plasminogen activator inhibitor-1 and cytokine modulation by the PROACT™ system*

plasminogen activator inhibitor-1 and cytokine modulation by the PROACT™ system*

FERTILITY AND STERILITY威 VOL. 79, NO. 4, APRIL 2003 Copyright ©2003 American Society for Reproductive Medicine Published by Elsevier Science Inc. Prin...

128KB Sizes 0 Downloads 36 Views

FERTILITY AND STERILITY威 VOL. 79, NO. 4, APRIL 2003 Copyright ©2003 American Society for Reproductive Medicine Published by Elsevier Science Inc. Printed on acid-free paper in U.S.A.

Plasminogen activator/plasminogen activator inhibitor-1 and cytokine modulation by the PROACT™ System* Marie-Louise Ivarsson, M.D., Ph.D.,a Michael P. Diamond, M.D.,b Peter Falk, M.A.,a and Lena Holmdahl, M.D., Ph.D.a Sahlgrenska University Hospital, Go¨teborg University, Go¨teborg, Sweden

Received June 20, 2002; revised and accepted October 14, 2002. Supported by NTERO Surgical, Inc., Palo Alto, California. *PROACT™ System, NTERO Surgical, Inc., Palo Alto, California. Reprint requests: Michael P. Diamond, M.D., Department of Obstetrics and Gynecology, Wayne State University, 4707 St. Antoine Boulevard, Detroit, Michigan 48201 (FAX: 313– 745–7037; E-mail: [email protected]). a Department of Surgery, Sahlgrenska University Hospital, Go¨teborg University. b Department of Obstetrics and Gynecology, Wayne State University School of Medicine, Detroit, Michigan. 0015-0282/03/$30.00 doi:10.1016/S0015-0282(02) 04851-3

Objective: To examine the effects of the PROACT treatment on the fibrinolytic system and inflammatory cytokines in human peritoneum. Design: Controlled clinical study. Setting: University hospital. Patient(s): Nine subjects undergoing laparotomy had peritoneal samples taken at the incision. Intervention(s): The PROACT applicator was inserted through the peritoneal incision, and treatment of peritoneum was performed twice. A peritoneal sample was taken from one treated area. At closure, the second treated sample and an additional control sample were taken. All four samples were snap frozen in liquid nitrogen. Samples were homogenized and protein content extracted. Main Outcome Measure(s): Concentrations of total and active transforming growth factor-beta 1 (TGF-␤1), tumor necrosis factor-alpha (TNF-␣), tissue-type plasminogen activator (t-PA), urokinase plasminogen activator (uPA), and plasminogen activator inhibitor 1 (PAI-1) were obtained. Result(s): Total TGF-␤1 at opening was 30% less in treated samples. At closure, active TGF-␤1 increased significantly (163%) in control samples and not in treated samples. Tumor necrosis factor alpha was detectable only in control samples at closure. During surgery, tPA levels showed a marked decrease in control samples vs. a small increase in treated samples. Levels of uPA increased significantly only in the control samples. In control samples, tPA/PAI-1 ratio was two thirds of treated sample ratio. Conclusion(s): Heating of the peritoneum with the PROACT™ System modulates the biologic tissue response to induce effects that would be consistent with inhibition of postoperative adhesion development. (Fertil Steril威 2003;79:987–92. ©2003 by American Society for Reproductive Medicine.) Key Words: Adhesions, incision line, postoperative adhesions, plasminogen activator, PROACT™ System, transforming growth factor-beta, plasminogen activator inhibitor-1

Peritoneal adhesions occur in most individuals who are undergoing intraabdominal surgery (1– 4). Their presence may be of particular concern when they connect the anterior abdominal wall incision line to the underlying structures, particularly bowel. In general surgery procedures, such adhesions have been identified in 94% of subjects (5). In retrospective series, rates of approximately 67% have been described with midline incisions from the xiphoid to the symphysis pubis, and rates of approximately 55–59% have been described for shorter midline incisions, of which almost one third have been identified to involve the bowel (6 – 8).

