- Email: [email protected]
Flow cytometric crossmatch and early pregnancy loss in women with a history of recurrent spontaneous abortions who underwent paternal leukocyte immunotherapy
Flow cytometric crossmatch and early pregnancy loss in women with a history of recurrent spontaneous abortions who underwent paternal leukocyte immunotherapy Tetsuo Maruyama, MD, Tsunehisa Makino, MD, PhD, Toshitaka Sugi, MD, Ken-ichi Iwasaki MD, Nobuaki Ozawa, MD, Hidehiko Matsubayashi, MD, and Shiro Nozawa, MD, PhD
Tokyo, Japan OBJECTIVE: The purpose of our study was to evaluate the clinical significance of flow cytometric crossmatch in managing women with unexplained recurrent spontaneous abortions who have been immunized with paternal leukocytes. STUDY DESIGN: Two-color flow cytometric crossmatch was performed after immunotherapy in 69 women with recurrent spontaneous abortions to detect maternal immunoglobulin G antibodies against paternal T lymphocytes. A positive flow cytometric cross match was determined by a ~ 20 channel increases (256 channel linear scale-log amplification) in the fluorescence intensity of the T-cell peak. We analyzed the clinical correlation between flow cytometric crossmatch results and subsequent pregnancy outcome in those patients using the Fisher exact test. RESULTS: In 46 subsequent pregnancies among 69 immunized women the incidence of early subsequent pregnancy wastage was only 6 (17.1 %) of 35 in flow cytometric crossmatch-positive patients and 7 (63.6%) of 11 in flow cytometric crossmatch-negative patients I.p < 0.01). CONCLUSIONS: These data indicate that a two-color T cell flow cytometric crossmatch predicts subsequent pregnancy outcome in patients undergoing immunotherapy for recurrent spontaneous abortion. (AM J QBSTET GVNECOL 1993;168:1528-36.)
Key words: Habitual abortion, flow cytometric crossmatch, immunotherapy
Since leukocyte-enriched plasma transfusions or subcutaneous leukocyte injections were first used to treat women with recurrent spontaneous abortions in 1981," 2 immunotherapy has become generally accepted as an efficacious treatment for unexplained repeated spontaneous abortions. Most trials of such immunotherapy have shown about a 70% success rate." 4 In the first randomized, controlled trial Mowbray et al. 5 reported a 77% success rate in patients who received immunotherapy with their spouses' leukocytes, whereas only 37% of controls immunized with their own leukocytes were delivered successfully: By contrast, Ho et al." and Cauchi et al.? recently reported two randomized controlled trials with the husband or third-party From the Department of Obstetrics and Gynecology, School of Medicine, Keio UniverSity. Supported in part by a research grant from Keio Gijuku Academic Development Funds (1990). Received for publication June 4, 1992; revised August 20, 1992; accepted October 20, 1992. Reprint requests: Tetsuo Maruyama, MD, Department of Obstetrics and Gynecology, School of Medicine, Keio University, 35 Shinanomachi, Shinjuku-ku, Tokyo 160, Japan. Copyright © 1993 by Mosby-Year Book, Inc. 0002-9378193 $1.00 + .20 6/1/43617
1528
blood lymphocyte immunotherapy; these trials failed to show any significant benefit. Because of these conflicting studies immunotherapy has generated much controversy. Clark and Daya4 and Coulam and Clark" pointed out several controversial issues in immunotherapy: sample sizes of such trials have been small, the populations eligible for the immunotherapy have been heterogeneous, and there have been some cointerventions by the control treatment that may affect success rates in the control group. In addition to these problems standard immunologic assays such as mixed lymphocyte culture reaction-blocking assay or microlymphocytotoxicity crossmatch, which are usually used to select populations for immunotherapy and to demonstrate sensitization to donor leukocytes, might be inappropriate for immunologic monitoring because they may be insensitive techniques. 9 Criteria based on conventional immunologic examinations may fail to select sensitized patients and, thereby, may select immunologically heterogeneous populations. Furthermore, the prognostic value of assays for blocking antibodies and cytotoxic anti paternal antibodies is limited because of these problems. 5 • 10-12 In renal transplantation flow cytometric crossmatching has proved to be a very sensitive method for
\'uillme 168. l'\lImber " Am J Obslel Gynecol
detecting T -cell- and B-cell-directed antibodies and has identified a subgroup of patients who, although negative by conventional microlymphocytotoxicity crossmatch, not only possess immunoglobulin (Ig) G antibodies directed against donor lymphocytes but are also more likely to reject their grafts. 13-18 Our study was designed to evaluate the clinical significance of flow cytometric crossmatch in managing women with unexplained recurrent spontaneous abortions who have been immunized with paternal leukocytes. Material and methods Patients. Between June 1990 and March 1992, 69 women with a history of two or more unexplained, consecutive, first-trimester spontaneous abortions with the same partner were included in this study. The couples were thoroughly investigated for possible causes of spontaneous abortion, including the presence of a genetic impairment, infectious disease, hormonal deficiency, metabolic disorder, uterine anomaly, and any autoimmune abnormality such as anticardiolipin antibody. There was no family history of recurrent loss and congenital anomalies in any couple. Mter those causes were ruled out, and the patients were thoroughly counseled about the risk of white blood cell administration, immunotherapy was performed with the spouse's leukocytes. All donors were screened and found to be negative for hepatitis, syphilis, and antibodies to HIV and HTLV-l. Immunization procedures. Mononuclear cells (usually 4 x 10 7 ) were prepared from 30 ml of heparinized whole blood taken from the husband, separated by density centrifugation, concentrated to 2 ml under sterile conditions, and injected subcutaneously into the forearms of the patient at six to eight sites. This procedure was repeated four times at intervals of 4 to 6 weeks before each couple was permitted to attempt pregnancy. Flow cytometric crossmatch was performed for each couple 2 to 4 weeks after the last immunization. Flow cytometric crossmatch. Flow cytometric crossmatch was performed with a modification of the technique proposed by Gilman-Sachs et al."· 19 in which maternal heat-inactivated serum and paternal lymphocytes were used to detect maternal anti paternal lymphocyte antibodies. Maternal sera were collected before treatment at each immunization and at 2 to 4 weeks after the last immunization, pooled, and then frozen until flow cytometric crossmatch was performed. For each husband, lymphocytes were separated from heparinized whole peripheral blood by Ficoll-Conray (IBL, Gunma, Japan) density gradient centrifugation. Five x 105 paternal lymphocytes were incubated with 20 IJ..i of each maternal serum and 100 ILl of phosphate
Maruyama et at.
1529
buffered saline solution for 15 minutes at room temperature. As a negative control, type AB serum (lCN Flow, Costa Mesa, Calif.) was used instead of maternal serum. The cells were washed twice at 4° C by centrifugation. Optimum amounts of F(ab)'2-fluorescein isothiocyanate-conjugated goat antihuman IgG (Cappel, West Chester, Pa.) and phycoerythrin-conjugated antihuman CD3 or CD 19 antibody (Becton Dickinson, Mountain View, Calif.) were then added to each tube to detect maternal antibody bound to the paternal T- or B-lymphocyte surface. The amount of antibody used in the assay was predetermined from titration curves. Samples were incubated for 15 minutes in the dark at 4° C. Mter two washes the cells were suspended in 300 ILl of phosphate buffered saline solution. The labeled samples (in phosphate buffered saline solution) were analyzed with a FACScan flow cytometer (Becton Dickinson). Data from 10,000 cells were collected in list mode files by using logarithmic amplification of fluorescence parameters. Data acquisition and analysis was performed with Consort 30 software (Becton Dickinson). The FACScan uses a four-decade log amplifier. Flow cytometric crossmatch analysis. For the purpose of our study a positive flow cytometric crossmatch was that in which a ~ 20 channel shift to the right of the T-cell peak ~ 14 channel shift of the B-cell peak in the test serum or compared with the control serum on a 256-channel linear scale. This cutoff point was chosen on the basis of the mean + SD of fluorescence intensity increase of the cell peak treated with the preimmunization serum compared with the control serum. These cutoff points are summarized in Table I. In this analysis the log scale channel was converted to a linear channel by the following formula: Linear channel = Channels-per-decade x Log (log channel) The channels-per-decade for 256-channel resolution data is 63.75 in the FACScan and the Consort 30 software system. A shift is the difference in flow cytometric crossmatch parameter between the test serum and the control serum. Mean channel reflects the average channel value of the horizontal parameter for the processed cells. Peak channel reflects the horizontal channel value of the point with the highest cell count. The percentage of positive lymphocytes reflects the proportion of cells that bound antibodies (Table I). Our study was performed in a double-blind fashion because the patients were not informed of their own flow cytometric crossmatch results unless their subsequent pregnancies ended in spontaneous abortions and because we did not know the subsequent pregnancy outcome of each patient when analyzing their data. Histocompatibility testing. HLA-A, -B, -C, -DR, and
