- Email: [email protected]
Interleukin-10 in amniotic fluid at midtrimester: Immune activation and suppression in relation to fetal growth
Interleukin-lO in amniotic fluid at midtrimester: Immune activation and suppression in relation to fetal growth Kent D. Heyborne, MD," James A. McGregor, MD," George Henry, MD,b Steven S. Witkin, PhD,c and John S. Abrams, PhD d Denver, Colorado, New York, New York, and Palo Alto, California OBJECTIVE: Low birth weight remains the leading cause of perinatal morbidity and mortality, but mechanisms mediating impaired fetal growth are poorly understood. To further define the role of abnormal immune activation and suppression in mediating impaired fetal growth, we measured levels of interleukin-10, a potent immunosuppressive cytokine not previously identified in association with pregnancy, in amniotic fluid samples obtained at genetic amniocentesis. STUDY DESIGN: In a case-control study with an enzyme-linked immunoassay we compared amniotic fluid levels of interleukin-1 0 in midtrimester samples obtained from appropriate-for-gestational age (n = 42) and small-for-gestational-age (n = 24) pregnancies. RESULTS: Interleukin-10 levels in small-for-gestational-age samples were elevated (median 78 pg/ml) compared with levels in control samples (median < 40 pg/ml), p = 0.02. In small-for-gestational-age pregnancies elevated levels were associated with nulliparity, p = 0.003. CONCLUSION: Our data support the role of abnormal immune activation, as opposed to inadequate immune suppression, in mediating impaired fetal growth. (AM J OBSTET GYNECOL 1994;171 :55-9.)
Key words: Low birth weight, immunology, interleukin-l0, intrauterine growth retardation
During placental invasion and growth the maternal immune system is challenged by fetal antigens present on trophoblasts. Current concepts suggest that the ensuing maternal immune response is complex and balanced. Some response to fetal antigens may enhance placental growth and development, whereas an abnormalone may lead to pregnancy loss or other perinatal complications. 1. 2 In recent years a group of protein messengers secreted by cells of the immune system and other tissues, referred to as cytokines, has been described. Cytokines may be thought of as pro inflammatory, immunosuppressive, or growth promoting. A balance between these
From the Department of Obstetrics and Gynecology, University of Colorado Health Sciences Center: Reproductive Genetics/ the Department of Obstetrics and Gynecology, Cornell University Medical Center,' and DNAX Research Institute! Supported by the Academic Enrichment Fund of the Department of Obstetrics and Gynecology of the University of Colorado Health Sciences Center, Denver. Presented at the Fortieth Annual Meeting of the Society for Gynecologic Investigation, Toronto, Ontario, Canada, March 31-April 3, 1993. Received for publication November 29, 1993; revised December 16, 1993; accepted December 28, 1993. Reprint requests: Kent D. Heyborne, MD, Department of Obstetrics and Gynecology, B-198, University of Colorado Health Sciences Center, Denver, CO 80262. Copyright © 1994 by Mosby-Year Book, Inc. 0002-9378/94 $3.00 + 0 6/1/54137
three cytokine activIties may influence placental and fetal growth. Second-trimester amniotic fluid levels of the cytokine colony-stimulating factor-l (CSF -1), a growth factor, are positively correlated with subsequent fetal growth." In the mouse maternal or fetal elaboration of CSF-l is required for successful pregnancy.4 Recently we determined that in human pregnancy increased midtrimester amniotic fluid levels of the proinflammatory cytokine tumor necrosis factor-a (TNF-a) are linked with poor fetal growth. 5 In this case-control study we investigated the relationship of other cytokines to fetal growth. To this end, we measured levels of second-trimester amniotic fluid interleukin-l0 (IL-I0, cytokine synthesis inhibitory factor), a potent immunosuppressive cytokine, in appropriate-for-gestational-age (AGA) and small-for-gestational-age (SGA) gestations. In addition, we measured levels of interleukin-6 (IL-6), which, like TNF-a, IS an inflammatory cytokine, in the same specimens.
Material and methods Amniotic fluid specimens were collected from January 1989 through September 1990 by midtrimester amniocentesis in a Denver private-practice genetics clinic where all patients had private health insurance. At amniocentesis demographic and medical information was obtained by genetic counselors. No patient in the study had a history of maternal immune disease;
55
56
July 1994
Heyborne et al.
Am J Obstet Gynecol
Table I. Maternal characteristics and indications for amniocentesis Maternal characteristics
Group 1 (SGA, n = 24)
(control, n = 44)
Significance
Age (yr) Gravidity Parity Percent previous spontaneous abortion
35.1 2.4 0.5 60
35.7 2.8 1.0 20
White Black Hispanic Other Smoker (> 0.5 pack/day) Gestational age at amniocentesis (wk) Indication for amniocentesis Maternal age High maternal serum a-fetoprotein Low maternal serum a-fetoprotein Other
20 1 1
41 1 1
NS* NSt p = 0.007t P = 0.003t NS:::
3 16.3
16.4
Race
Group 2
o
2
18 4 1 1
33
NS::: NSt
NSt
o
4 4
NS, Not significant. Age, gravidity, parity, gestational age at amniocentesis, and amniotic fluid a-fetoprotein are expressed as means. Other characteristics are expressed as numbers of patients. Percentage of previous pregnancies ending in miscarriage was calculated for each patient as follows: [Number of miscarriages -;- (Number of previous pregnancies - Number of elective terminations )]. *By Student t test. tBy Wilcoxon rank-sum test. tBy Fisher's exact test.
