Bronchopulmonary dysplasia (BPD) is a common neonatal respiratory disease affecting preterm neonates, particularly those who are extremely low birth weight (ELBW) (birth weight less than 1,000 g) (1–4). BPD is a significant cause of morbidity and mortality in this population, with prominent outcomes including cardiovascular and neurodevelopmental delay (5–7). Several factors contribute to BPD in these infants, including perinatal inflammation, sepsis, necrotizing enterocolitis (NEC), and the need for mechanical ventilation and oxygen supplementation (6, 7). These conditions collectively promote damage and inflammation in the alveoli and airways, resulting in abnormal lung development and BPD onset (4). Despite neonatal care advancements, BPD prevalence remains high, and previous therapies primarily focused on mitigating its effects rather than prevention or cure.

In recent years, studies have explored therapies aimed at reducing the BPD prevalence in preterm infants and improving their short-term and long-term outcomes (3–5). Among these therapies, postnatal steroids (PNS) have gained significant attention for their ability to reduce lung inflammation (8–14). PNS achieve this by suppressing lymphocytes and enhancing the effects of anti-inflammatory cytokines (9). PNS therapy is commonly used to treat and prevent BPD and to facilitate extubation from mechanical ventilation (12, 13). However, their use in neonates, especially preterm infants, is associated with potential adverse effects, including impaired growth and development, an increased risk of infection, and neurodevelopmental problems (2, 14–17, 18).

Studies have assessed the efficacy and safety of PNS, particularly dexamethasone, with various doses or durations in neonates (19–22). The most commonly studied protocol is the DART (Dexamethasone: A Randomized Trial) protocol, which involves a lower dose (total 0.89 mg/kg) over 10 days for the treatment of BPD in preterm infants (23). Despite its long-standing use, BPD remains a significant cause of neonatal morbidity. In recent years, clinicians have started to explore different dosing regimens and durations for dexamethasone to evaluate potential benefits (20, 21, 24).

The efficacy and safety of low-dose dexamethasone (0.08 mg/kg/day for seven days) compared to high dose (0.5 mg/kg/day for three days followed by 0.3 mg/kg/day for four days) was evaluated in a study involving preterm neonates with BPD (25). The study found that low-dose dexamethasone therapy was associated with lower incidence of adverse effects, including growth failure and hypertension, compared to conventional dexamethasone therapy with similar pulmonary outcomes. Another study used early dexamethasone therapy at a higher overall dose (0.5 mg/kg/day for seven days) vs. placebo (0.9% normal saline) (21). This study found that dexamethasone improved short-term pulmonary outcome compared to placebo, with no reported serious adverse effects in either group.

These findings underscore the variability in current research regarding the optimal duration and dose for dexamethasone. The American Academy of Pediatrics (AAP) state that doses ≥0.5 mg/kg are associated with adverse outcomes. However, insufficient data is available on the potential benefits and risks of lower doses of dexamethasone in neonates (26).

Overall, studies suggest that the optimal dexamethasone regimen is unclear. However, doses higher than those recommended in the low-dose DART protocol show promise in effectively reducing BPD without significant differences in adverse events. Neonatologists often favor the DART regimen due to the lack of established evidence for alternative dexamethasone dose regimens.

The objective of this study is to conduct a retrospective analysis to investigate the clinical efficacy of low-dose dexamethasone dose (DART protocol) compared to enhanced low-dose dexamethasone regimens in facilitating extubation, and to examine their effects on significant neonatal outcomes in ELBW infants with evolving or established BPD.

