1. Top of page
  2. Abstract
  6. Acknowledgements


Rheumatoid arthritis (RA) usually improves during pregnancy and recurs postpartum. Fetal cells and cell-free DNA reach the maternal circulation during normal pregnancy. The present study investigated dynamic changes in levels of fetal DNA in serum from women with RA and inflammatory arthritis during and after pregnancy to test the hypothesis that the levels of circulating fetal DNA correlate with arthritis improvement.


Twenty-five pregnant patients were prospectively studied. A real-time quantitative polymerase chain reaction panel targeting unshared, paternally transmitted HLA sequences, a Y chromosome–specific sequence, or an insertion sequence within the glutathione S-transferase M1 gene was used to measure cell-free fetal DNA. Results were expressed as fetal genomic equivalents per milliliter (gE/ml) of maternal serum. Physical examinations were conducted during and after pregnancy.


Levels of fetal DNA in women with improvement in or remission of arthritis were higher than those in women with active disease, especially in the third trimester. Overall, an inverse relationship between serum fetal DNA levels and disease activity was observed (P < 0.001). Serum fetal DNA increased with advancing gestation, reaching median levels of 24 gE/ml (range 0–334), 61 gE/ml (range 0–689), and 199 gE/ml (range 0–2,576) in the first, second, and third trimesters, respectively, with fetal DNA clearance observed postpartum. Arthritis improvement was initially noted in the first trimester for most patients, increased further or was sustained with advancing gestation, and was active postpartum.


Changes in serum fetal DNA levels correlated with arthritis improvement during pregnancy and recurrence postpartum. Immunologic mechanisms by which pregnancy might modulate RA activity are described.

Rheumatoid arthritis (RA) usually improves or even remits during pregnancy, while a disease flare regularly occurs within 3–4 months postpartum (1, 2). Arthritis improvement during pregnancy with active disease postpartum similarly has been reported in patients with juvenile idiopathic arthritis (JIA) (3, 4). Yet, little is known about the mechanisms involved in this clinically important phenomenon. In 2 previous studies, fetal–maternal HLA class II disparity was found to correlate with pregnancy-induced improvement in RA (5, 6). As a result of advances in techniques for prenatal diagnosis, it has recently been established that fetal cells and cell-free DNA routinely traffic into the maternal circulation during normal pregnancy (7, 8). Circulating fetal DNA is found in maternal serum/plasma in the first trimester of normal gestation, and the concentration of fetal DNA increases during the course of pregnancy (8). Rapid removal of fetal DNA occurs following parturition (9). Thus, recent studies of normal pregnancy, when considered along with previous observations of improvement in RA, led us to ask whether changes in serum levels of fetal DNA correlate with alterations in arthritis activity during and after pregnancy. Fetal DNA in the serum of pregnant RA patients was measured by real-time quantitative polymerase chain reaction (PCR), targeting unshared, paternally transmitted HLA sequences, a Y chromosome–specific sequence (DYS14), or an insertion sequence within the glutathione S-transferase M1 (GSTM1) gene.


  1. Top of page
  2. Abstract
  6. Acknowledgements


A total of 25 pregnant patients were evaluated prospectively. Seventeen were classified as having adult-onset RA according to the 1987 revised criteria of the American College of Rheumatology (ACR; formerly, the American Rheumatism Association) (10), 2 fulfilled the 1958 criteria for definite RA (11) but not the 1987 criteria, and 6 were classified as having JIA. Three of the patients with JIA had polyarticular-onset disease, 2 had systemic-onset disease, and 1 had oligoarticular-onset disease (12). Patients' ages at the time of the study pregnancy ranged from 23 to 43 years (median 33 years). All patients were white, and all index study pregnancies resulted in a single live birth. Seven patients were in their first pregnancy, 7 were in their second pregnancy, and 11 had been pregnant at least twice before. Among women with adult-onset RA (age >16 years), all pregnancies had occurred after disease onset in 63%. Prior to the study pregnancy, 12 women had had 1–3 live births (median 1 birth) after RA onset. Among the 19 patients fulfilling 1987 ACR or 1958 RA criteria, 13 were seropositive and 6 were seronegative for rheumatoid factor (RF). Of the 6 patients with JIA, 3 were RF positive and 1 was RF negative (2 unknown).

The study was approved by the institutional review board of the Fred Hutchinson Cancer Research Center. Informed consent was obtained from all participants.

Evaluation of disease activity and collection of blood samples.

