Non-invasive early prenatal diagnosis using fluorescent in situ hybridization on transcervical cells: comparison of two different methods for retrieval

European Journal of Obstetrics & Gynecology and Reproductive Biology 95 (2001) 37±41

Non-invasive early prenatal diagnosis using ¯uorescent in situ hybridization on transcervical cells: comparison of two different methods for retrieval Tolga ErgIÇn MDa,*, Volkan Baltacib, Hulusi B. ZeynelogÏlua, E. Hakan Durana, Mehmet H. ErgenelIÇa, SertacË BatiogÏlua a

Department of Obstetrics and Gynecology, Baskent University School of Medicine, Kubilay Sokak, No. 36, 06570 Maltepe, Ankara, Turkey b Department of Medical Genetics, Baskent University School of Medicine, Kubilay Sokak, No. 36, 06570 Maltepe, Ankara, Turkey Received 11 January 2000; received in revised form 29 March 2000; accepted 14 April 2000

Abstract Objective: We compared the ef®ciencies of uterine and endocervical lavage to retrieve fetal cells from ®rst trimester pregnancies for further analysis with ¯uorescent in situ hybridization (FISH). Study Design: Transcervical cell (TCC) samples were collected at 7±10 weeks of gestations by uterine lavage (13 women) and by endocervical lavage (12 women) who were scheduled for volunteer termination of pregnancy. A sample of placenta was also obtained for cytogenetic analysis to con®rm the sex or genotype in the end of the procedure. FISH was performed using probes for the chromosomes 18, X and Y in a three color hybridization protocol. The statistical analysis included w2analysis, and t-test. Results: Suf®cient cells were obtained in 12 of the 13 (92.3%) in uterine lavage and 10 of the 12 (83.3%) in endocervical lavage group for FISH procedures for fetal sex prediction. The mean success rate of signal detection for FISH procedure was 91.7% (range 83±97%). Fetal sex was correctly predicted in 11 of 12 (91.6%) with uterine lavage and 8 of 10 (80.0%) in endocervical lavage and the difference was statistically insigni®cant. Conclusion: This study demonstrated that there are available cells of fetal origin in the lower part of the uterus and these cells may be collected successfully as early as 7 weeks of the gestation. In addition, our results show that endocervical lavage method is as effective as uterine lavage. FISH has been successfully used to detect status of aneuploidy and sex of the fetus from TCC. # 2001 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Non-invasive prenatal diagnosis; Transcervical cells; Multicolor FISH

1. Introduction Chorionic villous sampling (CVS) and amniocentesis are the two principal techniques of sampling fetal material for prenatal diagnosis. Since both these techniques are invasive, require expertise and may jeopardize the fetus, there is an increasing demand for a non-invasive technique. Evaluation of the fetal cells in maternal circulation have generated interest, since it is a non-invasive method of collecting fetal genetic material for prenatal diagnosis, however, the number of these cells circulating in maternal blood is extremely low and their isolation and analysis are technically dif®cult and standards are yet to be established [1±3]. As early as 1971, Sheetles [4] postulated that trophoblasts are shed into the uterine cavity and suggested the use of these cells as a source of fetal deoxyribonucleic acid (DNA) * Corresponding author. Tel.: ‡90-312-232-44-00; fax: ‡90-312-232-39-12.

for prenatal diagnosis. Those shed cells may be retrieved by uterine lavage, cervical mucus aspiration, endocervical lavage or cytobrush [5±7]. These methods are likely to be less invasive and have been suggested as alternative fetal tissue sampling procedures [8], despite there are some discrepancies among the ef®ciencies and complications of these techniques [5,6,8]. We compared the ef®ciencies of two techniques to retrieve fetal cells from ®rst trimester pregnancies namely: uterine and endocervical lavages. The harvested cells were then processed with ¯uorescent in situ hybridization (FISH). 2. Materials and methods 2.1. Subjects Transcervical cell (TCC) samples were collected at 7±10 weeks of gestation from 25 women by uterine lavage (13

0301-2115/01/$ ± see front matter # 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 3 0 1 - 2 1 1 5 ( 0 0 ) 0 0 3 5 7 - 2

