Heliox Adjunct Therapy for Neonates With Congenital Diaphragmatic Hernia

Discussion

In this cohort review, we found that the adjunctive use of heliox corresponded with improved gas exchange and reduced ventilator support in patients with congenital diaphragmatic hernia. Effects were maintained at 24 h of heliox administration. The decrease in ventilator support over the first 24 h seemed to be tolerated as evidenced by the stable trend of P_aCO2 seen at 12 and 24 h. This supports a positive effect of heliox while on ventilation in our cohort.

We chose to focus on the initial 24 h of exposure given the acute effect of heliox and the limitations of a retrospective analysis. We hypothesized that this interval would allow enough time to demonstrate the clinical effects of heliox, while decreasing exposure to possible confounding variables that would increase over longer study periods. The acute effect of heliox on gas exchange is demonstrated by the significant change in P_aCO2 shown within 1–2 h after exposure (this was the first arterial blood gas drawn after initiation). This was the most dramatic change in P_aCO2 during the study time period and could only be explained by the addition of heliox because the ventilator settings were largely unchanged during this short time interval. This acute improvement in ventilation was not transitory but rather sustained over the 24-h period. The stable trend of P_aCO2 seen at the 12- and 24-h intervals was continued, even while ventilator support was decreasing. Although our analysis was limited to 24 h, it is our experience that heliox continues to improve airway dynamics beyond 24 h of use. Theoretically, there is no reason why this effect would be limited to 24 h.

The decrease in F_IO2 over this study period was small but statistically significant. This amount of change may have uncertain clinical importance; however, it is important to note that the improvement occurred over a relatively short period and despite concurrent reductions in ventilator settings. This reduced oxygen requirement occurred without an increase in mean airway pressure. This effect is likely multifactorial. Improved laminar flow leads to better gas exchange in the distal airways, whereas lower P_aCO2 and increased pH allow for vasodilatation, which improves pulmonary hypertension.

Although consensus from the literature indicates a minimum effective dose of 50–60% helium, there may be a larger therapeutic window for heliox.²⁹ Theoretically, there is no minimum concentration at which heliox should not provide some advantage because helium is lighter than nitrogen.¹³ Case reports¹² and a small proof-of-concept study demonstrated clinical effect with concentrations down to 20–25% helium.²⁹ In our study group, there was a wide range of helium concentrations used at initiation of treatment: 10–70%. Of note, no patients had to be taken off heliox due to acute hypoxia after initiation. Heliox concentrations could not be analyzed separately due to our study size and heterogeneity. However, helium levels as low as 10% were still associated with a progressive improvement in ventilation. Thus, heliox is a potentially viable treatment, even for patients with congenital diaphragmatic hernia who have a very high oxygen requirement.

The heliox group was biased to include a greater amount of high risk subjects, which is evident in the severity of pulmonary hypertension by echocardiography, inhaled nitric oxide use, age at repair, and defect type. It is our experience that this population often has issues of hypercarbia, independent of oxygenation. Heliox is typically initiated in patients who are not responding to standard treatment or who are deteriorating from a ventilation standpoint. The mean starting F_IO2 of 0.5 and the mean starting P_aCO2 of 68 mm Hg support this. Although the clinical severity differed between the group that received heliox and the group that did not, the requirement of extracorporeal membrane oxygenation and mortality rates were similar. In addition, the heliox group had a wider range in time of death. Due to the retrospective design of this study, these findings were limited to interpretation but were, at the very least, encouraging trends for this more severely affected congenital diaphragmatic hernia population.

Given the limitations of a retrospective cohort review with multiple confounding variables, including varied start times (ie, pre-, peri-, and postoperative) and F_IO2 requirements, we were not able to directly compare the heliox and non-heliox groups. We believed that the most rigorous way to assess the effect of heliox in this historical cohort was to use each individual as his or her own control. Thus, each episode occurred with the same confounding variables, which limited interpersonal variation. Most of the heliox episodes occurred in the first days of life or near the congenital diaphragmatic hernia repair, as expected, given that the respiratory status of these babies is quite tenuous during that time frame. Eight episodes occurred more than 14 d after congenital diaphragmatic hernia repair. These corresponded with acute deteriorations in respiratory status related primarily to infection or subsequent surgeries. Although these later exposures were not acutely related to the congenital diaphragmatic hernia itself, the underlying pathophysiology of the lung remained the same.

The initiation of heliox was clinically determined and varied in timing and duration. Therefore, it is unlikely that the changes in blood gas and ventilator data were simply due to natural history. The short study period and analysis of each individual exposure also diminished the likelihood that these results could be explained by any other intervention or change in clinical care. Concurrent therapies were administered regardless of heliox administration, including inhaled nitric oxide, steroids, and vasoactive medications.

Patients with congenital diaphragmatic hernia have multiple long-term issues, including chronic lung disease, feeding difficulty, and neurodevelopmental concerns. Recent data reconfirmed a connection between pulmonary status and developmental delay.^30,31 Therapies that minimize the degree and duration of ventilator support would be expected to decrease respiratory morbidity and improve outcomes, both during the initial hospitalization and after discharge.

Improved understanding of the unique pathophysiology present in the patient with congenital diaphragmatic hernia, coupled with strategies to prevent ventilator-induced lung injury, has led to dramatic changes in how we treat these patients. The adoption of these therapies has resulted in significant improvement in their survival. In building on this knowledge, heliox offers a novel strategy to improve flow dynamics and gas exchange through abnormal airways with peripheral obstruction, which allows for lower ventilator settings. Heliox may also improve oxygen delivery due to laminar flow dynamics and may enhance pulmonary vasodilatation by minimizing hypercarbia and normalizing pH. This effect could indirectly improve pulmonary hypertension without requiring higher ventilator settings or increasing oxygen exposure.

The limitations of this study included retrospective data collection and the lack of a formal protocol for heliox use, patient heterogeneity, variable timing and concentration of heliox therapy, and lack of a matched control group. Despite its limitations, this sort of retrospective analysis has been an important method for advancing care in this unique population.