Evaluation of a sprayable polyethylene glycol adhesion barrier in a porcine efficacy model

R. Ferland1,3, D. Mulani2 and P.K. Campbell2

1 Women and Infants Hospital, Providence, RI 02905 and 2 Confluent Surgical Inc., 101A First Avenue, Waltham, MA 02451, USA


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
BACKGROUND: The formation of adhesions following pelvic surgery remains one of the leading causes of infertility, small bowel obstruction and re-operation for pelvic pain. A novel hydrophilic polyethylene glycol based adhesion barrier (SprayGel) is formed by simultaneously spraying two liquid precursors onto surgical sites. The liquids polymerize to form a gel that effectively coats and adheres to tissue. After about 5 days, the hydrogel layer is absorbed and subsequently undergoes renal clearance. It is believed that the presence of such a barrier would inhibit the formation of adhesions following surgical insult. METHODS: A porcine adhesion model was developed wherein bilateral uterine horn transection and re-anastomosis, along with peritoneal side wall excision was performed via laparotomy. In each animal (n = 10, including the pilot study) one pelvic side wall was treated with adhesion barrier, while the contralateral side remained untreated. RESULTS: At second look laparoscopy, 90% of the untreated sites had adhesions, compared with 30% of the treated sites (P = 0.006). Also observed were statistically significant reductions in the adhesion extent (P = 0.029) and adhesion severity scores (P = 0.023) at the treated sites. However, if the pilot study was excluded (n = 8) the differences obtained were no longer significant. CONCLUSIONS: Polyethylene glycol (SprayGel) merits further investigation as an effective barrier to the formation of post-operative adhesions in this porcine model.

Key words: adhesion prevention/post-surgical adhesions/large animal models/laparoscopy/adhesion barriers


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Pelvic surgery often causes unavoidable tissue injury that can lead to the formation of post-surgical adhesions in more than 50% of patients (Trimbos-Kemper et al., 1985Go; Golan et al., 1995Go). Such injuries include mechanical trauma from retractors and tissue handling, ischaemia at suture sites and after electrocautery use, foreign bodies, tissue desiccation and infection (Stangel et al., 1984). These stimuli of adhesion formation occur in both open and laparoscopic approaches.

Post-operative adhesions are a common cause of small bowel obstruction and re-operation for pelvic pain, and adhesions involving the ovaries or Fallopian tubes are responsible for 15–20% of female infertility cases (Ray et al., 1998Go). In addition to these adverse clinical sequelae, the economic impact in the USA in 1994 was significant with a direct cost of $1.33 billion for all hospitalizations during which adhesiolysis was performed, based on Ray's analysis (Ray et al., 1998Go). Other studies have shown that of surgical patients, 35% were readmitted at least once for problems directly or possibly related to adhesions over a 10 year period (Ellis et al., 1999Go; Lower et al., 2000Go).

Although the ultimate solution to this problem will probably result from an increased understanding of the humoral agents and cellular events that control adhesion formation, current clinical needs could be met by an effective adhesion barrier that is easy to apply in both open and laparoscopic procedures.

The majority of models developed to study adhesion barriers utilize small animals, such as rats (Golan et al., 1995Go; Hellebrekers et al., 2000Go), mice (Haney and Doty, 1992Go) and rabbits (Marana et al., 1997Go). These models use a variety of means to mimic surgical injury, such as abrasion or electrocautery applied to a range of organs including uterine horns (Golan et al., 1995Go), the caecum, ovaries (Marana et al., 1997Go) and the pelvic sidewall to create a nidus for adhesion formation. These models have contributed greatly to our knowledge of adhesion formation and prevention. However, due to differences in scale between humans and these models, they are limited in their ability to predict clinical success.

Two studies have previously used porcine adhesion models (Montz et al., 1993aGo; Christoforoni et al., 1996Go), and in this study a porcine adhesion model has been developed with the intent to better mimic conditions involved in pelvic surgery. This model involves a clinically relevant surgical injury, in a species with organ size and weights similar to man. This model was used to evaluate the efficacy of a new synthetic absorbable polyethylene glycol product, SprayGelTM Adhesion Barrier (Confluent Surgical Inc., Waltham, MA, USA).


