Closure techniques for fetoscopic access sites in the rabbit at mid-gestation

Jan A. Deprest1,2,4, Nikolaos A. Papadopulos1, Herbert Decaluwé1, Hirotoshi Yamamoto1, Toni E. Lerut1,3 and Eduard Gratacós1

1 Centre for Surgical Technologies, Faculty of Medicine, Katholieke Universiteit Leuven and the Departments of 2 Obstetrics & Gynaecology and 3 Thoracic Surgery, University Hospitals, Leuven, Belgium


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Operative fetoscopy may be limited by its relatively high associated risk of preterm prelabour rupture of membranes. The objective of this study was to study closure techniques of the access site for fetoscopy in the mid-gestational rabbit. A total of 32 does (288 amniotic sacs) at 22 days gestational age (GA; term = 32 days) underwent 14 gauge needle fetoscopy, by puncture through surgically exposed amnion. Entry site was randomly allocated to four closure technique groups: myometrial suture (n = 14), fibrin sealant (n = 15), autologous maternal blood plug (n = 13), collagen plug (n = 14); 16 sacs were left unclosed (positive controls), and the unmanipulated 216 sacs were negative controls. Membrane integrity, presence of amniotic fluid and fetal lung to body weight ratio (FLBWR) were evaluated at 31 days GA. Following fetoscopy without an attempt to close the membranes, amniotic integrity was restored in 41% of cases (amniotic integrity in controls 94%; P = 0.00001). When the access site was surgically closed, the amnion resealed in 20–44% of cases, but none of the tested techniques was significantly better than the others or than positive controls. Permanent amniotic disruption was associated with a significantly lower FLBWR in all groups. In conclusion, the rate of fetoscopy-induced permanent membrane defects in this model did not improve by using any of the closure techniques tested here.

Key words: amnion/fetal membranes/fetoscopy/preterm premature rupture of the membranes/rabbit


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Invasive first and second trimester procedures are by definition associated with membrane trauma, but so far it is not known exactly how the fetal membranes remodel following such iatrogenic disruption. Rather than sealing, clinically obvious amnion leak or even frank preterm prelabour rupture of the membranes (PPROM) may occur. Chorio-amnionitis is the commonest complication of PPROM, which is a strong trigger for preterm labour but may even cause potentially lethal feto–maternal sepsis. It may also lead to chronic oligohydramnios and subsequent pulmonary hypoplasia, which is a cause of high perinatal mortality (Carrol et al., 1996).

Mid-gestational amniocentesis is associated with a 1–2% risk of persistent amniotic leak (Tabor et al., 1986Go; Gold et al., 1989Go; Reece, 1997Go). Diagnostic fetoscopy, as practised for diagnostic purposes in the 1970s, was associated with a 4–8% risk for PPROM (Holzgreve, 1988Go; Grannum and Copel, 1990Go). Fetoscopy is still rarely used in fetal medicine, but currently suggested as an access modality for in-utero diagnosis and surgery in selected conditions. Nd:YAG laser coagulation for feto–fetal transfusion syndrome (FFTS) is associated with a PPROM rate of ~10% (Ville et al., 1998Go). More complex fetoscopic procedures such as cord ligation are associated with a PPROM rate of >30% (Quintero et al., 1996bGo; Deprest et al., 1998bGo). Suggested predictive factors for membrane rupture are; experience of the operator, the duration of and extent of manipulation during the procedure, the diameter of the needle used and gestational age (GA) (Deprest et al., 1997Go). Wide application of operative fetoscopy will be limited by this relatively high risk for `iatrogenic' PPROM. In analogy to preterm labour following hysterotomy in case of so-called `open' fetal surgery, the former was called the Achilles heel of fetoscopic surgery (Adzick and Harrison, 1994Go; Deprest et al., 1996Go). Therefore research is needed to document risk factors and ultimately prevent these membrane-related side-effects of mid-trimester fetoscopy. The development of techniques for closing the membrane defect created at the access sites appears to be a logical approach. Different closure techniques are used or have been proposed for fetoscopic procedures (Genz, 1979Go; Anger, 1986Go; Baumgartem and Moser, 1986Go; Quintero et al., 1996aGo; Sener et al., 1997Go), but to our knowledge few of them have been evaluated in experimental conditions. We developed an experimental model for creation of membrane defects following fetoscopy (Gratacós et al., 1999Go). In the present paper, we use this model to study the efficacy of different techniques for closing fetoscopic access sites.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
A total of 32 time-dated pregnant New-Zealand does was housed 2 days prior to surgery. The housing was quiet, at normal room temperature and normal day light. They were allowed only water during the last 12 h prior to the primary intervention. At 22 days GA (range 21–23 days), the first operation was performed. After premedication with ketamin 50 mg/kg i.m. (Ketalin®; Apharmo, Arnhem, The Netherlands), promazinum hydrochloridium 5 mg/kg i.m. (Prazine®; Libamedi, Brussels, Belgium), and antibiotic prophylaxis with penicillin G 300 000 IU i.m., animals were put under general face mask anaesthesia using halothane (2–5%) in oxygen 1 l/min. Maternal heart rate and oxygen saturation were monitored with a pulse oximeter (Nellcor®N-20P; Nellcor Inc., Haasrode, Belgium). The animals were positioned in the supine position and the abdomen was shaved under continuous aspiration, disinfected with povidone iodine (Iso-Betadine®; Asta Medica, Brussels, Belgium) and draped in a sterile way. Interventions on the uterus and membranes were performed using an operation microscope (Carl Zeiss, Oberkochen, Germany; magnification x5–25) and with micro-instruments. Fetoscopy was done with a short 1.2 mm 10 000 pixels 0° fibre endoscope (Karl Storz, Tüttlingen, Germany) housed within a 14 gauge needle.

