|Home | About | Journals | Submit | Contact Us | Français|
Dehiscence or leakage following bowel anastomoses is associated with high morbidity and mortality. Perfusion and local tissue oxygenation (StO2), independent of systemic oxygen saturation, are fundamental determinates of anastomotic viability. As current technology is limited for monitoring local StO2 at bowel anastomoses, we aimed to construct a wireless pulse oximeter (WiPOX) to monitor real-time intraoperative tissue oxygenation, permitting identification of compromised anastomotic perfusion.
We have: (a) designed a handheld device capable of real-time monitoring of serosal and mucosal StO2 through endoscopic ports with wireless data transmission to standard intraoperative monitors, (b) constructed the WiPOX using materials meeting FDA regulations for intraoperative use and re-use, (c) performed accuracy testing in humans by comparing the WiPOX to standard pulse oximeters, and (d) tested WiPOX efficacy for detecting early tissue hypoxia in stomach, intestines, and kidneys in anesthetized rats and swine.
In humans, WiPOX demonstrated accuracy within 3% when compared to commercially available pulse oximeters. Application of the WiPOX in rats and swine demonstrated normal serosal and mucosal StO2 and pulse rates in healthy small bowel and stomach. Within 30 seconds of compromised perfusion, the WiPOX detected bowel hypoxia over a wide range of oxygen saturation (p<0.005). A greater degree of hypoxia was detected in mucosal versus serosal measurements during early ischemia, despite normal appearance of tissue. The onboard sensor-processor unit permitted non-invasive pulse oximetry and integration with current intraoperative monitoring. The contact pressure-sensing head allowed for consistent, high quality StO2 waveform readouts despite the presence of body fluids.
We have constructed, validated, and successfully tested a novel wireless pulse oximeter capable of detecting intraoperative tissue hypoxia in open or endoscopic surgery. This device will aid surgeons in detecting anastomotic vascular compromise and facilitate choosing an ideal site for bowel anastomosis by targeting well-perfused tissue with optimal healing capacity.
Anastomotic complications following gastrointestinal (GI) surgery are associated with substantial post-operative morbidity and mortality representing a significant health-care burden. Specifically, anastomotic leak is associated with decreased survival following GI surgery [1–2]. Anastomotic leak is identified following 0–50% of esophageal resections and reconstructions with an associated mortality of 35% [3–5]. Leakage from the pancreatico-jejunal anastomosis, which complicates up to 20% of pancreatic anastomotic operations, results in a 5–10% mortality rate [6–7]. Similar outcomes are noted with colonic and small bowel operations . Morbidity is significant even in survivors, who often require multiple reoperations, extended length of hospital stay, and decreased overall quality of life [2, 9–10]. In particular, anastomoses following rectal surgery are at high risk for postoperative complications, in which leakage has been associated with both decreased survival  and increased local cancer recurrence . Interventions to reduce anastomotic complications following GI surgery would have a significant impact on improving overall postoperative patient outcomes.
While the etiology of anastomotic breakdown is multifactorial, with contributing factors from both the patient and surgeon, the final common pathway frequently involves a combination of anastomotic tension, compromised tissue perfusion, and decreased oxygenation – factors shown to be key determinants of anastomotic viability [11–14]. These concepts were demonstrated in 1987 by Sheridan et al.  showing higher rates of colon anastomotic leakage in patients with decreased tissue oxygen tension. Subsequently, animal models have confirmed the importance of adequate blood flow and oxygenation for the creation of a sustainable anastomosis [11, 13]. Anegg et al.  demonstrated that decreased oxygen tension at the esophago-enteric anastomosis is associated with an increased leak rate in humans. Overall, these data suggest that intraoperative recognition of peri-anastomotic tissue oxygenation could help minimize post-operative anastomotic complications by guiding surgeons to perform maneuvers to increase tissue perfusion and/or fashion anastomoses at sites of optimal tissue healing.
Current standards of assessing viability at the anastomosis rely upon clinical determinants such as visual inspection of the resection margin and palpation of mesenteric arterial flow – notably subjective and crude parameters. Intraoperative doppler techniques have been shown to improve anastomotic failure rates by confirming adequate blood flow to the pancreatic side of the pancreaticojejunostomy during pancreatic resection . The utility of intraoperative doppler, however, is limited by arteriovenous shunting which causes non-nutrient flow, and lack of oxygen delivery data . Tissue tonometry, a frequently cited method for determining gastric mucosal oxygenation, in fact measures tissue pH and readouts are confounded by systemic processes that alter tissue CO2 such as sepsis or CO2 retention . Furthermore tonometry has demonstrated variable results and poor sensitivity in preclinical models of tissue ischemia . Pulse oximetry, routinely used intraoperatively to measure peripheral blood oxygen saturation (SpO2) uses optics to determine the changes in light absorption based on the volume of blood in the tissue, and is thus dependent upon the systemic blood circulation and does not necessarily reflect actual tissue oxygen tension. Furthermore, these devices are not technically amenable to placement on intra-abdominal or thoracic organs. Currently, pulse oximeters are not routinely available to measure local tissue oxygen saturation (StO2) at the anastomotic site. Table 1 elucidates the current oximeter technologies available for clinical use and their limitations in applying to intraoperative monitoring.
