Article Text
Abstract
Objective To evaluate incidence of minimally invasive surfactant therapy (MIST) failure, identify risk factors and assess the impact of MIST failure on neonatal outcome.
Design Retrospective cohort study. MIST failure was defined as need for early mechanical ventilation (<72 hours of life). Multivariate logistic regression analysis was performed to identify risk factors for MIST failure and compare outcomes between groups.
Setting Two tertiary neonatal intensive care centres in the Netherlands.
Patients Infants born between 24 weeks’ and 31 weeks’ gestational age (GA) (n=185) with MIST for respiratory distress syndrome.
Interventions MIST procedure with poractant alfa (100–200 mg/kg).
Main outcome measures Continuous positive airway pressure (CPAP) failure after MIST in the first 72 hours of life.
Results 30% of the infants failed CPAP after MIST. In a multivariate logistic regression analysis, four risk factors were independently associated with failure: GA <28 weeks, C reactive protein ≥10 mg/L, absence of antenatal corticosteroids and lower surfactant dose. Infants receiving 200 mg/kg surfactant had a failure rate of 14% versus 35% with surfactant dose <200 mg/kg. Mean body temperature was 0.4°C lower at neonatal intensive care unit admission and before the procedure in infants with MIST failure.
Furthermore, MIST failure was independently associated with an increased risk of severe intraventricular haemorrhage.
Conclusion We observed moderate MIST failure rates in concordance with the results of earlier studies. Absence of corticosteroids and lower surfactant dose are risk factors for MIST failure that may be modifiable in order to improve MIST success and patient outcome.
- surfactant
- preterm
- respiratory
Statistics from Altmetric.com
What is already known on this topic?
Minimally invasive surfactant therapy (MIST) is known to avoid mechanical ventilation and increases continuous positive airway pressure success rates.
MIST leads to less adverse events in preterm infants.
A considerable amount of the infants who received MIST needs intubation and mechanical ventilation during the first days of life.
What this study adds?
MIST failure is associated with lower gestational age and elevated C reactive protein levels 24 hours after birth.
MIST failure is also associated to lack of antenatal steroids and low surfactant dose. These factors may be modifiable.
MIST failure is related to decreased survival without serious adverse events, prolonged mechanical ventilation and more severe intraventricular haemorrhage (grade 3–4).
Introduction
Minimally invasive surfactant therapy (MIST) has been introduced as a new technique to administer surfactant in spontaneously breathing preterm infants with respiratory distress syndrome (RDS) without the need for intubation and mechanical ventilation.
MIST is considered to be safe and well tolerated.1 2 Several randomised controlled trials (RCTs) showed less need for mechanical ventilation and supplementary oxygen therapy in MIST-treated infants compared with surfactant administration after endotracheal intubation with or without the intubation–surfactant–extubation (INSURE) procedure and continuous positive airway pressure (CPAP) alone.3–5 Recently published systematic reviews show that infants treated with MIST have a reduced risk of bronchopulmonary dysplasia (BPD) compared with those receiving surfactant via an endotracheal tube.6 7 Moreover, MIST is associated with increased survival without serious adverse events and a lower incidence in severe intraventricular haemorrhage (IVH).2 5 7
Currently, a significant proportion of preterm infants can be managed on CPAP resulting in an improved neonatal outcome compared with invasive mechanical ventilation.8 9 Although the introduction of MIST has increased the success rate of primary CPAP treatment, a significant proportion of mainly extremely preterm infants fail CPAP and need to be intubated and mechanically ventilated.3 5 Risk factors for both CPAP failure and INSURE failure have been identified,8 10 but to the best of our knowledge, no studies have assessed risk factors for MIST failure. Identifying potential risk factors may help to improve MIST indications, technique and success rates in the future. Therefore, the objectives of our study were: (A) to determine the incidence of MIST failure defined as the need for mechanical ventilation in the first 72 hours of life; (B) to identify risk factors associated with MIST failure and (C) to assess if neonatal morbidity and mortality differ between infants failing MIST and infants who are successfully treated with MIST.
Methods
This study was an observational, retrospective cohort study conducted at two neonatal intensive care units (NICUs) in the Netherlands: Emma Children’s Hospital, Amsterdam University Medical Centre (AUMC) in Amsterdam and Máxima Medical Centre (MMC) in Veldhoven.
Surfactant indication and MIST procedure
The indication for surfactant treatment was similar in both centres: infants on CPAP level 6 cm H2O with fraction of inspired oxygen (FiO2) above 0.30 (preterm infants <32 weeks’ gestation) within the first 72 hours of life. MIST was the preferred procedure in all spontaneously breathing preterm infants requiring surfactant. However, the attending neonatologist could decide to administer surfactant via an endotracheal tube depending on the clinical condition of the infant. Contraindications for MIST were insufficient respiratory drive, haemodynamic instability needing inotropes or pneumothorax.