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
All surgical procedures were conducted in accordance with the regulations and approval of the Rhode Island Hospital Animal Care and Use Committee.

Animal Preparation and Operative Procedure
10 female virgin hogs weighing 50–75 lbs were received and acclimatized for a minimum of 2 days prior to surgery. They were induced with a combination of tiletamine-zolazepam (5 mg/kg), xylazine (2 mg/kg) and atropine (0.05 mg/kg) administered i.m. Pre-operatively, the subjects received 1 gm cefazolin i.v. Following induction of the general anaesthetic the animals were maintained on a mix of isoflurane and oxygen inhalation for anaesthesia for the duration of the procedure. The abdominal region was shaved, scrubbed and draped in preparation for sterile surgery.

The celiotomy was created via a single, midline abdominal incision from the umbilicus to the symphysis pubis. The subcutaneous tissue and fascia were divided using electrocautery (Valley Lab Force 2, 35 watts cutting current; Valley Labs, Boulder, CO, USA). The subjects were placed in a Trendelenberg position, and the bladder was aspirated by cystostomy with electrocautery and wall suction or 18-gauge needle. Dry surgical gauze, towels and retractors were used to obtain adequate exposure to the pelvic side wall during the injury process. Both uterine horns were sharply transected at their midpoint after coagulation with monopolar electrocautery (25 watts coagulating current), and the transected ends of each were then re-anastomosed (end-to-end) using two interrupted sutures (3–0 Vicryl; Ethicon, Sommerville, NJ, USA). The parietal peritoneum of the pelvic side wall opposed to each uterine horn was then sharply excised from the analogue of the round ligament to the infundibulopelvic ligament to expose an area about 5x4 cm on the pelvic sidewall. Monopolar electrocautery (25 watts coagulating current) was used to obtain haemostasis where needed. One subject required suture ligation of a large bleeder of the pelvic sidewall. Following bilateral peritoneal excision and uterine horn anastomoses, both sidewalls and uterine horns were irrigated with saline to ensure adequate haemostasis and reduce tissue drying. Figure 1Go shows an intraoperative view of a completed peritoneal injury, just prior to uterine horn transection and anastomosis.



View larger version (88K):
[in this window]
[in a new window]
 
Figure 1. Intraoperative view of a completed peritoneal injury, just prior to uterine horn transection and anastomosis.

 
Thereafter, a coin toss was used to randomly assign the left or right pelvic side wall to treatment with adhesion barrier, which was then applied to only the assigned side wall, with no application to the uterine horn. Following application to the assigned sidewall, the hydrogel barrier was rinsed with saline to ensure a moist, lubricious surface. The contralateral side received no further treatment and served as the internal control. After appropriate treatment was performed, the laparotomy was closed in layers with continuous braided polyester (O Ethibond; Ethicon). After closure, the abdominal incisions were injected with local anaesthetic (2% lidocaine) for analgesia. Animals were returned to their cages where they received food and water ad lib and i.m. pain medications, as deemed necessary based on their behaviour, with buprenorphine 0.01 mg/kg.

Application of adhesion barrier
The SprayGel Adhesion Barrier (Confluent Surgical) consists of two synthetic liquid precursors that, when mixed together, rapidly cross-link to form a solid, absorbable biocompatible hydrogel in situ. No external energy sources are required for polymerization, which is substantially completed within a few seconds with no heat evolution.

Both precursor solutions contain upwards of 90% water. The first precursor solution contains a modified polyethylene glycol (PEG) polymer with terminal electrophilic ester groups while the other precursor solution contains PEG that has nucleophilic amine end-groups. This second precursor solution also contains methylene blue, a colourant that is added to the formulation to facilitate visualization of the hydrogel. The SprayGel barrier is formulated to remain adherent to the site of application for approximately five days. At approximately that time the barrier breaks down by the process of hydrolysis, and the liberated water-soluble PEG molecules (<20 KDa) are absorbed and undergo renal clearance (Yamaoka et al., 1993Go). PEG molecules of this size have been shown to have a clearance half-life of about 15 min in mice (Yamaoka et al., 1993Go).

SprayGel has passed a complete battery of tests including cytotoxicity, genotoxicity, haemolytic potential, sensitization and irritation. It does not affect wound healing or potentiate infections, and has been shown to be non-toxic at 30 times the expected human dose.

