PERINATAL OUTCOMES IN CYSTIC FIBROSIS WOMEN: DATA FROM THE ITALIAN CYSTIC FIBROSIS REGISTRY.

Data from the Italian Cystic Fibrosis Registry (ICFR) concerning pregnancies in the period 2010-2015 were used to investigate the association between the pre-conception clinical status and perinatal outcomes of women with CF. The assessed clinical variables were genotype, pre-conception age, the body mass index (BMI), and the percentage of predicted forced expiratory volume in one second (ppFEV 1 ); the analysed outcomes were gestational age, birth weight (BW), and the frequency of caesarean deliveries. Complete information concerning mother and child was available for 56 pregnancies that came to term. The median pre-conception age was 30.8 years (18.7-42.3); median BMI was 21.5 kg/m 2 (16.5-26.8); and median ppFEV 1 was 73.9% (30-128%). Twenty-four women (42.8%) carried class I-III mutations, and 25 (44.8%) at least one mutation conferring residual CFTR function or having variable clinical consequences. The median duration of pregnancy was 37 weeks (31-41), and the frequency of prematurity was 28.30%. Median BW was 2,910 g (1,300-3,650). There was a positive correlation between BMI and ppFEV 1 in comparison with BW ( r =0.29 and r =0.31 by p=0.048 and p=0.032, respectively), but no correlation between age and BW. The overall frequency of caesarean sections was 63.64%, but it was 77.27% among the women with severe genotypes. A lower ppFEV 1 corresponded to a lower BW (p=0.047). Respiratory function had no effect on the percentage of caesarean deliveries. The nutritional status and respiratory function of CF mothers correlate with their newborn’s BW. Caesarean sections are very frequent, even in young women with normal lung function.


INTRODUCTION
Cystic Fibrosis (CF) is the most frequent autosomal recessive disease among Caucasians due to alterations in the CF transmembrane regulator (CFTR) gene. This multi-organ disease reduces the life expectancy of affected subjects. There have been noticeable improvements in the prognosis of CF patients over the last 20 years due to earlier diagnoses as a result of neonatal screening, improved patient management at specialised Centres, and the use of treatments that improve CFTR protein function, all of which have led to a median predicted survival of >50 years for subjects born after 2000 [1] and an increase in the adult CF population [2,3].
The first report of a full-term pregnancy in a woman with CF was published in 1960 [4], when early death at pre-school age was prevalent. However, the subsequent improvements in diagnostic and therapeutic methods leading to longer and healthier life have meant that women with CF could plan a pregnancy as a real possibility. A number of reports [5][6][7][8][9][10] have documented the effect of pregnancy on the clinical status of women with CF, but less is known about perinatal outcomes [11][12][13]. Other than the annual number of pregnancies, the American, British, Canadian and Australian CF registries provide no information in terms of gestational age or type of labour; only the French registry reports some data concerning perinatal outcomes [14,15].
Clinicians are consequently very interested in learning more about the management and outcomes of pregnancy in fertile women with airway disease and possibly reduced respiratory function [16], and so the aim of this study was to investigate the perinatal outcomes of women with CF who gave birth in the period 2010-2015 using data from the Italian CF Registry (ICFR).

METHODS
In order to evaluate the effects of the mothers' clinical status on perinatal outcomes, we retrospectively studied the clinical data of women with CF who had at least one pregnancy during the study period of 2010-2015. The considered variables were genotype, age at pregnancy, lung function as the percentage of predicted of forced expiratory volume in one second (ppFEV1), and nutritional status in terms of the body mass index (BMI) in the year of conception of their first pregnancy, and the analyzed outcomes were the type of labour, and the newborns' gestational age and birth weight (BW). Missing data were acquired by means of a dedicated phone interview and collected from each CF Centre.
All of the patients had given their informed consent before being included in the ICFR (http://www.registroitalianofibrosicistica.it/), and this study of anonymous patient data was approved by the Scientific and Steering Committees of the ICFR on 27 February 2017. The terms of use of the provided data are governed by the Italian law in accordance with European data protection legislation.

Statistical analysis
The continuous variables are recorded as mean values ± standard deviation (SD) when they were normally distributed, verified using the Shapiro-Wilk test, and as median values and ranges when they were not, and the correlations between them were measured using Spearman's coefficient. The application of the Box-Cox procedure not identified a suitable mathematical function to make the data normal. Comparisons between the continuous variables and subjects with absent/minimal function vs. those with residual function/variants with varying clinical consequences were made using the t test, normally distributed data, or the Mann-Whitney test, not normally distributed data.
Differences in the continuous variables in relation to ppFEV1 levels were tested using the Kruskal-Wallis test and the post hoc analysis was made using the Steel-Dwass test. Homoscedasticity was verified by Levene and Brown-Forsythe tests.
The categorical variables are recorded as absolute numbers and percentages, and the associations between them were evaluated using the c 2 test. The Cochran-Armitage test was used to verify the associations and trends of the categorical variables of ppFEV1 levels and chronic Pseudomonas aeruginosa infection.
The statistical analyses were made using SAS v. 9.4 and JMP v. 15 software (SAS Institute Inc., Cary, NC, USA). A p-value≤0.05 was considered statistically significant.

