Reduction of Post-Surgical Adhesion Formation with Tranilast

Journal of Surgical Research 141, 153–161 (2007) doi:10.1016/j.jss.2006.05.044

Reduction of Post-Surgical Adhesion Formation with Tranilast Kevin Cooper, Ph.D.,* Janel Young, B.S.,† Scott Wadsworth, Ph.D.,† Helen Cui,* Gere S. diZerega, M.D.,‡,1 and Kathleen E. Rodgers, Ph.D.‡ *Center for Biomaterials & Advanced Technologies, Medical Devices Group, A division of Ethicon, A Johnson & Johnson Company, Somerville, New Jersey; †Johnson & Johnson Pharmaceutical Research & Development, L.L.C., Raritan, New Jersey; ‡University of Southern California, Keck School of Medicine, Department of Obstetrics & Gynecology, Livingston Reproductive Biology Laboratories, Los Angeles, California Submitted for publication March 30, 2006

INTRODUCTION Background. Preclinical studies using the rabbit sidewall and double uterine horn models were used to assess time and dose response of tranilast delivered via subcutaneous pump, p.o., or as an intraperitoneal bolus in viscoelastic gels as well as an intraperitoneal biodegradable poly(p-dioxanone) fiber in reducing adhesions compared to vehicle controls. Materials and methods. New Zealand white rabbits underwent laparotomy followed by: 1) uterine horn abrasion and peripheral devascularization or 2) cecal abrasion and sidewall deperitonealization. Tranilast treatment using various vehicles and dosages was compared to vehicle alone versus no treatment. Animals were euthanized after 7 to 21 days. Adhesion formation was assessed by two independent observers. Results. There were reductions in adhesion formation when drug was delivered topically, but oral drug alone was not effective. When tranilast was given preoperatively, oral drug added to the adhesion reduction of intraperitoneal administered drug. Tranilast in a viscoelastic carrier as well as in a biodegradable fiber was effective at reducing adhesions in the double uterine horn model. The slow release of tranilast from a biodegradable rod produced overall the best results. There were no safety issues. Conclusion. Tranilast was effective in reducing adhesions when given in a variety of vehicles in different rabbit models of adhesion formation. Overall, the sustained intraperitoneal delivery of tranilast from biodegradable fibers was the most suitable for clinical testing. © 2007 Elsevier Inc. All rights reserved. Key Words: adhesion prevention; tranilast. 1 To whom correspondence and reprint requests should be addressed at Department of Obstetrics & Gynecology, University of Southern California, Keck School of Medicine, Livingston Reproductive Biology Laboratories, 1321 N. Mission Road, Los Angeles, CA 90033. E-mail: [email protected].

Adhesion prevention devices became available to gynecologic surgeons in 1989 with the FDA approval of Interceed Absorbable Adhesion Barrier (Gynecare, Somerville, NJ). Other site-specific barriers soon followed, including Preclude (WF Gore, Flagstaff, AZ) and Seprafilm Bioabsorbable Membrane (Genzyme, Cambridge, MA). These first generation adhesion prevention devices were principally used in laparotomy procedures [1–3]. FDA approved Intergel Adhesion Prevention Solution (LifeCore Biomedical, Chaska, MN) in 2001 for laparotomy use although it was later withdrawn [4, 5]. When Intergel was withdrawn from the market in 2003, the only instillate available in Europe for adhesion reduction was Adept (ML Laboratories, Leichester, United Kingdom) [6]. Development of site-specific adhesion prevention devices which could be easily delivered via laparoscopy led to initial clinical studies of SprayGel (Confluent, Waltham, MA), and more recently, Oxiplex/AP Gel (FzioMed, San Luis Obispo, CA) [7–10]. With the availability of these products, use of adhesion prevention adjuvants has become the standard of practice following a variety of conservative gynecologic surgical procedures [11– 14]. Nevertheless, adhesion formation remains the leading cause of failed surgical therapy in the peritoneal cavity. To further enhance effectiveness, new technologies are required especially in patients with severe pre-existing adhesions, stage IV endometriosis, active gynecologic infections, or extensive intraperitoneal carcinoma. Further benefit in adhesion reduction may require the use of bioactive additives or pharmaceuticals. A number of molecules have shown promise in preclinical studies of adhesion prevention [15]. Initially antifibrinolytics including tPA were shown to reduce adhesions at concentrations appropriate for clinical use.

153

0022-4804/07 $32.00 © 2007 Elsevier Inc. All rights reserved.

154

JOURNAL OF SURGICAL RESEARCH: VOL. 141, NO. 2, AUGUST 2007

Although anti-inflammatory drugs were shown to be promising in preclinical studies, initial clinical studies were disappointing. Paclitaxel-loaded crosslinked hyaluronic acid films [16], lexipafant [17], spermine combined with carbosymethylchitosan [18], and neurokinin 1 receptor antagonists [19], have all shown positive preclinical results and are the subjects of ongoing clinical studies. Leading challenges to general prophylactic use of a bioactive adhesion prevention technology are ease of use, general safety of the active therapeutic as well as cost. Over the last 8 years, tranilast, a drug with antifibrotic and anti-inflammatory activities, was evaluated in over 11,000 patients in a variety of clinical conditions [20, 21]. This drug inhibits TGF-␤ expression, a known contributor to post-surgical adhesion formation [22–26]. Previous studies by others have demonstrated reduction in adhesions in a rat model using p.o. administered concentrations within established human safety limits [27]. In this report, the effectiveness of tranilast in reducing adhesions is evaluated in established rabbit models. In addition, results of studies are presented that evaluate a delivery system that provides ease of use in laparoscopic surgery. MATERIALS AND METHODS All animal studies were conducted under NIH guidelines for the use of animals in research only after approval from an Institution Animal Care and Use Committee. Animals. Female New Zealand white rabbits, 2.4 to 2.7 kg, were used in these studies and quarantined for at least 2 days before use. The rabbits were housed on a 12:12 light:dark cycle with food and water available ad libitum. The rabbits were euthanized 7 to 21 days after the surgery by intramuscular anesthesia followed by intravenous EuthaSol.

