September 2016 | JAMES SCHLETT | Photonics Media —
In 1985, a team of researchers at Duke University Medical Center in Durham, N.C., pointed out the incongruity that, while the goal of intensive care for preterm newborns is to ensure their brains were receiving sufficient oxygen, neonatal intensive care units (NICUs) lacked a direct, noninvasive way to measure cerebral oxygenation. Instead, NICU staff had to rely on indirect methods, such as monitoring blood-oxygen content and transcutaneous oxygen at sites far from the brain, usually the chest or foot. That compelled the researchers to combine near-infrared (NIR) light and spectroscopy as a way to achieve this goal4.
In the three-plus decades since then, the need to measure neonate cerebral oxygenation has become no less urgent. Perinatal hypoxia, a deficiency of oxygen in tissue shortly before or after birth, can result in long-term neurological disorders such as cerebral palsy and mental retardation2. A 2005 study of infant mortality at an Egyptian NICU found that hypoxic-ischemic encephalopathy, also known as perinatal asphyxia, accounted for nearly one in five deaths (18.8 percent) during an 18-month period. Only deaths due to infections (56.25 percent) and respiratory distress syndrome (26.7 percent) had higher mortality rates3. Yet cerebral NIR spectroscopy (NIRS) monitors have not risen to the standard-of-care level in these settings, in part, because of the need for more accurate measurements and for a better understanding of how to respond to them (Figure 1).
Figure 1. The INVOS near-infrared spectroscopy cerebral oximetry system by Medtronic PLC gives physicians a first alert to critical changes in an infant’s brain oxygenation and end organ perfusion. Courtesy of Medtronic.
“I don’t think NIRS cerebral monitoring is standard of care in the NICU as in ‘neonatology,’ certainly not in the U.S. For our institution and many academic institutions, it’s been used as [a] research tool or [in] limited routine neonatal applications,” said Dr. Steve M. Liao, an assistant professor of pediatrics at Washington University School of Medicine in St. Louis.
However, researchers in Europe are pursuing large randomized clinical trials that promise to help NICUs better understand — and implement treatments based on — the measurements of cerebral NIRS monitors. At the same time, optoacoustic spectroscopy is emerging as a possibly more accurate alternative to NIRS in neonatal cerebral oximetry.
Cerebral NIRS oximeters’ ability to monitor infants’ cerebral oxygenation revolves around the Beer-Lambert law, which holds that the amount of light a substance, such as oxygen, absorbs can reveal how much of it is present. Consequently, the more oxygen that is present in tissue, the lower the intensity of the light transmitted through it4.
NIR light is easily able to transmit through skin, soft tissue and bone up to a depth of 8 cm in the brain5. An oximeter probe placed on the infant’s head emits the NIR light, which arcs through the brain (Figure 2). As it passes through the brain and is absorbed by the chromophores oxyhemoglobin, deoxyhemoglobin and cytochrome oxidase, the NIR light intensity changes. Detectors on the same sensor detect this change in intensity. The percentage of oxygen in the pathway of the light source can then be determined by using proprietary formulas that differentiate the chromophores6.
Figure 2. Medtronic’s disposable OxyAlert NIRSensor, together with INVOS technology, makes the monitoring of ischemic threats to infants safe and easy. Courtesy of Medtronic.
“There is a bit of misunderstanding about what exactly these devices are measuring. They are not measuring purely arterial blood nor purely venous blood. For this reason, using either venous or arterial samples as reference standards is not appropriate. These devices measure a unique space in the vascular system — the capillary bed of the tissue beneath the sensor — which is mostly venous blood but contains a portion of arterial blood,” said Doug Bartlett, vice president of global marketing for patient monitoring for Dublin-based Medtronic PLC.
Commercial cerebral oximeters
It was not until 2008 that the U.S. Food and Drug Administration (FDA) cleared a cerebral NIRS oximeter — the FORE-SIGHT by CAS Medical Systems Inc. in Branford, Conn. — for use in neonatal and infant populations. Other cerebral NIRS oximeters with FDA clearance for use in NICUs include the Equanox by Plymouth, Minn.-based Nonin Medical Inc., the INVOS 5100C by Medtronic and the NIRO 200NX by Hamamatsu Corp. in Hamamatsu City, Japan.
These cerebral oximeters vary in terms of, among other things, their algorithms, NIR light emission sources and wavelength. While the INVOS 5100C, for example, emits NIR light at two wavelengths (730 and 810 nm), the Equanox 7600 emits NIR light at three wavelengths (730, 810 and 880 nm) and the FORE-SIGHT emits it at four (from 670, 780, 805 and 850 nm)7.
