OBJECTIVE: To evaluate the efficacy and safety of clarithromycin treatment in preventing bronchopulmonary dysplasia (BPD) in Ureaplasma urealyticum–positive preterm infants.
PATIENTS AND METHODS: Nasopharyngeal swabs for U urealyticum culture were taken from infants with a birth weight between 750 and 1250 g in the first 3 postnatal days. Infants with a positive culture for U urealyticum were randomly assigned to 1 of 2 groups to receive either intravenous clarithromycin or placebo. All the patients were followed at least up to the 36th postmenstrual week.
RESULTS: A total of 224 infants met the eligibility criteria of the study. Seventy-four (33%) infants had a positive culture for U urealyticum in the first 3 day cultures. The rate of BPD development was significantly higher in patients with U urealyticum positivity (15.9% vs 36.4%; P < .01). However, multivariate logistic regression analysis failed to reveal a significant association between the presence of U urealyticum and BPD development (odds ratio: 2.4 [95% confidence interval: 0.9–6.3]; P = .06). Clarithromycin treatment resulted in eradication of U urealyticum in 68.5% of the patients. The incidence of BPD was significantly lower in the clarithromycin group than in the placebo group (2.9% vs 36.4%; P < .001). Multivariate logistic regression analysis confirmed the independent preventive effect of clarithromycin for the development of BPD (odds ratio: 27.2 [95% confidence interval: 2.5–296.1]; P = .007).
CONCLUSIONS: Clarithromycin treatment prevents development of BPD in preterm infants who are born at 750 to 1250 g and colonized with U urealyticum.
WHAT'S KNOWN ON THIS SUBJECT:
Despite numerous studies, controversy still exists about whether Ureaplasma urealyticum colonization or infection of the respiratory tract contributes to the development of bronchopulmonary dysplasia. Additional controversy exists on the use of macrolides for the treatment of U urealyticum.
WHAT THIS STUDY ADDS:
Clarithromycin treatment prevents development of bronchopulmonary dysplasia in preterm infants who are born at 750 to 1250 g and colonized with U urealyticum.
Bronchopulmonary dysplasia (BPD) is a significant cause of morbidity and mortality in premature infants and occurs in 30% of infants born at ≤28 weeks' gestation. The etiology of the BPD is likely to be multifactorial. Because of the serious long-term health consequences of BPD, prevention of the disease is the focus of major research.1
Ureaplasma urealyticum has long been implicated in the pathogenesis of chronic lung disease of prematurity, but despite numerous studies, reviews and meta-analyses, controversy still exists about whether U urealyticum colonization or infection of the respiratory tract contributes to the development of BPD. Inconsistence between the studies may be because of small sample sizes, vastly different inclusion criteria, methods of sampling and testing, and different diagnostic criteria for various outcomes including BPD. A meta-analysis by Schelonka et al,2 which included 23 studies and 2216 infants, revealed an odds ratio (OR) of 2.83 (95% confidence interval [CI]: 2.29–3.51) for the relationship between the presence of Ureaplasma and BPD diagnosed at 28 days of life. However, a causative role of U urealyticum could not be proven.
Additional controversy exists on the use of macrolides, particularly erythromycin, for the treatment of U urealyticum colonization. A recent Cochrane analysis3 identified 2 randomized studies. In neither trial was a significant effect of erythromycin shown on development of BPD or death. But both studies had a small sample size and underpowered to detect a true benefit. The lack of efficacy of erythromycin on the incidence of BPD may also be secondary to low rate of eradication of U urealyticum from the airway.4
In this study, we aimed to determine the relation of U urealyticum colonization to BPD in preterm infants with a birth weight of ≤1250 g and to evaluate, for the first time, the efficacy and safety of clarithromycin treatment in eradicating U urealyticum in the respiratory tract and preventing BPD in culture-positive preterm infants.
PATIENTS AND METHODS
This placebo-controlled, prospective, randomized study was conducted in Zekai Tahir Burak Maternity Teaching Hospital, which is one of the largest maternity hospitals in Turkey and performs 20 000 deliveries annually. The NICU includes 150 incubators and admits ∼4000 infants annually. Infants who were eligible for the study were determined as preterms with a birth weight between 750 and 1250 g. Exclusion criteria consisted of the presence of major congenital abnormalities, lack of parental informed consent, and intrauterine growth retardation with a birth weight at the <10th percentile for gestational age.