RESULTS
During the study period, 81 women underwent a total of 84 pregnancies; nine women (11.1%) experienced pregnancy interruptions, equally distributed between miscarriages and abortions.
Complete information concerning mother and child was available for 56 pregnancies that came to term: 52 single pregnancies and four twin pregnancies. The median pre-conception age of the women was 30.8 years (18.7-42.3); their median BMI was 21.5 kg/m 2 (16.5-26.8); and their median ppFEV1 was 73.9% (30-128%). Twenty-four women (42.8%) had two mutations leading to minimal CFTR function; 25 (44.8%) had at least one mutation conferring residual function (RF) or a variant with varying clinical consequences (VVCC); and remaining seven (12.5%) carried CFTR alterations of undetermined function.
Median gestational age was 37 weeks (31-41), and median birth weight was 2,910 grams (1,300-3,650). The frequency of prematurity (gestational age <37 weeks) was 28.30%, and the overall frequency of caesarean section was 63.64%. One newborn was diagnosed as having CF. Tables 1 and 2 show the mothers' clinical data and perinatal outcomes by CFTR function (minimal vs. RF/VVCC). The between-group difference in ppFEV1 was significant (p=0.0032), but there was no difference in perinatal outcomes except for the greater frequency of caesarian deliveries among the women with minimal CFTR function (77.27 vs. 47.83; p=0.042).   Tables 5 and 6 show the effect of ppFEV1 on the mothers' condition and perinatal outcomes. A younger age and the absence of Pseudomonas aeruginosa chronic infection significantly correlated with a better ppFEV1 (p=0.038 and p<0.0001). Post hoc analysis showed a significantly older maternal age between the mothers with a ppFEV1 of 40-70% (32.5 years) and those with a ppFEV1 of 70-90% (27 years) (p=0.046) (Fig. 1A).
There was a positive correlation between maternal BMI and ppFEV1 in comparison with BW (r=0.29 and r=0.31 by p=0.048 and p=0.032, respectively), but no correlation between maternal age and BW. BW was distributed by ppFEV1 class (p=0.047): Post hoc analysis showed a difference between the BW of the newborns of mothers with a ppFEV1 of <40% and that of the newborns of mothers with a ppFEV1 of >90% (p=0.041) (Fig. 1B). No significant relationship was found between the frequency of caesarean sections and the severity of lung function impairment.

DISCUSSION
As an extension of a study of pregnancies in Italian CF women [10], this paper describes for the first time the good perinatal outcomes documented by the Italian CF Registry (ICFR) data of 2010-

The previous study evaluated the effects of pregnancy on the women's clinical condition, and
showed that the majority of Italian CF women starting a pregnancy became mothers regardless of the severity of their genotype or lung function impairment; the only adverse effect was on their nutritional status. This suggests that pregnancy is feasible in CF women, and can be successfully managed by the integrated multidisciplinary care provided by specialised CF centres following specific recommendations [17].
The data analysed in the present study show that neither severe lung disease nor malnutrition affect successful pregnancies in Italian women with CF, and that most pregnancies come to term and lead to the birth of babies with a good neonatal weight.
Multiple logistic regression analyses of the data comparing more than one thousand pregnancies in CF women with pregnancies in non-CF women in the largest study of pregnancy outcomes [18] showed that the former were more likely to go into premature labour. Depending on the characteristics of the studied populations and the survey period, it has been reported that between 8% and 46% of the pregnancies of CF women end in premature births [11]. The frequency of prematurity (28.3%) in our population is in line with range, and significantly higher than the 7% observed in the general Italian population [19]. Premature birth usually affects birth weight but, despite the high percentage of prematurity in our pregnant CF women, the average birth weight is in line with that reported by Girault et. al [11]. Use of the general linear model (GLM) showed that prematurity was not related to the mothers' pre-conception respiratory function but, in addition to spontaneous pre-term labour, may be due to prudential medical indications for anticipating childbirth in order to avoid respiratory worsening.
In relation to the type of birth, it has been reported that 41.9% of CF women with a ppFEV1 of <50% [15] and 45.4% of those giving birth prematurely [13] undergo caesarean sections, whereas we found that a much higher proportion of cases (63.64%) were managed by means of a caesarean delivery. Our data show that there is no association between caesarean delivery and the severity of pre-conception respiratory function impairment, the age of the mother or her nutritional status by genotype, but there is a more significant association with a severe genotype than with the presence of at least one RF or VCC variant (p = 0.042).
The rate of caesarean deliveries in Italy is generally very high (up to 35%) [19], and this may be reflected in the case of CF women. However, a high percentage was also observed among younger women, those with normal respiratory function, and those with full-term pregnancies, which may be due to the widely held opinion of medical staff that CF women poorly tolerate vaginal birth.
Our study refers to a period preceding the availability of CFTR modulators in Italy, whereas a number of cases of successfully completed pregnancies in women receiving these drugs have recently been reported [20]. It is therefore realistic to believe that the innovative treatments of precision medicine will improve the clinical status and fertility of many CF women, thus making it easier for them to become mothers. One of the main aims of CF registers in the future will be to follow the pregnancy outcomes of women receiving CFTR modulators. [21] In conclusion, analysis of the ICFR data relating to this large cohort of mother-child dyads showed that CF women can successfully undergo pregnancy, and that this often leads to the birth a full-term newborn of adequate weight. The very high percentage of caesarean sections transversally related to various factors is often an a priori choice rather than being determined by the clinical condition of the mother, and deserves further investigation.
We would also like to thank Kevin Smart for the linguistic revision of the manuscript.

Availability of data
The dataset analyzed during the current study is available from the corresponding author on reasonable request.