Early Insights into the Neurobiology of Pain in Sickle Cell Disease: A Systematic Review of the Literature

Abstract

Novel insights into the neurobiology of sickle cell disease (SCD) pain have recently been discovered. We systematically reviewed the literature focusing on original research that examined the biology of pain in SCD and/or addressed assessment or treatment of neuropathic pain in SCD. This review of 15 articles that met inclusion criteria provides epidemiological, basic and clinical data that support central and/or peripheral nervous system abnormalities likely contribute to sickle cell pain. Continued basic and clinical investigation into pain neurobiology is imperative to translate these discoveries into novel ways to assess and treat neuropathic pain and decrease patient suffering.

Introduction

Published research in both the murine model and humans over the past 5 years has offered new and novel insights into factors that may contribute to the underlying neurobiology of sickle cell pain. This research suggests that sickle cell disease (SCD) pain has a neuropathic component whereas in the past the pain has been thought to be only nociceptive. Neuropathic pain is defined as pain that is initiated or caused by a primary dysfunction of the nervous system that affects the somatosensory system.[1] Compared to inflammatory or nociceptive pain, which is the result of tissue damage, neuropathic pain is the result of damage to the peripheral or central nervous system that in turn produces pain. Classic clinical characteristics of neuropathic pain include hypersensitivity, exaggerated pain from a stimulus that normally produces pain, and allodynia, pain precipitated by a stimulus that is not normally painful (i.e., extreme sensitivity to cool stimuli).[2, 3]

This basic, clinical, and translational research has contributed to the development of novel theories for sickle cell pain that go beyond mechanical vaso-occlusion as the sole explanation for pain. These emerging theories suggest pathological changes in the peripheral and/or central nervous system contribute to the development and maintenance of sickle cell pain. For example, theories focused on neuroplasticity, peripheral/central nervous system sensitization, and chronic inflammation address how these factors may ultimately result in the development of abnormal pain sensitivity and chronic neuropathic pain. Collectively, this recent work has opened up novel avenues to explore when investigating the underlying neurobiology of sickle cell pain that may one day lead to innovative treatments and/or preventative measures. To date, a systematic review of this significant and novel work has not been done. Thus, the objective of this manuscript is to systematically review the existing published literature focused on the neurobiology of pain in SCD in both murine and human models.

Methods

Data Sources and Searches

The following data sources were used for the literature search: 1) MEDLINE (1950-1/12/2015), 2) CINAHL (1982-1/12/2015), 3) PsycINFO (1806-1/12/2015). Ongoing research related to the topic was assessed using ClinicalTrials.gov. The search criteria for all databases included MeSH terms sickle cell and: pain, neuropathic pain, neurobiology, biology of pain, central/peripheral sensitization, pain sensitivity, quantitative sensory testing. To identify additional studies, the references from reviews and other manuscripts were reviewed and these studies were obtained as appropriate.

Study Selection

The selection of studies to include was done by the primary investigator (AB) and these were confirmed by a secondary investigator (RF). Any disagreement among these two investigators was settled by a third investigator (JP). The titles and abstracts were reviewed for eligibility based on the following inclusion and exclusion criteria defined a priori. Inclusion criteria were original basic or clinical research studies published in English in a peer-reviewed journal that: 1) examine biology of pain in SCD and/or 2) specifically address assessment and or treatment of neuropathic pain in SCD. Citations were excluded if they were primarily review articles, educational materials, commentaries or editorials, case reports, or hypothesis papers.

Results

Overall Search Methodology Results

The search of the 3 different databases identified 2328 published articles. These articles were screened by title/abstract and 2307 were subsequently eliminated leaving 21 full text articles to further assess for eligibility. Of these, six were excluded. Thus, 15 articles were included in the qualitative synthesis. Of these, there were six murine studies and nine human studies. Based on the content of the published articles, the manuscripts were grouped into the following categories: murine studies (n=6), patient-reported outcome studies (n=3), quantitative sensory testing studies (QST) (n=5; one QST study also included patient-reported outcome data thus this study is listed in both categories), epidemiological studies (n=1), and interventional studies (n=1). Search of ClinicalTrials.gov revealed four ongoing trials, three interventional and one mechanistic.