Disease activity of the prospectively studied pregnancy was assessed by a study rheumatologist (JLN) after the patient underwent a physical examination and symptoms and medication use were recorded. Patients were examined and disease activity was scored (from 0 to 3) by the study rheumatologist every 2–3 months throughout the pregnancy and after parturition. Disease activity was also graded on a scale of 0 (inactive) to 3 (severely active) by the patients themselves, month by month, disease activity was for the 6 months before pregnancy, during pregnancy, and 3 months after delivery.

Patients were categorized into 2 groups. Group A included women who, compared with the 6 months before pregnancy, experienced improvement in joint symptoms, discontinued or reduced medications, and had ≤15 minutes of morning stiffness and no evidence of joint swelling throughout the third trimester. Group B included women with both subjective evidence and objective evidence (swelling noted on examination) of arthritis activity during pregnancy (described in more detail in a previous report [5]). Twenty-four women had active disease in the 6 months prior to pregnancy and 1 had RA onset in the first trimester.

Medication use by patients in both groups is summarized in Table 1. No patient took a disease-modifying antirheumatic drug during pregnancy, and patients taking prednisone (n = 6) took no more than 10 mg/day, with 1 exception. If a study subject had a previous pregnancy after RA onset, information regarding disease activity during the prior pregnancy was also sought by review of medical records and ranking of disease activity by the patient.

Table 1. Medication use during the 6 months before pregnancy, during the third trimester of pregnancy, and within 3 months postpartum*
NoneNSAIDsPrednisoneDMARDsPrednisone plus DMARDs
  • *

    Group A included subjects whose arthritis improved or remitted during pregnancy; group B included subjects with active disease during pregnancy. Values are the number (%) of patients. NSAIDs = nonsteroidal antiinflammatory drugs; DMARDs = disease-modifying antirheumatic drugs.

Group A (n = 21)     
 Prepregnancy1 (5)7 (33)8 (38)4 (19)1 (5)
 Trimester 315 (71)2 (10)4 (19)0 (0)0 (0)
 Puerperium5 (24)5 (24)6 (29)2 (10)3 (14)
Group B (n = 4)     
 Prepregnancy0 (0)1 (25)1 (25)1 (25)1 (25)
 Trimester 31 (25)1 (25)2 (50)0 (0)0 (0)
 Puerperium0 (0)2 (50)2 (50)0 (0)0 (0)

Peripheral venous blood was drawn at 2–5 different time points per patient (median 3) during and after the study pregnancy. Postpartum samples were collected within 3 months of delivery. Fetal sex was noted and umbilical cord blood was sampled at the time of delivery.

HLA and GSTM1 genotyping.

DNA was extracted from whole heparinized blood or peripheral blood mononuclear cells (PBMCs) obtained from the patients and from umbilical cord blood. DNA-based HLA typing of all probands and their children was performed for HLA–B, HLA–DRB (DRB1, DRB3, DRB4, and DRB5), and HLA–DQB1 loci, as previously described (5, 13). HLA typing is currently conducted using the DRB RELI SSO Strip typing kit (Dynal, Oslo, Norway) for initial determination of DRB families, followed by identification of the specific allele by sequencing and of DQB1 and HLA–B with typing kits (Dynal). Mother–child pairs were also genotyped for the GSTM1 locus by a conventional PCR system in which DNA from individuals with the GSTM1 gene produces an 80-bp PCR product that is absent in a GSTM1-null homozygote (7).

Extraction of DNA from serum.

DNA was extracted from serum with a QIAamp DNA blood mini kit (Qiagen, Valencia, CA), following the “blood and body fluid protocol” recommended by the manufacturer. For DNA extraction, a 500-μl serum sample was used. The extracted DNA was eluted into a final volume of 100 μl.

Real-time quantitative PCR.

A panel of HLA-specific quantitative PCR assays targeting unshared, paternally transmitted HLA sequences was used to measure serum fetal DNA. The development and validation of the panel of quantitative PCR assays was recently reported (13). If there was no fetal-specific HLA polymorphism that was informative for any of the HLA-specific assays, a quantitative PCR assay specific for the Y chromosome sequence DYS14 was used for women carrying a male fetus (14), and an assay for the insertion allele of the GSTM1 gene was used if the baby had this allele and the mother had the null allele (7). Quantitative PCR reactions were set up in a 50-μl volume, with 5 μl of DNA extracted from serum. Each sample was tested in duplicate for fetal DNA, and a standard curve for the fetus-specific sequence was run in parallel and in duplicate to quantify the fetal DNA. The result was expressed as fetal genomic equivalents per milliliter (gE/ml) of maternal serum by using a conversion factor of 6.6 pg of DNA per cell. All samples were tested for a housekeeping gene, β-globin, to confirm the quality (ability to amplify) of the extracted DNA. To ensure intra- and interassay homogeneity, a fetal-specific standard curve and a β-globin standard curve were run simultaneously on all plates (14). Seventeen patients were tested using fetal HLA-specific assays, 7 were tested using the DYS14 assay, and 1 was tested using GSTM1 quantitative PCR.