No sedative medication was given to the infants prior to the procedure. Most infants received 10 µg/kg atropine. Early caffeine (loading dose 10 mg/kg; continuation dose 5 mg/kg) was started in all infants <30 weeks’ gestation but not always administered before the procedure. The recommended surfactant dose (poractant alfa, Curosurf, Chiesi Pharmaceuticals, Parma, Italy) was 100–200 mg/kg in both hospitals, according to Dutch protocol. Surfactant was administered via an umbilical catheter size 4–5 French, which was placed 1 cm below the vocal cords under direct visualisation of a laryngoscope with the use of Magill forceps. The laryngoscope was removed, and surfactant was instilled. During the MIST procedure, CPAP administration was continued. If oxygen need (FiO2 ≥0.30) persisted 1 hour after the procedure, MIST procedure was repeated or a second dose of surfactant was administered after endotracheal intubation. As surfactant with MIST might not be as effective as surfactant after endotracheal intubation, partly due to spill and lack of distribution, we provided the opportunity to redose with shorter interval in this study than recommended in the international guidelines based on surfactant after endotracheal intubation.11 12 Criteria for endotracheal intubation were: FiO2 ≥0.40; blood gas analysis with pH <7.20 or partial pressure carbon dioxide (pCO2) >8 kPa and significant apnoea requiring intervention. The target oxygen saturation was 88%–95% in both centres.
Study group and data collection
All spontaneously breathing preterm infants of 24–31 weeks’ gestation born between January 2014 to December 2015 (AUMC) and June 2014 to October 2016 (MMC) treated with surfactant were registered and included in the study. We excluded infants with primary surfactant administration via endotracheal intubation in the analysis. However, to put our findings on MIST into context, we present additional data on the baseline characteristics and outcome of primarily intubated infants.
Data of all included infants were retrospectively collected from the neonatal and maternal medical records. The potential risk factors for MIST failure were selected prior to analysis and based on clinical experience and former studies on respiratory failure.8 10 The RDS severity was estimated by the latest FiO2 measured before MIST procedure. Furthermore, we collected data on postsurfactant characteristics: respiratory support during NICU admission, mortality, need for supplementary oxygen at 28 days and serious adverse events. Serious adverse events in this study included pneumothorax, surgical ligation for patent ductus arteriosus, IVH grade 3–4,13 cystic periventricular leukomalacia,14 abdominal surgery for necrotising enterocolitis or spontaneous intestinal perforation and laser therapy or vitrectomy for retinopathy of prematurity.
Statistical analysis
Continuous variables were dichotomised prior to analysis when possible. In the primary statistical analysis, we performed a univariate logistic regression analysis on all potential risk factors for MIST failure. Differences in univariate analysis were expressed as crude OR with 95% CI. Subsequently, we included all risk factors with p value <0.1 of the univariate analysis in a multivariate forward stepwise logistic regression model in order to identify independent risk factors. On outcome factors, we conducted an univariate and multivariate logistic regression with gestational age (GA), RDS severity and antenatal corticosteroids as predefined confounders because of their well-known effects on neonatal outcome.8 12 15 16 The statistical analysis was performed using SPSS V.24 and statistical significance was accepted at a two-sided p value <0.05.
Results
During the study period, 185 infants born between 24 weeks’ and 31 weeks’ gestation were treated with MIST (figure 1). Baseline characteristics of these are given in table 1. During the study period, 117 excluded infants were primarily intubated before receiving surfactant (GA 27.7±2.3 weeks; birth weight [BW] 978±514 g). Figure 2 shows an overview of respiratory support within the first 72 hours of all infants <32 weeks’ gestation admitted to both NICUs during the study period. CPAP failure rates decrease with increasing GA.
Overall, 30% of the infants failed MIST and needed invasive mechanical ventilation in the first 72 hours of life. Both centres showed comparable incidences of MIST failure (AUMC: 30.8%; MMC: 29.9%). Twenty-four of 31 (77%) infants underwent a successful second MIST procedure. Extremely preterm infants (<28 weeks) and infants with severe RDS (defined as FiO2 ≥0.50) showed failure rates of 44% and 40%, respectively. As presented in figure 3, the proportion of MIST failure gradually decreased with increasing GA.
Mean GA of the MIST failure group was 27.0±1.7 weeks and 28.6±1.9 weeks in the MIST success group. Univariate analysis in table 2 shows six risk factors associated with MIST failure: absence of antenatal corticosteroids, GA <28 weeks’ gestation, hypothermia (temperature <36.5°C) before the MIST procedure, more severe RDS, elevated C reactive protein (CRP) values 24 hours after birth and surfactant dose <200 mg/kg. Temperature in the MIST failure group was statistically significant lower compared with the success group both at time of NICU admission (35.9°C±1.0°C vs 36.3°C±0.8°C, p=0.006) and shortly before the MIST procedure (36.3°C±0.7°C vs 36.7°C±0.6°C, p=0.003). Significantly more preterm infants with MIST success reached normothermia before the procedure.