Murine Studies Investigating Pain Neurobiology

There were six published murine studies that originated out of three independent basic science laboratories investigating the neurobiology of SCD pain (Table I).[4-9] The main objectives and outcomes of these studies are described in Table 1. In summary, all studies provided evidence that abnormalities exist in the peripheral and/or central nervous system that likely contribute to the pathobiology of SCD pain in a murine SCD model. One finding common to all studies was evidence of mechanical and thermal (cold, heat) hyperalgesia/hypersensitivity as measured by well validated pain behavior studies in the mouse. In addition, the interaction of the immune system and the nervous system, termed neurogenic inflammation, was shown to be an important contributing factor for this pain biology. Importantly, 3 studies demonstrated the ability to reverse the observed hyperalgesia/hypersensitivity with compounds or drugs by different mechanisms providing pre-clinical evidence for potential novel therapeutics to treat sickle cell pain.[4-6] In summary, one study revealed cannabinoid receptor agonists reversed the mechanical, cold and heat hypersensitivity.[4] Vincent et al. showed that targeting mast cells provided the ability to reverse hyperalgesia.[5] Specifically, imatinib, a mast cell inhibitor, decreased hyperalgesia and prevented hypoxia/reoxygenation induced hyperalgesia and cromolyn sodium, a mast cell stabilizer, also decreased hyperalgesia.[5] Finally, inhibition of the mechanical pain receptor TRPV1 by a targeted compound reversed mechanical hypersensitivity.[6] Continued translation of these basic science findings to humans with SCD is imperative to determine whether these abnormalities also exist in patients with SCD and will facilitate the translation of these findings to potential novel therapeutic interventions to treat or prevent pain.

Table I. Murine Studies Evaluating the Neurobiology of Pain in Sickle Cell Disease.

Reference	Hypothesis or Objectives	Mouse Model	Primary Findings	Pharmacologic or Therapeutic Application
Minnesota Laboratory
Kohli et al. Blood 2010	Objectives “Use behavioral measures to characterize pain in transgenic mice expressing human HbS” Assess the analgesic effects of opioids and cannabinoids” in sickle mice Investigate morphologic and biochemical changes in primary afferent neurons and in the spinal cord” of sickle mice	Sickle HbSS-BERK (homozygous) HbAS-hBERK1 (hemizygous) Control HbA-BERK	Sickle mice have evidence of hypersensitivity to mechanical, heat, and cold stimuli and deep musculoskeletal pain Hypersensitivity was reversed in sickle mice by both morphine and cannabinoid receptor agonists Evidence of peripheral neuropathy in skin of sickle mice manifested by morphologic changes Increased levels of CGRP and Substance P in skin of sickle mice Abnormal levels of neurochemicals (COX-2, IL-6, Toll-like receptor 4) in spinal cord of sickle mice providing supporting evidence for central sensitization	Cannabinoid compounds may provide pain relief in patients with SCD
Vincent et al. Blood 2013	Hypothesis “Mast cell activation and neuropeptide release (substance P and CGRP) contribute to the pro-inflammatory milieu and pain in sickle cell anemia.” Objective “Investigate the role of mast cell activation and neurogenic inflammation in the pathophysiology of pain in sickle cell anemia.”	Sickle HbSS-BERK Sickle Mast cell knockout “Sickle mice backcrossed with WBB6F1-KitW/KitW-v mice with point mutation in c-kit lacking mast cells”	Sickle cell pain is exacerbated by activation and degranulation of mast cells which stimulate neurogenic inflammation and stimulation of nociceptors after substance P is released from the skin and dorsal root ganglion Imatinib, inhibitor of mast cells, decreased hyperalgesia and prevented hypoxia/reoxygenation hyperalgesia Mast cell stabilizer, cromolyn sodium, also decreased hyperalgesia in combination with low dose morphine	Mast cell targeting may be novel approach for sickle cell pain treatment
Milwaukee Laboratory
Hillery et al. Blood 2010	Objective “To quantify SCD-associated touch sensitivity and pain behavior”	Sickle HbSS-BERK Controls HbAA on same mixed BERK background	Sickle cell mice have hypersensivity to mechanical, cold, and heat stimuli as measured by valid behavioral studies Mechanical hypersensivity was exacerbated when acute sickling was induced by hypoxia/reoxygenation A selective antagonist of TRPV1 decreased mechanical hypersensitity as measured by behavioral studies and reversed mechanical hypersensitivity at primary afferent terminals “TRPV1 receptor contributes to primary afferent mechanical sensitization and a substantial portion of behavioral mechanical hypersensitivy in SCD mice.”	TRPV1 targeted compounds could be used to treat pain in patients with SCD
Garrison et al. Molecular Pain 2012	Objectives To determine if low threshold mechanoreceptors are sensitized in sickle mice. To determine the mechanisms that drives this potential sensitization.	Sickle HbSS-BERK Controls HbAA on same mixed BERK background	Using repetitive applications of innocuous punctate and dynamic light touch force, sickle mice have increased responses to these stimuli compared to control mice. Increase in number of evoked action potentials were seen in cutaneous afferents from sickle mice suggesting sensitization
Zappia et al. Pain 2014	Objectives Assess cold sensitivity using a temperature preference assay Assess C fiber nociceptor cold response Determine evidence for altered response to TRPM8 and TRPA1 agonists in sensory neurons (sickle and controls) Measure levels of cold-sensitive ion channels and 83 pain-related genes via mRNA expression	Sickle HbSS-BERK Controls HbAA on same mixed BERK backgroundC57BL/6 wild-type	Sickle cell mice had pronounced cold hypersensitivity as measured by a cold preference test Cold hypersensitivity was exacerbated with increasing age C fibers in sickle cell mice had evidence of detection thresholds that were more sensitivity to cold Cold detection ion channels (TRPM8 and TRPA1) were evaluated by mRNA expression and were similar in sickle cell mice and control mice Increased mRNA expression of pain related genes of endothelin-1 and tachykinin receptor 1
Kenyon et al. Exp Biol Med 2014	Objective “to examine the effect of SCD, strain, genotype, age, and sex on somatosensory fiber function and nocifensive behavior in two strains of humanized SCD mice” using sine-wave stimulation	Sickle Townes strain HbSS-BERK Controls C57BL/6 wild-type	Sensitization of sensory nerve fibers exist in both the BERK and Townes mice compared to controls as measured by sine-wave stimulation There was evidence of abnormal thermal sensitivity in both mouse strains (BERK, Townes) Myelinated and unmyelinated nerves appear sensitized in sickle mice