After the completion of injuries and randomization, the surgeon applied the adhesion barrier only to the pelvic side wall injury site assigned to the treatment group. No material was applied to the uterine horn. Figure 2Go shows a typical application at a treated site. An air-assisted sprayer (Confluent Surgical), shown in Figure 3Go, was used to carry out the deposition. The barrier was applied to achieve a thickness so that fine tissue structures under the barrier such as small blood vessels or muscle fibres became difficult to visualize due to the methylene blue colourant in the gel. This was previously established to be a thickness of 1–2 mm. The barrier was applied to the exposed subperitoneal muscle, and extended beyond the cut border by 2–3 cm to ensure coverage of potential ischaemic areas. Approximately 5 ml of each precursor was needed to cover an area of 12x8 cm. Due to the hydrophilic nature of the hydrogel barrier, treated sidewalls were irrigated with normal saline to ensure a moist, lubricious surface prior to closure.



View larger version (65K):
[in this window]
[in a new window]
 
Figure 2. Application of SprayGel to the peritoneal injury site. Blue coloration identifies extent of adhesion barrier coverage.

 


View larger version (101K):
[in this window]
[in a new window]
 
Figure 3. The SprayGelTM; Adhesion Barrier System, consisting of an applicator with two syringes containing hydrogel precursor solutions. A small compressor (not shown) supplies air for atomization via the connecting tubing.

 
Adhesion evaluation
The first two animals enrolled were evaluated for adhesions at an earlier time point than the others (6 versus 14–16 days) to ensure adequate model response. At the time of evaluation, one umbilical and two lower quadrant trochars were introduced. The animals were placed in the Trendelenberg position and adhesion formation was scored laparoscopically by a surgeon blinded to treatment site assignment. Laparoscopic procedures were videotaped for additional evaluation if needed. The presence or absence of adhesions at each site was noted, and extent of adhesion coverage was scored as follows: 0, clean, no adhesions; 1, adhesions on <50% of the stripped sidewall area; and 2, adhesions on 50–100% of the stripped sidewall area. After this, adhesion severity was scored as: 0, clean, no adhesions; 1, filmy adhesions; and 2, dense vascular adhesions.

Following adhesion scoring, animals were euthanized with an i.v. potassium chloride injection, and a midline laparotomy was performed. From this approach the laparoscopically obtained adhesion scores could be confirmed, and representative tissues were retrieved for histological examination. Tissues were fixed in neutral buffered formalin, embedded, sectioned and stained using haematoxylin and eosin for light microscopy.

Statistics
To determine the statistical significance of adhesion formation incidence between the treated and control groups, a {chi}2 test was used. Differences in the extent and severity of adhesion formation scores were assessed using the Wilcoxon signed rank test for nonparametric data. All data were analyzed using the SPSS software package (SPSS Version 9.0; SPSS, Chicago, IL, USA). A value of P < 0.05 was considered to be statistically significant.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
All animals survived the procedures uneventfully. Adhesion scores for the treated and control sites of each animal, along with the number of days implanted prior to adhesion evaluation are listed in Table IGo. At the time of the adhesion evaluation procedure, residual adhesion barrier was not observed at any of the treated sites within the pelvic cavity, including the two animals evaluated at 6 days. This hydrogel barrier absorption allowed the surgeon evaluating adhesion formation to remain blinded. This finding, coupled with the fact that the peritoneal adhesion formation process is largely complete within 4–5 days (Holtz, 1984Go, Harris et al., 1995Go), justifies the inclusion of 6 day animal data in the larger data set.


View this table:
[in this window]
[in a new window]
 
Table I. Adhesion extent and severity scores
 
Application of the standardized surgical injury resulted in formation of adhesions between the uterine horn anastamosis and the pelvic sidewall in 9 of the 10 sites randomized to control.

One animal had no adhesions to either the control or treatment sites. Three animals had adhesions to both the control and treatment sites. No animals had adhesions to only the treatment site. Three of the 10 sites randomized to adhesion barrier treatment were involved in adhesions. One of these treated sites with adhesions had required suture ligation of a large arterial bleeder on the side wall during the initial surgery that could not be controlled with electrocautery alone prior to randomization to treatment. The remaining 7 of 10 treated sites were adhesion free.

