Vapotherm & The HIPSTER Trial: Review of Recent Clinical Literature
Published by Vapotherm®
In September of 2016, the report of the HIPSTER trial came to press in the New England Journal of Medicine.[1] The HIPSTER trial was a prospective, multicenter randomized controlled trial evaluating the outcomes for High Flow Nasal Cannula (HFNC) compared to nasal continuous positive airway pressure (nCPAP) when used as a primary means of respiratory support in neonates.
At the time when this study was conducted, it was common to conflate Hi-VNI Technology with HFNC which is why a minority of the products included in this trial under the umbrella of HFNC were Vapotherm devices delivering Hi-VNI Technology (only 6 of the 278 patients (2.2%) in the HFNC group were treated with Hi-VNI Technology). This conflation is understandable as Hi-VNI Technology is only recently coming to be recognized as Mask-Free NIV™ for spontaneously breathing patients, and thereby seen as a separate category from HFNC both in terms of indications for use and FDA classification.
In this article, we will go over the HIPSTER trial and additional literature that shows how the HIPSTER trial is not reflective of Hi-VNI Technology’s clinical outcomes. As none of these specific studies differentiate between Hi-VNI Technology and commodity high flow oxygen products, we will indicate which studies were exclusively conducted with Hi-VNI Technology.
The HIPSTER trial represented a next step to prior work from this author group, wherein their previous work reported non-inferiority between HFNC and nCPAP when used as post-extubation support for infants who required immediate intubation. In this current trial, patients were enrolled who were at least 28 weeks gestation (block randomized between infants < 32 weeks and infants ≥ 32 weeks), less than 24 hours old and had not previously received endotracheal intubation. The primary outcome was treatment failure within 72 hours based on pre-specified criteria, and patients who failed HFNC could be “rescued” with nCPAP. A secondary endpoint was intubation rate. The authors mitigated bias to consider the experimental arm (HFNC) a failure by excluding patients < 28 weeks gestation, where the greatest failures were seen in their previous trial of post-extubation use, and defining very specific criteria for failure.
According to the publication, the HIPSTER trial was halted by recommendation of the data safety monitoring committee based on a highly significant difference in failure rates of the assigned arms (p < 0.001), despite only a trend of differences in intubation rates. Only 583 of the intended 750 infants were enrolled. Failure rates for HFNC and nCPAP were 25.5% (71 or 278) and 13.3% (38 of 286), respectively; intubation rates were 15.5% (43 of 278) and 11.5% (33 of 286), respectively.
With this negative finding for the HIPSTER trial, Vapotherm wants to be clear regarding several specific points related to this trial and contrasting this trial against the other major literature that exists in the neonatal patient population comparing both Hi-VNI Technology and other devices to nCPAP.
It is important to reiterate that the HIPSTER trial was not a trial of Vapotherm Hi-VNI Technology. Although the paper states that both the Fisher & Paykel Optiflow Jr and the Vapotherm Precision Flow were used, only 6 of the 278 patients (2.2%) in the HFNC group were treated with Vapotherm (per the online supplemental materials). Hence, the HIPSTER trial was in effect a study of Fisher & Paykel Optiflow Jr without the power to draw conclusions related to Vapotherm.
Vapotherm’s Precision Flow® system and the Fisher & Paykel Optiflow Jr are not comparable devices. Vapotherm contends, with the support of clinical evidence [2-8] Hi-VNI® Technology has a more pronounced clinical impact.
For a trial of Hi-VNI Technology, we should look to:
The Lavizzari trial demonstrates the effectiveness of Hi-VNI Technology in a matched model. [2] The study by Lavizzari and colleagues came to press weeks before the HIPSTER trial, and using a substantially similar study design, demonstrated non-inferiority of Hi-VNI Technology to nCPAP and bi-level nCPAP in 316 infants. Although the authors refer to the technology as “heated, humidified high-flow” this trial was conducted exclusively with Hi-VNI Technology.
Lavizzari’s primary endpoint was intubation; unlike the HIPSTER trial there was no established protocol to rescue Hi-VNI Technology patients with nCPAP. Failure rates for Hi-VNI Technology and nCPAP/bi-level were 10.8% (17 or 158) and 9.5% (15 of 158), respectively. Like the HISPTER trial, enrollment criteria called for infants at least 28 weeks gestation who required non-invasive therapy as initial support for respiratory distress and the acceptable margin for non-inferiority was set to 10%.
Consider the following points of comparison between the Lavizzari and HIPSTER studies:
- HFNC and Hi-VNI Technology were applied similarly and appropriately in both studies. Lavizzari started flows at 4-6 L/min and escalated to a maximum of 6 L/min. HIPSTER started flows of 6-8 L/min and escalated to a maximum of 8 L/min.
- CPAP control group was essentially applied similarly in both studies. Lavizzari, et al started at 4-6 cmH2O and escalated to a maximum of 6 cmH2O. Roberts, et al started at 6-8 cmH2O and escalated to a maximum of 8cmH2O. However, Lavizzari used bi-level nCPAP when necessary for apnea in the nCPAP group. Hence, the comparator was more extreme in the LAVAIZARRI study, which may explain the slightly lower nCPAP group intubation rate in this study: 9.5% compared to 13.3% for HIPSTER.
