Abstract
Background: Less invasive surfactant administration (LISA) has become the preferred method of surfactant administration for spontaneously breathing babies on continuous positive airway pressure (CPAP). Summary: The development of LISA followed the need to combine CPAP and surfactant replacement as mainstay treatment options for respiratory distress syndrome, thereby avoided exposure to positive pressure ventilation. Key Messages: This review summarises the current knowns and unknowns of LISA including the physiological concept, its relevance for short-term and long-term outcomes and the challenges for practical implementation of LISA as part of a less invasive respiratory care bundle. Further, we provide an update of the evidence on alternatives to LISA, for example, nebulised surfactant administration, pharyngeal deposition of surfactant and delivery via supraglottic airway.
Introduction
The proof of efficacy of exogenous surfactant for respiratory distress syndrome (RDS) was established in clinical trials conducted in the 1980s, with a clear effect on mortality and pneumothorax risk. Surfactant replacement therapy became, and still is, a mainstay intervention for preterm infants [1, 2]. The conventional mode of surfactant replacement therapy has been via an endotracheal tube (ETT) to infants intubated in early life and mechanically ventilated. This care bundle, although affording the benefits of surfactant therapy, involved exposure to positive pressure ventilation and hence the risk of ventilation-induced lung injury. More recently, continuous positive airway pressure (CPAP) has become the preferred means of initial respiratory stabilisation for preterm infants below 30 weeks’ gestation. CPAP prevents the development of atelectasis, reduces work of breathing and preserves the endogenous surfactant pool. Further, large randomised controlled trials have shown that application of CPAP as the primary mode of respiratory support, with selective surfactant therapy, leads to outcomes that are comparable to primary intubation, mechanical ventilation, and surfactant administration [3‒6]. Indeed, a meta-analysis of pooled data from these trials shows, for the outcome of death or bronchopulmonary dysplasia (BPD), a small advantage of CPAP over primary intubation (relative risk [RR] 0.91, 95% confidence interval [CI] 0.83–0.99, number needed to treat: 25) [7].
It remains a clinical challenge, however, that only around one half of preterm infants below 30 weeks’ gestation can successfully be managed with CPAP alone in the first days of life, with the remainder requiring some other form of intervention, CPAP having failed to provide adequate respiratory support. Predicting the respiratory trajectory for an individual infant is difficult. This is mainly related to the heterogeneity of RDS severity and the diagnostic uncertainty based on clinical appearance (Silverman score), fraction of inspired oxygen (FiO2), lung ultrasound scores or rapid tests of lung maturation [8]. Perinatal factors such as pre-eclampsia, intrauterine growth restriction, exposure to corticosteroids and presence of chorioamnionitis additionally impact the respiratory course in early life [9]. Medical decision-making should take these factors into account, and find a balance between non-invasive and interventionist approaches to treatment. The option of continuing treatment with CPAP avoids lung injury mediated by positive pressure ventilation but also exposes the infant to the potential harms associated with delaying surfactant therapy. In that respect infants experiencing CPAP failure are more susceptible to adverse outcomes including pneumothorax and BPD [10, 11]. The alternative of administration of surfactant via an ETT imposes the risk of volutrauma and barotrauma associated with positive pressure ventilation. The Intubate SURfactant Extubate (INSURE) method of surfactant delivery via brief intubation provides the hope of minimising exposure to mechanical ventilation. However, extubation is sometimes delayed and positive pressure ventilation continued, particularly in the most vulnerable babies [5, 6].
A newer and less invasive approach now available is the administration of surfactant via a thin catheter to an infant spontaneously breathing on CPAP, without use of positive pressure ventilation. This procedure was first described by Henrik Verder in 1992 as part of a small study of surfactant therapy for infants supported with CPAP [12]. The method of using a thin catheter rather than an endotracheal tube for surfactant delivery was rediscovered a decade or more later [13] and came to be known as less invasive surfactant administration (LISA). In the years since, the evidence regarding the potential benefits of LISA has accumulated, in observational studies and in randomised controlled trials, including the AMV [14], TakeCare [15], NINSAPP [16], and OPTIMIST-A trials [17]. In the wake of these trials, there is growing interest in the LISA technique worldwide as part of a comprehensive non-invasive respiratory support bundle. European, British, and Canadian consensus guidelines now recognise LISA as the preferred mode of surfactant delivery for preterm infants with RDS managed with CPAP [2, 8, 18, 19] and the method is acknowledged as being potentially preferable to ETT administration in a consensus statement from the USA [20].
The Scientific Concept of LISA
The scientific concept of LISA is primarily based on intratracheal administration of surfactant with no exposure to positive pressure ventilation thereby preventing lung injury [21]. Second, the use of an introducing catheter much thinner than an ETT allows glottic function to be preserved. In addition, with a thin catheter CPAP is presumably better transmitted from upper to lower airway whilst surfactant administration is underway. Third, with LISA (if done well), surfactant is dispersed from the trachea to the distal airspaces under the effect of spontaneous breathing, rather than PPV. This clearly appears to have an advantage, although preclinical studies have not provided conclusive proof and the evidence is hence largely circumstantial. Finally, maintaining spontaneous breathing for surfactant treatment seems to allow a more physiological transition to extrauterine life with regard to circulation and regional oxygen supply [22‒24].
