Precision medicine in neonatal hemodynamics: Need for prioritization of mechanism of illness and defining population of interest

Precision medicine in neonatal hemodynamics: Need for prioritization of mechanism of illness and defining population of interest

Manuscript citation: Serrano RM, Madison M, Lorant D, Hoyer M, Alexy R. Comparison of ‘post-patent ductus arteriosus syndrome’ in premature infants after surgical ligation vs. percutaneous closure. J Perinatol. 2020;40(2):324-329.

Reviewed by: Adrianne Rahde Bischoff¹, MD, Regan E Giesinger¹, MD, Edward F Bell¹, MD and Patrick J McNamara¹, MD

¹ Division of Neonatology, Department of Pediatrics, University of Iowa Stead Family Children’s Hospital

Correspondence:

Patrick J McNamara, MD

Neonatology Division Director, University of Iowa Stead Family Children’s Hospital

Professor of Pediatrics and Internal Medicine, University of Iowa

8803 JPP, 200 Hawkins Drive

Iowa City, IA, 52242

T: 319-467-7435

[email protected]

Keywords: patent ductus arteriosus, post-ligation cardiac syndrome,

Type of investigation: Retrospective and prospective cohort study

Question: In very low birth weight (VLBW) infants, is percutaneous closure of the patent ductus arteriosus, compared with surgical ligation, associated with decreased incidence of post-patent ductus arteriosus (PDA) ligation syndrome in the first 24 hours after PDA closure?

Methods:

Design: Retrospective and prospective single center cohort study

Patients:

• Inclusion criteria: birth weight <1500g who underwent surgical ligation (retrospective) or percutaneous closure of the PDA (prospective)
• Exclusion criteria: additional hemodynamically significant congenital heart disease

Intervention: Surgical vs. percutaneous closure of the PDA

Outcomes:

• Primary outcome: Incidence of post-PDA ligation syndrome, defined as the need for hemodynamic support secondary to instability in the first 24 hours after PDA closure.
• Secondary outcomes: Changes in post-procedure variables including mean arterial pressure, FiO₂, heart rate and mean airway pressure.

Analysis and sample size: The study design was a convenience-sample chart review with no pre-study power or sample size calculations. The study had 80% power to detect effect size of 0.75 with 59 patients included in the surgical group and 24 patients in the percutaneous group, assuming a two-sided significance level of 5%. Categorical and continuous pre-procedure variable differences between the groups were assessed using Fisher’s exact test and Wilcoxon rank sum test, respectively. Analysis of covariance was used to analyze changes in variables pre- and post-procedure as well as to compare surgical patients who needed hemodynamic support in the first 24 hours post-ligation versus those who did not.

Main results: There were no significant differences in gestational age (25.9 vs 25.2 weeks), birthweight (0.79 vs 0.76 kg), or age and weight at intervention (58.5 vs 58.2 days and 1.93 vs 1.74 kg, respectively) between the groups. There were no differences in pre-procedure vital signs, fraction of inspired oxygen (FiO₂), mode of ventilation, need for hemodynamic support, or PDA size and hemodynamic significance based on the McNamara score. Despite no differences in post-intervention mean arterial pressure or heart rate, 12/59 (20%) of the infants in the surgical group required initiation of hemodynamic support in the first 24 hours versus none in the percutaneous group (p=0.016). Hydrocortisone was used in 8/12 (66%) patients who received hemodynamic support. Infants undergoing surgery had a greater absolute increase in FiO₂ post-procedure (0.23 vs 0.09) (p=0.008). Subgroup analysis showed that surgical patients requiring hemodynamic support had lower mean arterial pressure both before the procedure (45.2 vs 53.5 mmHg) and prior to initiation of support (described as mean of 39.6 mmHg) when compared to those surgical patients not receiving hemodynamic support. Absolute values and post-procedural changes in mean airway pressure and maximum FiO₂ were not statistically different between surgical patients who required hemodynamic support and those who did not. Decreased left ventricular function as assessed by echocardiogram was identified but not treated in 9/24 (37%) patients in the percutaneous group during the first 24 hours and spontaneous resolved on repeat assessment. 

Study Conclusion: The study concluded that VLBW infants who underwent percutaneous PDA closure had no immediate hemodynamic instability compared to 20% of historical controls who underwent surgical ligation.

Commentary: The conclusion of the study by Serrano et al¹ does not fully define the potential superiority of percutaneous closure over surgical ligation related to the incidence of post-ligation cardiorespiratory instability in light of the population that was included. Percutaneous closure of the PDA is becoming a widely used procedure, particularly since the introduction of newer devices that are suitable for <1000 g infants.² There is growing interest in the benefits of a less invasive approach, although comparisons between cohorts of surgical and percutaneous-based closures are still scarce in the literature, particularly as they pertains to post-ligation cardiorespiratory instability. Serrano and colleagues’ study compares the use of hemodynamic support in the post-ligation period between a historical cohort of surgically ligated VLBW infants and a prospective cohort of infants undergoing percutaneous-based closure.¹ The authors have performed a thoughtful comparison between the two techniques. However, the importance of the “population of interest” needs to be discussed. The latter is relevant in clinical conditions where the mechanism of illness, in this situation post-operative LV dysfunction, is related to maturation and postnatal age.

