Placental Lesions Associated With Cerebral Palsy and Neurologic Impairment Following Term Birth.
Unlike preterm infants, term infants with later NI are often born outside of tertiary care centers. For this reason, it is very difficult for any one center to accumulate a sufficient number of cases to study the myriad of potentially causal factors. Because of the large number of cases that result in litigation, expert witness placental pathologists are in a position to review a large number of well-documented cases of NI. Although a number of selection bias and study design problems exist, these cases are an important resource that should not be ignored. In this study, we have compared placental pathology in 40 cases of term infants with later NI to that seen in a consecutive series of 176 placentas from term infants with meconium-stained fluid.
MATERIALS AND METHODS
This study was a retrospective comparison between placental findings in 40 term infants with NI and 176 consecutive meconium-stained term infants. Cases were drawn from a database containing clinicopathologic data derived from the medicolegal consultation files of one of the authors (R.W.R) and were identifiable only by a code number and whether consultation was requested on behalf of the defense or plaintiff. All infants born between 1990 and 1997 at 37 0/7 weeks or greater gestational age, as determined by best obstetric estimate, with NI diagnosed at a minimum of 20 months after birth constituted the study population. Neurologic impairment was defined as cerebral palsy or a closely related, nonprogressive motor disorder, as determined by an expert neurologic consultation in the patient's medical record. Infants with major congenital malformations, congenital infections, or genetic disorders, and mothers with insulin-dependent diabetes, autoimmune disease, multiple pregnancy, and severe preeclampsia were excluded. Following application of these criteria, 40 of 126 cases in the database remained and constituted the study cases.
The comparison group (N = 176) was drawn from a separate database of clinical cases examined by the same observer between June 1, 1996, and January 3, 1999. This group included only the subset of infants at 37 0/7 weeks or greater gestational age with a clinical history of meconium-stained fluid and complete clinical data, and this group was subject to the same exclusion criteria previously listed. Meconium staining was chosen for 2 reasons. The first reason was increased ascertainment: while placental submission at 37 weeks or more is sporadic, a relatively high percentage (approximately 70-80%) of placentas from meconium-stained infants at our institution are sent to pathology. The second reason was increased comparability between the 2 groups, since meconium is generally considered to be a marker of hypoxia, usually mild and/or transient hypoxia, at or near the time of birth. Neurologic outcome data were not available for the comparison group. Data from several large population studies indicate that the prevalence of NI at term varies from 0.7 to 4.0 per 1000 births.[7-9] Meconium staining increases the overall risk by 1.6- to 3.0-fold.[7,10] Even the combination of meconium plus a 5-minute Apgar score of 6 or less increases the risk only sixfold. Using the most recent prevalence data and the higher estimated risk for meconium staining, 0.34/176 infants would have been misclassified. In the worst-case scenario, using the highest prevalence and assuming that all infants had 5-minute Apgar scores of 6 or less, only 4/176 infants would have been misclassified. Misclassification of infants in the comparison group as not having NI would be expected to decrease the observed associations, and hence the findings reported could be considered as conservative estimates of the actual relationships.
Previous data implicate a limited number of placental lesions as possible predictors of NI and chronic hypoxic injury.[11-17] We evaluated 10 of these lesions and separated them for the purposes of this study into 2 categories: recent lesions (generally considered to occur within days of the time of labor and delivery), including acute chorioamnionitis, meconium-associated changes, recent chorionic vessel thrombi, changes consistent with abruptio placenta, and increased nucleated red blood cells (NRBC), and chronic lesions (generally believed to have their onset long before labor and delivery), including avascular villi, chronic villitis, perivillous fibrin, changes consistent with maternal vascular underperfusion, and diffuse chorioamnionic hemosiderosis. Four hypotheses were tested: first, that 1 or more of the lesions would be overrepresented in cases; second, that a dose response and/ or threshold effect related to severity or extent would be demonstrable; third, that multiple lesions would act combinatorially to increase the risk of NI; and fourth, that among placentas with multiple lesions, those with lesions estimated to occur at differing times relative to labor and delivery would show a stronger relationship to NI than those occurring at the same time.
Placental lesions were diagnosed prospectively before the study, using standardized criteria as outlined in previous publications. To assess possible bias, a subsample of placentas (20 cases with NI and 40 cases from the comparison group) were rereviewed in a blinded manner by the same observer, and intra-observer variation was calculated for each lesion. Criteria for the lesions are briefly summarized.
