Pressurized Intra-Peritoneal Aerosol Chemotherapy (PIPAC): Does increased intraperitoneal pressure change distribution patterns and penetration depth of doxorubicin in a sheep model?

Background: Pressurized Intra-Peritoneal Aerosol Chemotherapy (PIPAC) is an innovative treatment against peritoneal carcinomatosis. Doxorubicin is a common intra-venous chemotherapy used for peritoneal carcinomatosis and for PIPAC. This study evaluated the impact of increased PIPAC intraperitoneal pressure on the distribution and cell penetration of doxorubicin in a sheep model. Methods: Doxorubicin was aerosolized using PIPAC into the peritoneal cavity of 6 ewes (pre-alpes breed): N=3 with 12mmHg intraperitoneal pressure (group 1) and N=3 with 20mmHg (group 2). Samples from peritoneum (N=6), ovarian (N=1), omentum (N=1) and caecum (N=1) were collected for each ewe. The number of doxorubicin positive cells was determined using the ratio between doxorubicine uorescence-positive cell nuclei (DOXO+) over total number of DAPI positive cell nuclei (DAPI+). Penetration depth (μm) was dened as the distance between the luminal surface and the location of the deepest DOXO+ nuclei over the total number of cell nuclei that were stained with DAPI. Penetration depth (μm) was dened as the distance between the luminal surface and the location of the deepest DOXO+ nuclei. Results:

therapeutic possibilities become rare and prognosis is poor [5]. Although the recent availability of bevacizumab treatment improved the survival rate of these patients, surgery is rarely feasible and the effects of chemotherapy remain limited. Thus, nding new therapeutics for these patients is urgent [6].
In most cases, ovarian cancer is restricted to the peritoneal cavity without distant organic metastasis (stade IIIC in FIGO classi cation) [7]. This is the ideal target for intra-peritoneal treatment. In 2012 a new method for intra-peritoneal administration of chemotherapy, Pressurized Intra-Peritoneal Aerosol Chemotherapy (PIPAC), was developed, where the chemotherapy is nebulized at body temperature in the intra-peritoneal cavity during laparoscopy [8]. The conversion of liquid chemotherapy into droplets is thought to enable homogeneous peritoneal distribution. Moreover, compared to a simple lavage, drug administration under the pressure used for the laparoscopy was shown to induce a better penetration of drugs in an in vitro model [9]. Finally, the plasmatic passage is negligible, thus limiting side effects of chemotherapy [10,11]. Standard intra-abdominal pressure used in the initial published protocol was 12 mmHg [12], which has been followed for clinical use.
So far, clinically, doxorubicin, used for chemotherapy of ovarian cancers, is also used with the PIPAC procedure. It acts through the inhibition of DNA transcription. Three interventions at 4-6 weeks interval each were shown to largely reduce peritoneal carcinomatosis [13,14]. Furthermore, the quality of life of patients being treated with PIPAC chemotherapy seems to be maintained [15,16]. These encouraging pioneer data prompt the needs for further evaluation and improvement.
In this context, the objective of our study was to compare the penetration and the distribution of doxorubicin administered with PIPAC using two distinct intra-peritoneal pressures (12 and 20 mmHg).
Experiments were carried out in sheep, of similar size and weight as humans so that the same equipment could be used. None of the large domestic animals spontaneously nor experimentally develop ovarian cancer similar to humans, so a healthy model was used.

Ethical statement
The project was approved by the local ethics committee (N°16 in the french registry of ethical committees) of animal experimentation of the National Veterinary School of Alfort and validated by the French Ministry of Research under registration "APAFIS" number 2016113016134972. Sheep were euthanized under general anaesthesia after PIPAC procedure and before sampling. This was performed by a trained team. All precautions were taken to limit anxiety and pain of the animals.

Experimental plan
Altogether, 10 non-pregnant multiparous ewes were used. The rst three animals were used for preliminary tests and development of the model. Thereafter, PIPAC was carried out as follows: (i) one control female with physiological serum, (ii) three females with a capnoperitoneum at 12mmHg (group 1), and (iii) 3 females with a capnoperitoneum at 20mmHg (group 2). To avoid a potential "day" effect, group 1 and 2 were performed alternatively (2-3 procedures/day). Animal characteristics are described in Table 1.