Materials The tranilast formulations were supplied by The Center for Biomaterials & Advanced Technologies, Johnson & Johnson, Ethicon. In studies where the tranilast was delivered throughout the postsurgical interval, Alzet miniosmotic pumps (10 ␮L/hour, 2 mL, Model 2ML1) were purchased from Durect. Polyethylene tubing [Clay Adams polyethylene tubing PE-60 ID 0.76 mm (0.030“) OD 1.22 mm (0.048”] was purchased from VWR (Irvine, CA) to deliver the material from the subcutaneous pump to the site of injury. Because of the solubility of the tranilast, a complex vehicle was used for the pump studies. The vehicle for these studies consisted of 70% Polyethylene glycol 400, 20% Tween 80, 10% N, N-dimethylacetamide (DMAC). Once the efficacy of tranilast was established, other formulations were developed for placement in the abdomen at the end of surgery. These involved incorporation of tranilast in a sodium carboxymethyl cellulose gel or a polymer rod vehicle. In the studies where tranilast was delivered as a bolus formulation, a sodium carboxymethyl cellulose (NaCMC) gel was used to deliver the drug to the surgical site. A 3% (w/v) aqueous solution of NaCMC in phosphate buffer was prepared and sterilized via autoclaving. Tranilast was milled to a fine powder of 5 to 10 microns, cobalt irradiated and mixed with the NaCMC gel under aseptic conditions to form the bolus preparations. Sustained release formulations of tranilast were also prepared and used in the studies. Tranilast was encapsulated in fibers of poly(pdioxanone) (PDS) via a two-stage process in which the drug was melt compounded with the molten polymer and then extruded into 300

micron diameter fibers. The fibers were cut into 1 mm length rods, cobalt irradiated and mixed under aseptic conditions with the NaCMC gel described previously to form the sustained release preparations. Sutures used to close the peritoneum and skin were 3-0 Vicryl suture (Ethicon, Somerville, NJ).

Methods

Sidewall Model Rabbits were anesthetized with a mixture of 55 mg/kg ketamine hydrochloride and 5 mg/kg Rompum intramuscularly. Following preparation for sterile surgery, a midline laparotomy was performed. A polyethylene catheter was introduced into the peritoneal cavity and sutured to the sidewall with 5-0 Ethilon. In some studies, a single pump, filled with one of three dosage levels (0.625, 6.25, or 62.5 mg/mL) of tranilast or placebo delivered for 7 days starting with the day of surgery was placed in the subcutaneous space. In other studies, the pump was replaced after 7 or 14 days to allow more prolonged delivery. The cecum and bowel were exteriorized and digital pressure was exerted to create subserosal hemorrhages overall surfaces. The damaged intestine was then lightly abraded with 4“ ⫻ 4” four-ply sterile gauze until punctuate bleeding was observed. The cecum and bowel were then returned to their normal anatomical position. A 5 ⫻ 3 cm area of peritoneum and transversus abdominous muscle were removed on the right lateral abdominal wall. The catheter was attached to the pump and the midline muscle incision was closed around the catheter. The incision was closed in two layers with 3-0 Vicryl.

Necropsy General observations were conducted on each animal (appearance, behavior, clinical signs, symptoms, morbidity, mortality, and physical examinations) before assignment to study and on the day of necropsy. On the day of necropsy the following were recorded with particular attention to inflammation, granulomatous material and adhesion formation at the treatment site. All biocompatibility results were recorded, and the level of reaction was noted (none, mild, moderate, severe). Efficacy was assessed by evaluating adhesion formation at the site of sidewall injury. The sidewall injury area was examined to determine the percentage of the area involved in adhesions, and the tenacity of the adhesions was scored using the following system: 0 1 2 3

⫽ ⫽ ⫽ ⫽

no adhesions mild, easily dissectible adhesions moderate adhesions; non-dissectible, does not tear the organ dense adhesions; non-dissectable, tears organ when removed

The difference in adhesion formation between the control and treated groups were analyzed by Student’s t-test.

Double Uterine Horn Model Rabbits were anesthetized with a mixture of 55 mg/kg ketamine hydrochloride and 5 mg/kg Rompum intramuscularly. Following preparation for sterile surgery, a midline laparotomy was performed. The uterine horns were exteriorized and traumatized by abrasion of the serosal surface with gauze until punctate bleeding developed. Ischemia of both uterine horns was induced by removal of the collateral blood supply. The remaining blood supply to the uterine horns was the ascending branches of the utero-vaginal arterial supply of the myometrium. At the end of surgery, no treatment, placebo (NaCMC), placebo (NaCMC) containing tranilast was administered. The horns were then returned to their normal anatomical position and the midline sutured with 3-0 Vicryl. After 7 or 14 days, the rabbits were terminated and the percentage of the area of the horns adherent to various organs determined. The tenacity of the adhesions was scored using the following system:

155

COOPER ET AL.: REDUCTION OF POST-SURGICAL ADHESION FORMATION ⫽ ⫽ ⫽ ⫽

no adhesions mild, easily dissectable adhesions moderate adhesions; non-dissectable, does not tear the organ dense adhesions; non-dissectable, tears organ when removed

The rabbits were scored by two independent observers that blinded to the prior treatment of the animal. If there was disagreement as to the score to be assigned to an individual animal, the higher score was given.