1530 Maruyama et al.
May 1993 Am J Obstet Gynecol
Table I. Summary of flow cytometric crossmatch parameters in women immunized with paternal leukocytes according to choice of cutoff point for each parameter T flow cytometric crossmatch * Mean ± SD
Mean channel shift Control vs preimmunization Control vs postimmunization Peak channel shift Control vs preimmunization Control vs postimmunization Shift of percentage of positive lymphocytes Control vs preimmunization Control vs postimmunization
I
B flow cytometric crossmatcht
I
Cutoff pointt
Mean ± SD
Cutoff point
5 ± 18 38 ± 29§
23
-1 ± 14 II ± 2311
13
5 ± 15 45 ± 27~
20
-5 ± 19 16 ± 29#
14
2 ± 10 43 ± 39**
12
1± 4 5 ± 10tt
5
Values were obtained for 256-channel, 2-parameter histogram. *55 Of 69 women immunized with patemalleukocytes were tested by T flow cytometric crossmatch before immunization. t48 of 69 women were tested by B flow cytometric crossmatch after immunization; 42 of 48 women were examined by B flow cytometric crossmatch before immunization. tChoice of cutoff point was based on mean + SO of each value ofT or B flow cytometric crossmatch (control vs preimmunization). §p < 0.0001, compared with mean channel shift of control T cells vs preimmunization. lIP < 0.001, 'compared with mean channel shift of control B cells vs preimmunization. ~p < 0.0001, compared with peak channel shift of control T cells vs preimmunization. #p < 0.0001, compared with peak channel shift of control B cells vs preimmunization. **p < 0.0001, compared with shift of percent of positive control T cells vs preimmunization. ttp < 0.005, compared with shift of percent of positive control B cells vs preimmunization.
-DQ typing was performed on 69 couples with the National Institutes of Health standard microlymphocytotoxicity test at the Clinical Laboratory Department of Keio University Hospital and a commercial laboratory (SRL, Hachioji, Tokyo). Statistical analysis. The Fisher exact test or the X2 test for independence was used to formulate two-by-two tables and evaluate the association between flow cytometric crossmatch results and subsequent pregnancy outcome. The Wilcoxon rank sum test or the impaired t test was used selectively to analyze differences in several parameters between the successful subsequent pregnancy outcome group and the unsuccessful group. The paired t test was used to analyze any change in the flow cytometric crossmatch value between the preimmunization and postimmunization period. A probability of p < 0.05 was chosen to represent statistical significance. Results
Of 69 patients who underwent immunotherapy, 49 became pregnant as this manuscript was being prepared. Patient characteristics are summarized in Table II. Included in the calculation of the pregnancy success rate were live births and pregnancies of > 24 gestational weeks. There were no second- or third-trimester losses. The flow cytometric crossmatch values and cutoff
points in women immunized with paternal leukocytes are summarized in Table I. After immunotherapy the production of maternal antibodies against paternal T or B lymphocytes increased gradually in most immunized patients (Fig. 1). Each postimmunization T- and Bflow cytometric crossmatch value increased significantly compared with the corresponding preimmunization flow cytometric crossmatch (Table I). Of 69 patients after immunization, 52 (75%) had a positive T flow cytometric crossmatch and 9 of 55 patients (16%) who were tested by T flow cytometric crossmatch before immunization were flow cytometric crossmatch-positive. Of 48 patients after immunization, 21 (44%) had a positive B flow cytometric crossmatch, and 5 of 42 patients (12%) had a positive B flow cytometric crossmatch before immunization. In 46 women with a subsequent pregnancy among 69 immunized patients, the incidence of early pregnancy loss was only 17.1% (6 of 35) in flow cytometric crossmatch-positive patients and 63.6% (7 of 11) in flow cytometric crossmatch-negative patients (P < 0.01) (Table III). Furthermore, among 31 of 46 pregnant patients with a history of three or more spontaneous abortions, the incidence of unsuccessful pregnancy was only 19.2% (5 of 26) in flow cytometric crossmatchpositive patients and 80% (4 of 5) in flow cytometric crossmatch-negative patients (p < 0.05) (Table IV). No correlation existed between subsequent pregnancy out-
Volume 168, Number 5 Am J Obstet Gynecol
Maruyama et al.
(A) AB SERUM
1531
(B) PRE-IMZ MEAN=71
MEAN=95 (SHIFT=24)
(0) POST-IMZ(2)
PEAK=93
MEAN=72
MEAN=115 (SHIFT=44) PEAK=109
(E) POST-IMZ(4)
MEAN=126 (SHIFT=55) PEAK=113
Fig. 1. Three-dimensional contour histograms illustrating green fluorescence intensity on horizontal axis (caused by labeling with FITC-conjugated anti-IgG to paternal T-cell population), versus red fluorescence intensity on vertical axis (caused by labeling with phycoerythrin-conjugated anti-CD3 antibody), Each axis is four-decade logarithmic scale. Gate was set to encompass T-lymphocyte population. Number of cells is proportional to height of signal. A, Mean green fluorescence and peak channels for paternal T cells incubated with type AB serum were 71 and 53, respectively. B, Mean green fluorescence channel for paternal T cells incubated with maternal preimmunization (IMZ) serum was 72 (shift of one channel); peak channel was 57 (shift of 4 channels). C, D, and E, Postimmunization serum obtained at 49, 80, or 153 days, respectively, after first immunization was used. Blood samples were collected simultaneously with immunotherapy. Panel representing use of serum obtained 140 days after first immunization is not shown. Mean and peak channel numbers are indicated in each panel. Mean channel shift and peak channel shift are indicated in parentheses in each panel.
come and the B-flow cytometric crossmatch result regardless of the cutoff point chosen. The successful and unsuccessful outcome groups did not differ significantly in age, number of spontaneous abortions, number of shared HLA-A, -B, or -C loci, or interval between the last immunization and subsequent conception. However, HLA-DR and -DQ loci were more likely to be common to both partners in the successful group than in the unsuccessful group (Tables III and IV). Comment
In renal transplantation, flow cytometry has recently come into use to supplement standard crossmatch methods to detect recipient pre sensitization to donor tissues. A positive flow cytometric crossmatch is associated with a greater incidence of rejection episodes, even when a conventional standard cross match IS negative. 13-18 In managing patients with unexplained recurrent
spontaneous abortions, mixed lymphocyte culture reaction-blocking assay and microlymphocytotoxicity crossmatch have been applied for immunologic assessment and monitoring. However, the association between these conventional assays and pregnancy outcome remains controversial. 6 • ID-I2. 20-22 Gilman-Sachs et al. 9 reported that the detection of antilymphocyte antibodies by flow cytometric crossmatch is dearly more sensitive than the standard microlymphocytotoxicity crossmatch, because twice as many positive couples were detected by flow cytometric crossmatch than by microlymphocytotoxicity crossmatch (66% vs 33%). Two-color flow cytometric crossmatch demonstrated a correlation between inhibition of binding of anti-CD3 antibodies to paternal lymphocytes and failure of immunotherapy to treat recurrent spontaneous abortions. I9 In kidney transplantation the degree of sensitization to donor tissues reflected by the shift of corresponding
1532 Maruyama et al.
May 1993
Am
J Obstet Gynecol
Table II. Characteristics of patients classified according to number of spontaneous abortions No. of spontaneous abortions
5
No.
Patients No. of patients with live births Pregnancy outcome Infants delivered Pregnant beyond 24 wk Aborted Fetal chromosomal analysis Normal Abnormal Ectopic Molar Awaiting conception No. of successful outcomes No. of total pregnant patients§ Success rate (%) Maternal age§ Mean ± SD Range Interval daysll Mean ± SD Range
3 0
No. T-FCXMpositive
4 No. T-FCXMnegative
3
1 1 0 0 0 0 0 1 2 2 100
1
2 2
100
No.