What Are the Current Knowns regarding LISA?
LISA Effectively Prevents Exposure to Mechanical Ventilation
Randomised controlled trials including a total of >1,000 infants below 30 weeks’ gestation have clearly demonstrated that LISA – as compared to surfactant delivery via ETT (NINSAPP trial) or continuation of CPAP (AMV, OPTIMIST-A, CaLi) – considerably reduces the subsequent requirement for mechanical ventilation, both in the first 72 h and during the whole stay in hospital [14, 16, 17, 25] (Table 1). In the majority of infants above 26 weeks’ gestation intubation can be completely avoided following LISA treatment. In the NINSAPP trial, for the most vulnerable subgroup of infants <25 weeks’ gestation, LISA was effective in keeping infants free from ventilation during the first day of life [16], although intubation was needed in more than 90% of these infants at some time due to apnoea and/or muscular fatigue.
Author/study design . | Gestational age group . | LISA group threshold/method . | Control group threshold/method . | Results LISA versus ETT surfactant (%, p value), aRD (95% CI) . |
---|---|---|---|---|
LISA versus ETT surfactant administration | ||||
Kribs et al. [16] 13 centres Enrolled ≤2 h of life Poractant alfa 100 mg/kg | 23–26 weeks GA, n = 211 | FiO2 >0.3, CPAP 5–8; Silverman score ≥5; intubation criteria: FiO2 ≥0.45 for ≥2 h | FiO2 >0.3, CPAP ≥5–8 cm H2O Silverman score ≥5 Surfactant via ETT | Primary endpoint: ↔ Survival without BPD (67 vs. 59%, p = 0.2), 8.6 (−5 to 22) |
Secondary endpoints: ↑ Survival without major complications (51 vs. 36%, p = 0.02), 15 (1.4–28) | ||||
↓ MV during NICU stay (75 vs. 99%, p ≤ 0.001), 24 (16–34) | ||||
↓ Pneumothorax (5 vs. 13%, p = 0.04) | ||||
↓ Grade 3–4 IVH (10 vs. 22%, p = 0.02) | ||||
↔ Surgery for NEC, SIP; weight gain; treatment for ROP | ||||
LISA versus continued CPAP, selective ETT, and surfactant administration | ||||
Dargaville et al. [17] 33 centres Enrolled ≤6 h of life Double blind Poractant alfa 200 mg/kg | 25–28 weeks GA, n = 485 | FiO2 ≥0.3, CPAP 5–8, or NIV, intubation criteria: FiO2 ≥0.45, persistent apnoea LISA permitted only once (second surfactant dose via ETT) | FiO2 ≥0.3, CPAP 5–8 cm H2O or NIV Sham procedure (gentle repositioning) Intubation criteria: FiO2 ≥0.45, persistent apnoea; threshold for intubation and surfactant treatment as per discretion of attending neonatologist surfactant via ETT | Primary endpoint: ↔ Death or BPD (44 vs. 50%, p = 0.10), −6.3 (−14.2 to 1.6) |
Secondary endpoints: ↓ BPD (37 vs. 45%, p = 0.03), −7.8 (−15 to −0.7) | ||||
↓ Pneumothorax (4.6 vs. 10%, p = 0.005), −5.8 (−10.2 to −1.4) | ||||
↔ Mortality (12 vs. 8%, p = 0.30), 2.1 (−3.6 to −7.8) | ||||
↓ PDA medical treatment (35 vs. 45%), −11 (−20 to −1) | ||||
↓ Intubation within 72 h of life (37 vs. 72%, p ≤ 0.001), −36 (−47 to −24) | ||||
↓ Intubation during NICU stay (55 vs. 81%) −27 (−40 to −14) | ||||
↓ Oxygen at discharge (15 vs. 22%; −7.1 (−11.6 to−2.5) | ||||
↔ IVH, surgery for NEC, SIP; treatment for ROP | ||||
Göpel et al. [14] 12 centres Enrolled ≤12 h of life Surfactant: 80% poractant alfa 18% beractant 2% bovactant | 26–28 weeks GA, n = 220 | FiO2 >0.3, CPAP ≥4 cm H2O | Local guideline/physician-dependent threshold for intubation and surfactant treatment, surfactant via ETT | Primary endpoint: ↓ Need for MV (or if not intubated PaCO2 >65 or FiO2 >60% for ≥2 h) between hour 25 and 72 of life (28 vs. 46%, p = 0.008), −0.18 (−0.30 to −0.05) |
Secondary endpoints: ↓ MV during NICU stay (33 vs. 73%), −0.40 (−0.52 to −0.27) | ||||
↔ Mortality, BPD, pneumothorax, oxygen at discharge | ||||
↔ IVH, surgery for NEC, SIP; treatment for ROP | ||||
Katheria et al. [25] 3 centres Enrolled ≤1 h of life Poractant alfa 200 mg/kg Caffeine 20 mg/kg as controlled co-intervention within 2 h of life | 24–29 weeks GA, N = 180 | No threshold for LISA CPAP 6–8; intubation criteria: FiO2 ≥0.40, arterial pH ≤7.15, arterial pCO2 > 65 mm Hg, continued apnoea/bradycardia LISA permitted only once (second surfactant dose via ETT) | CPAP 6–8 intubation criteria: FiO2 ≥0.40, arterial pH ≤7.15, arterial pCO2 > 65 mm Hg, continued apnoea/bradycardia | Primary endpoint: ↓ MV within 72 h of lifec (21 vs. 47%, p ≤ 0.001), 0.26 (0.14–0.49) |