Human Studies Investigating Neuropathic Pain

The evaluation of a patient with suspected neuropathic pain requires a multifaceted approach. This approach includes assessment of self-reported neuropathic pain symptoms with validated patient-reported outcome measures in addition to a quantitative evaluation of the somatosensory system using Quantitative Sensory Testing (QST). QST is further described below. Valid patient-reported screening tools for neuropathic pain exist to identify patients with a neuropathic pain phenotype.[10] These patient-reported outcome measures yield important data from the patient perspective about qualitative aspects of their pain and can help in phenotyping pain as being neuropathic or nociceptive. Further evaluation of pain can be done using psychophysical testing of the somatosensory system using QST to evaluate for changes that suggest an abnormality in pain sensitivity or pain processing. When assessing a patient for neuropathic pain, including both patient-reported assessment and QST allows for a comprehensive evaluation by: 1) defining the overall nature of a patient's pain (i.e., neuropathic vs. non-neuropathic) and 2) better defining the underlying neurobiology of a patient's pain (i.e., cold, heat or mechanical hypersensitivity) by QST which can direct targeted therapies.

Patient-reported Outcome Studies

Three studies utilized patient-reported outcome measures to qualitatively assess pain in patients with SCD (Table II).[11-13] The overlapping themes of these three studies were to define a cohort of patients with SCD who experience clinical characteristics of neuropathic pain and to identify risk factors for neuropathic pain. Wilkie et al., administered the PAINReportIt to 145 adults with SCD to systematically assess sensory pain characteristics.[11] The measure asked patients to choose words that best described their pain from a list of 78 verbal pain descriptors. These descriptors were a priori determined by the investigators to be neuropathic or nociceptive in origin. The investigators found patients chose a mean of 4.5±3.4 neuropathic pain descriptors (i.e., aching, burning, cold, numb, radiating) and 6.8±4.0 nociceptive pain descriptors (i.e., crushing, hurting, pounding, squeezing).[11] Brandow et al. administered a neuropathic pain screening questionnaire, the painDETECT questionnaire, to 56 patients with SCD 14 or more years of age.[12] Data from this study revealed 37% of patients had scores indicative of neuropathic pain.[12] This study also found that older age and females were associated with higher painDETECT scores suggesting that neuropathic pain may be more likely to occur in older and female patients.[12] As part of a larger study, Ezenwa et al. administered two different neuropathic pain screening questionnaires [Leeds Assessment of Neuropathic Symptoms and Signs (S-LANSS) and Neuropathic Pain Symptom Inventory (NPSI)] to 25 adults with SCD.[13] Data revealed 40% of patients met criteria for neuropathic pain based on the S-LANSS.[13] Scoring of the NPSI does not generate a cut-point to differentiate neuropathic from non-neuropathic pain thus the proportion could not be identified, however the mean score was 27±18.6 based on a potential range of 0-100 with higher scores representing increased likelihood of neuropathic pain.[13] In summary, patient-reported outcome data support the existence of neuropathic pain in patients with SCD. Importantly, future utilization of these self-report measures could direct pain treatment clinically and allow for the selection of the correct patient pain phenotype for inclusion in neuropathic pain treatment clinical trials.

Table II. Patient-reported Outcome Studies Evaluating Neuropathic Pain in Patients with Sickle Cell Disease.