We evaluated 94 children who received dexamethasone treatment in the NICU for inclusion in the study. After excluding 17 infants, the study included 77 patients who received a total of 121 doses of dexamethasone (Figure 1). The characteristics of study subjects are summarized in Table 1. The study population was divided between those who received low-dose dexamethasone (n = 49) vs. enhanced low- dose dexamethasone (n = 28). The GA of neonates in the low-dose group was 24.9 ± 1.6 weeks vs. 25.3 ± 1.2 weeks in the enhanced low-dose group. The BW of neonates in both groups was respectively 748 ± 114 vs. 765 ± 124 g. Majority of patients were male in both groups (65% vs. 71%; p = 0.581). There were various ethnicities amongst the groups such as African (6.1% vs. 10.7%) and European (8.3% vs. 4.6%) without any visible differences. However, Asian population was much higher in the enhanced low-dose group 42.9% vs. low-dose 22.4%, while Middle Eastern ethnicity was higher in the low-dose group 63.3% vs. 42.9%. Difference in APGAR scores at 1 and 5 min showed no statistical significance between the groups. Discharge weight was significantly higher in the enhanced low-dose dexamethasone group compared to low-dose group (4,080 vs. 3,305 g; p = 0.048).

Maternal characteristics between the groups were further analyzed in Table 1 with a median age of mothers being 29.1 ± 6.6 in the low-dose group vs. 31.2 ± 7.2 enhanced low-dose group. Multiple gestations were much higher in the low-dose group 34.7% vs. 14.3% in the enhanced low-dose group, while not significant p = 0.053. Medical conditions during pregnancy were also analyzed including maternal hypertension (12.2% vs. 21.4% p = 0.311), gestational diabetes (22.5% vs. 21.4%; p = 0.999) and positive GBS screening (10.6% vs. 21.4%; p = 0.2), with no statistical significance between the groups. Other maternal factors examined were chorioamnionitis, PPROM, and oligohydramnios, did not show significant difference. The majority of mothers in both groups received at least one dose of antenatal steroids, with similar percentages in both groups, 91.7% vs. 96.4%. Additionally, nearly two-thirds of mothers received 2 doses of antenatal steroids in each group.

The respiratory characteristics of infants that received dexamethasone regimens were analyzed between the groups in Table 2. No statistical significance was found between the number of surfactant doses received whether none, one or ≥ two used between the two groups (p-value = 0.156). Receiving mechanical ventilation was assessed at 28 days between the low-dose and enhanced low-dose (83.7% vs. 92.8%; p = 0.312) and at 36 weeks PMA (22.5 vs. 22.2; p = 0.859), showed no statistical significance. No statistical significance was also observed between median invasive ventilation days (p = 0.787). However, the low-dose group received more non-invasive ventilation days compared to the enhanced low-dose group (40 vs. 60; p-value = 0.026). Additionally, infants who received dexamethasone therapy in the first 28 days were 37.4% vs. 25%; p = 0.377. Other respiratory characteristics analyzed were median FiO₂ at 28 days and at 36 weeks (40 vs. 57; p-value = 0.210) and (25 vs. 29; p-value = 0.08), respectively. The proportions of infants who received repeated courses of dexamethasone (38.8% vs. 39.3%; p-value = 0.965), showed no differences between the two groups.

Results of respiratory outcomes associated with different dexamethasone regimens are summarized in Table 3. The low-dose group were administered 75 courses vs. 46 in the enhanced low-dose group. The dosing weight for all courses was 1,696 ± 1,174 g in the low-dose group vs. 1,907 ± 1,280 g in the enhanced low-dose group p = 0.244. The dosing weight in grams at the start of the first dose was also evaluated and showed no significant differences between the groups p = 0.154. Day of life in terms of all courses and first course were analyzed and showed no significance, p = 0.571 and p = 0.603, respectively. Assessment of PMA in weeks at the start of all doses showed no significant difference between the groups (33.7 ± 6.7 vs. 35.3 ± 6.6; p = 0.112). However, PMA at the first course was found to be significant amongst the groups (30.8 ± 3.4 vs. 32.1 ± 2.5; p = 0.014). The median course dose per kg (mg) received was found to be higher in the enhanced low-dose group (0.73 vs. 1.27; p = <0.001). Also, evaluation of the median course duration (in days) between the groups was lower in the low-dose group (7 vs. 10. days p = <0.001). Statistical significance was observed in the course cumulative doses (mg) between the groups, with the enhanced low-dose group having a higher median dose (0.79 vs. 1.65; p = 0.001), and a higher total cumulative doses (mg) received throughout the NICU stay (1.00 vs. 1.99 (p < 0.001).

