Effect of endometrioma cyst fluid exposure on peritoneal adhesion formation in a rabbit model

Transcription

1 Effect of endometrioma cyst fluid exposure on peritoneal adhesion formation in a rabbit model Laura Proud Smith, M.D., a Christopher D. Williams, M.D., b Joseph O Brien Doyle, M.D., a Wendy B. Closshey, M.D., a William K. Brix, M.D., c and Lisa M. Pastore, Ph.D. a Departments of a Obstetrics and Gynecology, b Obstetrics and Gynecology, Division of Reproductive Endocrinology and Infertility, and c Pathology, University of Virginia Health System, Charlottesville, Virginia Objective: To determine whether copious lavage and suction of human endometrioma fluid placed in the peritoneal cavity of rabbits reduces adhesion formation compared to no lavage. Design: Prospective, randomized, blinded study. Setting: Academic research environment. Animal(s): Twenty-four female New Zealand white rabbits. Intervention(s): Rabbits randomized into three groups: [1] laparoscopy with instillation of human endometrioma material, no lavage; [2] laparoscopy with instillation of human endometrioma material, followed by clearance of all visible endometrioma fluid by saline lavage and suction; [3] laparoscopy alone. Main Outcome Measure(s): Six weeks after laparoscopy, adhesions scored by laparotomy using standard visual assessment scoring system and histologic microscopic evaluation. Data evaluated using Kruskal-Wallis and median nonparametric tests. Result(s): For groups 1, 2, and 3, respectively, mean total clinical adhesion scores were 0.67 (95% confidence interval [CI] 0.87, 2.2), 3.67 (95% CI 1.27, 6.07), and 0 (95% CI 0, 0). Group 2 had statistically significantly higher mean adhesion scores compared to group 1. Histologic adhesion scores followed the trend of clinical adhesion scores. Conclusion(s): In this rabbit model, human endometrioma fluid exposure in the peritoneal cavity is not associated with adhesion formation. Instillation of endometrioma fluid followed by copious saline lavage is strongly associated with adhesion formation. (Fertil Steril 2007;87: by American Society for Reproductive Medicine.) Key Words: Adhesion, endometrioma, rabbit model, lavage Received November 16, 2005; revised and accepted August 11, Present addresses: Dr. Joseph O Brien Doyle, Brigham and Women s Hospital, Department of Obstetrics and Gynecology, Boston, Massachusetts; and Dr. Wendy B. Closshey, University of San Francisco, Department of Obstetrics and Gynecology, Division of Reproductive Endocrinology and Infertility, San Francisco, California. Reprint requests: Christopher D. Williams, M.D., Reproductive Medicine and Surgery Center of Virginia, Martha Jefferson Outpatient Care Center, 595 Peter Jefferson Parkway, Suite 390, Charlottesville, VA (FAX: ; Christopher.Williams@mjh.org). Endometriosis, a disease involving ectopic endometrial tissue, is widely prevalent, affecting 4% 44% of asymptomatic women, 40% 60% of women with dysmenorrhea, and 20% 30% of subfertile women (1 3). Endometriomas, cystic collections of endometriosis of the ovary, occur in 17% 44% of those with endometriosis (1, 3). These cysts can become large and painful, requiring surgical management. Multiple surgical techniques are used to treat endometriomas. A common element to these approaches involves intentional cyst rupture, spillage of the contents of the cyst into the pelvic cavity, removal of the cyst wall, followed by aggressive irrigation and suction (4, 5). Laparoscopic cystectomy is an effective and widely used surgical technique for treating endometriomas, with 92% of patients with endometriomas who undergo the procedure showing no evidence of ovarian endometrioma on second-look laparoscopy within 3 6 months (1, 6). Although no statistically significant difference in complications or recurrence rate is evident, cystectomy has been shown to have superior outcomes relative to drainage and coagulation for pain relief, cumulative postoperative pregnancy rates (PR), and the disease-free interval (7). The recurrence rate, regardless of surgical technique, appears to correlate with the duration of follow-up (8). Rupture of the cyst with spillage of the contents into the pelvic cavity is contrary to the management of some other types of ovarian cysts. For instance, avoidance of cyst rupture during resection of mature teratomas is considered important (9). The perception is that copious irrigation and suction of released endometrioma contents reduces adhesion formation; however, there is a paucity of clinical literature to evaluate this assumption. The goal of this study was to determine whether copious lavage and suction of human endometrioma fluid in the peritoneal cavity of rabbits reduces adhesion formation compared to no lavage. MATERIALS AND METHODS Subjects and Study Design Approval for the study was obtained from the Animal Care and Use Committee and the Institutional Review Board /07/$32.00 Fertility and Sterility Vol. 87, No. 5, May 2007 doi: /j.fertnstert Copyright 2007 American Society for Reproductive Medicine, Published by Elsevier Inc. 1173