Study	Primary Objectives and/or Hypotheses	Study Design	Study Population	Assessment tool	Demographics	Primary Outcomes	Conclusions
Wilkie et al.* J Natl Med Assoc 2010	Objective “to describe sensory pain (location, intensity, quality, pattern), patient-related pain barriers, and the analgesics used by adult outpatients with SCD”	Cross-sectional	≥18 years with SCD	PAINReportIt tool asks patients to choose from 78 pain descriptors that were grouped into neuropathic, nociceptive, or other	n=145 Mean (SD) age: 34 (11.5) yrs 67% female	Mean (SD) of 4.5 (3.4) neuropathic pain descriptors selected 8.3% of patients selected no neuropathic pain descriptors Mean (SD) of 6.8 (4) nociceptive pain descriptors selected Adjuvant drugs effective for neuropathic pain used by only 14% (n=20) of patients	Almost all patients selected pain descriptors that are considered both neuropathic and nociceptive Suggests that pain in SCD also has a neuropathic component Few patients took adjuvant drugs effective for neuropathic pain
Brandow et al. Pediatr Blood Cancer 2014	Objectives determine the presence of clinically evident neuropathic pain in patients with SCD describe the sensory symptoms that patients with SCD experience when in pain evaluate the relationship between neuropathic pain, age, and gender Hypothesis Primary: “at least 20% of patients with SCD will report experiencing neuropathic pain Secondary: “the presence of neuropathic pain will be associated with older age and be more likely in female patients”	Cross-sectional	≥14 years with SCD	painDETECT Scoring: Total score: 0-38 Scores ≥13 designated as having evidence of neuropathic pain	n=56 Median (IQR) age: 20.3 (17-29) yrs 77% female	37% of patients had scores ≥13 indicative of neuropathic pain Age was positively correlated with total score (r = 0.43; p = 0.001) Females had higher mean total scores (13 vs. 8.4; p = 0.04) 5% of patients taking a neuropathic pain drug	Almost 40% of patient sample has evidence of neuropathic pain Older age was associated with higher painDETECT scores Female gender was associated with higher painDETECT scores Very small proportion of patients taking drugs specifically targeted to treat neuropathic pain
Ezenwa et al. Pain Practice 2014	Objective “to determine the feasibility of utilizing neuropathic pain questionnaires among adults with SCD”	Cross-sectional	≥18 years with SCD	S-LANSS: The Leeds Assessment of Neuropathic Symptoms and Signs Scoring: Total score: 0-24 ≥12 indicative of neuropathic pain NPSI: Neuropathic Pain Symptom Inventory Scoring: Total score: 0-100 Higher scores increase likelihood of neuropathic pain No cutoff score established to differentiate neuropathic from non-neuropathic pain	n=25 Mean (SD) age: 38.5 (12.5) yrs 72% female	S-LANSS 40% of patients had scores of ≥12 indicative of neuropathic pain NPSI Mean (SD) score of 27±18.6 Score range 0-80 Mean NPSI scores in S-LANSS high group (scores ≥12) were significantly higher than mean NPSI scores in S-LANSS low group (scores <12) (39.5±25 vs. 18.7±20.3; p=0.01) S-LANSS and NPSI scores were correlated (r=0.66, p<0.001)	40% of patients with SCD meet criteria for neuropathic pain based on screening questionnaire (S-LANSS)

^*Only the portions of the study/data are presented that were most pertinent to our study aims/inclusion criteria

Quantitative Sensory Testing Studies

Five studies were identified that utilized quantitative sensory testing (QST) in patients with SCD (Table III).[13-17] QST is psychophysical testing used to evaluate the sensitivity of the somatosensory system in humans with a specific focus on thermal (heat, cold) and mechanical (pressure) sensitivity. QST can detect sensory loss (hyposensitivity) or gain (hypersensitivity) through the evaluation of two different thresholds called detection thresholds and pain thresholds. Detection thresholds describe when the subject first feels the sensation of cold, warm, or pressure after the application of a defined stimulus that progressively becomes warmer or colder or progressively exerts more pressure. Pain thresholds describe when the same progressively applied stimulus (i.e., cold, heat, pressure) becomes painful. The main findings of these five studies are outlined in Table III. Two studies compared QST findings between patients with SCD and race-matched African American controls[14, 15] and two studies performed QST in patients with SCD only.[13, 16] In summary, Brandow et al. found that patients with SCD have hypersensitivity to cold and heat stimuli as reflected by significantly decreased cold and heat detection thresholds and cold and heat pain thresholds comparted to African American controls.[14] These data parallel the findings in the murine model as outlined above and in Table 1. There were no differences in mechanical sensitivity between patients with SCD and controls in this study.[14] In contrast, O’Leary et al. found that patients with SCD have increased cold and heat detection thresholds but decreased cold pain thresholds providing evidence for cold hypersensitivity.[15] Mechanical sensitivity was not measured in this study. In a study without a control group, Jacob et al[16] found 27% of their study population had one of the tested thermal thresholds (detection and pain) outside of the 2.5-97.5% reference interval for healthy historical controls that were predominantly Caucasian, suggesting hyposensitivity to both cold and heat. The patients with abnormal thermal thresholds had increased pain to mechanical brush stimuli.[16] Finally, Ezenwa et al.[13] completed a study in adults with SCD without a control group designed to evaluate the safety of QST and was thus not powered to evaluate differences in sensitivity outcomes. They found QST was safe.[13] In addition, this preliminary analysis, compared to historical published data, found adults with SCD have evidence of hypersensitivity to cold, heat, and mechanical stimuli.[13] In a related study using different methodology, Hollins et al. evaluated whether increased temporal summation occurred in patients with SCD compared to healthy African American controls.[17] Temporal summation is defined as an increase in the intensity and unpleasantness of pain that is provoked by a noxious stimulus that is constantly applied and provides supporting evidence for central sensitization. This study used constant pressure as the noxious stimulus and did not assess heat or cold stimuli. The main study findings are in Table III. Briefly, temporal summation measures were not different between patients with SCD and controls. Within the SCD group, the intensity pain ratings of prior pain episodes significantly predicted the rate of pain unpleasantness (i.e. “pain builds up rapidly”).[17] The variability of findings in these studies underscore the challenges of studying pain in patients with SCD due to the immense phenotypic variability that likely manifests with different pain features. Furthermore, the variability in these data when patients were tested in their baseline state provides strong rationale for the need to study the sensory abnormalities in patients with SCD during an acute pain event. In conclusion, despite the variability, all of these studies provide data that consistently support patients with SCD have abnormalities in the peripheral and/or central nervous system (i.e., peripheral or central sensitization) that could contribute to the development of neuropathic pain in SCD. These data support the need for additional work in this area of investigation.

