Did Reduced Alveolar Delivery of Surfactant Contribute to Negative Results in Adults with Acute Respiratory Distress Syndrome?

To the Editor:

The efficacy of surfactant replacement therapy for pediatric and adult lung injuries has been an ongoing area of investigation for more than 25 years. Surfactant therapy is lifesaving in preterm infants and is also efficacious in term infants with meconium aspiration or pneumonia. However, despite a strong scientific rationale for surfactant therapy in acute respiratory distress syndrome (ARDS), it has not decreased mortality in controlled trials in adults (1, 2). Inhibitor-induced surfactant dysfunction is present in patients and animal models of ARDS and can be reversed effectively by exogenous surfactant supplementation in mechanistic biophysical and animal studies (1–3). Why then have controlled clinical trials of surfactant therapy in ARDS had less than stellar success? Although differences in surfactant drug activity and inhibition resistance, mechanical ventilation and treatment protocols, and etiology of lung injury have all contributed, this letter focuses on the major issue of impaired alveolar delivery of tracheally instilled surfactant based on recent fluid mechanical computational analyses (4). These sophisticated assessments indicate that alveolar delivery of instilled surfactant was likely suboptimal in recent controlled trials of surfactant therapy in ARDS from direct lung injury that found no decreases in patient mortality (5–7).

Although there are many thoughtful and comprehensive publications on the efficacy of surfactant therapy in adult, pediatric, and neonatal settings, few have dealt specifically with the role of instilled dose volume. It is typically just presumed that administered surfactant (mg/kg, the product of drug concentration and instilled dose volume/body weight) is distributed to the alveoli during mechanical ventilation via a well-mixed lung compartment. However, our recent modeling of the airway transport of intratracheal surfactant shows that instilled dose volume is a critical independent parameter (4). Any viscous fluid flowing over a surface leaves a coating film that increases in thickness as shear viscosity increases. Coating the conducting airway tree significantly reduces the amount of instilled surfactant arriving at the alveoli (i.e., the “coating cost”) (4). Instilled dose volumes must be well above the coating cost to deliver sufficient surfactant to the alveoli in a timely manner. Our model predicts that, in neonates, the coating cost is relatively small, and volumetric doses of 1–4 ml/kg all result in significant drug delivery to the alveolar site of action. This dose volume range coincides with what is used clinically in preterm infants (e.g., a standard 100 mg/kg surfactant dose requires 4 ml/kg of beractant [25 mg/ml], 3 ml/kg of calfactant [35 mg/ml], or 1.25 ml/kg of poractant [80 mg/ml]). In our model, the neonatal lung behaves very much like a well-mixed compartment (4).

Adult lungs and those of children, however, have substantially different airway distribution behavior, due to developmental changes in pulmonary anatomy. In particular, adult lungs have significantly more conducting airway generations (∼15 conducting airway generations in adults compared with ∼8 in neonates). This leads to a greatly increased conducting airway surface area for coating loss in adults compared with neonates. For illustrative purposes, calculations based on a simple Weibel lung model, with the trachea as generation n = 0 and 2ⁿ tubes at each bifurcating airway generation, show that an adult lung (tracheal diameter = 1.8 cm) has approximately 4,500 cm² of conducting airway surface compared with approximately 40 cm² in a neonatal lung (tracheal diameter = 0.4 cm). This greater than 100-fold increase makes adult lungs much more susceptible to losing instilled surfactant to the coating cost relative to neonatal lungs. An instilled surfactant volumetric dose of 1 ml/kg in adults leads to inadequate, including zero, calculated delivery efficiency (4) (Figure 1). Doses of 2–4 ml/kg in children and adults still must pay the coating cost but have enough volumetric surfactant material left for pharmacologically adequate delivery efficiency, as Figure 2 shows for 3 ml/kg. In addition to volumetric effects, delivery efficiency also decreases if the ventilator airflow rate impinging on the instilled surfactant increases (Figures 1 and 2) (4).

Figure 1.

