Demographics and characteristics of pregnant SLE patients and non-pregnant SLE controls
There were 90 pregnancies from 77 patients (1, 2 and 3 pregnancies in 66, 9 and 2 patients, respectively) during the study period. None of the pregnant SLE patients or controls were alcohol drinkers or smokers. Seven patients were diagnosed SLE at the time of conception; therefore, the activity of SLE at the time of conception also was used to determine SLE disease activity at -6M and − 3M. SLE was diagnosed at the time of conception in 7 pregnancies (7.78%), and within 5 years before and after SLE diagnosis in 49 (54.4%), and 34 (37.78%), respectively. Among the 90 pregnancies, 45 (50.00%) were a first pregnancy, 25 (27.78%) a second one, and 20 (22.22%) a third one or more. Thirty-three of the 90 pregnancies (36.67%) were considered active at the time of conception (mSLEDAI score > 0).
The demographics and characteristics of the pregnant SLE patients and controls are shown in Table 1. Patients in the pregnancy group had slightly, but significantly, lower age at SLE onset (21.63 ± 5.89 years vs. 24.05 ± 7.27 years, p = 0.015) than the control group. In addition, the pregnancy group had significantly higher cumulative prevalence of facial rashes and photosensitivity when compared with the controls. However, it had significantly lower prevalence of dyslipidemia, presence of anti-dsDNA antibodies and lupus anticoagulants. Although the SLE disease activity (mSLEDAI-2K) and medication used at -6M were no different between the two groups; the SLE disease activity, mean daily dose of prednisolone and use of immunosuppressive drugs at conception were significantly higher in the control group (3.57 ± 4.28 vs. 1.91 ± 3.44, p = 0.003, 12.24 ± 14.14 mg/day vs. 8.74 ± 11.37 mg/day, p = 0.018, and 45.56% vs. 24.44%, p = 0.003, respectively).
Table 1
Demographics and characteristics of pregnant SLE patients and controls
Characteristics
|
Pregnancy group
(N = 90)
|
Control group
(N = 90)
|
p-value
|
Age at SLE diagnosis (years)
|
21.63 ± 5.89
|
24.05 ± 7.27
|
0.015
|
Disease duration prior to pregnancy
|
5.36 ± 5.12
|
5.24 ± 4.45
|
0.674
|
Age at pregnancy (years)
|
26.94 ± 4.80
|
|
|
Duration of pregnancy (weeks)
|
31.41 ± 9.56
|
|
|
Co-morbidities
|
|
|
|
Hypertension, n (%)
|
23 (25.56)
|
15 (16.67)
|
0.144
|
Diabetes, n (%)
|
1 (1.11)
|
4 (4.44)
|
0.368
|
Dyslipidemia, n (%)
|
8 (8.89)
|
19 (21.11)
|
0.022
|
Old pulmonary TB, n (%)
|
1 (1.11)
|
0
|
1.000
|
Thalassemia, n (%)
|
7 (7.78)
|
4 (4.44)
|
0.351
|
Cumulative clinical manifestations at 6 months prior to conception
|
|
|
|
Facial rashes
|
61 (67.78)
|
46 (51.11)
|
0.023
|
Oral ulcers
|
31 (34.44)
|
28 (31.11)
|
0.634
|
Photosensitivity
|
36 (40.00)
|
19 (21.11)
|
0.006
|
Discoid rashes
|
34 (37.78)
|
34 (37.78)
|
1.000
|
Arthritis
|
50 (55.56)
|
52 (57.78)
|
0.764
|
Serositis
|
9 (10.00)
|
12 (13.33)
|
0.486
|
Neurological involvement
|
10 (11.11)
|
7 (7.78)
|
0.445
|
Hematologic involvement
|
57 (63.33)
|
67 (74.44)
|
0.107
|
Renal involvement
|
72 (80.00)
|
64 (71.11)
|
0.165
|
ANA, n (%)
|
89 (98.89)
|
89 (98.89)
|
1.000
|
Anti-dsDNAa, n (%)
|
50/85 (58.82)
|
65/84 (77.38)
|
0.010
|
Anti-Sma, n (%)
|
1/12 (8.33)
|
1/1 (100.00)
|
0.154
|
Anti-cardiolipin a, n (%)
|
4/58 (6.90)
|
5/42 (11.90)
|
0.388
|