Secondary endpoints: ↓ oxygen at discharge (21 vs. 35%, p ≤ 0.001), OR 2.14 (1.05–4.19) | ||||
↓ BPD (24 vs. 39%, p ≤ 0.049), not adjusted | ||||
↔ Mortality, grade 3–4 IVH, pneumothorax, PDA medical treatment, surgery for NEC, SIP; treatment for ROP |
Author/study design . | Gestational age group . | LISA group threshold/method . | Control group threshold/method . | Results LISA versus ETT surfactant (%, p value), aRD (95% CI) . |
---|---|---|---|---|
LISA versus ETT surfactant administration | ||||
Kribs et al. [16] 13 centres Enrolled ≤2 h of life Poractant alfa 100 mg/kg | 23–26 weeks GA, n = 211 | FiO2 >0.3, CPAP 5–8; Silverman score ≥5; intubation criteria: FiO2 ≥0.45 for ≥2 h | FiO2 >0.3, CPAP ≥5–8 cm H2O Silverman score ≥5 Surfactant via ETT | Primary endpoint: ↔ Survival without BPD (67 vs. 59%, p = 0.2), 8.6 (−5 to 22) |
Secondary endpoints: ↑ Survival without major complications (51 vs. 36%, p = 0.02), 15 (1.4–28) | ||||
↓ MV during NICU stay (75 vs. 99%, p ≤ 0.001), 24 (16–34) | ||||
↓ Pneumothorax (5 vs. 13%, p = 0.04) | ||||
↓ Grade 3–4 IVH (10 vs. 22%, p = 0.02) | ||||
↔ Surgery for NEC, SIP; weight gain; treatment for ROP | ||||
LISA versus continued CPAP, selective ETT, and surfactant administration | ||||
Dargaville et al. [17] 33 centres Enrolled ≤6 h of life Double blind Poractant alfa 200 mg/kg | 25–28 weeks GA, n = 485 | FiO2 ≥0.3, CPAP 5–8, or NIV, intubation criteria: FiO2 ≥0.45, persistent apnoea LISA permitted only once (second surfactant dose via ETT) | FiO2 ≥0.3, CPAP 5–8 cm H2O or NIV Sham procedure (gentle repositioning) Intubation criteria: FiO2 ≥0.45, persistent apnoea; threshold for intubation and surfactant treatment as per discretion of attending neonatologist surfactant via ETT | Primary endpoint: ↔ Death or BPD (44 vs. 50%, p = 0.10), −6.3 (−14.2 to 1.6) |
Secondary endpoints: ↓ BPD (37 vs. 45%, p = 0.03), −7.8 (−15 to −0.7) | ||||
↓ Pneumothorax (4.6 vs. 10%, p = 0.005), −5.8 (−10.2 to −1.4) | ||||
↔ Mortality (12 vs. 8%, p = 0.30), 2.1 (−3.6 to −7.8) | ||||
↓ PDA medical treatment (35 vs. 45%), −11 (−20 to −1) | ||||
↓ Intubation within 72 h of life (37 vs. 72%, p ≤ 0.001), −36 (−47 to −24) | ||||
↓ Intubation during NICU stay (55 vs. 81%) −27 (−40 to −14) | ||||
↓ Oxygen at discharge (15 vs. 22%; −7.1 (−11.6 to−2.5) | ||||
↔ IVH, surgery for NEC, SIP; treatment for ROP | ||||
Göpel et al. [14] 12 centres Enrolled ≤12 h of life Surfactant: 80% poractant alfa 18% beractant 2% bovactant | 26–28 weeks GA, n = 220 | FiO2 >0.3, CPAP ≥4 cm H2O | Local guideline/physician-dependent threshold for intubation and surfactant treatment, surfactant via ETT | Primary endpoint: ↓ Need for MV (or if not intubated PaCO2 >65 or FiO2 >60% for ≥2 h) between hour 25 and 72 of life (28 vs. 46%, p = 0.008), −0.18 (−0.30 to −0.05) |
Secondary endpoints: ↓ MV during NICU stay (33 vs. 73%), −0.40 (−0.52 to −0.27) | ||||
↔ Mortality, BPD, pneumothorax, oxygen at discharge | ||||
↔ IVH, surgery for NEC, SIP; treatment for ROP | ||||
Katheria et al. [25] 3 centres Enrolled ≤1 h of life Poractant alfa 200 mg/kg Caffeine 20 mg/kg as controlled co-intervention within 2 h of life | 24–29 weeks GA, N = 180 | No threshold for LISA CPAP 6–8; intubation criteria: FiO2 ≥0.40, arterial pH ≤7.15, arterial pCO2 > 65 mm Hg, continued apnoea/bradycardia LISA permitted only once (second surfactant dose via ETT) | CPAP 6–8 intubation criteria: FiO2 ≥0.40, arterial pH ≤7.15, arterial pCO2 > 65 mm Hg, continued apnoea/bradycardia | Primary endpoint: ↓ MV within 72 h of lifec (21 vs. 47%, p ≤ 0.001), 0.26 (0.14–0.49) |
Secondary endpoints: ↓ oxygen at discharge (21 vs. 35%, p ≤ 0.001), OR 2.14 (1.05–4.19) | ||||
↓ BPD (24 vs. 39%, p ≤ 0.049), not adjusted | ||||
↔ Mortality, grade 3–4 IVH, pneumothorax, PDA medical treatment, surgery for NEC, SIP; treatment for ROP |
First % value refers to LISA group and second value refers to control group.