In a multivariate logistic regression model, we identified four factors independently associated with MIST failure. Of these, extreme prematurity (GA <28 weeks) was most strongly associated with MIST failure (OR 4.7, 95% CI 1.9 to 11.6). Moreover, the risk of MIST failure increased with an OR of 4.2 (95% CI 1.1 to 15.6) with absence of antenatal corticosteroid therapy. Surfactant dosage <200 mg/kg was also strongly associated with failure (OR 4.3, 95% CI 1.4 to 13.2). In the infants who received 200 mg/kg surfactant, failure rate was 14%, compared with 35% failure rate in infants receiving <200 mg/kg surfactant. Subsequently, MIST failures had more often elevated CRP values (≥10 mg/L). There was no significant correlation between RDS severity and timing of surfactant therapy (p=0.912). We observed a shorter but non-significant median interval between birth and MIST procedure in the infants who failed MIST therapy (3.3 hours vs 4.0 hours).
Table 3 shows the postsurfactant characteristics on mortality and neonatal morbidity within the MIST population and provides outcome data on the intubated group that did not receive MIST. Univariate analysis shows fewer median days on mechanical ventilation and less need for supplemental oxygen on day 28 in favour of the MIST success group compared with the failure group. Although not statistically significant in multivariate analysis, survival without serious adverse events in the infants with MIST failure was almost 20% lower compared with the group who did not need invasive mechanical ventilation in the first 72 hours of life (57.1% vs 76.0%). There was no significant difference in mortality between both groups. However, significantly more severe IVH was found in the infants requiring early mechanical ventilation (14.3%, n=8) following MIST compared with MIST success (3.1%, n=4) (OR 3.8, 95% CI 1.0 to 14.3).
Discussion
This retrospective study reported moderate MIST failure rates in accordance with previous studies, showing the feasibility of this procedure in daily practice.3 5 8 Four independently associated risk factors with MIST failure were: extreme prematurity (GA <28 weeks), elevated CRP levels, absence of antenatal corticosteroids and surfactant dose <200 mg/kg. Of these, the last two are considered to be modifiable.
MIST failure increased with decreasing GA in accordance with a study investigating risk factors for INSURE failure.10 Younger infants are more prone to respiratory insufficiency due to surfactant deficiency and immature respiratory drive. We observed 46% avoidance of mechanical ventilation in the first 72 hours of life in preterm infants of 24 weeks’ gestation. Our study shows the feasibility of the MIST procedure in extremely preterm infants. Overall, our study showed that MIST might be more effective in avoiding mechanical ventilation as compared with previously reported failure rates for INSURE (30% vs 69% treatment failure), even in extremely preterm infants.10 We suggest that there are two possible explanations for this difference: a lower mean GA in the INSURE study and the use of sedatives in INSURE with a suppressive effect on respiratory drive.
Second, elevated CRP value at 24 hours was a risk factor for MIST failure. Perinatal infection and inflammation are known to be associated with a decreased effect of administrated exogenous surfactant due to surfactant inactivation and increased surfactant clearance in preterm lungs.17 18 However, CRP values are a non-specific marker of inflammation and higher CRP levels may also reflect more severe RDS, although we tried to adjust for this effect in our multivariate model.17 18
We also identified two independent risk factors that may be modifiable in clinical practice. First, MIST failure was strongly associated with absence of antenatal steroids. Antenatal corticosteroids are strongly associated with RDS development and the response to exogenous surfactant therapy.12 We found that, even after adjusting for multiple factors including GA, antenatal steroids resulted in more MIST success. This effect was apparent even in two centres with high rates of antenatal corticosteroid administration; centres that perform MIST and have lower antenatal steroid treatment rates should focus on increasing these rates.
The most readily modifiable risk factor identified was surfactant dose. We found a clear association between lower surfactant dose and MIST failure. As poractant alfa is more concentrated than other surfactants, doses administered via respiratory tube above 100 mg/kg were studied.19 Several studies have shown a dose–effect response of poractant alfa, where 200 mg/kg administration compared with 100 mg/kg resulted in better respiratory function.19 20 This dose–effect response might even be more prominent with the MIST procedure. Coughing and reflux are important side effects of MIST causing surfactant spill. Moreover, 11% of the administered phospholipids is lost in the thin MIST catheter, which is two to three times higher than in an endotracheal tube.21 MIST also showed a slightly impaired surfactant distribution in the lungs compared with surfactant via endotracheal intubation.22 In line with our results, Dargaville repeatedly discussed the possible dose–effect relation on MIST success.1 23 He observed 100% MIST success in all preterm infants who received a 200 mg/kg dose of surfactant.1 It can be debated whether these findings apply to other kinds of surfactant administered in higher volumes.
