Postoperative monitoring of colorectal anastomosis – experimental study

Summary

Introduction:
Inadequate blood supply is one of the major risk factors for colorectal anastomotic leak. Early postoperative detection of local ischemic changes can predict complicated healing and lead to better outcome. Microdialysis (MD) offers real-time evaluation of adequate bowel perfusion through monitoring of tissue metabolism.

The aim of this study was to assess the feasibility of MD for early detection of ischemic changes in colorectal anastomosis.

Method:
Five pigs with end-to-end colorectal anastomosis were included. MD catheter was placed intramurally 5mm from anastomotic edge. Occlusive ischemia was induced after 3 measurements and followed by another 3 hours of monitoring. Tissue levels of different metabolites were measured every 60 minutes before and after ischemia induction. Mann-Whitney test was used to compare pre and post ischemic changes.

Results:
The monitoring of colorectal anastomosis using MD was technically feasible and associated with no complications. Significant changes caused by local ischemia were observed in decreased levels of glucose or pyruvate and increased levels of lactate and glycerol. All metabolic changes were detectable already in first samples 60 minutes after ischemia induction.

Conclusion:
Postoperative ischemic changes in colorectal anastomosis can be detected by means of microdialysis.

Key words:
colorectal anastomosis − anastomotic leak − microdialysis

INTRODUCTION

Although due to the introduction of minimally invasive approaches along with modern perioperative care principles the recovery was significantly shortened, the incidence of postoperative complications, including anastomotic leak (AL), has not changed significantly in colorectal surgery [1]. The rate of AL after low anterior resection is reported (depending on the definition) in the range of 3−27% [2,3]. AL significantly increases the risk of re-laparotomy, permanent stoma, prolongs length of stay and increases overall costs and mortality (19−33%). AL is affecting not only the oncological results in patients with rectal cancer but also functional outcomes with a significant impact on the quality of life [4−9].

Clinically significant AL usually requires re-laparotomy diversion of anastomosis and colostomy. However, the restoration of continuity is possible only in 30% of patients [10]. The alternative is to try to rescue the anastomosis together with ileostomy formation if this has not been done during primary procedure. Reoperation in these patients is subject to high morbidity (80%) and mortality, which is mainly due to the late diagnosis of AL [11].

One of the options of postoperative monitoring of colorectal anastomosis and early detection of AL is microdialysis (MD). This method allows monitoring of metabolic changes in the tissue close to the anastomosis by means of a two-lumen catheter with a diameter of up to 1 mm, ending with a semipermeable membrane. The catheter is continuously perfused with dialysate solution. On the semipermeable membrane (20 kD), the substances are exchanged based on the osmotic gradient and the resulting solution is collected and analyzed [12]. Substances determining the ratio of aerobic and anaerobic metabolism (lactate, pyruvate and their ratio, glucose, glycerol) are usually analysed.

Experimental studies have shown that MD can detect small bowel ischemia not only by a catheter placed intramurally or intraluminally, but also by a catheter inserted into the peritoneal cavity [13]. Based on these experiments, several clinical trials were conducted to diagnose early AL by detecting metabolic changes in the peritoneal effusion [14−17]. However, changes in the peritoneal fluid are not specific for AL and some increase of the lactate/pyruvate ratio occurs also after uncomplicated intraabdominal surgery [14].

Moreover, significant difference in complicated patients was in most studies based on the mean of all values measured within the first five postoperative days (area under the curve) and therefore this doesn’t prove clearly that intraperitonal MD is able to detect AL early [17].

This raises the question whether monitoring with MD catheter placed intramurally close to the anastomosis would be more effective.

The aim of this work was to verify whether MD is suitable for monitoring of ischemic changes of colorectal anastomosis.

