- Email: [email protected]
Do placental histologic findings of chorion-decidual hemorrhage or inflammation in spontaneous preterm birth influence outcomes in the subsequent pregnancy?
Placenta 35 (2014) 58e63
Contents lists available at ScienceDirect
Placenta journal homepage: www.elsevier.com/locate/placenta
Do placental histologic findings of chorion-decidual hemorrhage or inflammation in spontaneous preterm birth influence outcomes in the subsequent pregnancy? D.N. Hackney a, *, R. Tirumala a, L.J. Salamone a, R.K. Miller a, P.J. Katzman b a
Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, University of Rochester School of Medicine, 601 Elmwood Ave, Rochester, NY 14642, USA Department of Pathology, University of Rochester School of Medicine, 601 Elmwood Ave, Rochester, NY 14642, USA
b
a r t i c l e i n f o
a b s t r a c t
Article history: Accepted 2 November 2013
Introduction: Spontaneous preterm birth (SPTB) is the common endpoint of different underlying etiologies, including chorion-decidual bleeding and inflammation. However, specific histologic findings from a prior pregnancy do not always inform clinical management in subsequent pregnancies secondary to few prior studies having evaluated the relationship between prior pregnancy pathology and subsequent outcomes in patients with SPTB. Methods: Included subjects had: 1) a SPTB with available placental pathology and 2) a subsequent consecutive delivery at >20 weeks gestational age at our institution. For included subjects archived placenta and membrane paraffin blocks from the index SPTB were cut, stained with Prussian Blue and evaluated by a perinatal pathologist for the presence of hemosiderin. The association between histologic findings and subsequent pregnancy outcomes were evaluated through logistic and linear regression. Results: A total of 131 subjects were included, of whom 39.7% had a recurrent SPTB. Funisitis at the time of preterm delivery significantly increased the risk of early (<34 weeks) recurrent preterm birth (OR 3.38, p ¼ 0.016), though this may have been confounded by gestational age at delivery. Several histologic features were significantly associated with reductions in birth weight in the subsequent pregnancies, even if they did not increase the risk of recurrent preterm birth. Discussion: The presence of chorion-decidual bleeding or inflammation in a prior pregnancy can signal an increased risk in a future pregnancy beyond the recurrent risk of SPTB itself. Conclusions: Placental histologic findings after SPTB maybe associated with differences in birth weight in a subsequent pregnancy. Ó 2013 Elsevier Ltd. All rights reserved.
Keywords: Preterm birth Chorion-decidual hemorrhage Hemosiderin Placental pathology Chorioamnioitis Intrauterine growth restriction
1. Introduction The last several years have seen significant advancements in the management of patients at risk for spontaneous preterm birth (SPTB), including the use of 17-a-hydroxyprogesterone caproate [1] and cervical length screening with possible cerclage [2]. Many of these treatments have proved most beneficial in patients whose * Corresponding author. Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, University Hospitals Case Medical Center/Case-Western Reserve University, MacDonald Women’s Hospital, 11100 Euclid Avenue, MAC 5034, Cleveland, OH 44106, USA. Tel.: þ1 216 844 3787; fax: þ1 216 201 5063. E-mail addresses: [email protected] (D.N. Hackney), Radhika. [email protected] (R. Tirumala), [email protected]. edu (L.J. Salamone), [email protected] (R.K. Miller), Philip_ [email protected] (P.J. Katzman). 0143-4004/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.placenta.2013.11.001
specific SPTB risk factor is the occurrence of an SPTB in a prior pregnancy, leading to increased clinical and academic interest in these patient populations [3]. Though all cases of SPTB share the common endpoint of the delivery of a premature neonate, a variety of underlying etiological pathways are present for the preterm deliveries of different women [4,5]. These would include excessive uterine stretch, infection and inflammation and chorion-decidual hemorrhage, though in many cases a specific etiology cannot be identified. Studies of patients with prior preterm birth will often characterize the prior deliveries by gestational age at delivery or the presence or absence of premature membrane rupture, though rarely by the known or presumed underlying etiology of the prior delivery. As a result, patients are often included together whose prior preterm births may have resulted from very different fundamental processes.
D.N. Hackney et al. / Placenta 35 (2014) 58e63
Placental pathology after a spontaneous preterm birth is a powerful tool for elucidating the underlying etiology of the delivery which has just occurred. Specifically there is a well-characterized association between SPTB and the histologic presence of chorioamnionitis [6], with funisitis and vasculitis being particularly associated with deliveries at increasingly early gestational ages. Prior chorion-decidual hemorrhage as a potential cause of SPTB is more difficult to ascertain on standard Hemotoxylin/Eosin (H þ E) staining, though hemosiderin deposition can be identified through the use of Prussian Blue staining. Prior studies using Prussian Blue staining have also demonstrated a strong association between hemosiderin deposition and SPTB [7]. Though placental evaluation after SPTB is highly informative with regards to the pregnancy which just concluded, its overall clinical usefulness can be limited by the fact that the affected pregnancy is completed at the time of analysis. A potentially significant clinical value for placental pathology would be if the histologic findings from a prior pregnancy were able to inform clinical management in a subsequent pregnancy in the same patient. However, very few prior studies have evaluated the relationship between histologic findings in a prior pregnancy and clinical outcomes in subsequent gestations [8]. This study evaluated the clinical outcomes of pregnancies which occurred after a prior SPTB with regards to the histologic findings from the index delivery. Specifically the association between subsequent pregnancy outcomes and chorion-decidual bleeding as a presumed underlying etiology of a prior SPTB as evaluated with Prussian Blue staining for hemosiderin has not previously been reported. 2. Methods 2.1. Subject identification and clinical variable abstraction We conducted a retrospective cohort of patients who had a SPTB at the University of Rochester Medical Center between 2004 and 2009 followed by a consecutive pregnancy which had also occurred at the same institution. Potential subjects were identified through a maternally linked birth certificate database as described in a prior publication of birth certificate accuracy [9]. Patients who met the clinical inclusion criteria were then cross-referenced against an electronic pathology database to determine the availability of archived paraffin blocks of the placenta parenchyma and membrane roll. Clinical inclusion criteria were an “index” SPTB between 20 and 36 6/7 weeks estimated gestational age (EGA) and a
59
“subsequent” delivery at >20 weeks EGA with both deliveries having occurred at the above institution and no record of an intervening delivery at an outside institution between those two. The “index” preterm birth did not necessarily have to also be the patient’s first pregnancy. The “subsequent” delivery could be either term or preterm, with the outcomes of that pregnancy being the outcome of interest for the study. Spontaneous preterm births were defined as having occurred secondary to preterm labor or preterm premature rupture of membranes. “Index” preterm deliveries which were performed iatrogenically secondary to pre-eclampsia, intrauterine growth restriction (IUGR) or another maternal or fetal indication were excluded. Exclusion criteria were multifetal gestation, thrombophilia, anticoagulation, major congenital anomaly, structural uterine anomaly or substance abuse (other than marijuana) in either pregnancy as determined by manual review of the prenatal and delivery records of both pregnancies. The following clinical variables were abstracted by manual review of the prenatal and delivery records of both pregnancies along with the interpregnancy (delivery to the last menstrual period of the subsequent pregnancy) interval: EGA at delivery, birth weight, tobacco use, maternal age, maternal race, gravidity and maternal pre-pregnancy BMI. Gestational age specific birth weight percentiles were determined using the Finger Lakes Region birth weight curves, which are based on the local birth record registry within the Finger Lakes delivery region. The study was approved by the University of Rochester Research Subjects Review Board. Clinical characteristics of the index pregnancies for pregnancies of placentas studied is presented in Table 1 [10]. 