2 (IRB) at the University of Virginia Health System. Twentyfour sexually mature female New Zealand white rabbits were obtained from a commercial source (Burleson Enterprises, Inc., Unionville, VA) and were observed for 8 days to assess general health status before experimentation. Animal weights ranged from 3,500 to 5,500 g. The animals were maintained on Teklad Global High Fiber Rabbit Diet (Harlan, Madison, WI) and water ad libitum. The rabbits were randomized into three experimental groups. The three experimental groups underwent the following procedures: [1] laparoscopy with pelvic instillation of 2 ml of human endometrioma material, no lavage; [2] laparoscopy with pelvic instillation of 2 ml of human endometrioma material, followed by clearance of all visible endometrioma fluid with copious normal saline lavage and suction; [3] laparoscopy without instillation of endometrioma material or lavage (surgical control). For our results to have 90% confidence and 80% power in detecting differences by treatment group, we determined the need for nine rabbits in group 1, nine in group 2, and six in group 3. This sample size was based on the assumption of an adhesion score of 1.0 for group 2 and an adhesion score of 3.6 for group 1 on a subscale that ranged up to 4.0. These assumptions were based on the adhesion results demonstrated by Fiedler et al. (10) and our a priori judgment. An ear tag bearing a unique four-digit identification number was placed on each animal by the supplier. These identification numbers were used throughout the study to identify each animal. The initial surgery was performed during a 2-day period. On the day of the surgery, the animal care coordinator, who was not involved in the surgery or adhesion scoring, randomly selected rabbits to be assigned to the three treatment arms. Each animal s ear tag was covered before delivery of the animal to the surgeons, effectively blinding them to the animal s identity during surgery. Laparoscopies were performed under sterile conditions with 2 ml of ketamine (100 mg/ml; Fort Dodge Animal Health, Fort Dodge, IA) and 1 ml of xylazine (20 mg/ml; Burns Veterinary Supply, Inc., Westbury, NY) anesthesia administered IM. Rabbits were masked with 100% oxygen during the procedure to avoid hypoxia and hypercapnia, and anesthesia was maintained with isofluorane to effect. Each rabbit received 1 ml of enrofloxacin (Bayer, Shawnee Mission, KS) as antibiotic prophylaxis before the surgery. The hair over the abdomen was shaved and the skin prepared for an aseptic procedure by three alternating scrubs of providone iodine and 70% alcohol. Either a 10-mm or 5-mm laparoscope with an attached videoscopic camera setup was used on an alternating basis from one rabbit to the next to expedite surgical turn-around time while equipment was sterilized. The laparoscope was introduced into the upper abdomen through an open technique. Manual elevation of the abdominal wall allowed adequate visual inspection of the entire abdominal cavity to confirm normal anatomy and the absence of adhesions. A 60-mL container of refrigerated endometrioma material (see section on Human Endometrioma Cyst Fluid) was warmed to room temperature. In groups 1 and 2, 2 ml of endometrial material were instilled using a 6-mL syringe, such that the rabbits were exposed to a weight-adjusted amount of endometrial tissue grossly equivalent to potential human exposure. The capped container of endometrioma fluid was mixed before aspiration into the syringe to ensure that all aliquots injected would be expected to have identical contents. The syringe was introduced through the suprapubic laparoscopic port and directed toward the pelvis. For the animals that received normal saline irrigation (group 2), a 60-mL syringe was introduced through the same laparoscopic port to perform repeated lavage using normal saline until there was no visual evidence of residual endometrioma material on laparoscopy. Each animal in group 2 underwent a total of 10 rounds of normal saline lavage. The force of suction lavage used on