Other parameters assessed in Table 3 included the median baseline FiO₂ (%) before dexamethasone course and it was significant between the groups (45 vs. 63; p = 0.042). On the other hand, median FiO₂ (%) after dexamethasone course showed no significance (35 vs. 40; p = 0.527). The vast majority (>90%) of patients in each group were on invasive ventilation, and there was no difference between the groups (p = 0.124). The proportions of infants extubated at different time points from dexamethasone treatment initiation were assessed between the low-dose and enhanced low-dose groups. There were no significant differences between the groups in extubation rates within 72 h (25% vs. 22.2%; p = 0.735) or following seven days from initiating therapy (41.2% vs. 31.1%; p = 0.278). Further analysis of potential extubation either at the end of treatment course or seven days post treatment showed no statistical difference between the groups (45.6% vs. 53.3%; p = 0.420 and (54.4% vs. 64.4%; p = 0.389). Reduction in respiratory support of any kind was not significant between the groups, p = 0.316. Infants who received enhanced low-dose were extubated at a mean age of 63 vs. 49 days of life compared to infants who received a lower dose of dexamethasone; however, this finding was not statistically significant (p = 0.174). Both the low-dose and enhanced low-dose groups had a high but non-significant incidence of VAP during treatment (21.3% vs. 30.4%; p = 0.260) and sepsis (13.3% vs. 8.7%; p = 0.439), while the presence of NEC was low (1.3% vs. 0.0%; p = 0.999), respectively.

Secondary outcomes associated with dexamethasone regimen usage were also analyzed throughout the study and are summarized in Table 4. Several neonatal-specific factors were assessed between the low-dose and enhanced low-dose groups, including VAP (61% vs. 61%; p = 0.965), sepsis (45% vs. 50%; p = 0.666), NEC (18.4% vs. 7.1%; p = 0.310), any ROP (71.4% vs. 74%; p = 0.805), severe ROP (40.8% vs. 37%; p = 0.747), treated ROP (8.8% vs. 19%; p = 0.408), medically treated PDA (53% vs. 53.6%; p = 0.966), with only two infants in each group requiring device closure of the PDA, and moderate and severe BPD (85.4% vs. 100%; p = 0.074). Additional analysis of long-term outcomes such as NDI showed no significant difference between the groups (30.7% vs. 29.1%; p = 0.893). There were 13 (16.9%) death cases during the NICU hospitalizations (22.4% vs. 7.1%; p = 0.117), with a median day of death of 92 vs. 145 days, p = 0.194 in the low-dose and enhanced low-dose groups. Infants who died before and after 36-weeks PMA were four and seven in the low-dose group and one and one in the enhanced low-dose group. Additionally, the incidence of death or severe BPD was 30.6% vs. 25%; p = 0.6 between the two groups.

Univariate and stepwise multivariate regression analyses were conducted to identify significant variables associated with successful extubation at various time points: 72 h post dexamethasone treatment, within seven days of treatment initiation, by the end of dexamethasone treatment, and within seven days of dexamethasone treatment completion, as summarized in Table 5.

Among the demographic and clinical variables examined, the following predictors were found to be associated with successful extubation at 72 h post dexamethasone treatment initiation in the univariate regression model: baseline FiO₂, duration of invasive ventilation, and course duration. However, in the multivariate analysis, only baseline FiO₂ (adjusted odds ratio [aOR]: 0.951, 95% confidence interval [CI]: 0.920–0.984, p = 0.004) and course duration (aOR: 0.840, CI: 0.721–0.980, p = 0.026) remained significantly associated with successful extubation at this time point.