2.2. Histologic evaluation All histologic evaluations were performed by a single subspecialty trained perinatal pathologist (PJK) who was blinded to all of the clinical variables including outcomes. Original H þ E slides and archived paraffin embedded placental parenchyma and membrane roll blocks from the index preterm birth were retrieved for all included subjects. H þ E slides were reviewed to determine two paraffin blocks that would be most appropriate for the evaluation of hemosiderin by Prussian Blue staining. The selected blocks, one block with at least one section of the maternal surface and one block with the membrane roll, were then retrieved and re-cut. After xylene incubation and hydration via graded alcohols, slides were incubated for 30 min in potassium ferrocyanide solution (1.0 g potassium ferrocyanide, 98 mL water and 2.0 mL of 10% Triton X-100 then mixed in a 1:1 ratio with 2% Hydrochloric acid), followed by Nuclear Red fast stain, dehydration and mounting. H þ E slides were scored using a binary scale for presence or absence of acute or chronic chorioamnionitis, acute funisitis, acute chorionic vasculitis, fetal vessel thrombosis, villous sclerosis, laminar necrosis of the membranes, intramembranous/ intradecidual thrombi, subchorionic thrombi, and intervillous thrombi. The maternal acute inflammatory and fetal acute inflammatory responses were further scored by stage (score 1-3) and grade (score 1-2) as detailed by an Amnionic Fluid Infection Nosology Committee study [11]. The Prussian Blue stains were scored for focality and intensity of staining in the basal plate on the maternal surface and in the fetal membranes. Focality was scored as either focal or multifocal and intensity was scored as either light or heavy. Stains were evaluated using 4, 10, and 20
Table 1 Clinical characteristics of included subjects. Parameter
Clinical characteristics of pregnancies for placentas studied
Gravidity Parity Gestational age (weeks) Maternal age (years) Race Ethnicity Prenatal medications Drugs Previous prenatal admission(s) Blood pressures <140/90 mm Hg Screened for diabetes Antibiotics in labor Beta strep status Antenatal steroids: Magnesium sulfate Anesthesia Cervical ripening agent Labor Delivery mode (N ¼ # patients)
Median ¼ 2 Median ¼ 1 Average ¼ 32 3/7 Average ¼ 25.9 Black 25%
25e75% ¼ 1e3 25e75% ¼ 0e2 SD ¼ 3.6 SD ¼ 5.9 White 56%
Iron ¼ 12.2% Cigarettes n ¼ 18.7% Yes, n ¼ 0 Yes, n ¼ 2 Yes ¼ 53.5% None, n ¼ 0 Positive Yes, n ¼ 74 Yes, n ¼ 61 Epidural: Unknown Prostaglandin E1 Yes ¼ 100% C-section, Repeat, no labor: n ¼ 0 Yes, n ¼ 0 Average ¼ 1993 Average¼ Female ¼ 41% Average¼
Albuterol: 10% Alcohol: Unknown No, n ¼ 0 No, n ¼ 129 No ¼ 16% Penicillin, n ¼ 0 Negative No, n ¼ 57 No, n ¼ 70 Narcotics: Unknown Prostaglandin E2 If yes, hours: Unknown C-section, Repeat, with labor: n ¼ unknown No, n ¼ 0 SD ¼ 711 SD¼ Male ¼ 59% SD¼
Maternal oxygen given at delivery? Birth weight (grams) Placental weight (grams) Baby’s sex Delivery to processing (min)
Range 1e9 Range 0e7 Range 16e38 Other 19% None (except vitamins): 76.3% Others: 0% (excluded) Diagnoses if yes: Unknown ¼ 0 Unknown ¼ 30.5% Other: (list) Unknown 100% If yes, week GA: Unknown Unknown ¼ 0 General Mechanical C-section, primary, no labor: n ¼ 0 Unknown ¼ 100% Unknown Unknown 0% Unknown
Unknown 0% Unknown 0% Unknown, n ¼ 100%
Unknown, n ¼ 100% Unknown ¼ 0 Other/none Other: Unknown: 100% C-section, primary, with labor: n ¼ unknown
60
D.N. Hackney et al. / Placenta 35 (2014) 58e63
objectives and focality was defined as one positive area in a 10 objective field. Multifocality was defined as several areas or one continuous area covering more than one 10 objective field. Light staining was defined as dot-like staining pattern in a 20 objective. Heavy staining was defined as staining of entire cells with a homogenous staining in a 10 objective. Representative photomicrographs are presented in Figs. 1 and 2. 2.3. Statistical analysis The association between individual histologic variables and the binary outcomes of recurrent preterm birth and SGA were evaluated with logistic regression with and without adjustment for potential clinical confounders. The association between individual histologic variables and the continuous variables of birth weight percentiles and birth weight adjusted for gestational age at delivery were evaluated through linear regression models with adjustments for potential clinical confounders. Clinical confounders included BMI, tobacco use, maternal ethnicity and interpregnancy interval. All analyses were performed on Stata 11 (College Station, TX, USA). A p-value of 0.05 was considered statistically significant.
3. Results Subjects (256) were initially identified as having had a singleton spontaneous preterm birth at the University of Rochester Medical Center followed by a subsequent singleton delivery at >20 weeks gestational age. Subjects (65 or 25.4%) were excluded secondary to the absence of archived placental samples. The majority (88%) of this group constituted patients with “late” (34e36 weeks) prior preterm births, with the median gestational age at delivery in the “index” pregnancy for excluded subjects being 36 2/7 weeks. Though not always explicitly stated in the delivery record, presumably in these cases the placenta was not sent to pathology at the discretion of the delivering provider secondary to the delivery having occurred near term. Of the remaining subjects with available placental specimens, 60 were excluded following medical record review for the following indications: major congenital fetal anomaly in 24, substance abuse in 24, thrombophilia in 9 and congenital maternal uterine anomalies in 3. Most cases of substance abuse were patients who had experienced spontaneous preterm labor in the context of cocaine intoxication. A final total of 131 subjects were included, the subsequent pregnancy outcome characteristics of which are provided in Table 2. 39.7% of included subjects had a recurrent SPTB. The index SPTB demonstrated chorioamnionitis in 42.7%, vasculitis or funisitis in 24.4%, funisitis alone in 22.1%, fetal membrane hemosiderin in 52.7%, basal plate hemosiderin in 26.7% and laminar necrosis in 14.5%. The association between histologic findings and gestational age in the index pregnancy is presented in Table 3. Many subjects had more than one histologic finding, with only 16.7% having no specific histologic findings. Hemosiderin staining was more common when thrombi were present in the membranes or decidua, though there was no association with thrombi in other locations. The following histologic findings occurred infrequently (<10%
incidence) and were thus not evaluated separately: chronic chorioamnionitis, “heavy” hemosiderin of the basal plate, fetal vessel thrombosis and villous sclerosis. Preterm birth outcomes according to index placental pathology are provided in Table 4. When adjusted for potential clinical confounders, the presence of funisitis in an index preterm birth was associated with a significant increase in subsequent preterm birth (aOR 3.46, p ¼ 0.02) and specifically early preterm deliveries (aOR 5.46, p ¼ 0.008). The combined group of funisitis and/or fetal vessel vasculitis also demonstrated a significant though smaller association with recurrent preterm birth as did stage 2 or 3 chorioamnionitis and preterm birth recurrence in adjusted analysis. The finding of multifocal hemosiderin in the fetal membranes was associated with a degreased odds of recurrent preterm birth, though only in the multivariate analysis and with moderate significance (p ¼ 0.041). No other histologic findings had an association with the risk of preterm birth recurrence, including markers of chorion-decidual bleeding (hemosiderin) or potential hypoxia (laminar necrosis). Patients with funisitis in the index pregnancy had delivered at earlier gestational ages in that pregnancy than subjects without this finding (median EGA 31 1/7 versus 34 1/7) and index gestational age was associated with subsequent gestational age when adjusted for the same confounders (p ¼ 0.007) When adjusted for gestational age in the index delivery, the association between funisitis and subsequent early preterm birth was no longer significant at a 0.05 level (aOR 2.5, p ¼ 0.08). Several histologic features were significantly associated with reductions in gestational age-adjusted birth weight and/or birth weight percentiles in the subsequent pregnancies as outlined in Table 5. Histologic lesions that were associated with decreased birth weight included some that had no association with preterm birth recurrence itself, including basal plate hemosiderin and chorioamnionitis. Unlike the risk of recurrent preterm birth there was no association in linear regression between gestational age in the index preterm birth and gestational age adjusted birth weight (p ¼ 0.89) or birth weight percentile (p ¼ 0.97) in the subsequent pregnancy.
4. Discussion Spontaneous preterm birth represents a common endpoint for a heterogeneous collection of underlying etiologies. For patients who have experienced a prior spontaneous preterm birth, outcomes in a subsequent pregnancy may be impacted by these underlying etiologies in addition to other known prognostic clinical variables. Specifically, the results of our study suggest that patients with prior SPTBs secondary to chorion-decidual hemorrhage or inflammation may be at an increased risk of decreased birth weight in a future pregnancy relative to patients who did not demonstrate these
Fig. 1. (A) Free membranes contain laminar necrosis (lower half of A) where there is heavy iron deposition seen on Prussian Blue stain (B) (original magnification, 200).
