A safety and efficacy study of a resorbable hydrogel for reduction of post-operative adhesions following myomectomy

Transcription

1 Human Reproduction Vol.23, No.5 pp , 2008 Advance Access publication on March 17, 2008 doi: /humrep/den080 A safety and efficacy study of a resorbable hydrogel for reduction of post-operative adhesions following myomectomy L. Mettler 1,7, J. Hucke 2, B. Bojahr 3, H.-R. Tinneberg 4, N. Leyland 5 and R. Avelar 6 1 Department of Obstetrics and Gynecology, University Clinics of Schleswig-Holstein, Campus Kiel, Michaelisstrasse 16, Kiel, Germany; 2 Bethesda Krankenhaus Wuppertal, Wuppertal, Germany; 3 Klinik für Minimal Invasive Chirurgie, Berlin, Germany; 4 Universitätsklinikum Giessen, D Giessen, Germany; 5 Obstetrics and Gynecology, St Joseph s Health Centre, Toronto, Ontario, Canada M6R 3B2; 6 Medical Affairs, Angiotech Pharmaceuticals, Inc., Vancouver, British Columbia, Canada V6A 1B6 7 Correspondence address. Tel: þ ; Fax: þ ; endo-office@ .uni-kiel.de; lmettler@ . uni-kiel.de BACKGROUND: This multicenter, randomized, single-blind study assessed the safety and efficacy of a resorbable hydrogel ( Hydrogel ) for the reduction of post-operative adhesion formation following myomectomy. METHODS: Women (n 5 71) who were undergoing laparoscopic (67.6%) or laparotomic myomectomy were randomized (2:1) to Hydrogel (sprayed over surgically treated areas prior to wound closure, n 5 48) or to control (standard care, n 5 23). Patients (38 Hydrogel, 20 control) returned 8 10 weeks later for a second look. Adhesions were graded using a modified American Fertility Society (mafs) scoring method. The primary efficacy measure was the posterior uterus mafs score. RESULTS: For Hydrogel and control patients, respectively, mean + SD mafs scores were and at baseline, and and at the second look. Similarly, mean changes from baseline were and (P ); 95% confidence intervals for these mean changes were ( ) and ( ). Adverse events were reported by 9.6 and 17.4% of Hydrogel and control patients, respectively. No intraabdominal infections or post-operative site infections were reported. CONCLUSIONS: This 71-patient study provides the first clinical evidence of the safety and efficacy of Hydrogel for the reduction of adhesions following myomectomy. The ClinicalTrials.gov Identifier is NCT Keywords: adhesions; gynecologic surgery; laparoscopy; laparotomy; polyethylene glycol Introduction Post-operative adhesions are a common medical complication of gynecologic and other pelvic surgeries (Monk et al., 1994). They are frequently associated with chronic or recurrent pelvic pain, intestinal obstruction and infertility (Diamond and DeCherney, 1987; Monk et al., 1994). Even with the recent adoption of microsurgical and laparoscopic techniques that reduce surgical trauma, pelvic adhesions have been reported to form in as many as 95% of patients within the first several weeks of major gynecologic surgery (Diamond and DeCherney, 1987). While laparoscopic lysis of adhesions is considered a worthwhile treatment for adhesion-related chronic pelvic pain (Steege and Stout, 1991), adhesions have been shown to reform at one or more sites in almost all patients who undergo adhesiolysis and at two-thirds of the actual sites where These data were presented at the 19th European Congress of Obstetrics and Gynecology (EBCOG) meeting, 5 8 April 2006 in Torino, Italy. adhesiolysis was performed (Operative Laparoscopy Study Group, 1991). In addition, between 12 and 51% of patients also develop de novo adhesions following laparotomy or operative laparoscopy (Diamond et al., 1987; Operative Laparoscopy Study Group, 1991; Canis et al., 1992). As a result, many patients are re-admitted more than once for disorders directly or probably related to adhesion formation (Ellis et al., 1999). In 1994, the total direct cost for hospitalizations related to adhesiolysis of abdominal adhesions in the USA (primarily for procedures on the digestive and female reproductive systems) was estimated to be $1.3 billion (Ray et al., 1998). Since the early 1980 s, a number of different approaches for the reduction and/or prevention of post-operative pelvic adhesions have been reported. Current research has focused on the anti-adhesion potential of various barrier membranes and materials. While several of the barriers developed to date have proven more efficacious than no intervention, adhesion formation is still reported to occur in an undesirably high percentage of # The Author Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved. For Permissions, please journals.permissions@oxfordjournals.org 1093