↑ Higher, ↓ lower, ↔no significant difference.
We only included trials who enrolled infants <30 weeks. A recent summary of trials on LISA in all gestational age groups has been published by Kakkilaya and Gauham [26].
GA, gestational age, MV, mechanical ventilation; PDA, hemodynamically significant PDA; OR, odds ratio, aRD, absolute risk difference.
Avoiding mechanical ventilation in the first 24 h by the use of LISA may be a key factor for reducing the known complications of prematurity that arise disproportionately in ventilated infants. In this regard, LISA appears to reduce the risk of intracranial haemorrhage [16, 27, 28]. Further, it seems clear that LISA can provide a lasting benefit arising from the lessening of exposure to volutrauma and barotrauma. Rates of BPD have been seen to be reduced in individual clinical trials (TakeCare, OPTIMIST-A) and also in meta-analyses of pooled data from clinical trials of LISA. A recent meta-analysis including the results of the OPTIMIST-A trial assessed 26 randomised controlled trials (RCTs) with a total of 3,349 infants [29]. Rates of BPD were considerably decreased in infants surviving to 36 weeks postmenstrual age who received LISA (RR 0.66; 95% CI: 0.51–0.85), and the composite outcome of death or BPD also favoured LISA (RR 0.71; 95% CI: 0.60–0.84). There was no significant difference in the outcome of in-hospital mortality.
In line with these findings in relation to BPD assessed during first hospitalisation, the 2-year follow-up of infants enrolled in the OPTIMIST-A trial has found advantages of LISA in relation to longer term pulmonary outcome, including a one-third reduction in the rate of re-hospitalisation with respiratory illness in the first 2 years, and a lessening in symptoms of chronic airway disease and need for bronchodilator therapy [30]. These are outcomes of considerable importance to the parents of children born preterm and have remarkable health-economic impact [31]. Infants in the LISA group of the NINSAPP trial followed up at 5–9 years had comparable lung function tests to control infants [32].
LISA Now Has an Established Safety Profile
Procedural Safety
As with standard intubation, the LISA procedure is known to invoke physiological instability, in large part related to laryngoscopy. Brief hypoxic episodes with oxygen saturation <80%, bradycardia with heart rate <100 beats/min and changes in cerebral oxygenation (as measured by near infrared spectroscopy) are regularly observed during the LISA procedure. Such instability can usually be managed with brief interruption of laryngoscopy attempts, and, if needed, a short period of positive pressure ventilation [33]. As previously stated, the use of a narrow bore catheter allows more effective airflow during spontaneous breathing, but on the other hand permits reflux of surfactant from the trachea into the pharynx, as has been noted with videolaryngoscopy (https://www.youtube.com/watch?v=IYf92NN1kV0).