Statistical Analysis The overall non-parametric scores were analyzed by rank order analysis and analysis of variance on the ranks. The percentage area of the horns involved in adhesions to the various organs was compared by analysis of variance. The incidence of adhesion free sites was evaluated by Fischer’s exact test.

In Vitro Release Kinetics of Tranilast-Containing Formulations The in vitro release of tranilast from the delivery system was studied for two groups using a liquid-liquid extraction technique followed by HPLC analysis. Analysis of the release kinetics from the tranilast in NaCMC gels and Tranilast-loaded PDO fibers in NaCMC gel was conducted in rat plasma to mimic physiological conditions. To determine the release from formulations, 1 mg tranilast-containing formulations were placed in a microcentrifuge tube containing 2 mL of rat plasma under aseptic conditions. All samples were maintained at 37°C up to 12 days. At each time point (1 h, 5 h, days 1, 2, 6, and 12), the tubes were centrifuged and the supernatant collected for tranilast analysis. Fresh medium was replaced at each time point. Percent cumulative release was determined by normalizing the cumulative release of tranilast at each time point with the total dose of tranilast over the course of 12 days.

RESULTS Sidewall Model (drug administered via pump)

After administration of placebo or tranilast via Alzet pump for 7 days and necropsy at 21 days, there were no signs of inflammation or granulomas. While there were reductions in adhesion formation at all doses of drug, the change was significant at the middle dose. The mean adhesion scores were: Placebo control: 100 ⫾ 0; 0.625 mg/mL tranilast: 82.9 ⫾ 8.1 (P ⫽ 0.078); 6.25 mg/mL tranilast: 67.1 ⫾ 7.1 (P ⬍ 0.001); 62.5 mg/mL tranilast: 78.6 ⫾ 9.6 (P ⫽ 0.065). There were also significant reductions in the tenacity of the adhesions formed at all doses of drug (by ANOVA on the ranks). For the placebo control rabbits, the tenacity of the adhesions was 3 for all rabbits. However, in rabbits treated with tranilast there was a reduction in the tenacity of adhesions in all animals (Fig. 1). In a second study, tranilast was delivered for 14 or 21 days in Alzet pumps. In some animals, there was inflammation noted both in placebo and tranilast-treated animals. As the response was present in the vehicle treated animals, the inflammatory response may be because of prolonged administration of the vehicle used to solubi-

3

2

1

100% Percentage With Tenacity Score

0 1 2 3

80%

60%

40%

20%

0% 0 n=6

0.625 6.25 n=7 n=7 Tranilast Dose (mg/ml)

62.5 n=7

FIG. 1. Rabbits underwent midline laparotomy with sidewall excision and bowel abrasion. Twenty-one days after surgery, rabbits euthanized and adhesions scored. The tenacity of adhesions (1–3) were scored. The percentage of rabbits with adhesions with that tenacity score is shown.

lize tranilast. Therefore, the efficacy results should be interpreted in light of the possible presence of vehicle associated inflammation. While there were reductions in adhesion formation at all doses of drug and both times, the change was significant at the high dose at 2 weeks and all doses at 3 weeks. The mean adhesion scores were: Placebo control: 100 ⫾ 0; 0.625 mg/mL tranilast 2 weeks: 75.1 ⫾ 14.5; 6.25 mg/mL tranilast 2 weeks: 78.8 ⫾ 12.0; 62.5 mg/mL tranilast 2 weeks: 72.5 ⫾ 9.8; 0.625 mg/mL tranilast 3 weeks: 48.6 ⫾ 8.8; 6.25 mg/mL tranilast 3 weeks: 77.1 ⫾ 12.1 and 62.5 mg/mL tranilast 3 weeks: 64.3 ⫾ 10.9. There were also significant reductions in the tenacity of the adhesions formed at all doses of drug (by ANOVA on the ranks). The score distribution can be found in Fig. 2. Sidewall Model (drug administered p.o. and via Alzet pump)

In a third study, tranilast was administered both p.o. and locally (5 days before surgery and/or throughout the postoperative period p.o. and 7 days after surgery locally). At necropsy, there were no clinical signs associated with the administration of tranilast. There were reductions in the area of adhesion formation in all groups that received local delivery of tranilast. Oral tranilast alone did not reduce the area of adhesion formation. The mean adhesion scores were: Placebo control: 100 ⫾ 0; 6.25 mg/mL tranilast locally: 72.9 ⫾ 11.7 (P ⫽ 0.039); pre-operative and post-operative tranilast p.o.: 100 ⫾ 0; pre-operative tranilast p.o. (60 mg/kg/day) with local placebo: 100 ⫾ 0 (P ⫽ 1.000) and [re-operative tranilast p.o. (60 mg/kg/day) with local 6.25 mg/mL tranilast : 46.7 ⫾ 14.1 (P ⫽ 0.002). There was a further reduction when tranilast was given pre-

