The reduction of postoperative adhesions by two different barrier methods versus control in cynomolgus monkeys: a prospective, randomized, crossover study

VoL 61. No.6. June 1994

FERTILITY AND STERILITY Copyright

~

Printed on acid-free paper in U S. A.

1994 The American Fertility Society

The reduction of postoperative adhesions by two different barrier methods versus control in cynomolgus monkeys: a prospective, randomized, crossover study

Daniel R. Grow, M.D.* Howard J. Seltman, M.D. Charles C. Coddington, M.D. Gary D. Hodgen, Ph.D. The Jones Institute for Reproductive Medicine, Department of Obstetrics and Gynecology, Eastern Virginia Medical School, Norfolk, Virginia

Objectives: To determine the effectiveness of two different surgical membranes in preventing postoperative adhesions as compared with control and to evaluate the benefit as compared with "second-look" adhesiolysis. Design: A randomized, prospective, crossover study. Setting: A colony of individually caged non-human primates, Department of Obstetrics and Gynecology, Eastern Virginia Medical School Norfolk, Virginia. Interventions: Hysterotomy or adhesiolysis at the time of exploratory laparotomy, with subsequent scoring of adhesions based on area, vascularity, tenacity, and adhesion score. Results: The combined crossover data reveal that both surgical barriers are superior in adhesion prevention to microsurgical technique alone. Expanded polytetrafiuoroethylene (Gore-Tex Surgical Membrane; WL Gore and Associates, Flagstaff, AZ) was better than oxidized regenerated cellulose (Interceed; Johnson and Johnson Medical, Inc., Arlington, TX) with respect to adhesion area, tenacity, and vascularity, with a significant improvement in the total adhesion score. Second-look adhesiolysis resulted in significant adhesion reduction in the control group, making second-look adhesiolysis statistically similar to the use of either barrier alone (without subsequent adhesiolysis). Gore-Tex removal does not result in adhesion formation as determined by third-look surgery. Conclusions: Both Interceed and Gore-Tex show a reduction in the prevention of postsurgical adhesions after hysterotomy incisions, as compared with microsurgical technique alone. Secondlook adhesiolysis is as effective as either barrier in the reduction of permanent pelvic adhesions. Fertil Steril1994;61:1141-6 Key Words: Pelvic adhesions, polytetrafiuoroethylene, oxidized regenerated cellulose

Postoperative adhesions that occur in 55% to 100% of fertility-enhancing procedures continue to grieve the reproductive surgeon (1, 2). These adhe-

Received June 15, 1993; revised and accepted February 14, 1994. * Reprint requests and present address: Daniel R. Grow, M.D., Department of Obstetrics and Gynecology, Baystate Medical Center, 759 Chestnut Street, Springfield, Massachusetts 01199 (FAX: (413) 784-3440). Vol. 61, No.6, June 1994

sions often negate the beneficial effects of the primary surgery and are sometimes responsible for serious long-term complications such as chronic pain and intestinal obstruction. Innumerable adhesion-preventing agents have been investigated throughout the history of surgery (3). Recent works have shown two barrier methods with potential for long-term benefit: oxidized-regenerated cellulose (Interceed; Johnson and Johnson Medical, Inc., Arlington, TX), and expanded polytetrafluoroethylene (PTFE, Gore-Tex Surgical Grow et al. Adhesion prevention with barriers