Post-ligation cardiac syndrome (PLCS) is defined as systolic arterial pressure less than the third percentile requiring hemodynamic support in the setting of ventilation and oxygenation failure. The incidence of PLCS ranges between 28 and 45%.^{3, 4, 5} At the time of PDA closure, sudden removal of the lower pressure pulmonary circuit significantly increases left ventricular afterload. The immature myocardium has diminished capability to adapt to such changes, and left heart systolic and diastolic dysfunction may result.⁶ A more careful evaluation of the data shows that the incidence of PLCS after surgery varies from as high as 60% when PDA closure is done in the first week of life to around 10% when performed after 4 weeks of age (Figure 1)⁷ and that the true phenotype of PLCS is most commonly seen in infants <1000 g at the time of the procedure.⁸

Very few reports describe the incidence and biologic nature of PLCS after percutaneous closure of the PDA. While two previous groups have reported faster recovery time and improvement in respiratory status,^{9, 10} to our knowledge only one study prior to Serrano et al has reported changes in left ventricular function post-percutaneous PDA closure.¹¹ Careful evaluation of the study data is warranted given that the number of high-risk patients for PLCS is small. It is possible that surgical ligation induces a more potent inflammatory cascade compared to percutaneous closure, with less augmentation in afterload and stress on the left ventricle. It is also plausible that manipulation of intrathoracic pressure and lung retraction during surgical ligation induces types of lung injury that a percutaneous approach avoids.⁶

The subjects in the study of Serrano et al¹ were of advanced postnatal age at intervention (mean of 58 days in both groups) with mean weight at procedure of 1700-1900 g. In the percutaneous group, the smallest infant was 920 g, and 21/24 (87%) were >1000 g; the youngest infant was 23 days at the time of the procedure, 23/24 (96%) were >28 days old and 21/24 (87%) were >42 days old. In the surgical group the incidence of PLCS was reported to be 20%, which is approximately 4 times the projected probability for infants with postnatal age of 58 days (Figure 1).⁷ It is difficult to ascertain the degree and phenotype of cardiorespiratory instability from the data presented. Post-ligation instability typically occurs 6-12 hours after closure and manifests as systolic hypotension and acidosis. Some infants may exhibit exclusive signs of left ventricular diastolic impairment without signs of low cardiac output state. These signs include systemic hypertension and pulmonary edema which results in impaired oxygenation and the need for greater ventilatory support.⁶ It would have been interesting for the authors to present serial systolic and diastolic arterial pressures as well as oxygenation and ventilation data over time, which may provide more physiologic insights. Additionally, the methods section did not provide sufficient information about when changes in oxygenation occurred, the indication(s) for hemodynamic support, or the rationale for choosing hydrocortisone as the most common agent for hemodynamic support. Unfortunately, and as acknowledged by the authors, the surgical cohort lacked echocardiographic assessment of heart function in the first 24 hours which would have better characterized the observed differences between the two groups. Furthermore, it is possible that there was variance in the monitoring and threshold for medical intervention in the percutaneous group which could be due to the perception of a less invasive procedure.

EBM Lesson: Definition of population of interest

When comparing two interventions, it is important to clearly delineate the population of interest and ensure that inclusion criteria are selected such that patients who are most likely to develop the outcome of interest are enrolled. This is the premise of the PICO(T)[1] framework, a mnemonic where P stands for the choice of the population/patient as a one of 5 key elements to consider when designing a high-quality evidence-based research investigation. If the population under study is rare and the probability of the outcome is low, the comparison between both interventions is unlikely to yield any conclusion that is generalizable to those at higher risk. This principle is particularly relevant in patients undergoing interventional PDA closure where the probability of the outcome of interest is inversely proportional to age at intervention. This variance in risk profiling has biologic plausibility for two reasons; first, the more immature myocardium is less tolerant of afterload; second, the magnitude in change in LV afterload is lower in older patients due to the effect of a chronic high-volume shunt and lung disease on pulmonary vascular resistance.⁸ Therefor, in this study, it may have been misleading to conclude that percutaneous PDA closure has a lower incidence of PLCS because most of the infants studied by Serano et al. were more than 28 days old at the time of intervention; specifically, the enrolled population intrinsically have a lower risk of PLCS by virtue to their maturity at the time of intervention, independent of the method used to close the PDA (surgery vs percutaneous device closure). Although pragmatic trials are important, to the extent that these dilute the population of interest or at risk, they may reduce the likelihood of identifying a treatment effect. The recent follow-up paper to the PDA Tolerate trial further emphasizes this point; specifically, due to lack of physician equipoise, patients who were younger, had a higher risk profile, and had more pathologic PDA shunts were less likely to be enrolled in the trial.^{12, 13} It is therefore not surprising that the likelihood of identifying the true effect of treatment in the PDA Tolerate trial was compromised by the fact that the enrolled patients represented a lower risk group. There is an urgent need for physiologists and clinical trialists to work more closely together with the goal of designing trials that include appropriate numbers of patients at or above some threshold of risk.