“The manufacturers of cerebral oximetry devices use an indirect reference called field saturation, which is a reference derived from a venous measurement … and an arterial sample. And each manufacturer uses a slightly different calculation of this reference. These factors make it impossible to reliably or accurately compare technology performance ‘in the laboratory setting.’ Performance under true clinical conditions is much more relevant,” said Bartlett.
Too much variation
On top of these subtle differences between these devices, there are also distinctions between cerebral oxygenation saturation measurement from NIRS oximeters and from blood samples obtained via gold standard, invasive measurement methods. These include jugular venous oximetry and brain tissue oxygen tension, both of which involve catheters inserted into the brain. For example, a 2011 study by Dr. Philip E. Bickler of the Hypoxia Research Laboratory at the University of California, San Francisco, comparing these four FDA-cleared devices found their standard deviation from an invasive reference ranged from 3.92 percent for the FORE-SIGHT and 9.72 percent for the INVOS 5100C.
“The problem with NIRS-based cerebral oximetry in neonates is the poor precision of the current technology. We know that the estimate of regional oxygenation varies dramatically with small changes in sensor position. … Sick neonates are often born in a state of hemodynamic instability and [it] is thus not possible to determine safe values with certainty,” said Dr. Simon Hyttel-Sørensen, of the Department of Neonatology at the National University Hospital (Rigshospitalet) in Copenhagen, Denmark.
On top of the uncertainty resulting from this inconsistency of cerebral NIRS oximeter measurements, there is also the lack of clear guidance on how to respond to them. “Oxygen is not always the best fix for this population,” said Heather Elser, a neonatal nurse practitioner in Raleigh, N.C. She noted that too much oxygen has a toxic effect on the eye development of infants (see sidebar at end of article).
At St. Louis Children’s Hospital, Liao said infants with suspected or confirmed congenital heart defects get continuous NIRS monitoring. He said that if NIRS values fall outside of the proposed normative ranges, it should alarm the primary care team to evaluate the infant for underlying etiology for disturbed cerebral oxygenation/metabolism. And at the University of Florida Health Jacksonville and Baptist Medical Center, both of which are in Jacksonville, Fla., Dr. Josef Cortez said infants’ cerebral oxygenation is monitored when they are undergoing therapeutic hypothermia or are on extracorporeal membrane oxygenation. No medical decisions are based on cerebral NIRS oximeter measurements alone. However, Bartlett noted that INVOS monitoring is a standard monitoring parameter used pre-operatively and post-operatively in the management of infants with congenital heart disease.
“There are ongoing studies and treatment algorithms being proposed that are promising but would need further validation before hitting primetime in the NICU,” said Cortez, who is an assistant professor in the pediatrics department at the University of Florida College of Medicine.
There are no official treatment guidelines for how physicians should respond to NIR spectroscopy readings. Even more, there is insufficient evidence showing how NIRS-based cerebral oximetry impacts the clinical management of preterm infants. In a 2011 paper in the Philosophical Transactions of the Royal Society, Dr. Gorm Greisen, a professor and chairman of pediatrics at the University of Copenhagen, posited that, without a stronger foundation of clinical evidence, cerebral NIRS oximeters “may become another randomly applied expensive technology”8.
As the principal investigator of the multicenter Safeguarding the Brains of Our Smallest Children (SafeBoosC) project based in Denmark, Greisen was ideally positioned to deliver such evidence and make NIRS-based cerebral oximetry a more routine clinical practice in NICUs. Greisen and 18 other researchers from a dozen European medical centers last December published the results of a two-year randomized clinical trial in the Archives of Disease in Childhood9. The trial, dubbed SafeBoosC-II, employed INVOS oximeters and explored whether cerebral oxygenation could be stabilized through the use of NIRS-derived regional tissue oxygen saturation of hemoglobin (rStO2) monitoring when coupled with evidence-based treatment guidelines.
For this trial, the consortium of medical centers enrolled 86 premature infants within the first 72 hours of birth. During the study, 67 of those infants experienced 1,107 alarm episodes of cerebral hypoxia or hyperoxia, with a quarter of them — or an average of four per infant — triggering treatment guideline interventions. A related study published in The BMJ in January 2015 showed that through the combination of NIRS monitoring and the treatment guidelines, the amount of time the infants spent in hypoxia during the first 72 hours after birth was reduced by 58 percent. The normal range for rStO2 was 55 percent to 85 percent10.
The SafeBoosC-II findings laid the foundation for an even larger randomized clinical trial. The consortium last year applied to the European Commission for Horizon 2020 funding for SafeBoosC-III, which would involve 93 hospitals in 17 countries and about 3,800 premature infants. However, Hyttel-Sørensen said the bid for that €6 million in funding failed. Initially, Greisen considered “trying to reshape the protocol to a cluster randomized design,” but he said that Eugene Dempsey, a consultant neonatologist at Cork University Maternity Hospital in Cork, Ireland, instead submitted an application for funding for a randomized study involving 1,200 neonates in Ireland. Dempsey and Greisen worked together on the SafeBoosC-II study.