D.N. Hackney et al. / Placenta 35 (2014) 58e63
61
Fig. 2. (A) Free membranes contain intramembranous thrombus (asterisks) with hemosiderin that is positive for heavy iron deposition on Prussian Blue stain (B) (original magnification, 200).
histologic findings, and this association may be independent of the risk of recurrent preterm birth itself. For these patients, the histologic anomalies in their index preterm births could signify intrinsic or chorionic conditions, such as a predisposition to inflammation or suboptimal placental vascular development, which in a subsequent pregnancy could generate complications which are insufficient to lead to recurrent preterm birth though sufficient to impair growth. This would be particularly plausible with regards to subclinical bleeding events at the chorion-decidual interface. To our knowledge this is the first study to evaluate the association between chorion-decidual hemorrhage in a prior pregnancy and clinical outcomes in subsequent deliveries or between histologic findings in SPTB and subsequent pregnancy birth weight. Strengths of the study include the evaluations of all histology slides by a single, subspecialty trained perinatal pathologist who was blinded to the clinical variables. We also performed Prussian Blue staining on all archived cases, thus allowing us to evaluate subclinical hemosiderin deposition which may not have been evident on H þ E staining. Because all subjects had both of their deliveries at the same institution, we also had access to all complete delivery records. The primary limitation of our study was that it was retrospective and limited to patients for whom archived placental blocks were available for re-staining. Specifically, our study is not necessarily representative of all patients who have had a prior SPTB, as many patients with “late” (34e36 weeks) preterm births had not originally had their placentas submitted for histologic evaluation. We also had a very high incidence of tobacco use in our cohort, though this was adjusted for in multivariate analysis. Additionally, though our regression models included the relevant clinical variables which were available to us, we could only evaluate those which were present in the available records. 17-alpha hydroxyprogesterone caproate is commonly used for the prevention of recurrent preterm birth [1] and would have thus served as an interesting source of secondary analysis. However, only 6 subjects were identified as having used this medication in their subsequent
Table 2 Subsequent pregnancy outcome characteristics of study population. Inter-pregnancy interval: 16 (1e66) Estimated EGA subsequent pregnancy (weeks): 37 4/7 (20 0/7e41 6/7) Birth weight subsequent pregnancy (g): 2870 (710e4530) BMI subsequent pregnancy: 25.7 (18e47.5) % Recurrent preterm birth in subsequent pregnancy: 39.7% % Early (<34 weeks) preterm birth in subsequent pregnancy: 16% Data presented as median (range). Inter-pregnancy interval defined as delivery date of index to last menstrual period of subsequent pregnancy rounded to months. BMI ¼ pre-pregnancy body mass index. EGA ¼ estimated gestational age.
pregnancies, which was insufficient for meaningful statistical analysis. Also, as most patients delivered at term in their subsequent pregnancies and thus did not undergo placental evaluations in their subsequent pregnancies, we could not perform a meaningful evaluation of placental histology in subsequent pregnancies in the same patient, which would have been an interesting evaluation. Though the presence of placental hemosiderin correlates with preterm birth [7], it is not usually possible histologically to determine at what time in pregnancy the bleeding event occurred and thus if it had an actual temporal relationship with the preterm delivery. Placentas from patients experiencing recurrent SPTB are more likely to demonstrate inflammatory findings than those from patients who are experiencing a first SPTB [12,13], which would also potentially be consistent with prior inflammatory findings increasing the risk of SPTB recurrence. Finally, though we were able to demonstrate statistically significant and potentially clinically meaningful differences in birth weight with various histologic findings, we were underpowered to demonstrate specific differences in birth weight at <10%. The a priori primary outcome of our study was the recurrence of preterm birth in a subsequent pregnancy. In this regard a significant association was demonstrated with the presence of many of the inflammatory markers in the index preterm birth, including, in particular funisitis. However, it is possible that the association between funisitis and recurrent preterm birth was confounded by the gestational age at delivery in the index pregnancy as both funisitis was associated with an earlier gestational age in the index pregnancy and decreasing index pregnancy gestational age is associated
Table 3 Histologic findings and median gestational age at delivery in the index pregnancy. Histologic finding
Median GA at delivery
p
All subjects Chorioamnionitis Chorioamnionitis w/vasculitis and/or funisitis: Funisitis Chorioamnionitis Stage 2 and 3 only Chorioamnionitis Grade 2 only Hemosiderin in any location Hemosiderin in fetal membranes Hemosiderin in fetal membranes - Multifocal only Hemosiderin in fetal membranes - Heavy only Hemosiderin in the basal plate Hemosiderin in the basal plate - Multifocal only
33 0/7 30 7/8
Referent <0.001
30 30 28 28 32 32
<0.001 <0.001 <0.001 <0.001 0.94 0.76
1/7 3/7 6/7 4/7 3/7 4/7
32 5/7
0.5
33 1/7 31 3/7
0.17 0.03
32 0/7
0.67
GA ¼ gestational age. p values determined by linear regression.
62
D.N. Hackney et al. / Placenta 35 (2014) 58e63
Table 4 Associations between histologic findings and the subsequent pregnancy outcomes. Histologic finding (number)
% Outcome
Negative (22): %Recurrent PTB 36.4% % Early PTB 18.1% Chorioamnionitis (56): %Recurrent PTB 42.9% % Early PTB 21.4% Chorioamnionitis with vasculitis and/or funisitis (32): %Recurrent PTB 46.8% % Early PTB 28.1% Chorioamnionitis with funisitis (29) %Recurrent PTB 51.7% % Early PTB 31% Chorioamnionitis Stage 2 and 3 only (24) %Recurrent PTB 54.5% % Early PTB 29.2% Chorioamnionitis Grade 2 only (14) %Recurrent PTB 35.7% % Early PTB 14.2% Hemosiderin in any location (89): %Recurrent PTB 40.4% % Early PTB 13.5% Hemosiderin in the fetal membranes (69) %Recurrent PTB 37.7% % Early PTB 13% Hemosiderin in the fetal membranes e Multifocal only (42) %Recurrent PTB 30.9% % Early PTB 11.9% Hemosiderin in the fetal membranes e Heavy only (32) %Recurrent PTB 43.8% % Early PTB 12.5% Hemosiderin in the basal plate (35) %Recurrent PTB 51.5% % Early PTB 17.1% Hemosiderin in the basal plate e Multifocal only (14) %Recurrent PTB 35.7% % Early PTB 14.3% Laminar necrosis (19): %Recurrent PTB 36.8% % Early PTB 5.2% Intramembranous/intradecidual thrombi (24) %Recurrent PTB 25% % Early PTB 8.3% Subchorionic thrombi (15) %Recurrent PTB 40% % Early PTB 20% Intervillous thrombi (18) %Recurrent PTB 33.3% % Early PTB 16.7%
OR (95% CI)
p-value
Referent Referent
aOR* (95% CI)
p-value
Referent Referent
1.26 (0.62e1.55) 2.0 (0.78e5.14)
0.32 0.15
1.32 (0.6e2.92) 1.8 (0.65e4.94)
0.49 0.26
1.48 (0.67e3.3) 2.84 (1.70e7.55)
0.22 0.035
2.37 (0.88e6.36) 3.61 (1.12e11.7)
0.09 0.032
1.88 (0.82e4.32) 3.38 (1.25e9.09)
0.14 0.016
3.46 (1.2e9.9) 5.46 (1.57e19)
0.02 0.008
2.06 (0.84e50.4) 2.74 (0.96e7.77)
0.11 0.06
3.36 (1.04e10.8) 1.75 (0.47e6.51)
0.043 0.4
0.83 (0.26e2.62) 0.86 (0.18e4.15)
0.75 0.85
0.9 (0.21e3.92) 0.26 (0.02e2.58)
0.89 0.25
1.1 (0.52e2.34) 0.57 (2.2e1.49)
0.48 0.18
1.3 (0.55e3.09) 0.91 (0.31e2.62)
0.54 0.86
0.84 (0.42e1.69) 0.62 (0.24e1.6)
0.62 0.33
0.84 (0.39e1.85) 0.93 (0.33e2.57)
0.67 0.89
0.57 (0.26e1.25) 0.61 (0.21e1.81)
0.16 0.38
0.37 (0.14e0.96) 0.48 (0.12e1.85)
0.041 0.28
1.25 (0.56e2.8) 0.69 (0.21e2.22)
0.59 0.53
0.92 (0.35e2.47) 0.7 (0.18e2.77)
0.88 0.61
1.93 (0.88e4.23) 1.11 (0.4e3.15)
0.1 0.83
2.17 (0.91e5.18) 1.4 (0.45e4.33)
0.08 0.56
0.82 (0.26e2.62) 0.86 (0.18e4.15)
0.75 0.85
0.61 (0.15e2.34) 0.57 (0.06e5.17)
0.47 0.62
0.89 (0.32e2.37) 0.26 (0.03e2.02)
0.5 0.2
0.87 (0.29e2.62) 0.19 (0.02e1.63)
0.81 0.13
0.44 (0.16e1.2) 0.42 (0.9e1.94)
0.11 0.27
0.57 (0.19e1.64) 0.42 (0.1e2.53) 0.42
0.29
1.01 (0.34e3.04) 1.36 (0.35e5.31)
0.98 0.66
0.95 (0.25e3.61) 1.42 (0.26e7.57)
0.95 0.68
0.73 (0.26e2.08) 1.06 (0.28e4.02)
0.55 0.94
0.95 (0.3e2.99) 1.17 (0.28e4.89)
0.93 0.82
CI ¼ confidence interval, OR ¼ Odds Ratio, *aOR ¼ adjusted Odds Ratio (adjusted for interpregnancy interval, race, tobacco use and maternal BMI), PTB ¼ preterm birth (gestational age at delivery <34 weeks). Statistically significant (p < 0.05) results are in bold.
