Background

For more than half a century, neonatologists around the world have been puzzled by searching for the most optimal strategies for the management of children with extremely low body weight (ELBW), especially those with a birth weight less than 750 grams. Survival of these children in the United States was only 20% in the mid-1980s [1], in Japan — up to 40% [2]. By the mid-90s, survival of children with a birth weight less than 750 grams was increased up to 40% in the United States. However, the authors emphasized high risk of disability among these patients [3]. In the Netherlands, survival rate in children with a birth weight less than 750 grams had already reached 88% by 2010 [4].

The main cause of morbidity and mortality in premature infants, especially those with a birth weight less than 750 grams, is respiratory disorders. The last ones are primarily caused by severe lung immaturity [5] followed by bronchopulmonary dysplasia (BPD) [6].

Various researchers have studied the physiological characteristics of lungs in newborns in order to understand lung properties and search for the most optimal approaches to mechanical ventilation in newborns. Thus, Karlberg P. et al. (1962) concluded that a full-term newborn has short first breaths with a high peak pressure that can contribute to the most optimal air distribution in the lungs and formation of functional residual capacity [7]. Arjan B te Pas et al. (2009) analyzed the first breaths in spontaneously breathing full-term and premature infants and found similar high tidal volumes (6.7 vs. 6.5 ml/kg, p = 0.5) [8]. Greenough A. et al. (2017) reported no need for a longer inflation time to increase expiratory tidal volume at a higher inspiratory pressure in premature infants with gestational age less than 34 weeks [9]. Another authors found that prolonged inhales during initial respiratory therapy were ineffective and did not affect the outcomes [10].

In the above-mentioned studies, we did not find data on the features of initial respiratory support and changes in mechanical properties of lungs in premature newborns with a birth weight less than 750 grams undergoing volume guarantee ventilation in the delivery room. Evolution of respiratory mechanics after surfactant administration has been previously described. However, the majority of studies included children with a birth weight over 750 grams and gestational age near 28 weeks [11]. In children with a birth weight less than 750 grams and extremely immature lungs, various respiratory strategies are currently based on individual characteristics in order to maintain adequate oxygenation. However, the most optimal modes of ventilation are still searched. The protocols and clinical guidelines on the management of newborns with respiratory distress syndrome request various options for invasive ventilation, in particular, traditional pressure-controlled ventilation with a respiratory rate of 40-60 per minute [12]. In most hospitals, ventilation of newborns in the delivery room is currently performed using manual ventilators with a T-connector and positive end expiratory pressure (PEEP) valve. This ventilation can ensure a fixed inspiratory pressure and smooth regulation of inhaled oxygen fraction (FiO₂) [13]. However, these devices do not allow tidal volume monitoring or volume guarantee ventilation. Only modern neonatal devices equipped with microprocessors ensure the latter. This equipment can detect extremely small changes in tidal volume (within 0.5 ml) [14, 15]. This feature is especially important in children with extremely low birth weight (ELBW), since tidal volume usually does not exceed 3.8 ml/kg in such patients (2.8-4.7 ml/kg) [16].

That is why earlier studies devoted to volume guarantee ventilation in newborns with ELBW were not possible, and the recommendations for ventilation were based on empirical choice of the main respiratory parameters.

We assumed that volume guarantee ventilation in newborns with a birth weight less than 750 grams would be able to minimize the risk of ventilator-induced lung injury. Minimum respiratory volume sufficient for adequate gas exchange is used in these cases. Considering unclear literature data on volume guarantee ventilation in newborns with a birth weight less than 750 grams in the delivery room, we conducted our own research. The aim was objective analysis of mechanical properties of lungs in children with ELBW since the first minutes of life.

Material and methods

A prospective clinical study was conducted in a single perinatal center between January 2019 and June 2020. Inclusion criteria: birth weight less than 750 grams, gestational age ≤ 25 weeks, strict compliance with the protocol of management in delivery room and later. Premature newborns without effective spontaneous breathing were intubated immediately after birth and ventilation in PC-AC + VG mode was initiated (Pressure Control Assist Control + Volume Guarantee) (Dräger Babylog VN500 ventilator, Dräger Medical, Lübeck, Germany) [17, 18]. Dräger Babylog VN500 ventilator is equipped with an accurate flow sensor (anemometer with a double heating coil). This sensor is able to detect the minimal changes in tidal volume during conventional ventilation and high-frequency oscillatory ventilation [19, 20]. Another peculiarity of Dräger Babylog VN500 ventilator is measurement of gas volumes in the exhaled air that increases an accuracy of guaranteed tidal volume even at minimum values [18].