We observed interesting differences in mean body temperature between both groups. Mean body temperature at NICU admission and before the MIST procedure was significantly lower in the MIST failure group. However, hypothermia both at NICU admission and shortly before the procedure was not independently associated with MIST failure. Hypothermia in preterm infants is known to be associated with higher incidence of RDS, mechanical ventilation and mortality.24–26 Whether our observed univariate association between body temperature and MIST failure is causal, can be debated, as a low admission temperature can also be a sign of worse clinical condition or complicated transition in the delivery room. An alternative explanation was provided by Shima et al,27 who reported an association between hypothermia and surfactant unresponsiveness, suggesting a direct negative effect of a lower admission temperature on having a successful MIST procedure.
The MIST population showed a more favourable outcome for infants with MIST success compared with MIST failure. Although it was not the aim of this study, we observed that the outcome of MIST failures was similar to the outcome of the primarily intubated infants without major differences in GA and BW.
Kribs et al 5 reported a 15% increase in survival without serious adverse events in MIST compared with primary endotracheal intubation. No separate data were presented on the outcome of infants intubated after MIST. We observed similar overall incidences in adverse events compared earlier studies and—more importantly—also found significant differences within the MIST treatment group.5 There was almost 20% higher survival without serious adverse events in infants successfully treated with MIST compared with those failing MIST. This difference was not statistically significant in multivariate analysis, most likely due to the effect of GA on outcome.
Univariate analysis showed significantly longer total duration of mechanical ventilation and more need for supplemental oxygen on day 28 in infants failing MIST. The avoidance of mechanical ventilation is known to be associated with a lower incidence in BPD. However, as infants were transferred before a postmenstrual age of 36 weeks to other hospitals and no oxygen reduction tests were performed, our study lacks data on BPD incidence and severity.28
The overall incidence of severe IVH in our study population (6.5%) was comparable with the incidence found in the intervention arm of the Avoidance of Mechanical Ventilation (AMV) trial (7%) and the Nonintubated Surfactant Application(NINSAPP) trial (10%).3 5 The latter study reported a 53% relative risk reduction in developing severe IVH in the MIST group compared with the intubated group.5 However, this is the first study to report IVH incidences within infants who underwent MIST, with significant differences in who failed (14.3%) and succeeded CPAP treatment (3.1%) hereafter. Previous studies have found an association between mechanical ventilation and the development of severe IVH.7 8 Our data suggest that the decrease in IVH after successful MIST might be due to avoidance of early mechanical ventilation. However, our study is not designed to confirm a causal relationship and mechanical ventilation as a consequence of and apnoea in developing IVH cannot be ruled out. Future studies should investigate whether this association is causal or reverse causal.
Our study has several limitations. First, the study is retrospective and explorative. The observational design impairs to draw causal conclusions, and therefore the identified risk factors need to be further investigated in RCTs. Second, there were strict criteria for surfactant administration, and the preferred method of surfactant therapy is MIST in both centres, but the decision to perform MIST was party at the discretion of the neonatologist. This may have led to the risk of selection bias. However, MIST requires adequate respiratory drive and as such not every infant is a candidate for MIST. Indeed, in the RCTs performed by Gopel and Kribs, a considerable number of infants were excluded, partly because they were primarily intubated.3 5 In our study, we investigated infants who failed and succeeded CPAP after MIST and did not aim to compare MIST with other treatments. Incidence of mechanical ventilation was comparable with the above RCTs, and therefore we expect to have included a similar and representative group of infants. The major strengths of this study include its multicentre design and broad GA range, which make the results applicable to a standard NICU population.
In conclusion, we observed moderate MIST failure rates even in extremely preterm infants and in infants with severe RDS. MIST failure was independently associated with lower GA, elevated CRP values, absence of antenatal corticosteroids and lower surfactant dosing. These last two risk factors may be modifiable in order to achieve higher success rates and improve patient outcome in the future.
References
Footnotes
Contributors LJ: study design, acquisition of data, analysis of data, interpretation of data, wrote first draft manuscript and final approval. JVDS: acquisition of data, analysis of data, input on first draft and final approval. AHvK: study design, analysis and interpretation of data, revision of manuscript and final approval. JD: input on study design, major input on statistical analysis, input on statistical aspects of manuscript and final approval PA: study design, interpretation of data and revision of manuscript final approval. WO: study design, interpretation of data, revision of manuscript and final approval. HN: conceived study, study design, analysis of data, interpretation of data, revision of manuscript and final approval.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Ethics approval The study protocol was approved by the Regional Committee on Medical Research Ethics at Máxima Medical Centre, Veldhoven.
Provenance and peer review Not commissioned; externally peer reviewed.
Patient consent for publication Not required.