Herein we describe the results of a novel intraoperative wireless pulse oximeter (WiPOX) that enables the surgeon to monitor tissue oxygenation in real-time at the anastomotic site, thereby maximizing the ability to create a successful, sustainable anastomosis by targeting well-perfused tissue.
The wireless pulse oximeter (WiPOX) was designed and constructed through collaboration between the Division of Thoracic Surgery at Memorial Sloan-Kettering Cancer Center (MSKCC) and the Department of Bioengineering at the City College of New York (CCNY). The initial proof-of-principle prototype was constructed as a senior bioengineering project by a group of eight students with mentoring from senior authors (PSA, NR and MB). The device was fabricated using materials already approved for use in humans in the operating room by the Food and Drug Administration (FDA). Reflectance oximetry technology was utilized in order to facilitate application to intra-abdominal organs in contrast to traditional transmission oximetry as is used in finger and earlobe oximeters, which requires transmission through tissues – a technical limitation when attempting to assess solid organs and bowel (Fig. 1). Care was taken to ensure size parameters appropriate for use via laparoscopic and thoracoscopic 12mm ports.
Preliminary human accuracy testing assessed the WiPOX device in healthy human volunteers. The WiPOX was evaluated against two commercially available pulse oximeters as gold standards: the Nonin Onyx® 9500 (Nonin Medical, Inc., MN)) and the SPO 5500™ (SPO Medical, CA) finger oximeters. For each of ten trials, an average was taken for the two commercial oximeters. The percent differences between oxygen saturation for the WiPOX device and the average measurements from the commercial devices were recorded.
The WiPOX was next tested in vivo during animal operations using adult Sprague-Dawley rats (Charles River, MA) and adult female Yorkshire swine (Archer Farms Inc., MD). Procedures were approved and performed in accordance with the standards set forth by the Institutional Animal Care and Use Committee (IACUC). Rats were anesthetized using inhaled Isoflurane anesthesia via nose cone. Swine were anesthetized with Ketamine and Xylazine, endotracheally intubated, and maintained on mechanical ventilation by veterinarians from the Veterinary Services Department at MSKCC. Following adequate anesthesia, midline laparotomy was performed in swine and the stomach and small bowel were mobilized. StO2 measurements were taken using the WiPOX device from healthy stomach and small bowel and compared to systemic SpO2. To demonstrate WiPOX ability to detect StO2 variations in tissue hypoxia, experiments were performed in rats. The stomach was mobilized and an anterior gastrostomy created by raising a flap of stomach wall, which remained attached by a narrow tissue connection to the lesser curvature, which we ensured had normal initial StO2 (95–100%); we then induced ischemia by placing a clamp across the flap base (Fig. 2a,b). StO2 was measured using the WiPOX device on both the serosal and mucosal surfaces of the healthy vascularized stomach and compared to the compromised tissue flap (Fig. 2c,d). To investigate whether WiPOX could be efficacious in detecting renal hypoxia, as in the case of warm ischemia during partial nephrectomy, the right kidney was mobilized and StO2 measured from the kidney surface using the WiPOX before and after clamping of the vascular hilum. For both stomach and kidney experiments, measurements were performed in triplicate in each of three rats with means and standard errors reported. Animals were euthanized while under general anesthesia following all procedures.
Clinical prototype devices were constructed by the collaborative CCNY/MSKCC group to measure intraoperative StO2. The design concept employs multiple photodetectors arranged around a single infrared/red LED. The probe contains the reflectance sensor housed in a cylindrical compartment designed to optimize functionality for both open and minimally-invasive use and ergonomics. The sensor acquires StO2 readings by direct contact with the tissue surface. The contact pressure-sensing head allowed for consistent, high quality SpO2 waveform readouts even in the presence of body fluids and blood. To prevent application pressure from compromising mucosal StO2 measurements, a new generation device employs a spring loaded sensor to allow for optimal pressure application. Local oxygen saturation levels are determined by applying the device to any point on the gastrointestinal tissue surface. The onboard micro-controller and LED array permit noninvasive pulse oximetry using modified signal processing algorithms for real-time intraoperative application. Additionally, wireless technology was employed to create a wireless communication between the hand-held device and the intraoperative monitors. The WiPOX device specifications meet the standards set forth by the FDA for device implementation in humans (see Table 2).
Accuracy testing comparing WiPOX measurements in healthy human volunteers to standard commercially available pulse oximeters was obtained in ten independent trials. The final average relative accuracy compared to standard pulse oximeters across all trials was 97%. Due to the inherent +/−2% deviation for each commercial oximeter, the total relative accuracy was adjusted to be 95% (Fig. 3).
The WiPOX efficacy was evaluated intraoperatively during bowel surgery in rats and swine. Application of the WiPOX in swine demonstrated normal StO2 and pulse rates in healthy small bowel and stomach in comparison to systemic SpO2 values. Within 30 seconds of mesenteric vessel occlusion, the WiPOX device detected bowel hypoxia in the devascularized segments despite normal appearance of tissue on gross inspection (data not shown).