Lupus anticoagulant a, n (%)
|
3/42 (7.14)
|
8/33 (24.24)
|
0.038
|
Anti-Roa, n (%)
|
21/46 (45.65)
|
7/18 (38.89)
|
0.624
|
Anti-Laa, n (%)
|
20/47 (42.56)
|
6/18 (33.33)
|
0.497
|
SLICC score
|
0.40 ± 0.72
|
0.29 ± 0.66
|
0.255
|
6 months prior to conception
|
|
|
|
Disease activity (mSLEDAI-2K)
|
1.72 ± 3.22
|
2.79 ± 4.26
|
0.098
|
Medication used
|
|
|
|
No medication, n (%)
|
14 (15.56)
|
6 (6.67)
|
0.058
|
Prednisolone, n (%)
|
73 (81.11)
|
80 (88.89)
|
0.144
|
Prednisolone dose, mg/day
|
8.19 ± 9.40
|
12.15 ± 14.24
|
0.110
|
Hydroxychloroquine, n (%)
|
36 (40.00)
|
43 (47.78)
|
0.293
|
Immunosuppressive drugsb, n (%)
|
27 (30.00)
|
38 (42.22)
|
0.088
|
At conception
|
|
|
|
Disease activity (mSLEDAI-2K)
|
1.91 ± 3.44
|
3.57 ± 4.28
|
0.003
|
Disease activity (mSLEPDAI )
|
1.90 ± 3.44
|
|
|
Medication used
|
|
|
|
No medication, n (%)
|
17 (18.89)
|
8 (8.89)
|
0.060
|
Prednisolone, n (%)
|
73 (81.11)
|
82 (91.11)
|
0.843
|
Prednisolone dose, mg/day
|
8.74 ± 11.37
|
12.24 ± 14.14
|
0.018
|
Hydroxychloroquine, n (%)
|
37 (41.11)
|
43 (47.78)
|
0.368
|
Immunosuppressive drugsb, n (%)
|
22 (24.44)
|
41 (45.56)
|
0.003
|
Data are expressed as mean ± SD unless indicated otherwise; a = number of positive tests/number of tested; b = some patients received more than one immunosuppressive drug, mSLEDAI-2K = modified Systemic Lupus Erythematosus Disease Activity Index − 2000, mSLEPDAI = modified Systemic Lupus Erythematosus Pregnancy Disease Activity Index. |
At the time of conception, 4 and 6 patients received mycophenolate mofetil and cyclophosphamide, respectively; and these 2 medicines were switched to azathioprine and cyclosporine, respectively. The pregnancies terminated in the 1st, 2nd and 3rd trimester in 8, 11 and 71 pregnancies, respectively. Nineteen pregnancies ended in pregnancy loss (spontaneous abortion in 12 [7 in the 1st trimester, and 5 in the 2nd ], induced abortion in 5 [1 in the 1st trimester, and 4 in the 2nd ], stillbirth in 1 [in the 2nd trimester], and dead fetus in the utero in 1 [in the 3rd trimester]). There were 28 full term and 43 premature (one was a twin pregnancy) deliveries and 1 post-term delivery.
Correlation between the mSLEDAI-2K and mSLEPDAI
The correlation between the mSLEDAI-2K and mSLEPDAI scores in pregnant SLE patients was determined from the time of conception to termination of pregnancy or delivery, and it was found that both scores showed very good correlation (slope 1.002, r = 0.988, p < 0.001). Thus, the mSLEDAI-2K was used to compare SLE disease activity between the pregnancy and control group.
SLE disease activity, active organ involvement, flares and flare rate during pregnancy
Changes in SLE disease activity among the all pregnancy and controls groups and their subgroup based on the degree of SLE disease activity (remission, mild, and moderate and high) from − 6M to the third trimester, are shown in Table 2 and Supplementary Fig. 1). As the number of patients with moderate and high disease activity was small, these 2 groups were combined into one for statistical analysis. Also, as there was a significant difference in the age at SLE onset, and the mSLEDAI-2K score, the daily dose of prednisolone and immunosuppressive drugs used at the time of conception, these 4 variables were controlled in the analysis.
