Background of the Invention SAPL is used clinically for the treatment of respiratory distress syndrome (RDS) in neonates. In this role, it has been assumed that the SAPL functions by reducing the high surface tension forces at the air-water interface within the alveoli, thereby reducing the pressure needed to expand the lungs, see Bangham et al., Colloids & Surfaces, 10 (1984), 337 to 341.

Post-surgical adhesions are the single greatest complication of many surgeries. Post- surgical adhesions are fibrous attachments between tissues that can form inside the body following surgery. Internal tissues that would normally be separate become joined by fibrous scar tissues called adhesions, as a result of the body's normal healing process. Complications from post-surgical adhesions can include chronic back or pelvic pain, intestinal obstruction and infertility. Complications can be severe enough to require re-operation, but adhesions can make subsequent surgeries more difficult to perform.

Surgical adhesions continue to pose a problem as a major cause of female infertility (Trimbos-Kemper T, Trimbos B, van Hall E. Etiological factors in tubal infertility.

Fertil. Steril. 1982: 37: 384-8), while intraperitoneal adhesions are the main cause of intestinal obstruction following surgery (Menzies D, Ellis H. Intestinal obstruction from adhesions: How big is the problem? M. . Co.-S-g. Eng7. 1990: 72: 60-3).

Postsurgical adhesions following back surgery are one of the leading causes of recurrent back pain. It is widely recognised that postsurgical adhesions play a major role in poor patient outcome following surgery, or failed back syndrome. Failed back surgery syndrome is seen in 10-40 percent of patients who undergo back surgery.

Postsurgical adhesions are also a significant problem after tendon surgery. Following tendon surgery, adhesions can inhibit the tendons'ability to glide, thereby limiting motion. In peripheral nerve surgery (e. g. carpal tunnel syndrome) adhesions may cause tethering and compression of nerve roots, leading to pain and loss of function.

A number of agents and methods have been employed to minimise adhesion formation with varying degrees of success, including the use of : corticosteroids (Hockel M, Ott S, Sieman U, Kissel T. Prevention of peritoneal adhesions in the rat with sustained intraperitoneal dexamethasone delivered by a novel therapeutic system. Ann. Chir. Gynaecol. 1987: 76: 306-13), non-steroidal anti-inflammatory drugs (DeSimone JM, Meguid MM, Kurzer M, Westervelt J. Indomethacin decreases carrageenan-induced peritoneal adhesions.

Surgery 1988: 104 : 788-95), calcium channel blockers (Steinleitner A, Lambert H, Montoro L, Kelly E, Swanson J, Sueldo C. The use of calcium channel blockade for prevention of postoperative adhesion formation. Fertil. Steril. 1988 : 50: 818-21), plasminogen activators (Menzies D, Ellis H. Role of plasminogen activator in adhesionprevention. Surg. Gynaecol. Obstet. 1991: 172: 326-66), and 'bioresorbable'mechanical barriers (Goldberg EP, Sheets JW, Habal MB. Peritoneal adhesions. Prevention with the use of hydrophilic polymer coatings. Arch. Surg.

1980: 115: 776-8; Best CI, Rittenhouse D, Vasquez C, Norng T, Subias E, Sueldo CE. Evaluation of Intercede (TC7) for reduction of postoperative adhesions in rabbits. Fertil. Steril. 1992 : 58 : 817-20 ; Burns JW, Colts MJ, Burgees LS, Skinner KC. Pre-clinical evaluation of Seprafilm bioresorbable membrane. Eur. J Surg.

1997: 577: S40-8), which can introduce their own problems.

WO 91/12026 (McNaught Medical) discloses a method of reducing surgical adhesions by means of coating tissue surfaces with a phospholipid, preferably

lecithin, in suspension or solution in an inert carrier, such as for example, water, saline, or propylene glycol, or mixtures thereof.

Also, WO 99/51244 (Britannia) describes the use of powdered phospholipids to prevent surgical adhesions.

US Patent 6133249 (McNaught Medical) describes a method of lubricating mammalian joints using a liquid composition comprising of phospholipids dispersed in propylene glycol.