with increasing risks of recurrent preterm birth. Thus, the association is potentially complicated by multicolinearity or mediation. Specifically if inflammatory changes in the index pregnancy lead to earlier preterm births and these in turn increase the risk of subsequent preterm delivery, then gestational age in the index pregnancy could be a mediator and not necessarily a confounder. When an adjustment for gestational age at index SPTB was performed, however, a non-significant trend (p ¼ 0.08) towards an increased independent risk of recurrent preterm birth with funisitis still existed. Additionally, the association between inflammatory vascular lesions of the placenta and preterm birth recurrence in future pregnancies would be consistent with the prior work of Himes and Simhan [8] and Ghidini [12]. Of note, unlike the potential association with recurrent preterm birth, the gestational age at index delivery was not associated with differences in either birth weight percentile or gestational age adjusted birth weight in the subsequent pregnancies. An association was also demonstrated between hemosiderin in the fetal membranes and preterm birth. However, the degree of
association, while significant at an alpha of 0.05, was not large (p ¼ 0.041) and all other measures of hemosiderin staining were not associated with recurrent birth risk. Thus the potential exists that this singular result was a spurious consequence of multiple comparisons, though the potential for an association would be an interesting focus of future prospective studies. The association between histologic features from an index preterm delivery and decreased birth weight in a subsequent pregnancy is in our opinion both novel and interesting. Fetuses that deliver prematurely are known to have an overall decreased weight percentile [14,15] and it is plausible that the etiology for the decreased growth could persist into a subsequent pregnancy even if the preterm delivery itself does not recur. These could include predispositions to subclinical intrauterine inflammatory events or chorion-decidual hemorrhage. Additionally, Espinoza et al. [16] have demonstrated that patients who experience an episode of preterm labor which resolves and allows that pregnancy to continue to term have an increased risk of delivering an SGA infant. Thus many patients with prior SPTB may have had preterm labor
D.N. Hackney et al. / Placenta 35 (2014) 58e63
63
Table 5 Associations between histologic findings in SPTB and birth weight in subsequent pregnancies. Histologic finding (number)
Gestational-age adjusted BW Coefficienta (95% CI)
BW percentile p-value
Coefficient* (95% CI)
p-value
0.006
L13.3% (L22.8% to L3.8%)
0.006
0.024
11.6% (23.9% to 0.6%)
0.063
0.002
L17.1% (L29.9% to L4.3%)
0.009
0.1
12.5% (26.1% to 1.2%)
0.07
0.043
15.3% (32.7% to 2.2%)
0.085
0.011
L11.2% (21.5% to L1%)
0.032
0.18
6% (15.6% to 3.6%)
0.22
0.11
8.4% (18.8% to 2.1%)
0.11
0.82
0.4% (11.3% to 12.1%)
0.94
0.017
L11.3% (L21.9% to L6.1%)
0.038
0.42
9.1% (25.3% to 7.1%)
0.27
0.78
0.9% (12.5% to 14%)
0.89
0.37
5.7% (6.4% to 17.8%)
0.35
0.041
13.2% (28.6% to 2.2)
0.09
0.83
6.2% (7.2% to 19.7%)
0.36
Chorioamnionitis (56): 203 g (L348 to L58) Chorioamnionitis with vasculitis and/or funisitis (32): L214 (L400 toL28) Chorioamnionitis with funisitis (29) L308 g (L502 to L114) Chorioamnionitis Stage 2 and 3 only (24) 174 (384 to 35) Chorioamnionitis Grade 2 only (14) 278 (546 to 9.2) Hemosiderin in any location (89): 201 g (L356 to L46) Hemosiderin in the fetal membranes (69) 99 g (246 to 47) Hemosiderin in the fetal membranes e Multifocal only (42) 133 (295 to 29) Hemosiderin in the fetal membranes e Heavy only (32) 21 (200 to 158) Hemosiderin in the basal plate (35) 197 g (L356 to L36) Hemosiderin in the basal plate e Multifocal only (14) 102 (352 to 147) Laminar necrosis (19): 29 g (174 to 233) Intramembranous/intradecidual thrombi (24) 83.6 (101 to 268) Subchorionic thrombi (15) L244 (L478 to L10.5) Intervillous thrombi (18) 21.7 (228 to 184)
BW ¼ birth weight, CI ¼ confidence interval. SPTB ¼ spontaneous preterm birth. a Results are from linear regression with adjustment for interpregnancy interval and maternal BMI, race and tobacco use. For gestational age adjusted birth weight, gestational age at delivery was included in the model.
episodes which did not result in preterm delivery per se though still had clinical consequences. A future prospective study would potentially provide both validation and further elucidation. If additional evaluations ultimately confirm an association between the clinical outcomes in subsequent pregnancies among patients with prior SPTB and their prior placental histology it could lead to both enhanced management of such patients and an expansion of the role of placental pathology. In such a scenario vigilant follow-up of the presence of acute chorioamnionitis and acute funisitis by pathologic examination and more liberal use of iron staining in these specimens may be helpful in identifying chorion-decidual bleeding or inflammation that may signal increased risks in a future pregnancy beyond the recurrent risk of SPTB itself. Acknowledgment This work was funded by WRHR K-12: HD001332-09. References [1] Meis P, Klebanoff M, Thom E, Dombrowski M, Sibai B, Moawad A, et al. Prevention of recurrent preterm birth by 17 alpha-hydroxyprogesterone caproate. N Eng J Med 2003;348:2379e85. [2] Berghella V, Odibo A, To M, Rust O, Althuisius S. Cerclage for short cervix on ultrasonography meta-analysis of trials using individual patient-level data. Obstet Gynecol 2005;106:181e9. [3] Spong CY. Prediction and prevention of recurrent spontaneous preterm birth. Obstet Gynecol 2007 Aug;110:405e15.
[4] Lockwood C, Kuczynski E. Risk stratification and pathological mechanisms in preterm birth. Paediatr Perinat Epidemiol 2001;15:78e89. [5] Romero R, Mazor M, Munoz H, Gomez R, Galasso M, Sherer D. The preterm labor syndrome. Ann N Y Acad Sci 1994;734:414e29. [6] Hillier SL, Martius J, Krohn M, Kiviat N, Holmes KK, Eschenbach DA. A casecontrol study of chorioamnionic infection and histologic chorioamnionitis in prematurity. N Engl J Med 1988 Oct 13;319:972e8. [7] Salafia C, López-Zeno J, Sherer D, Whittington S, Minior V, Vintzileos A. Histologic evidence of old intrauterine bleeding is more frequent in prematurity. Am J Obstet Gynecol 1995;173:1065e70. [8] Himes K, Simhan H. Risk of recurrent preterm birth and placental pathology. Obstet Gynecol 2008;112:121e6. [9] Hackney DN, Durie DE, Dozier AM, Suter BJ, Glantz JC. Is the accuracy of prior preterm birth history biased by delivery characteristics? Matern Child Health J 2012 Aug;16:1241e6. [10] Nelson DM, Burton GJ. A technical note to improve the reporting of studies of the human placenta. Placenta 2011;32:195e6. [11] Redline RW, Faye-Petersen O, Heller D, Qureshi F, Savell V, Vogler C. Amniotic infection syndrome: nosology and reproducibility of placental reaction patterns. Pediatr Dev Pathol 2003;6:435e48. [12] Ghidini A, Salafia CM. Histologic placental lesions in women with recurrent preterm delivery. Acta Obstet Gynecol Scand 2005;84:547e50. [13] Goldenberg RL, Andrews WW, Faye-Petersen O, Cliver S, Goepfert AR, Hauth JC. The Alabama Preterm Birth Project: placental histology in recurrent spontaneous and indicated preterm birth. Am J Obstet Gynecol 2006;195: 792e6. [14] Bukowski R, Gahn D, Denning J, Saade G. Impairment of growth in fetuses destined to deliver preterm. Am J Obstet Gynecol 2001;185:463e7. [15] Morken NH, Kallen K, Jacobsson B. Fetal growth and onset of delivery: a nationwide population-based study of preterm infants. Am J Obstet Gynecol 2006;195:154e61. [16] Espinoza J, Kusanovic JP, Kim CJ, Kim YM, Kim JS, Hassan SS, et al. An episode of preterm labor is a risk factor for the birth of a small-for-gestational-age neonate. Am J Obstet Gynecol 2007;196. 574 e1e5.