A liberal midline abdominal incision was made to expose the pregnant uterus. First, gestational sacs were counted and numbered. A maximum of one in three amniotic sacs, with exclusion of the gestational sacs above the cervix, was randomly assigned to five study groups, according to the closure technique to be tested (n = 72). The others served as negative controls (n = 216; Table IGo).


View this table:
[in this window]
[in a new window]
 
Table I. Amniotic sacs in different study or control groups
 
In the rabbit, duration of pregnancy is 32 days, and mean litter size is eight. The fetus lies within the amniotic cavity facing the placenta, and its cord is attached antimesometrial to the choriovitelline circulation. The extra-coelomic and allantoic spaces are at that time more than virtual spaces. The amniotic sac can theoretically be accessed by blind needle puncture, but an open access technique guarantees correct positioning of the fetoscope (Gratacós et al., 1999Go; Papadopulos et al., 1999Go). To do so, a 2–3 mm myometrial incision with micro-scissors and under microscopic vision was performed. This mini-hysterotomy was located on the anti-mesometrial and thus avascular side, opposite the placenta and over the back of the fetus. Subsequently the chorion was incised and gentle uterine pressure was given, to let the amniotic membrane bulge through the chorionic incision. The amniotic sac was then entered under microscopic control with the 14 gauge needle. Once in the amniotic cavity, the fetoscope was introduced into the needle to explore the amniotic cavity. During fetoscopy fetal extremities, the face, tail, placenta or umbilical cord were identified (Figure 1Go). The fetoscopy was standardized to a maximum inspection time of 2 min with the use of a maximum of 5 ml Hartmann amniodistention fluid at body temperature.



View larger version (83K):
[in this window]
[in a new window]
 
Figure 1. Fetoscopic view of fetal mouth and part of the two forelimbs at 22 days gestational age, confirming intra-amniotic position of the needle.