2 Mettler et al. treated patients. In addition, a number of limitations have been associated with their use e.g. difficulty in laparoscopic deployment (DeCherney and dizerega, 1997). The need for a highly effective, safe and readily deliverable anti-adhesion product remains unmet (Al-Jaroudi and Tulandi, 2004). CoSeal w Surgical Sealant is a resorbable hydrogel ( Hydrogel ), formed by mixing two synthetic polyethylene glycol (PEG) polymer solutions that are co-extruded with an activating solution from a syringe housing unit. Hydrogel, which acts as a physical barrier through cross-linking to itself and to the tissues it contacts, forms within seconds of application. It is totally resorbed at 30 days post-application. This product was originally developed by Cohesion Technologies, Inc., Palo Alto, California. The technology is now owned by Angiotech Pharmaceuticals, Inc. (Vancouver, Canada). It is approved in both the USA and Europe for use in sealing vascular reconstructions, and has been available in Europe since 2002 for the reduction of post-operative adhesion formation in patients undergoing cardiac surgery. The potential of Hydrogel as a barrier to prevent post-operative adhesion formation was first investigated in a rabbit model. Rabbits that had Hydrogel applied to an abraded epicardium had significantly reduced formation and tenacity of adhesions at necropsy compared with the surgical control group (Hendrix et al., 2001). Subsequently, in a small European clinical study (Konertz et al., 2003), Hydrogel was observed to be of benefit in reducing both the extent and tenacity of pericardial adhesions in infants undergoing surgical correction of congenital heart abnormalities. The applicator tip has since been adapted to also allow for delivery of Hydrogel to surgically treated sites during laparoscopic procedures. The current study was undertaken to investigate both the safety and efficacy of Hydrogel in a population known to be at high risk for post-operative adhesion formation and subsequent morbidity. Currently myomectomy is mostly performed by laparoscopy; however, laparotomy continues to be used in cases of multiple and very large fibroids (Mettler, 2006). Subserosal and intramural myomectomies are carried out for reasons of pain, infertility and visible growth to avoid later hysterectomies due to fibroid growth. Not only are women who undergo myomectomy at significant risk, the magnitude of the problem is amplified by the considerable number of women who undergo these types of procedures each year. Materials and Methods Conduct and design This prospective, randomized, controlled, single-blind study was conducted at six investigational sites: four in Germany, one in Canada and one in the Netherlands Antilles (Curaçao). Before any study-related procedures were performed, the investigator obtained written informed consent from each patient. All aspects of the study were governed by the applicable sections of the International Conference on Harmonisation Guidelines for Good Clinical Practice, ISO Clinical Investigation of Medical Devices for Human Patients, the Declaration of Helsinki of 1964 and its subsequent revisions, and by the requirements of each of the local ethics committees. This study was conducted in women, aged 18 years and above, who were scheduled to undergo myomectomy surgery either via 1094 laparoscopy or laparotomy. For those women who met all preoperative selection criteria, the surgeon used videography to perform a systematic visualization of all areas to evaluate baseline adhesion formation and to ensure that all intra-operative selection criteria were met. The surgeon evaluated the presence, extent and tenacity of adhesions at 15 anatomical sites: the anterior uterus, the posterior uterus, the anterior abdominal wall, the anterior cul-de-sac, the posterior cul-de-sac, the right pelvic side wall, the right ovary, the right tube, the left pelvic side wall, the left ovary, the left tube, the medial leaf of broad ligament, the small bowel, the large bowel and the omentum. The surgeon recorded whether adhesions were lysed, as well as the number, size and location of the myomas removed. At the end of the procedure, immediately prior to wound closure for laparotomy or removal of the endoscope for laparoscopy, patients who met all selection criteria were randomly assigned in a 2:1 ratio to the Hydrogel treatment group or to the standard of care control group. Treatment assignments had been randomly generated from a list of sequential numbers prior to enrolling the first patient. The randomization schedule was stratified by study center, and was performed in blocks to promote a consistent ratio at each center. Treatment assignments were prepared by Angiotech BioMaterials Corporation, Palo Alto, California, USA for each investigational site in sealed envelopes, one envelope per subject; envelopes were opened at the time of the surgery after the intra-operative assessment was conducted and continued participation was indicated. While study personnel were aware of the treatment assignation, study patients were not told which treatment they had received until they terminated the study. For those patients randomized to the treatment group, Hydrogel was sprayed over all myomectomy suture lines and all surgically treated areas using a gas-assisted delivery system. The PEG polymer solutions were mixed by a member of the surgical team and Hydrogel was used within 2 h; when assembled, the device contained 8 ml of solution. Hydrogel could be re-applied if the treated sites were disrupted. For those patients randomized to the control group, tissues were irrigated with Ringers Lactate solution and ml were left in the pelvic cavity prior to wound closure; no additional medications