1141

Membrane; WL Gore and Associates, Flagstaff, AZ). These barriers theoretically allow healing of the mesothelial layer without the formation of connective adhesions between organs in close proximity. Each barrier has its own unique advantages and disadvantages, and both have been shown to have some effectiveness (Sckiba K, abstract) (2). Interceed is easy to apply, does not require sutures, and is absorbable over time; however, though not recognized clinically, it appears to cause de novo adhesions and peritoneal injury as found in the mouse model (4). Gore-Tex must be sutured into place and thus is more time-consuming to apply, but it is nonreactive to peritoneal surfaces (4) and is particu1arly inert in its interaction with gametes (5). Few studies to date have directly compared Interceed with Gore-Tex and with control in a prospective, randomized fashion to prevent postoperative adhesions. The first (and possibly the only) study to do this employed the rat uterine horn model and showed Ringer's lactate superior to both barriers (6). If these barriers are to be used clinically, well-controlled studies must be performed. We present a controlled, randomized, prospective, crossover study using a hysterotomy incision in the non-human primate. The cynomolgus monkey has anatomic and phylogenetic similarity to the human, making it an excellent non-patient model, allowing a rigorously structured experimental protocol. This incision is similar to one that may be employed when performing an abdominal myomectomy. Myomectomies are commonly performed but are notorious when considering postoperative adhesions (possibly because of the difficulty in achieving absolute homeostasis). Several reports in recent years have indicated a possible role for "second-look" laparoscopy for the reduction of permanent peritoneal adhesions (712); however, only two studies have specifically addressed efficacy (11, 12) because this requires a "third-look" procedure to assess the benefit. No study to date has addressed the efficacy of secondlook adhesiolysis as compared with the ability of either Interceed or Gore-Tex surgical membrane alone for postoperative adhesion prevention. This study will address the following three issues: (1) Is either adhesion barrier superior to control in the prevention of adhesions? (2) How effective is adhesiolysis in reducing adhesion reformation, and how does it compare with surgical membrane alone? (3) Do adhesions form at the peritoneal surface covered by Gore-Tex, after Gore-Tex is removed? 1142

Grow et a1.

Adhesion prevention with barriers

MATERIALS AND METHODS

Subjects Adult female cynomolgus monkeys (Macaca fascicularis, 3.5 to 6.0 kg) were caged individually in a controlled light and temperature environment. They were fed a commercial primate diet and had unrestricted access to water under the supervision of an in-house veterinarian in our Division of Animal Resources. The procedures described were approved by the Animal Use and Care Committee at our institution. Thirty-two cynomolgus monkeys underwent a laparotomy (ketamine 20 mg/kg 1M, xylazine 1 mg/kg 1M) to perform a 2-cm fundal hysterotomy incision extending from cornu-to-cornu. The incision was closed in two layers, the second an imbricating 4-0 absorbable suture. Sterile techniques were maintained with attention to meticulous hemostasis and careful tissue handling. Before surgery, each primate was randomly assigned to a treatment group: oxidized regenerated cellulose (Interceed); expanded PTFE (Gore-Tex); or control. Interceed was applied so that the barrier covered all incision and suture materials with a l-cm margin, then moistened with sterile Ringer's lactate. Gore-Tex was applied in the same way, except the corners were attached with a single 6-0 nonabsorbable (Proline; Ethicon, Somerville, NJ) and three 6-0 absorbable sutures (Vicryl; Ethicon). Control monkeys received no surgical membrane or other adjuvant. Peritoneal irrigation with lactated Ringer's solution was performed during all procedures but was removed in all primates with gentle application of sterile absorbent gauze to the cul-de-sac before abdominal closure. Postoperative antibiotics were given to all animals as is the routine in the animal facility (Kefzol, 50 mg/kg 1M; Eli Lilly and Company, Indianapolis, IN). Adequate postsurgical pain control was provided (Nubain, 1 mg/kg; Dupont Multi-Source Products, Garden City, NY). Figure 1 illustrates the sequence of experimental events. One month after the original surgery, a second laparotomy was performed for the grading of adhesions to the uterine fundus. Adhesions were systematically analyzed for area, vascularity, and tenacity, with a total score for each animal tabulated (13). Area of adhesive involvement to the site of the incision was measured in cm 2 • The tenacity of the adhesions was evaluated according to the tension necessary in lysing the adhesions: 1, filmy, easily lysed; 2, moderate tension required; 3, sharp dissection Fertility and Sterility

o 12 CD

10

(J)

8

Adhesion formation after the first hysterotomy

t2J Ahesion reformation, in corresponding treatment groups after adhesiolysis and/or Gore-tex removal.

