Five Year Follow-Up of a Randomized Controlled Trial on Warming and Humidification of Insufflation Gas in Laparoscopic Colonic Surgery—Impact on Small Bowel Obstruction and Oncologic Outcomes

In laparoscopic surgery, the abdominal wall is commonly distended using carbon dioxide (CO₂) insufflation to provide pneumoperitoneum.¹ The gas is delivered at room temperature (19–21°C) with a relative humidity approaching 0% at the point of entry into the peritoneal cavity.² Early data suggested that unconditioned gas can cause structural and biochemical injury to the peritoneal mesothelium, and that warming and humidification of the insufflation CO₂ resulted in reduced postoperative pain after laparoscopic procedures.^3–7 However, more recent evidence from high quality, randomized controlled trials and a Cochrane meta-analysis have shown this not to be the case, with no difference in postoperative pain scores or opiate use with warming and humidification.^8–10

With no demonstrable difference in short-term clinical outcomes, attention has now shifted toward long-term outcomes, namely adhesion formation and oncological response. There is now laboratory evidence to suggest that conditioning of insufflation gas may in fact reduce postoperative adhesion formation,^11,12 and peritoneal tumor implantation.¹³ It is thought that this is because conditioning insufflation gas reduces the peritoneal inflammatory response. However, clinical evidence to confirm these findings has been lacking.

We previously published a multicenter, double-blinded, randomized controlled trial investigating warming and humidification of insufflation gas in laparoscopic colonic surgery.⁸ This study showed that warming and humidification did not confer any clinically significant short-term recovery benefit in laparoscopic colonic surgery. In light of the recent laboratory study findings mentioned above, we aimed to test the hypothesis that warming and humidification of insufflation CO₂ leads to reduced adhesion formation, and improved oncologic outcomes in laparoscopic colonic surgery.

Materials and Methods

Study Group

The study group received warm, humidified insufflation gas. Insufflation pressure was set at 12 to 15 mmHg with a variable flow rate. The gas used was carbon dioxide (carbon dioxide medical gas, BOC Ltd, Auckland, New Zealand), and this was warmed to 37°C and humidified to 98% RH using a laparoscopic humidification system (Fisher & Paykel MR860, Fisher & Paykel Healthcare, Auckland, New Zealand). This humidification system is specifically designed to deliver warm, humidified CO₂ to patients undergoing laparoscopic surgery, and has previously undergone independent testing by our group to confirm the effectiveness of gas conditioning.¹⁵ The gas is passed from the insufflator through a chamber which is filled with 30 mL of sterile water and sits on a heater plate. Water evaporates from the chamber into the gas that flows over it. The temperature of the gas is maintained as it travels along a heated tube to the laparoscopic port and into the patient's abdomen. The humidifier monitors the temperature and flow rate of the gas at the chamber outlet with a probe attachment, controlling the amount of power delivered to the heater plate to maintain the chamber set point temperature.¹⁶

Control Group

The control group received standard dry carbon dioxide for insufflation (carbon dioxide medical gas; BOC Ltd) delivered at room temperature (19°C) and 0% RH. Insufflation pressure was set at 12 to 15 mm Hg with a variable flow rate.

Objectives

The objective of this study was to test the hypothesis that warming and humidification of insufflation CO₂ would lead to reduced adhesion formation, and improve oncologic outcomes in laparoscopic colonic surgery.

Outcomes

All data were collected by a single, blinded investigator (TS) to ensure standardization of data collection.

Baseline data

Primary Outcomes

Patients were followed up for 5 years from the date of their surgery. Electronic patient records were reviewed, and admissions for small bowel obstruction were recorded, as well as whether management was operative or nonoperative. Local and systemic cancer recurrence as detected on cross-sectional imaging using CT, MRI, or PET scanning was recorded. Histological confirmation was required for local recurrence, but not for systemic recurrence, as biopsy may not have been clinically indicated if the management of metastatic disease was noninterventional. Five-year overall survival and cancer-specific survival rates were also measured.