Table II. Pregnancy outcomes by group Pregnancy outcome Gestational age at delivery (wk) Infant sex (M/F ratio) Birth weight (gm) Birth weight percentile Ponderal index (Birth weight x 100)/Length3 ) Cesarean section Stillbirth Neonatal death Preterm delivery ( < 37 wk)
Group 1 (SGA, n = 24)
(control, n = 44)
Group 2
35.9 10:14 1812 4.3 2.16
39.6 22:20 3490 69.7 2.47
0.02* NSt p < 0.01* P < 0.01* P < 0.01*
14 1 2 12
15
p = 0.03ot
o o o
Significance
p<
NSt NSt p < O.Olt
NS, Not significant. The groups were statistically different as indicated. *By Wilcoxon rank-sum test. tBy Fisher's exact test.
one patient in the SGA group had chronic hypertension. All patients had an ultrasonographic examination at amniocentesis, confirming gestational age as determined by last menstrual period (within 1 week) and excluding early evidence of growth restriction. Amniotic fluid specimens were centrifuged (500 revolutions/min for 10 minutes) to remove cellular components for karyotype determination, and supernatants were frozen at - 20° C for later cytokine assay. At the expected date of delivery follow-up questionnaires requesting date of birth, birth weight and length, infant sex, route of delivery, and occurrence of stillbirth or neonatal death were sent to all patients. Responses were obtained from 2580 patients, representing an 80% response rate, and study patients were selected from this cohort.
Patient information was stored on a computer database (Revelation Technologies, New York) for later retrieval. From this database 24 specimens were identified as coming from gestations resulting in SGA births (birth weight < 10% for gestational age as determined by the Colorado Intrauterine Growth Curve).6 Less than 1% of patients from this private practice population fulfilled this definition of SGA, indicating that the pregnancies studied suffered from severe growth retardation relative to their demographic group. Forty-two patients with AGA term infants were randomly selected according to patient number by computer as controls. All specimens used in the study yielded normal karyotypic analysis. Immunoreactive IL-IO and IL-6 levels were deter-
Heyborne et al.
Volume 171, Number 1 Am J Obstet Gynecol
57
80r---------------------------------------~
<40
40-100
100-200
200-300
>300
40r-----------------------------------------,
30
10
o
<40
40-100
100-200
200-300
>300
Amniotic fluid IL-10, pg/ml Fig. 1. Frequency distribution of amniotic fluid IL-IO in AGA (toP) and SGA (bottom) gestations,
mined by immunoenzymetric assay with paired rat monoclonal antibodies for each cytokine, as previously described. 7 Amniotic fluid samples were clarified by centrifugation through a 0.45 J.Lm cellulose acetate Spin-X centrifuge filter unit (Costar, Cambridge, Mass.) and diluted 1: 1 with RPMI 1640 with 10% fetal bovine serum containing 50 J.Lg/ml each of rat immunoglobulin G 1 and immunoglobulin G 2 to inhibit false positivity that is related to rheumatoid factor or heterophilic antibody. Student t test, Wilcoxon rank sum analysis, Fisher's. exact test, and Spearman rank correlation were used for statistical comparisons.
Results
Table I and II compare the SGA and AGA groups regarding maternal characteristics, indications for amniocentesis, and pregnancy outcome. Both groups had a low incidence of clinical and epidemiologic risk factors traditionally associated with SGA birth (black race, smoking > 0.5 pack per day, or elevated maternal serum a-fetoprotein). Only six patients in the SGA group had at least one of the above risk factors, whereas two patients in the AGA group had such risk factors. The groups did not differ in maternal factors and amniocentesis indications, with the exception of parity and occurrence of previous spontaneous abortion. Birth
58
Heyborne et al.