- There are no appreciable differences in the patient population between the Lavizzari and HIPSTER trials. The patient populations were similar in gestational age (GA) (means of 33 vs 32 weeks, respectively, with SD of approximately 2 weeks), Apgar score (range of 8-9), PaCO2 before enrollment (mean 58 vs 55 mmHg, respectively), and FiO2 before enrollment (0.24 vs 0.21 respectively). Average birth weight in the Lavizzari cohort was approximately 1950 g vs 1750 g in HIPSTER, however, the standard deviations were in the 500-600g range, at least 2.5-fold the difference in means.
- The two studies used very similar criteria for HFNC and Hi-VNI Technology treatment failure, which included FiO2 requirement of > 0.40, pH of 7.2 or less, PaCO2 greater than 60mmHg (>70mmHg for Lavizzari), episodes of apnea, or urgent need for MV as determined by the clinician. If the patient met failure criteria, Lavizzari protocol was to escalate to intubation while Roberts protocol was to first attempt to rescue with CPAP (if failed CPAP, Roberts protocol was then to intubate).
- Both studies applied an intention to treat model that included the discretionary use of surfactant. In each study, surfactant use was in less than half of the infants and not different between groups.
A matrix of the substantial trials in neonatology demonstrates this consistent trend for Hi-VNI Technology’s efficacy compared to the Fisher & Paykel system. The below study matrix and discussion points reinforce the superiority of Hi-VNI® Technology in similar clinical scenarios.
The above figure is a matrix of the major neonatal prospective, randomized clinical trials for HFNC. The columns divide the trials into Vapotherm Hi-VNI Technology vs Fisher & Paykel, while the rows divide the trials between post-extubation support, initial support (Lavizzari and HIPSTER) and a trial of Vapotherm compared to Nasal Intermittent Mechanical Ventilation (NIMV). Consider the following points:
- The Collins and Manley studies looked at post-extubation support. The study cohorts and protocol parameters were matched. Collins studied Hi-VNI Technology and Manley studied Fisher and Paykel. In the Collins trial, the failure rate for Hi-VNI Technology was non-significant and the lesser mean value compared to nCPAP clearly represents non-inferiority. In the Manley trial, HFNC met the preset criteria for non-inferiority, but the study drew criticism that the lesser absolute performance of HFNC was clinically meaningful.
- The Lavizzari and HISPTER trials match up as discussed above. Lavizzari studied Hi-VNI Techology. HIPSTER was a study of Fisher & Paykel.
- The Kugelman trial took Hi-VNI Technology to the next level and demonstrated equivalence as primary support compared to nasal ventilation.[8]
Conclusion
Taken together, there is a consistency in the clinical literature demonstrating Hi-VNI Technology to be at least equivalent to other non-invasive ventilatory support modalities for the symptoms of infant respiratory distress syndrome. On the contrary, Fisher & Paykel failed to meet non-inferiority criteria for initial support and arguably shows marginal non-inferiority as support following extubation.
REFERENCES
[1] Roberts et al. Nasal High-Flow Therapy for Primary Respiratory Support in Preterm Infants. 2016. N Engl J Med; 375(12): 1142-1151.
[2] Lavizzari A, Colnaghi M, Ciuffini F, Veneroni C, Musumeci S, et al. (2016) Heated, Humidified High-Flow Nasal Cannula vs Nasal Continuous Positive Airway Pressure for Respiratory Distress Syndrome of Prematurity: A Randomized Clinical Noninferiority Trial. JAMA Pediatr
[3] Collins CL, Holberton JR, Barfield C, Davis PG (2013) A randomized controlled trial to compare heated humidified high-flow nasal cannulae with nasal continuous positive airway pressure postextubation in premature infants. J Pediatr 162(5): 949-954 e941.
[4] McQueen M, Rojas J, Sun SC, Tero R, Ives K, et al. (2014) Safety and Long Term Outcomes with High Flow Nasal Cannula Therapy in Neonatology: A Large Retrospective Cohort Study. Pulm Respir Med 4(6): 1000216.
[5] Miller SM, Dowd SA (2010) High-flow nasal cannula and extubation success in the premature infant: a comparison of two modalities. J Perinatol 30(12): 805-808.
[6] Yoder BA, Stoddard RA, Li M, King J, Dirnberger DR, et al. (2013) Heated, humidified high-flow nasal cannula versus nasal CPAP for respiratory support in neonates. Pediatrics 131(5): e1482-1490.
[7] Frizzola M, Miller TL, Rodriguez ME, Zhu Y, Rojas J, et al. (2011) High-flow nasal cannula: impact on oxygenation and ventilation in an acute lung injury model. Pediatr Pulmonol 46(1): 67-74.
[8] Kugelman A, Riskin A, Said W, Shoris I, Mor F, et al. (2015) A randomized pilot study comparing heated humidified high-flow nasal cannulae with NIPPV for RDS. Pediatr Pulmonol 50(6): 576-583.