We read with interest the article by Kaempf et al.,1 which suggests that a ‘less aggressive’ approach to the management of the patent ductus arteriosus (PDA) may be acceptable. Their approach is based on a strategy of modest fluid restriction and relatively more rigorous clinical and echocardiography criteria for diagnosing significant PDA, which were then selected for treatment with indomethacin and surgical ligation. The authors report a reduction in indomethacin use, older age at indomethacin commencement and surgical ligation, but rates of surgical ligation remained unchanged. They conclude that despite this conservative approach to management of the ductus arteriosus, individual neonatal outcomes were not adversely affected, but an increase was seen in the combined outcome of mortality after 7 days and chronic lung disease.
We compliment the authors for sharing their experience of rationalizing PDA treatment and hence providing further evidence that not all patients with a PDA require early ductal closure. It is our opinion that the impact of PDA on patient outcomes is more likely to be a function of magnitude of shunt, rather than its mere presence. In this letter, we will like to highlight certain methodological and interpretative concerns regarding the data in this report. First, number of infants in era 2 is almost half of era 1, making comparison challenging. It can be argued that the study had insufficient power to reach statistical significance for various clinically relevant outcomes. For instance, the rates of pulmonary hemorrhage more than doubled, spontaneous perforation halved but mortality rose by 62% in era 2, none of which reached statistical significance. A carefully matched case–control study could have yielded a more appropriate comparison. Second, in more than 90% of infants, the diagnosis of PDA was confirmed using echocardiography but authors failed to provide any data to demonstrate that characteristics of the ductus arteriosus, particularly those related to the magnitude of the shunt, were comparable between epochs. The high baseline rate of indomethacin use during era 1 (79% of all very low birth-weight infants) is suggestive that some patients with small restrictive ducts or with low-volume shunts may have been treated. This is further supported by their finding of similar rates of PDA ligation in both eras. It could be argued that a more refined approach to treatment, based on enhanced definition of the hemodynamic impact of the shunt using clinical and echocardiography criteria, may have resulted in a better balance between avoidance of unnecessary treatment when the ductus is patent but the shunt volume is low and the approach to minimize morbidities associated with prolong exposure to high-magnitude ductal shunt. Third, the authors report a lack of change in grade III/IV intraventricular hemorrhage between epochs. This is not surprising, as the age at initiation of indomethacin in both eras was beyond the point at which intervention could have had an impact. We hypothesize that the reduction in intraventricular hemorrhage seen in trials of prophylactic indomethacin, early intervention trials and epoch-based reports of early targeted intervention relates to timely optimization of systemic hemodynamics.2, 3, 4 Prospective observational physiologyical studies of the transitional circulation demonstrate low systemic blood flow in the first few hours of life leading to relative cerebral hypo-perfusion.5 As pulmonary vascular resistance falls, with consequential increased transductal resistance, the magnitude of the ductal shut increases leading to higher left heart preload and relatively acute rise in preductal cardiac output. This may create the potential of a cerebral hypo-hyperperfusion state which, depending on the patient’s cerebral auto regulatory capacity, may contribute, at least in part, to the nature and extent of intraventricular hemorrhage. Finally, one of the rationales mentioned by authors for adopting the more tolerant approach was the associated risks of treatment. Considering the significant reduction in use of indomethacin (79% vs 26%), one would have expected a reduction in treatment-associated morbidities, but data is not provided to support this counter argument to use of indomethacin. Despite the above-mentioned methodological and interpretative limitations of this study, there remains an important non-reassuring association of increase in the composite outcome of chronic lung disease or mortality after day 7, which may support the notion that there are ‘some patients’ who may benefit from early ductal closure. The recommendation of modest fluid restriction as an alternative to ductal closure is premature and somewhat concerning. Biologically, such an approach may reduce pulmonary overcirculation but at the expense of compromising end-organ blood flow in patients already suffering from the ‘ductal steal’. The recent article by Buyst et al.6 emphasizes this point, where fluid restriction as a exclusive treatment strategy led to compromised systemic blood flow without any demonstrable beneficial effect on pulmonary hemodynamics or lung compliance.
This is a preview of subscription content, access via your institution
