Introduction

The early diagnosis of any congenital defect in the fetus is of vital importance, as it enables the implementation of the most appropriate measures, both during pregnancy and during childbirth, as a means of preventing unnecessary risks to the mother and child and of attempting to improve the prognosis of the newborn after birth¹.

All pregnant women in the first trimester are offered prenatal aneuploidy screening, which estimates the individual risk of chromosomopathy, combining the a priori risk due to the mother´s age with the biochemical analysis of PAPP-A and maternal blood and free β-HCG fraction and with an ultrasound measurement of Nuchal Translucency (NT)²,³.

Previous studies have revealed that the biochemical and ultrasound markers of first-trimester combined screening, abnormal PAPP-A, βHCG and NT in pregnancies without aneuploidy can be predictive of adverse perinatal outcomes⁴-⁶.

Elevated NT with normal karyotype is associated with fetal structural anomalies and several genetic syndromes, including heart defects, Noonan syndrome, Smith-Lemli-Opitz syndrome and skeletal dysplasias of the achondroplastic type and thanatophoric displasia⁷-¹².

Low PAPP-A is associated with placental defects and has been associated with abortions, antepartum fetal death, restricted intrauterine growth, hypertensive disease of pregnancy, prematurity, and low birth weight.

Moreover, due to the association with abnormal placentation, high elevated βHCG is associated with preeclampsia, restricted intrauterine growth, antepartum fetal death and prematurity¹³-¹⁶.

The aim of our study is to find out if there are any significant differences in adverse obstetric and perinatal outcomes in pregnant women with positive combined screening and normal karyotype in the group of cases, by risk group (0-50, 51-100, 101-150, 151-200, 201-250), comparing them with the control group.

Material and methods

Using the medical records database, we conducted a retrospective study of cases and controls in pregnant women that underwent combined screening in the first trimester in the prenatal diagnosis section of the department of obstetrics and gynecology at Miguel Servet University Hospital, Zaragoza, in the period from January to December 2012, both months included.

The following inclusion criteria were used: single pregnancies subjected to combined first-trimester screening and follow-up of pregnancy and delivery at our hospital, whereby the following were excluded from the study: twin pregnancies with aneuploidy and/or fetal malformations, and cases with an absence of data related to the mother and/or newborn.

To determine biochemical free β-HCG and PAPP-A, all the pregnant women were scheduled for blood collection between 9+^0]-10+^{0] week}s, and for an ultrasound measurement of nuchal translucency between 11+^0] -13+^{6 week}s and with a skull-caudal length (LCC) of between 45 and 84 mm.

The IMMULITE^® 2000 automatic analyser (Siemens Healthcare Diagnostics) was used for the measurement of the biochemical parameters. The PAPP-A values in mIU/L and β-HCG values in ng/ml (with the equivalence ng/ml = IU/L), were converted into MoM, dividing the individual value of the marker by the value of the population median for the gestational age expressed in days.

The same plane as for the LCC was used to measure Nuchal Translucency (NT), with the fetus in a neutral position in accordance with the Fetal Medicine Foundation (FMF) standard https://fetalmedicine.org/education/the-11-13-weeks-scan.

A median of MoM NT of between 0.90 - 1.10 MoM, or the standard deviation of the log10 MoM NT, of between 0.08-0.13, was regarded as adequate. Prisca software version 639 – 0106 2012 was used to calculate the risk.

All the pregnant women with positive combined screening, risk >1/250 and normal karyotype were selected from the case group, who were divided into risk groups: 1-50, 51-100, 101-150, 151-200, 201-250 and the pregnant women with normal combined screening, risk < 1/14000 in the control group. Data on maternal backgrounds, delivery characteristics and neonatal morbidity was collected.

The gestational age of newborns was measured in days and was calculated through an ultrasound performed in the first trimester of pregnancy, or based on the date of the last period in which the gestational age did not changed by more than two standard deviations (< 6 days).

Preterm birth refers to births that occurred spontaneously or were induced before 36⁶ weeks of gestation (≤ 259 days)¹⁷.