Contents lists available at ScienceDirect
Placenta journal homepage: www.elsevier.com/locate/placenta
Do placental histologic findings of chorion-decidual hemorrhage or inflammation in spontaneous preterm birth influence outcomes in the subsequent pregnancy? D.N. Hackney a, *, R. Tirumala a, L.J. Salamone a, R.K. Miller a, P.J. Katzman b a
Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, University of Rochester School of Medicine, 601 Elmwood Ave, Rochester, NY 14642, USA Department of Pathology, University of Rochester School of Medicine, 601 Elmwood Ave, Rochester, NY 14642, USA
b
a r t i c l e i n f o
a b s t r a c t
Article history: Accepted 2 November 2013
Introduction: Spontaneous preterm birth (SPTB) is the common endpoint of different underlying etiologies, including chorion-decidual bleeding and inflammation. However, specific histologic findings from a prior pregnancy do not always inform clinical management in subsequent pregnancies secondary to few prior studies having evaluated the relationship between prior pregnancy pathology and subsequent outcomes in patients with SPTB. Methods: Included subjects had: 1) a SPTB with available placental pathology and 2) a subsequent consecutive delivery at >20 weeks gestational age at our institution. For included subjects archived placenta and membrane paraffin blocks from the index SPTB were cut, stained with Prussian Blue and evaluated by a perinatal pathologist for the presence of hemosiderin. The association between histologic findings and subsequent pregnancy outcomes were evaluated through logistic and linear regression. Results: A total of 131 subjects were included, of whom 39.7% had a recurrent SPTB. Funisitis at the time of preterm delivery significantly increased the risk of early (<34 weeks) recurrent preterm birth (OR 3.38, p ¼ 0.016), though this may have been confounded by gestational age at delivery. Several histologic features were significantly associated with reductions in birth weight in the subsequent pregnancies, even if they did not increase the risk of recurrent preterm birth. Discussion: The presence of chorion-decidual bleeding or inflammation in a prior pregnancy can signal an increased risk in a future pregnancy beyond the recurrent risk of SPTB itself. Conclusions: Placental histologic findings after SPTB maybe associated with differences in birth weight in a subsequent pregnancy. Ó 2013 Elsevier Ltd. All rights reserved.
Keywords: Preterm birth Chorion-decidual hemorrhage Hemosiderin Placental pathology Chorioamnioitis Intrauterine growth restriction
1. Introduction The last several years have seen significant advancements in the management of patients at risk for spontaneous preterm birth (SPTB), including the use of 17-a-hydroxyprogesterone caproate [1] and cervical length screening with possible cerclage [2]. Many of these treatments have proved most beneficial in patients whose * Corresponding author. Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, University Hospitals Case Medical Center/Case-Western Reserve University, MacDonald Women’s Hospital, 11100 Euclid Avenue, MAC 5034, Cleveland, OH 44106, USA. Tel.: þ1 216 844 3787; fax: þ1 216 201 5063. E-mail addresses: [email protected] (D.N. Hackney), Radhika. [email protected] (R. Tirumala), [email protected]. edu (L.J. Salamone), [email protected] (R.K. Miller), Philip_ [email protected] (P.J. Katzman). 0143-4004/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.placenta.2013.11.001
specific SPTB risk factor is the occurrence of an SPTB in a prior pregnancy, leading to increased clinical and academic interest in these patient populations [3]. Though all cases of SPTB share the common endpoint of the delivery of a premature neonate, a variety of underlying etiological pathways are present for the preterm deliveries of different women [4,5]. These would include excessive uterine stretch, infection and inflammation and chorion-decidual hemorrhage, though in many cases a specific etiology cannot be identified. Studies of patients with prior preterm birth will often characterize the prior deliveries by gestational age at delivery or the presence or absence of premature membrane rupture, though rarely by the known or presumed underlying etiology of the prior delivery. As a result, patients are often included together whose prior preterm births may have resulted from very different fundamental processes.
D.N. Hackney et al. / Placenta 35 (2014) 58e63
Placental pathology after a spontaneous preterm birth is a powerful tool for elucidating the underlying etiology of the delivery which has just occurred. Specifically there is a well-characterized association between SPTB and the histologic presence of chorioamnionitis [6], with funisitis and vasculitis being particularly associated with deliveries at increasingly early gestational ages. Prior chorion-decidual hemorrhage as a potential cause of SPTB is more difficult to ascertain on standard Hemotoxylin/Eosin (H þ E) staining, though hemosiderin deposition can be identified through the use of Prussian Blue staining. Prior studies using Prussian Blue staining have also demonstrated a strong association between hemosiderin deposition and SPTB [7]. Though placental evaluation after SPTB is highly informative with regards to the pregnancy which just concluded, its overall clinical usefulness can be limited by the fact that the affected pregnancy is completed at the time of analysis. A potentially significant clinical value for placental pathology would be if the histologic findings from a prior pregnancy were able to inform clinical management in a subsequent pregnancy in the same patient. However, very few prior studies have evaluated the relationship between histologic findings in a prior pregnancy and clinical outcomes in subsequent gestations [8]. This study evaluated the clinical outcomes of pregnancies which occurred after a prior SPTB with regards to the histologic findings from the index delivery. Specifically the association between subsequent pregnancy outcomes and chorion-decidual bleeding as a presumed underlying etiology of a prior SPTB as evaluated with Prussian Blue staining for hemosiderin has not previously been reported. 2. Methods 2.1. Subject identification and clinical variable abstraction We conducted a retrospective cohort of patients who had a SPTB at the University of Rochester Medical Center between 2004 and 2009 followed by a consecutive pregnancy which had also occurred at the same institution. Potential subjects were identified through a maternally linked birth certificate database as described in a prior publication of birth certificate accuracy [9]. Patients who met the clinical inclusion criteria were then cross-referenced against an electronic pathology database to determine the availability of archived paraffin blocks of the placenta parenchyma and membrane roll. Clinical inclusion criteria were an “index” SPTB between 20 and 36 6/7 weeks estimated gestational age (EGA) and a
59
“subsequent” delivery at >20 weeks EGA with both deliveries having occurred at the above institution and no record of an intervening delivery at an outside institution between those two. The “index” preterm birth did not necessarily have to also be the patient’s first pregnancy. The “subsequent” delivery could be either term or preterm, with the outcomes of that pregnancy being the outcome of interest for the study. Spontaneous preterm births were defined as having occurred secondary to preterm labor or preterm premature rupture of membranes. “Index” preterm deliveries which were performed iatrogenically secondary to pre-eclampsia, intrauterine growth restriction (IUGR) or another maternal or fetal indication were excluded. Exclusion criteria were multifetal gestation, thrombophilia, anticoagulation, major congenital anomaly, structural uterine anomaly or substance abuse (other than marijuana) in either pregnancy as determined by manual review of the prenatal and delivery records of both pregnancies. The following clinical variables were abstracted by manual review of the prenatal and delivery records of both pregnancies along with the interpregnancy (delivery to the last menstrual period of the subsequent pregnancy) interval: EGA at delivery, birth weight, tobacco use, maternal age, maternal race, gravidity and maternal pre-pregnancy BMI. Gestational age specific birth weight percentiles were determined using the Finger Lakes Region birth weight curves, which are based on the local birth record registry within the Finger Lakes delivery region. The study was approved by the University of Rochester Research Subjects Review Board. Clinical characteristics of the index pregnancies for pregnancies of placentas studied is presented in Table 1 [10]. 2.2. Histologic evaluation All histologic evaluations were performed by a single subspecialty trained perinatal pathologist (PJK) who was blinded to all of the clinical variables including outcomes. Original H þ E slides and archived paraffin embedded placental parenchyma and membrane roll blocks from the index preterm birth were retrieved for all included subjects. H þ E slides were reviewed to determine two paraffin blocks that would be most appropriate for the evaluation of hemosiderin by Prussian Blue staining. The selected blocks, one block with at least one section of the maternal surface and one block with the membrane roll, were then retrieved and re-cut. After xylene incubation and hydration via graded alcohols, slides were incubated for 30 min in potassium ferrocyanide solution (1.0 g potassium ferrocyanide, 98 mL water and 2.0 mL of 10% Triton X-100 then mixed in a 1:1 ratio with 2% Hydrochloric acid), followed by Nuclear Red fast stain, dehydration and mounting. H þ E slides were scored using a binary scale for presence or absence of acute or chronic chorioamnionitis, acute funisitis, acute chorionic vasculitis, fetal vessel thrombosis, villous sclerosis, laminar necrosis of the membranes, intramembranous/ intradecidual thrombi, subchorionic thrombi, and intervillous thrombi. The maternal acute inflammatory and fetal acute inflammatory responses were further scored by stage (score 1-3) and grade (score 1-2) as detailed by an Amnionic Fluid Infection Nosology Committee study [11]. The Prussian Blue stains were scored for focality and intensity of staining in the basal plate on the maternal surface and in the fetal membranes. Focality was scored as either focal or multifocal and intensity was scored as either light or heavy. Stains were evaluated using 4, 10, and 20
