Investigating the Feasibility of a Modified Quantitative Sensory Testing Approach to Profile Sensory Function and Predict Pain Outcomes Following Intrathecal Baclofen Implant Surgery in Cerebral Palsy

Abstract

Objectives

Intrathecal baclofen (ITB) pumps used to manage spasticity in children with cerebral palsy (CP) also improve pain outcomes for some but not all patients. The purpose of this clinical feasibility study was to explore whether a quantitative sensory testing approach could a) be modified and used to subgroup individuals into sensory profiles and b) test whether the profiles were related to postimplant pain outcomes (i.e., pain responsive or pain persistent).

Subjects

A purposeful clinical sample of nine children with CP (mean age = 12.5 years, male = 56%) and complex communication needs participated.

Methods

A prospective within-subject design was used to measure proxy-reported pain before and after ITB implant. Pain response status was determined by proxy-reported pain intensity change (>50% change in maximum rated intensity). A modified quantitative sensory testing (mQST) procedure was used to assess behavioral responsivity to an array of calibrated sensory (tactile/acute nociceptive) stimuli before surgery.

Results

Seven individuals with presurgical pain had mQST differentiated sensory profiles in relation to ITB pain outcomes and relative to the two individuals with no pain. Presurgically, the ITB pain responsive subgroup (N = 3, maximum rated pain intensity decreased >50% after ITB implant) showed increased behavioral reactivity to an acute nociceptive stimulus and cold stimulus, whereas the ITB pain persistent subgroup (N = 4) showed reduced behavioral reactivity to cold and repeated von Frey stimulation relative to the no pain individuals.

Conclusion

Implications for patient selection criteria and stratification to presurgically identify individuals with CP “at risk” for persistent postprocedure pain are discussed.

Pain is a well-recognized but poorly understood public health problem in cerebral palsy (CP). CP is the most common childhood-onset motor disorder [1]. Epidemiological estimates of chronic pain in CP are exceedingly high, ranging from 40% to 60% in large representative samples [2–6]. Phenomenological accounts indicate profound effects on quality of life for the affected individual and their family at tremendous human and health care cost [7,8]. Despite decades of documentation, there is little evidence that pain as a secondary condition in CP is routinely assessed or treated consistently or effectively. The most common pain treatment approach is empiric [9,10]. Furthermore, despite advances in pain assessment in clinical populations with motor and communication impairments [11], our scientific understanding of the pathophysiological mechanisms underlying pain in CP remains inadequate. CP is associated with musculoskeletal, myofascial, and central nervous system (CNS) pathologies that may all contribute to pain [12].

Intrathecal baclofen (ITB), a GABA_B receptor agonist, is frequently used to treat spasticity associated with CP. Spasticity is reported in approximately 70% of those with CP and, depending on severity, results in chronic pain and interferes with function and quality of life. ITB was approved for the treatment of spasticity associated with CP in 1996; since that time, there have been well over 10,000 implants for spasticity treatment among adults and, increasingly, children with CP [13,14]. There is limited understanding, however, of individual differences in response to ITB and its effects on pain outcomes. Baclofen is chemically similar to gamma amino butyric acid (GABA), binds with GABA_B receptors, and results in inhibition of mono- and polysynaptic reflexes. There are well-established pathophysiological models showing that changes in GABAergic systems occur in and contribute to aberrant nociceptive (pain) responding [15,16]. Considering the chronic nature of the spasticity associated with CP, the relation between spasticity and muscle pain, the problem of pain reported among individuals living with CP, and the use of a pharmaceutical/surgical intervention targeting a system with known nociceptive regulatory properties, there may well be individual differences or subgroups within the CP population that respond differently to ITB in relation to their pain outcomes.

In this study, we began the initial steps of a subgroup approach by using a modified quantitative sensory testing (QST) protocol to evaluate sensory function in CP in relation to chronic pain outcomes. QST refers to psychophysical methods that collectively provide a test for determining patterns of sensory loss (thermal or mechanical hypoesthesia) or gain (thermal or mechanical hyperalgesia or allodynia). Defining subgroups of CP based on sensory response profiles generated by QST could help delineate functional differences in somatosensory regulation that may ultimately relate to the pathophysiology underlying chronic pain in these individuals [17,18]. Current pain assessment approaches used in CP are most often based on proxy report and will not, in our opinion, accomplish the goal of subgrouping CP patients according to somatosensory function [19]. Most of the sensory testing research conducted with CP has relied on sensory stimuli designed to assess impaired discriminative tactile abilities such as two-point discrimination, texture perception, and shape perception reflecting large-fiber afferent function [20]. There has been little work incorporating sensory testing approaches that simultaneously evaluate loss and gain of function reflecting both large- and small-fiber afferent function. QST approaches provide the opportunity to evaluate the expression of sensory signs in relation to gain or loss of function and reveal clues about pain pathophysiology.