Initial tidal volume was 4.0 ml/kg because this value is recommended as minimum for ventilation in neonates [21]. Effective spontaneous breathing was assessed immediately after birth. In case of ineffective breathing, tracheal intubation was followed by ventilation with the following parameters: PC-AC + VG mode, tidal volume 4.0 ml/kg, PIPmax = 50 cmH₂O (increase in PIPmax up to 100 cmH₂O if necessary), respiratory rate 60 per minute, inspiration time (Tin) 0.25 sec, FiO₂ 0.21. Vital monitoring including ECG was ensured by multifunctional monitor (IntelliVue MX500 Modular Monitor, Philips Healthcare, USA). FiO₂ was adjusted to maintain the target oxygen saturation in accordance with Dawson's criteria [22].

In 10 minutes after birth, various natural surfactants were injected (registered in the Russian Federation in accordance with the instructions): Poractant Alfa (Curosurf®, Chiesi Farmaceutici, Italy) — four children, Beractant (Survanta®, AbbVie Inc., USA) — five children, Bovaktant (Alveofact®, LYOMARK PHARMA, GmbH, Germany) — two children. Information about certain type of surfactant was unknown in comparative analysis of mechanical properties of lungs. Only the fact of surfactant administration was taken into account. All newborns were stabilized in infant incubator combined with open intensive care system and radiant heating (Modification-transformer "Giraffe Omnibed", GE Healthcare, USA or Incubator-transformer "Atom Dual Incu i", ATOM Medical, Japan). After that, children were transferred to the neonatal intensive care unit under traditional ventilation (Dräger Babylog VN500 ventilator) in the same mode. Body temperature was maintained within normothermia (36.5 — 37.5°C).

In the neonatal intensive care unit, ventilation parameters were registered within 120 minutes of life. All respiratory data were later copied and recorded in Microsoft Excel spreadsheet on a USB flash drive. In case of progressive respiratory failure (FiO₂ increase up to 0.3 and MAP over 7.0 cmH₂O to maintain target oxygen saturation 90–94%), repeated administration of surfactant and/or high-frequency oscillatory ventilation (SLE 5000 ventilator, SLE Limited, Great Britain) were applied. Dräger Babylog VN500 ventilator saves respiratory parameters every five minutes (not more often). Therefore, analysis of mechanical properties of lungs within 30 minutes of life included those indicators measured every five minutes. Since the changes were minimal after 30 minutes of respiratory therapy, we identified the following the most significant points for assessing the respiratory indicators in all patients: 5, 10, 15, 20, 25, 30, 60, 90, 120 minutes of life.

There were 21 live-born babies with a birth weight less than 750 grams throughout the entire study period. As a result, only 11 newborns were enrolled. Ten children were excluded from the study for various reasons: gestational age ≥ 26 weeks (n=5), emergency delivery without prepared equipment (n=1), violation of the study protocol (necessary guaranteed volume was not prescribed since the birth, n=1), lost patient's data (n=1, device change without keeping the trends), air leakage past the tube over 25% (n=2) (Fig. 1).

Fig. 1. Distribution of patients with birth weight less than 750 grams over the entire study period.

Histological examination of lungs was carried out in some deceased children to assess the effect of PC-A/C + VG ventilation on the lungs of children with birth weight less than 500 grams.

We used the Microsoft Excel database to summarize the data. Database was a depersonalized spreadsheet with a numerical expression of 57 criteria. For continuous variables with normal distribution, mean and standard deviation (SD) were determined. Distribution normality was assessed using the Shapiro-Wilk test. In case of abnormal distribution, we used the median and interquartile range. Paired Student’s t-test was used to compare mean values of continuous variables with normal distribution. In other cases, Wilcoxon test was applied. The relationship of features was analyzed using Spearman's correlation analysis, since this method can be used for abnormal distribution of continuous variables, as well as for rank and dichotomous variables. Differences were significant at p-value <0.05.

Results

Clinical characteristics of children are shown in Table 1.

Table 1. Demographic and clinical characteristics of patients

Note. * mean value ± standard deviation (SD) / median (minimum value — maximum value); ** the number out of total (%).