Statistical analysis.

To estimate the overall association between disease activity and fetal DNA, disease improvement/lack of improvement was treated as binary outcomes in a logistic regression model. The serum fetal DNA values were log-transformed to approximate normality. Measurements based on multiple blood draws per patient were entered into the analysis as repeated measures, with adjustment for possible correlation between values within a subject (15). A similar model was used to compare the likelihood of being in group A as opposed to group B, given fetal DNA levels in the third trimester. Age at disease onset and pregnancies after as well as before RA onset were considered as potential confounders. P values less than 0.05 were considered significant.


  1. Top of page
  2. Abstract
  6. Acknowledgements

Improvement or remission of arthritis occurred during pregnancy in 15 patients with adult-onset RA and in all 6 patients with JIA. These patients were assigned to group A. Among these women, 62% had experienced at least some improvement by the end of the first trimester. Once improvement took place, it was sustained or increased further during the course of gestation. Disease recurrence was observed in 90% of patients in group A by the third to fourth month after delivery. Four RA patients with active disease throughout pregnancy, including 1 who had disease onset within the first trimester, were assigned to group B, and the disease in these patients remained active postpartum. Consistent with previous reports (2), most women who had experienced improvement during previous pregnancies improved during the index study pregnancies, with 1 exception. Levels of fetal DNA were evaluated in each patient over the course of the study pregnancy and postpartum (i.e., each patient as her own reference), with an additional evaluation that compared all of group A with all of group B.

Levels of circulating fetal DNA in serum increased with advancing gestation (Figure 1), reaching a median of 24 gE/ml (range 0–334) in the first trimester, 61 gE/ml (range 0–689) in the second trimester, and 199 gE/ml (range 0–2,576) in the third trimester. Clearance of serum fetal DNA was observed in the postpartum period (median 0; range 0–166). Serum fetal DNA levels were significantly higher during the third trimester in group A (median 270 gE/ml; range 12–2,576) as compared with group B (median 78 gE/ml; range 0–224) (P < 0.001), with fetal DNA detectable in 100% of study subjects from group A and 50% of subjects from group B. No qualitatively or quantitatively significant differences in serum fetal DNA were identified between the 2 groups within 3 months after delivery.

thumbnail image

Figure 1. Changes in serum fetal DNA levels during and after pregnancy in patients with arthritis. Each line represents 1 patient.

Download figure to PowerPoint

Overall, there was an inverse relationship between the serum fetal DNA level and arthritis disease activity. As fetal DNA quantities doubled, the likelihood of arthritis improvement increased 1.2-fold (95% confidence interval 1.1–1.3; P < 0.001). There was an increased likelihood of improvement with increasing levels of serum fetal DNA (without suggestion of a specific threshold). Results did not differ between patients with adult-onset RA and those with juvenile-onset arthritis (nor for the 2 adult patients who had definite RA according to the 1958 criteria but not the 1987 revised criteria). Results were not substantially altered by adjustment for the potential confounders, as described in Patients and Methods.


  1. Top of page
  2. Abstract
  6. Acknowledgements

Fetal DNA has not previously been studied in women with RA during pregnancy. In the current study, we found a significant inverse correlation between arthritis activity and the serum fetal DNA concentration over the course of pregnancy and postpartum. Arthritis improved in 79% of patients with adult-onset disease and in all women with JIA, over a time course similar to the time course of rising serum fetal DNA levels, which increased as pregnancy progressed and dropped to undetectable or very low levels after delivery, coincident with arthritis recurrence. In the third trimester, serum levels of fetal DNA were significantly higher in patients with arthritis improvement as compared with those with active disease.

A number of hypotheses have been proposed to explain arthritis improvement during pregnancy, including increased pregnancy-associated levels of serum α2-glycoprotein (16), elevated levels of sex hormones (e.g., estrogen, progesterone) (17), and a shift in cytokine production from a Th1 to a Th2 profile (18). In 2 prior studies, fetal–maternal HLA class II disparity was associated with the pregnancy-induced amelioration of RA (5, 6). No association was described in another study (19). However, when identifying the pregnancy-induced amelioration of RA, it is important to compare disease during pregnancy with that prior to pregnancy, whereas this study did not include the requirement that a woman have active arthritis in the 6 months prior to conception. (The use of “no change” as a disease activity category can result in identical classification of a patient whose disease was in remission prior to pregnancy and remained in remission and a patient who had active disease and continued to have active disease during pregnancy.)