weight, birth weight percentile, delivery gestational age, and ponderal index were significantly lower in the SGA group, and preterm delivery and cesarean section were more common in the SGA group. IL·I0 levels and fetal growth. Fig. 1 illustrates the frequency distribution of amniotic fluid IL-I0 levels in SGA and AGA pregnancies. Levels of IL-I0 in SGA pregnancies (median 78, range < 40 to 832 pg/ml) were elevated relative to those in AGA pregnancies (median < 40, range < 40 to 476 pg/ml), p = 0.02 by two-tailed Wilcoxon rank-sum test. With a threshold of 40 pg/ml defining an abnormal test, elevated IL-I0 levels identified SGA gestations with a sensitivity of 63% and a specificity of 74%. IL-6 was detected in all samples (median 548, range 18 to 10,198 pg/ml) but did not correlate with fetal growth, p > 0.05. TNF-cdevels were previously measured in a subset of these samples (24 SGA and 24 AGA specimens). IL-lO levels correlated with previously reported TNF -ex levels (Spearman correlation coefficient 0.48, p < 0.001). Clinical and epidemiologic risk factors and amniotic fluid TNF -ex have sensitivities of 25% and 48% and specificities of 95% and 83%, respectively, in this study group. IL·I0 levels and nulliparity. SGA patients were noted to have lower panty (0.5 vs 1.0, P = 0.007) and a higher percentage of previous pregnancies ending in miscarriage (60% vs 20%, P = 0.003) than the randomly selected AGA control patients. We found no relationship between IL-I0 levels and parity or the occurrence of a previous miscarriage when the study population as a whole was analyzed. Interestingly, in a subgroup analysis of SGA patients, nulliparity was strongly associated with increased IL-I0 levels, p = 0.003, whereas IL-I0 levels in AGA multiparous patients and AGA nulliparous patients were equivalent, p = 0.65. Comment
In this report we found increased midtrimester amniotic fluid IL-I0 levels to be associated with poor subsequent fetal growth. The measurement of cytokines in amniotic fluid at mid trimester may provide a window into processes regulating placental growth and invasion. Placentation typically culminates during the midtrimester and is often inadequate in pregnancies complicated by poor fetal growth." The growth-promoting cytokine CSF -1 has been shown to be positively correlated with birth weight, whereas levels of the inflammatory cytokine TNF -ex have a negative association. 3 • 5 IL-I0 is a recently described cytokine with potent immunosuppressive activity.g, 10 Originally identified on the basis of its functional ability to inhibit cytokine synthesis in an in vitro culture of antigen-stimulated mouse T-cell clones, it was originally known as cytokine synthesis inhibitor factor. IL-lO is produced as part of
July 1994 Am J Obstet Gynecol
the inflammatory response, but levels tend to lag temporarily behind those of pro inflammatory cytokines such as TNF -ex. As such, IL-I0 is hypothesized to have an autoregulatory function through this negative feedback loop. \I Our findings regarding IL-I0 and fetal growth help define the role of immunosuppression in the local intrauterine environment in pregnancy in relation to fetal growth. Initially we suspected that SGA gestations might be characterized by inadequate immunosuppression, with low IL-I0 levels. This hypothesis seemed likely on the basis of our earlier findings of increased TNF -ex levels in these pregnancies. Instead, we found increased levels of both IL-I0 and TNF -ex in the SGA gestations. This correlation suggests a common stimulus and tissue source for IL-l 0 and TNF -ex and indicates that these two cytokines are likely produced as part of the same inflammatory response. The amniotic fluid compartment most likely integrates quantities of these two cytokines produced at different temporal points. Although amniotic fluid is known to possess immunosuppressive properties, this activity has not been previously linked with a well-characterized biologic messenger. 12 Transforming growth factor-J3, another immunosuppressive cytokine, has been identified in murine amniotic fluid, and the corresponding messenger ribonucleic acid and protein have been identified in the human placenta. 13- 15 However, we have been unable to detect measurable levels of transforming growth factor-J3 in human amniotic fluid (Heyborne KD. Unpublished observations). Tissue sources of amniotic fluid IL-I0 are currently unknown and are the subject of continuing investigations. IL-l 0 may be produced by macrophages or other inflammatory cells at the maternal-fetal interface. Other tissue sources (trophoblast, decidua, amnion, etc.) are also possible. The stimulus for the production ofIL-l 0 is likewise unknown. Increased levels were highly associated with nulliparity in the SGA group but not in the AGA group or the study population as a whole. Perhaps some nulliparous women, in response to the primary antigenic challenge of a first pregnancy, mount an enhanced immune response that is associated with increased levels of IL-I0 and faulty placentation. Other stimuli of enhanced IL-l 0 production unrelated to parity are also possible, including infectious processes or disordered cytokine regulation caused by other factors. IL-I0 may have both physiologic and pathophysiologic roles in human pregnancy. In addition to inhibiting cytokine synthesis by macrophages and T lymphocytes, IL-I0 down-regulates expression of major histocompatibility complex molecules and is permissive for tumor growth in vitro. 16, 17 Such activities may suppress the maternal immune response and promote trophoblast growth. The high amniotic fluid IL-I0 levels
Volume 171, Number 1 Am J Obstet Gynecol