Surgical procedure
All PIPAC procedures were performed in the surgery theatre of the Biomedical research center (CRBM) of the National Veterinary School of Alfort.

General anaesthesia
The anaesthesia was carried out by a trained team. Animals were fasted for 12-16 hours before surgery.
PIPAC: surgical procedure The PIPAC was performed according to the safety rules described by Solaß (2013). All precautions were taken to ensure staff safety: every operator wore a surgical blouse, gloves, protection glasses and a high protection breathing mask.
After clipping the anterior abdominal wall, points were drawn on the skin for trocar localization 6 cm (laparoscopic camera) and 18cm (nebulizer) below the umbilicus. Two 12mm-incisions were made at these localizations (open-laparoscopy) and two 12mm-balloon trocars (Medtronic ® , Autosuture 12mm, BTT, Covidien) were inserted, ensuring tightness of the abdomen and steadiness of the pressure ( Figure   1). A capnoperitoneum was established and a camera was introduced in the abdomen for a short exploration phase. The nebuliser (MIP ® , Reger Medizintechnik, Tottweil, Germany) was connected to the high-pressure injector using a high-pressure injection line (Medrad, Mark 7, Arterion ® , Bayer). The distal part of the nebulizer was positioned at a 1cm depth, as measured from the trocar end. The sheep was placed in Trendelenbourg position to raise the rumen and provide a better exposition of the pelvis. Three milligrams of doxorubicin (Mylan ® , 2mg/mL) diluted in 50mL saline were nebulized at a owrate of 30mL/minute with a maximum pressure of 200psi, as usually recommended in human patients [17]. After nebulization, the capnoperitoneum was maintained during 30 minutes. The abdomen was subsequently de ated using an airtight device equipped with a smoke lter and connected to the waste air system in order to avoid contamination of the surgical room with doxorubicin. Thirty additional minutes were allowed for optimum drug penetration in tissues before the animal was euthanized with pentobarbital (Dolethal®, Vetoquinol, 3.6g, i.e., 20ml, IV). A median laparotomy was performed and 9 samples (6 peritoneal, 1 ovarian, 1 omental and 1 ceacal) were collected ( Figure 2). One more sample (omentum) was collected just facing the nebulizer. In order to ensure the reproducibility of the sampling for each animal, positions of the peritoneal samples were annotated relatively to their distance to the nebulizer. Samples were immediately frozen in isopentane at -40°C after horizontal inclusion in Optimum Cutting Temperature (Tissue-Tek® O.C.T. Compound, Sakura® Finetek). Blocks were kept frozen at -80°C.

Microscopic analyses
All analyses were performed blindly. The natural uorescent properties of doxorubicin was used for its localization in the tissues [18]. Samples were handled in a dark room to avoid light exposure that may decrease uorescence. Sections (7 µm) were cut using a cryostat (Leica ® CM1950), then mounted with 25µL anti-fade mounting medium (Vectashield®, Vector laboratories) that contained with 4,6-diamidino-2phenylindole (DAPI) at 1/1000. They were kept at 4°C until observation.
Analyses were performed with a Carl Zeiss (Germany) AxioObserver Z1 uorescence microscope equipped with an ApoTome slider and coupled to  Table 2.
Since all images were in the same horizontal plane, uorescence was not decreased depending on tissue depth.

Statistical analyses
Statistical analyses were performed with data collected from the 6 doxorubicin PIPAC-treated sheep. All analyses were performed with SPSS v15.0 and Stata v12.0 software (Stata Corp., College Station, TX, USA). Effect of treatment was analyzed using individual sample location, distance to nebulizer (for peritoneum, distinguishing frontal, proximal and distal samples) and histological type as variables. removed from analysis. The drug penetration depth was analyzed for each histological type and sample location. Tissue drug penetration was classi ed in 2 categories: <100µm and ≥100 µm for group comparison. In order to take into account the correlation between samples from the same ewe, a GEE model (Generalized Estimating Equation) was used [19] to compare penetration depth between the two groups. When one single sample was collected from each animal (ovary, caecum and omentum), drug penetration was compared using a one tailed Chi2 test.