Meconium changes were separated into 4 groups: pigment in the amnion only, pigment in the amnion and decidua, a green-stained umbilical cord, and meconium-associated vascular necrosis of large fetal vessels.[18,19] Chorioamnionitis was separated into 3 groups: maternal neutrophils only, maternal neutrophils plus a mild to moderate fetal vascular response in chorionic plate vessels, and maternal neutrophils plus a severe fetal vascular response in chorionic plate vessels. Changes consistent with abruptio placenta were defined as retroplacental hematoma with intraplacental extension. Changes qualifying as intraplacental extension included indentation or rupture of the basal plate, diffuse intradecidual hemorrhage, or villous changes, such as recent infarction, villous stromal hemorrhage, or irregular intervillous thrombi. Diffuse chorioamnionic hemosiderosis was defined by hemosiderin-laden macrophages in the amnion and chorion of membranes and chorionic plate, plus old blood clot. Avascular villi were classified as mild to moderate when 2 to 9 foci of fewer than 5 villi and fewer than 45 total villi were affected, and as extensive when more than 9 foci of fewer than 5 villi or more than 45 total villi were identified. Chronic villitis was classified as focal when each focus contained fewer than 5 villi and fewer than 50 total villi were involved, and as diffuse when foci contained more than 5 villi or more than 50 villi were involved.
An increased NRBC was defined based on the findings in 1 full-thickness placental section measuring at least 2 x 2 cm and having an average density of terminal villi at 10x magnification. All areas with terminal villous capillaries were examined at 40x magnification until either 3 fields containing at least 2 unequivocal normoblasts in a single terminal villus were seen (increased NRBC) or the scan was completed. Slides having 3 fields that each had at least 1 terminal villus containing more than 2 normoblasts were classified as having moderate NRBC. All other cases of increased NRBC were classified as mild. Increased perivillous fibrin was diagnosed when all nonperipheral placental sections showed that 5% or more of the terminal villi in the half of the villous parenchyma bordering the basal plate were enmeshed in a fibrinoid matrix. Recent chorionic vessel thrombi were defined by laminated fibrin adherent to the endothelium of major vessels in the chorionic plate.[7,21] Changes consistent with maternal underperfusion included at least 1 of 3 features: increased syncytial knots, villous infarcts, and decidual arteriopathy. The number of placental sections varied from 3 to 5 in cases. Five placental sections were available for all controls. Most of the lesions evaluated were diffuse and relatively independent of sampling. For the other lesions, any undersampling in cases should have decreased the chances of seeing an association where one existed.
Because neurologic outcome was not known for infants in the comparison group, a strict statistical comparison between the 2 groups was not possible. However, since the expected number of misclassified cases was small (0.34-4.0 of the 214 cases) and misclassification would be expected to increase the magnitude of association, we used the following statistical tools in a descriptive fashion to estimate the actual relationships. Chi square analysis or Fisher exact tests were used for categorical variables. Continuous variables were compared using the 2-tailed t test. Placental lesions significant at the univariate level were studied by logistic regression to evaluate individual lesions and total number of lesions as independent predictors of NI. Results are expressed as odds ratios (ORs) with 95% confidence intervals (CIs).
The clinical characteristics of mothers and infants in the case and comparison groups are summarized in Table 1. No significant difference in maternal age or either infant or placental weight was observed between the 2 groups. Although mean gestational ages were equivalent, a significantly larger fraction of infants with later NI were born postdate (after 42 weeks gestation). Primigravidity and gestational diabetes were also more frequent among cases. The 2 groups differed most significantly with regard to indicators of perinatal asphyxia, with cases having markedly increased prevalences of emergency cesarean section and Apgar scores of less than 6 at 1 and 5 minutes of life. Many of these associations, including primigravidity, delivery after 42 weeks, and signs of perinatal asphyxia, have been seen in other studies of cerebral palsy and neonatal encephalopathy.[22-24] Two unanticipated findings among cases were increased gestational diabetes and decreased maternal fever.