At the least, adhesions to the anterior abdominal wall can prolong the surgical procedure, resulting in greater operating room costs (9). Also, such adhesions can result in injuries to underlying organs, which greatly increase patient morbidity, resulting in more extensive procedures and longer hospitalizations, and, as in the case of unidentified bowel injury, result in patient morbidity (10). Thus a great clinical need exists to reduce adhesions to the anterior abdominal wall. Recently, we reported a device that was able to significantly reduce midline adhesions in a porcine model from 7 (50%) of 14 animals in the control groups to zero of 20 animals in the 987

FIGURE 1 PROACT™ System controller and applicator (NTERO Surgical, Inc., Palo Alto, CA).

Ivarsson. PROACT™ System peritoneal treatment. Fertil Steril 2003.

NTERO PROACT System™–treated groups (11). This device briefly heats the peritoneum to a preset temperature for a preset time; the peritoneal incision is then made through the treated peritoneum. Although the mechanism of action of this device may be to limit postoperative exudation of blood and serosanguinous fluid from the cut edges of the peritoneum, we speculated that the device might also have a biologic effect on tissue factors that previously have been identified to be associated with adhesion development (12– 22). To examine this possibility, we undertook the present study to examine molecular biologic effects of the NTERO PROACT™ System in humans on these factors reported to affect postoperative adhesion formation.

MATERIALS AND METHODS Male and female subjects (n ⫽ 9) recruited for this study were ages 18 years and older and were undergoing nonemergent colorectal resections by laparotomy for clinically indicated purposes as recommended by their personal physician. Subjects were excluded if they had peritonitis, had a prior midline abdominal incision, had metastasizing intraabdominal cancer, or were pregnant. This study was conducted at Sahlgrenska University Hospital under a protocol approved 988

Ivarsson et al.

PROACT™ System peritoneal treatment

by the local ethics committee. All subjects gave voluntary, informed, written consent before entry into the study. The study consisted of collection of peritoneal tissue immediately adjacent to the anterior abdominal wall incision line. Specifically, the skin, subcutaneous tissue, and fascia were incised in the midline according to surgeon’s preference. Fat and fascia were retracted laterally, exposing ⱖ6 cm of peritoneum in the midline. After this point, abdominal entry was performed totally without the use of electrosurgical cutting or coagulation devices, except for controlling of bleeding vessels. Subjects served as their own controls. All treatments were performed using the PROACT™ System. The PROACT™ System (NTERO Surgical, Inc., Palo Alto, CA) consists of two main components: a mechanical applicator and an electronic controller (Fig. 1). The applicator is provided sterile and is intended for single patient use. At the appropriate time during surgery, after the fascia has been transected and the peritoneum exposed, the tissue to be treated is captured between the jaws of the applicator and heated using a heating element. The electronic controller regulates the treatment parameters. The controller automatically achieves and then maintains the desired temperature; subsequently it turns itself off at the end of the treatment. Correct operation is confirmed by audible signals and lights Vol. 79, No. 4, April 2003

TABLE 1 Concentration of total protein, total TGF-␤1, active TGF-␤1, TNF-␣, tPA, uPA, and PAI-1 in control tissues and tissues treated with the NTERO PROACT™ System, at the time of opening and closing of the abdominal cavity. Mean opening values Tissue factor

Mean closing values

Control tissues

Treated tissues

P value

Control tissues

Treated tissues

P value

31.1 1,144 54 0 19.1 0.1 0.1

38.6 799 63 0 8.3 0.02 0.1

NS NS NS NS ⬍.01 NS NS

46.4 955 142a 0.7 7.4b 0.4a 0.9c

41.3 978 108 0 9.1 0.03 0.6

NS NS ⬍.05 NS NS ⬍.02 NS

Total Protein (mg) Total TGF-beta (pg/mg) Active TGF-beta (pg/mg) TNF-alpha (pg/mg) tPA (ng/mg) uPA (ng/mg) PAI-1 (ng/mg) a

Significant intraoperative decrease (P⬍.01) compared to treated tissues. Significant intraoperative increase (P⬍.05) compared to treated tissues. c Significant intraoperative increase (P⬍.02) compared to treated tissues. b

Ivarsson. PROACT™ System peritoneal treatment. Fertil Steril 2003.