The weight of newborns was calculated in grams (g) at the time of delivery. The following category variables were taken into account: Low weight: Percentile (p < 10), Normal weight: Percentile: P10 – 90 and Large for gestational age (LGE): Percentile: p > 90. Weight according to gestational age was estimated using the Spanish tables of neonatal weights of the Spanish Society of Obstetrics and Gynecology (SEGO).

The SPSS programme was used for the statistical analysis of data. The description of the quantitative variables was carried out using mean and standard deviation. The chi-square test or Fisher’s exact test was used to analyse the qualitative variables. Multivariate logistic regression analysis was used to assess the prediction in the case group and by risk group, in addition to the control group of the outcome variables. p < 0.05 was regarded as significant.

Only pregnant women that gave their consent to perform combined screening in the first trimester participated in the study, pursuant to the standards of the Society of Gynecology and Obstetrics (SEGO) and our hospital´s quality commission.

Results

4,217 deliveries and 3,791 first-trimester combined screenings (FTCS) were performed at our hospital in 2012, with an FTCS/number of deliveries ratio of 89.9%. Of the total number of FTCS performed, 3515 FTCS were negative (92.7%) and 276 FTCS were positive (7.3%). Of the population screened in the year 2012, there were 30 true positives (TP), 246 false positives (FP), 3514 true negatives (TN) and 1 false negative (diagnosis through out-of-hospital amniocentesis). The sensitivity and specificity of FTCS in our sample were 96.8% and 93.5% respectively, with a false positive rate of 6%.

There were 219 (39.9%) pregnant women with positive combined screening and normal karyotype in the study period. A control group from the same period was selected with 330 (15.7%) pregnant women with similar demographic characteristics.

Of the 549 single pregnancies included in the study, 30 (5.7%) were born prematurely and 54 (10.2%) were born with low birth weight (percentile <10). There was a greater likelihood in the case group of newborns with low birth weight (AOR = 2.642, p = 0.013) and of prematurity (OR = 2.170, p = 0.041) and in the full-term newborn control group (AOR = 2.265). By risk group, births with low weight were more likely in the 1-50 group (AOR = 2.269, p = 0.036) and in the 100-150 group (AOR = 3.291, p = 0.008) (Table 1).

52 newborns (9.8%) were admitted and observed, with 36 newborns (17.6%) in the case group and 16 newborns (4.9%) in the control group (p < 0.001). Of the total admissions, 11 newborns (2.1%) were admitted to the ICU, 17 (3.2%) were admitted due to prematurity, 12 (2.3%) were admitted due to low weight and 5 (3%) were admitted due to prematurity-related complications. The prematurity-related complications observed in the newborns in question were: wet lung/lung maladaptation, hyaline membrane, bronchopulmonary dysplasia, retinopathy of prematurity and retinal hemorrhage.

We found a greater likelihood of admission to the ICU in the case group (OR = 7.4, p = 0.011) and to the Neonates department (AOR = 8.120, p = 0.008), the most likely causes of admission being prematurity (OR = 3.979, p = 0.011), low weight (OR = 18.371, p = 0.005), jaundice (OR = 4.808, p = 0.002), and respiratory distress (OR = 4.808, p = 0.001). Admission due to perinatal infection was more likely in the control group (AOR = 6.554) (Table 2).

The likelihood of the hospitalisation of newborns by risk group was: 1-50 group admission to the Neonates department (OR = 2.472, p = 0.019), due to prematurity (OR = 4.13, p = 0.01), due to prematurity-related complications (OR = 6.333, p = 0.046), due to low weight (AOR = 6.322, p = 0.049) and due to respiratory distress (OR = 4.8, p = 0.001); in the 50-100 segment admission due to jaundice (AOR = 3.278, p = 0.035), in the 100-150 group, admission to the neonates unit (OR = 2.704, p = 0.028), due to low weight (OR = 2.704, p = 0.015) and perinatal infection (AOR = 5.505, p = 0.029); in the 150-200 group, due to respiratory distress (AOR = 5.813, p = 0.01) and in the 200-250 group, admission to the ICU (AOR = 16.8, p = 0.034) and due to respiratory distress (OR = 3.125, p = 0.048) (Table 3).