T. ErgIÇn et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 95 (2001) 37±41

38

women) and by endocervical lavage (12 women) who were admitted for volunteer termination of pregnancy. The two groups of patients were randomized using a random number list. Written consents were obtained from all patients after given detailed information prior to the procedure and the hospital human studies committee approved the study. An ultrasound scan was performed to determine the correct gestational age and placental site. General anesthesia was established and the patient was placed in a dorsal lithotomy position, the cervix was cleaned with povidone iodine after placement of a speculum to the vagina. After a tenaculum was applied to the cervix, two methods were used for the collection of the fetal cells: 1. Method I: Uterine lavage A sterile flexible, catheter for intrauterine insemination (22 cm length and 2 mm i.d.; CCD Laboratories, Paris, France) was connected to a 20 ml syringe containing 10 ml of sterile saline. Under transabdominal ultrasound guidance, the catheter was gently inserted through the cervical canal, just past internal os, up to where it met resistance. A volume of 10 ml physiological saline was injected into the uterine cavity and 1±2 ml fluid was aspirated after an interval of 10±15 s. The catheter was then removed from the uterine cavity. 2. Method II: Endocervical lavage The same catheter was gently advanced to the level of the internal cervical os. When it was felt that the distal tip of the tube is at the level of the internal os, 5 ml

saline was slowly injected. Afterwards, the fluid was gently aspirated as much as possible while the catheter was gradually drawn out with some rotation. A sample of placenta was obtained for cytogenetic analysis to con®rm the sex or genotype at the end of the procedure. 2.2. Hybridization procedure The cell suspension was spun down at 1500 rpm for 8 min. The supernatant was then discarded and resuspended in 0.075 M KCl at 378C for 30 min. The sample was centrifuged at 1500 rpm for 5 min, supernatant discarded and the pellet was resuspended in 5 ml 1trypsin/EDTA at 378C water bath for 30 min. The tube was recentrifuged at 1500 rpm for 8 min, the supernatant was discarded and then the pellet was resuspended in 0.075 M KCl at 378C for 30 min for the second time. Then 2 ml of methanol:glacial acetic acid was added to the cells/hypotonic solution. The cell suspension was centrifuged and resuspended in 1 ml fresh ®xative and stored at 48C for at least 30 min. For FISH, 15±25 ml of cell suspension was dropped on to four super frost slides making four hybridization areas. Before the hybridization, slides were denatured at 738C in 70% formamide/2SSC and hybridized overnight at 378C, using probes for the chromosomes 18, X and Y in a three-color hybridization protocol. All probes were directly labeled with ¯uorophores and obtained commercially (Vysis Inc., Down-

Table 1 Details of study group and FISH and cytogenetic results Patient

Gestational week

Placental localization

Uterine lavage

2 4 5 6 10 11 13 15 18 19 22 24 25

9 9 8 10 7 8 8 9 8 8 9 7 9

Posterior Anterior Posterior Anterior Anterior Posterior Lateral Anterior Anterior Anterior Posterior Anterior Lateral

‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡

1 3 7 8 9 12 14 16 17 20 21 23

8 9 9 7 8 9 10 7 7 8 9 7

Anterior Anterior Anterior Lateral Posterior Anterior Posterior Posterior Anterior Posterior Posterior Posterior

Endocervical lavage

Signal detection of FISH (%) 96 93 97 ± 92 95 94 86 92 97 85 90 85

‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡

95 90 83 95 91 90 ± 95 92 90 96 ±

Cells with Y signal ‡ ‡

‡ ‡ ‡ ‡

‡ ‡ ‡ ‡

Karyotype 46, 46, 46, 46, 46, 46, 46, 46, 46, 46, 46, 46, 46,

XY XY XX XX XY XX XX XX XY XX XY XY XX

46, 46, 46, 46, 46, 46, 46, 46, 46, 46, 46, 46,

XY XX XX XY XY XY XX XY XY XY XX XX

T. ErgIÇn et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 95 (2001) 37±41

39

Table 2 Fetal sex definition in uterine lavage and endocervical lavage groups Chromosome analysis

FISH after uterine lavage XX

XY

XX XY

6 1

± 5

Total

7

5

er's Grove, IL, USA). The post-hybridization wash was performed in 0.4SSC/0.3% NP40 at 738C for 2 min. Slides were examined using an DAPI/FITS/Texas Red triple-band pass ®lterblock (Nikon) allowing the simultaneous visualization of orange, green and blue ¯uorophores. The images were obtained using `Applied Imaging Image Analyzer System with Cyto-Vision, Version 3.52, Santa Clara, CA, USA'. The statistical analysis included w2-analysis, and ttest. 3. Results Twenty-®ve sampling materials (13 uterine lavage and 12 endocervical lavage) were obtained for analysis (Table 1). The gestational ages were the same in both groups and were between 7±10 weeks (median 8 weeks). The age of the patients was between 21 and 34 years (median 27 years) and 23 and 35 years (median 29 years), respectively. Suf®cient cells were retrieved in 12 of the 13 (92.3%) in uterine lavage and 10 of the 12 (83.3%) in endocervical lavage group for FISH procedures for fetal sex prediction. The mean success rate of signal detection for FISH procedure was 91.7% (range 83±97%). Fetal sex was correctly