All patients had an extremely low birth weight and Apgar score at 1 and 5 minutes that indicated on severe immaturity and moderate-to-severe asphyxia at birth. Tracheal intubation was followed by ventilation with preset parameters in all children in the delivery room within 90 — 180 seconds after birth. Tidal volume was determined considering fetometry data before childbirth (4.0 ml/kg). Thus, there were approximately 120 — 180 hardware breaths in each child by the moment of recording the first respiratory values by Dräger Babylog VN500 ventilator. Inspiration pressure was determined by the need to deliver guaranteed tidal volume through a 2.0-2.5 mm endotracheal tube.

More than 90% (10 out of 11) of newborns required repeated administration of surfactants. Indications for repeated administration of surfactant were clinical signs of respiratory distress syndrome and severe respiratory failure (MAP over 7.0 cm H₂O, FiO₂ over 0.3 to maintain the target oxygen saturation 90–94%). Repeated administration of surfactant was carried out in 8.8 ± 3.9 hours after birth (median 9.5 hours, range 3.0 — 14.0). High-frequency oscillatory ventilation within 72 hours after birth was required in 63.6% of patients. This mode of ventilation was initiated in 9.9 ± 8.4 hours after birth (median 6.0 hours, range 2.0 — 24.0).

Mechanical properties of lungs in premature newborns within 2 hours of life are shown in Table 2.

Table 2. Mechanical lung properties in premature newborns within two hours of life

Note. * mean ± standard deviation (SD) / median (minimum — maximum); MVe — Expiratory minute volume; Cdyn — dynamic compliance; C20/Cdyn — the ratio of the last 20% of inspiration to the total dynamic compliance; Raw — resistance in the patient's airway; TC — time constant. † — p <0.05 compared to the first five minutes of the study (Wilcoxon T-test).

Percentile distribution of inspiratory pressure and its changes after 5, 10, 15, 20, 25, 30, 60, 90 and 120 minutes are shown in Fig. 2.

Fig. 2. Decrease in inspiratory pressure in premature newborns with birth weight less than 750 grams in the delivery room throughout 120 minutes of life (percentile distribution).

Surfactant administration time — the 10^th minute of life (A), mean surfactant action time (B) — in 10-15 min after administration.

According to these data, peak pressure gradually decreased over time in all children. PIP range was very wide (20.0 — 79.0 cmH2O) at the 5^th minute of life in premature newborns that was probably associated with individual characteristics of patients. By the 120^th minute, range of values was much less. Maximum PIP was also recorded at the 5^th minute of life with subsequent gradual decrease by 1.5-2.0 times by the 10^th minute. Thus, mean PIP at the fifth minute (PIP05) was 43.18 ± 15.99 cmH₂O and significantly decreased up to 31.18 ± 13.83 cmH₂O by the 10^th minute (p = 0.005; Wilcoxon test). Subsequently, all children had similar dynamics of PIP at all time points. PIP was significantly lower compared to the values at the 5^th minute of life. For example, PIP at the 15^th minute was 27.40 ± 13.25 cmH₂O (p = 0.005; Wilcoxon's test), at the 120^th minute — 18.90 ± 4.32 cmH₂O (p = 0.004; Wilcoxon’s test). Despite high peak pressure, children were not diagnosed with air leakage syndrome within 72 hours after birth.

It should be noted that PIP changes after administration of surfactant were similar in all children regardless the type of surfactant. Moreover, there was no significant decrease in PIP after administration of surfactant during volume guarantee ventilation (Fig. 3).

Fig. 3. Peak pressure depending on age with determination of the mean (cross) or median (horizontal line in a rectangle), as well as range of values for each time interval.

A numerical value marked with one asterisk (* — Student’s t-test) or two asterisks (** — Wilcoxon test) indicates significant differences between the peak pressure at the 5^th minute (a) and surfactant administration time at the 10^th minute (b) with other time intervals.

Maximum difference in peak pressure was noted between the 5^th and the 10^th minutes of life (p = 0.005), that cannot be said about subsequent changes in peak inspiratory pressure. Surfactant administration was followed by inspiratory pressure decrease. However, this process occurred with an interval of 15 min rather immediate as it happened before the 10^th minute of life. Significant differences with PIP at the 10^th minute of life after surfactant administration appear only after the 25^th minute (p = 0.045; Wilcoxon's test) and persist until the 120^th minute (p = 0.005; Wilcoxon's test).

Changes in respiratory monitoring were different and comparable with changes in peak pressure in some signs. The lowest compliance (Cdyn) was determined at the 5^th minute (0.15 ± 0.13 ml/cmH₂O/kg). Subsequently, surfactant substitution therapy and volume guarantee ventilation resulted almost 2-fold compliance increase by the 120^th min (0.26 ± 0.20 ml/cmH₂O/kg, p = 0.018; Wilcoxon test). Airway resistance (Raw) showed similar dynamics. High Raw was also determined within 5 minutes of life (mean 412.6 cmH₂O/L/sec by the 5^th minute, range 174 — 551 cmH₂O/L/sec). However, we observed significant 2-fold decrease of resistance by the 120^th minute in all children (226.9 cmH₂O/L/sec, p = 0.013; Wilcoxon test).

