Male Wistar albino rats were purchased from Laboratory Animal Services Medical School (The University of Adelaide, SA, Australia), 8 to 10 weeks old, with an average weight between 350 and 500 grams. Rats were housed 1 week prior to surgery under standard laboratory conditions (temperature 21°C ± 2°C, humidity 55% ± 10%, 12:12 h light-dark-cycle). Rats were housed in groups of 3 per cage and food and water were provided in a standard manner. Kaolin [Aluminum silicate Hydroxide, Al₂Si₂O₅(OH)₄] was purchased from Sigma-Aldrich, St. Louis, Missouri, United States.

Surgical procedures were performed by the same surgeons (RSV, CB) after a period of training and a maximum group size of five animals per day was used to ensure close monitoring during the immediate postoperative period. Anesthesia was achieved using a sealed chamber to deliver 2–3% Isoflurane after which the animal was positioned for surgery in supine position and anesthesia maintained with isoflurane over an open mask. Analgesia was provided pre-operatively by subcutaneous injection of Buprenorphine (0.05 mg/kg) and post-operative 8 hourly for 48 h. The surgery was conducted in aseptic manner and a prophylactic dosage of broad-spectrum antibiotic in the form of Amoxicillin Clavulanic acid 5 mg/kg (Clavulox^* Zoetis Australia, Rhodes, NSW, Australia) was also administered via subcutaneous injection.

Rats then underwent a laparotomy and a colon abrasion (9) or a colon abrasion with enterotomy. Briefly, the abdomen was shaved and prepared with alcohol and after drying, a 3 cm laparotomy was performed to gain access to the abdominal cavity (Figure 1A). In the caecal abrasion group, the caecum was delivered (Figure 1B) and kept moist with saline-soaked gauze whilst a dry gauze was used to rub the caecum repeatedly until sub serosal bleeding occurred over an area of 1 cm² (Figure 1C). The caecum was then returned to the abdomen and the abdominal wall closed in layers with a 3-0 Polyglactin suture. Prior to the placement of the final abdominal closure suture, rats were randomized to receive the following treatments:

The operation was limited to <20 min each rat so as to avoid air drying of the organs.

In a second stage, we used a colon abrasion with enterotomy model (n = 5) to simulate a colon resection with anastomosis performed at a different site on the caecum to the abrasion. Rats underwent a laparotomy as above followed by a caecum incision to create a full thickness enterotomy over a length of 1 cm away from the abrasion site. The enterotomy defect was then closed with a continuous 4-0 PDS suture (resorbable, monofilament) and the repair leak tested with a simple pressure test. 2 ml 0.005 g/mL Kaolin in saline was instilled over the abrasion (Figure 1D) and sutured site before closure of the abdominal wall. The rats in this group were monitored for 3 weeks as part of the larger experiment protocol.

Post-surgery, the animals were housed individually in separate cages. Animals were monitored postoperatively 8-hourly for the first 48 h to observe their weight, behavior, physical well-being and appearance by using the Clinical Record Sheet, as approved by AEC. Distress scores higher than 6 or weight loss >15% required that animals be humanely killed.

The animals were humanely killed on postoperative day 14 and scored based on the presence and severity of adhesions using a previously validated adhesion scoring system Table 1 (10). This scoring system takes into account the number, strength and distribution of adhesions formed (Figure 2). Pictures were taken by iPhone 8 12 mp f /1.8 aperture camera and also evaluated by a blinded observer.

The caecum, and adhesions between the caecal adventitia and adherent, adjacent intestinal serosal surfaces, and between the adventitial aspect of the caecum and the parietal peritoneum of the abdominal wall, were collected and immersion-fixed in 10% neutral buffered formalin. These tissues were then paraffin-embedded, cut at 6μm, and stained with haematoxylin and eosin (H&E) (Figures 3A–F). Duplicate sections were also stained by the Masson's trichrome technique to demonstrate collagen deposition in fibrous adhesions (Figures 3G–I).