Discussion

The “Inversion of the antenatal control pyramid” concept is a model of prenatal care in which, gathering epidemiological, clinical, biophysical, ultrasound and analytical data between 11 and 14 weeks of gestation enables us to individualise the risk of each pregnant woman for a wide spectrum of pathological conditions (fetal death and spontaneous abortion, premature birth, preeclampsia, gestational diabetes, fetal growth restriction, macrosomia), thereby establishing preventive measures and, in some cases, therapeutic measures¹⁸-²⁰.

Prematurity has major health implications due to high fetal morbidity and mortality, particularly at extreme gestational ages. Prematurity is the most frequent cause of perinatal mortality (2/3) and can cause significant long-term morbidity due to neurological, gastrointestinal and respiratory disorders²¹,²².

The prevalence of prematurity in Spain is 9.5%¹⁷, with the rate of late prematurity (34 to 36 weeks of gestation) being far more prevalent both in Europe and in the USA, where it can vary from 50% to 75%²³,²⁴.

Beta et al. conducted a predictive model of preterm delivery with the biochemical markers of the 1^st trimester, maternal characteristics and obstetric history and conclude that preterm delivery could be detected with a rate of 20% in nulliparous women and 38% in women with a history of preterm delivery ≥ 16 weeks, with 10% false positives²⁵. Spencer et al. conclude that low PAPP-A increases the risk of prematurity, but the same does not apply to βHCG. The lower the PAPP-A, the greater the risk of prematurity, even so the detection rate for low PAPP-A does not exceed 10%²⁶. Dugoff et al, in the FASTER study, found a statistical association of PAPP-A < P5 with prematurity⁵. Prematurity was found in 6% of pregnancies in our study, with a greater probability in the case group (OR = 2).

Several studies report higher morbidity²⁷ and perinatal mortality in newborns with low birth weight < P10²⁸.

Morris et al. found an association between low weight < P10 and PAPP-A < P5 with a likelihood OR 2.08 (CI 1.89-2.29), with a prediction rate of 13%, for 7% of false positives²⁹.

Peterson et al. conclude that PAPP-A is positively correlated with birth weight, with a higher risk of low weight for gestational age with PAPP-A < 0.57 MoM³⁰.

A study conducted by Spencer et al. concludes that fetuses with a low weight for their gestational age are associated with lower medians of PAPP-A²⁶.

In our study, 2% of newborns weighing less than the 10^th percentile were identified, being more likely in the case group (AOR = 2.642) and in the 1 to 50 (AOR = 2.269) and 100 to 150 (AOR = 3.291) risk groups.

With regard to morbidity in newborns, Kirkegaard et al. associate low levels of PAPP-A < 4 MoM and βHCG with a higher rate of admissions to neonatal intensive care units (ICU), regardless of gestational age and low weight³¹.

In our study, there was a greater probability of hospitalisation in the neonates unit and the ICU in the case group and in the 1 to 50 and 50 to 100 risk groups, with the most likely cause of admission being prematurity (OR = 4.13), prematurity-related complications (OR = 6.333) and low birth weight (AOR = 6.322). Several studies have followed up on adverse obstetric outcomes and the relationship with biochemical and ultrasound parameters in the 1^st trimester, but in general they have all been used on an isolated basis and therefore have less sensitivity and predictive value.

Our study has the major advantage of having used combined 1^st trimester screening as a predictive method of adverse obstetric outcomes during pregnancy, which increases the likelihood of detecting pregnant women with a higher risk of developing pathologies, which may enable us to act from the beginning of pregnancy, modifying the risk (for example, the administration of aspirin in platelet dysfunction) or conducting a close follow-up procedure with predictable action.

One limitation of our study is that the relevance of the predictive values is limited due to the fact it is a case-control study. Nevertheless, we have been able to clearly illustrate the importance of first-trimester combined screening as a model for predicting adverse perinatal outcomes, such as prematurity, low weight and admissions to the ICU and neonates unit.

Conclusion

Performing prenatal screening for aneuploidy, from 9+^0] to 13+^{6 week}s of gestation, is of paramount importance, not only in screening for Down syndrome, but also for the early prediction of pregnant women at a high risk of developing adverse fetal events such as: prematurity, low weight for gestational age and neonatal admissions.