At the same time, C20/Cdyn remained the same within 5-120 minutes (range 0.78 — 0.79, p = 0.851; Student's test). These features pointed on incomplete lung inflation in children. Time constant (TC) remained short throughout the entire follow-up period in all patients and was not significantly changed (0.043 sec at the 5^th minute, 0.045 sec at the 120^th minute, p = 0.631; Student's test) despite the changes in resistance and compliance. At the same time, inspiratory time variability was quite high (range 0.10 — 0.43 sec). Air leakage past the endotracheal tube was registered in 10–12% of cases (range 0 — 24%).

In our opinion, the girl M. with a birth weight of 472 grams and gestational age of 23 weeks deserves a special attention. Peak inspiratory pressure ensuring target tidal volume (4 ml/kg) reached 79 cmH₂O at the 5^th minute (Fig. 4).

Fig. 4. Mechanical lung properties of the girl M. (birth weight 472 grams, gestation age 23 weeks) in the delivery room.

a — birth time; b — surfactant administration time; c — oxygen decrease in 15 min after surfactant administration; d — compensatory automatic increase in peak pressure at the moment of progressive air leakage past the endotracheal tube in order to maintain a stable tidal volume (e).

The girl M. underwent tracheal intubation within 2 minutes of life and volume guarantee ventilation was initiated. Peak inspiratory pressure of 79 cmH₂O was required by the 5^th minute to maintain the tidal volume of 4.0 ml/kg. The outcome was favorable in this case. Air leakage syndrome and central nervous system damage were absent. The child did not need for additional oxygen therapy at 36 weeks of postconceptional age (PCA). The patient was discharged at the 50^th week of PCA.

At the same time, high PIP can result lung damage. Signs of injury may be visualized during histological examination. Histological specimens of lungs of the child K. with a birth weight of 470 grams and gestational age of 23 3/7 weeks are shown in Fig. 5. Maximum peak pressure at the 5^th minute was 42 cmH2O. The patient died after 5 days from early neonatal sepsis.

As one can see in Fig. 5, morphological picture of lungs is represented by immature structures in the form of rounded cavities with irregular shape covered by cube-like and cylindrical epithelium with vacuolization. These signs correspond to pseudostratified simple cylindrical endodermal epithelium typical for pseudoglandular structure (Fig. 5C).

Fig. 5. Microscopic image of lung specimens of a 470-gram premature infant, gestation age 233/7 weeks after 135-hour ventilation in A/C+VG mode.

Wide interalveolar septa, capillaries are located in the thickness of interalveolar septa, thickened vascular walls (a), pneumocytes order I and II do not have clear morphological differences (no light-optical signs of differentiation into pneumocytes type I and II), no hyaline membranes (b). a — right lung, middle lobe, magnification ×50; b — left lung, lower lobe, ×200 magnification; c — right lung, magnification ×400; d — lung areas typical for covering in pseudoglandular type of structure; no vascularization of mesenchyme, magnification ×400. Hematoxylin and eosin staining.

In most standard visual fields, lung tissue is represented by the structures typical for canalicular stage. However, there are visual fields with histoarchitectonics of pseudoglandular stage (depletion or absence of mesenchyme vascularization, respiratory tract is represented by cuboidal epithelium) (Fig. 5D). These patterns indicate an uneven rate of lung maturation. Thus, the expected maturity of lungs in accordance with obstetric gestational age is far from always the same at birth. Lung immaturity at birth may be more severe than expected that will affect the results of respiratory therapy and dictate the need for selection of individual ventilation parameters. There were no signs of ventilator-induced lung injury in this child.

Immediate outcomes in study children are summarized in Table 3.

Table 3. Outcomes in patients with birth weight less than 750 grams undergoing volume guaranteed ventilation at birth

Note. * IVH — intraventricular hemorrhages were assessed according to L. Papile scale (1978), where grade III — blood clot fills the lateral ventricle and results its dilatation; **Mean ± SD / Me (range).

