Cervical conization doubles the risk of preterm and very preterm birth in assisted reproductive technology twin pregnancies

STUDY QUESTION: Does cervical conization add an additional risk of preterm birth (PTB) in assisted reproduction technology (ART) singleton and twin pregnancies? SUMMARY ANSWER: Cervical conization doubles the risk of preterm and very PTB in ART twin pregnancies. WHAT IS KNOWN ALREADY: ART and cervical conization are both risk factors for PTB. STUDY DESIGN, SIZE, DURATION: In this national population-based controlled cohort study, we included all ART singletons and twin deliveries from 1995 to 2009 in Denmark by cross-linkage of maternal and child data from the National IVF register and the Medical Birth register. Furthermore, control groups of naturally conceived (NC) singletons and twins were extracted. Cervical diagnoses were obtained from the Danish Pathology register. Cervical conization included both cold knife cone and LEEP (loop electrosurgical excision procedure) but not cervical biopsies. The main outcomes measures were PTB (PTB <= 37 + 0 gestational weeks), very preterm birth (VPTB <= 32 + 0 gestational weeks) and preterm premature rupture of membranes (PPROM). PARTICIPANTS/MATERIALS, SETTING, METHODS: In all 16 923 ART singletons and 4829 ART twin deliveries were included. A random sample of NC singletons, 2-fold the size of the ART singleton group matched by date and year of birth (n = 33 835) and all NC twin deliveries (n = 15 112), was also extracted. Multiple logistic regression analyses were performed to adjust for the following confounders: maternal age, parity, year of child birth and sex of child. MAIN RESULTS AND THE ROLE OF CHANCE: Cervical morbidity (dysplasia and conization) was more often observed in ART pregnancies (6.2% of ART singletons and 5.4% ART twins) than in NC pregnancies (4.2% for NC singletons and 4.5% for NC twins), both for singletons and twins. In ART singleton deliveries, the PTB rate was 13.1 versus 8.2% in women with and without conization, respectively, with an adjusted odds ratio (aOR) of 1.56 [95% confidence interval (Cl) 1.21-2.01]. In ART twin deliveries, the prevalence of PTB was 58.2 versus 41.3% in women with and without conization, respectively, with an aOR 1.94 (95% CI 1.36-2.77), and the risk of VPTB was also doubled. Furthermore, previous dysplasia (without conization) increased the risk of VPTB in ART twins (aOR 1.74,95% CI 1.04-2.94). Cervical dysplasia did not increase the risk of any of the other adverse outcomes in ART singletons or twins. The risk of PPROM was increased in both in ART and NC singleton deliveries with conizadon versus no conization; however, this increased risk of PPROM after conization was not observed in either ART or NC twin pregnancies. LIMITATIONS, REASONS FOR CAUTION: We were not able to adjust for the height of the cervical cone or the severity of the cervical intraepithelial neoplasia (CIN) or the time window between diagnosis of CIN and ART treatment. The finding on an increased risk of VPTB in ART twin pregnancies after dysplasia without conization may be random as we found no other increased risk after dysplasia alone either in singletons or in twins. WIDER IMPLICATIONS OF THE FINDINGS: After ART and prior conization, 58% of twin pregnancies versus 13% of ART singleton pregnancies result in PT. There is a doubled risk of preterm delivery in ART twins with conization versus ART twins with no prior conization. Single-embryo transfer should always be recommended in women with prior conization irrespective of female age, embryo quality and prior number of ART attempts.