Recent findings in studies of normal pregnancy, when considered along with the results of the current study and previous studies of RA, led us to a hypothesis for the mechanism by which pregnancy induces improvement in RA. Whether serum fetal DNA has any direct biologic effects is unknown, and although a direct effect remains possible given reports of horizontal transfer of DNA in other conditions (20), in our hypothesis, fetal DNA measured in serum is an indicator of other processes, as described further. Villous trophoblast apoptosis (a normal physiologic phenomenon during placental growth) results in extrusion of a large quantity of apoptotic fetal trophoblasts into the maternal circulation, which is thought to be the most important source of serum fetal DNA (7). Although trophoblasts do not express HLA class II antigens, HLA–DR protein has recently been identified within the cytoplasm (21). Thus, increasing levels of serum cell-free fetal DNA may reflect a progressive release of fetal HLA class II alloantigens from cell lysis as pregnancy progresses. In addition, not all fetal DNA in serum is cell-free, since some derives from fetal cells in the process of apoptosis (14, 22).

The uptake and cross-presentation of soluble fetal paternally transmitted HLA peptides and/or peptides packaged within apoptotic fetal cells (23) by maternal antigen-presenting cells, such as dendritic cells, is likely to have immunologic effects that may reasonably be anticipated to also affect autoimmunity in RA patients, for example by induction of regulatory T cells or by altering the maternal peripheral T cell repertoire. After delivery, the maternal T cell repertoire would be expected to return to the prepregnancy “baseline” due to lack of a source of apoptotic fetal cells, coincident with the relapse of arthritis. In a recent experimental study, persistence of allogeneic donor microchimeric cells was found to be essential for maintenance of the unresponsiveness of host T cells by clonal deletion of the donor-specific T cell repertoire (24). Therefore, whether intact fetal cells could modulate RA disease activity during pregnancy is of additional interest.

The current study has a number of limitations. First, most women develop RA in their postreproductive years, (25), and as in prior prospective studies of RA and pregnancy, the number of study subjects is modest. Second, some study subjects took medications during gestation, and whether the medications influence levels of serum fetal DNA is unknown. To the extent that we could examine this possibility, prednisone did not appear to have an effect on serum fetal DNA levels (4 patients in group A took prednisone and 2 in group B took it during the third trimester of their pregnancies). Third, most of our quantitative assays are based on HLA differences, so we were not able to evaluate the additional question of whether fetal–maternal HLA class II compatibility influences levels of fetal DNA in the mother. Finally, our study subjects were identified when they were already pregnant and prepregnancy serum samples were not available, thus, the results could possibly be confounded by persisting fetal DNA from a previous pregnancy. However, serum fetal DNA is rapidly cleared after pregnancy (9), and a previous study described persistent fetal cells among PBMCs but not in serum (14). We did not observe any difference in the results according to the number of prior pregnancies.

In summary, the serum fetal DNA concentration increased throughout gestation and was effectively cleared after delivery, with an inverse correlation observed between changes in fetal DNA levels and arthritis activity. Whether the dynamic changes in fetal DNA reflect the potential for immune modulation of maternal arthritis, are a result of disease activity changes, or are not causally related cannot be determined from these studies. If the changes reflect immune modulation, further studies could generate new therapeutic strategies for RA.


  1. Top of page
  2. Abstract
  6. Acknowledgements

We are indebted to Jennifer Brackensick for study coordination and to Allison Porter for technical support.