Table 2
SLE disease activity (mSLEDAI-2K) and number of patients according to disease severity both in the pregnancy and control group, from 6 months prior to conception and in each trimester
SLE disease activity (mSLEDAI-2K score)
|
Period of evaluation
|
-6M
|
-3M
|
Conception
|
1st trimester
|
2nd trimester
|
3rd trimester
|
All cases
|
|
|
|
|
|
|
Pregnancy group, n
|
90
|
90
|
90
|
90
|
82
|
71
|
Mean ± SEM
|
1.72 ± 0.34
|
2.04 ± 0.40
|
1.91 ± 0.36
|
2.59 ± 0.41
|
2.78 ± 0.45a
|
2.58 ± 0.46b
|
Control group, n
|
90
|
90
|
90
|
90
|
82
|
71
|
Mean ± SEM
|
2.79 ± 0.45
|
2.98 ± 0.39
|
3.57 ± 0.45
|
2.57 ± 0.38h
|
2.43 ± 0.39i
|
2.38 ± 0.47j
|
p-value
|
|
0.309
|
0.019
|
0.453
|
0.379
|
0.262
|
Remission (mSLEDAI-2K = 0)
|
|
|
|
|
|
|
Pregnancy group, n (%)
|
61 (67.77)
|
61 (67.77)
|
61 (67.77)
|
61 (67.77)
|
55 (67.07)
|
49 (69.01)
|
Mean ± SEM
|
0
|
0.74 ± 0.41
|
0.74 ± 0.31
|
1.84 ± 0.47c
|
2.25 ± 0.55d
|
2.06 ± 0.56e
|
Control group, n (%)
|
52 (57.78)
|
52 (57.58)
|
52 (57.58)
|
52 (57.78)
|
49 (59.76)
|
41 (57.75)
|
Mean ± SEM
|
0
|
1.38 ± 0.38
|
2.67 ± 0.58
|
1.36 ± 0.38k
|
1.86 ± 0.48
|
1.83 ± 0.57
|
p-value
|
|
0.241
|
0.001
|
0.496
|
0.663
|
0.722
|
Mild (mSLEDAI-2K = 1–5)
|
|
|
|
|
|
|
Pregnancy group, n (%)
|
19 (21.11)
|
19 (21.11)
|
19 (21.11)
|
19 (21.11)
|
18 (21.95)
|
16 (22.53)
|
Mean ± SEM
|
3.21 ± 0.25
|
3.10 ± 0.42
|
3.95 ± 0.79
|
3.95 ± 0.82
|
3.44 ± 0.80
|
4.00 ± 1.05
|
Control group, n (%)
|
18 (20.00)
|
18 (20.00)
|
18 (20.00)
|
18 (20.00)
|
17 (20.73)
|
16 (22.53)
|
Mean ± SEM
|
3.11 ± 0.24
|
2.89 ± 0.46
|
3.11 ± 0.54
|
3.11 ± 0.56
|
2.94 ± 0.62
|
3.00 ± 0.93
|
p-value
|
|
0.777
|
0.363
|
0.363
|
0.599
|
0.314
|
Moderate and high (mSLEDAI-2K ≥ 6)
|
|
|
|
|
|
|
Pregnancy group, n (%))
|
10 (11.11)
|
10 (11.11)
|
10 (11.11)
|
10 (11.11)
|
9 (10.98)
|
6 (8.45)
|
Mean ± SEM
|
9.40 ± 0.99
|
8.00 ± 1.30
|
5.20 ± 1.53
|
4.60 ± 1.55
|
4.67 ± 1.60f
|
3.00 ± 1.12 g
|
Control group, n (%)
|
20 (22.22)
|
20 (22.22)
|
20 (22.22)
|
20 (22.22)
|
16 (19.51)
|
14 (19.72)
|
Mean ± SEM
|
9.75 ± 0.77
|
7.20 ± 0.89
|
6.30 ± 1.09
|
5.20 ± 1.10
|
3.62 ± 1.19l
|
3.28 ± 1.32 m
|
p-value
|
|
0.670
|
0.529
|
0.725
|
0.759
|
0.908
|
The repeated measures mixed model was used for analysis and controlled by age at SLE onset and mSLEDAI-2K, daily dose of prednisolone and use of immunosuppressive drugs. mSLEDAI-2K = modified SLE disease activity index, -6M = 6 months prior to conception, -3M = 3 months prior to conception. |
a-g = p value of the mSLEDAI-2K score in the pregnancy group, compared between the time of conception and 1st, 2nd and 3rd trimester (a, p = 0.040; b, p = 0.028; c, p = 0.020; d, p = 0.002; e, p = 0.005; f. p = 0.008; g, p = 0.002). |
h-m = p value of the mSLEDAI-2K score in the control group, compared between the time of conception and 1st, 2nd and 3rd trimester, corresponding to time in the pregnancy group (h, p = 0.010; i, p = 0.018; j, p = 0.015; k, p = 0.011; l, p = 0.008; m = 0.002). |