Summary of the Invention The present invention is based on the surprising discovery that liquid, semi-liquid or pasty compositions of certain phospholipids dispersed in a physiologically acceptable carrier are equal to or better than the compositions of WO 99/51244 in reducing the risk of surgical adhesions.

The powder compositions of WO 99/51244 have the advantage that they are easily administered into body cavities such as the peritoneum by simple"puffers"or other gas stream delivery devices. The advantage of the liquid, and especially the paste, compositions of this invention is that the surgeon can apply the composition as a directed, substantially non-mobile,"slug"to be spread manually using a gloved finger, or apply compositions manually using a gloved finger directly to the desired site, and immediately check visually that the intended area is covered. This is especially advantageous in open surgery, such as on flexor tendons of the hand, or spinal surgery, or peripheral nerve surgery.

In one aspect the present invention provides a method of reducing the risk of surgical adhesions which comprises applying a composition comprising a SAPL to surfaces adjacent an incision during surgery, characterised in that the SAPL is DPPPC, or a mixture of DPPC and PG, or DPPG, dispersed in a physiologically acceptable non- volatile carrier liquid.

In another aspect the present invention provides the use of a SAPL to prepare a medicament for reducing the risk of adhesions following surgery, characterised in that the SAPL is DPPC or a mixture of DPPC and PG or DPPG, dispersed in a physiologically acceptable non-volatile carrier liquid.

The carrier liquid is one which is substantially non-volatile or only sparingly volatile at body temperature. Suitable carriers include physiologically acceptable glycols, especially propylene glycol, polyethylene glycols and glycerol.

The SAPL may be dispersed in the carrier so as to form liquid, semi-liquid or pasty compositions. Semi-liquid or paste compositions are preferred because they can be applied and spread by a surgeon using a gloved finger, and are particularly suitable for use in open surgery where the surfaces abraded by surgery are well defined and easily accessible.

Brief Description of the Drawing The Figure shows the sites for administration of SAPL compositions of this invention during tendon surgery.

Detailed Description of the Invention Pastes can be prepared by simply dispersing a SAPL powder in the carrier, or when appropriate dissolving the SAPL in heated carrier and allowing the SAPL to precipitate as a powder on cooling, preferably at a loading that will form a paste. A thick paste of the SAPL and carrier is ideal to apply to open wounds to which it adheres well. It enables a much higher concentration of the SAPL to be applied to the incision site.

Propylene glycol is especially effective as a carrier because at room temperature SAPL may be dispersed in it as a paste, but at body temperature a mobile solution is formed. A paste of 400 mg/ml of DPPC in propylene glycol has been shown to give 93% protection against adhesions in surgical tests, as described in the experiments below.

If desired, the pasty compositions of this invention may be applied to surgical sites in conjunction with the powder compositions of WO 99/51244 (Britannia), the latter being used to coat and protect more peripheral areas.

Various dispersions of SAPL in propylene glycol are described in US Patent 6133249, the entire contents of which are incorporated herein by reference. Similarly the powder compositions of WO 99/51244 may be dispersed in a carrier such as propylene glycol, and the entire disclosure of WO 99/51244 is also incorporated herein by reference.

In other medical uses of SAPL, spreading agents, especially PG, have been believed to enhance or potentiate the binding of DPPC to an epithelial surface. Surprisingly in the present invention, compositions based on DPPC alone have out-performed compositions based on DPPC/PG which have been especially effective in the situations covered by the patent applications acknowledged above.

A further surprising finding is that pastes prepared by dispersing coarse SAPL particles, for example around 1 Cum in size, are more effective than when using fine SAPL particles, such as around 5um in size. More generally, the powdered SAPL may have a particle size in the range of 0.5 to loom, more suitably of 0.5 to 2011m, preferably 0.5 to 10am.

The compositions may also include preservatives where appropriate, such as fungicides, bactericides and anti-oxidants.

The solutions/dispersions/pastes of the present invention are especially suitable for surgical procedures where there is a potential for adhesion in areas that are difficult to access by powder sprays. The Figure of the accompanying drawing shows a situation arising in tendon surgery where both a tendon (1) and its sheath (2) are cut through and require stitching back together. By means of a catheter (3), a solution dispersion or paste composition of the invention may be introduced between the tendon and the sheath to prevent adhesions that impair mobility.