156

JOURNAL OF SURGICAL RESEARCH: VOL. 141, NO. 2, AUGUST 2007 3

1 2

1 2

0

Panel

A

50

40 Rank of Scores

80%

60%

30

20

40% 10

20%

0

0 n=8

0.625 n=7

6.25 n=8 2 Weeks

62.5 n=8

0.625 n=7

6.25 n=7 3 Weeks

62.5 n=7

FIG. 2. Rabbits underwent midline laparotomy with sidewall excision and bowel abrasion. Twenty-one days after surgery, rabbits euthanized and adhesions scored. The tenacity of adhesions (1–3) were scored. The percentage of rabbits with adhesions with that tenacity score is shown.

operatively p.o. and locally for 7 days after surgery compared with tranilast given locally only, but this difference was not significant (P ⫽ 0.177). There were also significant reductions in the tenacity of the adhesions formed in all groups of animals that received tranilast (by analysis of variance on the ranks) (Fig. 3 for distribution of scores). Double Uterine Horn Model (bolus delivery of drug administered at end of surgery)

In the first study, tranilast was administered at a volume of 15 mL per rabbit in NaCMC gel at doses of 1.5 to 15 mg/mL tranilast. No intraperitoneal inflam-

Panel

B

Control n=7

NaCMC n=7

1.5 mg/mL n=7

Control n=7

NaCMC 1.5 mg/mL n=7 n=7 Tranilast

15 mg/mL n=7

Control n=7

NaCMC n= 7

1.5 mg/mL n=7

Tranilast

15 mg/mL n=7

120 100

% with Score 1.5 or less

0%

80 60 40 20 0

Panel

C 70 60 % Sites Adhesion Free

Percentage With Tenacity Score

100%

50 40 30 20 10

3

2

1

0

Percentage with Tenacity Score

100%

.75 mg/mL n=7

Tranilast

FIG. 4. Rabbits underwent midline laparotomy with double uterine horn abrasion and devascularization. At the end of surgery, the designated formulations were administered. At necropsy, adhesions were scored as stated in Methods section. In panel A, the rank order of overall score is shown. In panel B, percentage of animals with an overall score of 1.5 or less is shown. In panel C, the percentage of sites scored that were free of adhesions is shown.

80%

60%

40%

20%

0% Placebo

Topical Tranilast

n=7

n=7

Oral Tranilast Pre/Post n=6

Oral Oral (Pre) Tranilast Tranilast (Pre) Topical Post n=6 n=6

FIG. 3. Rabbits underwent midline laparotomy with sidewall excision and bowel abrasion. Twenty-one days after surgery, rabbits were euthanized and adhesions scored. The tenacity of adhesions (1–3) were scored. The percentage of rabbits with adhesions with that tenacity score is shown.

mation, material or granuloma was noted at necropsy. The administration of 15 mL NaCMC reduced the overall adhesion score and seemed to have the most benefit at non-surgical sites (adhesions to bowel and bladder) (Fig. 4A and 4B). The decrease in the incidence of adhesion free sites was not statistically significant (Fig. 4C). Addition of tranilast, at both concentrations, to the NaCMC further reduced adhesion formation both at surgical and non-surgical sites. The reduction in the incidence of adhesions and the overall score was significant compared both with surgical and placebo con-

157

COOPER ET AL.: REDUCTION OF POST-SURGICAL ADHESION FORMATION

60

Panel 50

30

20

NaCMC n=7

B 120

NaCMC n=7

.3 mg/mL n=7 Tranilast

Control n=7

NaCMC n=7

.3 mg/mL n=7 Tranilast

1.5 mg/mL n=7

Intergel n=7

.3 mg/mL n=7 Tranilast

1.5 mg/mL n=7

Intergel n=7

B 100

.75 mg/mL n=6 Tranilast

1.5 mg/mL n=7

Intergel n=7

80

60

40

20

1.5 mg/mL n=7

Panel

100

C 50 % Sites Adhesion Free

% with Score 1.5 or less

Control n=7

0 Control n=7

80 60 40

40

30

20

10

20 0

0 Control n=7

NaCMC n=7

.75 mg/mL n=6 Tranilast

1.5 mg/mL n=7

C 50 % Sites Adhesion Free

30

0

0

Panel

40

10

10

Panel

50

20

40 Rank of Scores

A 70

% with Score of 1.5 or Less

A

Panel

Panel

Rank of Scores

trols (Fig. 4). In a second study, the doses of tranilast tested were 0.75 and 1.5 mg/mL. The administration of 15 mL NaCMC reduced the overall adhesion score and incidence of adhesions, but not significantly. The vehicle seemed to have the most benefit at non-surgical sites (adhesions to bowel and bladder). Addition of tranilast to the NaCMC further reduced adhesion formation both at surgical and non-surgical sites. The reduction in the incidence of adhesions and the overall score was significant compared both with surgical and placebo controls at the higher concentration (Fig. 5). In a third study, tranilast (0.3 and 1.5 mg/mL tranilast) in

Control n=7

NaCMC n=7

FIG. 6. Rabbits underwent midline laparotomy with double uterine horn abrasion and devascularization. At the end of surgery, the designated formulations were administered. At necropsy, adhesions were scored as stated in Methods section. In panel A, the rank order of overall score is shown. In panel B, percentage of animals with an overall score of 1.5 or less is shown. In panel C, the percentage of sites scored that were free of adhesions is shown.