each animal was kept constant by a single surgeon doing the lavage on every animal and the time of evacuation of each 60-mL syringe was likewise kept constant. After the procedure, the peritoneum and fascia were reapproximated with 3-0 polyglactin 910 sutures in a running full-thickness stitch. The skin was closed with 4-0 polyglactin 910 sutures (Ethicon Inc., Cornelia, GA) in a running subcuticular stitch with a buried knot. Bupivicaine 0.25% was injected locally for postoperative analgesia. After the animals were maintained for 6.5 weeks under standard husbandry conditions, each animal underwent a laparotomy for adhesion scoring. The order of the rabbits undergoing surgery was randomly selected, and surgeons remained blinded to the treatment group. Animals were anesthetized with 1.5 ml of ketamine (100 mg/ml) and 1.5 ml of xylazine (100 mg/ml) administered IM and euthanized by lethal intracardiac injection with 2 ml of Euthanasia Solution (5 g/ml, 1 ml/10 lbs; Virbac AH Inc., Fort Worth, TX), per standard rabbit euthanasia protocol. During the laparotomy, the clinical appearance of the pelvis was scored by three investigators (L.P.S., C.D.W., J.O.D.) according to the standard visual assessment scoring system of Fiedler et al. (10). Biopsies were taken from representative areas on the ovary, uterine horn, peritoneal sidewall, and bladder of each animal and preserved in formalin. Human Endometrioma Cyst Fluid Approval was obtained from the Human Investigation Committee at the University of Virginia Health System for the use of human endometrioma cyst fluid. Endometrioma cyst fluid was collected in the operating room under sterile conditions from a single large endometrioma from an intact human pathological specimen. A single human female gave informed consent and served as the source of all pathologically confirmed endometrioma cyst fluid. At the time of a routine laparoscopy for the excision of a large endometrioma, endometrioma cyst fluid was aspirated directly from the cyst cavity. The contents of the cyst were drained into a sterile container approved by the Institutional Biosafety 1174 Smith et al. Endometrioma cyst fluid and adhesions Vol. 87, No. 5, May 2007

3 TABLE 1 Criteria for clinical scoring of adhesions. Category Description Score Extent No intraabdominal adhesions 0 25% of abdomen involved 1 50% of abdomen involved 2 75% of abdomen involved 3 75% of abdomen involved 4 Type No adhesions 0 Filmy, transparent, avascular 1 Opaque, translucent, avascular 2 Opaque, capillaries present 3 Opaque, larger vessels present 4 Tenacity No adhesions 0 Adhesions essentially fall apart 1 Adhesions lysed with traction 2 Adhesions require sharp dissection 3 Inflammation None 0 Mild erythema, localized surface involvement 1 Moderate erythema and edema, localized surface involvement 2 Severe erythema and edema, localized surface involvement 3 Severe erythema and edema, widespread surface involvement 4 Maximum total score 15 Committee. The container of endometrioma cyst fluid was immediately stored at minus 80 C until surgery. Strict sterile technique was maintained during the handling of the material. Adhesion Scoring An abdominal laparotomy was carried out on each animal to evaluate and score adhesion formation according to the scoring system in Table 1, as used by Fiedler et al. (10). The same three surgeons performed the laparotomy and dissection on each animal and the surgeons remained blinded to each animal s assigned group (L.P.S., C.D.W., J.O.D.). These three investigators jointly scored the adhesions. Photographs were taken of the peritoneal cavity of each rabbit. Histologic Evaluation of Rabbit Biopsies After laparotomy and clinical assessment of adhesion scores, representative biopsies were taken from each animal, fixed in 10% buffered formalin, embedded in paraffin, cut at 4 m, and stained with hematoxylin and eosin (H & E). Biopsies were taken from representative areas on the ovary, uterine horn, fallopian tubes, abdominal sidewall, and bladder. The slides were reviewed by two pathologists (K.A. and W.B.) who were blinded to the procedure. All cases were assessed for acute and chronic inflammation, abscess formation, granulation tissue, and fibrosis. The histologic scoring system used by Fiedler et al. (10) was used. Acute inflammation was graded on a four-point scale: 0 inflammation absent; 1 focal mild inflammation with scattered, fewer than 10 neutrophils per 400 field; 2 moderate inflammation with neutrophils per 400 field; 3 widespread severe inflammation with greater than 100 neutrophils per 