The same situation applies in spinal surgery where the spinal cord is damaged within the spinal cord sheath, and with surgery on other nerves.

The viscosity of the solution/dispersion/paste may be varied to suit specific surgical sites or the preferences of individual surgeons. For example, in difficult to access areas (e. g. flexor tendons) a relatively low viscosity fluid may be used, whereas at a site where gravity will effect its distribution a higher viscosity composition may be preferred. Temperature will play a part in the viscosity of the compositions. Typically a range of 0.1-10 000 cP may be considered. However, formulations outside this range may be required for specific applications at the discretion of the surgeon.

Further details of the invention are illustrated in the following experiments.

MATERIALS Animals Eighty New Zealand white rabbits, weighing between 2 and 3kg were used since this breed has been found to be a good model for surgical adhesions (Bhandarkar DS, Nathanson LK, Hills BA. Spray of phospholipid powder reduces peritoneal adhesions in rabbits. Aust. N. Z. J Surg. 1999: 69: 388-90). They were given access to food pellets and water ad libitum before and after the operation.

Release agent/lllbricant The phospholipid mixture (Pumactant) was supplied by Britannia Pharmaceuticals Ltd, Redhill, U. K. Coarse Pumactant is a dry powder of 10pm particle size of a mixture of 70% DPPC and 30% egg PG co-precipitated from solution and, therefore, intimately mixed. Fine Pumactant has the same composition but has a particle size of 5lem. The DPPC is La-DPPC, which is the optical isomer which occurs naturally and is digested by phospholipases. La-DPPC was supplied by Lipoid GmbH, Ludwigshafen, Germany, who also supplied the dipalmitoyl phosphatidylglycerol (DPPG). The solvent used was clinical grade propylene glycol and all solutions were made to a concentration of400mg/mL of total phospholipid (i. e. DPPC + DPPG) per

mL at 60°C which reverted to pastes at room temperature. The compositions used were DPPC alone, DPPG alone, 7: 3 DPPC: DPPG and 1 : 1 DPPC: DPPG.

METHODS The protocol for the study was approved by the University of Queensland Animal Experimentation Ethics Committee.

Operation General anaesthesia was induced by a combination of intravenous ketamine (Parnell Laboratories, NSW) and xylazine (Ilium Xylazil: Troy Laboratories, NSW) and maintained by xylazine alone. Common to all procedures were aseptic operative conditions consisting of iodine for skin preparation, drapes to minimise contamination of the peritoneal cavity and non-powdered gloves worn by both the surgeon and assistant. Each animal underwent a lower midline laparotomy for creation of a standard separate 5 x 1 cm parietal peritoneal defect (marked with a silk suture at each end) sited to lie adjacent to a matching 5 x lom visceral peritoneal defect over the adjacent caecum. At the conclusion of the procedure the abdomen was closed with continuous 4/0 polybutester (Novafil, David Geck, Quebec) in two layers including the peritoneum.

Experiments Ten animals did not have any further intervention except as described above and provided the controls (Group 1). In 20 animals a lmm I. D. polyethylene tube was placed intraperitoneally adjacent to the caecal intraperitoneal defect and exteriorised through a subcutaneous tunnel. Ten of these rabbits (Group II) received an intraperitoneal puff of Fine ALEC (100mg) via a purpose-built device prior to abdominal closure. Another 10 animals (Group III) received Coarse Pumactant powder, i. e. the same material as used in a previous study (21).

In 10 animals (Group IV) 1mL of propylene glycol was applied directly to the defect on the surface of the large bowel by means of a hypodermic syringe to which no needle was fitted. This procedure was repeated in a further 40 rabbits in which one of four phospholipid compositions were substituted for the propylene glycol and applied

as paste. In the first 10 of these animals, the paste was La-DPPC (Group V). In the next 10 rabbits (Group VI) the paste was DPPG ; in the next 10 (Group VII) it was 7 : 3 DPPC: DPPG, while in the last 10 (Group XIII) it was 1 : 1 DPPC: DPPG.