The primary outcomes for this study were eradication of U urealyticum in the airways and the composite chronic lung disease and/or death. BPD was defined as a persistent oxygen requirement at 36 weeks' postmenstrual age. A physiologic test was used to confirm the need for oxygen at the time when BPD was being diagnosed. Infants with moderate dependency on oxygen at 36 weeks' postmenstrual age (<30% oxygen) were challenged with room air breathing to determine if the supplemental oxygen was in fact needed.
Perinatal characteristics of the patients including birth weight, gestational age, route of delivery, prenatal steroid use, the presence of premature rupture of membrane, chorioamnionitis, sepsis, respiratory distress syndrome (RDS), surfactant use, pneumonia, patent ductus arteriosus (PDA), as well as postnatal clinical parameters including duration of mechanical ventilation, nasal continuous positive airway pressure, O2 therapy, caffeine, and diuretic treatment were recorded.
Infants were diagnosed with RDS if they had tachypnea, grunting, and cyanosis within several hours of birth, required mechanical ventilation including continuous positive airway pressure and oxygen in the first hours of life, and typical radiographic findings on the chest radiograph.5 BPD was diagnosed by using the US National Institutes of Health diagnostic criteria for BPD.1 PDA diagnosis required an echocardiogram with Doppler verification and cardiology recommendation of ibuprofen therapy on the basis of significance of flow.6
Nasopharyngeal swabs for U urealyticum were taken in first 3 postnatal days and on the 12th day after the commencing of the treatment only in culture-positive infants, transported to the laboratory, and cultured for U urealyticum immediately in special medium. U urealyticum is detected according to method defined by Biernat-Sudolska et al7 and culture results were obtained in a maximum of 48 hours.
Infants with a positive culture for U urealyticum in the first 3 days were randomly assigned to 1 of 2 groups to receive either intravenous clarithromycin (10 mg/kg twice per day for 10 days) or placebo. Randomization was performed by the use of sealed envelopes with the allocation of the patients to either clarithromycin treatment or placebo. Colonization by other potential pathogens was treated as warranted, according to the attending physician. Repeat cultures were obtained in the clarithromycin group to determine U urealyticum eradication. All the patients were followed up to at least the 36th postmenstrual week of gestation, when the presence of BPD was assessed, or death.
Statistical analysis was performed by using SPSS 17.0 (SPSS Inc, Chicago, IL). It was calculated that a sample size of 250 infants was required to yield >90% power to detect a 7% reduction in BPD incidence resulting from clarithromycin treatment.8 Independent samples t test or Mann-Whitney U test was used to compare continuous variables and χ2 test or Fisher's exact test for categorical variables. Multivariate (backward) logistic regression analysis was performed to simultaneously measure the influence of the independent variables with BPD as the dependent variable. Variables that had a P value of <.25 in the univariate analysis were used in multivariate analysis as possible risk factors. A P value of <.05 was considered statistically significant.
The study was approved by the institutional local ethics committee, and informed parental consent was obtained from all participants.
During the study period, 272 infants with a birth weight between 750 and 1250 g were enrolled in the study. Forty-eight infants were excluded because of major congenital abnormalities, lack of parental informed consent, intrauterine growth retardation, or death. There were a total of 224 infants who met the eligibility criteria, and 150 infants were culture-negative for U urealyticum in the first 3 days, whereas 12 infants in the culture-negative group died by follow-up. Seventy-four (33%) infants had a positive culture for U urealyticum in the first 3 day cultures (Fig 1).
To search for an association between U urealyticum positivity and the development of BPD, we have assessed the characteristics and outcomes of the U urealyticum–negative patients and U urealyticum–positive patients who did not receive clarithromycin (Table 1). BPD development was significantly higher in patients with U urealyticum positivity (15.9% vs 36.4%; P < .01). However, multivariate logistic regression analysis including birth weight, gestational age, gender, presence of RDS, pneumonia and PDA, culture positivity for U urealyticum, caffeine treatment, and mechanical ventilation in model failed to demonstrate a significant association between the presence of U urealyticum and BPD development, although P bordered significance (OR: 2.4 [95% CI: 0.9–6.3]; P = .06). The only significant variable associated with BPD development was duration of mechanical ventilation (OR: 1.1 [95% CI: 1–1.2]; P = .01).