or antiadhesives were to be administered. Ringers Lactate solution was obtained by each investigational site from commercially available sources. Patients were to return within 6 8 weeks of the initial procedure for a second-look surgery to evaluate post-operative adhesion formation at the same 15 anatomical sites; this second procedure was always performed via laparoscopy. In every case, other procedures for the benefit of the patient were also undertaken as indicated. These included adhesiolysis, chromotubation, hysteroscopy and tubal sounding. Participants were scheduled to attend a total of five study visits: the initial pre-operative screening visit during which written informed consent was obtained; the primary surgery visit; a post-operative follow-up visit scheduled within days of the primary surgery; the secondlook surgery visit and, a follow-up and termination visit scheduled within days of the second procedure. Participants The patients had to meet pre-operative inclusion criteria to initially qualify for participation in the study as well as intra-operative criteria to qualify for continued participation. Pre-operatively, patients must have been 18 years of age, scheduled for myomectomy surgery, agreed to a second-look laparoscopic procedure to assess and lyse any adhesions formed at 6 8 weeks following myomectomy, provided voluntary written informed consent, and been willing to comply with all aspects of the treatment and evaluation schedule. Intra-operatively, the uterine incisions must have been at least 2 cm on the posterior uterine surface, which may have included the fundal surface. Patients who met any of the pre- or intra-operative

3 Hydrogel for adhesion reduction post-myomectomy exclusion criteria were not eligible for continued participation. Pregnancy, pelvic malignancy, pelvic inflammatory disease, an immune compromised condition, or the use of corticosteroids intra-operatively or during the course of the post-operative study follow-up were contraindications for the study. Intra-operatively, patients must not have had evidence of any of these same conditions, or received any adhesion prevention adjuvants or barriers, or peritoneal instillates containing corticosteroids, non-steroidal anti-inflammatory drugs or dextran. Outcome measures The primary safety measure was reported adverse events. Events were coded using the MedDRA adverse event dictionary, and summarized by system organ class and preferred term. Events were also summarized for each group by their incidence, severity, seriousness, relatedness to the investigational device and outcome. A serious adverse event was defined as one that caused death, was life threatening, permanently disabling or required and/or prolonged hospitalization or surgery. Adhesions were graded using the modified American Fertility Society (mafs) score that factored the presence, extent and tenacity of adhesions into a single score (Table I) for each of the 15 anatomical sites examined. According to the original study protocol, a composite mafs score was to have been calculated by averaging the mafs scores from those sites that had been treated for adhesions. In hindsight, it became apparent that use of this score would have biased the interpretation of results. For Hydrogel subjects, the composite mafs score would have been calculated as the average of the individual mafs scores from the 2 or 3 sites that were treated with Hydrogel; whereas, for control subjects, the composite score would have been calculated as the average of the individual mafs scores at all 15 sites even though most of these sites (e.g. small and large bowel) would not have had baseline adhesions nor would they be expected to form post-baseline adhesions as a result of the myomectomy procedure. Upon review of the myomectomy surgery record, it became apparent that the posterior uterus was the one anatomical site at which all patients were at risk for post-operative adhesion formation: most (84.5%) had one or more myomas removed from a posterior site and, as stipulated in the inclusion criteria, all (100%) had a uterine incision at a posterior site. In keeping with these findings, it was determined that the mafs score at the posterior uterus was the outcome best suited to serve as the primary measure of performance. Secondary measures related to adhesion formation were similarly restricted to the individual extent and tenacity scores at the posterior uterus site. Statistical considerations The sample size calculation for the current study was based on average adhesion scores of the first 16 patients (5 control, 11 Hydrogel) who completed the European cardiovascular pediatric study (mean + SD of and , respectively) and an observed common SD of 4.36 (Konertz et al., 2003). Based on these preliminary data, using a similar randomization schedule (2:1) and a similar common Table I. Modified mafs scoring method for adhesions. Finding mafs score Severity grade Tenacity none, extent none 0 None Tenacity filmy, extent,25% 1 Mild Tenacity filmy, extent 25 50% 2 Mild Tenacity filmy, extent 51% 3 Moderate Tenacity dense, extent,25% 4 Moderate Tenacity dense, extent 25 50% 5 Severe Tenacity dense, extent 