Table 1 Clinical characteristics of included subjects. Parameter
Clinical characteristics of pregnancies for placentas studied
Gravidity Parity Gestational age (weeks) Maternal age (years) Race Ethnicity Prenatal medications Drugs Previous prenatal admission(s) Blood pressures <140/90 mm Hg Screened for diabetes Antibiotics in labor Beta strep status Antenatal steroids: Magnesium sulfate Anesthesia Cervical ripening agent Labor Delivery mode (N ¼ # patients)
Median ¼ 2 Median ¼ 1 Average ¼ 32 3/7 Average ¼ 25.9 Black 25%
25e75% ¼ 1e3 25e75% ¼ 0e2 SD ¼ 3.6 SD ¼ 5.9 White 56%
Iron ¼ 12.2% Cigarettes n ¼ 18.7% Yes, n ¼ 0 Yes, n ¼ 2 Yes ¼ 53.5% None, n ¼ 0 Positive Yes, n ¼ 74 Yes, n ¼ 61 Epidural: Unknown Prostaglandin E1 Yes ¼ 100% C-section, Repeat, no labor: n ¼ 0 Yes, n ¼ 0 Average ¼ 1993 Average¼ Female ¼ 41% Average¼
Albuterol: 10% Alcohol: Unknown No, n ¼ 0 No, n ¼ 129 No ¼ 16% Penicillin, n ¼ 0 Negative No, n ¼ 57 No, n ¼ 70 Narcotics: Unknown Prostaglandin E2 If yes, hours: Unknown C-section, Repeat, with labor: n ¼ unknown No, n ¼ 0 SD ¼ 711 SD¼ Male ¼ 59% SD¼
Maternal oxygen given at delivery? Birth weight (grams) Placental weight (grams) Baby’s sex Delivery to processing (min)
Range 1e9 Range 0e7 Range 16e38 Other 19% None (except vitamins): 76.3% Others: 0% (excluded) Diagnoses if yes: Unknown ¼ 0 Unknown ¼ 30.5% Other: (list) Unknown 100% If yes, week GA: Unknown Unknown ¼ 0 General Mechanical C-section, primary, no labor: n ¼ 0 Unknown ¼ 100% Unknown Unknown 0% Unknown
Unknown 0% Unknown 0% Unknown, n ¼ 100%
Unknown, n ¼ 100% Unknown ¼ 0 Other/none Other: Unknown: 100% C-section, primary, with labor: n ¼ unknown
60
D.N. Hackney et al. / Placenta 35 (2014) 58e63
objectives and focality was defined as one positive area in a 10 objective field. Multifocality was defined as several areas or one continuous area covering more than one 10 objective field. Light staining was defined as dot-like staining pattern in a 20 objective. Heavy staining was defined as staining of entire cells with a homogenous staining in a 10 objective. Representative photomicrographs are presented in Figs. 1 and 2. 2.3. Statistical analysis The association between individual histologic variables and the binary outcomes of recurrent preterm birth and SGA were evaluated with logistic regression with and without adjustment for potential clinical confounders. The association between individual histologic variables and the continuous variables of birth weight percentiles and birth weight adjusted for gestational age at delivery were evaluated through linear regression models with adjustments for potential clinical confounders. Clinical confounders included BMI, tobacco use, maternal ethnicity and interpregnancy interval. All analyses were performed on Stata 11 (College Station, TX, USA). A p-value of 0.05 was considered statistically significant.
3. Results Subjects (256) were initially identified as having had a singleton spontaneous preterm birth at the University of Rochester Medical Center followed by a subsequent singleton delivery at >20 weeks gestational age. Subjects (65 or 25.4%) were excluded secondary to the absence of archived placental samples. The majority (88%) of this group constituted patients with “late” (34e36 weeks) prior preterm births, with the median gestational age at delivery in the “index” pregnancy for excluded subjects being 36 2/7 weeks. Though not always explicitly stated in the delivery record, presumably in these cases the placenta was not sent to pathology at the discretion of the delivering provider secondary to the delivery having occurred near term. Of the remaining subjects with available placental specimens, 60 were excluded following medical record review for the following indications: major congenital fetal anomaly in 24, substance abuse in 24, thrombophilia in 9 and congenital maternal uterine anomalies in 3. Most cases of substance abuse were patients who had experienced spontaneous preterm labor in the context of cocaine intoxication. A final total of 131 subjects were included, the subsequent pregnancy outcome characteristics of which are provided in Table 2. 39.7% of included subjects had a recurrent SPTB. The index SPTB demonstrated chorioamnionitis in 42.7%, vasculitis or funisitis in 24.4%, funisitis alone in 22.1%, fetal membrane hemosiderin in 52.7%, basal plate hemosiderin in 26.7% and laminar necrosis in 14.5%. The association between histologic findings and gestational age in the index pregnancy is presented in Table 3. Many subjects had more than one histologic finding, with only 16.7% having no specific histologic findings. Hemosiderin staining was more common when thrombi were present in the membranes or decidua, though there was no association with thrombi in other locations. The following histologic findings occurred infrequently (<10%
incidence) and were thus not evaluated separately: chronic chorioamnionitis, “heavy” hemosiderin of the basal plate, fetal vessel thrombosis and villous sclerosis. Preterm birth outcomes according to index placental pathology are provided in Table 4. When adjusted for potential clinical confounders, the presence of funisitis in an index preterm birth was associated with a significant increase in subsequent preterm birth (aOR 3.46, p ¼ 0.02) and specifically early preterm deliveries (aOR 5.46, p ¼ 0.008). The combined group of funisitis and/or fetal vessel vasculitis also demonstrated a significant though smaller association with recurrent preterm birth as did stage 2 or 3 chorioamnionitis and preterm birth recurrence in adjusted analysis. The finding of multifocal hemosiderin in the fetal membranes was associated with a degreased odds of recurrent preterm birth, though only in the multivariate analysis and with moderate significance (p ¼ 0.041). No other histologic findings had an association with the risk of preterm birth recurrence, including markers of chorion-decidual bleeding (hemosiderin) or potential hypoxia (laminar necrosis). Patients with funisitis in the index pregnancy had delivered at earlier gestational ages in that pregnancy than subjects without this finding (median EGA 31 1/7 versus 34 1/7) and index gestational age was associated with subsequent gestational age when adjusted for the same confounders (p ¼ 0.007) When adjusted for gestational age in the index delivery, the association between funisitis and subsequent early preterm birth was no longer significant at a 0.05 level (aOR 2.5, p ¼ 0.08). Several histologic features were significantly associated with reductions in gestational age-adjusted birth weight and/or birth weight percentiles in the subsequent pregnancies as outlined in Table 5. Histologic lesions that were associated with decreased birth weight included some that had no association with preterm birth recurrence itself, including basal plate hemosiderin and chorioamnionitis. Unlike the risk of recurrent preterm birth there was no association in linear regression between gestational age in the index preterm birth and gestational age adjusted birth weight (p ¼ 0.89) or birth weight percentile (p ¼ 0.97) in the subsequent pregnancy.
4. Discussion Spontaneous preterm birth represents a common endpoint for a heterogeneous collection of underlying etiologies. For patients who have experienced a prior spontaneous preterm birth, outcomes in a subsequent pregnancy may be impacted by these underlying etiologies in addition to other known prognostic clinical variables. Specifically, the results of our study suggest that patients with prior SPTBs secondary to chorion-decidual hemorrhage or inflammation may be at an increased risk of decreased birth weight in a future pregnancy relative to patients who did not demonstrate these
Fig. 1. (A) Free membranes contain laminar necrosis (lower half of A) where there is heavy iron deposition seen on Prussian Blue stain (B) (original magnification, 200).