When one compares the incidence of adhesion formation in the treated and control sites, a statistically significant reduction of 67% was observed in the treated sites (P = 0.006). Also, a statistically significant reduction in the adhesion extent score (P = 0.029) and adhesion severity score (P = 0.023) was observed in the treated sites. It is notable that when evaluated without the first two (6 day) cases the differences between incidence, extent and severity are no longer significant, probably because of the reduced sample size (n = 8). For representative purposes, Figure 4AGo shows a gross evaluation of a treatment site without adhesions, while Figure 4BGo shows a control site with adhesions.



View larger version (95K):
[in this window]
[in a new window]
 
Figure 4. Gross appearance of a treatment site (A) without adhesions, and a control site (B) with adhesions. B = border of previous sidewall injury; F = forceps; S = suture knot;U = uterine wall; arrowheads = adhesion to sidewall.

 
Reperitonealization of the treated sites was confirmed histologically (see Figure 5AGo). Adhesions that had formed in most control sites were vascular and some serosal thickening was evident by both gross as well as histological analysis. A microscopic view of an adhesion free control sidewall is shown in Figure 5BGo for comparison.



View larger version (129K):
[in this window]
[in a new window]
 
Figure 5. Histological analysis of a treated site (A) showingre-peritonealization, and of a control site (B) showing the healed surface (from an adhesion free area). Bars = 100 µm.

 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Even though numerous adhesion barriers have been proposed and tested both pre-clinically and clinically (Steinleiter et al., 1991Go; Haney and Doty, 1992Go; Hill-West et al., 1994Go; Marana et al., 1997Go; Becker et al., 1996Go) there still remains a real clinical need for an effective material to prevent site specific adhesion.

For a barrier to be clinically effective, it should be easy to use in both laparoscopic and open procedures, as well as adhere to the desired tissue long enough to prevent adhesions. Small animal models for adhesion barriers typically do not allow laparoscopic device application, and organ sizes and forces are insufficient to challenge the ability of a barrier to remain attached to the desired site.

Many small animal models (mice, rats, rabbits) commonly described in the literature are acceptable analogues of the biochemical processes involved in adhesion formation. Valuable information on adhesion formation, reformation and prevention has been obtained from these models. However, when compared with humans, these models have differences in scale, surgical procedure, organ size, organ weight and physiological forces. These differences can reduce the ability of these models to predict clinical success.

To this end, some have proposed large animal models such as a canine radical pelvic resection model that simulates radical hysterectomy (Montz et al., 1993bGo), a porcine incisional hernia repair model (Christoforoni et al., 1996Go), a porcine pelvic surgery model (Montz et al., 1993aGo) and a ewe hysterotomy model that simulates a myomectomy (Moll et al., 1992Go). No large animal models have been adequately described for the formation of adnexal adhesions with the pelvic sidewall.

In this study we present a porcine model for adhesion formation following tubal and ovarian surgery. An attempt is made to develop site-specific adhesions with high reproducibility using a surgical procedure analogous to adnexal surgery or myomectomy.

This model creates typical surgical conditions encountered in tubal and ovarian surgery, a clinically relevant procedure that is known to be at high risk for the development of post-operative adhesions. The inclusion of several adhesiogenic stimuli that are routine during surgery, such as de-peritonealization of tissues and ischaemic insults from electrocautery and sutures, creates a surgical environment optimal for evaluating adhesion barriers. In this manner, preclinical animal model testing of proposed prophylactic anti-adhesion materials can more accurately predict ultimate clinical efficacy in humans. In addition to developing this animal model, the present study also evaluated a specific adhesion barrier material, SprayGel.

Relative to the control side, there was a 67% reduction in the incidence of adhesions at the treated side. On the treated side, only the pelvic sidewall received the adhesion barrier, leaving the injured uterine horn and bladder unprotected. It is significant to note that adhesions were observed on the treated side between the uterine horn and the bladder, demonstrating that the barrier did indeed exclude the adhesiogenic horn from the sidewall, but not from untreated ipsilateral sites.