Table 1. Clinical Characteristics of Mothers and Infants(*) Cases (N = 40) Maternal age, mean [+ or -] SD, y 26.8 [+ or -] 6.6 Gestational age, mean [+ or -] SD, wk 40.3 [+ or -] 1.4 Infant weight, mean [+ or -] SD, g 3381 [+ or -] 595 Placental weight, mean [+ or -] SD, g 552 [+ or -] 127 Primigravida, No. (%) 25 (63) Preeclampsia, No. (%) 3 (8) Diabetes, No. (%) 4 (10) Postdate, No. (%) 8 (20) Maternal fever, No. (%) 2 (5) Emergency cesarean section, No. (%) 22 (55) Apgar score at 1 min <6, No. (%) 37 (93) Apgar score at 5 min <6, No. (%) 30 (75) Comparison Group (N = 176) P Maternal age, mean [+ or -] SD, y 27.2 [+ or -] 7.1 .74 Gestational age, mean [+ or -] SD, wk 40.1 [+ or -] 1.1 .28 Infant weight, mean [+ or -] SD, g 3441 [+ or -] 485 .53 Placental weight, mean [+ or -] SD, g 528 [+ or -] 119 .28 Primigravida, No. (%) 73 (42) ... Preeclampsia, No. (%) 8 (5) ... Diabetes, No. (%) 3 (2) ... Postdate, No. (%) 8 (5) ... Maternal fever, No. (%) 17 (10) ... Emergency cesarean section, No. (%) 9 (5) ... Apgar score at 1 min <6, No. (%) 56 (32) ... Apgar score at 5 min <6, No. (%) 11 (6) ... OR (95% CI) Maternal age, mean [+ or -] SD, y ... Gestational age, mean [+ or -] SD, wk ... Infant weight, mean [+ or -] SD, g ... Placental weight, mean [+ or -] SD, g ... Primigravida, No. (%) 2.3 (1.2-4.6) Preeclampsia, No. (%) 1.7 (0.4-6.6) Diabetes, No. (%) 6.4 (1.6-25.0) Postdate, No. (%) 5.3 (2.0-13.8) Maternal fever, No. (%) 0.5 (0.1-2.2) Emergency cesarean section, No. (%) 22.7 (10.7-48.2) Apgar score at 1 min <6, No. (%) 26.4 (10.5-66.3) Apgar score at 5 min <6, No. (%) 45.0 (21.3-95.0)
(*) OR indicates odds ratio; CI, confidence interval; and SD, standard deviation.
The placental pathology findings are presented in Table 2. The 10 placental processes were subdivided by degree of severity, as appropriate, and their frequencies in the case and comparison groups were compared. Nine lesions were significantly more common in cases at the univariate level. These included 4 lesions that were not graded for severity (recent chorionic vessel thrombi, changes consistent with abruptio placenta, diffuse chorioamnionic hemosiderosis, and perivillous fibrin deposition); 4 lesions in which only the most severe variant was more common in cases (meconium-associated vascular necrosis, severe fetal chorioamnionitis, extensive avascular villi, and diffuse chronic villitis); and 1 lesion (increased NRBC) for which both levels of severity were significantly more common in cases, but the odds ratio was higher for the more severe form. Multiple logistic regression was performed using 8 of the 9 lesions achieving significance at the univariate level (changes consistent with abruptio placenta were excluded because this lesion did not occur in the comparison group). Three lesions were independently related to NI: severe fetal chorioamnionitis (adjusted OR, 13.2; 95% CI, 1.2-144); extensive avascular villi (adjusted OR, 9.0; 95% CI, 1.6-51); and diffuse chorioamnionic hemosiderosis (adjusted OR, 74.8; 95% CI, 6.3-894). These 3 lesions are illustrated in Figure 1.
[Figure 1 ILLUSTRATION OMITTED]
Table 2. Placental Pathology(*) Comparison Cases Group ([dagger]) ([dagger]) (N = 40) (N = 176) Meconium Amnion 4 (10) 61 (35) Amnion and decidua 13 (33) 50 (28) Green-stained umbilical cord 5 (13) 29 (17) Vascular necrosis 5 (13) 3 (2) Chorioamnionitis Maternal only 3 (8) 19 (11) Fetal (mild to moderate) 3 (8) 14 (8) Fetal (severe) 4 (10) 2 (1) Increased NRBC Mild 10 (25) 13 (7) Moderate 17 (43) 2 (1) Total 27 (68) 15 (9) Avascular villi Mild to moderate 5 (13) 10 (6) Extensive 9 (23) 6 (3) Chronic villitis Focal 2 (5) 9 (5) Diffuse 5 (13) 6 (3) Recent chorionic vessel thrombi 6 (15) 1 (1) Changes consistent with 3 (8) 0 (0) abruptio placenta Diffuse chorioamnionic 3 (8) 1 (1) hemosiderosis Perivillous fibrin 9 (23) 5 (3) Changes consistent with 9 (23) 22 (13) maternal underperfusion Unadjusted OR (95% CI) Meconium Amnion 0.2 (0.1-0.6) Amnion and decidua 1.2 (0.6-2.5) Green-stained umbilical cord 0.7 (0.3-2.0) Vascular necrosis 8.2 (2.3-29) Chorioamnionitis Maternal only 0.7 (0.2-2.4) Fetal (mild to moderate) 0.9 (0.3-3.4) Fetal (severe) 9.7 (2.3-41) Increased NRBC Mild 4.2 (1.9-9.9) Moderate 64.3 (24-171) Total 22.3 (11-46) Avascular villi Mild to moderate 2.4 (0.8-7.2) Extensive 8.2 (3.1-22) Chronic villitis Focal 1.0 (0.2-4.7) Diffuse 4.1 (1.3-13) Recent chorionic vessel thrombi 30.9 (7.3-131) Changes consistent with ([double dagger]) abruptio placenta Diffuse chorioamnionic 14.2 (2.4-84) hemosiderosis