400 field. Chronic inflammation was graded on a four-point scale: 0 inflammation absent; 1 focal mild inflammation with scattered lymphocyte and plasma cells per 400 field; 2 moderate inflammation with less than 5 follicles per 400 field; 3 widespread severe inflammation with greater than 5 follicles per 400 field. Abscess formation, granulation tissue, and fibrosis were graded as present or absent. The scores in each category were added for each animal and a final histologic score assigned to each rabbit. Statistical Analysis The mean, 95% confidence interval around the mean, standard error, and median total adhesion scores by group were calculated. The groups were then further analyzed with the Kruskal-Wallis one-way analysis of variance by ranks, the median test, and 2 analysis with P values of less than.05 showing statistical significance. All analyses were done using SPSS version 11 (SPSS, Inc., Chicago, IL). These nonparametric tests were selected as appropriate because of the small sample size. Fertility and Sterility 1175

4 TABLE 2 Comparison of group means for clinical adhesion scores. Group 1 (n 9) a Group 2 (n 9) b Group 3 (n 6) c Mean (95% CI) 0.67 ( 0.87, 2.2) 3.67 (1.27, 6.07) 0 (0, 0) Median Standard error P value d.027 (1 vs. 2).017 (3 vs. 2).414 (3 vs. 1) a Laparoscopy and endometrioma fluid instillation. b Laparoscopy, endometrioma fluid, and lavage. c Laparoscopy alone. d Data evaluated with Kruskal-Wallis one-way analysis of variance. RESULTS All 24 animals survived the surgery. There were no surgical complications, significant postoperative morbidity or mortality before euthanasia. One rabbit experienced two episodes of diarrhea 2 days postoperatively and stool was sent for bacterial culture and parasites, both of which returned negative. This rabbit received 5 days of empiric sulfamethoxazole and trimethoprim (15 mg/kg; Alpharma USPD Inc., Baltimore, MD), and the diarrhea resolved. The mean and median clinical adhesion scores are shown in Table 2 as well as in Figure 1. The mean adhesion scores were significantly higher in group 2 than group 1 (P.027). The mean adhesion scores were lower in group 3 than group 1 FIGURE 1 Total clinical adhesion scores. Quantitative analysis of adhesions using system shown in Table 1. Results are given for each animal with animals divided by group: group 1: laparoscopy and endometrioma fluid instillation; group 2: laparoscopy, endometrioma fluid instillation, and saline lavage; group 3: laparoscopy alone. (P.414). The mean adhesion scores were significantly lower in group 3 than group 2 (P.017). The animals in group 3 were found to have no visible adhesions. The clinical adhesion score of rabbit 2 in group 2 was statistically significantly higher than that of any other animal in the group despite rigid consistency in amount of endometrioma fluid instilled and no difference in the content of endometrioma cyst fluid between animals. Histologic scores are shown in Table 3. The total histologic scores were higher in group 2 than group 1 (P.19). The total histologic scores were lower in group 3 than group 1 (P.11). The total histologic scores were lower in group 3 than group 2 (P.61). The fallopian tubes, ovaries, and tissues of the abdominal sidewall (skeletal muscle and the associated fat) were the sites most commonly affected by the procedure. The most common finding was a mild-to-moderate chronic inflammation. The histologic results follow the trend demonstrated in the clinical scoring of adhesions, but are not of statistical significance. DISCUSSION The most common surgical technique for treating endometriomas involves intentional rupture and spillage of the contents of the cyst into the pelvic cavity followed by extensive irrigation and suction (7, 11). In humans, the perception has been that copious irrigation and suction of the pelvis after spillage of endometrioma contents intraperitoneally reduces adhesion formation. However, there is a lack of evidence in the medical literature to support this assumption. No study has used an animal or human model to prospectively evaluate whether avoiding intraperitoneal spillage of endometrioma fluid or thorough lavage of endometrioma spillage reduces postoperative adhesion formation. A human study comparing adhesions and peritoneal implants of endometriosis before and after laparoscopic treatment of endometriomas showed that most new implants on second-look were on the pelvic floor as a result of chocolate-like material that could not be aspirated Smith et al. Endometrioma cyst fluid and adhesions Vol. 87, No. 5, May 2007