Table III. Quantitative Sensory Testing Studies Completed in Patients with Sickle Cell Disease.

Study	Primary Objectives, Research Questions and/or Hypotheses	Study Design	Testing Methodology	Data Analysis	Demographics	Site Tested	Cold Detection Threshold	Heat Detection Threshold	Mechanical Detection Threshold	Cold Pain Threshold	Heat Pain Threshold	Mechanical Pain Threshold	Comments/Additional Findings
		Inclusion Criteria	Primary Outcomes
Brandow et al. AJH 2013	Primary Objective: “To quantify the difference in sensitivity to thermal (cold, heat) and mechanical stimuli between patients with SCD and healthy African American controls.” Hypothesis: “Patients with SCD will exhibit hypersensitivity to mechanical and thermal stimuli compared to healthy race-matched controls as measured by decreased cold, heat, and mechanical pain and detection thresholds”	Cross-sectional	Thermal: Medoc TSA-IIˆ Method of limits* Mechanical: vonFrey monofilaments Method of limits*	Primary outcomes compared between patients with SCD and controls	SCD n=55 age: 15.4 (6.3) yrs 60% female Controls n=57 age: 16.3 (10.2) yrs 56% female	Thenar eminence nondominant hand	SCD patients were more sensitive to cold detection (29.5°C vs. 28.6°C p=0.012)	SCD patients were more sensitive to heat detection (34.5°C vs. 35.3°C p=0.02)	No difference between SCD and controls	SCD patients had significantly increased sensitivity to cold pain (21.1°C vs. 14.8°C p=0.01)	SCD patients had significantly increased sensitivity to heat pain (42.7°C vs. 45.2°C; p=0.04)	No difference between SCD and controls	Older age associated with lower cold heat and mechanical pain thresholds in SCD patients and controls. No association between thresholds and pain frequency, gender or markers of hemolysis.
		SCD All genotypes ≥7 yrs Controls African American ≥7 yrs	Detection Thresholds Cold (°C) Heat (°C) Mechanical (g) Pain Thresholds Cold (°C) Heat (°C) Mechanical (g)			Lateral dorsum of foot randomized to right or left	No difference between SCD and controls	No difference between SCD and controls	No difference between SCD and controls	No difference between SCD and controls	No difference between SCD and controls	No difference between SCD and controls
O’Leary et al. Clin J Pain 2014	Hypothesis: “Children with SCD who experience recurrent pain are more sensitive to thermal pain and sensory stimuli compared with healthy African American children.”	Cross-sectional	Thermal: Medoc TSA-IIˆ Method of limits* Mechanical: Testing not performed	Primary outcomes compared between patients with SCD and controls	SCD n=27 age: 14.8 (2.4) yrs 56% female Controls n=28 age: 14.4 (1.9) yrs 36% female	Thenar eminence of nondominant hand	SCD patients were less sensitive to cold detection (29.2°C vs. 30.0°C; p=0.015)	SCD patients were less sensitive to heat detection (38.6°C vs. 35.9°C; p=0.006)	Not tested	No difference between SCD and controls	No difference between SCD and controls	Not tested	Study was powered on Heat Pain Thresholds. Initial sample size was 41 and was decreased to 27 due to issues with recruitment May be underpowered for differences in cold pain as this was a secondary outcome Thermal perceptual sensitization also tested (see full reference)
		SCD All genotypes 10-18 yrs Controls African American 10-28 yrs	Detection Thresholds Cold (°C) Heat (°C) Pain Thresholds Cold (°C) Heat (°C)			Volar surface of dominant forearm	No difference between SCD and controls	No difference between SCD and controls	Not tested	SCD patients had significantly increased sensitivity to cold pain(19.4 vs. 5.8; p=0.03)	No difference between SCD and controls	Not tested
Jacob et al. J Pediatr Hematol Oncol 2014	Hypothesis: “Children with SCD have abnormal mechanical and thermal sensory patterns due to the repetitive vaso-occlusive events.”	Cross-sectional	Thermal: Medoc TSA-IIˆ Method of limits* Mechanical: Nonpainful stimulus: cotton ball, soft brush Painful stimulus: pin-prick, Neuropen (40g) Method of levels (VAS)#	Children defined as having abnormal sensory testing if values were outside 95% reference intervals (< 2.5% and > 97.5%) in healthy historical controls	SCD n=48 age: 13.7 (2) yrs 45.8% female	Thenar eminence of hand randomized to right or left Overall results: 13/48 (27%) had one of tested thermal outcomes (detection or pain) outside of 2.5-97.5% reference interval for healthy historical controls	Median SCD→ 23.9°C vs. historical controls→ 30.5°C No statistical comparison	Median SCD→ 37.8°C vs. historical controls→ 33.7°C No statistical comparison	Not tested	Median SCD→ 3.4°C vs. historical controls→ 14.9°C No statistical comparison	Median SCD→ 49.7°C vs. historical controls→ 41.7°C No statistical comparison	Not tested	No control group; used historical controls that were not African American No discussion of sample size- unclear why choose 48
		SCD All genotypes 10-17 yrs No control group (used published control data)	Detection Thresholds Cold (°C) Heat (°C) Pain Thresholds Cold (°C) Heat (°C)			Forearm randomized to right or left	Not tested	Not tested	Not tested	Not tested	Not tested	In 13 patients with abnormal thermal outcomes: More pain with brush stimulus No differences with cotton or pin-prick stimulus.
