The Use of DNA Flow Cytometry in the Diagnosis of Triploidy in Human Abortions M. W. VAN OVEN, MD, C. J, F. SCHOOTS, MD, J. W. OOSTERHUIS, MD, J. F. KEIJ, BS, A. DAM-MEIRING, BS, AND H. J. HUISJES, MD The ploidy of placental tissue from 59 abortions was determined using DNA flow cytometry. The specimens were also screened histologically for features considered to be characteristic of a partial mole. Triploidy was found in six cases; three of these cases presented histologically as partial moles. From the four specimens histologically classified as partial moles, three appeared to be triploid. The most specific histologic features of triploidy were cystic change of the placental villi and trophoblastic hyperplasia with vacuolation of the syncytiotrophoblast; however, these changes were also found in some of the nontriploid abortions. Irregularity of villous contours was less specific. HUM PATHOL 20:238-242. 0 1989 by W.B. Saunders Company.
mole is not absolute, in the sense that not all partial moles are triploid,1*4-7 and not all triploidies present themselves as partial moles,3 the diagnosis of triploidy cannot be made with certainty on morphologic grounds alone. The possibility of measuring DNA ploidy by flow cytometry provides a simple way to assess triploidy in placental tissue. In this investigation, the results of morphologic examination and the measurement of DNA ploidy are compared in a series of abortions. MATERIALS AND METHODS
A relatively common cause of human abortion appears to be a triploid chromosomal constitution of the conceptus, which in most cases is the result of fertilization of a haploid egg1 by two sperms.’ Triploidy is thought to occur in approximately 8% of all spontaneous abortions. * Triploid abortions are not associated with an increased risk of recurrence. The gross and microscopic appearances of the placentas of these abortions are the most recognizable of all chromosomal aberrations and are welldescribed by Szulman and colleagues.3-5 The majority of these abortions exhibit the features of a partial mole. On gross examination, partial moles are characterized by the presence of vesicles, a result of hydropic change that has occurred in a part of the chorionic villi. Microscopically, this hydropic alteration appears to be so extensive that cisterns are formed centrally in the villi. Furthermore, slight or evident focal proliferation of the trophoblastic lining of the villi is present, often with vacuolation of the syncytium. Another feature is scalloping of the villous outlines which can produce the formation of trophoblastic inclusions. The presence of parts of the embryo proper or erythroblasts, and the finding of both normal and cystic villi, are helpful in excluding a complete mole. Szulman and colleagues insist on the combination of hydropic change with central cisterns and hyperplasia of trophoblast before making the diagnosis of a partial mole. As the relationship between triploidy and partial
The investigated series was composed of both spontaneous abortions and suction curettage material obtained after intrauterine death which was detected by ultrasound examination. All slides from these abortions received between December 1986 and November 1987 were screened. The few cases in which the menstrual age exceeded 15 weeks were excluded in order to obtain a narrow age range. A further selection criterion was the presence of both placental and maternal (decidua) tissue in such amounts that representative DNA histograms could be anticipated from the paraffin blocks. Maternal tissue had to be present as the internal diploid control; however, the amount should be small enough to avoid obscuring a triploid fetal cell population. In some cases, both components were present in adequate amounts in one paraffin block. In other cases, only a part of one block was selected, and in a few cases, a combination of two blocks was used. Nuclear suspensions were made according to methods described by Hedley et al,B with minor modifications. Briefly, 50 km-thick slides were cut, dewaxed in xylene, rehydrated in decreasing concentrations of alcohol, and suspended in pepsin. The enzymatic digestion was stopped by adding ice-cold phosphatebuffered saline. The resultant naked nuclei were stained with propidium-iodide after RNase digestion, filtered through nylon gauze, and run through an FACS 440 flow cytometer. As the use of internal standard controls in the measurements of suspensions prepared from paraffinembedded tissue provide unsatisfactory result9 and hypodiploid cell populations should not be present in placental tissue, the first peak in the histograms was considered to represent diploid cells. A triploid cell population was assumed when two distinct peaks were present with the ratio of the DNA contents of the first and second peak being 1:1.4 to 1.6. Coefficients of variation (CV) were assessed with the use of the computer program “Lysys” (Becton-Dickinson, Mountain View, CA). The CV of the diploid peak varied between 4.0% and 15.1% (mean, 7.7%). All hematoxylin-eosin-stained slides of the selected abortions were screened by an experienced pathologist (C.J.F.S.), who had no knowledge of the results of the DNA measurements, using the following parameters: the presence and degree of hydrops, the prominence of the tropho-