on the controller that indicate when treatment is taking place and when treatment has been completed. Entry through the peritoneum into the abdominal cavity was performed by sharp dissection. A peritoneal sample approximately 1.5 cm2 was immediately taken. The lower jaw of the PROACT™ System was then inserted through the peritoneal incision, and treatment of approximately a 5-cm length of peritoneum was performed. A peritoneal sample similar in size to the initial untreated sample was then taken from the treated area. The peritoneal incision was completed in standard (e.g., scissors or scalpel) fashion by the attending surgeon, and the indicated surgical procedure was performed. At the conclusion of the surgical procedure, just before fascial closure, two final peritoneal samples (each approximately 1.5 cm2) were taken from the initial midline incision, one from an untreated site and one from a site initially treated with the PROACT™ System. In total, four peritoneal tissue samples, two untreated and two treated, were collected. All four tissue samples were snap frozen in liquid nitrogen immediately after sampling. Specimens were then stored at ⫺70°C until analysis, which was performed in batches. Samples were homogenized and the protein content extracted as previously described (12). Concentrations of total and active transforming growth factor-beta 1 (TGF-␤1), tumor necrosis factor-alpha (TNF-␣), tissue-type plasminogen activator (tPA), urokinase plasminogen activator (uPA), and plasminogen activator inhibitor 1 (PAI-1) were assayed using commercially available kits, and values were normalized to total protein content. Transforming growth factor-beta 1 was assayed using a kit from Promega (Madison, WI) with a lower detection limit of 32 pg/mL. The intraassay coefficient of variance (CV) was ⬍4.5%, and the interassay CV, ⬍19.1%. Tumor necrosis factor-alpha was measured with a kit from Pierce-EndoFERTILITY & STERILITY威

gen (Woburn, MA). The lower detection limit was 2 pg/mL; intraassay CV, ⬍3.2%; and interassay CV, ⬍7.2%. Tissuetype plasminogen activator, PAI-1, and uPA were analyzed with kits from Biopool (Umeå, Sweden). TintElize tPA has a lower detection limit of 1.5 ng/mL, with an intraassay CV of ⬍5.5% and an interassay CV of ⬍5.4%. The lower detection limit of TintElize PAI-1 is 0.5 ng/mL, intraassay CV of ⬍2.9%, and interassay CV of ⬍3.3%. TintElize uPA has a lower detection limit of 0.1 ng/mL and an intraassay CV of 10%; furthermore, we observed the interassay variation to be CV ⬍10% (n ⫽ 10). Statistical analysis was performed with Wilcoxon’s signed rank test, and correlation with operation time was done with Spearman’s rank correlation test. Significance was defined as P⬍.05. All data are expressed as mean ⫾ SEM.

RESULTS Peritoneal tissue samples were collected from nine subjects, males and females, with a mean age of 73 ⫾ 3 years (range, 57 to 86 years). Four specimens, representing treated and untreated peritoneum at the initiation and conclusion of the surgical procedure, were collected from each subject. Primary indications for the procedures were colorectal resections. Measurements of tissue levels of the inflammatory cytokines and fibrinolytic parameters are all expressed as nanograms or picograms per milligram of protein. The initial protein content of the tissues was 31.1 ⫾ 3.8 and 38.6 ⫾ 3.7 protein per milligram of tissue in the control and treated specimens, respectively (Table 1; nonsignificant). At the conclusion of the surgical procedure, the protein content had increased in the control specimens to 46.4 ⫾ 3.8 mg of protein per milligram of tissue, as opposed to 41.3 ⫾ 5.0 of protein per milligram of tissue in treated tissue (nonsignificant). 989

FIGURE 2 Positive correlation demonstrating the relationship between increasing surgical time (hours) with increasing TGF-␤1 concentrations.

Ivarsson. PROACT™ System peritoneal treatment. Fertil Steril 2003.