Ezenwa et al. Pain Practice 2014	Aim: “To determine the safety of a QST protocol in adults with SCD.” Hypotheses: “A sensory detection and pain threshold protocol would be safe.” “…the proportion of adults with SCD with positive neuropathic pain indicators from QST and self-report tools would be as large or larger than proportions of people with neuropathic pain detected in community-based general population studies.”	Cross-sectional Designed as safety study thus not powered to look at differences in sensitivity outcomes	Thermal: Medoc TSA-IIˆ Method of Limits* Mechanical: vonFrey monofilaments Method of Limits/Levels (0-10 scale)*#	Thermal Subjects defined as having abnormal thermal sensory testing if: pain threshold was ½ SD above/below published pain norms for age and site or Pain score to thermal stimulus was >3/10 Mechanical Subjects defined as having abnormal stimuli response if: pain was reported in response to 0.6-10g filament or pain was ≥3 in response to 26-60g filament.	SCD n=25 age: 38.5 (12.5) yrs 72% female	Nonpainful site upper body specific site not reported	Nonpainful site (upper body) 18-39y: 29.04°C ≥40y: 28.43°C	Nonpainful site (upper body) 18-39y: 35.89°C ≥40y: 35.81°C	Not specifically evaluated	Nonpainful site (upper body) 18-39y: 19.16°C ≥40y: 20.58°C	Nonpainful site (upper body) 18-39y: 40.46°C ≥40y: 41.62°C	Nonpainful site (upper body) 80% reported pain	QST was safe 1 subject reported normal responses to mechanical and thermal stimuli at all sites tested 44% of subjects had abnormal mechanical pain sensitivity at all 3 test sites 72% of subjects had abnormal cold pain sensitivity at all 3 test sites 76% of subjects had abnormal heat pain sensitivity at all 3 test sites Decision tree created to classify patients as having central sensitization (60%), peripheral sensitization (4%) or mixed pain (32%) (see reference for details)
		SCD All genotypes ≥18 years No control group (used published control data)	Detection Thresholds Cold (°C) Heat (°C) Mechanical (g) Thermal: Pain Thresholds Cold (°C) Heat (°C) Mechanical (g)			Painful sites upper body lower body Site chosen randomly by software but specific sites not reported	Painful site (upper body) 18-39y: 29.14°C ≥40y: 29.87°C Painful site (lower body) 18-39y: 29.54°C ≥40y: 28.78°C	Painful site (upper body) 18-39y: 35.38°C ≥40y: 36.09°C Painful site (lower body) 18-39y: 36.58°C ≥40y: 38.11°C	Not specifically evaluated	Painful site (upper body) 18-39y: 18.82°C ≥40y: 20.75°C Painful site (lower body) 18-39y: 22.84°C ≥40y: 23.23°C	Painful site (upper body) 18-39y: 40.48°C ≥40y: 42.71°C Painful site (lower body) 18-39y: 41°C ≥40y: 42.87°C	Painful site (upper body) 58% (11/19) with 1 or more painful sites in upper body reported pain Painful site (lower body) 86% (12/14) with 1 or more painful sites in lower body reported pain
Hollins et al. J Pain Symptom Manage 2012	Research Questions: “Do some SCD patients show signs of central sensitization or other disturbances of pain processing?” “Is central sensitization more likely to occur during a pain episode than between episodes, showing that it is dependent on short-term, transient factors?” “..is central sensitization related to long-term factors, such as a patient's age or a history of intense pain episodes?”	SCD 2 points in time baseline during pain episode (if treated as outpatient only) Controls Cross-sectional at baseline	Mechanical Local dull pressure using Forgione-Barber stimulator	Data compared between: patients with SCD and controls Patients with SCD at baseline and during pain Impact of age and gender on outcomes assessed in both groups Impact of prior pain history on outcomes assessed in SCD group	SCD n=22 age: 32.6 yrs 59% female Controls n=52 age: 29.7 yrs 56% female	Dorsal surface of finger: middle phalanx of index or middle finger	Not tested	Not tested	Not tested	Not tested	Not tested	Specific threshold not determined- see comments section for main study results	Patients with SCD were not different than controls in temporal summation measures Patients with SCD in acute pain were not different in temporal summation outcome measures compared to their baseline data Female gender was significantly associated with more intense and more rapidly increasing pain in both groups Older age was significantly predictive of more rapidly increased pain intensity and unpleasantness In SCD group, the intensity pain ratings of prior pain episodes significantly predicted the rate of pain unpleasantness (i.e. “pain builds up rapidly”)
		SCD All genotypes ≥18 years Discrete episodes of SCD pain with relatively pain-free intervals Controls African American ≥18 years	Temporal Summation% Varying pressure applied continuously and subject gives rating of pain intensity on 0-10 scale and pain unpleasantness on 0-10 scale at 20 second intervals for 9 ratings Temporal Summation Outcomes: Intensity of pain Intensity of unpleasantness Rate of increase in pain Rate of increase in unpleasantness

