The effect of furosemide on extremely premature infants treated with nonsteroidal anti-inflammatory drugs for persistent patent ductus arteriosus

BACKGROUND/RATIONALE

Patent ductus arteriosus (PDA) is associated with significant morbidity and mortality in some preterm infants [1]. Pharmacological options, including non-steroidal anti-inflammatory drugs (NSAIDs) or acetaminophen, promote ductal closure by inhibiting prostaglandin E2 (PGE2) synthesis. Furosemide, a loop diuretic, may increase PGE2 levels, possibly counteracting the effects of NSAIDs [2]. Previous studies on the impact of furosemide on PDA treatment have yielded conflicting results; while some studies suggest a negative impact, others found no difference [3–5]. This study aims to clarify the effects of furosemide in preterm infants with larger PDA.

METHODS

Retrospective cohort study of preterm infants (≤27 weeks’ gestational age) admitted to the Level IV NICU at Children’s Memorial Hermann Hospital between 2008 and 2018 (n = 858) with a confirmed PDA between 7 and 30 days of life who received NSAID treatment (n = 208). For this study, a large PDA was defined as size >2.5 mm or diameter >1.5 mm with an La:Ao ratio >2 (n = 64).

Data collected included demographics, NSAID type, echocardiography parameters, fluid intake, urine output, creatinine, and hemodynamic/respiratory support. Furosemide administration was considered if given within a time period of two days before the first echocardiogram up until the echocardiogram following NSAID treatment. Furosemide wasn’t initially used for PDA but for other issues like fluid buildup or breathing problems.

The study compared PDA cases exposed or not exposed to furosemide. The primary outcome was PDA size after NSAID treatment (diameter <1.5 mm). Baseline characteristics and outcomes after NSAID treatment (PDA closure rates, need for surgical closure, multiple NSAID courses, respiratory/hemodynamic/laboratory measures) were analyzed between groups. Our a priori hypothesis was that co-administration of furosemide with NSAIDs would decrease rates of PDA closure.

For statistical analysis, Wilcoxon rank sum test, Pearson’s Chi-squared test, and Fisher’s exact test were used for sample characteristics. Outcomes were calculated using generalized estimating equations.

RESULTS

Baseline characteristics were comparable between groups with no significant differences in gestational age, birth weight, race, type of NSAID received, and laboratory values. Infants who received furosemide had larger PDAs and a higher proportion with a PDA/weight ratio >3 mm/kg, though not statistically significant. Notably, the furosemide group exhibited a higher FiO₂ at baseline, suggesting the reason for the administration of furosemide (p < 0.05). There were no differences in rates of PDA closure [Relative Risk (RR): 0.88 (95% Confidence interval (CI), 0.28–2.75)] or surgical intervention [RR: 1.53 (95% CI, 0.37–3.55)] between groups.

Infants who received furosemide, however, demonstrated lower urine output during treatment (p = 0.04). No other differences were observed in secondary outcomes, including the need for additional NSAID courses or surgical interventions (Table 1).

Table 1.

Baseline characteristics and outcomes of infants during PDA treatment.

Baseline characteristics	No furosemide (n = 34)a	Furosemide (n = 30)a	p-valueb
Gestational age (weeks)	25.0 (0.9)	24.7 (1.2)	0.30
Female	20 (59%)	18.8 (60%)	>0.90
Birth weight (grams)	713.1 (121.9)	670.4 (127.0)	0.20
IUGR	3 (9%)	4 (13%)	0.70
Race			0.30
Black	18 (55%)	15 (50%)
White	6 (18%)	10 (33%)
NSAID received			0.90
Ibuprofen	21 (62%)	18 (60%)
Indomethacin	13 (38%)	12 (40%)
PDA size (mm)	2.7 (0.5)	3.0 (0.5)	0.11
PDA/weight			0.09
>3 mm/kg	26 (77%)	28 (93%)
1.5–3 mm/kg	8 (24%)	2 (7%)
Age of treatment (d)	14.6 (9.2)	13.0 (5.5)	0.90
Platelets	219.8 (109.2)	183.2 (99.5)	0.30
Hemoglobin (g/dL)	11.4 (1.6)	11.4 (1.2)	>0.90
CO2 (mmHg)	60.1 (7.3)	60.0 (11.4)	0.90
Total fluid goal (mL/kg/d)*	139.3 (13.1)	142.0 (21.5)	0.90
Creatinine (mg/dL)*	0.9 (0.2)	0.9 (0.3)	0.50
UOP (mL/kg/h)*	3.2 (0.9)	3.1 (0.9)	0.80
FiO2*	27.9 (6.4)	30.9 (6.7)	0.03
Mean airway pressure*	8.4 (2.3)	8.9 (2.0)	0.40
Inotropes use	5 (15%)	5 (17%)	>0.90