Total transforming growth factor beta-1 (TGF-␤1) levels at the time of peritoneal opening were 1,144 ⫾ 207 pg per milligram of protein in untreated areas, as compared with the case of areas immediately undergoing PROACT™ System treatment at the beginning of the procedure, which had a 30% reduction in TGF-␤1 expression to 799 ⫾ 101 pg per milligram of protein, although this fall did not reach significance (Table 1). Opening active TGF-␤1 levels were no different in control and treated specimens (54 ⫾ 22 vs. 63 ⫾ 15 pg per milligram of protein). At the completion of the surgical procedure, total TGF-␤1 levels were indistinguishable (995 ⫾ 133 vs. 978 ⫾ 139 pg per milligram of protein, respectively). However, in the control specimens at the completion of the procedure, there was a significant increase in active TGF-␤1, to 142 ⫾ 23 pg per milligram of protein (P⬍.05), an increase over baseline of 163%. In contrast, in treated specimens, active TGF-␤1 increased only 71%, to 108 ⫾ 26 pg per milligram of protein (nonsignificant vs. initial treated specimen). Interestingly, increasing surgical time from the initial to final tissue sampling was positively correlated with active TGF-␤1 concentration in untreated tissue (r ⫽ .7, P⬍.05; Fig. 2). Tumor necrosis factor-alpha (TNF-␣) was initially undetectable in both control and treated specimens. At the completion of the surgery, TNF-␣ levels were still undetectable in treated tissues, whereas in control specimens the level was 0.7 ⫾ 0.6 pg per milligram of protein. Tissue plasminogen activator levels at the initiation of the procedure at untreated sites were 19.1 ⫾ 3.5 ng per milligram of protein. Treatment with the PROACT™ System immediately before tissue sampling reduced tissue tPA levels to 8.3 ⫾ 2.0 ng per milligram of protein (Table 1; P⬍.01). At the conclusion of the procedure, tPA levels from 990

Ivarsson et al.

PROACT™ System peritoneal treatment

previously untreated and treated areas were 7.4 ⫾ 1.5 and 9.1 ⫾ 1.8 ng per milligram of protein, respectively. Thus, there was a marked difference in the change in tPA levels during surgery; in untreated peritoneum, tPA concentrations decreased 11.6 ⫾ 3.7 ng per milligram of protein, as compared with treated peritoneum, in which tPA levels increased slightly by 0.8 ⫾ 2.1 ng per milligram of protein (P⬍.01). Urokinase plasminogen activator levels were initially 0.1 ⫾ 0.04 ng per milligram of protein and decreased to 0.02 ⫾ 0.01 ng per milligram of protein after PROACT™ System treatment. Levels at the conclusion of the surgical procedure were 0.4 ⫾ 0.1 in the control group and 0.1 ⫾ 0.03 ng per milligram of protein in the treated group, representing a significant increase in uPA in the control (P⬍.05) but not in the treated group (nonsignificant), with significantly higher levels at closure in the control group (P⬍.02). Plasminogen activator inhibitor-1 levels initially were 0.1 ⫾ 0.1 and 0.1 ⫾ 0.04 at control and treated sites, respectively. At the conclusion of the surgical procedure, there was a marked increase in PAI-1 levels in untreated peritoneum, to 0.9 ⫾ 0.3 ng per milligram of protein (P⬍.02 vs. initial concentrations); the increase was attenuated in peritoneum treated with the NTERO PROACT™ System (0.6 ⫾ 0.2 ng per milligram of protein). If fibrinolytic activity is expressed as the ratio of tPA/ PAI-1, the initial peritoneal level of 58.5 ⫾ 26.9 is acutely reduced by use of the PROACT™ System, to 19.6 ⫾ 0.8. However, at the conclusion of the surgical procedure, a state that may be considered to better represent the actual milieu during the subsequent time period of peritoneal healing and adhesion development, the ratio in untreated peritoneum was 11.9 ⫾ 4.1, which was only two thirds of the ratio in treated areas of 17.5 ⫾ 5.1. Similarly, if uPA is considered, the Vol. 79, No. 4, April 2003

initial uPA–PAI-1 ratio of 1.0 ⫾ 0.6 was reduced by treatment with the PROACT™ System to 0.6 ⫾ 0.1. However, at the time immediately before peritoneal closure, the ratio in untreated tissue was 0.5 ⫾ 0.1 ng per milligram of tissue, as compared with 0.2 ⫾ 0.1 at treated sites.