5 TABLE 3 Comparison of group means for histologic adhesion scores. Histologic category Group 1 (n 9) a Group 2 (n 9) b Group 3 (n 6) c Acute inflammation Chronic inflammation Abscess Granulation Fibrosis Total histologic score P value d.19 (1 vs. 2).11 (3 vs. 2).61 (3 vs. 1) a Laparoscopy and endometrioma fluid instillation. b Laparoscopy, endometrioma fluid instillation, and lavage. c Laparoscopy alone. d Data evaluated with Kruskal-Wallis one-way analysis of variance. Adhesions were more frequent and extensive in the endometrioma excision group on second-look compared to the control group (5). In our study, rabbits were chosen as the study animal to avoid repeating baseline experimentation. One group of rabbits was used to control for inflammatory consequences of laparoscopy alone. A normal saline lavage control was not believed to be necessary because normal saline lavage has been demonstrated not to be associated with either clinical or histologic adhesion formation in a rabbit model (10). This study attempts to replicate the method of saline lavage done in the study by Fiedler et al. (10), but there were some differences in the techniques. In the Fiedler study, the lavage alone group was exposed to 10 rounds of 100 ml of normal saline lavage for a total of 1Lofirrigation through a Corson suction irrigator. The present study used a 60-mL syringe to expose animals to 10 rounds of normal saline irrigation for a total of 600 ml of lavage and we did not use a formal suction irrigator due to equipment malfunction. This difference is important because the smaller volume of irrigant used in the present study could potentially have allowed for residual endometrioma fluid. However, no endometrioma fluid was visible laparoscopically at the end of the procedure, the same standard that is currently used in the surgical treatment of endometriomas in humans. The type of irrigator may also have affected the formation of adhesions and it may be that the commercial laparoscopic suction irrigator used by Fiedler et al. created more force and more thoroughly rinsed endometrioma fluid than the hand-held syringe technique used in the present study. We proceeded despite the differences in suction irrigation technique because normal saline lavage had not been shown to be associated with adhesions in the Fiedler animal model. The results of the study were unexpected. The current surgical management of endometriomas in humans involves rupture with spillage of endometrioma fluid, followed by copious saline irrigation and aspiration to remove all visible fluid. In line with current practice, we hypothesized that the experimental group without copious lavage of endometrioma fluid (group 1) would have the highest mean clinical adhesion scores. To the contrary, this study demonstrates that the group with copious normal saline lavage of endometrioma fluid (group 2) had significantly more clinical adhesions. These results were entirely unexpected as saline lavage has been previously shown to decrease peritoneal adhesion formation in both animal and human studies evaluating adhesion formation and histologic evidence of inflammation after human dermoid fluid spillage (10, 12 14). The histologic adhesion scores follow a trend that supports the clinical adhesion scores. The differences between histologic scores are not statistically significant, but the study was not powered to see differences by tissue or by histologic category. A possible explanation for the higher mean clinical adhesion scores in the lavage group is that the saline lavage itself served to spread the endometrioma fluid more effectively in the abdominal cavity, thereby increasing the distribution of tissue contact and potential for adhesions compared to local spill. This does not reconcile why more extensive localized adhesions were not present in group 1. Another possibility is that the process of lavage mechanically irritates the peritoneal cavity and thereby causes adhesions by local tissue damage by the suction instrument. This seems unlikely because data by Fiedler et al. (10) in a rabbit model evaluating adhesion formation after peritoneal instillation of human dermoid material found no clinical or histologic evidence of adhesion formation in their group evaluating saline lavage alone. Interestingly, Fiedler et al. (10) state that some statistical comparisons suggested that saline lavage after dermoid exposure resulted in higher adhesion scores than control conditions, although overall their data indicate that normal saline lavage of dermoid cyst contents in the perito- Fertility and Sterility 1177