Safety Outcomes during First Hospitalisation
Clinical trials to date have not found an impact of LISA (positive or negative) on mortality during first hospitalisation. A recent meta-analysis including data from 20 trials reported mortality in infants receiving LISA to be similar to that in standard treatment groups (11.2% vs. 14.3%, RR 0.78, 95% CI: 0.58–1.05) [29]. In most of these trials infants in the comparator group received exogenous surfactant via ETT. For the two trials included in the meta-analysis in which the comparison was with continuation of CPAP and selective surfactant therapy if required (ref AMV, OPTIMIST-A), the effect on death during first hospitalisation if anything favoured the control group (8.8 vs. 6.7%, RR 1.31, 95% CI: 0.79–2.19). In the OPTIMIST-A trial, for the outcome of mortality, subgroup analysis found the treatment effect to favour the LISA group in the gestation range 27–28 weeks, and the control group at 25–26 weeks, with a significant interaction between group assignment and gestation stratum for this outcome. At all gestations, deaths in the OPTIMIST-A trial had multiple causes, occurred at a range of postnatal ages and were not considered to be directly related to the intervention [17]. There was no concern regarding a negative effect on mortality in the NINSAPP trial involving infants at 23–26 weeks gestation (overall result: LISA 9.3% vs. 11.5%, RR 0.81, 95% CI: 0.37–1.79) [16]. Favourable results regarding survival after LISA have been noted down to 22 weeks’ gestation in the German Neonatal Network (GNN) cohort [27]. These data were extracted from centres in one country with a relatively uniform approach to the management of imminent preterm birth and delivery room stabilisation for extremely preterm infants including a quasi-prophylactic approach of LISA in >80% of surfactant-treated infants <27 weeks‘ gestation. Furthermore, GNN sites have more than 15-year experience with the LISA approach. This reinforces the need for training and experiential learning for the full benefits of LISA to be realised in the most vulnerable infants, especially in settings in which baseline mortality for such infants remains high.
Clinical trial and observational data have also not raised major concerns for other adverse effects of LISA during first hospitalisation. In observational GNN data, a higher rate of spontaneous intestinal perforation (SIP) was noted following LISA amongst infants below 26 weeks’ gestation, with in regression modelling the rates being progressively higher as gestation decreased [34]. A possible explanation might be the extreme fragility of the gut wall, compounded by drug exposure (steroids, non-steroidal anti-inflammatory drugs) and further exacerbated by intestinal loop gaseous distension (“CPAP belly”), a frequent phenomenon when using non-invasive respiratory support in the most immature infants. An increase in risk of SIP was not detected in RCTs reporting this outcome [14, 16, 17, 25], where there were a combined total of 17 SIP cases out of 548 LISA-treated infants as compared to 16 SIP cases in 548 control group infants.
Safety Outcomes in Follow-Up
With regard to 2-year neurodevelopmental outcome, 3 RCTs (AMV, NINSAPP, OPTIMIST-A) have so far reported follow-up data based on neurodevelopmental examination (Bayley-Scales) and parental reports (PARCA-R) which are summarised in Table 2 [30, 35, 36]. These data suggest that the LISA intervention is not inferior to standard treatment. This is an important finding, given that LISA involves performing a procedure that may cause physiological destabilisation, including episodes of hypoxia and bradycardia, at a critical time of life. On the other hand, there was not a neurodevelopmental advantage, in spite of a somewhat lesser rate of BPD amongst infants receiving LISA, BPD being a known risk factor for adverse neurodevelopment outcome in some but not all studies.
Parameter . | Control, n/N (%) . | Intervention . | p value . |
---|---|---|---|
Death before 2 years | |||
AMV (26–28 weeks GA) | 5/112 (4.5) | 9/108 (8.3) | |
NINSAPP (23–26 weeks GA) | 14/104 (13.5) | 15/107 (14.0) | |
OPTIMIST-A (25–28 weeks GA) | 24/229 (10.5) | 29/224 (12.9) | |
Total | 43/445 (9.7) | 53/439 (12.0) | 0.25 |
Neurodevelopmental disability at 2 years postmenstrual age | |||
AMV | 15/69 (21.7) | 24/58 (41.3) | |
NINSAPP | 42/69 (60.9) | 33/68 (48.5) | |
OPTIMIST-A | 55/195 (28.2) | 49/186 (26.3) | |
Total | 112/333 (33.6) | 106/312 (33.9) | 0.93 |
Parameter . | Control, n/N (%) . | Intervention . | p value . |
---|---|---|---|
Death before 2 years | |||
AMV (26–28 weeks GA) | 5/112 (4.5) | 9/108 (8.3) | |
NINSAPP (23–26 weeks GA) | 14/104 (13.5) | 15/107 (14.0) | |
OPTIMIST-A (25–28 weeks GA) | 24/229 (10.5) | 29/224 (12.9) | |
Total | 43/445 (9.7) | 53/439 (12.0) | 0.25 |
Neurodevelopmental disability at 2 years postmenstrual age | |||
AMV | 15/69 (21.7) | 24/58 (41.3) | |
NINSAPP | 42/69 (60.9) | 33/68 (48.5) | |
OPTIMIST-A | 55/195 (28.2) | 49/186 (26.3) | |
Total | 112/333 (33.6) | 106/312 (33.9) | 0.93 |
Outcomes at 2 years in RCTs of LISA.
Neurodevelopmental disability defined as Mental Developmental Index or Psychomotor Developmental Index [37].
<85 for the AMV and NINSAPP trials and any of (i) moderate to severe cognitive or language impairment, MDI <80; (ii) cerebral palsy equivalent to Gross Motor Function Classification System [38] >2; (iii) visual impairment; or (iv) hearing impairment for the OPTIMIST-A trial.