D.N. Hackney et al. / Placenta 35 (2014) 58e63
61
Fig. 2. (A) Free membranes contain intramembranous thrombus (asterisks) with hemosiderin that is positive for heavy iron deposition on Prussian Blue stain (B) (original magnification, 200).
histologic findings, and this association may be independent of the risk of recurrent preterm birth itself. For these patients, the histologic anomalies in their index preterm births could signify intrinsic or chorionic conditions, such as a predisposition to inflammation or suboptimal placental vascular development, which in a subsequent pregnancy could generate complications which are insufficient to lead to recurrent preterm birth though sufficient to impair growth. This would be particularly plausible with regards to subclinical bleeding events at the chorion-decidual interface. To our knowledge this is the first study to evaluate the association between chorion-decidual hemorrhage in a prior pregnancy and clinical outcomes in subsequent deliveries or between histologic findings in SPTB and subsequent pregnancy birth weight. Strengths of the study include the evaluations of all histology slides by a single, subspecialty trained perinatal pathologist who was blinded to the clinical variables. We also performed Prussian Blue staining on all archived cases, thus allowing us to evaluate subclinical hemosiderin deposition which may not have been evident on H þ E staining. Because all subjects had both of their deliveries at the same institution, we also had access to all complete delivery records. The primary limitation of our study was that it was retrospective and limited to patients for whom archived placental blocks were available for re-staining. Specifically, our study is not necessarily representative of all patients who have had a prior SPTB, as many patients with “late” (34e36 weeks) preterm births had not originally had their placentas submitted for histologic evaluation. We also had a very high incidence of tobacco use in our cohort, though this was adjusted for in multivariate analysis. Additionally, though our regression models included the relevant clinical variables which were available to us, we could only evaluate those which were present in the available records. 17-alpha hydroxyprogesterone caproate is commonly used for the prevention of recurrent preterm birth [1] and would have thus served as an interesting source of secondary analysis. However, only 6 subjects were identified as having used this medication in their subsequent
Table 2 Subsequent pregnancy outcome characteristics of study population. Inter-pregnancy interval: 16 (1e66) Estimated EGA subsequent pregnancy (weeks): 37 4/7 (20 0/7e41 6/7) Birth weight subsequent pregnancy (g): 2870 (710e4530) BMI subsequent pregnancy: 25.7 (18e47.5) % Recurrent preterm birth in subsequent pregnancy: 39.7% % Early (<34 weeks) preterm birth in subsequent pregnancy: 16% Data presented as median (range). Inter-pregnancy interval defined as delivery date of index to last menstrual period of subsequent pregnancy rounded to months. BMI ¼ pre-pregnancy body mass index. EGA ¼ estimated gestational age.
pregnancies, which was insufficient for meaningful statistical analysis. Also, as most patients delivered at term in their subsequent pregnancies and thus did not undergo placental evaluations in their subsequent pregnancies, we could not perform a meaningful evaluation of placental histology in subsequent pregnancies in the same patient, which would have been an interesting evaluation. Though the presence of placental hemosiderin correlates with preterm birth [7], it is not usually possible histologically to determine at what time in pregnancy the bleeding event occurred and thus if it had an actual temporal relationship with the preterm delivery. Placentas from patients experiencing recurrent SPTB are more likely to demonstrate inflammatory findings than those from patients who are experiencing a first SPTB [12,13], which would also potentially be consistent with prior inflammatory findings increasing the risk of SPTB recurrence. Finally, though we were able to demonstrate statistically significant and potentially clinically meaningful differences in birth weight with various histologic findings, we were underpowered to demonstrate specific differences in birth weight at <10%. The a priori primary outcome of our study was the recurrence of preterm birth in a subsequent pregnancy. In this regard a significant association was demonstrated with the presence of many of the inflammatory markers in the index preterm birth, including, in particular funisitis. However, it is possible that the association between funisitis and recurrent preterm birth was confounded by the gestational age at delivery in the index pregnancy as both funisitis was associated with an earlier gestational age in the index pregnancy and decreasing index pregnancy gestational age is associated
Table 3 Histologic findings and median gestational age at delivery in the index pregnancy. Histologic finding
Median GA at delivery
p
All subjects Chorioamnionitis Chorioamnionitis w/vasculitis and/or funisitis: Funisitis Chorioamnionitis Stage 2 and 3 only Chorioamnionitis Grade 2 only Hemosiderin in any location Hemosiderin in fetal membranes Hemosiderin in fetal membranes - Multifocal only Hemosiderin in fetal membranes - Heavy only Hemosiderin in the basal plate Hemosiderin in the basal plate - Multifocal only
33 0/7 30 7/8
Referent <0.001
30 30 28 28 32 32
<0.001 <0.001 <0.001 <0.001 0.94 0.76
1/7 3/7 6/7 4/7 3/7 4/7
32 5/7
0.5
33 1/7 31 3/7
0.17 0.03
32 0/7
0.67
GA ¼ gestational age. p values determined by linear regression.
62
D.N. Hackney et al. / Placenta 35 (2014) 58e63
Table 4 Associations between histologic findings and the subsequent pregnancy outcomes. Histologic finding (number)
% Outcome
Negative (22): %Recurrent PTB 36.4% % Early PTB 18.1% Chorioamnionitis (56): %Recurrent PTB 42.9% % Early PTB 21.4% Chorioamnionitis with vasculitis and/or funisitis (32): %Recurrent PTB 46.8% % Early PTB 28.1% Chorioamnionitis with funisitis (29) %Recurrent PTB 51.7% % Early PTB 31% Chorioamnionitis Stage 2 and 3 only (24) %Recurrent PTB 54.5% % Early PTB 29.2% Chorioamnionitis Grade 2 only (14) %Recurrent PTB 35.7% % Early PTB 14.2% Hemosiderin in any location (89): %Recurrent PTB 40.4% % Early PTB 13.5% Hemosiderin in the fetal membranes (69) %Recurrent PTB 37.7% % Early PTB 13% Hemosiderin in the fetal membranes e Multifocal only (42) %Recurrent PTB 30.9% % Early PTB 11.9% Hemosiderin in the fetal membranes e Heavy only (32) %Recurrent PTB 43.8% % Early PTB 12.5% Hemosiderin in the basal plate (35) %Recurrent PTB 51.5% % Early PTB 17.1% Hemosiderin in the basal plate e Multifocal only (14) %Recurrent PTB 35.7% % Early PTB 14.3% Laminar necrosis (19): %Recurrent PTB 36.8% % Early PTB 5.2% Intramembranous/intradecidual thrombi (24) %Recurrent PTB 25% % Early PTB 8.3% Subchorionic thrombi (15) %Recurrent PTB 40% % Early PTB 20% Intervillous thrombi (18) %Recurrent PTB 33.3% % Early PTB 16.7%
OR (95% CI)
p-value
Referent Referent
aOR* (95% CI)
p-value
Referent Referent
1.26 (0.62e1.55) 2.0 (0.78e5.14)
0.32 0.15
1.32 (0.6e2.92) 1.8 (0.65e4.94)
0.49 0.26
1.48 (0.67e3.3) 2.84 (1.70e7.55)
0.22 0.035
2.37 (0.88e6.36) 3.61 (1.12e11.7)
0.09 0.032
1.88 (0.82e4.32) 3.38 (1.25e9.09)
0.14 0.016
3.46 (1.2e9.9) 5.46 (1.57e19)
0.02 0.008
2.06 (0.84e50.4) 2.74 (0.96e7.77)
0.11 0.06
3.36 (1.04e10.8) 1.75 (0.47e6.51)
0.043 0.4
0.83 (0.26e2.62) 0.86 (0.18e4.15)
0.75 0.85
0.9 (0.21e3.92) 0.26 (0.02e2.58)
0.89 0.25
1.1 (0.52e2.34) 0.57 (2.2e1.49)
0.48 0.18
1.3 (0.55e3.09) 0.91 (0.31e2.62)
0.54 0.86
0.84 (0.42e1.69) 0.62 (0.24e1.6)
0.62 0.33
0.84 (0.39e1.85) 0.93 (0.33e2.57)
0.67 0.89
0.57 (0.26e1.25) 0.61 (0.21e1.81)
0.16 0.38
0.37 (0.14e0.96) 0.48 (0.12e1.85)
0.041 0.28
1.25 (0.56e2.8) 0.69 (0.21e2.22)
0.59 0.53
0.92 (0.35e2.47) 0.7 (0.18e2.77)
0.88 0.61
1.93 (0.88e4.23) 1.11 (0.4e3.15)
0.1 0.83
2.17 (0.91e5.18) 1.4 (0.45e4.33)
0.08 0.56
0.82 (0.26e2.62) 0.86 (0.18e4.15)
0.75 0.85
0.61 (0.15e2.34) 0.57 (0.06e5.17)
0.47 0.62
0.89 (0.32e2.37) 0.26 (0.03e2.02)
0.5 0.2
0.87 (0.29e2.62) 0.19 (0.02e1.63)
0.81 0.13
0.44 (0.16e1.2) 0.42 (0.9e1.94)
0.11 0.27
0.57 (0.19e1.64) 0.42 (0.1e2.53) 0.42
0.29
1.01 (0.34e3.04) 1.36 (0.35e5.31)
0.98 0.66
0.95 (0.25e3.61) 1.42 (0.26e7.57)
0.95 0.68
0.73 (0.26e2.08) 1.06 (0.28e4.02)
0.55 0.94
0.95 (0.3e2.99) 1.17 (0.28e4.89)
0.93 0.82
CI ¼ confidence interval, OR ¼ Odds Ratio, *aOR ¼ adjusted Odds Ratio (adjusted for interpregnancy interval, race, tobacco use and maternal BMI), PTB ¼ preterm birth (gestational age at delivery <34 weeks). Statistically significant (p < 0.05) results are in bold.