Perivillous fibrin 9.9 (3.7-27) Changes consistent with 2.0 (0.9-4.8) maternal underperfusion Adjusted OR (95% CI) Meconium Amnion ... Amnion and decidua ... Green-stained umbilical cord ... Vascular necrosis 3.9 (0.5-29) Chorioamnionitis Maternal only ... Fetal (mild to moderate) ... Fetal (severe) 13.2 (1.2-144) Increased NRBC Mild ... Moderate ... Total 27.6 (0.4-81) Avascular villi Mild to moderate ... Extensive 9.0 (1.6-51) Chronic villitis Focal ... Diffuse 1.9 (0.2-19) Recent chorionic vessel thrombi 4.9 (0.4-57) Changes consistent with ... abruptio placenta Diffuse chorioamnionic 74.8 (6.3-894) hemosiderosis Perivillous fibrin 5.9 (1.0-34.3) Changes consistent with ... maternal underperfusion
(*) OR indicates odds ratio; CI, confidence interval; and NRBC, nucleated red blood cells.
([dagger]) Values are expressed as number (%) positive.
([double dagger]) P = .006. Odds ratio could not be calculated because no placentas in the comparison group had changes consistent with abruptio placenta.
Three possible sources of bias were investigated. First, since cases were not evaluated in a blinded manner at diagnosis, it was possible that the case group was more carefully examined than the comparison group. To address this concern all available placentas from cases (n = 20) and 40 randomly selected placentas from the comparison group were rereviewed in a blinded manner for the variables significant in the initial analysis. Overall agreement ranged from 75% to 100% (cases, 86 [+ or -] 7%; comparison group, 91 [+ or -] 9%). Underdiagnosis in the first relative to the second review ranged from 0% to 24% (cases, 9 [+ or -] 7%; comparison group, 9 [+ or -] 9%), whereas overdiagnosis ranged from 0% to 10% (cases, 5 [+ or -] 5%; comparison group, 0%). All lesions remained more common in the case group than in the comparison group in the second analysis. Odds ratios were increased for 3 lesions (meconium vascular necrosis, extensive avascular villi, and perivillous fibrin), decreased for 3 lesions (increased NRBC, chorionic vessel thrombi, and diffuse villitis), and not evaluable for the other 3 lesions (no lesions in controls). A second potential source of bias was that cases reviewed for the defense might have yielded more lesions than those reviewed for the plaintiff, since, in general, placental lesions tend to strengthen the defense case. We stratified the analysis into cases reviewed for the defense (n = 18) versus plaintiff (n = 22) and found no significant difference in prevalence for any of the 9 lesions: 5 of the lesions were more common in defense cases and 4 of the lesions were more common in plaintiff cases. Finally, since the comparison group contained only meconium-stained placentas (as compared with 30/40 placentas in the case group), a third source of bias could have been introduced if the cases without meconium staining constituted a distinct group that was not controlled for in the study design. We stratified the analysis for the presence or absence of meconium staining and again found no significant difference in prevalence for any of the 9 lesions: 6 of the lesions were more common in the meconium-stained group and 3 of the lesions were more common in the non-meconium-stained group.
The effect of multiple placental lesions was studied by logistic regression using the sum of abnormalities as a predictor of NI (Figure 2). The odds ratio for each unit increase in sum was 10.1 (95% CI, 5.1-20) and the area under the receiver operating curve plotting sensitivity versus (1-specificity) was very high (0.89). Placentas from cases were less likely than those from the comparison group to have no lesions (OR, 0.04; 95% CI, 0.02-0.10) and more likely to have any 1 lesion (difference was not significant) (Table 3). Placentas with 2 or more lesions were significantly more common among cases with NI in both the original sample and the rereviewed subsample. The frequency of multiple lesions did not differ in the substrata of plaintiff versus defense or meconium-stained versus non--meconium-stained placentas (results not shown). Finally, the effect of timing was studied by comparing cases with multiple lesions believed to have their onset at different times with those with multiple lesions occurring within the same time frame. The odds ratio for NI was higher in the group with both recent and chronic lesions (OR, 94.2; 95% CI, 11.9-747) than in the groups with multiple chronic lesions (OR, 22.9; 95% CI, 1.1-487) and multiple recent lesions (OR, 43.8; 95% CI, 5.3-362).