From the Departments OS Pathology, Clinical Immunology, and Obstetrics andf Gynaescology, University Hospital of Croningen, The Netherlands. Accepted for publication August 2, 1988. A.D.M. and J.W.O. are partly supported by “Pediatric Oncology Foundation Groningen”(SKOG). Key Words: DNA flow cytometry, triploidy, abortion, partial mole. Address correspondence and reprint requests to M. W. van Oven, MD, Department of Pathology, University Hospital of Groningen, Oostersingel 63, 9713 EZ Groningen, The Netherlands. 0 1989 by W.B. Saunders Company. 0046-8177/89/2003-0008$5.00/O
238
DNA FLOW CYTOMETRYAND TRIPLOIDY(vanOven et al)
blast, the regularity of villous contours, and the presence of embryonal tissue or erythroblasts (Table 1). If the degree of villous edema was scored as 3 (central cisterns present) and the status of the trophoblast was
three triploid abortions did not: in one of these, villous edema with cisterns was present, but the trophoblast was considered to be normal, and the remaining two did not exhibit clear features of a partial mole (nor did the tetraploid abortion). Of the four partial moles, one was diploid. Hydropic villous change with cisterns was found in four of the 52 diploid abortions and in four of the six triploid abortions. Trophoblastic hyperplasia was found in seven diploid abortions and in three triploid abortions. Trophoblastic hyperplasia accompanied by vacuolation of the syncytiotrophoblast was found in three diploid cases and in three triploid cases. Irregular villous contours were found in 19 diploid abortions, four of which were scored as highly irregular. Irregular villous contours were found in all triploid abortions, but was scored as very irregular in only one of these cases.
scored as 2 or 3 (hyperplasia present), the diagnosis of partial mole was made. Because this was a retrospective investigation, no attempt was made to evaluate gross features of the placentas.
RESULTS (Table 2) DNA Flow Cytometry The series selected for DNA measurement consisted of 59 cases. Interpretable DNA histograms were obtained from all samples. Fifty-two histograms showed one high peak considered to be diploid GO/ Gl cells and occasionally showed a small second peak representing the G2/M cells (Fig 1A). A high peak was found in the G2/M region in only one case (Fig 1B). In view of the low mitotic activity in the specimen and the presence of an octaploid population, albeit small, the tetraploid peak was considered to represent tetraploid GO/G1 placental cells. The remaining six DNA histograms showed triploidy (Fig IC-H). The DNA indices of the triploid peak varied between 1.44 and 1.56. In none of the histograms were additional peaks found outside the triploid or tetraploid regions.
DISCUSSION The method of DNA measurement by flow cytometry is simple, inexpensive, and suitable to the assessment of triploidy in placental tissue in routine practice. The use of paraffin-embedded tissue makes microscopic selection of relevant material possible. Careful selection of material is important because both maternal and placental tissues should be present. Our experience in this investigation, and in earlier measurements of cytogenetically proven triploid placentas, is that the placental cell population is more easily underrepresented than the maternal cell population. This is probably due to inflammatory infiltrate and blood of maternal origin and to the high cellularity of decidual tissue compared with the placental villi. The possibility of detecting a very small triploid cell population depends on the CV of the “peak” and on the amount of nuclear fragments that influence baseline noise.g Figure 1H shows a triploid cell population that is about the smallest that can be reliably distinguished. In this investigation, we completed a careful histologic sampling and assume that we did not miss triploid placentas. There is much disagreement between investigators about the specificity