Acknowledgment: 

The Journal club is a collaboration between the American Academy of Pediatrics- Section of Neonatal Perinatal medicine and the International Society of Evidence- based neonatology (EBNEO.org)

References: 

1. Serrano RM, Madison M, Lorant D, Hoyer M, Alexy R. Comparison of 'post-patent ductus arteriosus ligation syndrome' in premature infants after surgical ligation vs. percutaneous closure. J Perinatol 2019.

2. Backes CH, Giesinger RE, Rivera BK, Berman DP, Smith CV, Cua CL, et al. Percutaneous Closure of the Patent Ductus Arteriosus in Very Low Weight Infants: Considerations Following US Food and Drug Administration Approval of a Novel Device. J Pediatr 2019.

3. Clyman RI, Wickremasinghe A, Merritt TA, Solomon T, McNamara P, Jain A, et al. Hypotension following patent ductus arteriosus ligation: the role of adrenal hormones. J Pediatr 2014, 164(6): 1449-1455 e1441.

4. Harting MT, Blakely ML, Cox CS, Jr., Lantin-Hermoso R, Andrassy RJ, Lally KP. Acute hemodynamic decompensation following patent ductus arteriosus ligation in premature infants. J Invest Surg 2008, 21(3): 133-138.

5. Moin F, Kennedy KA, Moya FR. Risk factors predicting vasopressor use after patent ductus arteriosus ligation. Am J Perinatol 2003, 20(6): 313-320.

6. Giesinger RE, Bischoff AR, McNamara PJ. Anticipatory perioperative management for patent ductus arteriosus surgery: Understanding postligation cardiac syndrome. Congenit Heart Dis 2019, 14(2): 311-316.

7. Teixeira LS, Shivananda SP, Stephens D, Van Arsdell G, McNamara PJ. Postoperative cardiorespiratory instability following ligation of the preterm ductus arteriosus is related to early need for intervention. J Perinatol 2008, 28(12): 803-810.

8. McNamara PJ, Stewart L, Shivananda SP, Stephens D, Sehgal A. Patent ductus arteriosus ligation is associated with impaired left ventricular systolic performance in premature infants weighing less than 1000 g. The Journal of thoracic and cardiovascular surgery 2010, 140(1): 150-157.

9. Abu Hazeem AA, Gillespie MJ, Thun H, Munson D, Schwartz MC, Dori Y, et al. Percutaneous closure of patent ductus arteriosus in small infants with significant lung disease may offer faster recovery of respiratory function when compared to surgical ligation. Catheter Cardiovasc Interv 2013, 82(4): 526-533.

10. Sathanandam S, Balduf K, Chilakala S, Washington K, Allen K, Knott-Craig C, et al. Role of Transcatheter patent ductus arteriosus closure in extremely low birth weight infants. Catheter Cardiovasc Interv 2019, 93(1): 89-96.

11. Zahn EM, Peck D, Phillips A, Nevin P, Basaker K, Simmons C, et al. Transcatheter Closure of Patent Ductus Arteriosus in Extremely Premature Newborns: Early Results and Midterm Follow-Up. JACC Cardiovasc Interv 2016, 9(23): 2429-2437.

12. Clyman RI, Liebowitz M, Kaempf J, Erdeve O, Bulbul A, Hakansson S, et al. PDA-TOLERATE Trial: An Exploratory Randomized Controlled Trial of Treatment of Moderate-to-Large Patent Ductus Arteriosus at 1 Week of Age. J Pediatr 2019, 205: 41-48 e46.

13. Liebowitz M, Katheria A, Sauberan J, Singh J, Nelson K, Hassinger DC, et al. Lack of Equipoise in the PDA-TOLERATE Trial: A Comparison of Eligible Infants Enrolled in the Trial and Those Treated Outside the Trial. J Pediatr 2019, 213: 222-226 e222.

Figure 1: Probability of inotropes in the post-operative period of PDA ligation. Adapted from Teixeira LS, et al. Postoperative cardiorespiratory instability following ligation of the preterm ductus arteriosus is related to early need for intervention. J Perinatol. 2008;28(12):803-810

[1] The acronym PICO(T) is used as a mnemonic for drafting a focused clinical question.  P stands for Patient/Population; I for Intervention/Indicator; C for Compare/Control; O for Outcome; and T for Time or Type of study or question.