Given the patient population we are addressing (individuals with CP as part of a larger class of individuals with neurodevelopmental disorders), it is important to acknowledge some issues specific to applying QST to vulnerable populations with cognitive, motor, and communication impairments [21,22]. Overall, there have been only a handful of studies with patient samples with developmental disorders. Collectively, results suggest the following: 1) QST procedures can be feasible, but adaptations may be necessary; 2) reaction time confounded by motor impairment is not trivial; 3) there are no definitive guidelines recommending which of the two dominant approaches (method of limits vs levels) should be used; 4) when language is impaired, establishing tactile/nociceptive thresholds is an extraordinary challenge using either method.

Because of the above issues with applying QST, we took an alternative approach for nonverbal populations with complex communication needs by modifying QST, such that behavioral reactivity is recorded (including any vocal responses) and time-locked to application of stimuli without requiring repeated applications. Such an approach sacrifices the ability to measure cutaneous sensory thresholds of various modalities. We have initiated such an approach in multiple clinical and field-based settings with multiple populations that have developmental disabilities and neurodegenerative diseases with evidence of excellent validity and very good psychometric properties in different applications, but we have not used it as a tool to understand pain outcomes in relation to interventions [23–29].

The approach, described in detail below, uses a baseline-controlled calibrated sensory trial procedure in which fine-grained behavioral recording (facial activity) [30] and more global response properties (vocal, motor, etc.) are used in a protocol in which combinations of calibrated tactile and nociceptive stimuli including light touch, deep pressure, pin prick, warm, cold, and repeated mechanical stimuli are compared with a baseline trial application (coders are blinded to stimulus trial type). Using this approach, we can compare within and across participants and within subjects pre- and post-treatment.

There were two specific objectives for this preliminary investigation. One objective was to apply a modified QST approach to test for the presence and severity of positive and/or negative tactile/nociceptive sensory behavior response phenomena [18]. We hypothesized that there would be individuals characterized by “positive sensory” profiles with increased behavioral reactivity response patterns and individuals with “negative sensory” profiles with decreased response patterns [17,31]. The second objective was a preliminary test of the value of sensory subtyping in relation to understanding clinical pain outcomes associated with ITB treatment. To accomplish this objective, we defined pain responsive and nonresponsive outcomes and then investigated whether subgroups characterized by different sensory profiles were associated with different pain outcomes following ITB implant.

Methods

Participants

Nine individual patients with CP (mean age [range] = 12.5 [7.9–26.6] years, male = 55.6%) participated in the study. With regard to communication abilities, three individuals used an augmentative alternative communication (AAC) device, with two of those individuals being considered otherwise nonverbal (no spoken language) and one having some verbal skills. Eight individuals had mild to moderate cognitive impairment, and one individual had no cognitive impairment. Due to the variability in chronological age, cognitive capacity, and language abilities, parent report (described below) was elicited for all participants to maintain consistency. Cerebral palsy diagnoses included quadriplegia (N = 5), diplegia (N = 3), and hemiplegia (N = 1). Study inclusion criteria were a) confirmed diagnosis of cerebral palsy, b) scheduled for a first-time intrathecal baclofen pump implant surgery, and c) parent/guardian consent for participation. Study exclusion criteria included a) parent/guardian or patient did not consent/assent to the study); b) existing cerebral shunt; c) compounded dosing (i.e., opioid adjunctive to baclofen) through their pump; d) comorbid psychiatric disorders (e.g., major depression—routinely screened for clinically as part of the presurgical evaluation); or d) co-occurring chronic pain condition such as juvenile rheumatoid arthritis (routinely screened for clinically as part of the presurgical evaluation). Approval for the study was given through the University of Minnesota Institutional Review Board (IRB), and all participants gave informed consent in accordance with the Declaration of Helsinki.