One week after the operation, the rabbits were killed by an overdose of pentobarbitone (Lethabarb, Virbac, NSW) and a postmortem performed to assess the extent of adhesion formation. An adhesion is defined as any connection between bowel and the healed 5 x lcm parietal peritoneum marked by the silk sutures.

Adhesions formed occasionally on the undersurface of the wound, but were not analysed. Treatment was performed by one surgeon who also performed the postmortems.

Statistics The various groups were compared using a simple Student's t-test since there were equal numbers of subjects in each comparison-see Table 2. A p-value of 0.05 or below was regarded as significant.

RESULTS The results for the eight groups are given in Table 1 and depicted with errors of the mean in Figure 1. The protection rates (PR) quoted in Table 1 are calculated as: PR = 100 [ 1 - (adhesion length/adhesion length of controls)]%.................... (1) The following features warrant special mention: 1. The Coarse Pumactant produced a highly significant (p= 0. 005) reduction in adhesions to 8.4 3. 6 (mean sem) compared to controls of 29 5.4, corresponding to a protection rate of 71% 2. The Fine Pumactant was less effective than the Coarse Pumactant, the protection rate of 37% failing to reach 95% significance over controls.

3. Propylene glycol alone offered no protection relative to controls.

4. The pastes all gave significant protection from adhesions apart from pure DPPG, the efficacy increasing as the proportion of DPPG was decreased.

5. The best performance was that of DPPC paste which amounted to a protection rate of 93% relative to controls.

6. This performance was better than that of Coarse Pumactant, although a direct comparison did not reach statistical significance (p = 0.15).

DISCUSSION The particularly encouraging feature of the results for dry powders was the reproducibility of the protection rate of 71% for Coarse Pumactant by comparison with an earlier study (Bhandarkar et al, supra) performed by a different surgeon using the identical material. A dry powder which can be'puffed'on to the incision and surrounding area still offers a much simpler mode of application than approaches requiring very careful positioning of a thin digestible film as a gross physical barrier.

The results for Fine Pumactant (5u. m) were surprising in so far as a larger surface area of DPPC/PG might have been expected to prove more protective than its Coarse (10 urn) competitor. The lower efficacy could be attributed to the use of a co-powder mixture of DPPC and Egg PG as opposed to Coarse Pumactant in which co- precipitation from an organic solvent ensures mixing at the molecular level. Another factor might be wider dispersion of the same dose of powder, leaving less in the test area. Wide dispersion could prove a benefit in cases of abdominal trauma where bile salts escaping from the gut could have stripped the peritoneal cavity of its protective lining of SAPL, just as they can in the stomach (Hills BA (1989). Oligolamellar lubricationofjointsbysurface-activephospholipid. ARheumatol. 16: 82-91).

Where DPPC and PG are used in the same proportion, viz. 7 : 3, there is not much difference (71% vs 80%) in protection between powder and paste. This could be explained by the difference in dose, which was higher for the paste. However comparable the protection rates, they amount to a ratio of 29 versus 20 in the number of residual adhesions, i. e. a 30% difference.

Pursuing the theme of residual adhesions, the ratio extends to 84 for Coarse Pumactant versus 21 for DPPC paste, i. e. a ratio of 4: 1-see Table 1. Moreover only 1 animal in 10 displayed any adhesions at all with the paste. This is an important consideration in the clinical setting where any adhesions at all could result in a patient returning to surgery for their resection.

Hence, for cases where the boundaries of surgical intervention are well defined the DPPC paste offers an exciting and possibly more efficient means of preventing surgical adhesions. It is very simple to apply either by a hypodermic syringe without a needle or by simply smearing on the material with a gloved finger or any other mode of application which the individual surgeon may prefer. The DPPC paste is easy to autoclave (20 mins. at 121°C) in its container which can be a vial or hypodermic syringe. It can also be sterilised in its container by y-irradiation.