METHOD

A total of 5 laboratory pigs were included in this study. After general anaesthesia was induced (Isoflurane - 1.5% in combination with Fentanyl – 3−5 ml/h i.v.) internal jugular vein was cannulated and low mid-line laparotomy was performed. The left colon was mobilized and divided at the proximal rectum, leaving the mesocolon intact. Subsequently, hand-sewn end-to-end single layer colorectal anastomosis was performed. One 0.9 mm MD catheter with a 30 mm long membrane (63 MD catheter, 40/30, M Dialysis AB, Sweden) was used to monitor the metabolism in the site of anastomosis. The catheter was inserted intramurally with the end of the membrane placed 5 mm from the anastomosis and fixed (Figure 1). The catheter was perfused with dialysate solution (T1 peripheral perfusion fluid, M Dialysis AB, Sweden) at a rate of 1 μl/min using a pump (107 MD pump, M Dialysis AB, Sweden) which allows analysis every 60 minutes.

After three baseline measurements, the inferior mesenteric artery was tied and the colon was skeletonized 4 cm proximal and distal to the anastomosis. After induction of ischemia, the monitoring continued for another 3 hours. The levels of the following substances were determined using the analyser (ISCUS flex MD analyzer, M Dialysis AB, Sweden): lactate, pyruvate, glucose and glycerol. Blood gas analysis was performed regularly and homeostasis was maintained by balanced crystalloids or glucose solution.

The experiment was terminated after the last measurement by closing the laparotomy and the animal was euthanized.

The protocol was approved by the Expert Commission for the Protection of Animals (Academy of Sciences of the Czech Republic) and study performed in accordance with applicable EU legal regulations (No. 246/1992, 207/2004).

With respect to the type of data distribution, these are presented as a median with ranges. A Mann-Whitney non-parametric test was used to compare the values ​​before and after ischemia induction.

RESULTS

Both surgery and MD monitoring were performed without complications and major technical difficulties. Only once the MD membrane was damaged due to manipulation when inducing the ischemia and the catheter had to be replaced for the second phase of the experiment. All animals remained haemodynamically stable throughout the whole experiment. Glycaemia had to be corrected with a bolus of glucose in 2 cases. All MD parameters remained stable for the first three hours before ischemia induction. Only nonsignificant decrease in levels of tissue glucose was recorded.

The values ​​before and after ischemia are shown in Graphs 1–4. Significant change in local metabolism caused by the ischemia were observed based on the decreased levels of glucose – 2.5 (0.2−8.3) vs. 0.1 (0−1.4) mmol/l; (p<0.0001) or pyruvate – 217 (113−394) vs. 79 (15−619) umol/l; (p<0.0001) and increased levels of lactate – 6.4 (3.9−9.4) vs. 13.4 (3.5−18.8) mmol/l; (p<0,0001) and glycerol 35 (14−80) vs. 118 (15−322) μmol/l (p<0.0001).

All metabolic changes were detectable and significant already in first samples 60 minutes after ischemia induction.

DISCUSSION

AL represents a severe complication after low anterior resection and is usually diagnosed between the 7th and 11th postoperative days. At this time the local and systemic inflammatory response is fully developed with diffuse peritonitis and the anastomosis may be morphologically altered so that its salvage is not possible [18]. It has been shown that each hour of delay in diagnosis and adequate treatment significantly increases the risk of mortality. Early detection of AL is therefore of extreme clinical relevance [19].

There are some known risk factors associated with AL and these can help the surgeon when considering protective stoma after low anterior resection. Current research is focused on the intraoperative evaluation of the viability of resection margins. These techniques, however, lead to a change in the surgical strategy only in a small number of patients [20]. The options of postoperative anastomosis monitoring are still limited and usually based on clinical observation or on changes in levels of systemic inflammatory markers [21].

However, recent studies on MD or intramucosal pH metrics showed that these techniques allow evaluation of the environment in close proximity to the anastomosis [14−17,22]. The outcomes so far do not allow wider implementation of these methods into clinical practice. Nevertheless, the idea of early postoperative detection of ischemic changes in the site of anastomosis which would allow immediate response and possibly increase a chance of anastomotic salvage is still tempting.