No. T-FCXMpositive
o
13 3*
12 2
o o
2 5 1
2 5 1
o o o
31.5 ± 3.5 28-35 11 ± 9 2-20
1 0 0 0 5 7 8 87.5
3 No. T-FCXMnegative
4 7 8 87.5
No.
No. T-FCXMpositive
No. T-FCXMnegative
33
It
25 1
8 0
0 0 0
8 5 8
7 5 4
1 0 4
0
6 2 1 0 11 13 21 61.9
2 2 0
4 0 1
9 12 16 75.0
2 1 5 20.0
1 0 0
31.3 ± 2.9 26-35
31.4 ± 3.1 27-38
44.5 ± 26.0 1-83
44.7 ± 30.1 0-108
FCXM, Flow cytometric crossmatch.
*None of the three patients conceived after immunization. tShe conceived after immunization. tEach patient had one child before immunization. §Molar or ectopic pregnancy outcome and patients awaiting conception were excluded. IIInterval days between the last immunization and subsequent conception. Conception date was determined to be seventh day of high phase of basal body temperature.
cells incubated with recipient serum can reportedly be quantitated by various parameters of the flow cytometric crossmatch such as mean channel shift, 16 peak channel shift, [3. 14. 18 percent relative mean channel number shift, 15 and the percentage of positive cells. l4 • 17 We could not find correlation between subsequent pregnancy outcome and flow cytometric crossmatch results in the shift of percentage of positive CD3 cells or mean channel shift, regardless of the cutoff point chosen. One other method by which the cutoff point may be established is from the receiver-operator characteristic curve."3 Usually the best cutoff point is where the receiver-operator characteristic curve "turns the corner."23 Furthermore, the better a parameter the closer its curve is to the upper left corner."3 We compared the characteristics of curves representing different parameters and then used the peak channel shift on flow cytometric crossmatch of T cells as a significant prognosticator of subsequent pregnancy outcome and an indicator of maternal antibody against paternal lym-
phocytes. In evaluating the significance of flow cytometric crossmatch in patients who had three or more spontaneous abortions, the sensitivity of this test was 44% (95% confidence interval; 26% to 62%), whereas its specificity was 95% (80% to 100%). The receiver-operator characteristics curve of this parameter is shown in Fig. 2, A. Assuming that 30% of women who undergo immunotherapy have an unsuccessful pregnancy outcome (a prior probability of 30%), the predictive value of a positive test (flow cytometric crossmatch-negative) was calculated to be 79% and the predictive value of a negative test (flow cytometric crossmatch-positive) 80%. We also tested the significance of flow cytometric crossmatch in 46 patients, including 13 women with a history of only two consecutive recurrent spontaneous abortions. The sensitivity was 54%, and specificity was 88% (Fig. 2). If pregnancies that ended in a chromosomally abnormal spontaneous abortion (47,XX,+22; 47,XY,+ 18; and 47 ,XX, + 16) were not included in tabulations of pregnancy outcome, the flow cytometric crossmatch
Maruyama et al.
Volume 168. Number 5 Am .I Obstet Gynecol
1533
No. of spontaneous abortions
2
No.
No. T-FCXMpositive
Total No. T-FCXMnegative
No.
No. T-FCXMpositive
No. T-FCXMnegative
20 It
12 0
8 1
69 5t
52 3
17 2
6 5 4
4 4 1
2 1 3
17 16 13
14 15 6
3 1 7
3 1 1 1 3 11 15 73.3
0 1 0 1 2 8 9 88.9
3 0 1 0 1 3 6 50.0
10 3 2 1 20 33 46 71.7
3 3 0 1 16 29 35 82.9
32.9 :±: 4.6 25-40
31.9 ± 3.7 25-40
45.1 :±: 29.6 0-105
43.3 ± 29.4 2-108
7 0 2 0 4 4 11 36.3
Table III. Summary of parameters according to pregnancy outcome in women with a history of two or more unexplained recurrent spontaneous abortions Pregnancy outcome Successful (n = 33)
Maternal age No. of spontaneous abortions HLA sharing A locus B locus C locus DR locus DQ locus T-FCXM (postimmunization) Negative Positive Interval days
31.6 ± 3.5 3.0 ± 0.9 0.6 0.3 0.2 1.5 0.8
± 0.5 ± 0.5 ± 0.4
± 0.8 ± 0.6
4 29
46.8 ± 29.2
I
Unsuccessful (n = 13)
Significance
32.5 ± 4.0 2.8 ± 0.6
NS NS
± 0.5 ± 0.5
NS NS NS
0.7 0.4 0.4 0.8 0.4
± 0.5
± 0.9 ±
7 6·
0.5
45.2 ± 27.3
P=
0.023
P = 0.023
p=
0.0058
NS
Values are expressed as mean ± SD. NS, Not significant; FCXM, flow cytometric crossmatch. "Fetal chromosomal abnormalities were found in three patients (47,XX, + 22; 47,XY, + 18; and 47,XX, + 16).
result became a stronger prognosticator of pregnancy outcome. The sensitivity of the parameter increased from 44% to 57% in patients with at least three repeated spontaneous abortions; the specificity remained constant. In those patients, including 13 women with an episode of two repeated spontaneous abortion. the'
sensitivity increased from 54% to 70%. The receiveroperator characteristic curves of flow cytometric crossmatch after unsuccessful cases with fetal chromosomal abnormalities were excluded are shown in Fig. 2. B. We cannot ascertain that the cause of spontaneous abortion with fetal chromosomal abnormalities differed from
1534
Maruyama et al.
May 1993
Am
100
(A)
100
.---e SPAB~3
(B)
0 - 0 SPABSi:2
.... '>
>-
80
£
....
60
>-
'>
c:
CD
60 (20)
'in
'in
tJ)
(90) (100)
0 - 0 SPABSi:2
80
£
.---e SPAB~3
J Obstet Gyneco1
c:
40
CD
(20)
tJ)
(30) (40) (50) (60)
(30) (40) (50)
(60)
},,"
40
(10)
(10)
20
(10)
20
o
20
40
60
80
o
100
40
20
60
80
100
I-Specificity (%)
I-Specificity (%)
Fig, 2, Receiver-operator characteristic curves display maximum crossmatch detection rate of 44% and minimum false-positive rate of 5% when flow cytometric crossmatch test was used to predict unsuccessful pregnancy outcome in 31 patients with history of three or more recurrent spontaneous abortions (SPAB) (solid circle). However, maximum sensitivity was 54% and minimum false-positive rate was 12% in 46 patients who had history of two or more recurrent spontaneous abortions (open circle) (A). Receiver-operator characteristic curves formulated after excluding three cases of fetal chromosomal abnormalities are shown in B. Maximum detection rate is increased compared with that evident in previous analysis (A). Cutoff values are indicated in parentheses. Receiver-operator characteristic curve "breaks" when cutoff point of peak channel shift was 20.
Table IV. Summary of parameters according to pregnancy outcome in women with a history of three or more unexplained recurrent spontaneous abortions Pregnancy outcome Successfol (n = 22)
Maternal age No. of spontaneous abortions HLA sharing A locus B locus C locus DR locus DQ locus T-FCXM (postimmunization) Negative Positive Interval days
31.1 ± 2.7 3.5 ± 0.7
0.5 ± 0.5 0.4 ± 0.6 0.3 ± 0.5 1.5 ± 0.8 0.9 ± 0.6 1 21
39.6 ± 20.1
I
Unsuccessfol (n = 9)
Significance
32.0 ± 3.9 3.1 ± 0.3
NS NS
0.8 0.2 0.2 0.8 0.3
± 0.4
± 0.4
± 0.4
P=
NS NS NS
± 0.8 ± 0.5
P = 0.035
4 5*
p
49.4 ± 29.7
=
0.043
0.017
NS
NS, Not significant; FCXM, flow cytometric crossmatch. *Fetal chromosomal abnormalities were found in two patients (47,XX, + 22 and 47,XY, + 18).