Similarly, for successful extubation at seven days post treatment initiation, the univariate regression model identified the following predictors: baseline FiO₂, duration of invasive ventilation, ethnicity, dose per kg, cumulative dose, and course duration. In the multivariate analysis, baseline FiO₂ (aOR: 0.966, CI: 0.941–0.992, p = 0.011), duration of invasive ventilation (aOR: 0.962, CI: 0.939–0.987, p = 0.003), cumulative dose (aOR: 1.864, CI: 1.270–2.736, p = 0.001), and course duration (aOR: 0.759, CI: 0.652–0.884, p < 0.001) remained significant.

Additionally, regarding extubation by the end of the dexamethasone course treatment, out of the significant predictors found in the univariate regression analysis (baseline FiO₂, duration of invasive ventilation, parity, ethnicity, dose per kg, cumulative dose, and course duration), only baseline FiO₂ (aOR:0.971, CI: 0.948–0.994, p = 0.016), dose per kg (aOR:11.71, CI: 3.070–44.691, p < 0.001), duration of invasive ventilation (aOR:0.977, CI: 0.961–0.993, p = 0.005), and course duration (aOR:0.852, CI: 0.736–0.987, p = 0.033) remained significant in the multivariate analysis.

Finally, the regression analysis within seven days of completing dexamethasone treatment showed a significant association in the univariate analysis for several factors, including baseline FiO₂, duration of invasive ventilation, parity, ethnicity, maternal age, dose per kg, gravidity, and cumulative dose. From these predictors, baseline FiO₂ (aOR:0.960, CI: 0.93–0.982, p < 0.0001), duration of invasive ventilation (aOR:0.974, CI: 0.95–0.989, p < 0.001), and dose per kg (aOR:7.361, CI: 2.42–22.37, p < 0.001) remained significant in the multivariate analysis.

BPD remains a challenging respiratory outcome among ELBW neonates in the NICU (1–4). The administration of PNS can be considered as a crucial therapy for these neonates to improve parenchymal lung disease and facilitate extubation (3–6, 8–10). Although DART protocol is the mainstay therapy in our unit, higher dose regimens such as enhanced low-dose dexamethasone, are currently used on a case-by-case basis by neonatologists. This practice aligns with the common approach of utilizing this dose of dexamethasone (DART protocol) for evolving or established BPD suggested by Doyle LW et al. (23). The same author conducted a meta-analysis of 23 RCTs on the use of systemic PNS for the prevention of BPD in preterm neonates (27). Their analysis revealed that two-thirds of the studies included in the meta-analysis utilized dexamethasone (n = 21) for BPD prevention.

Neonates in both groups showed similar neonatal-specific characteristics. However, it was observed that neonates receiving the enhanced low-dose dexamethasone regimen had a higher weight at the time of hospital discharge, required higher FiO₂, and remained longer on non-invasive respiratory support. Furthermore, they were extubated later compared to infants who received a lower dose of dexamethasone, although this latter finding was not statistically significant. These observations suggest a more severe manifestation of BPD. Thus, providing a potential explanation for the increased utilization of the enhanced low-dose dexamethasone regimen by attending neonatologists in this group. According to our study, the enhanced low-dose dexamethasone regimen resulted in a significantly longer duration of therapy and a higher cumulative dose compared to the low-dose dexamethasone protocol. This was expected as the enhanced low-dose regimen involves a higher overall dose and/or a longer duration of therapy. Bonnie L Marr et al. conducted a randomized controlled trial (RCT) in which they compared a 42-day course of dexamethasone to a nine-day course for the treatment of BPD in extremely preterm neonates (28). The study found that the 42-day group had a significantly higher overall dose of dexamethasone compared to the 9-day group (P < 0.001). These findings are expected with all studies that utilize a higher dexamethasone dose than the standard DART protocol dose.

