Unlike adults, infants can’t tell you if they’re in pain. Instead, clinicians must interpret behaviors such as crying and physiological measures such as heart rate to determine what a newborn is experiencing. Since these can occur for reasons unrelated to nociception, the pain field has long sought more objective ways to measure pain in this nonverbal population. Now, in a new study, investigators have identified pain-related brain activity in infants that could be measured with a simple electroencephalogram (EEG) recording and used the activity to create an EEG template that allowed them to test the efficacy of an analgesic.
A team led by Rebeccah Slater, University of Oxford, UK, found that the EEG template of brain activity correlated with the presence and intensity of pain-related behavior and validated the template across four independent samples of infants. Intriguingly, a topical analgesic dampened the brain signal, showing how the new approach could be used to assess the effect of pain medications in infants undergoing painful procedures.
“Measuring pain in infants is quite difficult. With so few researchers in this field, it’s great to see this level of progress, especially by demonstrating clinical utility through determining the efficacy of an analgesic,” said Manon Ranger, Columbia University, New York, who was not involved in the study.
The work was published online May 3 in Science Translational Medicine.
Knowing when a baby is in pain
Determining whether infants are experiencing pain is no easy task, since they have not yet developed language (Ranger et al., 2007). But relying on crying or changes in facial expression, for instance, to achieve this goal is problematic.
“Clearly, these behaviors are not specific to pain,” explained Slater. “These behaviors can be observed for other reasons, like when a baby is hungry.”
Furthermore, when studying infant pain or testing the efficacy of an analgesic in this population, it is critical to have an objective and clear way to determine an infant’s pain level.
“Pain is a subjective experience which an adult can describe. With infants, we have to rely on surrogate measures. We want to develop better ways of measuring and treating infant pain, because pain early in life can lead to changes in how an infant’s brain develops,” said Slater (see PRF related story).
To address this issue, first author Caroline Hartley and colleagues asked whether a template of brain activity could be created that represents a pattern of brain activity evoked by nociceptive stimuli. “Pain manifests in the brain, so in the absence of language, one way of trying to understand it is by looking at what’s happening within the brain,” said Slater.
Eighteen infants between one and nine days old were recruited for the study. These babies were already scheduled to undergo a heel lance in which a small blood sample is obtained from a prick of the heel. For a more controlled noxious stimulus, the researchers used a punctate probe that applies force to the surface of the foot and is described by adults as mildly painful. The brain activity induced by these painful procedures could then be compared to that elicited from a non-noxious heel lance control procedure or simply from gentle tactile stimulation of the heel. Key to the study was the use of EEG rather than other techniques such as functional magnetic resonance imaging (fMRI).
“Progress has been made with fMRI, but it is challenging because the baby needs to be taken away from the parent to be put in the scanner,” explained Slater. “EEG can be performed while the parent is holding the child.”
Eight EEG electrodes were placed on the scalp, and brain activity was recorded for 1,000 milliseconds after noxious or non-noxious stimuli. Between 446 and 611 milliseconds, the researchers found a noticeable pattern of activity at a single electrode sitting on the midline of the scalp that differed between noxious and non-noxious stimuli.
“We can’t tell specifically with EEG where the signal is coming from, but we have a broad idea based on studies in adults that points to areas such as the thalamus, somatosensory cortex, anterior cingulate cortex, and the insula,” said Slater.
Recording and comparing this signal across the babies, the researchers used a statistical approach called principal component analysis to generate a predefined template that could be applied to other groups of babies. The template was then validated in four independent samples of infants. In the future, with the template in hand, researchers could avoid having to identify and characterize a pain signal each and every time they begin a new study.
A consistent and sensitive measure
Premature babies are an even more sensitive infant population, in which it is also difficult to assess pain. So the team again made EEG recordings, this time from 14 preterm infants also undergoing a heel lance procedure. Using the template, the researchers identified increased brain activity in response to heel lance compared to the non-noxious heel lance control procedure.
Hartley and colleagues were then curious about how well the brain activity characterized by the template correlated with pain-related behaviors. They recruited 28 infants who required a heel lance and compared the intensity of the nociceptive signal to a facial expression score that served as a measure of pain intensity.
As predicted, when the facial expression score was high, so was the brain signal; when the score was low, so, too, was the signal. Intriguingly, five of 11 infants with no changes in facial expression showed a pain signal nonetheless, highlighting the sensitivity of the new brain activity template.
Is the signal specific to pain?
Pain is an arousing stimulus, so the team asked whether the noxious-evoked brain activity was specifically generated by the noxious input rather than only an indication of arousal. To test this, 14 more infants were recruited and exposed to non-noxious auditory, visual, or tactile stimuli. None of these stimuli evoked the pattern of brain activity seen with noxious stimuli.
Next, “to measure the effect of arousal, we took all of the trials where these non-noxious stimuli evoked a big change in heart rate, similar to that evoked by noxious stimuli, and looked again for a change in pain-related brain activity,” Slater said.
Although heart rates in response to non-noxious stimuli could match those seen in response to noxious stimuli, the noxious-evoked brain activity was absent.
“We still didn’t see this activity, and that’s important because it suggests that it is not just a measure of arousal,” said Slater.
Testing analgesic efficacy
Conflicting reports exist as to whether certain topical local anesthetics are effective in infants, and it is known that drugs can behave differently in infants compared to adults (Kearns et al., 2003). With their newly identified pain signal, the researchers could directly address this question.
They did so by applying a 4 percent tetracaine gel to one foot of 12 infants and seeing if it reduced noxious-induced brain activity in infants undergoing a clinically required blood draw. Using the punctate probe, they stimulated the treated foot and untreated foot (before the blood draw). Simultaneously answering an important clinical question and validating the usefulness of the template, the pain signal could be observed when stimulating the untreated foot, but not when stimulating the treated foot.
“We found that this analgesic did indeed reduce the magnitude of the noxious-evoked brain activity,” said Slater.
Applying the template to other studies
Further work is needed to determine if the new template representing a specific pattern of nociceptive brain activity is also seen in other forms of infant pain, which can often be more intense than that evoked by the noxious stimuli used in the current experiments.
The new study is “really important work and furthers our knowledge of pain in infants, but it will certainly need to be validated for other noxious stimuli and other age groups. In the analgesic experiment, for example, they only tested whether brain activity induced by the minimally painful [punctate stimulus] was changed,” said Ranger.
Still, the identification and validation of the noxious-induced pattern of brain activity provides for the first time a predefined template of activity that can be used in future studies. Already the team is working on a clinical trial (Procedural Pain in Premature Infants - Poppi) that will use the template to assess the efficacy of morphine to reduce pain during an eye exam that’s critical for preterm infants.
“Babies who are born early are at high risk of developing retinopathy of prematurity. If left untreated, it can lead to blindness. The procedure to test for this is considered to be painful, so we will see if morphine given to these babies prior to the exam reduces the nociceptive brain activity and improves the physiological stability of the infants after the procedure,” said Slater.
The Poppi trial is currently recruiting infants at Oxford and is anticipated to be completed in the fall of 2018.
Nathan Fried is a postdoctoral fellow at the University of Pennsylvania, Philadelphia, US.