that of reproductive immune aberrations because such fetal chromosomal abnormalities may not necessarily be fatal. Thus although the result of the analysis after exclusion of chromosomally abnormal cases was not
statistically significant, it supports an association between pregnancy outcome and the production of maternal anti paternal lymphocyte antibodies. It is unclear why successful patients are more likely to
Volume 168, l\umber 5 Am J Obstet Gynecol
have similar HLA-DR or -DQ loci, One possibility is that couples who are similar in HLA-DR or -DQ respond to immunization treatment particularly because donor-specific blood transfusion (another form of immunotherapy) can induce a marked donor-specific reduction in the frequency of cytotoxic T-lymphocyte precursors provided that the blood donor and the recipient have in common one HLA haplotype or at least one HLA-B and one HLA-DR antigen. 24 The effectiveness of immunotherapy remains controversial. 5 . 7 This controversy may be caused partly by the lack of sensitivity of conventional immunologic assays. Although 52 of 69 patients (75%) are flow cytometric crossmatch - positive after immunotherapy, other reports that failed to demonstrate an association between pregnancy outcome and anti paternal antibody showed a 26% to 63% seroconversion rate."' 10. II In these reports 60% to 75% microlymphocytotoxity crossmatchnegative couples succeed in a subsequent pregnancy, although in our analysis 25% of flow cytometric crossmatch-negative couples succeed. Some microlymphocytotoxicity crossmatch-negative patients who ultimately have a successful pregnancy outcome may already be sensitizied to the surface antigen of paternal lymphocytes, which can be detected only by flow cytometric crossmatch. We cannot ascertain that antilymphocyte antibodies, including blocking antibodies, are essentially required to maintain pregnancy in those who experience recurrent spontaneous abortion. In the process of producing such antibodies, to maintain pregnancy maternal immunity may be altered followed by reduction of cytotoxic T cells,2' increase of suppressor T cells,"5 generation of antiidiotypic antibodies against T-cell receptorrecognizing surface antigen of paternal leukocytes,25 or reduction of cytotoxic T-lymphocyte precursors. 24 In summary, our study indicates that a positive twocolor T-cell flow cytometric crossmatch predicts successful subsequent pregnancy outcome in patients undergoing immunotherapy for recurrent spontaneous abortion. Also, flow cytometric crossmatch IS more quantitative and methodologically easier in the clinical setting than either mixed lymphocyte culture reactionblocking assay or microlymphocytotoxicity crossmatch, and T flow cytometric crossmatch in particular is a useful test for managing recurrent spontaneous abortion. We thank Akira Sonoda, Department of Clinical Laboratory, Keio University Hospital, for the helpful suggestions and Hiroko Otsuka for assistance throughout this study. REFERENCES 1. Taylor C, Faulk WP. Prevention of recurrent abortion with leucocyte transfusions. Lancet 1981 ;2:68-70.
Maruyama et al.
1535
2. Beer AE, Quebbeman ]F, Ayers ]wr, Haines RF. Major histocompatibility complex antigens, maternal and paternal immune responses, and chronic habitual abortions in humans. AM] Ossn:T GYNECOl. 1981;141:987-99. 3. Makino T, Sakai A, Sugi T, et al. Current comprehensive therapy of habitual abortion. Ann N Y Acad Sci 1991 ;626: 597-604. 4. Clark DA, Daya S. Trials and tribulation in the treatment of recurrent spontaneous abortion. Am] Reprod Immunol 1991 ;25: 18-24. 5. Mowbray ]F, Gibbings C, Liddell H, Reginald PW, Underwood ]L, Beard RW. Controlled trial of treatment of recurrent spontaneous abortion by immunisation with paternal cells. Lancet 1985;1:941-3. 6. Ho HN, Gill T] III, Hsieh H-], Jiang ]-], Lee T-Y, Hsieh C-Y. Immunotherapy for recurrent spontaneous abortions in a Chinese population. Am] Reprod Immunol 1991 ;25: 10-5. 7. Cauchi MN, Lim D, Young DE, Kloss M, Pepperell RJ. Treatment of recurrent aborters by immunization with paternal cells-controlled trial. Am ] Reprod Immunol 1991;25:16-7. 8. Coulam CB, Clark DA. Report from the ethics committee for immunotherapy, American Society for the Immunology of Reproduction. Am ] Reprod Immunol 1991;26: 93-5. 9. Gilman-Sachs A, Luo S-P, Beer AE, Beaman KD. Analysis of anti-lymphocyte antibodies by flow cytometry or microIymphocytotoxicity in women with recurrent spontaneous abortions immunized with paternal leukocytes.] Clin Lab Immunol 1989;30:53-9. 10. Smith ]B, Cowchock FS. Immunological studies in recurrent spontaneous abortion: effects of immunization of women with paternal mononuclear cells on Iymphocytotoxic and mixed lymphocyte reaction blocking antibodies and correlation with sharing of HLA and pregnancy outcome.] Reprod Immunol 1988; 14:99-113. II. Gatenby PA, Moore H, Cameron K, Doran 'IJ, Adelstein S. Treatment of recurrent spontaneous abortion by immunization with paternal lymphocytes: correlates with outcome. Am] Reprod Immunol 1989;19:21-7. 12. Cowchock FS, Smith]B, David S, Scher], Batzer F, Corson S. Paternal mononuclear cell immunization therapy for repeated miscarriage: predictive variables for pregnancy success. Am] Reprod Immunol 1990;22:12-7. 13. Garovoy MR, Rheinschmidt MA, Bigos M, et al. Flow cytometry analysis: a high technology crossmatch technique facilitating transplantation. Transplant Proc 1983; 15:1939-44. 14. Lazda VA, Pollak R, Mozes MF,]onasson O. The relationship between flow cytometer cross match results and subsequent rejection episodes in cadaver renal allograft recipients. Transplantation 1988;45:562-5. 15. Iwaki Y, Cook D], Terasaki PI, et al. Flow cytometry crossmatching in human cadaver kidney transplantation. Transplant Proc 1987; 19:764-6. 16. Mahoney RJ, Ault KA, Given SR, et al. The flow cytometric crossmatch and early renal transplant loss. Transplantation 1990;49:527-35. 17. Talbot D, Givan AL, Shenton BK, et al. The prospective value of the preoperative flow cytometric crossmatch assay in renal transplantation. Transplantation 1990;49:809-10. 18. Bou-Habib ]C, Krams S, Colombe BW, et al. Flow cytometric crossmatching and long-term kidney allograft survival in donor-specific transfusion patients. Transplantation 1991 ;51 :543-5. 19. Gilman-Sachs A, Harris D, Beer A, Beaman KD. Inhibition of binding of anti-CD3 antibodies to paternal lymphocytes correlates with failure of immunotherapy for treatment of recurrent spontaneous abortions. ] Reprod Immunol 1990;17:41-51. 20. Carp H.JA, Toder V, Gazit E, et al. Immunization by paternal leukocytes for prevention of primary habitual
1536
Maruyama et al.
abortion: results of a matched controlled trial. Gynecol Obstet Invest 1990;29: 16-2l. 21. Reznikoff-Etievant MF, Durieux I, Huchet j, Salmon C, Netter A. Paternal leucocyte injections in recurrent spontaneous abortion. Lancet 1987;2:1460. 22. Takakuwa K, Goto S, Hasegawa I, et al. Result of immunotherapy on patients with unexplained recurrent abortion: a beneficial treatment for patients with negative blocking antibodies. Am j Reprod Immunol 1990;23:3741. 23. Browner WS, Newman TB, Cummings SR. Designing a
May 1993 Am J Obstet Gynecol
new study: III. Diagnostic tests. In: Hulley SB, Cummings SR, eds. Designing clinical research. Baltimore: Williams & Wilkins, 1988:87-97. 24. van Twuyver E, Mooijaart RJD, ten Berge IJM, et al. Prdransplanlalion blood transfusion revisiled. N Engl J Med 1991;325:1210-3. 25. Sugi T, Makino T, Maruyama T, Kim WK, Iizuka R. A possible mechanism of immunotherapy for patients with recurrent spontaneous abortions. Am J Reprod Immunol 1991 ;25: 185-9.
Bound volumes available to subscribers Bound volumes of the AMERICAN JOURNAL OF OBSTETRICS Ar\D GVr\ECOLOGY are available to subscribers (only) for the 1993 issues from the Publisher, at a cost of $74.50 for domestic, $102.72 for Canada, and $97.50 for international for Vol. 168 (January-June) and Vol. 169 (July-December). Shipping charges are included. Each bound volume contains a subject and author index and all advertising is removed. Copies are shipped within 60 days after publication of the last issue in the volume. The binding is durable buckram with the JOURr\AL name, volume number, and year stamped in gold on the spine. Payment must accompany all orders. Contact Mosby, Subscription Services, 11830 Westline Industrial Drive, St. Louis, MO 63146-3318, USA; phone (800) 453-4351 or (314) 453-4351. Subscriptions must be in force to qualify. Bound volumes are not available in place of a regular JOURNAL subscription.