observed in the SGA pregnancies may have pathologic effects. For example, IL-lO inhibitis elaboration of nitric oxide (by inhibiting nitric oxide synthetase production),18 and nitric oxide is felt to be an important vasodilator of the fetal-placental circulation. 19. 20 Elevated levels of IL-l 0 measured in the SGA pregnancies may compromise fetal-placental blood flow, resulting in poor fetal and placental growth. Alternatively, IL-I0 may inhibit the release of trophic cytokines such as CSF-1, thereby interfering with placental growth. Levels of amniotic fluid cytokines have also been measured and reported in association with intraamniotic infection."1 In contrast to levels of cytokines in association with infection, cytokine levels in secondtrimester specimens from SGA gestations are markedly reduced, although IL-6 in some of our samples was in a range consistent with intraamniotic infection. Additionally, essentially all cytokines measured to date in association with intraamniotic infection have been markedly elevated. By contrast, cytokine levels in association with SGA may be elevated (IL-I0, TNF-a), reduced (CSF-l), or unchanged (lL-6), suggesting more subtle and intricate regulation of cytokines in association with SGA birth. In summary, we report here for the first time the presence of IL-I0 in amniotic fluid and show that increased IL-I0 levels at mid trimester are associated with subsequent fetal growth impairment. We have suggested possible mechanisms leading to increased IL-I0 production and potential effects of the increased levels in mediating SGA birth. Our data continue to support the role of abnormal immune activation, as opposed to deficient immune suppression, in mediating impaired fetal growth. REFERENCES 1. Athanassakis I, Bleckley RC, Paetkau V, Guilbert L, Barr P, Wegmann T. The immunostimulatory effect ofT cells and T cell lymphokines on murine fetally derived placental cells. J Immunol 1987;138:37-44. 2. Mowbray J, Underwood J, Jalaei G. Immunologic processes of abortion. In: Chaouat G, ed. Immunology of pregnancy. Boca Raton, Florida: CRC Press, 1993. 3. Peng TCC, Kauma SW, Turner TT. Second trimester amniotic fluid colony stimulating factor-l (CSF-l) levels in appropriate (AGA) and small for gestational age (SGA) fetuses. In: Proceedings of the thirty-ninth annual meeting of the Society for Gynecologic Investigation, San Antonio, Texas, March 18-21, 1992. San Antonio: Society for Gynecologic Investigation, 1992:140. 4. Pollard Jw, Hunt JS, Wiktor-Jedrzejczak W, Stadley ER. A pregnancy defect in the osteopetrotic (oP/op) mouse demonstrates the requirement for CSF-l in female fertility. Dev BioI 1991;148:273-83.
Heyborne et al.
59
5. Heyborne K, Witkin SS, McGregor JA. Tumor necrosis factor-a in midtrimester amniotic fluid is associated with impaired intrauterine fetal growth. AM J OBSTET GYNECOL 1992;167:920-5. 6. Battaglia FC, Lubchenco LO. A practical classification of newborn infants by weight and gestational age. J Pediatr 1967;71:159. 7. Abrams JS, Roncarolo MG, Yssel H, Andersson U, Gleich GJ, Silver J. Strategies of anti-cytokine monoclonal antibody development: immunoassay of IL-I0 and IL-5 in clinical samples. Immunol Rev 1992;127:5. 8. Sheppard BL, Bonnar J. An ultrastructural study of uteroplacental spiral arteries in hypertensive and normotensive pregnancy and fetal growth retardation. Br J Obstet Gynaecol 1981;88:695. 9. Florentino DF, Bond Mw, Mosmann TR. Two types of mouse helper T cells. IV. Th2 clones secrete a factor that inhibits cytokine production by Thl clones. J Exp Med 1989;170:2081. 10. Moore KW, VIelra P, Florentino DF, Trounstine ML, Khan T, Mosmann TR. Homology of cytokine synthesis inhibitory factor (IL-IO) to the Epstein-Barr virus gene BCRFI. Science 1990;248: 1230. 11. de Waal MR, Abrams J, Bennett B, Figdor CF, de Vries JE. Interleukin 10 (IL-IO) inhibits cytokine synthesis by human monocytes: an autoregulatory role ofIL-1O produced by monocytes. J Exp Med 1991;174:1209-20. 12. Anderson D, Yunis E. The elusive immunosuppressive factors of pregnancy. AmJ Reprod ImmunoI1985;9:91-2. 13. Graham CH, LysiakJJ, McCrae KR, Lala PK. Localization of transforming growth factor-beta at the human fetalmaternal interface: role in trophoblast growth and differentiation. BioI Reprod 1992;46:561-72. 14. Altman DJ, Schneider SL, Thompson DA, Cheng HL, Tomasi TB. A transforming growth factor J3 (TGF-J3-2) like immunosuppressive factor in amniotic fluid and localization ofTGF-J3 2 mRNA in the pregnant uterus. J Exp Med 1990;172:1391-401.' 15. Dungy L, Siddiqi TA, Khan S. Transforming growth factor-J31 expression during placental development. AM J OBSTET GYNECOL 1991;165:853-7. 16. de Waal MR, Haanen J, Spits H, et al. Interleukin 10 (IL-IO) and viral IL-I0 strongly reduce antigen-specific human T cell proliferation by diminishing the antigenpresenting capacity of monocytes via downregulation of class II major histocompatibility complex expression. J Exp Med 1991;174:915-24. 17. Bogdan C, Vodovotz Y, Nathan C. Macrophage deactivation by interleukin 10. J Exp Med 1991;174:1549-55. 18. Cunha FQ, Moncada S, Liew FY. Interleukin-l0 (IL-IO) inhibits the induction of nitric oxide synthase by interferon-gamma in murine macrophages. Biochem Biophys Res Commun 1992;182:1155-9. 19. Chang JK, Roman C, Heymann MA. Effect of endothelium-derived relaxing factor inhibition on the umbilicalplacental circulation in fetal lambs in utero. AM J OBSTET GYNECOL 1992;166:727-34. 20. Myatt L, Brewer AS, Langdon G, Brockman DE. Attenuation of the vasoconstrictor effects of thromboxane and endothelin by nitric oxide in the human fetal-placental circulation. AM J OBSTET GYNECOL 1992;166:224-30. 21. Gibbs RS, Romero R, Hillier SL, Eschenbach DA, Sweet RL. A review of premature birth and subclinical infection. AMJ OBSTET GYNECOL 1992;166:1515-28.