High inter-animal variability in animal models has led to the suggestion of using each animal as its own control (Ordonez et al., 1997Go). This of course can only be performed with non-regional adhesion prevention strategies. Since in this study the adhesion barrier is applied locally by spraying and does not redistribute throughout the pelvis like a liquid or viscous gel, such a model can be used resulting in distinct statistical advantages. Moreover, the need for internal controls is particularly important when using large animals not specifically bred for genetic similarity.

Despite the presence of several large animal adhesion formation models in the literature, there is a need for a reliable large animal model of adhesion formation following tubal and ovarian surgery. Given the histological similarities between porcine uterine horns and human Fallopian tubes, this model seems appropriate to assess post-operative adhesions following adnexal surgery.

This porcine model also allowed the creation of conditions that are relevant to the human surgical environment in terms of organ size and forces exerted on the adhesion barrier. These conditions are needed to better evaluate important barrier features such as ease of placement and resistance to migration from the desired site.

In order to obtain consistent adhesions at the control site, it was determined during model development that the injury process needed to be performed via laparotomy. Therefore, the adhesion barrier was applied during the same open procedure. In the future, laparoscopic material application following the laparotomy closure may be performed. Even though in this model the adhesion barrier was not deployed laparoscopically, the sprayable nature of this barrier allows for the easy deposition in both laparoscopic and open surgical scenarios. The presence of a colourant allows for an easy visualization of the hydrogel and precise placement. The transformation of the precursor solutions to a tissue adherent hydrogel takes place within seconds and large denuded or traumatized areas can be expeditiously protected. An undetectable amount of heat is liberated during the gelation process, and due to the high water content of the components a lubricious surface is presented to surrounding tissues and organs after deposition.

PEG is a poor substrate for bacteria due to its non-biological origin. Thus, along with the rapid barrier absorption rate (less than 1 week), the barrier material does not lend itself readily to the promotion or potentiation of bacterial infection. Despite the emergence of several regional adhesion prevention instillates, there is a clear need for an efficacious, easy to use site-specific adhesion barrier that can be used laparoscopically. Thus, SprayGel will potentially address the need of the laparoscopic surgeon who needs to protect site-specific injuries that are susceptible to post-surgical adhesion formation.

The results of this study lead us to conclude that this porcine model of adhesion formation is appropriate for the investigation of site-specific adhesion formation and prevention in a clinically relevant surgical procedure. The promising efficacy demonstrated by the PEG adhesion barrier in this and other (Dunn et al., 2001Go) models of adhesion formation warrant the further investigation of this adhesion barrier material in a larger animal study.


    Acknowledgements
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
The sponsors would like to thank Tom Frayne and Jim Clarke of the Rhode Island Hospital for their expert assistance, Ogan Gurel, M.D. for his expert critical review of this manuscript and Erin R.Campbell, Ed.D. for statistical assistance. This work was funded with a research grant from Confluent Surgical Inc.


    Notes
 
3 To whom correspondence should be addressed at: 695 Eddy St., Providence, RI 02905, USA. E-mail: roger_ferland{at}brown.edu Back

Statement of commercial interest

D.Mulani is a former employee while P.K.Campbell is a current employee of Confluent Surgical Inc. R.Ferland is a consultant to Confluent Surgical Inc.


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Becker, J.M., Dayton, M.T., Fazio, V.W. et al. (1996) Prevention of postoperative abdominal adhesions by a sodium hyaluronate-based bioresorbable membrane: a prospective, randomized, double-blind multicenter study. J. Am. Coll. Surg., 183,297–306.[ISI][Medline]

Christoforoni, P.M., Kim, Y.B., Preys, Z. et al. (1996) Adhesion formation after incisional hernia repair: a randomized porcine trial. Am. Surg., 62, 935–938.[ISI][Medline]

Dunn, R., Lyman, M.D., Edelman, P.G. et al. (2001) Evaluation of the SprayGel adhesion barrier in the rat cecum abrasion and rabbit uterine horn adhesion models. Fertil. Steril., 75, 411–416.[ISI][Medline]

Ellis, H., Moran, B.J., Thompson, J.N. et al. (1999) Adhesion-related hospital readmissions after abdominal and pelvic surgery: a retrospective cohort study [see comments]. Lancet, 353, 1476–1480.[ISI][Medline]