[Figure 2 ILLUSTRATION OMITTED]
Table 3. Number and Timing of Placental Lesions(*) No. of Recent Chronic Lesions ([dagger]) ([double dagger]) 0 0 0 1 0 or 1 0 or 1 [is greater than 0 [is greater than or equal to] 2 or equal to] 2 [is greater than [is greater than 0 or equal to] 2 or equal to] 2 [is greater than [is greater than [is greater than or equal to] 2 or equal to] 1 or equal to] 1 Comparison Group No. of Cases([sections]) ([sections]) Lesions (N = 40) (N = 176) 0 5 (13)([dagger]) 141 (80) 1 11 (28) 33 (19) [is greater than 2 (5) 0 (0) or equal to] 2 [is greater than 8 (20) 1 (1) or equal to] 2 [is greater than 14 (35) 1 (1) or equal to] 2 No. of OR Lesions (95% CI) 0 0.04 (0.02-0.10) 1 1.6 (0.8-3.6) [is greater than 22.9 (1.1-487) or equal to] 2 [is greater than 43.8 (5.3-362) or equal to] 2 [is greater than 94.2 (11.9-747) or equal to] 2
(*) OR indicates odds ratio; CI, confidence interval.
([dagger]) Recent lesions (generally considered to occur within days of labor and delivery): meconium-associated vascular necrosis, severe fetal chorioamnionitis, increased nucleated red blood cells, recent chorionic vessel thrombi, and changes consistent with abruptio placenta.
([double dagger]) Chronic lesions (generally believed to occur long before labor): extensive avascular villi, diffuse chronic villitis, diffuse chorioamnionic hemosiderosis, and perivillous fibrin.
([sections]) Values are expressed as number (%) positive.
The causes of cerebral palsy and related forms of NI occurring after term birth are heterogeneous and poorly understood. Broadly speaking, at least 3 distinct processes may be distinguished: teratogenic insults (infections, toxins, dietary deficiencies, or inborn errors of metabolism) affecting primary brain development; remote antenatal hypoxic-ischemic injury leading to developmental brain abnormalities; and acute hypoxic-ischemic injury occurring at the time of labor and delivery. When labor and delivery are unremarkable, it is generally assumed that 1 or both of the first 2 processes are involved. When labor and delivery are complicated, 2 competing scenarios have been suggested: acute ischemia at birth (perinatal asphyxia) directly causing neonatal brain injury, and prior brain injury leading to signs and symptoms of perinatal asphyxia at the time of labor and delivery. The scenario that best explains the majority of cases of NI at term in the setting of apparent perinatal asphyxia has been hotly debated since the late 1800s, when the protagonists were W. J. Little, who wrote the classic description of cerebral palsy and championed the former view, and Sigmund Freud, the famed psychoanalyst and neuroanatomist, who championed the latter view. A third possibility, that multiple early and more recent insults act together to increase the risk of brain injury, has gained favor in recent years.[24,29]
Placental pathology has the potential to shed light on these issues because of its ability to identify and quantitate distinct pathologic processes occurring throughout gestation. Although many studies have investigated maternal risk factors and antenatal events as predictors of neurologic outcome in term infants, few have addressed the correlation between placental findings and neurologic outcome. Cases of well-documented, long-term impairment following term birth are sufficiently rare that only 2 approaches are feasible for studying the contribution of placental pathology. The first is a large multicenter study with uniform placental submission, such as the Collaborative Perinatal Project conducted in the 1960s.[30,31] For various reasons, including an overwhelming number of cases and a focus on placental variables that differed from the variables evaluated in this study, the relationship between placental lesions and outcome was not completely defined. The chances of repeating such a study are remote. The second approach is the one adopted in this study: retrospective review of cases ascertained on the basis of NI and comparison with a group with a low incidence of NI. In the United States today, a high percentage of cases with adverse neurologic outcomes result in litigation. Placental pathologists become involved in such cases as expert witnesses and have the opportunity to review both the original hospital records and later evaluations by child neurologists, in addition to placental slides and reports. Although subject to several possible biases that have been previously discussed, these cases provide a valuable resource.