50 <:

0

'iii CD .r:

6

first barrier adhesion score with the second procedure adhesion score and for paired comparisons in the crossover analysis. Linear regression was used to test the propensity of individual subjects to form adhesions by comparing the adhesions formed after the first control surgery with adhesions formed after the crossover control surgery.

"C

«

4

RESULTS Interceed • p (

a01 indicating

a significant improYement after adhesioiysis.

Figure 1 Adhesion scores after the first hysterotomy for each treatment group, and the score on the corresponding subjects after second-look adhesiolysis and/or Gore-Tex removal. D, Adhesion formation after the first hysterotomy; ~, adhesion reformation, in corresponding treatment groups after adhesiolysis and/or Gore-Tex removal. *P < 0.01 indicates a significant improvement after adhesiolysis.

required. Vascularity was scored as follows after adhesiolysis: 1, no bleeding; 2, punctate to mild bleeding; 3, bleeding that required a suture for hemostasis. When grade 2 vascularity was found, it was observed until hemostasis occurred, then irrigated with Ringer's lactate. An adhesion score was calculated as score-area (cm2 ) X (vascularity + tenacity). Adhesiolysis was then performed with removal ofthe Gore-Tex barrier when present. Three months later, a third surgery was performed for evaluation of adhesions and a second hysterotomy. The second hysterotomy incision and closure were performed as before, with each monkey prospectively assigned a different barrier or control. Animals were either cycling regularly, on GnRH agonist (leuprolide acetate, Lupron depot; TAP Pharmaceuticals, Deerfield, IL), or on mifepristone (Roussel UCLAF, Paris, France) from the time of the second surgery until the end of the study protocol. Randomization took place after hormonal assignments were made, leading to an even distribution of barriers within hormonal groups. Photographs were taken of all adhesions before and after lysis for objective evaluation. All operative procedures were performed by a single operator for consistency of technique. Statistical evaluation was performed with analysis of variance, least-square multiple comparison test for area, and the nonparametric Kruskal-Wallis test for the score data with the appropriate Bonferoni correction for multiple comparisons. Wilcoxon's signed rank test was used for comparing the Vol. 61, No.6, June 1994

No primates suffered undue complications. The results of the adhesion assessments and scoring are tabulated below. Table 1 (top) shows the results of the adhesion assessment after the first hysterotomy. Adhesions after Gore-Tex placement were significantly less vascular (P < 0.02), less tenacious (P < 0.01), and occupied less area (P < 0.02) for a lower adhesion score (P < 0.006) when compared with the control group. The adhesions found after Interceed placement appear to have a better score than the control group, but statistical significance was not reached. Table 1 (middle) shows the result of the adhesion assessments that took place after adhesiolysis and/ or Gore-Tex removal (this is synonymous with a third-look assessment to examine the effectiveness of adhesiolysis and to determine whether the GoreTex site predisposes to adhesion formation after removal). After adhesiolysis, there was no statistical difference in adhesion quality or quantity between the three groups studied. The control group received significant benefit after adhesiolysis, as evidenced by a significant reduction in the area of adhesion involvement (P < 0.025) and in the adhesion score (P < 0.025), making them similar to the other two groups (Fig. 1). The Interceed group had a nonsignificant reduction in adhesions after adhesiolysis, with no net change in the Gore-Tex group. Gore-Tex was removed at the second-look evaluation and sent for histologic evaluation. At times it was covered with a layer of mesothelial cells and was quite firmly applied to the uterine fundus. Table 1 (bottom) shows the results of the adhesion assessment in the second arm of the crossover, with each subject receiving a different barrier or control after the second hysterotomy. (Please note that the time to adhesion scoring was 1 month in the first arm of the crossover and 3 months in the second arm, potentially altering the adhesion characteristics and scores). The second arm of the crossover is quite similar to the first, with Gore-Tex showing a significant benefit in each characteristic as compared with control. In this comparison, Grow et al.