Distribution patterns of doxorubicin
No nuclear uorescence in the >520nm wavelength (corresponding to the uorescence signal emitted by doxorubicin) was observed in any tissue collected in the control ewe. Doxorubicin was observed in 47 samples of the 54 collected (87%). Pressure increase had no effect on the distribution patterns of doxorubicin regardless of the tissue or peritoneal localization ( Figures 5 and 6). Cell nuclei distribution patterns of doxorubicin were heterogeneous in the peritoneal tissue. Almost all omental nuclei were DOXO+ (99%) whereas the caecum rarely stained positive (17%). Interestingly, in 4 of the 6 ovaries, DOXO+ cells were only found on one side of the ovary and not on the other (Figure 7).

Penetration depth of doxorubicin
Similar to cell distribution, penetration depth of doxorubicin was heterogeneous in the peritoneum with no signi cant difference between groups (p=0,69) when analysed altogether. Penetration depth was >100 µm in all group 1 ovarian samples versus 55% in group 2. There was a signi cant difference in penetration depth in the caecum between the 2 groups (100% for group 1 versus 22% for group 2). Regarding the omentum, 100% of sampled tissues showed a penetration depth > 100µm, regardless of the intra-abdominal pressure. These results are summarized in Table 3.

Discussion
In this study, a sheep model of PIPAC-doxorubicin was developed to evaluate the impact of intraperitoneal pressure on two parameters, namely the number of doxorubicin-positive cells and their localization relatively to the surface of the tissue (penetration depth). The sheep is human-sized model and the same parameters and surgical conditions are used as in human patients, making it very relevant for clinical practice.
This is the rst report assessing the impact of increased intra-abdominal pressure on penetration depth of chemotherapy. Penetration depth in the ovaries and caecum was signi cantly increased with a pressure at 20mmHg compared to 12mmHg but this increase was not consistent over all peritoneal samples. In the mouse model, Jacquet and Sugarbaker evaluated the effect of intra-abdominal pressure (12, 20 and 30 mmHg) on doxorubicin concentration in peritoneal tissues after the abdominal cavity was treated with doxorubicin as a simple lavage [20]. They showed that a higher pressure signi cantly increased doxorubicin penetration into the tissue. Nevertheless, 30 mmHg intra-abdominal pressure induced toxic effects, especially on digestive organs (necrosis). The impact of increased pressure (5, 10, 15 and 20 mmHg) was also studied in vitro using colon adenocarcinoma cells [21], with cytotoxic effects being signi cantly increased and proportional to pressure. The same team evaluated the effect of increased pressure on penetration depth of doxorubin in an ex vivo study (fresh porcine peritoneal tissue in a hermetically closed chamber) and did not demonstrate any signi cant effect [22]. These experiments suggest that peritoneal cells may be less permeable to doxorubicin that other cell types, as also observed in the present study. The formation of a liquid lm on the peritoneum after PIPAC may also contribute to the poorer effects of increased intra-abdominal pressure pressure on the peritoneum [23]. tissues that are positioned beneath other organs, as exempli ed with our observations for ovaries where doxorubicin only reached the ovarian side exposed to the nebulization. This observation could have important consequences in clinical practice. Nowadays, patients with recurrent peritoneal carcinomatosis from ovarian cancer often undergo an initial treatment with large abdominal surgery. These surgeries currently induce adherences between organs, thus potentially reducing access to many surfaces at the time when PIPAC is used. In any case, in practice, changing the direction of the trocar during the nebulization may help reach more peritoneal surface.
The data and conclusions drawn from this study deserve to be con rmed with a larger number of animals. Nevertheless, a signi cant effect of increased intra-abdominal pressure on penetration depth of doxorubicin was observed, suggesting that this should be further explored in clinical conditions. Furthermore, the experiments were performed on healthy tissues and the effect of pressure on doxorubicin penetration could be different on cancerous cells. Peritoneal carcinomatosis is not currently observed in domestic animals and large animal models of peritoneal carcinomatosis are required because laparoscopy and PIPAC could not be performed on rodent nor rabbit models.

Conclusion
Increased pressure was shown to increase penetration depth of doxorubicin in healthy abdominal tissues, suggesting that increased pressure may improve the e ciency of PIPAC on tumoral tissues in clinical practice. In order to con rm these encouraging results, large animal modesl such as sheep or pigs with peritoneal carcinomatosis should be developed for the bene t of oncologic research and especially PIPAC.     Standardized location of peritoneal samples (P1 to P6) according to distance to nebulizer   Description and comparison of intra-peritoneal distribution pattern of doxorubicin for each peritoneal localization