51% 6 Severe SD for both groups (4.25), it was determined that 40 Hydrogel and 20 control patients would be sufficient to declare a difference of 3.4 in mean adhesions scores statistically significant (a ¼ 0.05, 80% power). Assuming that 10% of the patients would be excluded from the analysis, a minimum of 66 patients were to be enrolled (44 Hydrogel, 22 control). Baseline characteristics were summarized by treatment group for all patients who were randomized and received the treatment to which they were randomized. The primary population for efficacy analyses was the intention-to-treat (ITT) population which included all patients who were randomized, received the treatment to which they were randomized, and for whom there was a post-baseline assessment (i.e. they underwent the second-look procedure for the assessment of postoperative adhesions). Comparative analyses between Hydrogel and control groups were performed on both the raw mafs scores at the primary and second-look surgeries, and on the change from baseline scores, where the change from baseline ¼ second-look surgery score2primary surgery score. Analysis of variance was used to assess treatment group differences in the means after blocking for center. Various post hoc subset analyses were performed based on baseline characteristics. These included comparisons between Hydrogel and control groups for the categories of: primary surgery via laparotomy and via laparoscopy; removal of single and multiple myomas, and previous and no previous abdominal surgeries. The paired t-test was used to test the null hypothesis that the mean change from baseline ¼ 0. The Cochran Mantel Haenszel chi-square was used to assess treatment group differences after controlling for center for categorical variables. Safety outcomes were summarized by treatment group for all patients who were randomized and received the treatment to which they were randomized; in addition, all non-randomized test patients for the application of Hydrogel were included in the safety population. Results Patients Seventy-six women were enrolled in this study and underwent the primary surgery (Fig. 1). The study was initiated on 14 July 2003 the date of the first visit of the first subject; the study completed on 25 January 2005 the date of the last visit of the last subject enrolled. Four non-randomized patients served as test patients for the application of Hydrogel. One patient was mistakenly randomized and withdrawn prior to receiving treatment. A total of 71 patients were randomized and received the intervention to which they were randomized. Of these, 48 patients received Hydrogel and 23 received the standard of care (i.e. the control intervention). Ten Hydrogel and 3 control patients withdrew or were withdrawn from the study prior to attending the second-look surgery; most of them (9 Hydrogel, 2 control) withdrew consent to undergo the second-look procedure. As a result, post-baseline efficacy data were available for 58 patients (38 Hydrogel, 20 control) i.e. 81.7% of those who were randomized and received the study intervention. Patients in the Hydrogel and control groups had similar baseline characteristics (Table II). Participants had a mean age of 35 years; 42.3% (30/71) had prior abdominal or pelvic surgery that may have affected adhesion formation. Only 3 patients had taken any medications within the week prior to surgery that might have affected adhesion formation (1 subject in 1095

4 Mettler et al. Figure 1: CONSORT flow diagram Table II. Summary of demographic, medical and surgical history. Characteristic Hydrogel (n ¼ 48) Control (n ¼ 23) Age (years), Mean + SD Age range (years), min max Prior abdominal/pelvic surgery No, n (%) 27 (56.3) 14 (60.9) Yes, n (%) 21 (43.8) 9 (39.1) Surgeries reported by 4 Patients Laparoscopy 8 7 Laparotomy 2 2 Hysteroscopy 2 2 Myomectomy 3 1 Appendectomy 3 1 Conditions that may affect adhesion 0.94 formation No, n (%) 40 (83.3) 18 (78.3) Yes, n (%) 8 (16.7) 5 (21.7) Conditions reported by 2 Patients Asthma 3 1 Extra uterine pregnancy 1 3 Uterine myoma each group had used a steroid, 1 Hydrogel subject had used a medication listed as other ). Most patients underwent the primary surgery by laparoscopy: 34/48 Hydrogel patients (70.8%) and 14/23 control patients (60.9%). Drains were used at the treatment site in 30% of cases. The myomas removed were most commonly described as subserosal or intramural. Most patients had at least one myoma removed from a posterior site: 41/48 Hydrogel patients (85.4%) and 19/23 control patients (82.3%). Adhesiolysis was more commonly performed among patients in the Hydrogel group (15/48, 31.3%) than in the control group (3/23, 13.0%). Six Hydrogel patients (12.5%) had baseline adhesions at the posterior uterus; all six underwent adhesiolysis at this site. No control patients had baseline adhesions at this site. Only two Hydrogel patients (4.2%) were treated at all 15 anatomical sites examined; most (34/48, 70.8%) were treated at three or fewer sites. The most commonly treated site was the posterior uterus: 97.9% of Hydrogel patients were treated there. Considerably fewer patients were treated at any of the remaining anatomical sites.