Procedures

Participation in the study protocol included completion of the sensory testing (specific procedures described below) and parents completing a set of pain assessment measures (described below) pre (day of) and post (approximately three months) ITB implant surgery.

Modified Quantitative Sensory Test

The mQST was conducted during the initial preoperative visit. The mQST is a modification of quantitative sensory testing that scores behavioral responses in nonverbal participants. We have shown that modified quantitative sensory testing can be performed across diverse pediatric patient populations with significant disabilities and communicative impairments (global developmental delay, severe intellectual disability, autism, and Rett syndrome) [23–29].

The mQST protocol (adapted, in part, from Freeman et al. [32] and following the same protocol as Barney et al. [25]) consisted of six calibrated stimuli including light touch, pin prick, cold, deep pressure, repeated application of a von Frey monofilament, and heat. Sensory testing was performed by applying the stimuli to the participants’ right and left foot dorsa while reclined in a quiet preoperative room. Light touch was applied by hand with a von Frey monofilament (2.0 g) pressed against the skin until the filament bent. The filament was touched lightly to the skin five times in five seconds. Second, a light pin prick was applied for less than one second with a plastic pin made for use during neurological exams (Medipin; US Neurologicals). Third, a cold touch was applied lightly to the participant’s skin for five seconds using a cold thermal probe that maintained room temperature (approximately 22°C; Tip Therm, US Neurologicals). Fourth, deep pressure (4.0 lbs) was applied for five seconds using an algometer (Wagner Instruments). Fifth, a von Frey monofilament (60 g) was applied by hand repeatedly to the skin at approximately 1 Hz for 30 seconds with timing guided by a digital timer. Finally, a 3-mm electronic thermal heat probe (WR Medical Electronics) was applied for five seconds (or withdrawal, whichever occurred first) at a temperature of 50°C.

The stimuli were always applied in the same order, and each stimulus application/removal was audibly signaled. Stimuli were presented in order of increasing intensity while also attempting to group stimuli by the primary sensory afferent (A-delta, A-beta, etc.) being tested. The order of application from non-noxious to noxious was designed to reduce the likelihood of carryover effects. To measure behavior during the sensory test, responses of the participant were recorded continuously with a digital camera 2.5 meters from the participant and orthogonal to the body. Stimuli were applied sequentially across the right and left foot dorsa in the pre-ordered sequence of test stimuli. Any codable behavioral reactivity was scored accordingly. For our purposes, behavioral reactivity to testing at either foot was sufficient for our interest in lower extremity sensory function (i.e., reactivity on either foot was coded and summed into one score for each behavioral class for each stimulus).

mQST Behavioral Expression Coding

Because we were unable to obtain reliable self-reports across all participants, we were unable to determine with certainty whether any one child experienced pain per se in relation to the application of a stimulus, nor were the behaviors we measured specific to pain. Thus, we selected the Pain and Discomfort Scale (PADS) [33], a coding system that provides a measure of nonverbal behavioral expression during a standardized patient exam (similar to a range of motion exam), which was derived from the Non-Communicating Children’s Pain Checklist (NCCPC) [34], as the basis for our observational coding. The PADS items and anchors comprehensively code behavioral reactivity in relation to the sensory testing protocol. In our previous work with the PADS (N = 65), the mean inter-rater reliability of the PADS across four raters was 0.82, indicating excellent reliability for clinical screening [23,33–36]. The content validity of the PADS is also excellent, as the items were derived directly from the NCCPC.

In the current coding system, modified slightly from the original PADS, individual behavioral events (i.e., reactions) were grouped by behavior class (vocal, upper face, lower face, gross motor, and physiology). Each behavior class was coded on a scale ranging from 0 to 3 for each stimulus, based on specific operational definitions for each code, then summed to give a total score [33]. Behavioral coders were trained to a minimum 85% criterion across all observational codes using practice videos and were kept blind to which stimuli corresponded to which audible signal (materials were behind the patient, out of view of the camera). All videos were coded independently by two trained observers, and scores were then sampled and compared from the independent scoring for approximately half of the sessions (55%) based on random sampling. Disagreements between coders were consensus-coded to resolution. Interobserver agreement (IOA) was calculated as agreements/agreements + disagreements. Coding agreements were scored one point each (when the same PADS item scores were assigned by both independent coders). Half-points were given when PADS item scores were within one point of one another. Disagreements were scored zero points when PADS item scores between coders differed by more than one point. Preconsensus IOA averaged 96.3% (range = 92.9–100%). Across all participants, it was feasible to code body, physiological, and vocal behavior classes throughout all stimulus applications. However, 7.9% of the time during stimulus application, facial codes could not be assigned due to the face being obstructed during video recording. Additionally, two participants did not experience the heat stimulus due to technical difficulty.