Randomization

Blinding

The patient, study investigators, surgeon, and medical staff responsible for patient care were all blinded to patient allocation. This was achieved by having the humidifier connected to the insufflation apparatus and power supply regardless of allocation, and covered with a specially designed plastic casing which concealed its LCD screen and water chamber. This was to ensure that none of the theatre occupants were able to tell if the humidifier was switched on or not. A research assistant not involved in patient management, study design, data collection, data analysis or results write-up was responsible for setting up the humidifier. After the patient was asleep, the assistant opened an opaque envelope with allocation instructions and set up the humidifier away from view of the theatre staff and investigator. If the patient was in the study group, 30 mL of sterile water was added to the chamber and the humidifier was switched on and muted so that it did not make any noise. If the patient was in the control group, water was not added and humidifier was not switched on.

The blinding protocol was practiced in simulation several times and tested in March 2008 (prior to study commencement) on a consented patient undergoing a laparoscopic colonic resection at the Counties Manukau DHB. The patient, study investigators, surgeon, and all medical staff responsible for patient care were blinded successfully.

Data analysis was also blinded. The investigator undertaking statistical analysis on study completion was only allowed access to modified data tables with the allocation concealed. These specified patients as being allocated to “group 1” and “group 2.”

Statistical Methods

Results were analyzed using statistical software (SPSS for Windows, version 22.0, IBM Corp, Armonk, New York). Continuous variable parametricity was tested using the Shapiro-Wilk test. Results are presented as mean (standard deviation) for parametric data and median (interquartile range) for nonparametric data. Groups were compared using the Fisher's exact or χ² test for categorical variables, the Mann-Whitney U test for nonparametric continuous variables, and the t-test for parametric continuous variables. Statistical significance was accepted at the 0.05 level.

Ethics Approval and Trial Registration

Ethics approval was granted by the Ministry of Health, Northern X Regional Ethics Committee. Approval was also granted by the clinical boards of all three DHBs. The trial was prospectively registered with ClinicalTrials.gov (trial identifier: NCT00642005, US National Library of Medicine, 8600 Rockville Pike, Bethesda, MD 20894, USA).

Results

Participant recruitment and flow

Detailed patient flow is shown in the CONSORT diagram (Fig. 1).²⁰ Between April 2008 and June 2009, 111 patients were screened for inclusion. Of these, 82 patients were randomized equally between the study and control groups. Six patients in the study group were excluded after randomization: 2 patients had a rectal lesion below 15 cm found intraoperatively, 2 patients did not have any colon resected despite initial plan, 1 patient had an unplanned diverting ileostomy performed due to a positive anastomotic air-leak test, and in 1 case the investigator was unblinded when a nurse inadvertently lifted the plastic cover off the humidifier during surgery. Two patients in the control group were excluded after randomization: 1 patient did not have any colon resected despite initial plan, and 1 patient suffered a severe anaphylactic reaction on induction of anesthesia, and the procedure was abandoned.

View Figure

• Download as Powerpoint
• Download Figure

Baseline characteristics

Groups were well matched at baseline, with no significant differences in age, sex, BMI, ASA, Cr Possum, previous abdominal surgery, operation performed, diagnosis, or histological stage (Table 1).

Table 1 Baseline characteristics

View Fullscreen

The operative time, pneumoperitoneum time and volume of CO₂ used were similar between groups (Table 2). There was a higher conversion rate in the control group (15.4 versus 5.7%), but this was not statistically significant. Out of 6 conversions in the control group, 2 were due to extensive intraoperative adhesions preventing dissection, 1 due to invasive disease requiring en-bloc resection, 1 due to intraoperative bleeding from inadvertent injury to a gonadal vessel, 1 due to inability to localize the lesion tattoo, and 1 due to failure of the bean bag used to secure the patient to the operating table. The first conversion in the study group was due to difficulty extracting a large specimen, and the second due to invasive disease requiring partial bladder resection. The total wound size was similar in both groups despite this. Complication rates and grades were equivalent.

Table 2 Intraoperative parameters and post-operative complications

View Fullscreen

Small bowel obstruction

There was no statistically significant difference in the rate of admission for small bowel obstruction between the two groups (Table 3). Only 2 patients were admitted with a small bowel obstruction, and both of these were in the study group. Both patients were given gastrograffin but proceeded to open adhesiolysis as the obstruction did not resolve with nonoperative management.