of the individual morphologic features of triploidy. Szulman and colleagues found villous edema to be common in abortions with normal karyotype and in trisomic abortions, but cistern formation was practically never observed. Nor did they encounter trophoblastic hyperplasia or scalloping of villous outlines in these abortions.3 Geisler and Kleinebrecht performed a histologic analysis of the placentas of 107 spontaneous abortions with known karyotype. lo The seven triploidies in this series all presented as a partial mole, (partial mole being defined as cystic degeneration without hypertrophy of the trophoblastic epithelium). Partial moles were not found among other chromosomal aberrations or abortions with normal karyotype. Honore et al found cystic change in 39 of 66
Microscopic Examination Four cases fulfilled the morphologic criteria of a partial mole (Figs 2 and 3). In three of these cases, embryonal tissue or erythroblasts were found. In the fourth case, embryonal tissue or erythroblasts were lacking; however, the focal character of the hydropic and hyperplastic changes ruled out the possibility of a complete mole. Relationship Between Ploidy and Histologic Features Of the six triploid abortions, three appeared to fulfill the criteria of a partial mole. The remaining TABLE 1. Criteria Used in the Microscopic Evaluation of Studied Abortions Degree of villous edema
0. No edema 1. Little edema 2. Extensive edema in at least part of the villi, but no central cisterns
3. Extensive edema in at least part of the villi and presence of central cisterns Trophoblast 0. Hypoplasia
1. Normal 2. Hyperplasia 3. Hyperplasia
with vacuolated syncytium
Villous contours 0. Regular 1. Slightly irregular 2. Very irregular Embryonal tissue or erythroblasts
+ , Present - , Absent 239
HUMANPATHOLOGY TABLE 2. Case No. 1 2 3 4 5 6 7 8 1: (PM) 11 12 13 14 15 16 17 18 :: :; 23 (PM) 24 25 26 27 28 29 30 31 32 33 34 35 36 :: (PM) 39
Volume 20. No. 3 (March1989)
Results of Microscopic Evaluation and Ploidy Measurements of 59 Abortions
Degree of Villous Edema
Villous Contours
Trophoblast
2 1 2 1 2 1 2 1 2 3 2 2 : : 1 2 2 2 2 2 3 1 0 1 : 2 1 : 1 : 2 1 3
1 1 1 1 1 2 3 3 1 3 3 1 1 1 1 1 1 1 1 1 1 1 3 1 1 1 1 1 1 1 1 2 1 1 1 1 1 3 1
0 0
0 0 0 1 0 1 0 1 1 1 2 0 0 0 1 0 0 1 0 0 1 0 0 0 0 1 0 2 0 1 0 0 0
2
0 2 1 1 1 1 1 0 0 1 1 0 0 0 1 0 0 0 0 1 0 1 2
: t; (PM) 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59
3 3 2 2 2 2 0 2 2 3 3 2 2 2 2 0 1 3 2
: 1 2 1 1 1 1 1 1 0 1 1 1 1 0 1 1 1 1
Embryonal Tissue on Erythroblasts
_ + _ _ + + + + _ + + + _ + + _ + + + + + + + _ + + + _ + + + _ + + + + + + + _ + + + + + + + +
Ploidy Didoid Diploid Diploid Diploid Diploid Diploid Diploid Diploid Diploid Triploid Diploid Diploid Diploid Diploid Diploid Diploid Diploid Diploid Diploid Triploid Diploid Diploid Triploid Diploid Diploid Diploid Diploid Diploid Diploid Diploid Diploid Diploid Diploid Diploid Diploid Diploid Diploid Triploid Diploid Diploid Diploid Diploid Diploid Diploid Diploid Diploid Triploid Diploid Diploid Diploid Triploid Diploid Diploid Diploid Tetraploid Diploid Diploid Diploid Diploid
NOTE. Refer to Table 1 for definitions of results. Abbreviation: PM, partial mole.
trisomic placentas, and described the cytotrophoblast and syncytiotrophoblast of these placentas as more pronounced than in triploid placentas. l l Vassilakos et al karyotyped 49 partial moles, identified by grossly recognizable swelling of part of the villi, and found nine normal karyotypes, 22 triploidies, and 17 triso-
mies (especially trisomy 16). In their microscopic descriptions, Vassilakos et al mentioned that slight hyperplasia of the syncytiotrophoblast was present in only a minority of the cases.12 In our series, neither hydropic change with central cistern, trophoblastic hyperplasia, nor irregular240
DNA FLOW CYTOMETRY AND TRIPLOIDY (van Oven et al] 11ou
A
-z..,l
0
w
EI
-.
~‘~~l’~~~I’t’~I~‘~~l~” lCl0
2ww
0
400
100
200
700
0 420
E00
100 ,
200
0
100
200
,
F
E
300
FIGURE 1. Examples of the three kinds of DNA histograms. Vertical axis, number of counted events; horizontal axis, channel number, representing the relatlve DNA content, [A) Normal diploid DNA histogram. One high peak is considered to be diploid _ maternai and placental cell populations. The small peak represents the G2/M cells. [B] DNA histogram expressing tetraploidy. The sizes of the tetraploid peak and the attending small octoploid population are inconsistent with normal G2/M cells. (C-H] DNA histograms expressing triploidy. The first peak represents maternal diploid cells and the second peak represents placental cells with a triploid DNA content.