Tokyo, Japan OBJECTIVE: The purpose of our study was to evaluate the clinical significance of flow cytometric crossmatch in managing women with unexplained recurrent spontaneous abortions who have been immunized with paternal leukocytes. STUDY DESIGN: Two-color flow cytometric crossmatch was performed after immunotherapy in 69 women with recurrent spontaneous abortions to detect maternal immunoglobulin G antibodies against paternal T lymphocytes. A positive flow cytometric cross match was determined by a ~ 20 channel increases (256 channel linear scale-log amplification) in the fluorescence intensity of the T-cell peak. We analyzed the clinical correlation between flow cytometric crossmatch results and subsequent pregnancy outcome in those patients using the Fisher exact test. RESULTS: In 46 subsequent pregnancies among 69 immunized women the incidence of early subsequent pregnancy wastage was only 6 (17.1 %) of 35 in flow cytometric crossmatch-positive patients and 7 (63.6%) of 11 in flow cytometric crossmatch-negative patients I.p < 0.01). CONCLUSIONS: These data indicate that a two-color T cell flow cytometric crossmatch predicts subsequent pregnancy outcome in patients undergoing immunotherapy for recurrent spontaneous abortion. (AM J QBSTET GVNECOL 1993;168:1528-36.)
Key words: Habitual abortion, flow cytometric crossmatch, immunotherapy
Since leukocyte-enriched plasma transfusions or subcutaneous leukocyte injections were first used to treat women with recurrent spontaneous abortions in 1981," 2 immunotherapy has become generally accepted as an efficacious treatment for unexplained repeated spontaneous abortions. Most trials of such immunotherapy have shown about a 70% success rate." 4 In the first randomized, controlled trial Mowbray et al. 5 reported a 77% success rate in patients who received immunotherapy with their spouses' leukocytes, whereas only 37% of controls immunized with their own leukocytes were delivered successfully: By contrast, Ho et al." and Cauchi et al.? recently reported two randomized controlled trials with the husband or third-party From the Department of Obstetrics and Gynecology, School of Medicine, Keio UniverSity. Supported in part by a research grant from Keio Gijuku Academic Development Funds (1990). Received for publication June 4, 1992; revised August 20, 1992; accepted October 20, 1992. Reprint requests: Tetsuo Maruyama, MD, Department of Obstetrics and Gynecology, School of Medicine, Keio University, 35 Shinanomachi, Shinjuku-ku, Tokyo 160, Japan. Copyright © 1993 by Mosby-Year Book, Inc. 0002-9378193 $1.00 + .20 6/1/43617
1528
blood lymphocyte immunotherapy; these trials failed to show any significant benefit. Because of these conflicting studies immunotherapy has generated much controversy. Clark and Daya4 and Coulam and Clark" pointed out several controversial issues in immunotherapy: sample sizes of such trials have been small, the populations eligible for the immunotherapy have been heterogeneous, and there have been some cointerventions by the control treatment that may affect success rates in the control group. In addition to these problems standard immunologic assays such as mixed lymphocyte culture reaction-blocking assay or microlymphocytotoxicity crossmatch, which are usually used to select populations for immunotherapy and to demonstrate sensitization to donor leukocytes, might be inappropriate for immunologic monitoring because they may be insensitive techniques. 9 Criteria based on conventional immunologic examinations may fail to select sensitized patients and, thereby, may select immunologically heterogeneous populations. Furthermore, the prognostic value of assays for blocking antibodies and cytotoxic anti paternal antibodies is limited because of these problems. 5 • 10-12 In renal transplantation flow cytometric crossmatching has proved to be a very sensitive method for
\'uillme 168. l'\lImber " Am J Obslel Gynecol
detecting T -cell- and B-cell-directed antibodies and has identified a subgroup of patients who, although negative by conventional microlymphocytotoxicity crossmatch, not only possess immunoglobulin (Ig) G antibodies directed against donor lymphocytes but are also more likely to reject their grafts. 13-18 Our study was designed to evaluate the clinical significance of flow cytometric crossmatch in managing women with unexplained recurrent spontaneous abortions who have been immunized with paternal leukocytes. Material and methods Patients. Between June 1990 and March 1992, 69 women with a history of two or more unexplained, consecutive, first-trimester spontaneous abortions with the same partner were included in this study. The couples were thoroughly investigated for possible causes of spontaneous abortion, including the presence of a genetic impairment, infectious disease, hormonal deficiency, metabolic disorder, uterine anomaly, and any autoimmune abnormality such as anticardiolipin antibody. There was no family history of recurrent loss and congenital anomalies in any couple. Mter those causes were ruled out, and the patients were thoroughly counseled about the risk of white blood cell administration, immunotherapy was performed with the spouse's leukocytes. All donors were screened and found to be negative for hepatitis, syphilis, and antibodies to HIV and HTLV-l. Immunization procedures. Mononuclear cells (usually 4 x 10 7 ) were prepared from 30 ml of heparinized whole blood taken from the husband, separated by density centrifugation, concentrated to 2 ml under sterile conditions, and injected subcutaneously into the forearms of the patient at six to eight sites. This procedure was repeated four times at intervals of 4 to 6 weeks before each couple was permitted to attempt pregnancy. Flow cytometric crossmatch was performed for each couple 2 to 4 weeks after the last immunization. Flow cytometric crossmatch. Flow cytometric crossmatch was performed with a modification of the technique proposed by Gilman-Sachs et al."· 19 in which maternal heat-inactivated serum and paternal lymphocytes were used to detect maternal anti paternal lymphocyte antibodies. Maternal sera were collected before treatment at each immunization and at 2 to 4 weeks after the last immunization, pooled, and then frozen until flow cytometric crossmatch was performed. For each husband, lymphocytes were separated from heparinized whole peripheral blood by Ficoll-Conray (IBL, Gunma, Japan) density gradient centrifugation. Five x 105 paternal lymphocytes were incubated with 20 IJ..i of each maternal serum and 100 ILl of phosphate
Maruyama et at.
1529
buffered saline solution for 15 minutes at room temperature. As a negative control, type AB serum (lCN Flow, Costa Mesa, Calif.) was used instead of maternal serum. The cells were washed twice at 4° C by centrifugation. Optimum amounts of F(ab)'2-fluorescein isothiocyanate-conjugated goat antihuman IgG (Cappel, West Chester, Pa.) and phycoerythrin-conjugated antihuman CD3 or CD 19 antibody (Becton Dickinson, Mountain View, Calif.) were then added to each tube to detect maternal antibody bound to the paternal T- or B-lymphocyte surface. The amount of antibody used in the assay was predetermined from titration curves. Samples were incubated for 15 minutes in the dark at 4° C. Mter two washes the cells were suspended in 300 ILl of phosphate buffered saline solution. The labeled samples (in phosphate buffered saline solution) were analyzed with a FACScan flow cytometer (Becton Dickinson). Data from 10,000 cells were collected in list mode files by using logarithmic amplification of fluorescence parameters. Data acquisition and analysis was performed with Consort 30 software (Becton Dickinson). The FACScan uses a four-decade log amplifier. Flow cytometric crossmatch analysis. For the purpose of our study a positive flow cytometric crossmatch was that in which a ~ 20 channel shift to the right of the T-cell peak ~ 14 channel shift of the B-cell peak in the test serum or compared with the control serum on a 256-channel linear scale. This cutoff point was chosen on the basis of the mean + SD of fluorescence intensity increase of the cell peak treated with the preimmunization serum compared with the control serum. These cutoff points are summarized in Table I. In this analysis the log scale channel was converted to a linear channel by the following formula: Linear channel = Channels-per-decade x Log (log channel) The channels-per-decade for 256-channel resolution data is 63.75 in the FACScan and the Consort 30 software system. A shift is the difference in flow cytometric crossmatch parameter between the test serum and the control serum. Mean channel reflects the average channel value of the horizontal parameter for the processed cells. Peak channel reflects the horizontal channel value of the point with the highest cell count. The percentage of positive lymphocytes reflects the proportion of cells that bound antibodies (Table I). Our study was performed in a double-blind fashion because the patients were not informed of their own flow cytometric crossmatch results unless their subsequent pregnancies ended in spontaneous abortions and because we did not know the subsequent pregnancy outcome of each patient when analyzing their data. Histocompatibility testing. HLA-A, -B, -C, -DR, and