One great advantage of the paste is our past experience of using the same product (at half concentration) as a lubricant in the joint for treating osteoarthritis (OA) in 10 humans (Vecchio P, Thomas K, Hills BA. (1999). Surfactant treatment for osteoarthritis. Rheumatology. 38: 1020-1021) and its equine equivalent- degenerative joint disease-in 130 horses. We have had only two adverse reactions- one human and one equine. Both cases were"gouty reactions"attributed to "unidentified crystals"found in the synovial fluid aspirated from the joints. In those investigations the problem was attributed to crystals of DPPC separated from solution in propylene glycol. In subsequent clinical trials the paste was autoclaved within 2 days of injection and this has eliminated the problem. We have subsequently studied the crystallisation of DPPC from its supersaturated solutions (200mg/mL propylene glycol); although crystals may not prove such a problem between surfaces which are not weight-bearing. Incidentally propylene glycol is FDA-approved as a lipid vehicle for i. v. and i. m. administration.

Finally there are many clinical studies of the intraperitoneal injection of SAPL in its various forms into the peritoneal cavity with the intention of restoring ultrafiltration in CAPD as reviewed elsewhere (Hills BA, (2000). Role of surfactant in peritoneal dislysisParit Dial Internat 20: 503-515) In that review we could find no cases

of inflammatory reaction reported, although all except one animal study used SAPL as an aqueous suspension.

In conclusion it can be stated that, whereas dry Pumactant powder offers a good, safe and reproducible means of preventing about 70% of adhesions, the paste formulation offers a cheaper means of reducing those residual adhesions a further four-fold.

TABLE 1: NUMBER OF ADHESIONS FORMED PER ABRADED AREA (5x1cm) IN RABB8ITS Group I II III IV V VI VII VIII DPPC:DPPG DPPC:DPPG TEREATMENT CONTROL Fine Coarse propylene DPPC DPPG 7:3 1:1 Pumactant Pumactant powder powder Glycol only paste paste paste paste 38 0 0 0 0 0 0 0 36 0 0 0 0 0 0 0 25 0 23 50 0 0 0 0 36 50 0 0 21 0 0 0 Adhesion Length (mm) 46 0 19 0 0 0 0 0 30 37 0 50 0 0 37 0 29 0 28 50 0 0 00 50 27 0 50 0 37 0 0 0 28 14 50 0 43 0 32 0 40 0 0 0 45 22 0 Mean 29 18.2 8.4 25 2.1 12.5 5.9 3.2 SD 17 20.2 11.4 26.4 6.6 20.2 12.9 10.1 SEM 5.4 6.4 3.6 8.3 2.1 6.4 4.1 3.2 PROTECTION RATE* - 37% 71% 145 93% 57% 80% 89% *Defined by Equation 1 TABLE 2: STATISTICAL ANALYSIS OF DATA BY THE t-fest COMPARISON p value Control vs Fine Pumactant 0. 2123 Control vs Coarse Pumactant 0. 0052 Control vs propylene glycol 0.6916 Control vs DPPC paste 0.0002 Control vs DPPG paste 0.0639 Control vs DPPC : DPPG, 7: 3 0.0031 Control vs DPPC : DPPG, 1 : 1 0.0006 propylene glycol vs DPPC paste 0.0158 propylene glycol vs DPPG paste 0.2495 propylene glycol vs DPPC: DPPG, 7: 3 0.0544 propylene glycol vs DPPC: DPPG, 1: 1 0.0251 Fine ALEC vs Coarse Pumactant 0.1978 DPPC paste vs DPPG paste 0.1397 7: 3 paste vs 1: 1 paste 0.6094 DPPC paste vs 7: 3 paste 0.4193 DPPC paste vs 1 : 1 paste 0.1207 DPPG paste vs 7: 3 paste 0.3960 DPPG paste vs 1 : 1 paste 0.2098 DPPC paste vs Fine Pumactant 0.0277 DPPC paste vs Coarse Pumactant 0.1477 DPPG paste vs Fine Pumactant 0.5361 DPPG paste vs Coarse Pumactant 0.5832 7: 3 paste vs Fine Pumactant 0.1222 7: 3 paste vs Coarse Pumactant 0. 6517