The mSLEDAI-2K score in the pregnancy group was slightly lower than that of the controls at -6M and − 3M, but it became lower with statistical significance at the time of conception (1.91 ± 0.36 vs. 3.57 ± 0.45, p = 0.019). The mSLEDAI-2K score in the pregnancy group slightly increased in the 1st, 2nd and 3rd trimester when compared with the time of conception, however, it did not show any statistically significant difference from the control group. In the subgroup analysis, those who were in clinical remission at -6M (mSLEDAI-2 k = 0) had an mSLEDAI-2K score that increased from − 3M to the 3rd trimester in both groups, but it was significantly higher in the control group than in the pregnancy group only at the time of conception (0.74 ± 0.31 vs. 2.67 ± 0.58, p = 0.001). There was no significant difference in mean mSLEDAI-2K score between the pregnancy and control group from − 6M to the 3rd trimester among patients with mild, or moderate and high disease activity.
The effect of pregnancy on SLE disease activity was determined. The disease activity (mSLEDAI-2K score) at the time of conception was compared with the 1st, 2nd and 3rd trimester in both the pregnancy and control group, and the results are shown in Table 2. The mSLEDAI-2K score in the all pregnancy group increased slightly and gradually after conception and became statistically higher at the 2nd and 3rd trimester. Pregnant patients with disease in remission at -6M had a slightly and progressively increasing mSLEDAI-2K score which was significantly higher in all the trimesters than at the time of conception. The mSLEDAI-2K score in those with mild disease activity at -6M did not show any significant changes after conception throughout the pregnancy period. It was interesting that the mSLEDAI-2K score in pregnant patients with moderate and high disease activity at -6M showed a progressive decline, which was significantly lower than at the time of conception at the 2nd and 3rd trimester. On the other hand, all of the patients in the control group had progressive decrease, but significantly, in the mSLEDAI-2K score at the 1st, 2nd and 3rd trimesters, when compared with the time of conception. The controls with disease in remission at -6M had a lower mSLEDAI-2K score after conception, which was statistically lower than that at the time of conception at the 1st trimester only. There was no significant change in the mSLEDAI-2K score in the control group with mild disease activity at -6M after conception during the pregnancy period. The mSLEDAI-2K score among the controls with moderate and high disease activity at -6M declined progressively after conception and was significantly lower at the 2nd and 3rd trimester. These mSLEDAI-2K changes within the pregnancy and control group, although statically significance, ranged between 0.87–2.20 and 1.31–3.02, respectively. Results of these comparisons (a to m) are shown at the footnote in Table 2.
Active organ involvement seen during the study from − 6M to the 3rd trimester, both in the pregnancy and control groups, are shown in Supplementary Table 1. Renal involvement (mostly proteinuria), mucocutaneous lesions, and hematologic abnormalities were the 3 most common organs involved and seen in 21.11–36.62%, 11.27–18.89% and 0-6.67% of the pregnancies, respectively. However, only hematologic abnormalities were seen to be significantly more common at -6M in the control than in the pregnancy group (6.67% vs. 0%, p = 0.029).