40

30

20

10

0 Control n=7

NaCMC n=7

.75 mg/mL n=6

Tranilast

1.5 mg/mL n=7

FIG. 5. Rabbits underwent midline laparotomy with double uterine horn abrasion and devascularization. At the end of surgery, the designated formulations were administered. At necropsy, adhesions were scored as stated in Methods section. In panel A, the rank order of overall score is shown. In panel B, percentage of animals with an overall score of 1.5 or less is shown. In panel C, the percentage of sites scored that were free of adhesions is shown.

NaCMC was compared with Intergel (15 mL volume). Addition of 1.5 mg/mL tranilast to the NaCMC further reduced adhesion formation both at surgical and nonsurgical sites (predominately in the non-surgical sites). The reduction in the incidence of adhesions and the overall score was significant compared both with surgical and placebo controls. The lower dose of tranilast did not significantly reduce adhesions by either measure (Fig. 6). Addition of both volumes of Intergel reduced adhesion formation both at surgical and nonsurgical sites (predominately in the non-surgical sites).

158

JOURNAL OF SURGICAL RESEARCH: VOL. 141, NO. 2, AUGUST 2007

The reduction in the incidence of adhesions and the overall score was significant compared with surgical controls (Fig. 6a and 6b). Double Uterine Horn Model (sustained release drug delivery administered at end of surgery)

A slow release system was developed for the delivery of tranilast in the form of PDS fibers filled with tranilast. The efficacy of tranilast delivered in these fibers was evaluated 14 days after surgery. Fibers were seen in the peritoneum (on the parietal peritoneum, retroperitoneal space, bowel, bladder, and horns) in the majority

Panel

A

40

Rank Order

30

20

10

0

Panel

Control

NaCMC

n=7

n=7

Control

NaCMC

Tranilast Tranilast Tranilast in Fibers in Fibers in in NaCMC and NaCMC NaCMC n=7 n=7 n=7

Fibers in in NaCMC w/o Drug n=7

Tranilast Tranilast Tranilast in Fibers in Fibers in in NaCMC and NaCMC NaCMC n=7 n=7 n=7

Fibers in in NaCMC w/o Drug n=7

B

% with Score 1.5 or Less

120 100 80 60 40 20 0

n=7 Panel

n=7

C

% Sites Adhesion Free

50 40 30 20 10 0 Control

n=7

NaCMC

n=7

Tranilast

Tranilast

Tranilast

in Fibers in Fibers in in NaCMC and NaCMC NaCMC n=7 n=7 n=7

Fibers in

in NaCMC w/o Drug n=7

FIG. 7. Rabbits underwent midline laparotomy with double uterine horn abrasion and devascularization. At the end of surgery, the designated formulations were administered. At necropsy, adhesions were scored as stated in Methods section. In panel A, the rank order of overall score is shown. In panel B, percentage of animals with an overall score of 1.5 or less is shown. In panel C, the percentage of sites scored that were free of adhesions is shown.

of animals. The administration of 15 mL NaCMC reduced the overall adhesion score and seemed to have the most benefit at non-surgical sites (adhesions to bowel and bladder) (Fig. 7). The decrease in the incidence of adhesion free sites was statistically significant in all groups compared to surgical controls. NaCMC or PDS did not decrease the overall adhesion scores (Fig. 7A). Addition of tranilast to the formulations further reduced adhesion formation both at surgical and nonsurgical sites (Fig. 7B). The reduction in the incidence of adhesions was significant for all formulations containing tranilast compared to their respective placebo controls (Fig. 7C). Admixing tranilast with the NaCMC in addition to tranilast in the PDS fibers did not significantly further reduce the incidence of adhesions (P ⫽ 0.428). All formulations containing tranilast had significant decreases in their overall adhesion score compared to surgical control (P ⬍ 0.001), NaCMC control (P ⬍ 0.05) or PDS control (P ⬍ 0.05). There were no significant differences between the tranilast containing solutions with regards to overall score. However, the lowest ranks were observed with tranilast in both the PDS and NaCMC (not significant). In a final study, various doses of tranilast in PDS slow release rods were evaluated. All animals survived until necropsy and no intraperitoneal inflammation or granuloma was noted at necropsy. Fibers were seen in the peritoneum (on the parietal peritoneum, retroperitoneal space, bowel, bladder, and horns) in the majority of animals that received PDS in their formulation. The scores in the surgical control animals were consistent with previous studies. The administration of 15 mL NaCMC slightly reduced the overall adhesion score (Fig. 8A). Addition of tranilast to the formulations further reduced adhesion formation both at surgical and nonsurgical sites. All formulations containing tranilast reduced the incidence of adhesions compared with surgical controls (Fig. 8B). The reduction in the incidence of adhesions was significant for formulations containing tranilast in NaCMC with and without 1X PDS compared with their placebo controls. Addition of 2⫻ or 4⫻ PDS reduced the incidence of adhesions compared with placebo, but not significantly. All formulations containing tranilast (except that containing 4X PDS) had significant decreases in their overall adhesion score compared to surgical control (P ⬍ 0.01) (Fig. 8C). Only the formulation containing tranilast in NaCMC with 1X PDS had significantly lower overall scores than the NaCMC control (P ⬍ 0.001). The only significant difference between the tranilast containing solutions with regards to overall score was between that containing 1X and 4X PDS. As the efficacy of the formulation decreased with increasing PDO rods, it may be that increasing the amount of foreign material reduces efficacy.