1530 Maruyama et al.
May 1993 Am J Obstet Gynecol
Table I. Summary of flow cytometric crossmatch parameters in women immunized with paternal leukocytes according to choice of cutoff point for each parameter T flow cytometric crossmatch * Mean ± SD
Mean channel shift Control vs preimmunization Control vs postimmunization Peak channel shift Control vs preimmunization Control vs postimmunization Shift of percentage of positive lymphocytes Control vs preimmunization Control vs postimmunization
I
B flow cytometric crossmatcht
I
Cutoff pointt
Mean ± SD
Cutoff point
5 ± 18 38 ± 29§
23
-1 ± 14 II ± 2311
13
5 ± 15 45 ± 27~
20
-5 ± 19 16 ± 29#
14
2 ± 10 43 ± 39**
12
1± 4 5 ± 10tt
5
Values were obtained for 256-channel, 2-parameter histogram. *55 Of 69 women immunized with patemalleukocytes were tested by T flow cytometric crossmatch before immunization. t48 of 69 women were tested by B flow cytometric crossmatch after immunization; 42 of 48 women were examined by B flow cytometric crossmatch before immunization. tChoice of cutoff point was based on mean + SO of each value ofT or B flow cytometric crossmatch (control vs preimmunization). §p < 0.0001, compared with mean channel shift of control T cells vs preimmunization. lIP < 0.001, 'compared with mean channel shift of control B cells vs preimmunization. ~p < 0.0001, compared with peak channel shift of control T cells vs preimmunization. #p < 0.0001, compared with peak channel shift of control B cells vs preimmunization. **p < 0.0001, compared with shift of percent of positive control T cells vs preimmunization. ttp < 0.005, compared with shift of percent of positive control B cells vs preimmunization.
-DQ typing was performed on 69 couples with the National Institutes of Health standard microlymphocytotoxicity test at the Clinical Laboratory Department of Keio University Hospital and a commercial laboratory (SRL, Hachioji, Tokyo). Statistical analysis. The Fisher exact test or the X2 test for independence was used to formulate two-by-two tables and evaluate the association between flow cytometric crossmatch results and subsequent pregnancy outcome. The Wilcoxon rank sum test or the impaired t test was used selectively to analyze differences in several parameters between the successful subsequent pregnancy outcome group and the unsuccessful group. The paired t test was used to analyze any change in the flow cytometric crossmatch value between the preimmunization and postimmunization period. A probability of p < 0.05 was chosen to represent statistical significance. Results
Of 69 patients who underwent immunotherapy, 49 became pregnant as this manuscript was being prepared. Patient characteristics are summarized in Table II. Included in the calculation of the pregnancy success rate were live births and pregnancies of > 24 gestational weeks. There were no second- or third-trimester losses. The flow cytometric crossmatch values and cutoff
points in women immunized with paternal leukocytes are summarized in Table I. After immunotherapy the production of maternal antibodies against paternal T or B lymphocytes increased gradually in most immunized patients (Fig. 1). Each postimmunization T- and Bflow cytometric crossmatch value increased significantly compared with the corresponding preimmunization flow cytometric crossmatch (Table I). Of 69 patients after immunization, 52 (75%) had a positive T flow cytometric crossmatch and 9 of 55 patients (16%) who were tested by T flow cytometric crossmatch before immunization were flow cytometric crossmatch-positive. Of 48 patients after immunization, 21 (44%) had a positive B flow cytometric crossmatch, and 5 of 42 patients (12%) had a positive B flow cytometric crossmatch before immunization. In 46 women with a subsequent pregnancy among 69 immunized patients, the incidence of early pregnancy loss was only 17.1% (6 of 35) in flow cytometric crossmatch-positive patients and 63.6% (7 of 11) in flow cytometric crossmatch-negative patients (P < 0.01) (Table III). Furthermore, among 31 of 46 pregnant patients with a history of three or more spontaneous abortions, the incidence of unsuccessful pregnancy was only 19.2% (5 of 26) in flow cytometric crossmatchpositive patients and 80% (4 of 5) in flow cytometric crossmatch-negative patients (p < 0.05) (Table IV). No correlation existed between subsequent pregnancy out-
Volume 168, Number 5 Am J Obstet Gynecol
Maruyama et al.
(A) AB SERUM
1531
(B) PRE-IMZ MEAN=71
MEAN=95 (SHIFT=24)
(0) POST-IMZ(2)
PEAK=93
MEAN=72
MEAN=115 (SHIFT=44) PEAK=109
(E) POST-IMZ(4)
MEAN=126 (SHIFT=55) PEAK=113
Fig. 1. Three-dimensional contour histograms illustrating green fluorescence intensity on horizontal axis (caused by labeling with FITC-conjugated anti-IgG to paternal T-cell population), versus red fluorescence intensity on vertical axis (caused by labeling with phycoerythrin-conjugated anti-CD3 antibody), Each axis is four-decade logarithmic scale. Gate was set to encompass T-lymphocyte population. Number of cells is proportional to height of signal. A, Mean green fluorescence and peak channels for paternal T cells incubated with type AB serum were 71 and 53, respectively. B, Mean green fluorescence channel for paternal T cells incubated with maternal preimmunization (IMZ) serum was 72 (shift of one channel); peak channel was 57 (shift of 4 channels). C, D, and E, Postimmunization serum obtained at 49, 80, or 153 days, respectively, after first immunization was used. Blood samples were collected simultaneously with immunotherapy. Panel representing use of serum obtained 140 days after first immunization is not shown. Mean and peak channel numbers are indicated in each panel. Mean channel shift and peak channel shift are indicated in parentheses in each panel.
come and the B-flow cytometric crossmatch result regardless of the cutoff point chosen. The successful and unsuccessful outcome groups did not differ significantly in age, number of spontaneous abortions, number of shared HLA-A, -B, or -C loci, or interval between the last immunization and subsequent conception. However, HLA-DR and -DQ loci were more likely to be common to both partners in the successful group than in the unsuccessful group (Tables III and IV). Comment
In renal transplantation, flow cytometry has recently come into use to supplement standard crossmatch methods to detect recipient pre sensitization to donor tissues. A positive flow cytometric crossmatch is associated with a greater incidence of rejection episodes, even when a conventional standard cross match IS negative. 13-18 In managing patients with unexplained recurrent
spontaneous abortions, mixed lymphocyte culture reaction-blocking assay and microlymphocytotoxicity crossmatch have been applied for immunologic assessment and monitoring. However, the association between these conventional assays and pregnancy outcome remains controversial. 6 • ID-I2. 20-22 Gilman-Sachs et al. 9 reported that the detection of antilymphocyte antibodies by flow cytometric crossmatch is dearly more sensitive than the standard microlymphocytotoxicity crossmatch, because twice as many positive couples were detected by flow cytometric crossmatch than by microlymphocytotoxicity crossmatch (66% vs 33%). Two-color flow cytometric crossmatch demonstrated a correlation between inhibition of binding of anti-CD3 antibodies to paternal lymphocytes and failure of immunotherapy to treat recurrent spontaneous abortions. I9 In kidney transplantation the degree of sensitization to donor tissues reflected by the shift of corresponding
1532 Maruyama et al.
May 1993
Am
J Obstet Gynecol
Table II. Characteristics of patients classified according to number of spontaneous abortions No. of spontaneous abortions
5
No.
Patients No. of patients with live births Pregnancy outcome Infants delivered Pregnant beyond 24 wk Aborted Fetal chromosomal analysis Normal Abnormal Ectopic Molar Awaiting conception No. of successful outcomes No. of total pregnant patients§ Success rate (%) Maternal age§ Mean ± SD Range Interval daysll Mean ± SD Range
3 0
No. T-FCXMpositive
4 No. T-FCXMnegative
3
1 1 0 0 0 0 0 1 2 2 100
1
2 2
100
No.
No. T-FCXMpositive
o
13 3*
12 2
o o
2 5 1
2 5 1
o o o
31.5 ± 3.5 28-35 11 ± 9 2-20
1 0 0 0 5 7 8 87.5
3 No. T-FCXMnegative
4 7 8 87.5
No.