DISCUSSION The process of peritoneal healing and adhesion development after surgical injury includes initiation of an inflammatory response with increased elaboration of cytokines, including TGF-␤1 and TNF-␣ (12–17). In humans, TGF-␤1 expression occurs both in the mesothelial cells, which line the peritoneal cavity, as well as underlying fibroblasts (14, 18, 23). That elevation of TGF-␤1 is associated with adhesion development is suggested by increased concentrations in peritoneal fluid and adhesions specimens of animals and subjects with adhesions. The ability of TGF-␤1 to increase adhesion development in animals has also been shown (14, 16, 17, 19, 24). Similarly, TNF-␣ has been identified in increased concentrations in the presence of peritoneal adhesions (12, 13, 15). It is noteworthy that TGF-␤1 exists in a latent or inactive form, as well as in an active form that is responsible for its actions, which include stimulation of extracellular matrix (ECM) production, enhancement of angiogenesis, and impairment of both matrix metalloproteinase and plasminogen activity. As a result of the latter two functions, there is enhancement of deposition of ECM, which contributes to adhesion development. In this study the use of the PROACT™ system resulted in no significant effect on total or active TGF-␤1 at the beginning of surgery. However, by the completion of the surgical procedure, peritoneum untreated with the PROACT™ system had a significant, marked increase in TGF-␤1 protein concentration. In fact, a significant correlation existed between increasing operative time and the degree of elevation of TGF-␤1, such that prolongation of the operative procedure would be expected to increase the inflammatory response and subsequent adhesion development. These findings in our small sample of patients lend support for conducting a larger study to evaluate the impact of surgical time on levels of molecular tissue markers. Consistent with this observation, TNF-␣ levels that were initially undetectable in both control and treated peritoneum became present in untreated (but not treated) peritoneum by the conclusion of the surgical procedure. Thus, use of the PROACT™ System appears to diminish elaboration of inflammatory cytokines, which in turn would be expected to be associated with reduced postoperative adhesion development. The pathogenesis of adhesions is also believed to be critically dependent on plasminogen activator activity (PAA). More than three decades ago, Buckman et al. (25, 26) demonstrated that normal peritoneum manifested PAA, as evidenced by creation of a zone of hemolysis in a blood FERTILITY & STERILITY威

auger plate, and that this activity can be markedly reduced by tissue hypoxia. These observations were consistent with the results of a series of rat studies (27–32) by Raftery, which included light microscopy, election microscopy, and enzymatic elevations. More recently, Holmdahl et al. (20, 33–35) and Ivarsson et al. (36) have conducted a series of studies in humans that have confirmed a central role of tPA and PAI-1 in peritoneal fibrinolysis and its consequences on postoperative adhesion development. Interestingly, although PAA has classically been considered to reside in the mesothelial cells that line the peritoneal cavity, recent studies have suggested that PAA also exists in fibroblasts from the peritoneum (37). Additionally, it is now understood that PAA is a function both of plasminogen activators, which are serine proteases, and its inhibitors. More specifically, the protease activity of peritoneum is thought primarily to reside in tPA as opposed to uPA (38, 39). Of the two established inhibitors of tPA, PAI-1 rather than PAI-2 is thought to be most responsible for limiting tPA proteolytic activity, which is accomplished by binding in a one-to-one molar ratio. In the absence of PAA, the proteinaceous mass, which develops after surgery from bleeding and oozing of serosanguinous fluids, congeals and becomes cross-linked to mature fibrin. This clot then forms a scaffold for ingrowth of fibroblasts from the underlying injured peritoneal surfaces, with deposition of collagen and other forms of extracellular matrix material. Mesothelial cells subsequently reepithelialize the clot, resulting in the creation of an adhesion. Over time, and particularly in the presence of angiogenic compounds like vascular endothelial growth factor, vessels may develop as a means of resupplying oxygen and nutrients to tissues devascularized by surgery. However, if PAA persists after surgery, the proteolytic activity will degrade fibrin, resulting in a reduction or absence of congealing of the proteinaceous mass, and thus limiting or eliminating the scaffold required by fibroblasts for migration. As a consequence, ECM deposition does not occur so as to connect adjoining tissues, and healing (including mesothelial cell reepithelialization) can occur independent of adhesion development. Consequently, for the purposes of reducing postoperative adhesion development, on a molecular biologic level, it would be desirable to enhance tPA and/or minimize PAI-1. The net effect of such alterations would be a rise in the tPA/PAI-1 ratio and an increase in PAA. The PROACT™ System heats the peritoneal surface, thereby potentially limiting bleeding and exudation of serosanguinous fluids. In this human study, we sought to determine whether its mechanism of action could also include a biologic effect, namely enhancement of the tPA/PAI-1 ratio. When examining the samples collected just before abdominal closure, it is notable that there has been a marked increase in this ratio at treatment sites, which primarily 991