Golan, A., Maymon, R., Winograd, I. et al. (1995) Prevention of post surgical adhesion formation using aspirin in a rodent model: a preliminary report. Hum. Reprod., 10, 1797–1800.[Abstract]

Haney, A.F. and Doty, E. (1992) Murine peritoneal injury and de novo adhesion formation caused by oxidized-regenerated cellulose (Interceed TC7) but not expanded polytetrafluorethylene (Gore-Tex surgical membrane). Fertil. Steril. 57,202–208.[ISI][Medline]

Harris, E.S., Morgan, R.F. and Rodeheaver, G.T. (1995) Analysis of the kinetics of peritoneal adhesion formation in the rat and evaluation of potential antiadhesive agents. Surgery, 117, 663–669.[ISI][Medline]

Hellebrekers, B.W., Trimbos-Kemper, T.C., Trimbos, J.B. et al. (2000) Use of fibrinolytic agents in the prevention of postoperative adhesion formation. Fertil. Steril., 74, 203–212.[ISI][Medline]

Hill-West, J.L., Chowdhury, S.M., Dunn, R.C. et al. (1994) Efficacy of a resorbable hydrogel barrier, oxidized regenerated cellulose, and hyaluronic acid in the prevention of ovarian adhesions in a rabbit model. Fertil. Steril., 62, 630–634.[ISI][Medline]

Holtz, G. (1984) Prevention and management of peritoneal adhesions. Fertil. Steril., 41, 497–507.[ISI][Medline]

Lower, A.M., Hawthorn, R.J., Ellis, H. et al. (2000) The impact of adhesions on hospital readmissions over ten years after 8849 open gynaecological operations: an assessment from the Surgical and Clinical Adhesions Research Study. Br. J. Obstet. Gynaecol., 107, 855–862.[ISI]

Marana, R., Catalano, G.F., Caruana, P. et al. (1997) Postoperative adhesion formation and reproductive outcome using Interceed after ovarian surgery: a randomized trial in the rabbit model.Hum. Reprod., 12, 1935–1938.[Abstract]

Moll, H.D., Wolfe, D.F., Schumacher, J. et al. (1992) Evaluation of sodium carboxymethylcellulose for prevention of adhesions after trauma in ewes. Am. J. Vet. Res., 53, 1454–1456.[ISI][Medline]

Montz, F.J., Monk, B.J., Lacy, S.M. and Fowler, J.M. (1993a) Ketorolac tromethamine, a non-steroidal anti-inflammatory drug: ability to inhibit post-radical pelvic surgery adhesions in a porcine model. Gynecol. Oncol., 48, 76–79.[ISI][Medline]

Montz, F.J., Monk, B.J. and Lacy, S.M. (1993b) Effectiveness of two barriers at inhibiting post-radical pelvic surgery adhesions. Gynecol. Oncol., 48, 247–251.[ISI][Medline]

Ordonez, J.L, Dominguez, J., Evrard, V. et al. (1997) The effect of training and duration of surgery on adhesion formation in the rabbit model. Hum. Reprod., 12, 2654–2657.[Abstract]

Ray, N.F., Denton, W.G., Thamer, M. et al. (1998) Abdominal adhesiolysis: inpatient care and expenditures in the United States in 1994. J. Am. Coll. Surg., 186, 1–9.[ISI][Medline]

Stangel, J.J., Nisbet, J.D. 2nd. and Settles, H. (1984) Formation and prevention of postoperative abdominal adhesions. J. Reprod. Med., 29, 143–156.[ISI][Medline]

Steinleiter, A., Lambert, H., Kazensky, C. et al. (1991) Poloxamer 407 as an intraperitoneal barrier material for the prevention of post-surgical adhesion reformation. Obstet. Gynaecol., 77, 48–52.[Abstract]

Trimbos-Kemper, T.C., Trimbos, J.B. and van Hall, E.V. (1985) Adhesion formation after tubal surgery: results of the eighth-day laparoscopy in 188 patients. Fertil. Steril., 43,395–400.[ISI][Medline]

Yamaoka, T., Tabata, Y. and Ikada, Y. (1993) Distribution and tissue uptake of poly(ethylene glycol) with different molecular weights after intravenous administration to mice.J. Pharm. Sci., 83, 601–606.[ISI]

Submitted on October 27, 2000; accepted on August 28, 2001.