All statistics were performed using R statistical software (R Foundation for Statistical Computing, Vienna, Austria) through the Jupyter notebook interface. The R package “MASS” (11) was used for ordinal regression. The “polr” function from MASS was used to fit a proportional odds logistic regression model for the ordinal outcome variable (the adhesion score as scored by the primary surgeons), with the treatment as the explanatory variable. A Likelihood ratio test (using the R function “anova”) was used to compare the model with a null ordinal regression model. The means of the ordinal response (interpreted as a numeric value from 1 to the number of classes) were calculated and post-hoc pairwise contrasts for each pair of levels of the treatment variable were compared using the “emmeans” package (11), using the Tukey method for p-value correction. Statistical significance was taken at the traditional 0.05 level.

Control rats receiving colon abrasion and saline and had variable adhesion scores with a mean adhesion score of 1(SD1) and 2/5 having an adhesion score of 0 (no adhesions) (Figure 2A).

Four rats died in the treatment groups with high Kaolin doses, 2 in the group treated with 4 ml 0.25 g/ml and 2 in the group receiving 2 ml 0.1 g/ml Kaolin (Figures 2B,C). Post-mortem evaluation showed severe adhesions with complications of intestinal obstruction, thought to be the likely cause of demise. The remaining rats which lasted the full 14 days showed mean adhesion grades of 4 (SD 0.44) (Figure 3A) and 4.6 (SD 0.6324) (Figure 3B) for 4 ml 0.25 and 2 ml 0.1 g/mL respectively.

Five rats received 2 ml 0.005 g/mL of Kaolin. These rats tolerated the procedure well with no significant morbidity or mortality at the end of 14 days recovery period. The resultant adhesions were mean grade 3.4 SD 0.54 (Figure 2D). The grade of adhesions was significantly greater in the 0.005 g/mL Kaolin treated rats compared to saline treated rats (p < 0.0001). Similarly, the abrasion with enterotomy group, treated with Kaolin 0.005 g/ml showed much thicker and vascularized adhesions consistently over the enterotomy site in comparison to the abrasion site. These rats had mean adhesion grade of 4(SD 0.816) and was significantly higher than the Kaolin 0.005 g/ml treated abrasion alone model with adhesion grade 3.4(SD 0.54) (p < 0.0001).

Microscopic analysis of the various grades of adhesions formed in the presence of Kaolin showed classical foreign body (FB) reaction with granular activity at the epicenter of inflammation (Supplementary Figure 1I) which was not seen in the saline treated caecum. The FB reaction was in the form of numerous invading macrophages (Supplementary Figure 1III), which contained phagocytosed kaolin, with active fibrovascular granulation tissue formation with numerous proliferating fibroblasts and supportive micro-vessels (Supplementary Figure 1II). There were mature adhesions with abundant compact collagen and fewer fibroblasts in the 0.1 and 0.25 g/ml kaolin treated rats. The presence of adhesions was predominantly confined to the abrasion site and one rat to the abdominal wall at the suture site (Supplementary Figure 1V). Masson's trichrome stain (MT stain) demonstrated a clear pattern of adhesion formation due to fibroblastic activity at various stages (Supplementary Figures 1VI–IX). The adhesions from the rats that were treated with higher concentrations of Kaolin (0.25 and 0.1 g/ml) showed a very irregular pattern of collagen distribution (Supplementary Figure 1VI) compared to the uniform nature in the lower dosage group of Kaolin 0.005 g/ml (Supplementary Figure 1IX).

A proportional odds logistic regression model for the ordinal outcome variable was fitted and the means of the ordinal response were calculated as described in the Methods. The means (and standard errors) resulting from fitting the ordinal regression model were plotted in Figure 4. Post-hoc pairwise comparisons showed a significantly higher mean adhesion grades for Kaolin doses 0.1 g/ml and 0.005 g/ml compared to normal saline (p < 0.0001; Figure 4).

Another ordinal model was fitted to compare the adhesions caused by Kaolin 0.005 g/ml in the Laparotomy model with abrasion vs. the same dose in the abrasion with enterotomy model. The Enterotomy model showed a higher mean ordinal response (difference in mean ordinal response = 1.3; p < 0.0001).

PW is on the patent for Chitogel and combination therapy. SV is on the patent for combination therapy. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.