5 Hydrogel for adhesion reduction post-myomectomy Efficacy outcomes Hydrogel and control patients were similar in the median number of days between primary and second-look surgeries: 54 and 57 days for Hydrogel and control patients, respectively. Hydrogel and control patients did not differ significantly in their mean baseline mafs scores recorded at the primary surgery in any of the data sets examined (data only shown for the overall analysis). In the overall analysis (Fig. 2), the mean change from baseline mafs score (+SD) experienced by Hydrogel patients was less than that experienced by control patients ( versus , P ¼ 0.01); 95% confidence intervals for these mean changes were ( ) and ( ), respectively. Similarly, for all subsets examined (Fig. 3), the mean changes from baseline mafs scores for Hydrogel patients were less than those of control patients; differences between groups were statistically significant for four of the six subsets examined. Significant differences were evident among those who underwent the primary surgery via laparotomy, those who had a single myoma removed, those who had multiple myomas removed and those who had no previous abdominal surgeries. Although these P-values are considered descriptive in nature, they are indicative of the overall treatment effect of Hydrogel. At the time of the primary surgery, the extent and tenacity of adhesions at the posterior uterus site were categorized as none for almost all patients. Only 3 patients who attended the second-look surgery had adhesions at baseline of any extent or tenacity: all were in the Hydrogel group (7.9 versus 0.0% in the control group). At the second-look surgery, most Hydrogel patients (25/38, 65.8%) still had adhesions with an extent and a tenacity categorized as none. In contrast, only 35% of control patients (7/20) had adhesions with an extent and tenacity categorized as none. While the difference in the distribution of patients among the four extent score categories (none,,25, 25 50, 51%) between Hydrogel and control groups was not statistically significant (P ¼ 0.138), the difference in the distribution of patients among the three tenacity score categories (none, filmy, dense) was (P ¼ 0.032). Figure 2: Mean mafs score at the posterior uterus, overall analysis ITT population (n ¼ 58). Analysis of variance was used to assess treatment group differences in the means after blocking by center. The primary surgery data are presented for all patients including those who discontinued prior to attending the second-look surgery. ITT, intention-to-treat; mafs, modified American Fertility Society. Control-light shading, Hydrogelheavy shading Safety outcomes The Safety Population (n ¼ 75) for this study included the 71 patients who were randomized and received the treatment to which they were randomized, and the 4 non-randomized patients who served as test patients for the application of Hydrogel. That is, the safety population consisted of 52 Hydrogel patients (48 who were randomized and received the intervention, and 4 test patients who received the Hydrogel prior to randomization) and 23 control patients. A total of 11 adverse events were reported for 9 patients during the course of this study (Table III): 6 events were reported for 5 patients (5/52, 9.6%) in the Hydrogel group; and 5 events were reported for 4 patients (4/2317.4%) in the control group. Among the Hydrogel group, dizziness and vomiting were reported for one subject, vomiting alone was reported for two patients, turning dizziness was reported for one subject and low grade fever was reported for one subject. Among the control group, fever and thrombocytopenia (note that thrombocytes had returned to normal within 12 h) were reported for one subject, fever alone was reported for one subject, a urinary tract infection was reported for one subject and laceration of a small collateral vessel at the time of the second-look surgery was reported for one subject. The most commonly reported events were vomiting for Hydrogel patients (5.8%) which represented 50% of all reported events in this group and fever for control patients (8.7%) which represented 40% of all reported events for this group. All events were rated either mild or moderate in severity and all were judged by the investigator to be either definitely not or unlikely related to the investigational device (Table IV). All events were reported as resolved with treatment. One event in each group was identified as serious: turning dizziness in a Hydrogel subject and fever in a control subject. Both patients, for whom serious adverse events