Key words: Low birth weight, immunology, interleukin-l0, intrauterine growth retardation
During placental invasion and growth the maternal immune system is challenged by fetal antigens present on trophoblasts. Current concepts suggest that the ensuing maternal immune response is complex and balanced. Some response to fetal antigens may enhance placental growth and development, whereas an abnormalone may lead to pregnancy loss or other perinatal complications. 1. 2 In recent years a group of protein messengers secreted by cells of the immune system and other tissues, referred to as cytokines, has been described. Cytokines may be thought of as pro inflammatory, immunosuppressive, or growth promoting. A balance between these
From the Department of Obstetrics and Gynecology, University of Colorado Health Sciences Center: Reproductive Genetics/ the Department of Obstetrics and Gynecology, Cornell University Medical Center,' and DNAX Research Institute! Supported by the Academic Enrichment Fund of the Department of Obstetrics and Gynecology of the University of Colorado Health Sciences Center, Denver. Presented at the Fortieth Annual Meeting of the Society for Gynecologic Investigation, Toronto, Ontario, Canada, March 31-April 3, 1993. Received for publication November 29, 1993; revised December 16, 1993; accepted December 28, 1993. Reprint requests: Kent D. Heyborne, MD, Department of Obstetrics and Gynecology, B-198, University of Colorado Health Sciences Center, Denver, CO 80262. Copyright © 1994 by Mosby-Year Book, Inc. 0002-9378/94 $3.00 + 0 6/1/54137
three cytokine activIties may influence placental and fetal growth. Second-trimester amniotic fluid levels of the cytokine colony-stimulating factor-l (CSF -1), a growth factor, are positively correlated with subsequent fetal growth." In the mouse maternal or fetal elaboration of CSF-l is required for successful pregnancy.4 Recently we determined that in human pregnancy increased midtrimester amniotic fluid levels of the proinflammatory cytokine tumor necrosis factor-a (TNF-a) are linked with poor fetal growth. 5 In this case-control study we investigated the relationship of other cytokines to fetal growth. To this end, we measured levels of second-trimester amniotic fluid interleukin-l0 (IL-I0, cytokine synthesis inhibitory factor), a potent immunosuppressive cytokine, in appropriate-for-gestational-age (AGA) and small-for-gestational-age (SGA) gestations. In addition, we measured levels of interleukin-6 (IL-6), which, like TNF-a, IS an inflammatory cytokine, in the same specimens.
Material and methods Amniotic fluid specimens were collected from January 1989 through September 1990 by midtrimester amniocentesis in a Denver private-practice genetics clinic where all patients had private health insurance. At amniocentesis demographic and medical information was obtained by genetic counselors. No patient in the study had a history of maternal immune disease;
55
56
July 1994
Heyborne et al.
Am J Obstet Gynecol
Table I. Maternal characteristics and indications for amniocentesis Maternal characteristics
Group 1 (SGA, n = 24)
(control, n = 44)
Significance
Age (yr) Gravidity Parity Percent previous spontaneous abortion
35.1 2.4 0.5 60
35.7 2.8 1.0 20
White Black Hispanic Other Smoker (> 0.5 pack/day) Gestational age at amniocentesis (wk) Indication for amniocentesis Maternal age High maternal serum a-fetoprotein Low maternal serum a-fetoprotein Other
20 1 1
41 1 1
NS* NSt p = 0.007t P = 0.003t NS:::
3 16.3
16.4
Race
Group 2
o
2
18 4 1 1
33
NS::: NSt
NSt
o
4 4
NS, Not significant. Age, gravidity, parity, gestational age at amniocentesis, and amniotic fluid a-fetoprotein are expressed as means. Other characteristics are expressed as numbers of patients. Percentage of previous pregnancies ending in miscarriage was calculated for each patient as follows: [Number of miscarriages -;- (Number of previous pregnancies - Number of elective terminations )]. *By Student t test. tBy Wilcoxon rank-sum test. tBy Fisher's exact test.