U urealyticum–positive patients were prospectively randomly assigned to receive either clarithromycin (n = 37) or placebo (n = 37). Characteristics of the U urealyticum–positive patients who received clarithromycin or placebo are shown in Table 2. According to the repeat cultures 2 days after the discontinuation of the treatment, clarithromycin treatment resulted in eradication of U urealyticum in 68.5% of the patients. Groups did not have significant difference in incidences of PDA, sepsis, and pneumonia, whereas the incidence of BPD was significantly lower in the clarithromycin group (2.9% vs 36.4%; P < .001). Deaths in the groups had occurred before the evaluation of BPD at the 36th postmenstrual week, and they were not related to pulmonary problems; therefore, we did not choose to use the composite outcome of death or BPD. However, when statistical analysis was performed for the composite outcome defined as death or BPD, there was still a highly significant difference between the treatment and control groups (P = .001).
Variables that had a P value of <.25 in the univariate analysis were used in multivariate analysis as possible risk factors in subgroup analysis. Multivariate logistic regression analysis including birth weight, presence of pneumonia and PDA, caffeine and clarithromycin treatment, mechanical ventilation, and nasal continuous positive airway pressure in model confirmed the independent preventive effect of the clarithromycin (OR: 27.2 [95% CI: 2.5–296.1]; P = .007) for the development of BPD. No serious adverse event related to clarithromycin use was documented.
We demonstrated that 33% of the preterm infants with birth weight between 750 and 1250 g had U urealyticum positivity in nasopharyngeal swabs. U urealyticum positivity was associated with higher rate of BPD development, but we failed to demonstrate this association in multivariate analysis, probably because of low number of patients; the P value and 95% CI bordered significance. Treatment of these infants with clarithromycin resulted in lower rates of BPD than those treated with placebo.
Ureaplasma comprises 2 species (U parvum and U urealyticum) and 14 serovars. Ureaplasma is a commensal in the adult female genital tract with low virulance. However, previous studies have shown association of U urealyticum with infertility, adverse pregnancy outcomes, and perinatal death.9,–,12 Vertical transmission rate is inversely related to gestational age (18%–55% in term infants versus 29%–60% in preterm infants)10,–,13 and increases with duration of rupture of membranes.14,–,16 Respiratory tract colonization was detected in 20% of infants born at <1500 g by cultures and 25% to 48% by polymerase chain reaction. Culture positivity was 30% in our study, consistent with the literature. Differences in colonization rates may be a result of differences in communities or in specimen collection and processing.
Since the 1980s, many studies have searched for the link between U urealyticum colonization and BPD development.17,–,19 Although not all the studies have confirmed the association, the meta-analysis by Wang et al20 in 1995 showed that the relative risk for the development of BPD in U urealyticum colonized infants was 1.72 (95% CI: 1.5–1.96). These findings were confirmed by Schelonka et al2 in their meta-analysis in 2005, which revealed a significant association between Ureaplasma infection and development of BPD, at both 28 days (P < .001) and 36 weeks (P < .001) by definition.
In a number of subsequent trials, the therapeutic efficacy of erythromycin therapy was searched for in the first few weeks of life, given the in vitro sensitivity of U urealyticum to erythromycin. In 1 study, Lyon et al21 randomly assigned 75 infants born at <30 weeks' gestation and ventilated from birth for lung disease to receive erythromycin (n = 34) intravenously for 7 days or to no treatment (n = 41). Nine infants (13%) were positive for U urealyticum by culture or polymerase chain reaction. Those treated with erythromycin had a similar incidence of BPD as untreated infants. A small number of infected patients made it impossible to analyze infected infants alone. In the second study, Jonsson et al22 investigated colonization with U urealyticum in 155 infants born at <30 weeks' gestation and assessed the effect of treatment with erythromycin. The rate of colonization was 29 of 155 (19%). More colonized infants needed supplemental oxygen at 36 weeks' postconceptional age compared with noncolonized infants. Colonized infants were randomly assigned to treatment with erythromycin 40 mg/kg per day, intravenously or orally (n = 14), or no treatment (n = 14). Erythromycin treatment eradicated colonization in 12 of 14 (86%) treated infants. However, there was no significant difference between treated and untreated infants in the rate of BPD at 36 weeks' postconceptional age (64% in treated infants versus 71% in untreated group). These 2 randomized trials of erythromycin had low number of infected infants and no specific analysis in this subgroup could be made in the first trial, whereas only 14 patients could be evaluated in both arms in the second trial, which increases the possibility to miss a true benefit of erythromycin because of type II error. Another reason why erythromycin was ineffective in preventing BPD development might be relatively lower anti-inflammatory activity of erythromycin in preterm infants.4,21,23