reflects a 50% reduction in the rise in PAI-1 at treated sites as compared with at control sites in the same subjects. It is important to note that errors in tissue collection at the treatment site, such as collection from an untreated area or collection of a tissue width exceeding the zone of treatment, would have minimized differences between control and treated groups, thus lending additional credence to the observed differences. The increased uPA–PAI-1 ratio in control, as opposed to treated tissues, may provide additional support for the lack of uPA contribution to peritoneal fibrinolysis. In summary, this clinical trial has provided at least one biologic mechanism to support the porcine findings that heating of the peritoneum with the PROACT™ System reduces postoperative adhesion development (11). These observations lend additional support for undertaking a clinical trial with the PROACT™ System to determine its efficacy in reducing postoperative adhesion development. References 1. Diamond MP, Feste J, McLaughlin DS, Martin DC, Daniell JF. Pelvic adhesions at early second-look laparoscopy following carbon dioxide laser surgery. Infertility 1984;7:39 –44. 2. Diamond MP, Daniell JF, Johns DA, Hill GA, Reich H, Martin DC, et al. Postoperative adhesion development following operative laparoscopy: evaluation at early second-look procedures. Fertil Steril 1991;55: 700 –4. 3. Wiseman DM, Trout JR, Diamond MP. The rates of adhesion development and the effects of crystalloid solutions on adhesion development in pelvic surgery. Fertil Steril 1998;70:702–11. 4. Menzies D, Ellis H. Intestinal obstruction from adhesions— how big is the problem? Am R Coll Surg Engl 1990;72:60 –3. 5. Becker JM, Dayton MT, Fazio VW, Beck DE, Stryker SJ, Wexner SD, et al. Prevention of postoperative adhesions by a sodium hyaluronatebased bioresorbable membrane: a prospective, randomized doubleblind multicenter study. J Am Coll Surg 1996;183:297–306. 6. Brill AL, Nezhat F, Nezhat CH, Nezhat C. The incidence of adhesions after prior laparotomy: a laparoscopic appraisal. Obstet Gynecol 1995; 86:269 –72. 7. Levrant SG, Bieber EJ, Barnes RB. Anterior abdominal wall adhesions after laparotomy or laparoscopy. J Am Assoc Gynecol Laparosc 1997; 4:353–6. 8. Audebert AS, Gomel V. Role of microlaparoscopy in the diagnosis of peritoneal and visceral adhesions and in the prevention of bowel injury associated with blind trocar insertion. Fertil Stertil 2000;73:631–5. 9. Coleman MG, McLain AD, Moran BJ. Impact of previous surgery on time taken for incision and division of adhesions during laparotomy. Dis Colon Rectum 2000;43:1297–9. 10. van der Krabben AA, Dijkstra FR, Nieuwenhuijzen M, Reijnen MM, Schaapveld M, Van Goon H, et al. Morbidity and mortality of inadvertent enterotomy during adhesiotomy. Br J Surg 2000;87:467–71. 11. Diamond MP, Stecco K, Paulson AJ. Use of the PROACT System for reduction of postsurgical peritoneal adhesions. Fertil Steril 2003;79: 198 –202. 12. Chegini N, Simms J, Williams RS, Masterson BJ. Identification of epidermal growth factor, transforming growth factor-alpha and epidermal growth factor receptor in surgically induced pelvic adhesions in the rat and intraperitoneal adhesions in the human. Am J Obstet Gynecol 1994;171:321–7. 13. Kaidi AA, Gurchumelidze T, Nazzal M, Figert P, Vanterpool C, Silva Y. Tumor necrosis factor-alpha: a marker for peritoneal adhesion formation. J Surg Res 1995;58:516 –8.

992

Ivarsson et al.