had been reported, attended the second-look surgery and completed the study. The event of turning dizziness in a Hydrogel subject occurred eight days after the primary surgery. The investigator has since clarified that the event lasted 5 min, resolved without treatment, and that the subject chose to remain in hospital of her own volition. The event was considered by the investigator to be definitely not related to the investigational device, and was coded as a nervous system disorder and listed by the preferred term of dizziness. The event of a408c fever in a control subject occurred on the day of the primary surgery which was performed by laparotomy and during which four myomas had been removed. The event was considered by the investigator to be unlikely related to treatment, possibly related to the laparotomy procedure, and was coded as a general disorder and listed by the preferred term of pyrexia. The event resolved within 18 days, following the draining of a subcutaneous hematoma. No intra-abdominal infections and no post-operative infections at the operative site were reported. No patients were withdrawn from this study due to an adverse event. There were no deaths. Discussion The ideal anti-adhesion product has been described by various authors as one that prevents de novo adhesion formation as well 1097

6 Mettler et al. Figure 3: Mean change from baseline mafs score at the posterior uterus, six subset analyses, ITT population (n ¼ 58). Analysis of variance was used to assess treatment group differences in the means after blocking by center. Control-light shading, Hydrogel-heavy shading as adhesion reformation following adhesiolysis, without promoting an inflammatory tissue response, supporting bacterial growth or affecting normal wound healing. Post-operative adhesions in myomectomy and infertility patients are of great importance. They may interact in the fertilization process, hinder ovulation, block tubal patency, form pseudo cysts around the ovary and cause pelvic pain. In addition, the agent should be resorbable, and deliverable by the laparotomic and laparoscopic approaches. With these criteria in mind, a resorbable hydrogel barrier product was designed that could be dispersed directly and via a laparoscopic delivery system. Data from the current study suggest that the use of Hydrogel represents a significant advance in the field of post-operative adhesion prevention. The Hydrogel proved to be an effective intervention for the reduction of adhesion formation following myomectomy. Based on the revised primary efficacy measure of the mafs score at the posterior uterus site, adhesion formation among Hydrogel patients was 3-fold less than that among control patients ( versus ) a Table III. Adverse events by system organ class and preferred term. Hydrogel (n ¼ 52) Control (n ¼ 23) Number of reported adverse events 6 5 Number of patients reporting adverse events, n (%) 5 (9.6) 4 (17.4) Blood and lymphatic system disorders, n (%) 0 (0.0) 1 (4.3) Thrombocytopenia, n (%) 0 (0.0) 1 (4.3) Gastrointestinal disorders, n (%) 3 (5.8) 0 (0.0) Vomiting NOS, n (%) 3 (5.8) 0 (0.0) General disorders and administrative site 1 (1.9) 2 (8.7) conditions, n (%) Pyrexia, n (%) 1 (1.9) 2 (8.7) Infections and infestations, n (%) 0 (0.0) 1 (4.3) Urinary tract infection NOS, n (%) 0 (0.0) 1 (4.3) Nervous system disorders, n (%) 2 (3.8) 0 (0.0) Dizziness, n (%) 2 (3.8) 0 (0.0) NOS, not otherwise specified. All percentages are column percentages based on the number of patients in each group of the Safety population (n ¼ 75). The event of laceration of a small collateral vessel, which occurred at the time of the second-look surgery in one control subject, was not coded. difference that was statistically significant (P ¼ 0.01). Although not always statistically significant, similar trends were observed in all subsets examined. Hydrogel appeared to have been of some benefit in reducing adhesion formation in circumstances where the degree of trauma was relatively high and thus the propensity for scarring was greatest i.e. in surgeries performed via open laparotomy, in cases where multiple myomas were removed, and in patients who had undergone previous surgeries (mean changes from baseline mafs score for Hydrogel patients compared with controls were, respectively, 2.7, 21 and 2.2-fold less, P ¼ 0.03, 0.03 and 0.19, respectively). Similarly, Hydrogel appeared to have been of some benefit in situations where the degree of surgical trauma was relatively small i.e. in laparoscopic surgeries, in cases where a single myoma was removed, and in first surgeries (mean changes from baseline mafs score for Hydrogel Table IV. Adverse events by severity, relationship to treatment and outcome. Hydrogel (n ¼ 52) Control (n ¼ 23) Number of patients reporting adverse events, 5 (9.6) 4 (17.4) n (%) Number of reported adverse events 6 5 Number of serious adverse events 1 1 Number of adverse events by severity, n (%) Mild 2 (33.3) 4 (80.0) Moderate 4 (66.7) 1 (20.0) Severe 0 (0.0) 0 (0.0) Number of adverse events by relationship to treatment, n (%) Definitely 0 (0.0) 0 (0.0) Likely 0 (0.0) 0 (0.0) Unlikely 1 (16.7) 2 (40.0) Definitely not 5 (83.3) 3 (60.0) Number of adverse events by outcome, n (%) Resolved with treatment 5 (83.3) 5 (100.0) Ongoing treatment 1 (16.7) 0 (0.0) Death 0 (0.0) 0 (0.0) Unknown 0 (0.0) 0 (0.0) The percentage of patients reporting adverse events is based on the number of patients in each group of the safety population (n ¼ 75). All other percentages are based on the number of reported adverse events for each group. 1098

7 Hydrogel for adhesion reduction post-myomectomy patients compared with controls were, respectively, 3, 2.2 and 4.8-fold less, all P ¼ 0.03). Almost twice as many Hydrogel patients as control patients were adhesion-free at the second-look procedure (65.8 versus 35.0%). That is, despite all patients having had a minimum 2 cm incision at the posterior uterus and most (85%) having had myomas removed from the posterior uterus, only about one-third of Hydrogel patients had any evidence of postoperative adhesion formation at this site within a mean of 60 days after the primary surgery. While our analyses focused specifically on the posterior uterus, it is conceivable that similar outcomes would have been achieved at other anatomical sites had all patients been at equally high risk for postoperative adhesion formation at these sites as well. That 66% of Hydrogel-treated patients in our study were adhesion-free at the posterior uterus site is particularly interesting in light of the findings of Tulandi et al. (1993). In their study in which no adhesion prevention interventions were undertaken, 93.7% of myomectomy incisions on the posterior uterine wall were associated with adnexal adhesions i.e. only 6.3% of patients were adhesion-free at this site. Also of interest, women with incisions at this site were at significantly greater risk for adhesions than those with incisions on the fundal or anterior uterine wall. As a point of reference, the incidence of post-operative adhesion formation associated with a similar type of adhesion barrier (i.e. SprayGel TM ) that is also deployable laparoscopically may be of interest. In this report of women who had undergone open or laparoscopic myomectomy, the incidence of post-operative uterine adhesions was 68.2% in treated patients and 88.9% in control patients, 3 16 weeks after the primary surgery (Mettler et al., 2004); these incidences were not statistically significantly different. In addition, the median adhesion covered area of the uterus was similar in control and treatment patients. However, the mean adhesion tenacity scores as well as the severity and the area affected were three times lower in the treatment group than in controls. It should be noted that, while the incidence of post-operative adhesion formation associated with SprayGel TM was approximately twice that reported for Hydrogel, the data are not strictly comparable since, in the current study, reporting of adhesions was limited to the posterior uterus, and did not include other uterine surfaces. The Hydrogel proved to be a safe intervention for the reduction of adhesion formation following myomectomy. Not only were the adverse events reported in this study unremarkable but, unlike some of the early investigational anti-adhesion products studied, no intra-abdominal infections or postoperative infections at the operative site were associated with Hydrogel use. One major limitation of this probative study was the lack of full blinding. Although participants in this study were blinded to their treatment, the study investigators who performed the surgeries, applied the randomized treatment, and evaluated adhesion formation during both the primary and second-look surgeries, were not. Blinding of the investigator was impractical in this type of study. A post hoc attempt to provide a fully blinded assessment of adhesions, using the videotapes recorded at the time of the surgeries, proved to be impossible. Many of the videotapes were not sufficiently clear at all 15 anatomical sites, including the posterior uterus, for this task to be performed with a high