Calculated exogenous surfactant profile for a low instilled volumetric dose of 1 ml/kg in the airway tree of a 70-kg adult based on the model in Reference 4. A 12-generation Weibel airway model is used for visualization, with surfactant levels at 4,096 airway branch point terminations, color coded based on the amount delivered to the termination of the 12th generation, which is near the start of the alveolar level. Dose volume (V_D) = 70 ml (1 ml/kg). The surfactant mixture is instilled in two aliquots, 35 ml with the patient in the right lateral decubitus (RLD) position, and 35 ml in the left lateral decubitus (LLD) position. Instilled surfactant aliquots enter the distal trachea as a plug forced by airflow. For a given airflow rate, distributional graphics are shown for one lung for simplicity, and the contralateral lung would be the mirror image (not shown). Right (R): 250 ml/s airflow rate, calculated delivery efficiency (η) = (total delivered volume 4.3 ml to airway terminations/dose volume 70 ml) × 100 = 6.2%, and calculated coating cost volume (V_CC) = 65.7 ml. Left (L): 500 ml/s airflow, calculated η = 0, V_CC = 70 ml, that is, zero alveolar delivery. Surfactant physical properties for simulations were those used in Reference 4. Comparative surfactant delivery and coating cost calculations for a higher V_D at the same airflow rates are shown in Figure 2.

Figure 2.

Calculated exogenous surfactant profile for a higher instilled volumetric dose (3 ml/kg) in the airway tree of a 70-kg adult based on the model in Reference 4. Parameters are the same as in Figure 1, except for the total instilled dose volume (V_D) = 210 ml (3 ml/kg) that was given in two aliquots of 105 ml in the right lateral decubitus (RLD) position and 105 ml in the left lateral decubitus (LLD) position. Right (R): 250 ml/s airflow rate, calculated delivery efficiency (η) = (116.3 ml/210 ml) × 100 = 55.4%, coating cost volume (V_CC) = 93.7 ml. Left (L): 500 ml/s airflow rate, η = (76.9 ml/210 ml) × 100 = 36.6%, V_CC = 133.1 ml. Compared with Figure 1, the 3-ml/kg dose provides much higher delivery efficiency than a 1-ml/kg dose because the dose volume substantially exceeds the coating cost, V_D $\gg$ V_CC. Calculations in both Figures 1 and 2 also show that the higher impinging airflow rate increases V_CC and reduces delivery efficiency (mechanistically, the surfactant coating layer in each airway generation becomes thicker as impinging airflow increases and deposits more material on the airway walls). Additional model calculations (data not shown) document that delivery efficiency also decreases if surfactant viscosity increases, because this similarly increases coating film thickness, and hence V_CC.

The above principles have direct ramifications for controlled trials of surfactant therapy in patients with ARDS. In an initial trial in 1999, Willson and colleagues (8) treated pediatric patients up to age 18 years with calfactant (35 mg/ml) at a dose of 80 ml/m². This translates to a volumetric dose of 2–3 ml/kg from the oldest to the youngest patients studied. Oxygenation was improved by calfactant, resulting in earlier extubation and decreased intensive care unit care. A subsequent randomized, controlled pediatric calfactant trial by Willson and colleagues (9) using the same dose protocol reported significantly decreased mortality in patients with direct ARDS. In both these successful studies (8, 9), calfactant administration was associated with rapid improvement in oxygenation. In contrast, two recent controlled trials with calfactant in pediatric (6) and adult patients (7) with direct ARDS disappointingly reported no clinical benefits, including no improvements in oxygenation. Because calfactant has documented alveolar activity in enhancing respiratory function, this implies that it was not present in the alveoli in pharmacologic concentrations. Importantly, in these unsuccessful trials (6, 7), calfactant was concentrated and given in a reduced volume dose of approximately 1.3 ml/kg to lower instilled fluid load in ventilated patients. Computational modeling indicates that this lower volumetric dose led to a reduced delivery efficiency of instilled surfactant (Figure 1). The higher surfactant concentration may also have increased viscosity at low shear rates, adversely affecting delivery. Two additional unsuccessful controlled trials of surfactant therapy in adults with ARDS by Spragg and colleagues (5, 10) also used concentrated low-volume doses (1 ml/kg) of instilled recombinant surfactant protein C–based surfactant (Venticute) that likely had reduced alveolar delivery based on our modeling. In addition, Spragg and colleagues (5) also reported surfactant stability issues that may have affected drug activity.

An important implication from our computational modeling (4) is that future trials of surfactant therapy in ARDS would benefit from animal models similarly sized to the patient and prospective in silico simulations, specifically to assess drug delivery and aid in detailed protocol development to enhance study replicability. Instilled surfactant volumes obviously cannot be overly large in patients with ARDS, but, within acceptable patient care limits, they must generate a pharmacologically adequate alveolar dose in clinical trials. Otherwise, intrinsic therapeutic efficacy is not really being tested.