Modeling open surgery in mice to explore peritoneal damage, carbon dioxide humidification and desmoidogenesis

Mouse surgery studies

Female BALB/c mice aged 7–12 weeks old, weighing 17–25 g were used according to the Institutional Animal Ethics Committee approval and the National Health and Medical Research Council of Australia as previously described [5]. Mice were purchased from the ARC (Perth, Australia) and housed in the PMCC facility with a 12 h light-dark cycle allowing ad libitum access to food and water. The delivery of HWCO₂ was via a HumiGard (Fisher and Paykel Healthcare, New Zealand) recalibrated to produce a pressure of 2 mmHg appropriate to mice (Figure 1(a)).

Figure 1:

Mouse surgery set up, engineered devices and effect of humidified-warm CO_2.

Normal BALB/c mice were anesthetized, intubated and placed on a warming pad as for insufflated mice used in previous laparoscopic studies [6] with or without a miniaturized gas diffuser designed for open surgery studies (Figure 1(b)). For these laparotomy investigations, mice were randomized to two groups: (i) open surgery with exposure to active airflow and (ii) open surgery with active airflow plus H-WCO₂ diffusion using a purpose engineered operating suite (Figure 1(c)). Replicating surgical trauma on the peritoneal abdominal wall has been described [5]. Briefly, peritoneal wall damage was simulated using sterile plastic serrated forceps on one side of the peritoneum. This was performed by a sole investigator whereby the mini gas diffuser was removed and the sterile plastic forceps used to abrade the mesothelial layer with five strokes to the right-sided wall at the 30 min time point standardized to the middle of the peritoneum approximately 1 cm from the abdominal incision. Peritoneal wall from the left side was used as a non-surgical intervention control. C57BL/6 mice (see further details below) were similarly managed to above but were ventilated with a nose cone (instead of intubated) due to the small size of the Apc^min/+:p53^+/−mice.

Equipment

To establish operating room conditions that accurately replicate that employed in the operating theatre a custom miniature laminar flow system was developed. Laminar flow was created with a perforated ceiling and a computer-cooling fan mounted directly above the mouse operating space. Flow velocity tested with a hot wire anemometer and calibrated to be 0.5 m/sec in accordance with the DIN standard 1946 for hospital ventilation. Fisher and Paykel Healthcare engineered an open-sided box constructed from Perspex with the aim of positioning the fan over the anesthetized mouse. Thus, the “mouse” operating suite recapitulated the environment used for patients (Figure 1(c)). Mouse vital signs were monitored as previously described as well as tissue analyses [5, 6]. Briefly, for perfusion indices presented in this study these were determined using an inbuilt pulse oximeter and SpO₂ sensor (PhysioSuite™; Kent Scientific, Torrington, CT, USA) which analyzes the rate of change between oxygenated hemoglobin and reduced hemoglobin as arterial oxygen saturation expressed as a percentage.

Desmoidogenesis studies

A model of desmoidogenesis was developed by crossing of Apc^Min/+ mice [15, 16] with p53^+/− mice on a C57Bl/6 background based upon a similar model using a different Apc mutation reported previously [9]. Due to breeding constraints due to reduced viability of female p53^−/− mice [17] only male mice were evaluated. Genotyping by polymerase chain reactions for the p53 exon deletion used the following primers: p53x6s 5ʹ-TTATGAGCCACCCGAGGT-3ʹ; p53x7as 5ʹ-TATACTCAGAGCCGGCCT-3ʹ; neo18.5 5ʹ-TCCTCGTGCTTTACGGTATC-3ʹ, reaction ratio 1:1:2 to generate product of 600 bp (KO) and 450 bp (WT), respectively. For the Apc^min mutation detection PCRs employed the following primers: MAPCMT 5ʹ-TGAGAAAGACAGAAGTTA-3ʹ; MAPC15 5ʹ-TTCCACTTTGGCATAAGG-3ʹ; MAPC9 5ʹ-GCCATCCCTTCACGTTAG-3ʹ to generate 600 bp (WT) and 300 bp (min allele) products, respectively. In these studies, no tissue was removed and the only intervention was either laparoscopy or laparotomy in order to determine the impact of surgical approach on subsequent desmoid risk. Desmoid burden was quantified macroscopically (number of lesions present) as well as microscopically at sites of surgical trauma (i. e. laparotomy wound).