Our experimental study has shown that MD responses promptly to ischemic changes in colorectal anastomosis. All detected changes were significant during the first hour after ischemia. The monitoring was performed with catheter inserted intramurally close to the mucosa which is considered as the most metabolically active layer of the bowel due to its absorptive and secretory functions [23]. We proved this technique as simple, feasible and safe. The catheter can be later pulled through the abdominal wall and connected to the pump, which means that this technique is suitable for clinical use, where due to safety reasons only intraperitoneal catheters analysing metabolites in peritoneal effusion have been tested so far [12−17]. Some authors reported the changes in tissue metabolism resulting from the trauma of catheter insertion itself [24]. In our experiment, however, such deviations were not observed in pre-ischemic period.

We observed a relatively rapid decrease in tissue glycemia after ischemia induction, indicating an early depletion of glucose stores. However, its tissue level decreased even before ischemia, although the difference was not significant. This can be explained by either local trauma caused by manipulation when anastomosis was constructed or systemically as a result of perioperative stress, although systemic hypoglycemia was detected in only two animals. Similar results have been reported by the Danish experimental study published in 2004, when the authors used MD to monitor small bowel ischemia. Levels of tissue glucose were very variable before and dropped to zero after ischemia induction [13]. Based on our outcomes and those from the cited study, we can conclude that the tissue glucose concentration does not reflect the state of anastomosis accurately, and this parameter is therefore not ideal for prediction of AL.

Pyruvate is transformed into lactate during anaerobic metabolism, resulting in an increase of its tissue concentration and also in increase of the lactate/pyruvate ratio. Such changes can occur as a result of an inflammatory reaction at the anastomosis area, but also due to accelerated metabolic processes in postoperative period [12, 25]. The lactate level during our experiment responded rapidly to the induced ischemia and therefore it can be assumed that early anastomosis perfusion disorders could be detected effectively using this metabolite in the postoperative period. Data from Matthiesen et al confirmed increased lactate/pyruvate ratio in peritoneal effusion in patients with AL. The difference, however, became significant first on 5^th postoperative day when other markers like C-reactive protein or white blood count are usually already raised [15].

Glycerol is released from the complex with free fatty acids during lipolysis, and its concentration has been shown to correlate with the rate of cytolysis [26]. Elevated tissue glycerol levels were also observed during small bowel ischemia [27]. In our study the glycerol levels were relatively low and stable before ischemia. They increased significantly within the first hour after ischemia induction. Intraperitoneal glycerol concentrations in patients after major abdominal surgery were also studied in a Swedish trial published in 2011. The glycerol level was, on the contrary, lower in patients with severe postoperative complications (16 out of 60 patients - 27%). The authors explained this by influence of insulin and proinflammatory cytokines released during the inflammatory response and by higher consumption of glycerol as an energy source in anaerobic metabolism [16]. It can be assumed from these conclusions that the metabolic changes in the peritoneal effusion detected by MD are non-specific and can be affected by the overall state of the organism. Moreover, the values are also influenced by the position of the catheter [28]. On the other hand, catheter placed near the anastomosis reflects only local changes and could therefore be more beneficial in early AL diagnosis.

The limitation of our work is the model of ischemic anastomosis that does not fully correspond to reality. In our experiment it was not technically possible to adjust the degree of ischemic changes and a complete skeletisation of the bowel was performed. The ischemia of anastomosis is usually incomplete in real world. Since the changes detected by MD were significant (mostly highly) within 60 minutes after ischemia, it can be assumed that this method will be able to detect even minor hyperperfusion. This can only be verified in a human trial.

CONCLUSION

The experimental study has confirmed that MD is technically feasible for postoperative monitoring of colorectal anastomosis and allows early detection of perfusion changes.

This technique, in the authors’ opinion, is ready for evaluation in a clinical trial.

List of abbreviations

AL – anastomotic leak

MD – microdialysis

The work was created with the support of grant AZV 16 - 31806 A, project No. LO1609, with financial support from the Ministry of Education, Youth and Sports within the framework of programs NPP RVO 7985904 and MO 1012.

Conflict of interests

The authors of this study declare that they are not in conflict with the emergence of this study and that this study has not been published in any other journal.

MUDr. Jaroslav Kalvach

Department of Surgery,

2^nd Faculty of Medicine,

Charles University and Central Military 

Hospital, Prague

U Vojenské nemocnice 1200

169 02 Praha 6

e-mail: jaroslav.kalvach@uvn.cz