Bartlett at Medtronic said, “It’s quite surprising to realize that today no such guidelines exist for our smallest and most vulnerable patients. It’s absolutely critical that guidelines be put into place to guide practice for neonates, so I’m heartened to see that SafeBoosC is pushing to make this a reality. It’s sadly very overdue.”
If cerebral NIRS oximeters fail to prove their clinical value in NICUs through initiatives such as the SafeBoosC trials, an alternative technology based on optoacoustic spectroscopy may be positioned to meet that need.
“All NIR spectroscopic techniques rely on returning or transmitted light; therefore, they have limited ability to separate the signal derived from venous saturation, which represents tissue oxygen uptake, and arterial saturation, which represents a component of oxygen supply. Although NIRS techniques have provided vital information on neonatal cerebral circulation, they have not yet proven useful for routine clinical monitoring,” said Graham Randall, CEO of Noninvasix Inc. in Galveston, Texas.
Figure 3. Noninvasix Inc. has developed an optoacoustic monitor for measuring the amount of oxygen in preterm infants’ brains. Courtesy of Nonivasix.
Noninvasix last February filed a presubmission for a Food and Drug Administration (FDA) 510(k) marketing clearance for an optoacoustic monitor for measuring the amount of oxygen in preterm infants’ brains (Figure 3). Once approved, the device will compete with cerebral NIRS oximeters for bedside spaces in NICUs. Randall said a key advantage the optoacoustic monitor may have over NIRS monitors is a stronger correlation with cerebral oxygenation measurements taken with invasive oximetry techniques.
Figure 4. Noninvasix’s monitoring system involves a head strap with an optoacoustic probe consisting of a compact nanosecond Nd:YAG laser that sits above the superior sagittal sinus (SSS), the largest dural venous sinus that runs down the middle of the brain. Courtesy of Nonivasix.
Noninvasix’s monitoring system involves a head strap with an optoacoustic probe consisting of a compact nanosecond laser diode that sits above the superior sagittal sinus (SSS), the largest dural venous sinus that runs down the middle of the brain (Figure 4). Taking advantage of the still-forming sections of the infant’s skull, referred to as fontanelles, the probe emits near-infrared (NIR) light into the SSS, where it is absorbed by hemoglobin, which thermally expands as it is oxygenated (Figure 5).
Figure 5. Noninvasix’s probe emits NIR light into the superior sagittal sinus (SSS) (a), where it is absorbed by hemoglobin, which thermally expands as it is oxygenated (b). The pulse of NIR light generates a measuraurable acoustic signal from the oxygenated hemoglobin, which exhibits a thermoelastic optoacoustic pressure profile distinct from deoxygenated hemoglobin. Courtesy of Noninvasix.
The pulse of NIR light generates a measurable acoustic signal from the oxygenated and deoxygenated hemoglobin. This acoustic wave is detected by an acoustic transducer, and the depth of the signal’s origin in the brain is determined by time resolution. Software analyzes the signal to calculate continuously and in real time the oxygen saturation in the SSS. The ratio of the amplitudes of the oxygenated and deoxygenated hemoglobin signals yields the percent of hemoglobin saturation, which is the fraction of available hemoglobin that is carrying oxygen.
This technique amounts to optoacoustic, or photoacoustic, spectroscopy, according to Noninvasix’s vice president of research and development, Rinat Esenaliev. Randall said the company plans to complete a production prototype, perform clinical trials and obtain FDA 510(k) clearance for use in NICUs by early 2018.
1. J.E. Brazy, et al. (February 1985). Noninvasive monitoring of cerebral oxygenation in preterm infants: preliminary observations. Pediatrics, Vol. 75, pp. 217-225.
2. M. Lai and S. Yang (2011). Perinatal hypoxic-ischemic encephalopathy. J Biomed Biotechnol, Vol. 2011.
3. I. Seoud, et al. (January 2005). Predictors of neonatal mortality in intensive care unit in Children’s Hospital, Cairo University. Alexandria Journal of Pediatrics, Vol. 19, pp. 93-97.
4. H.E. Elsner (2012). The development of cerebral oxygenation in premature infants. Ph.D. dissertation. Duke University School of Nursing.
5. N. Ionita, et al. (October-December 2013). Near-infrared spectroscopy in the neonatal intensive care unit — a literature review. Jurnalul Pediatrului, Vol. 16, pp. 70-73.