Table II. Pregnancy outcomes by group Pregnancy outcome Gestational age at delivery (wk) Infant sex (M/F ratio) Birth weight (gm) Birth weight percentile Ponderal index (Birth weight x 100)/Length3 ) Cesarean section Stillbirth Neonatal death Preterm delivery ( < 37 wk)
Group 1 (SGA, n = 24)
(control, n = 44)
Group 2
35.9 10:14 1812 4.3 2.16
39.6 22:20 3490 69.7 2.47
0.02* NSt p < 0.01* P < 0.01* P < 0.01*
14 1 2 12
15
p = 0.03ot
o o o
Significance
p<
NSt NSt p < O.Olt
NS, Not significant. The groups were statistically different as indicated. *By Wilcoxon rank-sum test. tBy Fisher's exact test.
one patient in the SGA group had chronic hypertension. All patients had an ultrasonographic examination at amniocentesis, confirming gestational age as determined by last menstrual period (within 1 week) and excluding early evidence of growth restriction. Amniotic fluid specimens were centrifuged (500 revolutions/min for 10 minutes) to remove cellular components for karyotype determination, and supernatants were frozen at - 20° C for later cytokine assay. At the expected date of delivery follow-up questionnaires requesting date of birth, birth weight and length, infant sex, route of delivery, and occurrence of stillbirth or neonatal death were sent to all patients. Responses were obtained from 2580 patients, representing an 80% response rate, and study patients were selected from this cohort.
Patient information was stored on a computer database (Revelation Technologies, New York) for later retrieval. From this database 24 specimens were identified as coming from gestations resulting in SGA births (birth weight < 10% for gestational age as determined by the Colorado Intrauterine Growth Curve).6 Less than 1% of patients from this private practice population fulfilled this definition of SGA, indicating that the pregnancies studied suffered from severe growth retardation relative to their demographic group. Forty-two patients with AGA term infants were randomly selected according to patient number by computer as controls. All specimens used in the study yielded normal karyotypic analysis. Immunoreactive IL-IO and IL-6 levels were deter-
Heyborne et al.
Volume 171, Number 1 Am J Obstet Gynecol
57
80r---------------------------------------~
<40
40-100
100-200
200-300
>300
40r-----------------------------------------,
30
10
o
<40
40-100
100-200
200-300
>300
Amniotic fluid IL-10, pg/ml Fig. 1. Frequency distribution of amniotic fluid IL-IO in AGA (toP) and SGA (bottom) gestations,
mined by immunoenzymetric assay with paired rat monoclonal antibodies for each cytokine, as previously described. 7 Amniotic fluid samples were clarified by centrifugation through a 0.45 J.Lm cellulose acetate Spin-X centrifuge filter unit (Costar, Cambridge, Mass.) and diluted 1: 1 with RPMI 1640 with 10% fetal bovine serum containing 50 J.Lg/ml each of rat immunoglobulin G 1 and immunoglobulin G 2 to inhibit false positivity that is related to rheumatoid factor or heterophilic antibody. Student t test, Wilcoxon rank sum analysis, Fisher's. exact test, and Spearman rank correlation were used for statistical comparisons.
Results
Table I and II compare the SGA and AGA groups regarding maternal characteristics, indications for amniocentesis, and pregnancy outcome. Both groups had a low incidence of clinical and epidemiologic risk factors traditionally associated with SGA birth (black race, smoking > 0.5 pack per day, or elevated maternal serum a-fetoprotein). Only six patients in the SGA group had at least one of the above risk factors, whereas two patients in the AGA group had such risk factors. The groups did not differ in maternal factors and amniocentesis indications, with the exception of parity and occurrence of previous spontaneous abortion. Birth
58
Heyborne et al.
weight, birth weight percentile, delivery gestational age, and ponderal index were significantly lower in the SGA group, and preterm delivery and cesarean section were more common in the SGA group. IL·I0 levels and fetal growth. Fig. 1 illustrates the frequency distribution of amniotic fluid IL-I0 levels in SGA and AGA pregnancies. Levels of IL-I0 in SGA pregnancies (median 78, range < 40 to 832 pg/ml) were elevated relative to those in AGA pregnancies (median < 40, range < 40 to 476 pg/ml), p = 0.02 by two-tailed Wilcoxon rank-sum test. With a threshold of 40 pg/ml defining an abnormal test, elevated IL-I0 levels identified SGA gestations with a sensitivity of 63% and a specificity of 74%. IL-6 was detected in all samples (median 548, range 18 to 10,198 pg/ml) but did not correlate with fetal growth, p > 0.05. TNF-cdevels were previously measured in a subset of these samples (24 SGA and 24 AGA specimens). IL-lO levels correlated with previously reported TNF -ex levels (Spearman correlation coefficient 0.48, p < 0.001). Clinical and epidemiologic risk factors and amniotic fluid TNF -ex have sensitivities of 25% and 48% and specificities of 95% and 83%, respectively, in this study group. IL·I0 levels and nulliparity. SGA patients were noted to have lower panty (0.5 vs 1.0, P = 0.007) and a higher percentage of previous pregnancies ending in miscarriage (60% vs 20%, P = 0.003) than the randomly selected AGA control patients. We found no relationship between IL-I0 levels and parity or the occurrence of a previous miscarriage when the study population as a whole was analyzed. Interestingly, in a subgroup analysis of SGA patients, nulliparity was strongly associated with increased IL-I0 levels, p = 0.003, whereas IL-I0 levels in AGA multiparous patients and AGA nulliparous patients were equivalent, p = 0.65. Comment