We preferred to use clarithromycin in our study instead of erythromycin. In vitro activity of clarithromycin against U urealyticum is much higher than erythromycin.24,25 Second, penetration of clarithromycin into bronchial mucosa, bronchial secretions, and epithelial lining as reflected by the tissue/fluid-to-serum ratio is higher than that for erythromycin. Twice-daily use and a better toxicity profile are other advantages of clarithromycin over erythromycin.26
In a recent study by Ballard et al,27 220 infants who had a birth weight of <1250 g were randomly assigned to azithromycin 10 mg/kg per day for 7 days followed by 5 mg/kg per day for a maximum of 6 weeks (n = 111) or placebo (n = 109) within 12 h of beginning mechanical ventilation and within 72 h of birth. The incidence of BPD in the U urealyticum–positive subgroup was 73% in those treated with azithromycin versus 94% in the placebo group (P = .03). The authors concluded that early treatment of U urealyticum colonized/infected patients might be beneficial.27 Our study also yielded benefit from treatment in culture-positive infants, similar to this study. BPD incidence was overall lower in our study than in the study of Ballard et al.27 This might be secondary to the differences in patient characteristics of studies. Mean birth weight was 803 g in their study versus 1013 g in our study, and mean gestational age was 25.7 vs 27.9 weeks, respectively.
Several studies in the literature show that macrolides affect many inflammatory processes including migration of neutrophils, the oxidative burst in phagocytes, and production of pro-inflammatory cytokines.28 Clarithromycin inhibits superoxide production by activated neutrophils and also has a membrane-stabilizing activity.29 Clarithromycin was also found to suppress interleukin 1β gene expression in human nasal epithelial cells.30 We believe that anti-inflammatory effects of clarithromycin also contribute to overall lower incidence of BPD in the treatment group, but we could not evaluate pure anti-inflammatory effect of clarithromycin in our study because no culture-negative infants received clarithromycin. However, the benefit of azithromycin was observed selectively in culture-positive infants but not in the whole group in the study of Ballard et al.27 Thus, anti-microbial activity seems to be more considerable compared with anti-inflammatory activity in prevention of BPD. Prospective trials are needed to confirm these hypotheses. Because of an overall low rate of BPD in patients treated with clarithromycin, we could not assess whether failure to eradicate U urealyticum was associated with higher incidence of BPD.
There are a number of limitations to our study. First, ours was a single-center study, and our findings should be confirmed by additional well-designed clinical trials in different settings. Second, although there were not significant differences in incidences of PDA, sepsis and pneumonia between groups, there was a tendency to lower rates of these parameters in treatment group, which might affect BPD development. Third, this study was a pioneering one, and it seems that there is a requirement for additional studies with larger sample size estimates for the cohort of infants randomly assigned to the treatment and placebo groups.
Clarithromycin treatment for U urealyticum–positive very low birth weight infants was shown to reduce the incidence of BPD development in this study. The benefit of clarithromycin in U urealyticum–negative infants is currently unclear. Given the advantage of early use of antibiotics in these patients, early treatment after obtaining respiratory cultures for U urealyticum, and discontinuation of treatment in culture-negative patients, might be a feasible strategy. Additional randomized studies are needed to establish routine use of macrolides in this high-risk population.
- Accepted August 15, 2011.
- Address correspondence to Ramazan Ozdemir, MD, Neonatal Intensive Care Unit, Zekai Tahir Burak Maternity Teaching Hospital, 06110 Hamamonu, Ankara, Turkey. E-mail:
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
- BPD —
- bronchopulmonary dysplasia
- OR —
- odds ratio
- CI —
- confidence interval
- RDS —
- respiratory distress syndrome
- PDA —
- patent ductus arteriosus
Statistical power calculators. DSS Research. Available at: www.dssresearch.com/toolkit/spcalc/power_a2.asp. Accessed May 4, 2009
- Waites KB,
- Katz B,
- Schelonka RL
- Kafetzis DA,
- Skevaki CL,
- Skouteri V,
- et al
- Theron AJ,
- Feldman C,
- Anderson R
- Copyright © 2011 by the American Academy of Pediatrics