Semi-quantitative analyses

Mice were culled at 24 h and 10 days. Scanning electron microscopy (SEM) was used to evaluate changes to morphology and alterations quantified using a customary scale adapted from the H-score method. Essentially, microvilli were evaluated per cell as normal, damaged or shortened and absent. The proportions were expressed as a percentage of the total. For mesothelial cells, integrity cells were scored as normal, retracted or bulging expressed as a percentage of the total [6].

Statistical analyses

Data are expressed as mean±standard error of the mean evaluated using GraphPad^® Prism 6 (La Jolla, CA, USA) and analyzed using 1- or 2-way ANOVA with Tukey’s multiple comparisons test or two-tailed unpaired t-test. A p-value of less than 0.05 was considered statistically significant.

Design of animal operating suite

Laminar flow ventilation is standard in current day operating rooms. Investigations have demonstrated the efficacy of the system to reduce the number of particles and bacteria in the operating room air. To achieve effective laminar flow, highly filtered air is blown with uniform velocity in single parallel flow lines over the target area. The air may be “changed” more than 300 times per hour at a flow velocity of 0.2–0.5 m/s. This continual flow is a relevant variable when recapitulating the operating environment in animal models of open abdominal surgery as it may impact the local temperature of the wound and the evaporative capacity of the operating environment has a desiccating effect on abdominal surfaces [5, 18].

To explore different modes of open surgery for testing the effects of HWCO₂ on the peritoneum mice were subjected to intubation and anesthesia as previously described [6] with an option to deliver HWCO₂ via a proprietary humidification device shown in Figure 1(a). A laparotomy model was developed where mice were exposed to ambient and passive airflow with and without provision of HWCO₂ delivered via a gas diffusion device (Figure 1(b)) [5]. To recapitulate active airflow indicative of a modern operating theatre a Perspex^® box structure was engineered to provide a constant down draft of ambient air to mice undergoing laparotomy (Figure 1(c)) with or without the gas diffusor device.

Temperature maintenance under active airflow

As all mice were the same strain (BALB/c) and balanced for gender the effects of various surgery modalities could be robustly compared within new, and against previously published data [5]. A rectal probe was employed to as a reliable method of recording core body temperature over time [19]. The current study focused on the active airflow surgery model finding that HWCO₂ allows the maintenance of statistically higher stable core body temperature for the first 15 min for each 5-min interval from the beginning and overall for all time points of the procedure compared to active airflow alone (Figure 1(d)). However, the temperatures measured previously in mice undergoing laparoscopy [6] and laparotomy [5] were all higher by approximately 1 °C (37.5 °C) where HWCO₂ was used. These data indicate that the active flow of air into an open abdomen is most effective at lowering core body temperature in mice.

Total body perfusion under active airflow

We monitored vital signs in all mice over the procedure and collected perfusion at the distant hind paw as previously described [5] to find that HWCO₂ allowed for significantly increased perfusion compared to active airflow alone (Figure 1(e)). The degree of perfusion was reduced by approximately one-third when compared to passive flow data but comparable to the passive flow with additional HWCO₂ [5].

Measuring peritoneal integrity after laparotomy under active airflow

At 24 h, post-surgery SEM micrographs reveal extensive loss of microvilli under active flow conditions (Figure 2(a)) while microvillus integrity was essentially comparable to baseline when HWCO₂ was employed (Figure 2(b)). The quantitative measures of extent of damage are presented in Figure 2(c). Historical data from passive airflow work [5] have been included to emphasize the effects of active vs. passive flow. Although this previous study by definition was conducted at an earlier time the strain of mouse, age range, gender mix and experimental team were comparable. When mechanical damage was introduced to one side of the peritoneal wall and evaluated 24 h later all groups displayed maximal microvilli damage (Figure 2(d)).

Figure 2:

Comparing peritoneal damage following laparotomy at 24 h post-treatment.