6 neal cavity reduces inflammation and adhesion formation to near control levels. A separate concern is that although normal saline lavage has been shown to have no association with either clinical or histologic adhesion formation in a rabbit model, this baseline assumption may not be correct (10). Review of the literature reveals few prospective studies with divergent results that compare normal saline lavage to lactated Ringer s solution or other crystalloid, and associated adhesion formation. There is a body of literature asserting that normal saline irrigation in rats does not cause peritoneal inflammation or adhesions compared with a variety of other irrigants (15 17). But there is also evidence that peritoneal irrigation with lactated Ringer s solution is superior to irrigation with normal saline and that lactated Ringer s solution irrigation is protective against adhesions in both rat and rabbit models (18, 19). Furthermore, van Westreenen et al. (20) found enhanced peritoneal adhesion formation (P.0001) in rats lavaged with all solutions tested, including normal saline, compared to a surgical control. Close evaluation of the surgical methods described in many articles on the surgical management of ovarian endometriomas in humans exposes both the inconsistency in irrigation solutions used as well as the frequent complete absence of description of the manner or degree of irrigation other than simply that lavage was done (3 6, 11). Other limitations include that the use of a rabbit model may not offer data that extrapolates to humans. Given that human endometrioma fluid is foreign to the rabbit, the xenographic rejection phenomenon is a concern. Fiedler et al. (10) addressed this concern in a rabbit model evaluating adhesion formation after human dermoid tumor spillage by a study arm instilling human follicular fluid (FF) into the rabbit peritoneum as an antigenic control. The animals used in that study demonstrated no histologic evidence of inflammation or adhesions in those exposed to the human FF. Those animals who received intraperitoneal instillation of dermoid material survived the instillation of this material without evidence of organ failure, fevers, or other morbidity. We anticipated that instillation of human endometrioma fluid into the rabbit peritoneum would have less antigenicity than that of the dermoid tumor, which contains many more human cell types. In our study, none of the rabbits experienced adverse health effects attributable to the endometrioma fluid and none demonstrated evidence of tissue rejection. It is possible, however, that rabbits do not mount an immune response against the cells in human endometrioma fluid. A lack of immune response would explain the lack of inflammation and adhesions in the group with endometrioma fluid without lavage. Similarly, the use of a single human source of the endometrioma fluid may have affected the antigenicity of the endometrioma fluid if there were individual variability in the content of the cyst and the amount of inflammatory factors. It may not be the case that all human endometrioma fluid is equal in its ability to incite an inflammatory reaction in the rabbit. We were purposeful in our decision to use a single large endometrioma from a single human source for all endometrioma fluid to maintain constancy in this variable across the groups. It is also important to note that the capped container of endometrioma fluid was mixed before aspiration into the syringe and instillation into the animals studied to ensure that all aliquots injected would be expected to have identical contents. Further investigation into the variability of cellular activity in human endometrioma fluid would, however, be valuable. Although this study is subject to limitations, it raises an important question: what if irrigation and suction of endometrioma cyst contents did prove to cause greater adhesion formation in humans? It is technically possible to minimize or avoid spillage of endometrioma contents during surgical resection and thereby avoid the need for irrigation and suction. Use of careful dissection, aspiration of the endometrioma fluid before rupture, or a sterile bag collection device to avoid spillage of the endometrioma fluid are possible substitute techniques to the currently accepted surgical approach. One must consider the surgical context when discussing practice changes, and it is critical to recall that endometriomas in humans are often associated with pelvic adhesions even before surgical