The Methodology of LISA Is Highly Variable
Many different modifications of the LISA procedure now exist. The LISA catheter may be introduced orally or even nasally, with or without use of Magill forceps. The original description and evaluation of LISA involved the placement of a thin and very flexible catheter through the glottis during direct laryngoscopy, with the aid of Magill forceps. The Hobart method is an alternative for neonatal settings where oral intubation without Magill forceps is more popular, using a less flexible (“semi-rigid”) vascular catheter which can be guided into the trachea without using forceps. In one study, a rigid catheter was found to be superior than a soft catheter in relation to speed of performance [39]. Other investigators reported the use of a very short but flexible catheter that was placed in the trachea without Magill forceps [15]. Nowadays most centres use purpose-built catheters designed for LISA or umbilical catheters (3.5–6 French diameter). Insertion depth beyond the vocal cords can be adjusted according to gestational age, with a pragmatic approach being: <27 weeks: 1.5 cm; ≥27 weeks: 2 cm [40, 41]. Catheters which have depth markings provide an advantage in guiding depth beyond the cords [42]. Several studies suggest the use of videolaryngoscopy to improve the procedural success of LISA [43, 44]. Operator competence is important and laryngoscopy attempts should be limited to 60 s or less if there is hypoxia or bradycardia. Once the catheter is inserted and the infant is breathing comfortably the speed of surfactant administration can be adjusted to the infant’s respiratory stability. In most cases surfactant application can be completed in 1–2 min and the catheter can be removed. Whatever method is chosen there needs to be training focussing on both the LISA procedure and less invasive respiratory care as a team approach [40, 44]. Video demonstrations of the LISA procedure are available online: https://www.youtube.com/watch?v=OUvgJ57FQR8 (Cologne method); https://youtu.be/9FvSktdn8Hg (Hobart method); https://youtu.be/C74-MLI6Rac (videolaryngoscopy and LISA).
What Are the Current Uncertainties about LISA?
LISA: Indications and Timing
Previous studies of (evolving) RDS have defined indications for surfactant treatment based on clinical signs of respiratory distress (tachypnoea, grunting, retractions, Silverman score) and oxygen requirement. The threshold for LISA treatment in many centres has moved down to a FiO2 of 0.30 (30%) which is mainly based on observations of CPAP failure rates [2]. Taken together, the available studies of CPAP failure argue for both gestation and age-specific FiO2 thresholds for surfactant administration [2, 9, 33, 45, 46]. There remains ongoing debate about the influence of airway pressure on oxygen requirement, and hence whether applied pressure should be included in the decision-making algorithm. As an adjunct to clinical decisions regarding timing of LISA, bedside tests of surfactant deficiency such as gastric aspirate analysis are available at point-of-care but need to be subjected to clinical trials [8, 47]. Recently, lung ultrasound has been extensively studied and may offer an alternative way of accurately predicting the need of surfactant replacement at an earlier stage and guiding individualised management [48, 49].
The LISA procedure should be considered as part of a complex non-invasive respiratory strategy. For its successful application LISA requires an optimal setting, comprehensive perinatal care, and team competence. Key indicators include prenatal transport to a tertiary level care centre, administration of antenatal steroids at the point of viability, Caesarean section with local anaesthesia as preferred mode of delivery in extremely preterm infants and delayed cord clamping to facilitate physiological cardiopulmonary transition [33]. A crucial feature for successful LISA is CPAP transmission using a functional interface with continuous monitoring of pressure delivery without substantial pressure leaks. In the Cologne protocol dedicated to the most immature infants, criteria for consideration of LISA after initial stabilisation are: (a) breathing spontaneously with CPAP support, (b) heart rate greater than 120 beats per minute, and (c) oxygen saturation greater than 85% [50]. A strong balance between awareness of the infant’s vulnerability and harnessing the infant’s vitality (capacity to breathe spontaneously) is required.
The book of prophylactic surfactant appeared to be closed a decade ago when several studies noted that in infant populations with high rates of antenatal steroid exposure and the routine use of CPAP prophylactic surfactant (by intubation) gained no benefit from prophylactic as compared to selective surfactant treatment [51]. However, with the broad implementation of LISA across German NICUs, a large proportion of neonatologists would now consider administration of surfactant in the delivery room in infants <27 weeks’ gestation [27, 52]. This is reinforced by data of the CaLI trial where one-third of the infants in the intervention group received LISA at an FiO2 <0.3 in a quasi-prophylactic fashion. The CaLI trial included 180 infants born at 24–29 weeks’ gestation in which caffeine was given as controlled co-intervention in the first 2 h of life. For the primary endpoint, avoidance of mechanical ventilation or pre-defined respiratory failure criteria in the first 72 h of life, caffeine, CPAP plus LISA proved to be advantageous over caffeine, CPAP and selective surfactant treatment (23 vs. 53%; OR 0.23, 95% CI: 0.11–0.46). The benefit of caffeine and LISA persisted with decreasing gestational age – need for intubation (the primary endpoint) was reduced in the subgroup of 57 infants <27 weeks’ gestation (44 vs. 80%). As with most studies on LISA, apart from the OPTIMIST-A-trial, CaLI was not blinded, a fact which deserves careful interpretation since endotracheal intubation and mechanical ventilation decisions were made with the attending physicians aware of whether surfactant had already been administered. Although intubation criteria were pre-defined, there was a potential bias toward providing surfactant to the CPAP group. With regard to the timing of LISA the CaLI trial may re-open the book on prophylactic surfactant in the LISA era based on theoretical reasoning regarding the bolstering of the surfactant pool before the symptomatology of RDS develops, and the potential for more even surfactant distribution when given early. Further evidence will come from the pro.LISA trial, which is currently investigating whether prophylactic LISA in the delivery room could be of advantage for infants <31 weeks’ gestation, with particular interest in lung function during childhood [53].