Figure 2(a) and 2(e) highlights the separation of cells from each other across the mesothelial cell sheet that typically lines the peritoneum. This effect was most pronounced in the active airflow group and the delaminating/retraction of cell was reduced when HWCO₂ was used (Figure 2(f) and (g)). By contrast, the extent of mesothelial damage was pervasive in all groups when mechanical damage was assessed (Figure 2(h)).

Longer term measures of peritoneal integrity

By day 10, microvilli and mesothelial cell integrity is essentially normal in all groups of mice regardless of HWCO₂ delivery (Figure 3(a) and (c)). However, as previously discovered the introduction of mechanical damage at the time of surgery has an enduring impact on microvilli restoration and mesothelial cell integrity (Figure 3(b) and 3(d)). By both cellular indicators, the use of HWCO₂ appears to accelerate or perhaps allow tissue repair such that base line normal cellular SEM micrographs were observed at day 10. These data suggest that HWCO₂ provides protection to both exposed and additionally mechanically damaged peritoneum over the 1 h of operating time.

Figure 3:

Comparing peritoneal damage following laparotomy at 10-day post-treatment.

Development of a desmoid tumor model in mice

To generate a model of desmoidogenesis Apc^Min/:p53^−/− mice were bred on a C57BL/6 background. The mouse model allowed one hundred percent penetrance of desmoid tumors. Mice were culled due to illness that most typically involved anemia associated with bloody stools or morbid behavior. This pathology is typical of Apc^Min/+ mice that have developed intestinal adenomas which were subsequently confirmed by examination of the large and small intestines. Co-incident with adenoma formation were the development of desmoid tumors on the abdominal walls of Apc^Min/:p53^−/− (Figure 4(a)) but not Apc^Min/:p53^+/− or Apc^Min/+ mice. Such tumors are highlighted by staining fresh specimens with Fast Green (Figure 4(b)) and were further characterized by scanning electron microscopy (Figure 4(c) and 4(d)). Clinical features of mice evaluated in this study are reported in Table 1.

Figure 4:

Development of a mouse model for the study of the effects of surgery on desmoid tumorigenesis.

Apc^Min/+:p53^−/− mice had a slightly reduced time to development of disease compared to Apc^Min/+ mice while the p53^+/− genotype afforded extended disease-free survival (Figure 4(e)) and more rapidly than expected for mice on a p53^+/− or p53^−/− background alone [17]. The small and large intestinal adenomas were observed in all mice with the mutant Apc allele. However, only p53^−/− genotypes had the additional burden of desmoid tumors (Figure 4(f)). Of note only p53^−/− male mice were investigated in these studies due to the general low yield of p53^−/− female mice. This sex bias presents a limitation of this pre-clinical study because desmoid tumors are twice as frequent in females vs. males [20].

Effect of surgery approach on desmoid tumor formation

To evaluate the effect of laparoscopic and laparotomy surgery on the degree of desmoid formation in Apc^Min/+:p53^−/− mice animals were culled when declared by an independent animal technician blinded to the genotype and surgery approach. Similar tumor burdens were found in the no-intervention and surgery groups (Figure 4(f)).

The humidified-warm (HW) and dry-cold (DC) CO₂ groups were pooled according to surgery approach whereby there were no significant relationship between desmoid tumor burden and age at cull (Figure 5(a)). However, when the desmoid tumors were evaluated in terms of days post-surgery there was a trend toward an increased burden in relation to time following surgery. By contrast, adenoma burden in Apc^Min/+:p53^−/− mice increased with age as expected as well as with time from surgery (Supplemental Figure 1). There was no significant relationship between adenoma tumor burden, CO₂ type or surgery modality although there was perhaps a trend toward more adenomas in mice that received HWCO₂ but these mice were generally older at the time of cull consistent with an increase in adenomatogenesis with age in the Apc^min/+ model. When all genotypes were considered collectively vs. HWCO₂ and either DCCO₂ or passive airflow there was no significant differences in time to cull (data not shown).

Figure 5:

Desmoid tumor burden increases with time following surgery but not age.