One concern is that medicolegal cases represent only a small subgroup of infants with cerebral palsy--those with signs of severe perinatal asphyxia. In fact, this subgroup is not really that small, constituting approximately 30% of term infants with later NI in one large population-based study. An estimate of the overrepresentation of cases with severe perinatal asphyxia in our population can be obtained by comparing the prevalence of markers for asphyxia in our study with those reported in another recent population-based study of 84 infants with cerebral palsy following birth at term. Apgar scores of less than 6 at 5 minutes were assigned in 75% of our cases versus 21% of the cases in the population-based study, and an umbilical cord blood pH [is less than] 7.0 was found in 18% of our cases versus 6% of the population-based cases (data not shown). In our study, this approximately threefold enrichment in patients with signs of severe perinatal asphyxia was somewhat advantageous because the frequency of such signs in cases with significant placental lesions strengthens the view that perinatal asphyxia can be a consequence of rather than the cause of NI. On the other hand, our failure to find significant differences in placental pathology between cases with and without meconium staining suggests that the underlying causes of NI with and without signs of perinatal asphyxia may be similar.
The pathogenesis, pathophysiology, and possible mechanism of action of the 9 findings achieving significance in the univariate analysis have been discussed elsewhere.[28,33] Briefly, abruptio placenta and chorioamnionic hemosiderosis (chronic abruption) reflect interruption of maternal perfusion. Chronic villitis and perivillous fibrin effectively increase the diffusion distance between the maternal and fetal circulations. Four of the remaining 5 lesions affect large fetal vessels. Meconium-associated vascular necrosis and severe fetal chorioamnionitis cause chemical- and/or cytokine-mediated damage of the vessel wall.[11,34-38] Extensive avascular villi are related to upstream vascular occlusion, which can be caused by chorionic vessel thrombosis.[11,21] These 4 processes are all potential causes of fetal hypoxia by virtue of either obstruction or predisposition to vasospasm in major vessels that deliver fetal blood to and from the site of gas exchange. Increased NRBC, on the other hand, is distinct from the other findings because these cells are a marker for rather than a potential cause of hypoxia. Significant hypoxia leads to erythropoietin release and subsequent release of red blood cell precursors from the fetal bone marrow.
Three lesions were independently related to NI in our study. Extensive avascular villi are part of the spectrum of fetal thrombo-occlusive disease, which has been associated with NI in several previous studies.[12,14,21] Possible underlying causes include inherited thrombophilia, chronic villitis with stem villous vasculitis, and chronic stasis due to fetal positioning or umbilical cord compression.[13-15,21,39] Severe fetal chorioamnionitis in the placenta was also a risk factor for NI in our previous study of very low-birth weight infants, and maternal infection at term has been identified as a risk factor in studies without detailed placental pathology.[25,30] Diffuse chorioamnionic hemosiderosis as an indicator of chronic abruption and abnormal first- and second-trimester bleeding has also been previously identified as a risk factor for cerebral palsy. Our previous negative findings suggest that this association may apply primarily to term infants.[11,20] It is noteworthy that increased NRBC, despite a very high odds ratio in univariate analysis, was not a significant factor after logistic regression. This finding is consistent with the fact that increased NRBC are a result rather than a cause of significant hypoxia and hence may have been seen in combination with many of the other lesions studied.
Two possible criticisms of our findings are that one or more lesions may have achieved significance by chance, on the basis of multiple comparisons, and that the number of cases with each individual lesion is small, and the confidence intervals are wide. Three arguments help to mollify these valid concerns. First, the lesions studied were not randomly selected, but rather were chosen a priori, based on past work.[11-16,19-21,35] Second, although numbers were small the magnitude of the odds ratios was high. Third, a clear dose-response or threshold effect was demonstrated for each of the 5 processes where an effect was evaluated. Nevertheless, given the inherent methodologic problems and the large number of variables analyzed, the association with NI of any one of these lesions remains tentative and needs to be strengthened in future studies.
Perhaps the most important conclusion of this study relates to the number and timing of lesions in cases with NI. There was a strong and highly significant direct relationship between the number of lesions identified in a given placenta and NI. This association argues strongly for a combinatorial or synergistic relationship between lesions and is similar to our previous results in very low-birth weight infants with NI and other data relating to stillbirths.[11,40] Finally, the relationship between NI and multiple placental lesions was strongest when placentas had lesions of differing durations. This latter observation provides evidence in support of clinical, animal, and in vitro data indicating that previous insults may decrease the threshold required for more recent events to cause brain injury.[24,29,41,42]
[1.] Hagberg B, Hagberg G, Olow I, Vonwendt L. The changing panorama of cerebral palsy in Sweden. 7. Prevalence and origin in the birth year period 1987-90. Acta Paediatr. 1996;85:954-960.