The incidence and categories of flares among the pregnancy and control group in each trimester are shown in Table 3. The overall number of flares was higher numerically in the pregnancy than in the control group, but this did not reach statistical significance (37 in 90 pregnancies [41.11%] vs. 26 in 90 pregnancies [28.89%], p = 0.070). Subgroup analysis of flares in patients who were in remission, and had mild, or moderate and high disease activity at the time at conception, did not show any significant differences in flare rate. The 37 flares in the pregnancy group occurred in the 1st, 2nd and 3rd trimester in 19 of 90 (21.11%), 28 of 82 (34.15%), and 28 of 71 (39.44%) pregnancies, respectively.
Table 3
Prevalence and categories of flares from the time of conception and in trimesters among pregnant SLE patients and their controls (according to their disease activity at the time of conception)
|
|
|
1st trimester
|
|
2nd trimester
|
|
3rd trimester
|
|
Overall
|
|
|
N/N1
|
n1 (%) : n2 (%)
|
N/N1
|
n1 (%) : n2 (%)
|
N/N1
|
n1 (%) : n2 (%)
|
N/N1
|
n 1(%) : n2 (%)
|
All cases
|
Pregnancy group
|
19/90
|
3 (15.79) : 16 (84.21)
|
12/82
|
5 (41.67) : 7 (58.33)
|
6/71
|
1 (16.67) : 5 (83.33)
|
37/90
|
9 (24.32) : 28 (75.68)
|
|
Control group
|
7/90
|
2 (28.57) : 5 (71.43)
|
12/82
|
6 (50.00) : 6 (50.00)
|
7/71
|
4 (57.14) : 3 (42.86)
|
26/90
|
12 (46.15) : 14 (53.85)
|
|
p - value
|
|
0.588
|
|
0.682
|
|
0.266
|
|
0.070
|
Remission
|
Pregnancy group
|
9/57
|
0 : 9 (100.00)
|
10/50
|
5 (50.00) : 5 (50.00)
|
3/45
|
0 : 3 (100.00)
|
22/57
|
5 (22.73) : 17 (77.27)
|
|
Control group
|
5/40
|
1 (20.00) : 4 (80.00)
|
7/38
|
3 (42.86) : 4 (57.14)
|
2/32
|
1 (50.00) : 1 (50.00)
|
14/40
|
5 (35.71) : 9 (64.29)
|
|
p - value
|
|
0.357
|
|
1.000
|
|
0.400
|
|
0.396
|
Mild
|
Pregnancy group
|
6/22
|
3 (50.00) : 3 (50.00)
|
1/21
|
0 : 1 (100.00)
|
3/19
|
1 (33.33) : 2 (66.67)
|
10/22
|
4 (40.00) : 6 (60.00)
|
|
Control group
|
1/28
|
0 : 1 (100.0)
|
5/25
|
3 (60.00) : 2 (40.00)
|
3/20
|
2 (66.67) : 1 (33.33)
|
9/28
|
5 (55.56) : 4 (44.44)
|
|
p - value
|
|
1.000
|
|
1.000
|
|
1.000
|
|
0.656
|
Moderate
and high
disease activity
|
Pregnancy group
|
4/11
|
0 : 4 (100.00)
|
1/11
|
0 : 1 (100.00)
|
0/7
|
0 : 0
|
5/11
|
0 : 5 (100.00)
|
Control group
|
1/22
|
1 (100.00) : 0
|
0/19
|
0 : 0
|
2/19
|
1 (50.00) : 1 (50.00)
|
3/22
|
2 (66.67) : 1 (33.33)
|
p - value
|
|
0.200
|
|
|
|
|
|
0.107
|
N = number of flares, N1 = number of pregnancies at each trimester, n1 (%): n2 (%) = number of mild or moderate flares (%) : number of severe flares (%). |
Effect of pregnancy termination or delivery on SLE disease activity and flares
To determine the effect of pregnancy termination or delivery on SLE disease activity, the mSLEDAI-2K score at the time of last visit prior to pregnancy termination (either by spontaneous or induced abortion, stillbirth or dead fetus in the utero) or delivery was compared with that of the post-partum visit (Table 4). There was no significant difference in the mSLEDAI-2K score between that at the last visit prior to pregnancy termination or delivery and that in the post-partum period in all of the pregnancy and control group, no matter whether they had a flare or not. There also was no significant difference in the mSLEDAI-2K score between the last visit prior to pregnancy termination or delivery and the post-partum period among those with successful pregnancy or pregnancy loss, no matter whether there was a flare or not.