6. H.E. Elsner, pp. 16-18.
7. L.M.L. Dix, et al. (2013). Comparing near-infrared spectroscopy devices and their sensors for monitoring regional cerebral oxygen saturation in the neonate. Pediatr Res, Vol. 74, pp. 557-563.
8. G. Greisen, et al. (2011). Has the time come to use near-infrared spectroscopy as a routine clinical tool in preterm infants undergoing intensive care? Phil Trans R Soc A, Vol. 369, pp. 4440-4451.
9. J. Riera, et al. (December 2015). The SafeBoosC phase II clinical trial: an analysis of the interventions related with the oximeter readings. Arch Dis Child Fetal Neonatal Ed, Vol. 101, Issue 4, pp. 333-338.
10. S. Hyttel-Sørensen, et al. (January 2015). Cerebral near infrared spectroscopy oximetry in extremely preterm infants: Phase II randomised clinical trial. BMJ, Vol. 350, g7635.
Advances in Wide-Field Imaging Help With ROP Detection
While many studies using near infrared spectroscopy (NIRS) oximeters on infants have focused on conditions associated with hypoxia — a shortage in the amount of oxygen in cerebral tissue that can result in long-term neurological disorders such as cerebral palsy — hyperoxia, an excess of oxygen, is of equal concern. Such an oversaturation of oxygen, which may be delivered to a premature infant in an incubator, can halt the vascular endothelial growth factor and trigger a potentially blinding eye disorder known as retinopathy of prematurity (ROP)1. The gold standard for diagnosing ROP has long been binocular indirect ophthalmoscopy (BIO), in which the pupils are dilated and the retina are observed with a binocular indirect ophthalmoscope. Although wide-field digital retinal imaging (WFDRI), such as that provided by the RetCam from Clarity Medical Systems Inc. in Pleasanton, Calif., has been growing in popularity, a Turkish research team concluded in a June 2013 study published in Eye that this newer method “cannot completely replace BIO.” WFDRI, which provides a 130-degree field of view compared to BIO’s 30-degree field of view, should instead be used as an “adjunctive method” to the more technically challenging BIO2.
“Babies do not fixate, and their eyes are constantly moving, so [it is] difficult to image out peripherally,” said David S. Yeh, general manager at Fremont, Calif.-based Visunex Medical Systems Inc.
Recent years have seen several interesting developments in WFDRI, particularly in the area of noncontact digital imaging of ROP. These advances, listed below, promise to reduce the risk of infection posed by other imaging systems that require contact with the infant’s eye. These developments include the following:
• Demonstrations of noncontact digital imaging of ROP using an ultrawide-field scanning laser ophthalmoscope with a 200° field of view3 from Optos PLC in Dunfermline, Scotland, and a noncontact digital fundus camera with a 45° field of view4 by Volk Optical Inc. in Mentor, Ohio, have been reported.
• The e-ROP Cooperative Group published the findings of a two-year, large multicenter, National Eye Institute-funded clinical study to gauge the effectiveness of a telemedicine system for the evaluation of acute-phase ROP5. The study involved more than 1,200 U.S. and Canadian infants whose WFDRI scans were taken by nonphysician imagers and remotely evaluated by trained nonphysician readers. The researchers concluded the study showed there was “strong support” for the validity for such a ROP telemedicine system.
• The U.S. Food and Drug Administration last November granted 510(k) marketing clearance for the Fremont, Calif.-based Visunex Medical Systems Inc.’s PanoCam LT wide-field imaging system — the first completely wireless hardware and software solution for neonatal vision screening.
Yeh said “the goal is to screen all newborn babies, not just premature. Since there are limited retinal specialists to read fundus images, a remote review telemedicine system needs to be in place and a simple camera for a non-ophthalmic imager to be able to capture the images needed for review.”
1. O.D. Saugstad (2006). Oxygen and retinopathy of prematurity. J Perinatol, Vol. 26, pp. S46-S50.
2. M.A. Sekeroglu, et al. (September 2013). Alternative methods for the screening of retinopathy of prematurity: binocular indirect ophthalmoscopy vs wide-field digital retinal imaging. Eye (London), Vol. 27, pp. 1053-1057.
3. C.K. Patel, et al. (May 2013). Non-contact ultra-widefield imaging of retinopathy of prematurity using the Optos dual wavelength scanning laser ophthalmoscope. Eye (London), Vol. 27, pp. 589-596.
4. S.G. Prakalapakorn, et al. (August 2014). Retinal imaging in premature infants using the Pictor noncontact digital camera. JAAPOS, Vol. 18, pp. 321-326.
5. G. E. Quinn, et al. (October 2014). Validity of a telemedicine system for the evaluation of acute-phase retinopathy of prematurity. JAMA Ophthalmol, Vol. 132, pp. 1178–1184.
This content appears as it was originally published on Photonics Media.