In this report we found increased midtrimester amniotic fluid IL-I0 levels to be associated with poor subsequent fetal growth. The measurement of cytokines in amniotic fluid at mid trimester may provide a window into processes regulating placental growth and invasion. Placentation typically culminates during the midtrimester and is often inadequate in pregnancies complicated by poor fetal growth." The growth-promoting cytokine CSF -1 has been shown to be positively correlated with birth weight, whereas levels of the inflammatory cytokine TNF -ex have a negative association. 3 • 5 IL-I0 is a recently described cytokine with potent immunosuppressive activity.g, 10 Originally identified on the basis of its functional ability to inhibit cytokine synthesis in an in vitro culture of antigen-stimulated mouse T-cell clones, it was originally known as cytokine synthesis inhibitor factor. IL-lO is produced as part of
July 1994 Am J Obstet Gynecol
the inflammatory response, but levels tend to lag temporarily behind those of pro inflammatory cytokines such as TNF -ex. As such, IL-I0 is hypothesized to have an autoregulatory function through this negative feedback loop. \I Our findings regarding IL-I0 and fetal growth help define the role of immunosuppression in the local intrauterine environment in pregnancy in relation to fetal growth. Initially we suspected that SGA gestations might be characterized by inadequate immunosuppression, with low IL-I0 levels. This hypothesis seemed likely on the basis of our earlier findings of increased TNF -ex levels in these pregnancies. Instead, we found increased levels of both IL-I0 and TNF -ex in the SGA gestations. This correlation suggests a common stimulus and tissue source for IL-l 0 and TNF -ex and indicates that these two cytokines are likely produced as part of the same inflammatory response. The amniotic fluid compartment most likely integrates quantities of these two cytokines produced at different temporal points. Although amniotic fluid is known to possess immunosuppressive properties, this activity has not been previously linked with a well-characterized biologic messenger. 12 Transforming growth factor-J3, another immunosuppressive cytokine, has been identified in murine amniotic fluid, and the corresponding messenger ribonucleic acid and protein have been identified in the human placenta. 13- 15 However, we have been unable to detect measurable levels of transforming growth factor-J3 in human amniotic fluid (Heyborne KD. Unpublished observations). Tissue sources of amniotic fluid IL-I0 are currently unknown and are the subject of continuing investigations. IL-l 0 may be produced by macrophages or other inflammatory cells at the maternal-fetal interface. Other tissue sources (trophoblast, decidua, amnion, etc.) are also possible. The stimulus for the production ofIL-l 0 is likewise unknown. Increased levels were highly associated with nulliparity in the SGA group but not in the AGA group or the study population as a whole. Perhaps some nulliparous women, in response to the primary antigenic challenge of a first pregnancy, mount an enhanced immune response that is associated with increased levels of IL-I0 and faulty placentation. Other stimuli of enhanced IL-l 0 production unrelated to parity are also possible, including infectious processes or disordered cytokine regulation caused by other factors. IL-I0 may have both physiologic and pathophysiologic roles in human pregnancy. In addition to inhibiting cytokine synthesis by macrophages and T lymphocytes, IL-I0 down-regulates expression of major histocompatibility complex molecules and is permissive for tumor growth in vitro. 16, 17 Such activities may suppress the maternal immune response and promote trophoblast growth. The high amniotic fluid IL-I0 levels
Volume 171, Number 1 Am J Obstet Gynecol
observed in the SGA pregnancies may have pathologic effects. For example, IL-lO inhibitis elaboration of nitric oxide (by inhibiting nitric oxide synthetase production),18 and nitric oxide is felt to be an important vasodilator of the fetal-placental circulation. 19. 20 Elevated levels of IL-l 0 measured in the SGA pregnancies may compromise fetal-placental blood flow, resulting in poor fetal and placental growth. Alternatively, IL-I0 may inhibit the release of trophic cytokines such as CSF-1, thereby interfering with placental growth. Levels of amniotic fluid cytokines have also been measured and reported in association with intraamniotic infection."1 In contrast to levels of cytokines in association with infection, cytokine levels in secondtrimester specimens from SGA gestations are markedly reduced, although IL-6 in some of our samples was in a range consistent with intraamniotic infection. Additionally, essentially all cytokines measured to date in association with intraamniotic infection have been markedly elevated. By contrast, cytokine levels in association with SGA may be elevated (IL-I0, TNF-a), reduced (CSF-l), or unchanged (lL-6), suggesting more subtle and intricate regulation of cytokines in association with SGA birth. In summary, we report here for the first time the presence of IL-I0 in amniotic fluid and show that increased IL-I0 levels at mid trimester are associated with subsequent fetal growth impairment. We have suggested possible mechanisms leading to increased IL-I0 production and potential effects of the increased levels in mediating SGA birth. Our data continue to support the role of abnormal immune activation, as opposed to deficient immune suppression, in mediating impaired fetal growth. REFERENCES 1. Athanassakis I, Bleckley RC, Paetkau V, Guilbert L, Barr P, Wegmann T. The immunostimulatory effect ofT cells and T cell lymphokines on murine fetally derived placental cells. J Immunol 1987;138:37-44. 2. Mowbray J, Underwood J, Jalaei G. Immunologic processes of abortion. In: Chaouat G, ed. Immunology of pregnancy. Boca Raton, Florida: CRC Press, 1993. 3. Peng TCC, Kauma SW, Turner TT. Second trimester amniotic fluid colony stimulating factor-l (CSF-l) levels in appropriate (AGA) and small for gestational age (SGA) fetuses. In: Proceedings of the thirty-ninth annual meeting of the Society for Gynecologic Investigation, San Antonio, Texas, March 18-21, 1992. San Antonio: Society for Gynecologic Investigation, 1992:140. 4. Pollard Jw, Hunt JS, Wiktor-Jedrzejczak W, Stadley ER. A pregnancy defect in the osteopetrotic (oP/op) mouse demonstrates the requirement for CSF-l in female fertility. Dev BioI 1991;148:273-83.