Adhesion prevention with barriers

1143

Table 1 Adhesion Characteristics After Placement of the First Barrier or Control, After Adhesiolysis and/or Gore-Tex Removal, and After the Crossover to a Second Barrier or Control* Adhesion characteristics Group

Vascularity

Tenacity

Adhesion score

2.04 ± 0.51 0.46 ± 0.15t 1.15 ± 0.48

2.0 ± 0.30 1.18 ± 0.12t 1.73 ± 0.27

2.1 ± 0.31 1.0 ± O.ot 1.55 ± 0.28

9.9 ± 2.6 1.19 ± 0.46t 3.03 ± 2.2

0.63 ± 0.24:j:

1.67 ± 0.33

1.89 ± 0.31

2.91 ± 1.05:j:

0.45 ± 0.16

1.20 ± 0.20

1.40 ± 0.16

1.07 ± 0.32

0.30 ± 0.15

1.91 ± 0.26

1.18 ± 0.33

1.25 ± 0.73

1.37 ± 0.35 0.30 ± 0.12§ 1.06 ± 0.23

2.0 ± 0.30 1.0 ± 0.33§ 2.10 ± 0.31

2.09 ± 0.31 0.89 ± 0.26§ 2.10 ± 0.31

7.16 ± 2.0 1.07 ± 0.55§ 5.41 ± 1.25

Area em2

First hysterotomy Control (n = 10) Gore-Tex (n = 11) Interceed (n = 11) Second-look Control (after adhesiolysis) (n = 9) Gore-Tex (after removal) (n = 10) Interceed (after adhesiolysis (n = 11) Crossover (second hysterotomy) Control (n = 11) Gore-Tex (n = 9) Interceed (n = 10)

:j: p §p

* Values are means ± SD. t p ::0; 0.01 versus control.

Gore-Tex also shows advantages over Interceed, although statistical significance is not quite reached. Table 2 shows the combined data from both arms of the crossover. With 20 or more subjects in each group, both Gore-Tex and Interceed showed a significant advantage in the adhesion score when compared with control. Gore-Tex showed an advantage over both Interceed and control with respect to area, vascularity, and tenacity adhesion score. Figure 2 shows the results of the crossover study for the first barrier or control, the second barrier or control, and the combined data. Performing a linear regression analysis on the crossover data revealed that there was no correlation in the adhesion scores in individual animals between the first hysterotomy and the second hysterotomy. This implies that the difference in adhesion-forming propensity was greater between the

Table 2

::0; ::0;

0.01 indicates a significant reduction in adhesions. 0.05 versus control.

adhesion prevention methods than any difference that existed in an individual's predisposition to adhesion formation. If certain subjects had a higher propensity to form adhesions, we would have expected a significant correlation. None was found, even if considering only the Interceed and control groups. DISCUSSION

Cynomolgus monkeys have phylogenetic and anatomic similarity to the human, perhaps making them ideal subjects for a rigorously structured experiment for the study of postoperative adhesions. We studied the efficacy of two of the most promising barrier methods (Gore-Tex and Interceed) to control in a randomized, prospective, crossover experiment using a uniform surgical incision in all

Table Showing Overall Results, Combining Data From Both Arms of the Crossover Study* Adhesion characteristics Group

Area

Control Gore-Tex Interceed

1.69 ± 0.30 0.39 ± 0.10t:j: 1.11 ± 0.27

Vascularity

Tenacity

Adhesion score

2.0 ± 0.30 1.10 ± 0.16t 1.90 ± 0.21

2.1 ± 0.31 0.95 ± 0.11 t§ 1.81 ± 0.21

9.9 ± 2.6 1.13 ± 0.35t:j: 5.21 ± 1.2811

em'

* Values are means ± SD. t p < 0.02 versus control. :j: P < 0.02 versus Interceed.

1144

Grow et al. Adhesion prevention with barriers

§ P < 0.04 versus Interceed. II P < 0.01 versus control.

Fertility and Sterility

14

o Adhesion formation after first hysterotomy in subjects receiving barrier or control indicated.

12

ILl Adhesion formation after second (crossover) hysterotomy. subjects receiving barrier or control indicated.

10

"5 0

en c: 0 'in

..,so"

[JJ

Combined crossover data.