therapy, therefore any animal model must be followed by human investigation before changing surgical technique. In light of these possibilities, further studies are important. This study demonstrates that release of human endometrioma fluid alone into the peritoneal cavity of rabbits caused minimal adhesion formation. Release of human endometrioma fluid followed by copious normal saline lavage with removal of all visible endometrioma fluid caused significantly higher mean adhesion scores. The reasons for these results are unclear. The study raises the concern that present surgical techniques in the treatment of endometriomas causes increased peritoneal adhesion formation. Further investigation of the abdominal and peritoneal adhesive effects of endometrioma fluid are needed to assure that the present surgical management of endometriomas in humans is optimal. Acknowledgments: The authors are grateful to Patricia Foley for her veterinarian assistance and to Gina Wimer for her expertise with animal care. We thank Laurel Rice for the use of her laboratory facilities. We thank Tyler Shackelford for assistance with laparoscopic equipment. We thank Philip Smith for editorial review. REFERENCES 1. Busacca M, Vignali M. Ovarian endometriosis: from pathogenesis to surgical treatment. Curr Opin Obst Gynecol 2003;15: Rawson, JM. Prevalence of endometriosis in asymptomatic women. J Repro Med 1991;37: Exacoustos C, Zupi E, Amadio A, Szabolcs B, De Vivo B, Marconi D, et al. Laparoscopic removal of endometriomas: sonographic evaluation of residual functioning ovarian tissue. Am J Obstet Gynecol 2004;191: Smith et al. Endometrioma cyst fluid and adhesions Vol. 87, No. 5, May 2007

7 4. Donnez J, Smets M, Jadoul P, Pirard C, Squifflet J. Laparoscopic management of peritoneal endometriosis, endometriotic cysts, and rectovaginal adenomyosis. Ann NY Acad Sci 2003;997: Fayez JA, Vogel MF. Comparison of different treatment methods of endometriomas by laparoscopy. Obstet Gynecol 1991;78: Canis M, Mage G, Wattiez A, Chapron C, Pouly JL, Bassil S. Secondlook laparoscopy after laparoscopic cystectomy of large ovarian endometriomas. Fertil Steril 1992;58: Beretta P, Franchi M, Ghezzi F, Busacca M, Zupi E, Bolis P. Randomized clinical trial of two laparoscopic treatments of endometriomas: cystectomy versus drainage and coagulation. Fertil Steril 1998;70: Busacca M, Marana R, Caruana P, Candiani M, Muzii L, Calia C, et al. Recurrence of ovarian endometrioma after laparoscopic excision. Am J Obstet Gynecol 1999;180: Pantoja E, Noy MA, Axtmayer RW, Colon FE, Pelegrina I. Ovarian dermoids and their complications. Comprehensive historical review. Obstet Gynecol Surv 1975;30: Fiedler EP, Guzick DS, Guido R, Kanbour-Shakir A, Krasnow JS. Adhesion formation from release of dermoid contents in the peritoneal cavity and effect of copious lavage: a prospective, randomized, blinded, controlled study in a rabbit model. Fertil Steril 1996;65: Chapron, C, Vercellini P, Barakat H, Vieira M, Dubuisson J-B. Management of ovarian endometriomas. Hum Repro Update 2002;8: Campo S, Garcea N. Laparoscopic conservative excision of ovarian dermoid cysts with and without an endobag. J Am Assoc Gynecol Laparosc 1998;5: Hessami SH, Kohanim B, Grazi RV. Laparoscopic excision of benign dermoid cysts with controlled intraoperative spillage. J Am Assoc Gynecol Laparosc 1995;2: Milad MP, Olson E. Factors that increase the risk of leakage during surgical removal of benign cystic teratomas. Hum Repro 1999;14: Larsson B, Perbeck L. The possible advantage of keeping the uterine and intestinal serosa irrigated with saline to prevent intraabdominal adhesions in operations for fertility. An experimental study in rats. Acta Chirurg Scand 1986: Roberts LM, Sanfilippo JS, Raab S. Effects of laparoscopic lavage on adhesion formation and peritoneum in an animal model of pelvic inflammatory disease. J Am Assoc Gynecol Laparosc 2002;9: Al-Took S, Murray C, Tulandi T. Effects of pirfenidone and dermoid cyst fluid on adhesion formation. Fertil Steril 1998;69: Sahakian V, Rogers RG, Halme J, Hulka J. Effects of carbon dioxidesaturated normal saline and Ringer s lactate on postsurgical adhesion formation in the rabbit. Obstet Gynecol 1993;82: Pagidas K, Tulandi T. Effects of Ringer s lactate, Interceed (TC7) and Gore-Tex Surgical Membrane on postsurgical adhesion formation. Fertil Steril 1992;57: van Westreenen M, van den Tol PM, Pronk A, Marquet RL, Jeekel J, Leguit P. Perioperative lavage promotes intraperitoneal adhesion in the rat. Eur Surg Res 1999;31: Fertility and Sterility 1179