No. T-FCXMpositive
No. T-FCXMnegative
33
It
25 1
8 0
0 0 0
8 5 8
7 5 4
1 0 4
0
6 2 1 0 11 13 21 61.9
2 2 0
4 0 1
9 12 16 75.0
2 1 5 20.0
1 0 0
31.3 ± 2.9 26-35
31.4 ± 3.1 27-38
44.5 ± 26.0 1-83
44.7 ± 30.1 0-108
FCXM, Flow cytometric crossmatch.
*None of the three patients conceived after immunization. tShe conceived after immunization. tEach patient had one child before immunization. §Molar or ectopic pregnancy outcome and patients awaiting conception were excluded. IIInterval days between the last immunization and subsequent conception. Conception date was determined to be seventh day of high phase of basal body temperature.
cells incubated with recipient serum can reportedly be quantitated by various parameters of the flow cytometric crossmatch such as mean channel shift, 16 peak channel shift, [3. 14. 18 percent relative mean channel number shift, 15 and the percentage of positive cells. l4 • 17 We could not find correlation between subsequent pregnancy outcome and flow cytometric crossmatch results in the shift of percentage of positive CD3 cells or mean channel shift, regardless of the cutoff point chosen. One other method by which the cutoff point may be established is from the receiver-operator characteristic curve."3 Usually the best cutoff point is where the receiver-operator characteristic curve "turns the corner."23 Furthermore, the better a parameter the closer its curve is to the upper left corner."3 We compared the characteristics of curves representing different parameters and then used the peak channel shift on flow cytometric crossmatch of T cells as a significant prognosticator of subsequent pregnancy outcome and an indicator of maternal antibody against paternal lym-
phocytes. In evaluating the significance of flow cytometric crossmatch in patients who had three or more spontaneous abortions, the sensitivity of this test was 44% (95% confidence interval; 26% to 62%), whereas its specificity was 95% (80% to 100%). The receiver-operator characteristics curve of this parameter is shown in Fig. 2, A. Assuming that 30% of women who undergo immunotherapy have an unsuccessful pregnancy outcome (a prior probability of 30%), the predictive value of a positive test (flow cytometric crossmatch-negative) was calculated to be 79% and the predictive value of a negative test (flow cytometric crossmatch-positive) 80%. We also tested the significance of flow cytometric crossmatch in 46 patients, including 13 women with a history of only two consecutive recurrent spontaneous abortions. The sensitivity was 54%, and specificity was 88% (Fig. 2). If pregnancies that ended in a chromosomally abnormal spontaneous abortion (47,XX,+22; 47,XY,+ 18; and 47 ,XX, + 16) were not included in tabulations of pregnancy outcome, the flow cytometric crossmatch
Maruyama et al.
Volume 168. Number 5 Am .I Obstet Gynecol
1533
No. of spontaneous abortions
2
No.
No. T-FCXMpositive
Total No. T-FCXMnegative
No.
No. T-FCXMpositive
No. T-FCXMnegative
20 It
12 0
8 1
69 5t
52 3
17 2
6 5 4
4 4 1
2 1 3
17 16 13
14 15 6
3 1 7
3 1 1 1 3 11 15 73.3
0 1 0 1 2 8 9 88.9
3 0 1 0 1 3 6 50.0
10 3 2 1 20 33 46 71.7
3 3 0 1 16 29 35 82.9
32.9 :±: 4.6 25-40
31.9 ± 3.7 25-40
45.1 :±: 29.6 0-105
43.3 ± 29.4 2-108
7 0 2 0 4 4 11 36.3
Table III. Summary of parameters according to pregnancy outcome in women with a history of two or more unexplained recurrent spontaneous abortions Pregnancy outcome Successful (n = 33)
Maternal age No. of spontaneous abortions HLA sharing A locus B locus C locus DR locus DQ locus T-FCXM (postimmunization) Negative Positive Interval days
31.6 ± 3.5 3.0 ± 0.9 0.6 0.3 0.2 1.5 0.8
± 0.5 ± 0.5 ± 0.4
± 0.8 ± 0.6
4 29
46.8 ± 29.2
I
Unsuccessful (n = 13)
Significance
32.5 ± 4.0 2.8 ± 0.6
NS NS
± 0.5 ± 0.5
NS NS NS
0.7 0.4 0.4 0.8 0.4
± 0.5
± 0.9 ±
7 6·
0.5
45.2 ± 27.3
P=
0.023
P = 0.023
p=
0.0058
NS
Values are expressed as mean ± SD. NS, Not significant; FCXM, flow cytometric crossmatch. "Fetal chromosomal abnormalities were found in three patients (47,XX, + 22; 47,XY, + 18; and 47,XX, + 16).
result became a stronger prognosticator of pregnancy outcome. The sensitivity of the parameter increased from 44% to 57% in patients with at least three repeated spontaneous abortions; the specificity remained constant. In those patients, including 13 women with an episode of two repeated spontaneous abortion. the'
sensitivity increased from 54% to 70%. The receiveroperator characteristic curves of flow cytometric crossmatch after unsuccessful cases with fetal chromosomal abnormalities were excluded are shown in Fig. 2. B. We cannot ascertain that the cause of spontaneous abortion with fetal chromosomal abnormalities differed from
1534
Maruyama et al.
May 1993
Am
100
(A)
100
.---e SPAB~3
(B)
0 - 0 SPABSi:2
.... '>
>-
80
£
....
60
>-
'>
c:
CD
60 (20)
'in
'in
tJ)
(90) (100)
0 - 0 SPABSi:2
80
£
.---e SPAB~3
J Obstet Gyneco1
c:
40
CD
(20)
tJ)
(30) (40) (50) (60)
(30) (40) (50)
(60)
},,"
40
(10)
(10)
20
(10)
20
o
20
40
60
80
o
100
40
20
60
80
100
I-Specificity (%)
I-Specificity (%)
Fig, 2, Receiver-operator characteristic curves display maximum crossmatch detection rate of 44% and minimum false-positive rate of 5% when flow cytometric crossmatch test was used to predict unsuccessful pregnancy outcome in 31 patients with history of three or more recurrent spontaneous abortions (SPAB) (solid circle). However, maximum sensitivity was 54% and minimum false-positive rate was 12% in 46 patients who had history of two or more recurrent spontaneous abortions (open circle) (A). Receiver-operator characteristic curves formulated after excluding three cases of fetal chromosomal abnormalities are shown in B. Maximum detection rate is increased compared with that evident in previous analysis (A). Cutoff values are indicated in parentheses. Receiver-operator characteristic curve "breaks" when cutoff point of peak channel shift was 20.
Table IV. Summary of parameters according to pregnancy outcome in women with a history of three or more unexplained recurrent spontaneous abortions Pregnancy outcome Successfol (n = 22)
Maternal age No. of spontaneous abortions HLA sharing A locus B locus C locus DR locus DQ locus T-FCXM (postimmunization) Negative Positive Interval days
31.1 ± 2.7 3.5 ± 0.7
0.5 ± 0.5 0.4 ± 0.6 0.3 ± 0.5 1.5 ± 0.8 0.9 ± 0.6 1 21
39.6 ± 20.1
I
Unsuccessfol (n = 9)
Significance
32.0 ± 3.9 3.1 ± 0.3
NS NS
0.8 0.2 0.2 0.8 0.3
± 0.4
± 0.4
± 0.4
P=
NS NS NS
± 0.8 ± 0.5
P = 0.035
4 5*
p
49.4 ± 29.7
=
0.043
0.017
NS
NS, Not significant; FCXM, flow cytometric crossmatch. *Fetal chromosomal abnormalities were found in two patients (47,XX, + 22 and 47,XY, + 18).