8 6

c(

4 2

Gore-Tex

Control a p < 0.002 vs cOlllro/ b p ( 0.05 vs control c p ( a002 vs conlro/

a 02 vs Interceed e p ( 0.01 vs control

d p ,

Figure 2 Adhesion score for each treatment group: after the first barrier or control, after the second barrier or control, and the combined crossover data. D, adhesion formation after first hysterotomy in subjects receiving barrier or control indicated; ~, adhesion formation after second (crossover) hysterotomy, subjects receiving barrier or control indicated; IZI, combined crossover data. a, P < 0.002 versus control; b, P < 0.05 versus control; c, P < 0.002 versus control; d, P < 0.02 versus interceed; e, P < 0.01 versus-control.

subjects and following a strict time frame. We show that both barriers are superior to control when using microsurgical techniques alone and that GoreTex has a significant advantage over both Interceed and control, regardless of whether it is removed or left in place. The efficacy of second-look adhesiolysis was judged by a third-look adhesion assessment and found to be as effective in reducing adhesions as either of the barrier techniques alone. Gore-Tex, despite its clear advantages, is not without other problems. It is more difficult to apply than Interceed because it must be sutured in place with two types of 6-0 suture material, as recommended by the manufacturer. The PTFE is not water soluble like oxidized regenerated cellulose and thus does not easily conform nor adhere to irregular peritoneal surfaces. Application time on a spherical uterine fundus during laparotomy is approximately 10 minutes. Interceed showed signs of complete absorption at second look (1 month). Its application was less time-consuming and technically less demanding than Gore-Tex. It lies flat on any contour because of its ability to absorb water. The adhesions that formed in the Gore-Tex group were mostly on the edge of the membrane itself. These adhesions were uniformly filmy and avascular requiring only gentle traction for complete removal. Rarely, omentum would migrate under the barrier to reach the hysterotomy site. Uterine adhesions in the Interceed group were of Vol. 61, No.6, June 1994

mixed type and unpredictable, sometimes scant, filmy and avascular, and other times, extensive, tenacious, and vascular. The reasons for this are unclear, but the author suspects that migration ofthe Interceed barrier may occur (12). Support for the migration theory comes from the fact that when subjects form heavy adhesions with Interceed, they usually form scant adhesions with Gore-Tex. Suturing Interceed in place, however, would cause it to lose its ease of application. Our suggestion is to use a larger piece than one would think necessary to cover the denuded peritoneum, thus augmenting its adhesive surface area. Although not performed, it would have been interesting to also apply Interceed with sutures to test the migration theory and to compare the efficacy of the different barriers with identical application technique. Also, please note that the surgical staging of adhesions took place at 1 month in the first arm of the study and at 3 months in the second arm. Although the adhesions graded at both times were permanent in nature, this time difference could potentially alter the adhesion scores. Studies of second-look adhesiolysis are difficult in humans because of the need for a third-look to assess efficacy. Two groups have previously shown second-look adhesiolysis to be effective (11, 12), and we concur. Because of the rigorous design here, we can also compare the adhesion scores after second-look adhesiolysis (Table 2) with scores after either barrier method alone (Fig. 2). This approach shows that there is no statistical difference between the results of combining a hysterotomy procedure with second -look adhesiolysis and performing a hysterotomy using a mechanical barrier without later adhesiolysis. Second-look adhesiolysis in the Interceed group resulted in a 75% reduction in the adhesion score; however, because of the relatively large variation, this did not reach statistical significance. Removing the Gore-Tex barrier did not increase the mean adhesion score. At times, it was covered with a layer of mesothelial cells and was quite firmly applied to the incision site. Other times, it was attached loosely, held in place by the suture ligatures. Its removal left in place a layer of mesothelium that had a low likelihood to later adhere to surrounding structures. Several of the subjects in the Gore-Tex group had more adhesions after barrier removal, but the mean adhesion scores were not different. Gore-Tex is inert and routinely left in vivo after aortic graft procedures. The risk of leaving it undisturbed on the uterine surface would not likely cause difficulty should pregnancy ensue. Much speculation has taken place throughout Grow et al.