[2.] Robertson CMT, Svenson LW, Joffres MR. Prevalence of cerebral palsy in Alberta. Can J Neurol Sci. 1998;25:117-122.
[3.] Pharoah POD, Cooke T, Johnson MA, King R, Mutch L. Epidemiology of cerebral palsy in England and Scotland, 1984-9. Arch Dis Child. 1998;79:F21-F25.
[4.] Nelson KB, Grether JK. Selection of neonates for neuroprotective therapies--one set of criteria applied to a population. Arch Pediatr Adolesc Med. 1999; 4:393-398.
[5.] Perlman JM. Intrapartum hypoxic-ischemic cerebral injury and subsequent cerebral palsy: medicolegal issues. Pediatrics. 1997;6:851-859.
[6.] MacLennan A. A template for defining a causal relation between acute intrapartum events and cerebral palsy: international consensus statement. BMJ. 1999;319:1054-1059.
[7.] Nelson KB, Ellenberg JH. Obstetric complications as risk factors for cerebral palsy or seizure disorders. JAMA. 1984;251:1843-1848.
[8.] Johnson A, King R. A regional register of early childhood impairments: a discussion paper. Community Med. 1989;11:352-363.
[9.] Cummins SK, Nelson KB, Grether JK, Velie EM. Cerebral palsy in four northern California counties, births 1983 through 1985. J Pediatr. 1993;123:230-237.
[10.] Gaffney G, Sellers S, Flavell V, Squier M, Johnson A. Case-control study of intrapartum care, cerebral palsy, and perinatal death. BMJ. 1994;308:743-750.
[11.] Redline RW, Wilson-Costello D, Borawski E, Fanaroff AA, Hack M. Placental lesions associated with neurologic impairment and cerebral palsy in very low-birth-weight infants. Arch Pathol Lab Med. 1998;122:1091-1098.
[12.] Sander CH, Kinnane L, Stevens NG. Hemorrhagic endovasculitis of the placenta: a clinicopathologic entity associated with adverse pregnancy outcome. Compr Ther. 1985;11:66-74.
[13.] Grafe MR. The correlation of prenatal brain damage with placental pathology. J Neuropathol Exp Neurol. 1994;53:407-415.
[14.] Kraus FT. Cerebral palsy and thrombi in placental vessels of the fetus. Hum Pathol. 1997;28:246-248.
[15.] Scher MS, Trucco GS, Beggarly ME, Steppe DA, Macpherson TA. Neonates with electrically confirmed seizures and possible placental associations. Pediatr Neurol. 1998;19:37-41.
[16.] Kraus FT, Acheen VI. Fetal thrombotic vasculopathy in the placenta: cerebral thrombi and infarcts, coagulopathies, and cerebral palsy. Hum Pathol. 1999;30:759-769.
[17.] Burke CJ, Tannenberg AE. Prenatal brain damage and placental infarction--an autopsy study. Dev Med Child Neurol. 1995;37:555-562.
[18.] Benirschke K, Kaufman P. Pathology of the Human Placenta. 3rd ed. New York, NY: Springer Verlag; 1995:290-296.
[19.] Altshuler G, Arizawa M, Molnar-Nadasdy G. Meconium-induced umbilical cord vascular necrosis and ulceration: a potential link between the placenta and poor pregnancy outcome. Obstet Gynecol. 1992;79:760-766.
[20.] Redline RW, Wilson-Costello D. Chronic peripheral separation of placenta: the significance of diffuse chorioamnionic hemosiderosis. Am J Clin Patho!. 1999; 111:804-810.
[21.] Redline RW, Pappin A. Fetal thrombotic vasculopathy: the clinical significance of extensive avascular villi. Hum Pathol. 1995;26:80-85.
[22.] Gaffney G, Flavell V, Johnson A, Squier M, Sellers S. Cerebral palsy and neonatal encephalopathy. Arch Dis Child. 1994;70:F195-F200.
[23.] Badawi N, Kurinczuk JJ, Keogh JM, et al. Antepartum risk factors for newborn encephalopathy: the Western Australian case-control study. BMJ. 1998;317: 1549-1553.