FISH after endocervical lavage Total

XX

XY

Total

6 6

3 2

± 5

3 7

12

5

5

10

predicted in 11 of 12 (91.6%) with uterine lavage and 8 of 10 (80.0%) in endocervical lavage and the difference was statistically insigni®cant (p>0.05). (Table 2) The FISH procedures did not detect male sex signals in any sample from pregnancies with female fetuses. There was no association between placental location and success rate of retrieved procedures. No aneuploidies of chromosomes X, Y or 18 were observed by FISH analysis and these data were also con®rmed by in vitro cytogenetics of placental cultures. Interestingly, in one cell of one sample, monosomy of chromosome 18 was determined by FISH (Fig. 1) and it was con®rmed by placental culture. 4. Discussion In the present study, we compared uterine and endocervical lavage to retrieve fetal cells for early prenatal diagnosis. Although our data is limited, we provided suf®cient cells for FISH procedures which is relevant to the fact that trophoblastic cellular elements were immunologically and morphologically proven to be available within (transcervical samples) the endocervical canal between 7 and 13 weeks of gestation [1,9±11]. The presence of trophoblasts and chor-

Fig. 1. Triple color fluorescence in situ hybridization (FISH) for X chromosome (green), Y chromosome (orange) and chromosome 18 (aqua) on a transcervical cell. The signal of chromosome 18 shows monosomy.

40

T. ErgIÇn et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 95 (2001) 37±41

Table 3 Incidence of fetal cells in transcervical cell samples retrieved by different methods Methods for collection

Test (FISH)

References

Endocervical lavage Endocervical lavage Cytobrush Endocervical lavage Aspiration Endocervical lavage Uterine lavage Uterine lavage Uterine lavage

7 in 7 5 in 6 4 in 6 6 in 6 2 in 5 1 in 1 8 in 8 5 in 5 10 in 12

[17] [16] [16] [8] [8] [9] [12] [15] [11]

ionic villi in the uterine cavity may be explained by evidences of migratory nature of trophoblasts or degeneration of the decidua capsularis layer to expose the chorionic villi to the uterine cavity [4±6]. There are two major factors to in¯uence the rate of fetal cell detection in TCC sampling; ®rst is the retrieval method and the second is molecular genetic technique. Both factors should accurately work to obtain a satisfactory result. There are numerous studies reporting modi®cations of the retrieval methods with varying successes [5±7], nevertheless, the methods using cotton swabs have been repeatedly shown to be unreliable [8]. Table 3 shows that the incidence of fetal cells in TCC samples retrieved by different methods. Both methods employed in our study are very simple and may be performed in minutes. The major concern of these methods for prenatal diagnosis is the risks to the mother and the fetus. Endocervical lavage has been reported to be simple, safe, and effective without any increase of the risks for infection or abortion [5,7,8,10,12,13]. However, there were three reported cases of limb defects in uterine lavage cases [14]. Although placental biopsy is not part of the technique of routine uterine lavage, placental damage, puncture of the amniotic sac can not be ruled out without ultrasound guidance, that may increase extremity abnormalities. Chang et al. de®ned uterine lavage as a minimally invasive and safe procedure and he reported no maternal or fetal complications in the follow-up of 13 ongoing pregnancies that underwent uterine lavage procedures [15]. One of the advantages of TCC sampling for prenatal diagnosis is collection of the samples at early stages of gestation and if the specimen fails to contain any fetal cells, then a second sample may be collected 1 week later [12]. We have shown that there are available cells of fetal origin in the lower part of the uterus and these cells may be collected successfully as early as 7 weeks of the gestation. In addition, our results show that endocervical lavage method is as effective as uterine lavage. Since uterine lavage is more likely to cause complications such as infection, abortion, and limb defects, endocervical lavage may serve better. Grif®th-Jones et al. [13], employed a polymerase chain reaction (PCR) assay to detect chromosome-Y-derived