PROACT™ System peritoneal treatment

14. Saed GM, Zhang W, Chegini N, Holmdahl L, Diamond MP. Transforming growth factor beta isoforms production by human peritoneal mesothelial cells after exposure to hypoxia. Am J Reprod Immunol 2000;43:285–91. 15. Saba AA, Godziachvili V, Mavani AK, Silva YJ. Serum levels of interleukin 1 and tumor necrosis factor alpha correlate with peritoneal adhesion grades in humans after major abdominal surgery. Am Surg 1998;64:734 –6. 16. Chegini N, Kotseos K, Zhao Y, Bennett B, Frederick W, McLean F, et al. Differential expression of TGF-␤1 and TGF-␤3 in serosal tissues of human intraperitoneal organs and peritoneal adhesions. Hum Reprod 2001;16:1291–300. 17. Holmdahl L, Kotseos K, Bergstrom M, Falk P, Ivarsson ML, Chegini N. Over-production of transforming growth factor beta 1 (TGF-␤1) is associated with adhesion formation and peritoneal fibrinolytic impairment. Surgery 2001;129:626 –32. 18. Saed GM, Zhang W, Chegini N, Holmdahl L, Diamond MP, the PHAMUS Group. Alteration of type I and III collagen expression in human peritoneal mesothelial cells in response to hypoxia and transforming growth factor-beta 1. Wound Rep Regen 1999;7:504 –10. 19. Freeman ML, Saed GM, Diamond MP. Increased TGF-␤1/␤3 ratio following surgically induced peritoneal injury. J Soc Gynecol Invest 2001;8:229A. 20. Holmdahl L, Eriksson E, Al-Jabreen M, Risberg B. Fibrinolysis in human peritoneum during operation. Surgery 1996;119:701–5. 21. Holmdahl L, Falkenberg M, Ivarsson ML, Risberg B. Plasminogen activators and inhibitors in peritoneal tissue. APMIS 1997;105:25–30. 22. Ivarsson ML, Holmdahl L, Falk P, Molne J, Risberg B. Characterization and fibrinolytic properties of mesothelial cells isolated from peritoneal leverage. Scand J Clin Lab Invest 1998;58:195–204. 23. Saed GM, Zhang W, Diamond MP. Molecular characterization of fibroblasts isolated from human peritoneum and adhesions. Fertil Steril 2001;75:763–8. 24. Chegini N, Gold KI, Williams RS. Localization of transforming growth factor beta isoforms TGF-␤1, TGF-␤2, and TGF-␤3 in surgically induced pelvic adhesions in the rat. Obstet Gynecol 1994;83:449 –54. 25. Buckman RF, Buckman PD, Hufnagel HV, Gervin AS. A physiologic basis for the adhesion-free healing of deperitonealized surfaces. J Surg Res 1976;21:67–76. 26. Buckman RF, Woods M, Sargent L, Gervin AS. A unifying pathogenetic mechanism in the etiology of intraperitoneal adhesions. J Surg Res 1976;20:1–5. 27. Raftery AT. Regeneration of parietal and visceral peritoneum. A light microscopical study. Br J Surg 1973;60:293–9. 28. Raftery AT. Regeneration of parietal and visceral peritoneum. An electron microscopical study. J Anat 1973;115:375–92. 29. Raftery AT. An enzyme histochemical study of mesothelial cells in rodents. J Anat 1973;115:365–73. 30. Raftery AT. Noxythiolin (Noxyflex), aprotinin (Trasylol) and peritoneal adhesion formation: an experimental study in the rat. Br J Surg 1979;66:654 –6. 31. Raftery AT. Regeneration of peritoneum. A fibrinolytic study. J Anat 1979;129:659 –64. 32. Raftery AT. Effect of peritoneal trauma on peritoneal fibrinolytic activity and intraperitoneal adhesion formation: an experimental study in the rat. Eur Surg Res 1981;13:397–401. 33. Holmdahl L, Eriksson E, Risberg B. Measurement of fibrinolytic components in human tissue. Scand J Clin Lab Invest 1997;57:445–52. 34. Holmdahl L, Al-Jabreen M, Risberg B. The role of fibrinolysis in formation of postoperative adhesions. Wound Repair Regen 1994;7: 171–6. 35. Holmdahl L, Falkenberg M, Ivarsson ML, Risberg B. Plasminogen activators and inhibitors in peritoneal tissue. APMIS 1997;105:25–30. 36. Ivarsson ML, Holmdahl L, Falk P, Molne J, Risberg B. Characterization and fibrinolytic properties of mesothelial cells isolated from peritoneal leverage. Scand J Clin Lab Invest 1998;58:195–204. 37. Saed GM, Diamond MP. Modulation of the expression of tissue plasminogen activator and its inhibitor by hypoxia in human peritoneal and adhesion fibroblasts. Fertil Steril 2003;79:164 – 8. 38. Rout UK, Diamond MP. Plasminogen activators during the progression of uterine-peritoneal adhesions in the rat. Fertil Steril 2003;79:138 – 45. 39. Holmdahl L, Eriksson E. Fibrinolysis in human peritoneum during operation. Surgery 1996;119:701–5.

Vol. 79, No. 4, April 2003