Outcomes after PDA treatment	No furosemide (n = 34)a	Furosemide (n = 30)a,c	RR (95% CI)	p-value
PDA closure	5 (15%)	4 (13%)	0.88 (0.28, 2.75)	0.83
PDA < 1.5 mm	10 (29%)	7 (23%)	0.78 (0.35, 1.73)	0.54
PDA size (mm)	1.8 (1.0)	2.0 (1.0)	0.226 (−0.23, 0.69)d	0.34
PDA/weight			1.08 (0.8, 1.46)	0.63
<1.5 mm/kg	6 (18%)	7 (23%)
1.5–3 mm/kg	20 (59%)	12 (40%)
>3 mm/kg	8 (24%)	11 (37%)
>1 NSAID course	13 (38%)	11 (39%)	1.01 (0.55, 1.85)	0.97
Surgical closure	6 (18%)	8 (27%)	1.53 (0.61–3.84)	0.36
Inotropes use	5 (15%)	5 (17%)	1.15 (0.37, 3.55)	0.81
Hydrocortisone	1 (3%)	1 (3%)	1.17 (0.08, 17.35)	0.91
Creatinine (mg/dL)**	0.9 (0.3)	0.9 (0.3)	−0.04 (−0.19, 0.11)d	0.63
UOP (mL/kg/h)**	2.7 (0.6)	2.4 (0.7)	−0.35 (−0.68, −0.01)d	0.04
FiO2*	33.9 (12.2)	36.1 (11.3)	2.18 (−3.55, 7.92)d	0.46
Mean airway pressure	8.2 (3.0)	9.0 (2.0)	0.81 (−0.42, 2.04)d	0.20
Total fluid goal**	132.4 (9.6)	130.4 (11.7)	−2 (−7.41, 3.4)d	0.47
Change in weight (g)	62.9 (70.9)	69.9 (73.8)	6.8 (−29, 43)d	0.7

CI confidence, IUGR intrauterine growth restriction, n number infants, RR relative risks, SD standard difference, UOP urine output.

Statistically significant p-values are in bold.

^aMean (SD); n(%).

^bWilcoxon rank sum test, T-test; Pearson’s Chi-squared test; Fisher’s exact test.

^cNumber of furosemide doses (1 mg/kg/dose intravenous or 2 mg/kg/dose enteral) Mean (SD) 2.3 (2.5).

^dMean difference.

^*Average of measurements 3 days before treatment.

^**Average of measurements during treatment.

DISCUSSION

Contrary to our initial hypothesis, we found no difference in PDA closure rates or other major outcomes between infants who received furosemide and those who did not.

Our findings contribute to the ongoing debate regarding the role of furosemide in PDA management. Previous studies, which often focused on a broader range of patients with a diagnosis of PDA, have reported conflicting results 3–5. Our study specifically focused on patients with presumed higher volume shunt PDA cases and did not observe a greater need for medical or surgical intervention. These results provide reassurance that furosemide may not compromise the effectiveness of NSAIDs in treating PDA. Nevertheless, the risk-benefits of empiric furosemide require thoughtful consideration especially in extremely preterm infants at risk of nephrocalcinosis, pneumothoraces, or bone fractures. The unexpected lower urinary output in the furosemide-treated group could relate to the additive effects of fluid restriction in infants with larger PDAs leading to greater renal hypoperfusion.

This study has limitations. It is a small, single-center retrospective review where treatment was based on clinician preference, introducing potential variability in patient selection, as evidenced by the higher fraction of inspired oxygen in the furosemide group. Furthermore, the study’s retrospective design limited our ability to control furosemide’s timing, dosing, and duration, potentially impacting the results. In addition, the transition from indomethacin to ibuprofen during the study period may have influenced the results. Actual delineation of hemodynamic significance and whether these findings are applicable to patients with small-moderate volume shunts requires further investigation. It is possible that any negative effects on PDA closure are less noticeable in larger or more pathological shunts that are less likely to respond to NSAID treatment. The low rate of PDA closure in both groups (<15%) supports this hypothesis. It remains, therefore, unclear whether a threshold exists that the benefits on lung compliance are offset by a negative effect on PDA closure. Regardless, our study provides valuable insights into the interplay between furosemide and NSAID therapy in PDA management.

CONCLUSION

Furosemide administration did not influence PDA closure rates or other major outcomes in this cohort of preterm infants with PDA. While further research is needed, our findings suggest that furosemide may be a safe adjunctive therapy in patients with large PDA and pulmonary edema.