[24.] Badawi N, Kurinczuk JJ, Keogh JM, et al. Intrapartum risk factors for newborn encephalopathy: the Western Australian case-control study. BMJ. 1998;317: 1554-1558.
[25.] Grether JK, Nelson KB. Maternal infection and cerebral palsy in infants of normal birth weight. JAMA. 1997;278:207-211.
[26.] Panel of the Australian and New Zealand Perinatal Societies. The origins of cerebral palsy--a consensus statement. Med J Aust. 1995;162:85-90.
[27.] Little WJ. On the influence of abnormal parturition, difficult labour, premature birth, and asphyxia neonatorum on the mental and physical condition of the child, particularly in relation to deformities. Trans Obstet Soc London. 1862; 3:293-344.
[28.] Freud S. Infantile cerebrallabumung. Nothnagel's Specialle Pathologie und Therapie 9. Vol 12. Vienna, Austria: A. Holder.
[29.] Mallard EC, Williams CE, Johnston BM, Gluckman PD. Neuronal damage in the developing brain following intrauterine asphyxia. Reprod Fertil Dev. 1995; 7:345-351.
[30.] Nelson KB, Ellenberg JH. Antecedents of cerebral palsy. Multivariate analysis of risk. N Engl J Med. 1986;315:81-86.
[31.] Naeye RL, Peters EC, Bartholomew M, Landis RJ. Origins of cerebral palsy. Am J Dis Child. 1989;143:1154-1161.
[32.] Kaplan C, Lowell DM, Salafia C. College of American Pathologists Conference XIX on the examination of the placenta: report of the working group on the definition of structural changes associated with abnormal function in the maternal/fetal/placental unit in the second and third trimesters. Arch Pathol Lab Med. 1991;115:709-716.
[33.] Redline RW. Placental pathology. In: Fanaroff A, Martin R, eds. Neonatal-Perinatal Medicine. 6th ed. St Louis, Mo: CV Mosby; 1997:350-363.
[34.] Miller FC, Sacks DA, Yeh SY, et al. Significance of meconium during labor. Am J Obstet Gynecol. 1975;122:573-579.
[35.] Naeye RL, Localio AR. Determining the time before birth when ischemia and hypoxemia initiated cerebral palsy. Obstet Gynecol. 1995;86:713-719.
[36.] Nelson KB, Dambrosia JM, Grether JK, Phillips TM. Neonatal cytokines and coagulation factors in children with cerebral palsy. Ann Neurol. 1998;44: 665-675.
[37.] Grether JK, Nelson KB, Dambrosia JM, Phillips TM. Interferons and cerebral palsy. J Pediatr. 1999;134:324-332.
[38.] Suebji A, Altshuler G. Meconium-induced umbilical vascular necrosis in abortuses and fetuses: a histopathologic study for cytokines. Obstet Gynecol. 1999;94:415-420.
[39.] Nelson KB, Grether JK. Potentially asphyxiating conditions and spastic cerebral palsy in infants of normal birth weight. Am J Obstet Gynecol. 1998;179: 507-513.
[40.] Driscoll SG. Autopsy following stillbirth: a challenge neglected. In: Ryder OA, Byrd ML, eds. One Medicine. Berlin, Germany: Springer-Verlag; 1984:20-31.
[41.] Tan S, Zhou F, Nielsen VG, Wang ZW, Gladson CL, Parks DA. Increased injury following intermittent fetal hypoxia-reoxygenation is associated with increased free radical production in fetal rabbit brain. J Neuropathol Exp Neurol. 1999;58:972-981.
[42.] Mallard EC, Williams CE, Gunn AJ, Gluckman PD. When and why do brain cells die? Dev Med Child Neurol. 1992;34:1010-1021.
Accepted for publication July 14, 2000.
From the Departments of Pathology (Dr Redline) and Pediatrics (Ms O'Riordan), Case Western University and University Hospitals of Cleveland, Cleveland, Ohio.
Reprints: Raymond W. Redline, MD, Department of Pathology, University Hospitals of Cleveland, 11100 Euclid Ave, Cleveland, OH (e-mail: email@example.com).
|Printer friendly Cite/link Email Feedback|
|Author:||Redline, Raymond W.; O'Riordan, Mary Ann|
|Publication:||Archives of Pathology & Laboratory Medicine|
|Date:||Dec 1, 2000|
|Previous Article:||Parotid Gland Malignant Melanomas.|
|Next Article:||Utility of Flow Cytometry Immunophenotyping for the Diagnosis and Classification of Lymphoma in Community Hospital Clinical Needle...|