^ˆMedoc Thermal Sensory Analyzer (TSA-II);

^*Method of limits: Stimulus applied and subject asked to report when stimulus is first detected (detection threshold) and when stimulus is painful (pain threshold);

^#Method of levels: Stimulus applied and subject asked whether stimulus is painful and if answers yes, subject asked to rate pain intensity on scale of 0-10 or with visual analog scale (VAS);

^%Increase in the intensity and unpleasantness of pain that is provoked by a noxious stimulus that is constantly applied

Epidemiological Study

One epidemiological study focused on the assessment of neuropathic pain utilized a large database, the Pediatric Health Information System (PHIS) (Table IV).[18] This study sought to investigate the use of neuropathic pain drugs in hospitalized children with SCD and to determine patient-level factors associated with the use of these drugs. Data revealed that only 2.9% of patients with a diagnosis of SCD were prescribed a neuropathic pain drug.[18] Those that were prescribed these drugs were significantly more likely to be older, female and have a longer length of hospital stay[18] further defining a cohort of patients that may be at risk for the development of neuropathic pain.

Table IV. Epidemiological Study Investigating Neuropathic Pain in Patients with Sickle Cell Disease.

Study	Primary Objectives and/or Hypotheses	Study Design	Inclusion Criteria	Primary Outcomes	Demographics	Results	Conclusions
Brandow et al. J Pediatr Hematol Oncol2015	Objectives “describe the use of neuropathic pain drugs in children with SCD” “determine patient-level factors associated with the use of these drugs” “determine the association between the use of neuropathic drugs and length of hospital stay” Hypothesis “older age and female sex are associated with increased use of neuropathic drug use and the use of these drugs is associated with longer length of stay”	Retrospective cohort Data Source: Pediatric Health Information System (PHIS) database (2004-09)	0-18 yrs Inpatient visit with any SCD discharge diagnosis as per ICD-9 code	“receipt of neuropathic pain drug during hospital admission including the total number of neuropathic pain drugs received “the association between neuropathic drug use and age, sex, and length of hospital stay”	53,557 patient visits identified Mean (SD) age: 9.6 (5.7) yrs 49.1% female	2.9% of patient visits had ≥1 neuropathic pain drug prescribed Odds of receiving a neuropathic pain drug increased significantly with older age (reference group, 0 to 4 y: 5 to 10, odds ratio [OR], 5.7; 11 to 14: OR, 12.5; 15 to 18: OR, 22.8; all P<0.0001) Odds of receiving a neuropathic pain drug increased significantly with female gender (OR, 1.5; p=0.001) Use of neuropathic drugs was associated with longer length of hospital stay (RR 8.3; p<0.0001)	The use of neuropathic pain drugs in children with SCD is associated with: Older age Female gender Longer length of hospital stay

Interventional Study

There was only one study investigating a novel intervention for pain in humans with SCD and characteristics of neuropathic pain (Table V).[19] This was an open-label phase I dose-escalation study that investigated the safety of, trifluoperazine, an antipsychotic that inhibits Ca²⁺calmodulin protein kinase IIα (CAMKIIα inhibitor), in 18 adult patients with SCD.[19] CaMKIIα is found in abundance in the central nervous system and is associated with the development of neuropathic pain.[19] As a phase I study, the primary outcomes were safety and toxicity; however, secondary outcomes included reduction in pain intensity. The investigators had previously shown that trifluoperazine based inhibition of CaMKIIα can reverse neuropathic and inflammatory pain in a rodent model of non-SCD chronic pain.[20-22] Among other inclusion criteria, patients were eligible if they reported chronic pain with ≥4 neuropathic pain descriptors. This study revealed the drug was safe and provided supporting evidence for reduction in pain intensity.[19] Based on these data, the authors conclude that trifluoperazine should be tested in a larger Phase I/II trial as an adjuvant pain treatment in patients with SCD and evidence of neuropathic pain.