 
After withdrawal of the fetoscope, the access site was closed according to the study group to which the gestational sac had been assigned (Table IGo). In group I (n = 14) only the myometrial layers were microsurgically closed with one nylon 6/0 suture (Figure 2Go). In study group II (n = 15), the fetoscopic access site was covered with 0.3 ml of human fibrin sealant over the membrane defect, also plugging the myometrium (Tissucol®; Immuno, Paris, France) (Figure 3Go). In group III (n = 13), homologous maternal fresh, full blood was taken from the ear vein, and applied as in group II. In group IV (n = 14) the gestational sac was closed with a collagen plug (Colgen®; Inter-Phar, Paris, France) (Figure 4Go). Before the operation, sterile dry collagen was precut and sized to the needle's lumen, with a maximal plug length of ~0.5 cm. To position the plug, we made a purpose-designed pusher of a few mm shorter than the length of the needle. After withdrawal of the fetoscope from the needle, a plug was loaded within the lumen. The plug and custom-made pusher were one after the other introduced into the needle, advancing the plug exactly to the tip of the needle point (Luks et al., 1997Go). Once in place, the plug inflates by absorbing surrounding fluid, and locks the access site. The myometrium was left untouched in groups II–IV. In the positive control group (n = 16) fetoscopic access sites were not closed at all. Unmanipulated amniotic sacs (n = 216) were negative controls.



View larger version (14K):
[in this window]
[in a new window]
 
Figure 2. Schematic representation of myometrial suture (treatment group I). M = myometrium; C = chorion; A = amnion.

 


View larger version (18K):
[in this window]
[in a new window]
 
Figure 3. Schematic representation of sealing fetoscopic access site with gel-like materials, e.g. Tissucol or blood patch (treatment groups II and III respectively). M = myometrium; C = chorion; A = amnion.

 


View larger version (18K):
[in this window]
[in a new window]
 
Figure 4. Schematic representation of a collagen plug (treatment group IV). M = myometrium; C = chorion; A = amnion.

 
After reposition of the uterus, the abdomen was closed in layers with polyglactine 3/0 (Vicryl ; Ethicon, Dilbeek, Belgium) for the fascia and intracutaneous nylon 2/0 suture (Ethilon®; Ethicon) for the skin. Postoperative uterine relaxation consisted of medroxyprogesterone-acetate 4.5 mg i.m. (Depo-Provera®; Upjohn, Puurs, Belgium). The animals were housed for the next 8 days in the same conditions as prior to surgery.

At 30–31 days GA does and fetuses were euthanized with a T61 injection (Hoechst, Brussels, Belgium), to undergo a second-look laparotomy. A myometrial incision was made on the antimesenterial side, starting over the first cornual sac. The incision was continued over the full length of the horn, both over treated and control sacs. Under microscopic control, the myometrium was gently dissected using microsurgical instruments. Main outcome measurements were integrity of both chorionic and amniotic membranes and presence or absence of amniotic fluid, as observed under the microscope. In case of an intact amniotic membrane, integrity was double checked by an intra-amniotic injection of 5–10 ml of saline, dyed with methylene blue. The injection was done with a 30 gauge needle through the membrane opposite the former fetoscopic access site. With this injection any possible leak accidentally missed before, would be visualized. The fetus was then delivered and for each fetus it was noted if it was alive or not. They were weighed, and dissected to assess the wet fetal lung weight (FLW) and to calculate the wet fetal lung-to-body-weight ratio (FLBWR). Macerated stillborn fetuses were noted as non-surviving and not further included in statistical processing. Initially, there were two more rabbits included in this study, but one of them aborted and one died. No obvious reason could be found at the time of necropsy. These animals were excluded from the study.

All animals were treated in accordance with the current guidelines on animal welfare and the experiments were approved by the Ethical Committee for Animal Experimentation of the Faculty of Medicine of the Katholieke Universiteit Leuven. Where appropriate, statistics were done with {chi}2 Fisher's exact two-tailed test for nominal variables, and unpaired Student's t-test or analysis of variance (ANOVA) for continuous variables using a SAS software package (SAS Institute Inc., Haasrode, Belgium). Membrane integrity of groups I–IV were compared to negative and positive controls and against each other. FLW and FLBWR of intact and open gestational sacs were compared.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The fetal loss rate in negative controls, or background fetal loss, was 28% (Table IIGo). There was no additional fetal loss due to fetoscopy, as observed in positive controls (28%; not significant = NS). Fetal survival was neither affected by any of the closure techniques used.