CPAP or Non-Invasive Positive Pressure Ventilation Combined with LISA?
The use of different modes of non-invasive respiratory support during LISA has not yet been well studied. A survey of Maiwald et al. [52] reported, amongst all modes of respiratory support in use, nasal CPAP was the most frequently used (67%), followed by non-synchronised intermittent positive pressure ventilation (17%) and synchronised intermittent positive pressure ventilation (14%). Small studies have noted that by using non-invasive positive pressure ventilation (NIPPV) as primary respiratory support, LISA and INSURE had comparable outcome [54] while LISA+NIPPV was either superior to NIPPV alone [55] or to INSURE+CPAP [56]. Oncel et al. [57] compared CPAP with NIPPV with the selective use of LISA and found that NIPPV was associated with a lower rate of positive pressure ventilation and surfactant treatment.
Dose of Surfactant?
The AMV trial used a surfactant dose of 100 mg/kg body weight and allowed all 3 different surfactants that were available in Germany at that time [14]. This was in line with a variability in phospholipid concentration (25–80 mg/mL) and bolus volumes (1.25–4.0 mL per kg body weight). The OPTIMIST-A trial and CaLI trial used a surfactant dose of 200 mg/kg. Theoretically, larger surfactant volumes administered intra-tracheally with CPAP support might interfere with spontaneous breathing, although in practice this problem is infrequently encountered. The NINSAPP trial used a whole vial approach (within the range 100–200 mg/kg body weight [16] which may continue to apply for surfactant administered in the delivery room. While retrospective data suggest that surfactant doses of 200 mg/kg body weight [58, 59] are associated with reduced rate of mechanical ventilation or re-treatment, randomised trials on the optimum surfactant dose for the LISA approach are needed [60].
Repeat Surfactant Dosing Using LISA
For infants showing a rebound in oxygen requirement after an effective first LISA procedure, a second LISA treatment may become necessary. The need for further surfactant dosing should prompt consideration of whether the second surfactant dose can be given via LISA, or alternatively whether intubation and surfactant administration via ETT should be opted for. Klijkers et al. [61] retrospectively analysed a cohort of 209 preterm infants requiring a second surfactant dose, including 132 in which the LISA procedure was repeated. Subsequent CPAP failure was observed in 42% of cases receiving the second dose via LISA and was associated with lower gestational age and an oxygen requirement above 50%. Notably, infants receiving repeat LISA had a lower risk of death or BPD compared to infants intubated for the second surfactant dose [61]. Timing of a second dose is controversial. Current survey results imply that most neonatologists would repeat LISA 4 h after a first dosing, if that proved to be effective. Where a LISA procedure appeared to be unsuccessful in eliciting a surfactant response, one-third of survey respondents stated they would repeat LISA immediately, and one-half indicated they would redose after 1 h [52].
To Sedate or Not to Sedate during LISA?
Current recommendations on non-emergent endotracheal intubation include the use of premedication. There is an ongoing debate about whether this recommendation should also apply to LISA [62, 63]. In most RCTs and observational studies on LISA, non-pharmacological measures (intra-oral sucrose, swaddling, facilitated tucking) have been applied successfully to improve the tolerance of the procedure [64]. Different drugs such as propofol, ketamine, fentanyl, and morphine have been evaluated in comparison to non-pharmacological measures. Systematic reviews conclude that premedication can reduce pain/discomfort as assessed by scores such as COMFORT-Neo but increases the risk of adverse effects such as desaturation, apnoea, and failure to maintain the concept of spontaneous breathing. Therefore, an individualized approach is needed based on gestational age, vigorous appearance, and postnatal age [65, 66]. The authors of this review would not routinely recommend pre-medication, however, in selected infants with high levels of discomfort a slow bolus infusion (5–10 min) of fentanyl (0.5–1 micg/kg) would be a widely acceptable option for analgesia. Recently, the trial protocol for a non-pharmacological approach to during LISA, the NONA-LISA trial (NCT05609877) has been published [67].