with increasing risks of recurrent preterm birth. Thus, the association is potentially complicated by multicolinearity or mediation. Specifically if inflammatory changes in the index pregnancy lead to earlier preterm births and these in turn increase the risk of subsequent preterm delivery, then gestational age in the index pregnancy could be a mediator and not necessarily a confounder. When an adjustment for gestational age at index SPTB was performed, however, a non-significant trend (p ¼ 0.08) towards an increased independent risk of recurrent preterm birth with funisitis still existed. Additionally, the association between inflammatory vascular lesions of the placenta and preterm birth recurrence in future pregnancies would be consistent with the prior work of Himes and Simhan [8] and Ghidini [12]. Of note, unlike the potential association with recurrent preterm birth, the gestational age at index delivery was not associated with differences in either birth weight percentile or gestational age adjusted birth weight in the subsequent pregnancies. An association was also demonstrated between hemosiderin in the fetal membranes and preterm birth. However, the degree of
association, while significant at an alpha of 0.05, was not large (p ¼ 0.041) and all other measures of hemosiderin staining were not associated with recurrent birth risk. Thus the potential exists that this singular result was a spurious consequence of multiple comparisons, though the potential for an association would be an interesting focus of future prospective studies. The association between histologic features from an index preterm delivery and decreased birth weight in a subsequent pregnancy is in our opinion both novel and interesting. Fetuses that deliver prematurely are known to have an overall decreased weight percentile [14,15] and it is plausible that the etiology for the decreased growth could persist into a subsequent pregnancy even if the preterm delivery itself does not recur. These could include predispositions to subclinical intrauterine inflammatory events or chorion-decidual hemorrhage. Additionally, Espinoza et al. [16] have demonstrated that patients who experience an episode of preterm labor which resolves and allows that pregnancy to continue to term have an increased risk of delivering an SGA infant. Thus many patients with prior SPTB may have had preterm labor
D.N. Hackney et al. / Placenta 35 (2014) 58e63
63
Table 5 Associations between histologic findings in SPTB and birth weight in subsequent pregnancies. Histologic finding (number)
Gestational-age adjusted BW Coefficienta (95% CI)
BW percentile p-value
Coefficient* (95% CI)
p-value
0.006
L13.3% (L22.8% to L3.8%)
0.006
0.024
11.6% (23.9% to 0.6%)
0.063
0.002
L17.1% (L29.9% to L4.3%)
0.009
0.1
12.5% (26.1% to 1.2%)
0.07
0.043
15.3% (32.7% to 2.2%)
0.085
0.011
L11.2% (21.5% to L1%)
0.032
0.18
6% (15.6% to 3.6%)
0.22
0.11
8.4% (18.8% to 2.1%)
0.11
0.82
0.4% (11.3% to 12.1%)
0.94
0.017
L11.3% (L21.9% to L6.1%)
0.038
0.42
9.1% (25.3% to 7.1%)
0.27
0.78
0.9% (12.5% to 14%)
0.89
0.37
5.7% (6.4% to 17.8%)
0.35
0.041
13.2% (28.6% to 2.2)
0.09
0.83
6.2% (7.2% to 19.7%)
0.36
Chorioamnionitis (56): 203 g (L348 to L58) Chorioamnionitis with vasculitis and/or funisitis (32): L214 (L400 toL28) Chorioamnionitis with funisitis (29) L308 g (L502 to L114) Chorioamnionitis Stage 2 and 3 only (24) 174 (384 to 35) Chorioamnionitis Grade 2 only (14) 278 (546 to 9.2) Hemosiderin in any location (89): 201 g (L356 to L46) Hemosiderin in the fetal membranes (69) 99 g (246 to 47) Hemosiderin in the fetal membranes e Multifocal only (42) 133 (295 to 29) Hemosiderin in the fetal membranes e Heavy only (32) 21 (200 to 158) Hemosiderin in the basal plate (35) 197 g (L356 to L36) Hemosiderin in the basal plate e Multifocal only (14) 102 (352 to 147) Laminar necrosis (19): 29 g (174 to 233) Intramembranous/intradecidual thrombi (24) 83.6 (101 to 268) Subchorionic thrombi (15) L244 (L478 to L10.5) Intervillous thrombi (18) 21.7 (228 to 184)
BW ¼ birth weight, CI ¼ confidence interval. SPTB ¼ spontaneous preterm birth. a Results are from linear regression with adjustment for interpregnancy interval and maternal BMI, race and tobacco use. For gestational age adjusted birth weight, gestational age at delivery was included in the model.
episodes which did not result in preterm delivery per se though still had clinical consequences. A future prospective study would potentially provide both validation and further elucidation. If additional evaluations ultimately confirm an association between the clinical outcomes in subsequent pregnancies among patients with prior SPTB and their prior placental histology it could lead to both enhanced management of such patients and an expansion of the role of placental pathology. In such a scenario vigilant follow-up of the presence of acute chorioamnionitis and acute funisitis by pathologic examination and more liberal use of iron staining in these specimens may be helpful in identifying chorion-decidual bleeding or inflammation that may signal increased risks in a future pregnancy beyond the recurrent risk of SPTB itself. Acknowledgment This work was funded by WRHR K-12: HD001332-09. References [1] Meis P, Klebanoff M, Thom E, Dombrowski M, Sibai B, Moawad A, et al. Prevention of recurrent preterm birth by 17 alpha-hydroxyprogesterone caproate. N Eng J Med 2003;348:2379e85. [2] Berghella V, Odibo A, To M, Rust O, Althuisius S. Cerclage for short cervix on ultrasonography meta-analysis of trials using individual patient-level data. Obstet Gynecol 2005;106:181e9. [3] Spong CY. Prediction and prevention of recurrent spontaneous preterm birth. Obstet Gynecol 2007 Aug;110:405e15.
[4] Lockwood C, Kuczynski E. Risk stratification and pathological mechanisms in preterm birth. Paediatr Perinat Epidemiol 2001;15:78e89. [5] Romero R, Mazor M, Munoz H, Gomez R, Galasso M, Sherer D. The preterm labor syndrome. Ann N Y Acad Sci 1994;734:414e29. [6] Hillier SL, Martius J, Krohn M, Kiviat N, Holmes KK, Eschenbach DA. A casecontrol study of chorioamnionic infection and histologic chorioamnionitis in prematurity. N Engl J Med 1988 Oct 13;319:972e8. [7] Salafia C, López-Zeno J, Sherer D, Whittington S, Minior V, Vintzileos A. Histologic evidence of old intrauterine bleeding is more frequent in prematurity. Am J Obstet Gynecol 1995;173:1065e70. [8] Himes K, Simhan H. Risk of recurrent preterm birth and placental pathology. Obstet Gynecol 2008;112:121e6. [9] Hackney DN, Durie DE, Dozier AM, Suter BJ, Glantz JC. Is the accuracy of prior preterm birth history biased by delivery characteristics? Matern Child Health J 2012 Aug;16:1241e6. [10] Nelson DM, Burton GJ. A technical note to improve the reporting of studies of the human placenta. Placenta 2011;32:195e6. [11] Redline RW, Faye-Petersen O, Heller D, Qureshi F, Savell V, Vogler C. Amniotic infection syndrome: nosology and reproducibility of placental reaction patterns. Pediatr Dev Pathol 2003;6:435e48. [12] Ghidini A, Salafia CM. Histologic placental lesions in women with recurrent preterm delivery. Acta Obstet Gynecol Scand 2005;84:547e50. [13] Goldenberg RL, Andrews WW, Faye-Petersen O, Cliver S, Goepfert AR, Hauth JC. The Alabama Preterm Birth Project: placental histology in recurrent spontaneous and indicated preterm birth. Am J Obstet Gynecol 2006;195: 792e6. [14] Bukowski R, Gahn D, Denning J, Saade G. Impairment of growth in fetuses destined to deliver preterm. Am J Obstet Gynecol 2001;185:463e7. [15] Morken NH, Kallen K, Jacobsson B. Fetal growth and onset of delivery: a nationwide population-based study of preterm infants. Am J Obstet Gynecol 2006;195:154e61. [16] Espinoza J, Kusanovic JP, Kim CJ, Kim YM, Kim JS, Hassan SS, et al. An episode of preterm labor is a risk factor for the birth of a small-for-gestational-age neonate. Am J Obstet Gynecol 2007;196. 574 e1e5.