Pain Outcome Measure

At the time of a hospital visit before ITB surgery, the child’s primary caregivers completed the Dalhousie Pain Interview (DPI) based on a one-week recall period [37]. The DPI is presented in an interview/script format consisting of 10 open- and close-ended questions that provide a standardized proxy measure of pain. This measure has been adapted from the methodology used in previous research for obtaining pain information via proxy report when self-report is not possible or is otherwise difficult to obtain. Specific items are anchored to whether there has been pain in the past week, its general description, possible cause, duration (cumulative hours, minutes, and seconds), frequency (number of episodes), and intensity (0–10; 0 means “no pain at all” and 10 means the “worst pain ever”). The DPI also consists of items examining chronic pain, including the type of chronic pain and an intensity score tied to pain “on a bad day” (not specific to the one-week recall; scored 0–10). For this study, as per Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials (IMMPACT) guidelines (described below), our primary pain outcome variable was percent reduction in maximum proxy-reported pain intensity pre– and post–ITB implant (mean follow-up [range] = 4.2 [3–6.8] months).

We based our analyses of pain outcomes on the IMMPACT Consensus Statement on interpreting the clinical importance of treatment outcome in chronic pain trials [38]. IMMPACT is a recurring internationally recognized consensus group of pain experts from academia, government agencies, the pharmaceutical industry, and advocacy organizations. The IMMPACT group provided provisional benchmarks for identifying clinically important changes in specific outcome measures to be used for outcome studies of chronic pain treatments. On the basis of their recommendations, we evaluated the magnitude of change in proxy-reported maximum pain intensity in a one-week recall period as our primary efficacy analysis outcome. Specifically, reductions in pain intensity in individuals of at least 10% to 20% reflected minimal improvement, reductions of ≥30% reflected moderate improvement, and reductions of ≥50% reflected substantial improvement.

mQST Sensory Profiling

From the mQST, we derived individual reactivity scores by summing the frequency of coded behavior for each stimulus trial. We then aggregated the reactivity scores across individuals by pain outcome status. The behavioral reactivity scores were then normalized by z-transforming the reactivity scores against the “no pain” CP individuals (individuals who had no reported pain pre- or post-implant). A z-transformation was calculated for each sensory test variable, based in part on the German Research Network on Neuropathic Pain analysis approach [17,39]. The sign of the z-score was adjusted such that z-values >0 indicate increased reactivity, reflecting more sensitivity to the test stimuli compared with no pain comparison individuals, whereas z-score values <0 reflect a loss of function, indicating lower sensitivity to the test stimuli compared with no pain comparison individuals. A z-score of 0 represents a value corresponding to the mean of the “no pain” same disability comparison patients. Z-scores of 0 +/-1.96 represent the expected range, which would include 95% of no pain control data (i.e., the 95% confidence interval [CI]). If any individual z-score values were outside of the 95% confidence interval of the reference group (i.e., z-scores >1.96 or <1.96), the values were designated as atypical absolute values and considered either greater or lower reactivity, respectively.

Results

Initial Sensory and Pain Characterization

Based on proxy report through the DPI, seven of nine participants were judged to have pain presurgically and were also reported to be living in chronic pain (pain lasting greater than three months). Based on the mQST approach, individuals with reported pain were able to be profiled in relation to behavioral reactivity by stimulus type (Figure 1a). There were clear individual response profiles with respect to patterns of behavioral reactivity to the calibrated sensory stimuli and marked interindividual variability.