159

COOPER ET AL.: REDUCTION OF POST-SURGICAL ADHESION FORMATION

Panel

A

Control NaCMC

40

released by day 2. In the fiber formulations, the burst release was reduced and delayed with the majority of the tranilast released by day 12. Hence, incorporating tranilast into PDO fiber delays the burst effects and gives a sustained release for approximately 2 weeks. This result corresponds well with the results of the pump studies that showed that a sustained release improves the efficacy of the drug.

NaCMC Containing Tranilast NaCMC Containing Tranilast with 1x PDS Tranilast NaCMC Containing Tranilast with 2x PDS Tranilast NaCMC Containing Tranilast with 4x PDS Tranilast

Rank Order

30

20

DISCUSSION

10

0

Panel

B

n=6

n=7

n=7

Tranilast

n=7

n=7

n=7

Control NaCMC

NaCMC Containing Tranilast NaCMC Containing Tranilast with 1x PDS Tranilast NaCMC Containing Tranilast with 2x PDS Tranilast NaCMC Containing Tranilast with 4x PDS Tranilast

n=6

n=7

Control NaCMC

NaCMC Containing Tranilast NaCMC Containing Tranilast with 1x PDS Tranilast NaCMC Containing Tranilast with 2x PDS Tranilast NaCMC Containing Tranilast with 4x PDS Tranilast

1.4 140

% 1.5 or less

1.2 120 1 100 0.8 80 0.6 60 0.4 40 0.2 20 0

Panel

C

% Adhesion Free Sites

60

n=7

Tranilast

n=7

n=7

n=7

50 40 30 20 10 0

n=6

n=7

n=7

Tranilast

n=7

n=7

n=7

FIG. 8. Rabbits underwent midline laparotomy with double uterine horn abrasion and devascularization. At the end of surgery, the designated formulations were administered. At necropsy, adhesions were scored as stated ion Methods section. In panel A, the rank order of overall score is shown. In panel B, percentage of animals with an overall score of 1.5 or less is shown. In panel C, the percentage of sites scored that were free of adhesions is shown.

The purpose of the studies reported here was to evaluate the use of tranilast in reduction of postoperative peritoneal adhesions. Taken together, these studies show that tranilast reduces adhesion in a variety of preclinical models. Tranilast (N-(3=4=-dimethoxycinnamoyl) anthranilic acid), has been used p.o. to treat allergy and asthma in Japan for over 20 years. There are many pharmacological actions of tranilast that theoretically should contribute to adhesion reduction. Tranilast was originally identified as a mast cell stabilizer, capable of inhibiting mast cell degranulation in response to various stimuli [28]. Mast cells are important early mediators of intraperitoneal inflammation after surgery [29, 30]; mast cell stabilization may be an important part of the mechanism of action of tranilast in adhesion prevention. Tranilast was subsequently found to inhibit cell proliferation [23, 24, 31, 32], including vascular smooth muscle cells, which lead to its preclinical [25] and clinical testing as an anti-restenosis agent. In addition to its anti-proliferative effects, tranilast can inhibit various inflammatory responses [33–35], TGF-␤ production [26], and collagen production [36], all of which would be expected to negatively impact adhesion formation [37, 38]. This same reduction in TGF-␤ and type 1 collagen was demonstrated in the kidneys of rats following streptozotocin induced diabetic nephropathy as well as in the heart of mRen-2 diabetic rats [39, 40]. Tranilast dependent reduction in fibrosis was also found in mice

100.00% 90.00% 80.00%

Drug Delivery Kinetics

70.00% 60.00%

With the drug concentration and time of delivery parameters evaluated, clinically appropriate devices for drug delivery were tested. The in vitro release of tranilast from NaCMC gel and tranilast-containing fiber in NaCMC gel into rat plasma was determined using liquid-liquid extraction followed by HPLC analysis. The releasate was analyzed at 1 h, 5 h, 1, 2, 6, and 12 days, and the results are shown in Fig. 9. Within 1 day, the admixed tranilast demonstrated a high burst release (67.9 ⫾ 7.9%) with the majority of the tranilast

50.00% 40.00% 30.00% 20.00% 10.00% 0.00% 0

50

100

150

200

250

300

350

Time (hrs)

FIG. 9. Tranilast release in rat plasma from admixed in NaCMC gel and from tranilast-laden fibers in NaCMC gel. (Color version of figure is available online.)