In rats, hypoxia was detected in the devascularized stomach compared to healthy stomach by applying the WiPOX device to the stomach serosa (74 vs 96%, p<0.005); this difference was magnified when mucosal measurements were obtained (52 vs 88%, p<0.005) (Fig. 2 and and4).4). Notably, the WiPOX detected serosal and mucosal hypoxia prior to visible signs detectable by the surgeons. These observations were reproduced in repeat experiments. In addition, following clamping of the hilum, kidney hypoxia was detected by the WiPOX as compared to normal kidney (p<0.005), which similarly occurred prior to overt signs of ischemia (Fig. 5).
Anastomotic leakage continues to occur with a high frequency following gastrointestinal surgery and portends a poor overall patient outcome. Patients experiencing anastomotic breakdown suffer prolonged hospitalization, decreased quality of life, increased mortality, and consequently increased health-care costs. Attempts to reduce the rate of post-operative anastomotic leakage following esophagogastrectomy have been largely unsuccessful as illustrated by the essentially unchanged leak rates following esophagogastrectomy over the past two decades. However, evidence suggests that peri-anastomotic oxygenation contributes to anastomotic healing [11, 15–16, 22]; thus, there is compelling evidence that ensuring maximal tissue oxygenation will likely improve anastomotic viability. The WiPOX device represents a novel technology capable of real-time assessment of tissue oxygenation intraoperatively and, thereby, aiding surgeons to select an optimal site for fashioning gastrointestinal anastomoses.
The WiPOX device is designed for use during open or minimally-invasive procedures and integrates seamlessly with current intraoperative monitoring systems. Care has been taken to optimize the WiPOX device for application in various intraoperative conditions. In our observations during the course of the rat and pig experiments, we found that body fluids, such as blood, ascites, or bile, do not interfere with WiPOX efficacy; this is in contrast to other intraoperative oximeters in which these common intraoperative fluids have been reported to potentially decrease oximeter accuracy. In addition, light-emitting oximeters, such as the T-Stat® 303 Microvascular Tissue Oximeter (Spectros, Portola Valley, CA) are vulnerable to signal interference from the endoscopic light used during minimally-invasive surgical procedures. These devices are also expensive, with the T-Stat® oximeter base-unit costing over $29,000 and non-reusable probes each costing $350 .
The preclinical animal studies described in this report demonstrate the efficiency and accuracy of the WiPOX in measuring tissue oxygenation intraoperatively. More importantly, we have shown that the WiPOX is capable of detecting early tissue hypoxia at a point when no overt evidence of ischemia was observed by the surgeon. This is particularly important considering that tissue ischemia is a progressive process and subtle unrecognized tissue hypoxia could develop into significant tissue ischemia postoperatively. The data herein suggest that the use of WiPOX may detect otherwise occult peri-anastomotic ischemia and allow the surgeon to correct the anastomosis prior to completing the case. Similarly, in renal surgery, the use of WIPOX could help minimize the occurrence of ischemia induced renal dysfunctional following partial nephrectomy. These preclinical studies are, however, limited by the fact that the animals were not followed postoperatively to assess the association between WiPOX oxygenation data and rate of anastomotic leakage or tissue healing. These evaluations were not pursued as our goal was to expeditiously obtain “proof-of-principle” data for a timely application to begin a clinical trial using the WiPOX device in humans.
In order to validate the efficacy of the WiPOX in reducing anastomotic complications in humans, we have designed a prospective clinical trial in patients undergoing esophagogastrectomy for esophageal or GEJ cancer and expect the trial to be open by June 2010. The primary objective of this clinical trial is to determine whether introduction of the WiPOX to guide surgeon choice of anastomotic site will decrease the post-operative anastomotic leak rate compared to historical controls.
In conclusion, the novel WiPOX device has demonstrated highly accurate and reproducible real-time assessment of tissue oxygenation in human volunteers as well as intraoperatively during gastrointestinal surgery in animals. Furthermore, the WiPOX reliably detected tissue hypoxia following bowel resection in rats and swine prior to overt clinical signs of ischemia. A clinical trial to investigate the efficacy of the WiPOX in human patients is underway. Routine intraoperative use of the WiPOX for the real-time detection of peri-anastomotic tissue oxygenation may help surgeons reduce anastomotic complications following gastrointestinal surgery. In addition, this device is easily applicable to use in other indications such as intra- and post-operative monitoring of tissue oxygenation in microsurgery flaps.
The authors would like to thank the following bioengineering students from The City College of New York for their work in designing the device prototype: Christopher Hue, Jaafar Tindi, Rohan Shah, Saumya Banerjee, Farah Khan, Alina Levchuk, and Chinedu Chukuigwe. We also thank Dr. Stephen Solomon from MSKCC for his assistance with the large animal experiments as well as Desiree Ehleiter from the Research Animal Resource Center (RARC) at MSKCC.