Table 4
SLE disease activity at the time of last visit prior to pregnancy termination or delivery and at the time of post-partum visit in the pregnancy and control group
Group
|
Flare status
|
SLE disease activity (mSLEDAI-2K)
|
Last visit prior to pregnancy
termination or
delivery
(Mean ± SEM)
|
post-partum period
(Mean ± SEM)
|
p- value
|
Pregnancy group
|
N = 90
|
2.61 ± 0.40
|
2.50 ± 0.38
|
0.713
|
|
No flare (n = 48)
|
1.33 ± 0.42
|
1.25 ± 0.42
|
0.312
|
|
Flare (n = 42)
|
4.07 ± 0.64
|
3.93 ± 0.61
|
0.823
|
Control group
|
N = 90
|
2.81 ± 0.42
|
2.26 ± 0.39
|
0.074
|
|
No flare (n = 60)
|
2.13 ± 0.48
|
2.10 ± 0.48
|
0.654
|
|
Flare (n = 30)
|
4.17± 0.76
|
2.57 ± 0.69
|
0.073
|
Successful pregnancy
|
N = 71
|
2.58 ± 0.46
|
2.27 ± 0.43
|
0.350
|
|
No flare (n = 37)
|
0.97 ± 0.42
|
0.97 ± 0.42
|
|
|
Flare (n = 4)
|
4.32 ± 0.75
|
3.68 ± 0.71
|
0.347
|
Pregnancy loss
|
N = 19
|
2.74 ± 0.75
|
3.37 ± 0.82
|
0.355
|
|
No flare (n = 11)
|
2.54 ± 1.11
|
2.18± 1.13
|
0.294
|
|
Flare (n = 8)
|
3.00 ± 1.00
|
5.00 ± 1.00
|
0.052
|
The repeated measures mixed model was used for analysis and controlled by age at SLE onset, mSLEDAI-2K score, daily dose of prednisolone and use of immunosuppressive drugs. N = number of pregnancies, n = number of events, mSLEDAI-2K = modified the Systemic Lupus Erythematosus Disease Activity Index − 2000. |
SLE disease activity during the post-partum period was compared between the pregnancy and control group (Table 5). There was no significant difference in the mSLEDAI-2K score between the two groups. Although the number of flares was higher numerically in the pregnancy group than in the controls, it did not reach a statistically significant difference (46.67% vs. 33.33%, p = 0.091). Similarly, in the subgroup analysis according to flare categories, the pregnancy group had numerically more severe flares than the controls, but they did not reach statistical significance (76.19% vs. 53.33%, p = 0.095).
Table 5
Disease activity and flares at the post-partum period among pregnant SLE patients and their controls
|
|
p-value
|
SLE disease activity (mSLEDAI-2K score)
|
|
|
Pregnancy group, n = 90
|
Mean ± SEM
|
2.50 ± 0.38
|
0.354
|
Control group, n = 90
|
Mean ± SEM
|
2.26 ± 0.39
|
|
Any post-partum flares, n (%)
|
|
|
Pregnancy group
|
No. flares/No. of pregnancy (%)
|
42/90 (46.67)
|
0.129
|
Control group
|
No. flares/No. of pregnancy (%)
|
30/90 (33.33)
|
|
Flare category, n (%)
|
|
|
Pregnancy group
|
Mild or moderate flare
|
10 (23.81)
|
0.095
|
|
Severe flare
|
32 (76.19)
|
|
Control group
|
Mild or moderate flare
|
14 (46.67)
|
|
|
Severe flare
|
16 (53.33)
|
|
The ANCOVA for continuous response and logistic regression for binary response was used for analysis and controlled by age at SLE onset, mSLEDAI-2K score, daily dose of prednisolone, and use of immunosuppressive drugs. mSLEDAI-2K = modified Systemic Lupus Erythematosus Disease Activity Index − 2000. |