degree of accuracy. Furthermore, in some cases, videotapes were never taken. Thus, we had to rely solely on the assessments performed at the time of the surgery; inter-rater reliability testing of the surgeons scoring of adhesions was not performed. Given that patients were blinded throughout the study, it is unlikely that bias would have been introduced in the reporting of adverse events i.e. the primary endpoint for this study. However, the lack of full blinding, combined with the subjective nature of the adhesion assessment process, could have potentially introduced an element of bias in the efficacy outcomes. In future studies, it would be preferable to ensure that high-quality videographic tapes were sent to blinded readers who would independently assess adhesion formation. In terms of its clinical implications, this study adds significantly to the body of work in the ongoing search for effective interventions that will ultimately eliminate post-operative adhesion formation. Given the high morbidity and costs associated with this problem, investigation in this field continues to be a high priority. It will be of interest to see whether Hydrogel proves to be safe and effective for the reduction of postoperative adhesion formation following other types of uncomplicated abdominal surgeries. Conclusion This 71-patient study provides the first clinical evidence for the safety and efficacy of this resorbable hydrogel for the reduction of post-operative adhesions in women undergoing myomectomy surgery, performed either by laparotomy or laparoscopy. Acknowledgements The authors gratefully acknowledge the contributions of the study co-investigators: Dr Jacques Capello, Dr Marek Zygmunt, Dr Ulrich Fuellers and Dr Dominik Leitsch. CoSealw is a registered trademark belonging to an Angiotech company. Funding This study was funded by Angiotech Pharmaceuticals, Inc. References Al-Jaroudi D, Tulandi T. Adhesion prevention in gynecologic surgery. Obstet Gynecol Survey 2004;59: Canis M, Mage G, Wattiez A, Chapron C, Pouly JL, Bassil S. Second-look laparoscopy after laparoscopic cystectomy of large ovarian endometriomas. Fertil Steril 1992;58: DeCherney AH, dizerega GS. Clinical problem of intraperitoneal postsurgical adhesion formation following general surgery and the use of adhesion prevention barriers. Surg Clin North Am 1977;77: Diamond MP, DeCherney AH. Pathogenesis of adhesion formation/ reformation: application to reproductive pelvic surgery. Microsurg 1987;8:

8 Mettler et al. Diamond MP, Daniell JF, Feste J, Surrey MW, McLaughlin DS, Friedman S, Vaughn WK, Martin DC. Adhesion reformation and de novo adhesion formation after reproductive pelvic surgery. Fertil Steril 1987;47: Ellis H, Moran BJ, Thompson JN, Parker MC, Wilson MS, Menzies D, McGuire A, Lower AM, Hawthorn RJ, O Brien F et al. Adhesion-related hospital readmissions after abdominal and pelvic surgery: a retrospective cohort study. Lancet 1999;353: Hendrix M, Mees U, Hill AC, Egbert B, Coker GT, Estridge TD. Evalulation of a novel synthetic sealant for inhibition of cardiac adhesions and clinical experience in cardiac surgery procedures. Heart Surg Forum 2001;4: Konertz WF, Kostelka M, Mohr FW, Hetzer R, Hübler M, Ritter J, Liu J, Koch C, Block JE. Reducing the incidence and severity of pericardial adhesions with a sprayable polymeric matrix. Ann Thorac Surg 2003;76: Mettler L. Manual for Laparoscopic and Hysteroscopic Gynecological Surgery. New Delhi, India: Jaypee Brothers Medical Publishers (P) Ltd, Mettler L, Audebert A, Lehmann-Willenbrock E, Schive-Peterhansl K, Jacobs VR. A randomized, prospective, controlled, multicenter clinical trial of a sprayable, site-specific adhesion barrier system in patients undergoing myomectomy. Fertil Steril 2004;82: Monk BJ, Berman ML, Montz FJ. Adhesions after extensive gynecologic surgery: clinical significance, etiology, and prevention. Am J Obstet Gynecol 1994;170: Operative Laparoscopy Study Group. Postoperative adhesion development after operative laparoscopy: evaluation at early second-look procedures. Fertil Steril 1991;55: Ray NF, Denton WG, Thamer M, Henderson SC, Perry S. Abdominal adhesiolysis: inpatient care and expenditures in the United States in J Am Coll Surg 1998;186:1 9. Steege JF, Stout AL. Resolution of chronic pelvic pain after laparoscopic lysis of adhesions. Am J Obstet Gynecol 1991;165: Tulandi T, Murray C, Guralnick M. Adhesion formation and reproductive outcome after myomectomy and second-look laparoscopy. Obstet Gynecol 1993;82: Submitted on July 18, 2007; resubmitted on January 22, 2008; accepted on February 22,