Figure 1.

a) Individual response profiles in modified quantitative sensory testing (mQST) reactivity—sham (no touch), light touch with von Frey, pin prick, cold thermal; deep pressure, repeated von Frey, and warm thermal. b) Three relevant pain phenotype outcome subgroups—no chronic pain reported pre/postprocedure (solid grey), pain responsive (i.e., pain intensity reduced >50% postprocedure; horizontal hatch lines), and pain persistent (i.e., pain intensity increased, stayed the same, or reduced <20% postprocedure; diagonal hatch lines). c) Presurgical sensory subtyping through mQST demonstrated a differentiated sensory profile relative to the two individuals with no pain. The subgroup whose proxy-reported pain decreased after intrathecal baclofen (ITB) implant (i.e., maximum rated pain intensity was reduced by >50% after ITB implant; solid line) had presurgical sensory profiles characterized by increased behavioral reactivity to the acute nociceptive stimulus and the cold stimulus (outside 95% confidence interval [CI] of the reference data). The presurgical sensory profile for the ITB pain persistent subgroup (<20% change or no change, i.e., “pain persistent”; broken line) was characterized by reduced behavioral reactivity for cold and repeated von Frey. All other sensory test values from the two groups were within the 95% CI.

Pain Intensity Outcome

Using the 50% reduction in pain intensity criterion, after ITB placement, three of the nine individuals in the sample had substantial improvements in their proxy-reported maximum pain intensity (mean reduction [range] = 82% [75%–89%]), four of the nine individuals in the sample had worsening or minimal to no change (maximum pain intensity increased [2], no change [1], or <20% reduction in proxy-rated pain intensity [1]), and two of the nine individuals in the sample had no proxy-reported pain before or after ITB implant surgery. All participants with pain had been living with chronic pain for longer than three months, and their pain was considered to be musculoskeletal, originating from bones, joints, and/or muscle spasticity or spasm. Chronic pain intensity ratings for “a bad day” (not necessarily observed in the previous seven days) before and after ITB pump implant followed the same pattern (i.e., pain persisted, improved, no pain pre/post) as the pain intensity ratings recalled for the previous week.

Pain Intensity Outcome by Sensory Subtypes

Of the seven individuals with presurgical pain, mQST demonstrated a differentiated sensory reactivity profile in relation to ITB pain outcomes and relative to the two individuals with no pain (Figure 1c). The ITB pain responsive subgroup (i.e., maximum rated pain intensity decreased >50% after ITB implant) had presurgical sensory profiles characterized by greater behavioral reactivity for the acute nociceptive stimulus and cold stimulus (outside 95% CI of the reference data). The presurgical sensory profile for the ITB pain nonresponder subgroup (<20% change or no change, i.e., “pain persistent”) was characterized by less behavioral reactivity to cold and to the repeated von Frey. All other sensory test values from the two groups were within the 95% CI of the referent “no pain” individuals (Figure 1c).

Discussion

This was a feasibility study exploring the value of adapting a QST approach to nonverbal individuals with CP to apply a sensory profiling procedure and improve our understanding of pain outcomes after surgical and drug intervention for spasticity. We successfully utilized a modified QST protocol in a CP sample. This adds a novel patient population to prior work with individuals with neurodevelopmental disabilities and extends its application to a pre/postintervention context [26–32]. The individual differences in pain outcomes for the CP participants appeared to be meaningfully subgrouped in relation to sensory function consistent with a “gain/loss” perspective (but see the limitations noted below). This is consistent with the large body of conventional QST clinical research within other pain populations [18].

McLaughlin et al. also applied QST to the lower extremities in young children with CP before and after a surgical procedure (selective dorsal rhizotomy [SDR]) [40]. They demonstrated feasibility, found group differences from healthy controls in some sensory modalities, and identified no detectable changes in sensory reactivity after the SDR. Our study differs from the McLaughlin study in two important respects. Their study was before and after SDR, whereas ours was before and after ITB pump implant; these tend to be different CP patient groups. Unfortunately, we did not complete the mQST postimplant, so we do not know whether the scores produced by the mQST changed for the participants. The McLaughlin et al. sensory testing was aimed largely at impaired discriminative tactile function (i.e., loss of function) and depended primarily on stimuli designed to elicit large-fiber sensory afferent activity. The stimuli used in our study were aimed at sensory function involving both large- and small-fiber regulated modalities.