160

JOURNAL OF SURGICAL RESEARCH: VOL. 141, NO. 2, AUGUST 2007

following bleomycin treatment for the induction of pulmonary fibrosis [41]. Multiple studies have shown that tranilast reduces TGF-␤- induced fibrosis and excessive formation of extracellular matrix in a variety of in vitro models [36, 42]. The current studies show that tranilast effectively reduced adhesion formation in two challenging adhesion formation models. This reduction was dose-responsive when given via intraperitoneal pumps. The optimal biological dose, established in pumps, was shown to be effective when the drug was loaded into NaCMC gel or polymer rods, amenable to clinical use via laparoscopy. Previously, Adachi et al., reported the reduction of post-surgical peritoneal adhesions in male Donryu rats with tranilast [27]. Tranilast was administered p.o. pre- and post-operatively. Serum levels of bFGF and TGF-␤1 were also reduced. The current study confirms that p.o. administered tranilast can decrease adhesion formation, but only in the presence of concomitant locally delivered drug. Therefore, in contrast to the results of Adachi et al., our data suggest that systemic tranilast alone is not effective when delivered p.o. Additionally, the doses of tranilast used in the current studies were at concentrations well within previously established limits of clinical safety [20, 21]. Evaluation of a pharmaceutical for intraperitoneal adhesion prevention requires a delivery system for screening the drug and dose after surgical injury. The initial studies of tranilast were performed using an Alzet pump attached to an indwelling catheter sutured over the sidewall injury. Subcutaneous placement of the pump allows for easy access to inject drug or occlude the catheter for time response studies without disrupting the intraperitoneal cavity [15]. This approach allows for evaluation of the drug effect as well as vehicle selection. Preloading of tissues with drug was evaluated by giving oral tranilast daily for 5 days before surgery. Tranilast treatment continued by both oral and Alzet pump routes an additional 7 days after surgery. Preoperative tranilast together with postoperative treatment by both oral and intraperitoneal pump provided additional reduction in adhesion formation. These studies of surgical injury to the sidewall were followed by more challenging studies in the double uterine horn model. In these studies, tranilast was formulated in a sodium carboxymethylcellulose gel at an osmolality and pH appropriate for intraperitoneal administration. Various concentrations of tranilast administered via NaCMC gel confirmed effects seen by Alzet pump at similar drug concentrations. Dose response effects of tranilast were confirmed in two different models of adhesion formation and by two different routes of drug delivery. The in vitro release of tranilast from. Tranilastcontaining fiber provided slow release of tranilast over the appropriate postoperative time. In fiber formula-

tions, burst release was reduced and delayed with the majority of tranilast released by day 12. Hence, incorporating tranilast into PDS fiber delays the burst effects and gives a sustained release for approximately 2 weeks. This result corresponds well with the results of the pump studies that showed a sustained release improves the efficacy of the drug. In conclusion, tranilast, either an in NaCMC gel or an extended release PDS system was effective in reducing postoperative peritoneal adhesions. Further studies of dose scheduling, including preoperative dosing, may demonstrate additional benefits. In addition, longer term studies of tranilast in biocompatible drug-delivery system may demonstrate further adhesion reduction benefits to a wide variety of surgical therapies. REFERENCES 1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11. 12. 13. 14.

Interceed (TC7) Adhesion Barrier Group. Prevention of post surgical adhesions by intercede (TC7), an absorbable adhesion barrier: A prospective randomized multicenter clinical study. Fertil Steril 1989;51:933. Diamond MP. Reduction of adhesions after uterine myomectomy by Seprafilm (HAL-F): A blinded, prospective, randomized, multicenter clinical study. Seprafilm Adhesion Study Group. Fertil Steril 1996;66:904. Surgical Membrane Study Group. Prophylaxis of pelvic sidewall adhesion formation with Gore-Tex surgical membrane: A multicenter clinical investigation. Fertil Steril 1992;57:921. Johns DB, Keyport GM, Hoehler F, et al. Intergel adhesions prevention study group: Reduction of PDStsurgical adhesions with Intergel adhesion prevention solution. A multicenter study of safety and efficacy after conservative gynecologic surgery. Fertil Steril 2001;76:595. Lundorff P, Donnez J, Korell M, et al. Clinical evaluation of a viscoelastic gel for reduction of adhesions following gynaecological surgery by laparoscopy in Europe. Fertil Steril 2005;20:514. diZerega GS, Verco SJS, Young P, et al. A randomized, controlled pilot study of the safety and efficacy of 4% icodextrin solution in the reduction of adhesions following laparoscopic gynaecological surgery. Hum Reprod 2002;17:1031. Johns DA, Ferland R, Dunn RI. Initial feasibility study of a sprayable hydrogel adhesion barrier system in patients undergoing laparoscopic ovarian surgery. J Am Assoc Gynecol Laparosc 2003;10:339. Lundorff P, Donnez J, Korell M, et al. Clinical evaluation of a viscoelastic gel for reduction of adhesions following gynaecological surgery by laparoscopy in Europe. Fertil Steril 2005;20:514. Young P, Johns A, Templeman C, et al. Reduction of PDStoperative adhesions following laparoscopic gynecological surgery with Oxiplex/AP Gel: A pilot study. Fertil Steril 2005;84:1450. Mettler L, Audebert A, Lehmann W, et al. Prospective clinical trial of SprayGel as a barrier to adhesion formation: An interim analysis. J Am Assoc Gynecol Laparosc 2003;10:339. Tulandi T. Introduction-prevention of adhesion formation: The journey continues. Hum Reprod Update 2001;7:545. Trew G. Consensus in adhesion reduction management. Obstet Gynecol 2004;6:1. Al-Musawi D, Thompson JN. Adhesion prevention: State of the art. Gynecol Endoscopy 2001;10:123. Canis M, Botchorishvili R, Tambiurro SM. Adhesion prevention in the surgical treatment of pelvic endometriosis. Gynecol Endoscopy 2001;10:99.

COOPER ET AL.: REDUCTION OF POST-SURGICAL ADHESION FORMATION 15.

16.