Heyborne et al.
59
5. Heyborne K, Witkin SS, McGregor JA. Tumor necrosis factor-a in midtrimester amniotic fluid is associated with impaired intrauterine fetal growth. AM J OBSTET GYNECOL 1992;167:920-5. 6. Battaglia FC, Lubchenco LO. A practical classification of newborn infants by weight and gestational age. J Pediatr 1967;71:159. 7. Abrams JS, Roncarolo MG, Yssel H, Andersson U, Gleich GJ, Silver J. Strategies of anti-cytokine monoclonal antibody development: immunoassay of IL-I0 and IL-5 in clinical samples. Immunol Rev 1992;127:5. 8. Sheppard BL, Bonnar J. An ultrastructural study of uteroplacental spiral arteries in hypertensive and normotensive pregnancy and fetal growth retardation. Br J Obstet Gynaecol 1981;88:695. 9. Florentino DF, Bond Mw, Mosmann TR. Two types of mouse helper T cells. IV. Th2 clones secrete a factor that inhibits cytokine production by Thl clones. J Exp Med 1989;170:2081. 10. Moore KW, VIelra P, Florentino DF, Trounstine ML, Khan T, Mosmann TR. Homology of cytokine synthesis inhibitory factor (IL-IO) to the Epstein-Barr virus gene BCRFI. Science 1990;248: 1230. 11. de Waal MR, Abrams J, Bennett B, Figdor CF, de Vries JE. Interleukin 10 (IL-IO) inhibits cytokine synthesis by human monocytes: an autoregulatory role ofIL-1O produced by monocytes. J Exp Med 1991;174:1209-20. 12. Anderson D, Yunis E. The elusive immunosuppressive factors of pregnancy. AmJ Reprod ImmunoI1985;9:91-2. 13. Graham CH, LysiakJJ, McCrae KR, Lala PK. Localization of transforming growth factor-beta at the human fetalmaternal interface: role in trophoblast growth and differentiation. BioI Reprod 1992;46:561-72. 14. Altman DJ, Schneider SL, Thompson DA, Cheng HL, Tomasi TB. A transforming growth factor J3 (TGF-J3-2) like immunosuppressive factor in amniotic fluid and localization ofTGF-J3 2 mRNA in the pregnant uterus. J Exp Med 1990;172:1391-401.' 15. Dungy L, Siddiqi TA, Khan S. Transforming growth factor-J31 expression during placental development. AM J OBSTET GYNECOL 1991;165:853-7. 16. de Waal MR, Haanen J, Spits H, et al. Interleukin 10 (IL-IO) and viral IL-I0 strongly reduce antigen-specific human T cell proliferation by diminishing the antigenpresenting capacity of monocytes via downregulation of class II major histocompatibility complex expression. J Exp Med 1991;174:915-24. 17. Bogdan C, Vodovotz Y, Nathan C. Macrophage deactivation by interleukin 10. J Exp Med 1991;174:1549-55. 18. Cunha FQ, Moncada S, Liew FY. Interleukin-l0 (IL-IO) inhibits the induction of nitric oxide synthase by interferon-gamma in murine macrophages. Biochem Biophys Res Commun 1992;182:1155-9. 19. Chang JK, Roman C, Heymann MA. Effect of endothelium-derived relaxing factor inhibition on the umbilicalplacental circulation in fetal lambs in utero. AM J OBSTET GYNECOL 1992;166:727-34. 20. Myatt L, Brewer AS, Langdon G, Brockman DE. Attenuation of the vasoconstrictor effects of thromboxane and endothelin by nitric oxide in the human fetal-placental circulation. AM J OBSTET GYNECOL 1992;166:224-30. 21. Gibbs RS, Romero R, Hillier SL, Eschenbach DA, Sweet RL. A review of premature birth and subclinical infection. AMJ OBSTET GYNECOL 1992;166:1515-28.