View this table:
[in this window]
[in a new window]
 
Table II. Evaluation of the amniotic sacs of liveborn rabbit fetuses at 30 days of gestational age
 
Integrity of the amniotic and chorionic membranes of unmanipulated sacs near term was 94.3%. Following fetoscopy with the 14 gauge needle without closure (positive controls), amniotic integrity as evaluated 8 days later was 41.7% (P < 0.000001 compared with negative control) and chorionic integrity 25% (P < 0.01). The presence of amniotic fluid followed similar trends. None of the studied closure techniques significantly improved or worsened the restoration of amniotic integrity. In absolute numbers, myometrial suturing performed the best with 44%, while covering the access site with maternal blood resulted in only 20% amniotic integrity. Chorionic integrity rates followed the same trend, with significantly (P < 0.05) lower rates for all treatment groups versus negative controls. Membrane defects looked somewhat larger than the initial trauma in all treatment groups (data not shown). No intra-amniotic adhesions or amniotic bands were seen in any animal.

FLBWR in sacs with and without membrane integrity were compared (Table IIIGo). In gestational sacs with persistent membrane defects at term, FLBWR was significantly lower than in fetuses from treated but intact gestational sacs by the time of the second-look operation [0.0256 (SD 0.005) versus 0.0288 (SD 0.003); P < 0.05]. This remained significant when comparing to negative controls with intact membranes [0.030; (SD 0.017) P < 0.05]. Numbers were too low to allow statistical analysis of pulmonary parameters within individual study groups.


View this table:
[in this window]
[in a new window]
 
Table III. Comparison of fetal body, lung and lung-to-body-weight ratio at term in sacs with intact and disrupted membranes
 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Thanks to advances in endoscopic surgery and the development of high resolution millimetric fibre-endoscopes, fetoscopy has been reconsidered for operations on the fetus or placenta. Key-hole access to the amniotic cavity might overcome the risk for preterm labour, the limiting problem of open fetal surgery (Adzick and Harrison, 1994Go). However, fetoscopy is associated with a substantial risk of PPROM. While the occurrence of this complication may be minimal when very small endoscope diameters are used, such as in early pregnancy diagnostic fetoscopy (Schwartzler et al., 1998Go), there is still a substantial risk of iatrogenic PPROM following operative fetoscopic procedures performed with larger endoscope diameters and/or more than one port (Deprest et al., 1998aGo).

A potential way to prevent iatrogenic PPROM might be to close or seal the membrane defect at the time of the intervention. Several techniques are used or have been suggested to close the fetoscopic access site, but their effect on the risk of PPROM is unclear. The uterine access mode may influence the technique used. Only a few fetoscopic procedures are done with partial or complete exposure of the uterine wall through maternal laparotomy. In this case, the uterus may be closed by a purse string, which arrests immediate fluid leak or potential uterine wall haemorrhage (Luks et al., 1997Go; Deprest et al., 1998bGo). However, most fetoscopies today are done with a percutaneous approach. After withdrawal of the endoscope, the entry site is left unclosed and the myometrial defect probably closes by uterine contraction. When considering sealing the entry site percutaneously, the use of liquid agents or plugs is intuitively the most feasible approach.

The concept of sealing the site of disruption with fibrin glue had already been suggested for spontaneous PPROM (Genz, 1986Go), but this technique was never widely implemented because of lack of efficacy (Anger, 1986Go; Baumgartem and Moser, 1986Go; De Lia et al., 1990Go). In our early clinical experience, we used this glue to seal the trocar site following a fetoscopic cord ligation, but an amniotic band syndrome occurred in that patient (Deprest et al., 1996Go). Fear of a potential causal relationship and the lack of experimental evidence precluded further implementation of this technique. Successful treatment of PPROM following fetoscopic cord ligation has been reported using an intra-amniotic platelet-cryoprecipitate plug (Quintero et al., 1996aGo). Following the same line, full maternal blood has been used successfully to patch the entry site in a case of persistent amniotic fluid leak after amniocentesis at 16 weeks (Sener et al., 1997Go). However, additional clinical experience confirming the successful outcome described in these case reports continues to be lacking.