I
ity of villous contours appeared to be specific for triploidy. Central cisterns were associated as many times with triploidy as with diploidy. These diploid cases might have been trisomic abortions, which cannot be assessed by DNA flow cytometry. Trophoblastic hyperplasia is a far less specific change, which might, in part, be the result of the fact that the assessment of the presence of hyperplasia is difficult and subjective. The trophoblast in young placentas may be very active, depending on the localization of the villi. A prominent proliferating trophoblast is present especially in the anchoring villi. Furthermore, activation
of trophoblast is described as an aspecific result of ischemic change in the placenta.13 Trophoblastic hyperplasia accompanied by vacuolation of the syncytium is more specific for triploidy, but was also found in some diploid abortions. Irregularity of villous contours was frequently found in diploid abortions, often to a more extensive degree than that found in triploid abortions. We consider partial moles and triploid abortions to be different entities. The former is a morphologic diagnosis; the latter diagnosis can be made by cytogenetics or DNA measurement. The diagnosis of par241
HUMAN PATHOLOGY
Volume 20, No. 3 (March 1989)
gators, l4 we do not propose that the finding of triploidy should influence the morphologic diagnosis. Demonstrating triploidy establishes the cause of abortion. A reasonable approach is to measure those abortions in which hydropic villi with central cisterns are present. This strategy would have revealed four of the six triploidies in our series. REFERENCES
FIGURE 2. Hydropic villi with central cisterns. Case no. 23. (Hematoxylin-eosin stain; magnification x 25.)
tial mole can be difficult to make which cystic villi are found without ation of trophoblast or the reverse; trophoblast is evident, but cisterns are lacking. However, in contrast
in those cases in evident proliferie, hyperplasia of in hydropic villi to other investi-
FIGURE 3. Focal proliferation of trophoblast accompanied by vacuolation of syncytiotrophoblast. Case no. 23. (Hematoxylin-eosin stain; magnification x 40.)
1. Lawler SD, Fisher RA: Genetic studies in hydatidiform mole with clinical correlations. Placenta 8:77-88, 1987 2. Huisjes HJ: Spontaneous Abortion. Edinburgh, Churchill Livingstone, 1984, p 59 3. Szulman AE, Philippe E, BouC JG, et al: Human triploidy: Association with partial hydatidiform moles and nonmolar conceptuses. HUM PATHOL 12:1016-1021. 1981 4. Szulman AE, Surti U: The syndromes of hydatidiform mole. I. Cytogenetic and morphologic correlations. Am J Obstet Gynecol 131:655-671, 1978 5. Szulman AE, Surti U: The syndromes of hydatidiform mole. II. Morphologic evolution of the complete and partial mole. Am J Obstet Gynecol 132:20-27, 1978 6. Teng NNH, Ballon SC: Partial hydatidiform mole with diploid karyotype: Report of three cases. Am J Obstet Gynecol 150:961-964, 1984 7. Hemming JD, Quirke P, Womack C, et al: Diagnosis of molar pregnancy and persistent trophoblastic disease by flow cytometry. J Clin Pathol 40:615-620. 1987 8. Hedley DW, Friedlander ML, Taylor IW, et al: Method for analysis of cellular DNA content of paraffin-embedded pathological material using flow cytometry. J Histochem Cytochem 31:1333-1335, 1983 9. Stephenson RA, Gay H, Fair WR, et al: Effect of section thickness on quality of flow cytometric DNA content determinations in paraffin-embedded tissues. Cytometry 7:41-44, 1986 10. Geisler M, Kleinebrecht J: Cytogenetic and histologic analyses of spontaneous abortions. Hum Genet 45:239-251, 1978 11. Honor+ LH, Dill FJ, Poland BJ: Placental morphology in spontaneous human abortions with normal and abnormal karyotypes. Teratology 14: 151-166, 1976 12. Vassilakos P, Riotton G, Kajii T: Hydatidiform mole: Two entities. A morphologic and cytogenetic study with some clinical considerations. Am J Obstet Gynecol 127:167-170, 1977 13. Fox F: Histological abnormalities of the placenta, in Bennington JL (ed): Major Problems in Pathology, vol 7. Philadelphia, Saunders, 1978, pp 149-197 14. Lage M, Driscoll SC, Yavner DL, et al: Hydatidiform moles. Application of flow cytometric in diagnosis. Am J Clin Path01 89:596-600, 1988