Table V. Interventional Study Targeted at the Treatment of Neuropathic Pain in Patients with Sickle Cell Disease.

Study	Primary Objectives and/or Hypotheses	Study Design	Inclusion Criteria	Intervention	Primary Outcomes	Demographics	Results	Conclusions
Molokie et al. Eur J Pharmacol 2014	“to determine safety and potential pain relief effect in adults with SCD and well characterized pain”	Phase I Open-label Repeated measures dose escalation	Main criteria: ≥18 years HbSS genotype Pain ≥3 (0-10 scale) from SCD Chronic pain described with ≥4 neuropathic pain descriptors	Study drug: Trifluoperazine 6 doses: 0.5, 1.2, 5, 7.5, 10 mg	Safety: extrapyramidal symptom rating scale, adverse effects checklist Pain intensity: visual analog scale (VAS)	n=18 Mean (SD) age: 35.8 (8.9) yrs 83% female	Safety: 10mg found to be minimum toxic dose (sedation, dystonia) Sedation: most common severe adverse effect, other effects were mild Pain: 44% (n=8) of subjects had ≥50% pain reduction from baseline without other analgesics other than study drug	Drug has good safety profile Evidence of reduction of pain intensity Preliminary evidence to support larger Phase I/II trial for patients with SCD and evidence of neuropathic pain

Ongoing Research

A search of ClinicalTrials.gov for ongoing research studies pertaining to the topic revealed three interventional trials and one mechanistic trial. Two of the interventional trials are focused on the investigation of acute pain and one is focused on chronic pain. The first study is a Phase II double-blind placebo-controlled clinical trial investigating the effect of gabapentin (Neurontin) in addition to standard pain management on the outcomes of acute pain in children with SCD treated in the acute care setting.[23] The second study is a Phase II trial investigating the efficacy and safety of topical lidocaine in children ages 6-18 years admitted with an acute pain event that have failed standard pain treatment in France.[24] This study also includes pediatric patients without SCD who suffer from neuropathic pain. The only study focused on chronic pain is investigating the role of vaporized cannabis for chronic pain in adults with SCD.[25] The study will evaluate the effect of cannabis alone, potential synergism between cannabis and opioids, side-effects associated with use of cannabis and opioids together, and anti-inflammatory effects of cannabis in SCD.[25] One of the co-investigators on this trial has published basic pain research that interrogated the cannabinoid receptor in the sickle cell murine model as outlined above.[4] These pre-clinical data provide justification for this translational work. In addition to the interventional trials, there is one registered mechanistic trial in patients with SCD. This study is utilizing functional imaging through EEG and fMRI to evaluate pain in adults with SCD.[26] This study uses heat to cause pain with subsequent measurements of pain processing in the brain via EEG and fMRI studies.[26]

Conclusions and Future Directions

In conclusion, this systematic review provides the evidence to support the productivity in the last 5 years that has led to an increased understanding of the biology of pain in SCD. Epidemiological, basic science and clinical data support the existence of nervous system abnormalities that could contribute to sickle cell pain. Importantly, parallel work in the murine model and humans reveals congruent findings of both heat and cold hypersensitivity in SCD supporting potential abnormalities in the central and/or peripheral nervous system that could explain the neurobiology of SCD pain. Ongoing work is attempting to elucidate the underlying mechanism of these initial findings. Furthermore, these parallel data also underscore that a true translational model exists in which to study pain in SCD. Patient-reported outcome measures have also documented the presence of neuropathic pain characteristics in patients with SCD using validated neuropathic pain screening measures. These data support the potential use of these measures to phenotype pain in order to identify patients for inclusion in pain clinical trials. These measures could also be used as outcomes in clinical trials aimed at treating neuropathic pain in patients with SCD. Importantly, research and hypotheses from other pain conditions have been applied to SCD which has led to the generation of novel hypotheses that are under active investigation. There has been scientific interaction among multiple different disciplines (i.e., hematologists, neurobiologists, clinical pain investigators, immunologists) that has led to interdisciplinary research that further strengthens scientific discovery. Continued clinical/translational research done in close parallel to basic research is vital to replicate or refute findings in the murine model and decrease the lag time for laboratory discoveries to reach patients. In conclusion, continued interdisciplinary approach to investigation is imperative to move the field of sickle cell pain forward with the ultimate hope of translating these discoveries into novel ways to both assess and treat pain in order to decrease the insurmountable suffering that patients with SCD experience.