  1. Top of page
  2. Abstract
  6. Acknowledgements
  • 1
    Hench PS. The ameliorating effect of pregnancy on chronic atrophic (infectious rheumatoid) arthritis, fibrositis, and intermittent hydrarthrosis. Mayo Clin Proc 1938; 13: 1617.
  • 2
    Nelson JL, Ostensen M. Pregnancy and rheumatoid arthritis [review]. Rheum Dis Clin North Am 1997; 23: 195212.
  • 3
    Ostensen M. Pregnancy in patients with a history of juvenile rheumatoid arthritis. Arthritis Rheum 1991; 34: 8817.
  • 4
    Musiej-Nowakowska E, Ploski R. Pregnancy and early onset pauciarticular juvenile chronic arthritis. Ann Rheum Dis 1999; 58: 47580.
  • 5
    Nelson JL, Hughes KA, Smith AG, Nisperos BB, Branchaud AM, Hansen JA. Maternal-fetal disparity in HLA class II alloantigens and the pregnancy-induced amelioration of rheumatoid arthritis. N Engl J Med 1993; 329: 46671.
  • 6
    Van der Horst-Bruinsma IE, de Vries RR, de Buck PD, van Schendel PW, Breedveld FC, Schreuder GM, et al. Influence of HLA-class II incompatibility between mother and fetus on the development and course of rheumatoid arthritis of the mother. Ann Rheum Dis 1998; 57: 28690.
  • 7
    Lo YM, Lau TK, Chan LY, Leung TN, Chang AM. Quantitative analysis of the bidirectional fetomaternal transfer of nucleated cells and plasma DNA. Clin Chem 2000; 46: 13019.
  • 8
    Lo YM, Tein MS, Lau TK, Haines CJ, Leung TN, Poon PM, et al. Quantitative analysis of fetal DNA in maternal plasma and serum: implications for noninvasive prenatal diagnosis. Am J Hum Genet 1998; 62: 76875.
  • 9
    Lo YM, Zhang J, Leung TN, Lau TK, Chang AM, Hjelm NM. Rapid clearance of fetal DNA from maternal plasma. Am J Hum Genet 1999; 64: 21824.
  • 10
    Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988; 31: 31524.
  • 11
    Ropes MW, Bennett GA, Cobb S, Jacox R, Jessar RA. 1958 revision of diagnostic criteria for rheumatoid arthritis. Bull Rheum Dis 1958; 9: 1756.
  • 12
    Petty RE, Southwood TR, Manners P, Baum J, Glass DN, Goldenberg J, et al. International League of Associations for Rheumatology classification of juvenile idiopathic arthritis: second revision, Edmonton, 2001. J Rheumatol 2004; 31: 3902.
  • 13
    Lambert NC, Erickson TD, Yan Z, Pang JM, Guthrie KA, Furst DE, et al. Quantification of maternal microchimerism by HLA-specific real-time polymerase chain reaction: studies of healthy women and women with scleroderma. Arthritis Rheum 2004; 50: 90614.
  • 14
    Lambert NC, Lo YM, Erickson TD, Tylee TS, Guthrie KA, Furst DE, et al. Male microchimerism in healthy women and women with scleroderma: cells or circulating DNA? A quantitative answer. Blood 2002; 100: 284551.
  • 15
    Liang KY, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika 1986; 73: 1322.
  • 16
    Unger A, Kay A, Griffin AJ, Panayi GS. Disease activity and pregnancy associated α 2-glycoprotein in rheumatoid arthritis during pregnancy. Br Med J (Clin Res Ed) 1983; 286: 7502.
  • 17
    Masi AT, Feigenbaum SL, Chatterton RT. Hormonal and pregnancy relationships to rheumatoid arthritis: convergent effects with immunologic and microvascular systems [review]. Semin Arthritis Rheum 1995; 25: 127.
  • 18
    Ostensen M, Forger F, Nelson JL, Schuhmacher A, Hebisch G, Villiger PM. Pregnancy in patients with rheumatic disease: anti-inflammatory cytokines increase in pregnancy and decrease post partum. Ann Rheum Dis 2005; 64: 83944.
  • 19
    Brennan P, Barrett J, Fiddler M, Thomson W, Payton T, Silman A. Maternal-fetal HLA incompatibility and the course of inflammatory arthritis during pregnancy. J Rheumatol 2000; 27: 28438.
  • 20
    Holmgren L, Szeles A, Rajnavolgyi E, Folkman J, Klein G, Ernberg I, et al. Horizontal transfer of DNA by the uptake of apoptotic bodies. Blood 1999; 93: 395663.
  • 21
    Ranella A, Vassiliadis S, Mastora C, Valentina M, Dionyssopoulou E, Athanassakis I. Constitutive intracellular expression of human leukocyte antigen (HLA)-DO and HLA-DR but not HLA-DM in trophoblast cells. Hum Immunol 2005; 66: 4355.
  • 22
    Van Wijk IJ, de Hoon AC, Jurhawan R, Tjoa ML, Griffioen S, Mulders MA, et al. Detection of apoptotic fetal cells in plasma of pregnant women. Clin Chem 2000; 46: 72931.
  • 23
    Albert ML, Sauter B, Bhardwaj N. Dendritic cells acquire antigen from apoptotic cells and induce class I-restricted CTLs. Nature 1998; 392: 869.
  • 24
    Bonilla WV, Geuking MB, Aichele P, Ludewig B, Hengartner H, Zinkernagel RM. Microchimerism maintains deletion of the donor cell–specific CD8+ T cell repertoire. J Clin Invest 2006; 116: 15662.
  • 25
    Silman AJ. Rheumatoid arthritis. In: SilmanAJ, HochbergMC, editors. Epidemiology of the rheumatic diseases. New York: Oxford; 2001. p. 3144.