Table 3 Five-year follow-up

View Fullscreen

Discussion

We conducted a 5-year follow-up of a multicenter, double-blinded, randomized controlled trial investigating warming and humidification of insufflation carbon dioxide in laparoscopic colonic surgery. There was no statistically significant difference in rates of admission for small bowel obstruction, oncological outcome, or survival between groups.

There are no other published randomized trials investigating long-term outcomes of warming and humidification of insufflation gases in laparoscopic surgery. There is only one published clinical trial evaluating conditioning insufflation gas and postoperative adhesions in laparoscopic endometriosis surgery. However, comparisons with this study cannot be made as there was significant cointervention in the treatment group with humidification, cooling, addition of other gases, intravenous steroid administration, and pharmacological barrier application all administered simultaneously.²¹ It is therefore unclear which of these interventions led to the reported reduction in adhesion scoring at second look laparoscopy. In addition, the clinical relevance of the adhesion reduction with this strategy was not evaluated with long-term follow-up.

Early data suggested that cold, dry gas can cause structural and biochemical injury to the peritoneal mesothelium, which results in a local inflammatory and cytokine response that increases postoperative pain, and potentially contributes to adhesion formation tumor seeding or systemic progression.^3–7 Animal models have actually demonstrated a reduction in peritoneal adhesion formation with humidification of insufflation gas which is at odds with our study findings.^11,22,23 There may be several reasons for this disparity. First, several animal models have not accurately emulated the clinical situation with insufflation flow rates and pressures that are exaggerated.²⁴ Furthermore, in most animalmodels, no intraoperative surgery is actually performed, and it may be that the peritoneal inflammatory response after colonic surgery is so extensive, that it overshadows any immunological effects caused by the insufflation gas per se.^25–29 This is supported by the finding in our original RCT that the inflammatory response as measured by local and systemic cytokine concentrations was equivalent in both groups,⁸ and by the fact that the difference in the inflammatory response between laparoscopic and open surgery (particularly in colonic resections) is minimal.³⁰ Another important consideration is that adhesion scores in animal models are notoriously unreliable, and do not necessarily correlate with adhesions that manifest as clinically significant (i.e., cause a small bowel obstruction).^31,32

There is also laboratory evidence that conditioning of insufflation gas reduces tumor implantation.¹³ Once again, translation of these findings to the bedside is difficult because of inherent limitations of the animal model. However, this and other studies by Binda et al. have shown that there may be additional benefit with cooling and supplementation with oxygen and nitrous oxide (which also reduce adhesion formation). It may be that this method should be further investigated in the setting of a clinical trial.^23,33,34

We note some weaknesses of our study. There was a nonsignificant trend toward more right sided cases in the study group, and more left-sided cases and conversions in the control group. Importantly, there were no resulting differences in operating time, pneumoperitoneum time or volume of gas used, and had these seemingly random differences been important, they would all have actually favored the study group in terms of adhesions and recurrence. Another weakness is that this study was plainly not powered for the endpoints measured at 5 years and given the very low rate of bowel obstruction and recurrence in this group of patients it is likely that the study is significantly underpowered. A retrospective power calculation suggests that 750 patients would have been required in each arm to have an 80% power to detect a 50% decline in adhesion or recurrence from 6 to 3%. However, it must also be noted that there was not even a trend in favor of gas conditioning. In fact, the reverse is true, with a trend favoring unconditioned insufflation gas for both primary outcomes. Even if this is taken as a pilot study in this context (which is probably appropriate), it is unlikely that this trend justifies a large multicenter RCT given the numbers and resources that would be required to achieve this. A third weakness is that we were not able to assess intra-abdominal adhesions visually in this context, but relied instead on the surrogate measure of adhesive small bowel obstruction. Arguably, however, adhesions in the absence of small bowel obstruction are not necessarily clinically relevant.

Conclusion

Warming and humidification of insufflation CO₂ in laparoscopic colonic surgery does not appear to confer a clinically significant long term benefit in terms of adhesion reduction or oncological outcomes, although a much larger RCT would be require to confirm this.