Anecdotal experience is clearly not enough to support wide implementation of sealing techniques and therefore experimental work to estimate the real efficacy and potential complications of any proposed technique seems justified. We have previously described the mid-gestational rabbit model to test out closure techniques of hystero-amniotomy sites as performed in open fetal surgery (Papadopulos et al., 1998Go). In that study, a 1 cm hystero-amniotomy resulted in a permanent large membrane defect with anhydramnios in 100% of cases. Among the various techniques evaluated, closure of the myometrium was associated with the highest rate of postoperative amniotic integrity. We have later used a comparable protocol in the same animal model to produce fetoscopy-induced membrane defects (Gratacós et al., 1999Go). In that study, we demonstrated that fetoscopy induced a persistent membrane rupture rate of ~40% which was accompanied by pulmonary hypoplasia. Thus the model resembled the complications of oligohydramnios in humans and was suitable for studies aiming to modify the rate of membrane defects following fetoscopy.

In the present study, we have used the rabbit model to assess the efficacy of some techniques that may be considered for closing the access site following fetoscopy. Most of these techniques are applicable by a percutaneous approach (groups II–IV). Myometrial closure (group I) was also tested, since it yielded the best results in our previous study on hystero-amniotomy incisions (Papadopulos et al., 1998Go). However, none of the tested techniques improved the spontaneous rate of restored membrane integrity. The fetal pulmonary complications of oligohydramnios were also present in fetuses from treated sacs but with persistent membrane defects. There may be different reasons for this relatively poor performance. It is possible that plugs became dislodged in the postoperative course (groups II–IV). The lack of improvement associated to the use of myometrial suture could reflect that this approach does not address the membrane defect adequately.

Finally, as with any animal experiment, there are limitations for extrapolating results to human pregnancy. In any case, these discouraging results illustrate the need for experimental work testing any proposed technique prior to its clinical application. Combinations of techniques, more appropriate platelet preparations or more species-specific fibrin sealant or collagen plugs could be also tested.

In conclusion, clinical experience with operative fetoscopy stresses the need for techniques to prevent or treat membrane defects. None of the surgical techniques evaluated in a pregnant rabbit model could improve spontaneous resealing rates of fetal membranes following an invasive procedure such as fetoscopy. Among the possible directions for future research, a combination of different techniques may be considered, but additional research should go deeper into the mechanism of spontaneous membrane remodelling following iatrogenic trauma, both at the morphological and the molecular level.


    Acknowledgments
 
The authors thank T.Adriaenssens, C.Ghysel, T.Philips, I.Laermans and R.Kinnart for technical assistance and D.van Schoubroeck for clinical input and fruitful discussions. Maurice Bagot d'Arc, Medical Director of Immuno France, is thanked for providing the Tissucol® and Colgen® samples. Gerard Barki and Mrs Storz-Reling are thanked for providing the fetoscopic equipment. The other members of the Eurofoetus group (Y.Ville, Y.Arboy, T.H.Bui, K.Hecher, Y.Dumez, U.Nicolini) are acknowledged for their efforts to set up the Eurofoetus project (http://www.eurofoetus.org).

This study was supported by the Biomed 2 Programme of the European Community (`Eurofoetus`, grant N° PL 962383). E.G. is the recipient of a research fellowship from the European Commission.


    Notes
 
4 To whom correspondence should be addressed at: Center for Surgical Technologies, Minderbroedersstraat 17, B-3000 Leuven, Belgium Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Adzick, N.S. and Harrison, M.R. (1994) Fetal surgical therapy. Lancet, 343, 897–902.[ISI][Medline]

Anger, H. (1986) Fibrin sealing in premature rupture of membranes in pregnancy. Gynaecol. Obstet. Urol., 3, 95–97.

Baumgarten, K. and Moser, S. (1986) The technique of fibrin adhesion for premature rupture of the membranes during pregnancy. J. Perinat Med., 14, 43–49.[ISI][Medline]

Carroll, S., Sebire, N. and Nicolaides, K. (1996) Preterm prelabour amniorrhexis. Fetal Maternal Med. Rev., 8, 173–186.

De Lia, J.E., Cruikshank, D.P. and Keye, W.R. (1990) Fetoscopic neodymium:YAG laser occlusion of placental vessels in severe twin–twin transfusion syndrome. Obstet. Gynecol., 75, 1046–1053.

Deprest, J., Evrard, V., Van Schoubroeck, D. and Vandenberghe, K. (1996) Fetoscopic cord ligation. Lancet, 348, 890–891.