Alternatives to LISA
There are alternative less-invasive means of giving surfactant, with the three methods to date being nebulisation, pharyngeal administration and surfactant via laryngeal mask or supraglottic airway (SALSA). Surfactant nebulisation has been studied for many years as an attractive, truly non-invasive option of surfactant delivery as it requires no instrumentation of the airways. It has been proven feasible with different types of nebulisers (e.g., pneumatically driven, vibrating membranes, capillary-driven). However, achieving an adequate rate of alveolar deposition of surfactant remains a technical challenge, in particular in infants <28 weeks’ gestation with low tidal volumes. Current studies have not yet shown a net benefit on avoiding mechanical ventilation or BPD prevention [68]. Technical advances like breath-synchronised nebulisation and increasing tidal volumes by stimulation of breathing may be of relevance for optimising the method of surfactant nebulisation.
Oropharyngeal delivery of surfactant has been evaluated in the POPART study [69]. This RCT failed to demonstrate a difference in the primary outcome of mechanical ventilation during the first 120 h with no other benefits in secondary outcomes, and the concern raised by an increased risk of pneumothorax in infants receiving pharyngeal surfactant. Potential explanations for these findings include the timing of the procedure (at birth, with the lungs opened but still filled with fluids), a lack of sufficient spontaneous breathing as well as inadequate surfactant deposition in the lung. These influences require further study [70].
The SALSA method has been evaluated in several RCTs in comparison to either CPAP alone or with INSURE and has proven to have benefits with regard to avoidance of mechanical ventilation and reduced oxygen requirement [71, 72]. The SALSA approach raised no safety issues in terms of risk of mortality, BPD and pneumothorax and is therefore regarded as effective alternative to LISA in infants >1,000 g birth weight (2). The approach can be easily performed by less-experienced clinicians and has been successfully used in low- and middle-income countries. A LISA catheter could be inserted intratracheally through a laryngeal mask [73]. Further research is necessary to perform adequately powered RCTs to detect possible differences in BPD as well as head-to-head comparisons with LISA. One such study is underway in preterm infants >1,250 g: SURFSUP trial (ACTRN12620001184965) [74].
Future Directions
Given the increasing focus on value-based care, health economic effects of LISA and alternatives will have to be considered. Federici et al. [75] noted that LISA, given at FiO2 ≥0.3 is expected to reduce resource consumption and costs compared with CPAP alone for preterm infants. In their calculation, cost savings of -GBP 5,146 (95% credible interval, -GBP 22,403 to GBP 13) were anticipated for infants born at 25–28 weeks’ gestation and -GBP 176 (95% credible interval -GBP 4,279 to GBP 339) for infants born at 29–32 weeks’ gestation. Probabilities of cost savings were 85–97%, with the prevention of short-term complications such as IVH or BPD being the most influential parameters on health-economic outcomes. Recently, a nation-wide cohort study from Taiwan described a higher risk of serious infections in the first 12 month of life in infants exposed to antenatal steroid treatment in the time-frame between 24 and 34 weeks of gestation [76]. This adds to the discussion that potential long term risks of antenatal steroids should be carefully weighed against the well acknowledged benefits of steroids in the perinatal period. It also emphasizes that early lung maturation measurements including lung ultrasound [47, 48] and LISA are increasingly important means of balanced, personalised care. More research is needed on benefits of the LISA approach for moderate and late preterm infants [26, 77]. Another important question remains in defining the subgroup of infants for whom LISA will not avoid CPAP failure. Adjusted analyses of cohort data suggest that low gestational age and the presence of early onset infection (as evidenced by increased levels of C-reactive protein) are critical risk factors for treatment failure after LISA [78]. Defining different infant RDS endotypes including genetic predisposition [79] and risk profiles with consideration of human (team) factors related to LISA administration will be crucial for targeting individual trajectories. In line with that, Kloonen et al. [80] recently reported exploratory data on the use of artificial intelligence to predict LISA failure in 18 out of 51 infants on CPAP. The best machine learning model to predict LISA failure within 30 min after administration included gestational age and (possibly modifiable) vital signs such as heart rate variability, respiration rate, and SpO2. These tools will support personalised approaches to guide LISA in vulnerable infants.
Conflict of Interest Statement
C.H., A.K., W.G., and E.H. have received honoraria for presentations and travel support from Chiesi Farmaceutici, a surfactant producer. C.H. received research support for a study on LISA/CPAP failure from Chiesi Farmaceutici. C.H. and E.H. served as advisors for Draeger Medical, a company producing incubators, monitors and ventilators. C.H. and E.H. contributed to German national guidelines related to the care of vulnerable newborn infants. PAD reported receiving personal fees from Chiesi Farmaceutici (advisory board consultancies 2021) and provision of surfactant at reduced cost and support for conference travel from Chiesi Farmaceutici; in addition, PAD has a patent for a catheter design (USD752215S) issue.
Funding Sources
This review summarises in part scientific presentations and interactive plenary sessions of the 3rd International LISA workshop (LISA III) in Lübeck, Germany which took place on Sept 24th/25th 2023 and was supported by the German Research Council (DFG-HE-2072/5-1).
Author Contributions
C.H., A.K., W.G., P.A.D., and E.H. debated the series of articles, studies and presentations at the LISA III workshop, made editorial comments and assured the validity of the content of this review. C.H. drafted this article initially, with comments and involvement from all authors.