Our study findings showed no difference between the groups in terms of development of any adverse outcomes including NDI at 18–24 months. These findings suggest that the safety profile of dexamethasone at an enhanced low-dose is comparable to that of the low-dose (DART protocol). A meta-analysis by Ramaswamy et al. assessed the safety of 14 different PNS regimens used for the prevention of BPD in neonates and secondary outcomes included NDI (18–24 months) (29). It was found that moderately early (2–4 mg/kg) of dexamethasone was potentially the most appropriate regimen for preventing BPD. Also, none of the interventions were associated with an increased or decreased risk of NDI. Linan Zeng et al., conducted a meta-analysis which included 47 RCTs with 6,747 participants and focused on the efficacy and safety of high dose dexamethasone (>3 mg/kg) compared to other corticosteroids for prevention of BPD in preterm infants (30). The meta-analyses concluded that higher-doses are effective, but not without the potential increased risk of adverse neurodevelopment effects. A systematic review conducted in 2017 by Onland W. et al., which compared different corticosteroid regimens on mortality and neurodevelopmental outcomes (31). In a subgroup analysis of the study, it was concluded that there were no differences between high and low-dose regimens on death or any abnormal neurodevelopmental outcomes. Other studies have suggested that the use of dexamethasone at various high doses could potentially be harmful for overall brain development (32, 33). In our study, enhanced low-dose dexamethasone, which is considered a lower cumulative dose compared to those referred to as “moderate to high” in the previously mentioned studies carries a considerably less risk for NDI at 18–24 months as suggested by the results. However, a longer-term follow-up beyond two years on NDI is warranted to further confirm the findings of our study.

Our study did not show that an enhanced low-dose dexamethasone is associated with a higher rate of extubation at various time points. Nonetheless, regression analysis showed that a higher dose per kilogram or higher cumulative dose with a shorter duration are associated with successful extubation. These findings are supported by recommendations by Zeng et al, who suggested aggressive initiation seems more effective (30). However, higher doses for longer durations are not recommended based on safety. Moreover, the regression analysis also showed that enhanced low dose dexamethasone dose significantly correlated with higher chances of neonatal extubation within different time frames of receiving dexamethasone course. This is an essential finding as it further sheds light on the potential importance of increasing dexamethasone dosing within ranges that are still considered to be safe (34). Studies have found similar findings where moderately high doses of dexamethasone were associated with a significant improvement in FiO₂ (30, 35, 36). Despite the higher extubation rates associated with the enhanced low dose in our study, no differences were found between the rates of BPD outcomes between the groups.

Similarly, death was not clinically significant although higher in the low dose dexamethasone group. These findings are supported by a meta-analysis and other studies which showed no difference of composite death between various dexamethasone regimens (26, 37, 38).

This study has several strengths as it is the first of its kind to be conducted in Qatar. It also provides valuable insight on the use of dexamethasone in ELBW neonates with BPD as it compares two different dose regimens low dose (DART) and enhanced low-dose. The study focused on a vulnerable and specific population, ELBW neonates, for whom the optimal recommended dose of dexamethasone is still in question. In addition, matching was conducted to help control for potential confounding factors by matching patients in terms of birth weight and gestational age. This enhances the validity of the comparison between the groups. Regression analysis was used to help identify any predictors of successful extubation. This statistical approach adds depth to the findings by highlighting factors that influence extubation outcomes in ELBW infants. The consideration of secondary outcomes, such as death or BPD, provides a comprehensive view of the safety and efficacy of the treatments. Despite its strengths this study contained limitations such as its retrospective nature which has inherent biases with data collection and dependence on existing medical records. Recall bias also may potentially impact the accuracy of the data recorded. Although the study attempted to control for potential confounding factors however, some variables which were not accounted for could potentially influence the extubation outcomes. In addition, the study may be underpowered, due to the small sample size, to fully form conclusions on safety and efficacy outcomes. The need for further research to evaluate the findings of this study with larger sample sizes and longer follow-up data are warranted to help draw more clear conclusions.

Our findings showed no significant differences in extubation success rates between the two regimens at various time points following treatment initiation. However, we identified important predictors of successful extubation, including baseline FiO₂, course duration, and duration of invasive mechanical ventilation, which negatively impacted extubation success. Conversely, received dose per kilogram and cumulative dose positively correlated with successful extubation at different time points. These results underscore the need for a personalized approach to dexamethasone therapy in ELBW infants with BPD, considering individual patient characteristics and clinical factors.