17. 18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

Rodgers KE, diZerega GS. Developing Pharmacologic agents for adhesions prevention. In: diZerega GS, Ed. Peritoneal surgery. New York: Springer-Verlag, 2000:441. Jackson JK, Skinner KC, Burgess L, et al. Paclitaxed-loaded crosslinked hyaluronic acid films fro the prevention of PDStsurgical adhesions. Pharm Res 2002;19:411. Ozugn H, Cevikel MH, Kozaci LD, et al. Lexipafant inhibits PDStsurgical adhesion formation. J Surg Res 2002;103:141. Duran B, Ak D, Cetin A, et al. Reduction of PDStoperative adhesions by N, O-Carboxymethylchitosan and spermine NONOate in rats. Esp Anim 2003;52:267. Reed KL, Brent-Fruin A, Gower AC, et al. A neurokinin 1 receptor antagonist decreases PDStoperative peritoneal adhesion formation and increases peritoneal fibrinolytic activity. PNAS 2004;101:9115. Holmes DR Jr., Savage M, LaBlanche JM, et al. Results of Prevention of REStenosis with Tranilast and its Outcomes (PRESTO) trial. Circulation 2002;106:1243. Garot P, Lefevre T, Savage M, et al. Nine-month outcome of patients treated by percutaneous coronary interventions for bifurcation lesions in the recent era: A report from the Prevention of Restenosis with Tranilast and its Outcomes (Presto) trial. J Am Coll Cardiol 2005;46:606. Topley N. Cytokine responses to peritoneal inflammation. In: diZerega GS, Ed. Peritoneal surgery. New York: SpringerVerlag, 2000:75. Yasukawa T, Kimura H, Dong J, et al. Effect of tranilast on proliferation, collagen gel contraction, and transforming growth factor beta secretion of retinal pigment epithelial cells and fibroblasts. Ophthalmic Res 2002;34:206. Miyazawa K, Kikuchi S, Fukuyama J, et al. Inhibition of PDGF- and TGF-beta 1-induced collagen synthesis, migration and proliferation by tranilast in vascular smooth muscle cells from spontaneously hypertensive rats. Atherosclerosis 1995; 118:213. Ward MR, Agrotis A, Kanellakis P, et al. Tranilast prevents activation of transforming growth factor-␤ system, leukocyte accumulation, and neointimal growth in porcine coronary arteries after stenting. Arterioscler Thromb Vasc Biol 2002;22:940. Ward MR, Sasahara T, Agrotis A, et al. Inhibitory effects of tranilast on expression of transforming growth factor-beta isoforms and receptors in injured arteries. Atherosclerosis 1998; 137:267. Adachi S, Maruyama T, Kondo T, et al. The prevention of PDStoperative intraperitoneal adhesions by tranilast: N-(3=,4=dimethoxycinnamoyl) anthranilic acid. Surg Today 1999;29:51. Koda A, Kurashina Y, Nakazawa M. The inhibition mechanism of histamine release by N-(3,4-dimethoxycinnamoyl) anthranilic acid. Int Arch Allergy Appl Immunol 1985;77:244.

29. 30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

161

Liebman SM, Langer JC, Marshall JS, et al. Role of mast cells in peritoneal adhesion formation. Am J Surg 1993;165:127. Xu X, Rivkind A, Pappo O, et al. Role of mast cells and myofibroblasts in human peritoneal adhesion formation. Ann Surg 2003;236:593. Fukuyama J, Miyazawa K, Hamano S, et al. Inhibitory effects of tranilast on proliferation, migration, and collagen synthesis of human vascular smooth muscle cells. Can J Physiol Pharmacol 1996;74:80. Shime H, Kariya M, Orii A, et al. Tranilast inhibits the proliferation of uterine leiomyoma cells in vitro through G1 arrest associated with the induction of p21waf1 and p53. J Clin Endocrinol Metabol 2002;87:5610. Inoue H, Ohshima H, Kono H, et al. Suppressive effects of tranilast on the expression of inducible cyclooxygenase (COX2) in interleukin-1␤-stimulated fibroblasts. Biochem Pharmacol 1997;53:1941. Suzawa H, Kikuchi S, Ichikawa K, et al. Inhibitory action of tranilast, an anti-allergic drug, on the release of cytokines and PGE2 from human monocytes-macrophages. Jpn J Pharmacol 1992a;60:85. Chikaraishi A, Hirahashi J, Takase O, et al. Tranilast inhibits interleukin-1␤-induced monocyte chemoattractant protein-1 expression in rat mesangial cells. Eur J Pharmacol 2001;427: 151. Suzawa H, Kikuchi S, Arai N, et al. The mechanism involved in the ingibitory action of tranilast on collagen biosynthesis of keloid fibroblasts. Jpn J Pharmacol 1992;60:91. diZerega GS. Peritoneum, peritoneal healing, and adhesion formation. In: diZerega GS, Ed. Peritoneal surgery. New York: Springer-Verlag, 2000:3. Yamada H, Tajima S, Nishikawa T, et al. Tranilast, a selective inhibitor of collagen synthesis in human skin fibroblasts. J Biochem 1994;116:892. Akahori H, Ota T, Torita M, et al. Tranilast prevents the progression of experimental diabetic nephropathy through suppression of enhanced Extracellular matrix gene expression. Pharmacol Exp Ther 2005;27:147. Martin J, Kelly DJ, Mifsud SA, et al. Tranilast attenuates cardiac matrix dePDSition in experimental diabetes: Role of transforming growth factor-beta. Cardiovasc Res 2005;65:694. Mori H, Tanaka H, Kawada K, et al. Suppressive effects of tranilast on pulmonary fibrosis and activation of alveolar macrophages in mice treated with bleomycin: Role of alveolar macrophages in the fibrosis. Jpn J Pharmacol 1995;67:279. Isaji M, Aruga N, Naito J, et al. Inhibition by tranilast of collagen accumulation in hypersensitive granulomatous inflammation in vivo and of morphological changes and functions of fibroblasts in vitro. Life Sci 1994;55:PL287.