Deprest, J., Lerut, T. and Vandenberghe, K. (1997) Operative fetoscopy: new perspective in fetal therapy? Prenat. Diagn., 17, 1247–1260.[ISI][Medline]

Deprest, J.A., Evrard, V.A., Van Ballaer, P.P. et al. (1998a) Fetoscopic cord ligation. Eur J. Obstet Gynecol. Reprod. Biol., 81, 157–164.[ISI][Medline]

Deprest, J., Van Schoubroeck, D., Van Ballaer, P. et al. (1998b) Alternative access for fetoscopic Nd: YAG laser in TTS with anterior placenta. Ultrasound Obstet. Gynecol., 347–352.

Genz, H.J. (1979) Die Behandlung des vorzeitigen Blasensprungs durch Fibrinklebung. Med. Welt., 30, 1557–1559.[ISI][Medline]

Genz, H.J. (1986) Fibrin sealing in premature rupture of membranes. Gynaecol. Obstet. Urol., 3, 98–102.

Gold, R.B., Goyert, G.L., Schwartz, D.B. et al. (1989) Conservative management of second-trimester post-amniocentesis fluid leakage. Obstet. Gynecol., 74, 745–747.[Abstract]

Grannum, P.A. and Copel, J.A. (1990) Invasive fetal procedures. Radiol. Clin. N. Am., 28, 217–226.[ISI][Medline]

Gratacós, E., Yamamoto, H., Papadopulos, N. et al. (1999) The mid-gestational rabbit as a model for the creation of membrane defects following needle fetoscopy. Am. J. Obstet. Gynecol., in press.

Holzgreve, W. (1988) Endoskopie in pränataler Diagnose und Therapie. Z. Kinderchir., Suppl I, 6–7.

Luks, F.I., Deprest, J.A., Gillchrist, B.E. et al. (1997) Access techniques in endoscopic fetal surgery. Eur. J. Pediatr. Surg., 7, 131–134.[ISI][Medline]

Papadopulos, N., Van Ballaer, P.P., Ordonez, J.L.O. et al. (1998) Fetal membrane closure techniques following mid-gestational hystero-amniotomy in the rabbit model. Am. J. Obstet. Gynaecol., 178, 938–942.[ISI][Medline]

Papadopulos, M.A., Dumitrascu, I., Ordanez, J.L., et al. (1999) Fetoscopy in the pregnant rabbit at mid-gestation. Fet. Diagn. Ther., 14, 118–121.[ISI]

Quintero, R.A., Romero, R., Reich, H. et al. (1996a) In utero percutaneous umbilical cord ligation in the management of complicated monochorionic multiple gestations. Ultrasound Obstet. Gynecol., 8, 16–22.[ISI][Medline]

Quintero, R.A., Romero, R., Dzieczkowski, J. et al. (1996b) Sealing of ruptured amniotic membranes with intra-amniotic platelet-cryoprecipitate plug. Lancet, 347, 1117.[ISI][Medline]

Reece, A.E. (1997) Early and midtrimester genetic amniocentesis. Safety and outcome. Fet. Diagn. Ther., 24, 71–81.

Schwartzler, P., Moscoso, G., Senat, M.V. et al. (1998) The cobweb syndrome: first trimester sonographic diagnosis of multiple amniotic bands confirmed by fetoscopy and pathological examination. Hum. Reprod., 13, 2966–2969.[Abstract/Free Full Text]

Sener, T., Özalp, S., Hassa, H. et al. (1997) Maternal blood clot patch therapy: a model for postammniocentesis amniorrhea. Am. J. Obstet. Gynecol., 177, 1535–1536.[ISI][Medline]

Tabor, A., Madsen, M., Obe, E.B. et al. (1986) Randomised controlled trial of genetic amniocentesis in 4606 low risk women. Lancet, i, 287–293.

Ville, Y., Hecher, K., Gagnon, A. et al. (1998) Endoscopic laser coagulation in the management of severe twin-to-twin transfusion syndrome. Br. J. Obstet. Gynaecol., 105, 446–453.[ISI][Medline]

Submitted on December 29, 1998; accepted on April 1, 1999.