that of reproductive immune aberrations because such fetal chromosomal abnormalities may not necessarily be fatal. Thus although the result of the analysis after exclusion of chromosomally abnormal cases was not
statistically significant, it supports an association between pregnancy outcome and the production of maternal anti paternal lymphocyte antibodies. It is unclear why successful patients are more likely to
Volume 168, l\umber 5 Am J Obstet Gynecol
have similar HLA-DR or -DQ loci, One possibility is that couples who are similar in HLA-DR or -DQ respond to immunization treatment particularly because donor-specific blood transfusion (another form of immunotherapy) can induce a marked donor-specific reduction in the frequency of cytotoxic T-lymphocyte precursors provided that the blood donor and the recipient have in common one HLA haplotype or at least one HLA-B and one HLA-DR antigen. 24 The effectiveness of immunotherapy remains controversial. 5 . 7 This controversy may be caused partly by the lack of sensitivity of conventional immunologic assays. Although 52 of 69 patients (75%) are flow cytometric crossmatch - positive after immunotherapy, other reports that failed to demonstrate an association between pregnancy outcome and anti paternal antibody showed a 26% to 63% seroconversion rate."' 10. II In these reports 60% to 75% microlymphocytotoxity crossmatchnegative couples succeed in a subsequent pregnancy, although in our analysis 25% of flow cytometric crossmatch-negative couples succeed. Some microlymphocytotoxicity crossmatch-negative patients who ultimately have a successful pregnancy outcome may already be sensitizied to the surface antigen of paternal lymphocytes, which can be detected only by flow cytometric crossmatch. We cannot ascertain that antilymphocyte antibodies, including blocking antibodies, are essentially required to maintain pregnancy in those who experience recurrent spontaneous abortion. In the process of producing such antibodies, to maintain pregnancy maternal immunity may be altered followed by reduction of cytotoxic T cells,2' increase of suppressor T cells,"5 generation of antiidiotypic antibodies against T-cell receptorrecognizing surface antigen of paternal leukocytes,25 or reduction of cytotoxic T-lymphocyte precursors. 24 In summary, our study indicates that a positive twocolor T-cell flow cytometric crossmatch predicts successful subsequent pregnancy outcome in patients undergoing immunotherapy for recurrent spontaneous abortion. Also, flow cytometric crossmatch IS more quantitative and methodologically easier in the clinical setting than either mixed lymphocyte culture reactionblocking assay or microlymphocytotoxicity crossmatch, and T flow cytometric crossmatch in particular is a useful test for managing recurrent spontaneous abortion. We thank Akira Sonoda, Department of Clinical Laboratory, Keio University Hospital, for the helpful suggestions and Hiroko Otsuka for assistance throughout this study. REFERENCES 1. Taylor C, Faulk WP. Prevention of recurrent abortion with leucocyte transfusions. Lancet 1981 ;2:68-70.
Maruyama et al.
1535
2. Beer AE, Quebbeman ]F, Ayers ]wr, Haines RF. Major histocompatibility complex antigens, maternal and paternal immune responses, and chronic habitual abortions in humans. AM] Ossn:T GYNECOl. 1981;141:987-99. 3. Makino T, Sakai A, Sugi T, et al. Current comprehensive therapy of habitual abortion. Ann N Y Acad Sci 1991 ;626: 597-604. 4. Clark DA, Daya S. Trials and tribulation in the treatment of recurrent spontaneous abortion. Am] Reprod Immunol 1991 ;25: 18-24. 5. Mowbray ]F, Gibbings C, Liddell H, Reginald PW, Underwood ]L, Beard RW. Controlled trial of treatment of recurrent spontaneous abortion by immunisation with paternal cells. Lancet 1985;1:941-3. 6. Ho HN, Gill T] III, Hsieh H-], Jiang ]-], Lee T-Y, Hsieh C-Y. Immunotherapy for recurrent spontaneous abortions in a Chinese population. Am] Reprod Immunol 1991 ;25: 10-5. 7. Cauchi MN, Lim D, Young DE, Kloss M, Pepperell RJ. Treatment of recurrent aborters by immunization with paternal cells-controlled trial. Am ] Reprod Immunol 1991;25:16-7. 8. Coulam CB, Clark DA. Report from the ethics committee for immunotherapy, American Society for the Immunology of Reproduction. Am ] Reprod Immunol 1991;26: 93-5. 9. Gilman-Sachs A, Luo S-P, Beer AE, Beaman KD. Analysis of anti-lymphocyte antibodies by flow cytometry or microIymphocytotoxicity in women with recurrent spontaneous abortions immunized with paternal leukocytes.] Clin Lab Immunol 1989;30:53-9. 10. Smith ]B, Cowchock FS. Immunological studies in recurrent spontaneous abortion: effects of immunization of women with paternal mononuclear cells on Iymphocytotoxic and mixed lymphocyte reaction blocking antibodies and correlation with sharing of HLA and pregnancy outcome.] Reprod Immunol 1988; 14:99-113. II. Gatenby PA, Moore H, Cameron K, Doran 'IJ, Adelstein S. Treatment of recurrent spontaneous abortion by immunization with paternal lymphocytes: correlates with outcome. Am] Reprod Immunol 1989;19:21-7. 12. Cowchock FS, Smith]B, David S, Scher], Batzer F, Corson S. Paternal mononuclear cell immunization therapy for repeated miscarriage: predictive variables for pregnancy success. Am] Reprod Immunol 1990;22:12-7. 13. Garovoy MR, Rheinschmidt MA, Bigos M, et al. Flow cytometry analysis: a high technology crossmatch technique facilitating transplantation. Transplant Proc 1983; 15:1939-44. 14. Lazda VA, Pollak R, Mozes MF,]onasson O. The relationship between flow cytometer cross match results and subsequent rejection episodes in cadaver renal allograft recipients. Transplantation 1988;45:562-5. 15. Iwaki Y, Cook D], Terasaki PI, et al. Flow cytometry crossmatching in human cadaver kidney transplantation. Transplant Proc 1987; 19:764-6. 16. Mahoney RJ, Ault KA, Given SR, et al. The flow cytometric crossmatch and early renal transplant loss. Transplantation 1990;49:527-35. 17. Talbot D, Givan AL, Shenton BK, et al. The prospective value of the preoperative flow cytometric crossmatch assay in renal transplantation. Transplantation 1990;49:809-10. 18. Bou-Habib ]C, Krams S, Colombe BW, et al. Flow cytometric crossmatching and long-term kidney allograft survival in donor-specific transfusion patients. Transplantation 1991 ;51 :543-5. 19. Gilman-Sachs A, Harris D, Beer A, Beaman KD. Inhibition of binding of anti-CD3 antibodies to paternal lymphocytes correlates with failure of immunotherapy for treatment of recurrent spontaneous abortions. ] Reprod Immunol 1990;17:41-51. 20. Carp H.JA, Toder V, Gazit E, et al. Immunization by paternal leukocytes for prevention of primary habitual
1536
Maruyama et al.
abortion: results of a matched controlled trial. Gynecol Obstet Invest 1990;29: 16-2l. 21. Reznikoff-Etievant MF, Durieux I, Huchet j, Salmon C, Netter A. Paternal leucocyte injections in recurrent spontaneous abortion. Lancet 1987;2:1460. 22. Takakuwa K, Goto S, Hasegawa I, et al. Result of immunotherapy on patients with unexplained recurrent abortion: a beneficial treatment for patients with negative blocking antibodies. Am j Reprod Immunol 1990;23:3741. 23. Browner WS, Newman TB, Cummings SR. Designing a
May 1993 Am J Obstet Gynecol
new study: III. Diagnostic tests. In: Hulley SB, Cummings SR, eds. Designing clinical research. Baltimore: Williams & Wilkins, 1988:87-97. 24. van Twuyver E, Mooijaart RJD, ten Berge IJM, et al. Prdransplanlalion blood transfusion revisiled. N Engl J Med 1991;325:1210-3. 25. Sugi T, Makino T, Maruyama T, Kim WK, Iizuka R. A possible mechanism of immunotherapy for patients with recurrent spontaneous abortions. Am J Reprod Immunol 1991 ;25: 185-9.
Bound volumes available to subscribers Bound volumes of the AMERICAN JOURNAL OF OBSTETRICS Ar\D GVr\ECOLOGY are available to subscribers (only) for the 1993 issues from the Publisher, at a cost of $74.50 for domestic, $102.72 for Canada, and $97.50 for international for Vol. 168 (January-June) and Vol. 169 (July-December). Shipping charges are included. Each bound volume contains a subject and author index and all advertising is removed. Copies are shipped within 60 days after publication of the last issue in the volume. The binding is durable buckram with the JOURr\AL name, volume number, and year stamped in gold on the spine. Payment must accompany all orders. Contact Mosby, Subscription Services, 11830 Westline Industrial Drive, St. Louis, MO 63146-3318, USA; phone (800) 453-4351 or (314) 453-4351. Subscriptions must be in force to qualify. Bound volumes are not available in place of a regular JOURNAL subscription.