Adhesion prevention with barriers

1145

the years regarding whether particular patients are more prone to adhesive disease than other patients, given the same peritoneal insult (Jones HW Jr., personal communication). This issue is very difficult to formally study, but evidence ofthis intuition has been the analogy with keloid formation. The crossover design allows one to determine whether the propensity of an individual to form adhesions is greater than the ability of a barrier to prevent them. We did not find this to be the case. Subjects with high adhesion scores in the first control group had lower scores when crossed over to either barrier. Regression analysis reveals no correlation in the adhesion scores in individual animals over time. We conclude that both Interceed and Gore-Tex show a significant reduction in adhesion formation when compared with control and that Gore-Tex leads to fewer and less tenacious adhesions than Interceed. Second-look adhesiolysis is efficacious when not using an adhesion barrier, and adhesion reformation is similar to adhesions formed when using either barrier method alone. Acknowledgments. The authors express their appreciation to Mr. Bruce Lamar for his professionalism in managing one of the finest animal facilities available; to Ms. Pauline M. Clynes for her editorial assistance and advice; and to Ms. Martha Wilson for the graphics. The authors also thank The Howard and Georgeanna Jones Institute for Reproductive Medicine, Norfolk, Virginia, for fostering an environment of academic and intellectual excellence. This project was performed independently of any corporate financial support.

REFERENCES 1. Adhesion Study Group. Reduction of postoperative pelvic adhesions with intraperitoneal 32% dextran 70: a prospective, randomized clinical trial. Fertil Steril 1983;40:612-9.

1146

Grow et aI.

Adhesion prevention with barriers

2. Interceed (TC7) Adhesion Barrier Study Group. Prevention of postsurgical adhesions by Interceed (TC7), an absorbable adhesion barrier: a prospective, randomized multicenter clinical study. Fertil Steril 1989;51:933-8. 3. diZerega GS. The peritoneum: postsurgical repair and adhesion formation. In: Rock JA, Murphy AA, Jones HW, editors. Female reproductive surgery. Baltimore: Williams and Wilkins, 1992:2-18. 4. Haney AF, Doty E. Murine peritoneal injury and de novo adhesion formation caused by oxidized-regenerated cellulose (lnterceed [TC7]) but not expanded polytetrafluoroethylene (Gore-Tex Surgical Membrane). Fertil Steril 1992;57:202-8. 5. Hunter SK, Scott JR, Hull D, Urry RL. The gamete and embryo compatibility of various synthetic polymers. Fertil Steril1988;50:110-6. 6. Pagidas K, Tulandi T. Effects of Ringer's lactate, Interceed (TC7) and Gore-Tex Surgical Membrane on postsurgical adhesion formation. Fertil Steril 1992;57:199-201. 7. Surrey MW, Friedman S. Second-look laparoscopy after reconstructive pelvic surgery for infertility. J Reprod Med 1982;27:658-60. 8. Raj SG, Hulka JF. Second-look laparoscopy in infertility surgery: therapeutic and prognostic value. Fertil Steril 1982;38:325-9. 9. DeCherney AH, Mezer HC. The nature of posttuboplasty pelvic adhesions as determined by early and late laparoscopy. Fertil Steril 1984;41:643-6. 10. Diamond MP, Daniell JF, Martin DC, Feste J, Vaughn WK, McLaughlin DS. Tubal patency and pelvic adhesions at early second-look laparoscopy following intraabdominal use of the carbon dioxide laser: initial report of the intraabdominal laser study group. Fertil Steril 1984;42:717-23. 11. Jansen RPS. Early laparoscopy after pelvic operations to prevent adhesions: safety and efficacy. Fertil Steril 1989;49:26-31. 12. Trimbos-Kemper TCM, Trimbos JB, van Hall EV. Adhesion formation after tubal surgery: results of the eighth-day laparoscopy in 188 patients. Fertil Steril 1985;43:395-400. 13. Boyers SP, Diamond MP, DeCherney AH. Reduction of postoperative pelvic adhesions in the rabbit with Gore-Tex surgical membrane. Fertil SteriI1988;49:1066-70.

Fertility and Sterility