There were 8 newborns with gestational age of less than 24 weeks at birth. Two of these children (both died) had gestation period less than 22 weeks, 6 patients — 22 — 24 weeks. Considering severe immaturity and body weight less than 750 grams, neonatal mortality rate was 54.5%. Early neonatal sepsis occurred in the majority of cases (10 out of 11 children). This process probably contributed to IVH grade III almost in every third child in addition to severe immaturity.

We used Spearman's rank correlation analysis to identify the relationship between ventilation parameters and clinical data. We were most interested in relationship between the maximum peak pressure at the 5^th minute of life, PIP changes, other ventilation parameters and clinical data. High PIP at the 5^th minute was not associated with mortality, IVH grade III and air leakage syndrome. At the same time, higher peak pressure at the 5^th minute was accompanied by shorter time constant (TC) at the 120^th minute (r = –0.852, p = 0.001). The need for a higher peak inspiratory pressure at the 5^th minute was combined with higher Raw at the 30^th minute of life (r = 0.629, p = 0.038). Newborns with early neonatal sepsis required higher peak pressure at the 5^th minute of life (r = 0.659, p = 0.027).

Discussion

The need for this study was dictated by few reports on volume guarantee ventilation in children with ELBW (especially those with a body weight less than 750 grams) in the delivery room. We have analyzed the PubMed Central (PMC) database for more than 40 years and found only few studies. The authors describe respiratory changes during A/C + VG ventilation (Dräger Babylog 8000) in less premature newborns after introduction of surfactant [23]. Scopesi F. et al. (2007) compared ventilation with and without guaranteed volume in conditions of “withdrawal” from respiratory support in 10 preterm infants with gestational age less than 32 weeks. The authors confirmed that the algorithm of traditional volume guarantee ventilation is highly effective in premature infants [24].

This study demonstrated quite unexpected results of volume guarantee ventilation in the delivery room. Indeed, respiratory parameters are characterized by certain variability within 120 minutes of life in children with a birth weight less than 750 grams despite some common trends. Thus, these changes reflect extremely individual characteristics of lungs in each case. In the context of individual approach, inspiratory pressure deserves a special attention. In addition to a wide range, this value may be very high (up to 79 cmH₂O). Routine clinical practice in newborns with ELBW in the delivery room does not imply such respiratory parameters worldwide as prescribed in clinical guidelines. PIP 20—40 cmH₂O is recommended as a rule [12, 25]. Considering our own data and these recommendations, we can assume that some children will not receive the proper tidal volume since they can require a peak pressure over 40 cmH₂O. On the other hand, some patients will receive excess tidal volume because they would have enough 20 cmH₂O.

In this context, the concept of prolonged inspiration with a unified algorithm in all children with ELBW in the delivery room [26] seems questionable. This approach may be traumatic, at least for the lungs, and not entirely effective, since the technology assumes strict respiratory parameters, for example, inspiratory pressure and time. At the same time, volume guarantee ventilation early after birth in children with ELBW ensures selection of individual ventilation parameters and minimizes the risk of complications. Moreover, the need for a high PIP in the first minutes of life to achieve targeted tidal volume is nothing more than lung recruitment. This process occurred before administration of surfactant in our study. Perhaps, this aspect explains no significant changes in peak pressure after administration of surfactant, since maximum lung recruitment had already occurred earlier. Probably, earlier administration of surfactant (within 10 min of life), especially before the first ventilator-induced inspiration [27], may be valuable to avoid extremely high PIP for achievement of targeted tidal volume. However, this hypothesis requires new and well-designed study.

Conclusion

Thus, volume guarantee ventilation from the first minutes after birth of children weighing less than 750 grams is a safe and effective strategy for respiratory support. This strategy can be recommended for initial ventilation in the delivery room since it ensures an individual approach. A prospective randomized controlled trial with a large sample size is required.

Ethics

Ethical approval was not required as the research is descriptive. No additional interventions were performed, except for those described in federal and local protocols for the treatment of premature infants with extremely low birth weight.

Acknowledgments

The authors would like to thank the obstetrician-gynecologist Nikolai Yurievich Karpov (Yaroslavl) for the collaboration in the statistical analysis, as well as pathologist Zhakota Dmitry Anatolyevich (Moscow) for consultative assistance.

Author contribution

Concept and design of the study — A.V. Mostovoi, A.L. Karpova

Collection and analysis of data — A.V. Mostovoi, A.L. Karpova, R.A. Burenkov

Statistical analysis — A.V. Mostovoi, A.L. Karpova

Writing the text — A.V. Mostovoi, A.L. Karpova, N.N. Volodin

Editing — N.N. Volodin

The authors declare no conflicts of interest.