
All animals survived until day 5 after mesentery ligation. On postop day 5, AL was assessed by exploratory laparotomy. The rate of AL was 0% (0/3) in CTL, 60% (3/5) in the 50% ligation group, and 100% (3/3) in the 100% ligation group (Figure 4 and 5). An average was calculated by using three StO 2 readings at the respective efferent or afferent sites to assess potential ischemia. With 40% StO 2 set as the cut-off value, there was a clear correlation between low StO 2 readings and AL formation (p = .015).

Determination of Cut-off Point to Detect an Association Between Low StO 2 and AL
As shown in Table 1, post-anastomotic StO 2 reading ≤ 40% indicate AL potential with 100% sensitivity, 80% specificity, a positive predictive value of 85.7% and a negative predictive value of 100%.

Discussion
This is the first study to establish a positive correlation between the extent of tissue ischemia as assessed by the novel Intra.OxÔ device and risk of intestinal AL formation. While the StO 2 baseline was in a stable range prior to mesenteric ligation, significant differences in StO 2 values were detected in all groups following anastomosis and application of mesenteric ligation. By creating different degrees of ischemia, we were able to determine a reliable cut-off value (StO 2 ≤ 40%) which indicates strong potential for AL formation. While previous studies have suggested that mixed tissue StO 2 measured by NIRS can be a reliable and sensitive parameter reflecting tissue perfusion status and tissue viability, 24 our study is the first to confirm the efficacy of Intra.OxÔ in providing sensitive and accurate mixed tissue StO 2 measurements at the anastomotic site.  Interestingly, even in control animals, StO 2 levels at the anastomotic site were significantly decreased following immediate placement of side-to side anastomosis (53.1 ± 8.3 vs 64.2 ± 5.5, P = .001). This observation could be attributed to potential tissue ischemia caused by GIA staple use during intestinal anastomosis. To eliminate potential pre-anastomotic ischemia caused by GIA staplers as well as to allow a small reperfusion effect, we delayed the first tissue measurements for 20 minutes after completion of side-to-side anastomosis.
Since the Intra.OxÔ device is equipped with a fine sensor tip, we were able to measure well-defined areas adjacent to the staple line, including serosal surfaces in less than 1 cm proximity to the staple line. These measurements were crucial for the assessment of anastomotic microcirculation.
While AL occurred in locations with low oxygenation (StO 2 ≤ 40%), not all locations with low oxygenation resulted in anastomotic failure, suggesting that AL formation is most likely a complex, multifactorial event. There is growing evidence that gut bacteria can affect the anastomotic healing process, and that specific intraluminal microorganisms are strongly associated with the development of colorectal AL. [25][26][27][28][29][30] Results published in previous studies, as well as the outcome of this study, indicate that local tissue oxygen saturation remains a key determinant in predicting AL formation. [31][32][33] In a previous study including over 400 patients, intraoperative oxygen saturation from the colonic wall was measured during colorectal procedures and the values were analyzed with respect to AL formation. 31 Beside demonstrating a clear correlation between low intraoperative colonic StO 2 values and AL formation, the authors were also able to present a cut-off value of StO 2 ≤ 90 to indicate an increased risk of AL formation. 31 Despite these important findings, the study included only 1 set of measurements following the division of the mesentery prior to bowel resection, which could explain the high and rather unspecific StO 2 cut-off value. In contrast, our study included multiple sets of StO 2 measurements following mesenteric ligation as well as post-anastomosis at 6 distinct points, which allowed us to further narrow the scope and establish a valid, more reliable cut-off value for AL formation of ≤ 40 StO 2 .
While our first data is encouraging, the results should be interpreted with caution with respect to the limitations of this study, including an overall relatively small sample size as well as the use of small intestinal rather than colon anastomosis. Thus, further in-vivo studies are required to closely investigate the applicability and accuracy of Intra. Ox TM to predict AL following colorectal anastomosis.

Conclusion
In conclusion, our study was able to demonstrate the value of Intra.OxÔ in assessing local perfusion and indicate the association between low StO 2 and AL by providing accurate, reliable, quantitative, real-time, noninvasive tissue oxygenation measurements at small intestine anastomotic sites. These findings validate the centrality of oxygen saturation as a key determinant of anastomotic healing and pave the road for further investigations in NIRS evaluation of bowel anastomoses.

Author Contributions
TK: performance of operative procedures, data acquisition and interpretation, drafting of the manuscript, final approval for publication SL: performance of operative procedures, data acquisition and interpretation, drafting of the manuscript, final approval for publication ES: preparation and performance of operative procedures, data acquisition and final approval for publication HC: performance of operative procedures, data acquisition and interpretation, final approval for publication DN: performance of operative procedures, revision of the manuscript, final approval for publication VK: data interpretation and analysis, drafting of the manuscript, final approval for publication JMD: performance of operative procedures, data interpretation, critical revision for important intellectual content of the manuscript, final approval for publication AP: study design and conceptualization, performance of operative procedures, supervision of the study, drafting and critical revision for important intellectual content of the manuscript.
All authors read and approved the final manuscript.

Declaration of conflicting interests
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Dr Alessio Pigazzi and Dr Mehraneh Dorna Jafari were paid consultants to ViOptix. Tanja Khosrawipour, Shiri Li, Earl Steward, Haris Chaudhry, Danielle Nguyen and Veria Khosrawipour do not have any disclosures.

Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This project was completed with funding from Vi-Optix. The company did not have any control over the editing of this article.

Ethics approval
To conduct this study, approval of the local board on animal welfare was obtained from the local Institutional Animal Care and Use Committee (IACUC) at the University of California, Irvine.

Availability of data and material
The data that support the findings of this study are available from the corresponding author upon reasonable request.