There are several important limitations to this study that should be pointed out to further contextualize the results. First, this was a convenience sample, and as such the findings may not be representative of the population of children with CP. The sample size in this exploratory, proof-of-principle study was small. The pain outcome measures relied on proxy report; although this is a limitation, it is the accepted approach for assessment of pain in nonverbal and minimally verbal patients with neurodevelopmental disabilities with associated motor, communicative, and intellectual impairments [41]. We recognize and acknowledge that some percentage of the variance in any one child’s pain score is likely accounted for by variance in the proxy. Last, because we have no way of knowing for certain the type of pain being reported on by proxy, it could be that the children whose pain was more responsive to baclofen had pain primarily of musculoskeletal origin, whereas those who were less responsive had other types of pain. Thus, the pain outcomes may not be dependent on sensory gain or loss but rather the type of pain that determines the pain outcome of the implant. Although this is possible, it is likely that the majority of reported pain was in some way related to spasticity (i.e., of musculoskeletal origin) given that it—spasticity—was and is the primary presenting concern that leads a CP patient to be a surgical candidate (as opposed to accidental pain or gastrointestinal pain, etc.). These limitations, considered collectively, suggest appropriate caution in interpreting the pain outcome findings.

In addition to the preliminary nature of the pain outcome findings, the results in relation to the sensory subgrouping are also preliminary and bound to this sample of children living with CP. In this regard, it is important to point out that we did not have a no pain nondisability (i.e., healthy control) comparison group, nor did we have a within-subject no pain referent site to which our mQST protocol was applied. Because of this, we were not able to make definitive claims about sensory gain or loss in a normative sense. The issue of appropriate controls for severe physical disability is not always easy to resolve; one way forward would be to include verbal and nonverbal children with CP to generate further validity evidence for our modified QST protocol and measurement approach. Future work seems warranted, and studies should be designed to include appropriate controls to improve the conceptual precision of the measurement approach and bring it more fully in line with contemporary QST approaches that test for dynamic and static mechanical allodynia, punctate hyperalgesia, thermal allodynia, and temporal summation. Interestingly, original observations from 1963 seemed to indicate that children with CP with greater motor impairment were more (not less) sensitive to pressure pain [42], which underscores our interest in more fully documenting sensory loss and gain in CP. Doing so may help to shed new light on pain pathology in CP.

Improved understanding of pain mechanisms in CP would facilitate personalized strategies for pain management. One of the clinical problems created by the limited scientific study of mechanism as it relates to treatment outcomes is a corresponding lack of patient selection criteria designed to optimize outcome by producing decision aids. Little is known about the relationship between sensory impairments and chronic pain in CP. By contrast, pain phenotypes in other chronic pain patient populations (e.g., fibromyalgia, polyneuropathy, spinal cord injury, radiculopathy) are often characterized by the presence of spontaneous, ongoing, and evoked pain, which may be a function, in part, of some form of central sensitization. Evoked pain properties include allodynia (pain elicited by a non-noxious stimulus) and hyperalgesia (increased pain response to a noxious stimulus). In CP, deep tissue inflammation may lead to enhanced excitability in somatosensory neurons, in particular nociceptors and dorsal horn neurons in the spinal cord [43,44]. Experimental evidence from joint and tissue mechanics and muscle morphology studies suggest that spasticity alters muscle tissue [45]. Theoretically, the initial local ischemia from sustained muscle fiber contraction may have several potential deleterious “downstream” effects including pathophysiological inflammatory and nociceptive consequences, affecting excitatory or inhibitory circuits peripherally but also centrally.

Overall, there remains a paucity of evidence regarding the effect of ITB on pain in children with CP [46,47]. Relative to no pain comparison individuals, we found that ITB pain responders were more responsive to an acute nociceptive stimulus and a cold stimulus pre–ITB implant, whereas ITB pain nonresponders showed patterns of less responsivity to cold and repeated light touch (von Frey). Whether this procedure and approach would have clinical value by identifying in advance of surgery those “at risk” for persistent post–ITB implant pain would require a much larger sample and a prospective study design. The pattern of increased sensory responsivity in a subgroup of our cohort is similar to other chronic widespread pain populations and is consistent with central sensitization [32]. Central sensitization is a neuroplasticity phenomenon established in preclinical models in which repetitive activation of nociceptive primary afferents leads to an activity-dependent increase in the excitability of neuronal circuitry in the central nervous system that can contribute to chronic pain. Further work is needed to determine whether central sensitization contributes to chronic pain in CP. Discovering and defining sensory subtypes that align with underlying central transduction and transmission mechanisms would be a critical step forward toward meaningfully subtyping CP patients in relation to sensory function.

Funding sources: Funded, in part, by Eunice Kennedy Shriver NICHD Grant No. 73126, the Gillette Children's Foundation, and the Mayday Fund.

Conflicts of interest: There are no conflicts of interest to report.