sequences in TCC retrieved from mothers bearing male fetuses. However, subsequent studies expressed a frequency of false negative or positive results [10,16]. The ampli®cation of DNA sequences from contaminating cells derived from the seminal ¯uid of male partners may interfere with the correct detection. Recently, the FISH studies were found effective to detect fetal cells with Y chromosome signals in TCC samples retrieved from mothers with male fetuses [17] with an advantage of allowing enumeration of the incidence of fetal cells in each transcervically collected samples. In addition, double color FISH with X and Y probes allows to rule out the contamination by spermatozoa that have only one X or Y [6]. Multicolor FISH application not only detects fetal sex but also enables to detect other chromosomal anuploidies such as trisomies. We have shown that the ®rst trimester prenatal diagnosis from endocervical samples has potential as a non-invasive alternative procedure. Standardization of the procedure will render endocervical lavage as a potential alternative method to CVS. 5. Conclusion The ef®ciencies of uterine and endocervical lavage was compared to retrieve fetal cells for early prenatal diagnosis with FISH.

References [1] Simpson JL, Elias S (Eds.). Fetal cells in maternal blood: prospects for noninvasive prenatal diagnosis. Ann NY Acad Sci 1994;731:1± 270. [2] Adinolfi M. Breaking the blood barrier. Nature Genet 1992;1:316±8. [3] Adinolfi M. Non- or minimally invasive prenatal diagnostic tests on maternal blood samples or transcervical cells. Prenat Diagn 1995;15:889±97. [4] Shettles LB. Use of the Y chromosome in prenatal sex determination. Nature 1971;230:52. [5] Rodeck C, Tutschek B, Sherlock J, Kingdom J. Methods for the transcervical collection of fetal cells during the first trimester of pregnancy. Prenat Diagn 1995;15:933±42. [6] Adinolfi M, Sherlock J. First trimester prenatal diagnosis using transcervical cells: an evaluation. Hum Reprod Update 1997;4:383± 92. [7] Pertl B, Yau SC, Sherlock J, Davies AF, Mathew CG, Adinolfi M. Rapid molecular method for prenatal detection of Down syndrome. Lancet 1994;343:1197±8. [8] Adinolfi M, Sherlock J, Tutschek B. Detection of fetal cells in transcervical samples and prenatal diagnosis of chromosomal abnormalities. Prenat Diagn 1995;15:943±51. [9] Adinolfi M, Sherlock J, Kemp T. Prenatal detection of fetal RhD DNA sequences in transcervical samples. Lancet 1995;345:318±9. [10] Massari A, Novelli G, Colosimo A, Sangiuolo F, Pulka G, Calabrese G. Non-invasive early prenatal molecular diagnosis using retrieved transcervical trophoblast cells. Hum Genet 1996;97:150±5. [11] Daryani YP, Penna LK, Patton MA. Detection of cells of fetal origin from transcervical irrigations. Prenat Diagn 1997;17: 243±8.

T. ErgIÇn et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 95 (2001) 37±41 [12] Ishai D, Amiel A, Duikman R, Cogan O, Liehtenstain Z, Abramovici H, Fejgin MD. Uterine cavity lavage: adding FISH to conventional cytogenetics for embryonic sexing and diagnosing common chromosomal aberrations. Prenat Diagn 1995;15:961±7. [13] Griffith-Jones MD, Miller D, Lilford RJ, Scott J, Bulmer J. Detection of fetal DNA in trans-cervical swabs from the first trimester of pregnancies by gene amplification: a new route to prenatal diagnosis? Br J Obstet Gynaecol 1992;99:508±11. [14] Chou MM, Lin SK, Hoe SC. Severe limb reduction defects after uterine lavage at 7±8 week's gestation. Prenat Diagn 1997;17:77±80.

41

[15] Chang SD, Lin SL, Chu KK, Hsi BL. Minimally-invasive early prenatal diagnosis using fluorescence in situ hybridization on samples from uterine lavage. Prenat Diagn 1997;17:1019± 25. [16] Kingdom J, Sherlock J, Rodeck C, Adinolfi M. Detection of trophoblast cells in transcervical samples collected by lavage or cytobrush. Obstet Gynecol 1995;86(2):283±8. [17] Adinolfi M, Davies A, Sharif S, Soothill P, Rodeck C. Detection of trisomy 18 and Y derived sequences in fetal nucleated cells obtained by transcervical flushing. Lancet 1993;342:403±4.