Perception Threshold Variations of Pain Area with Herpetic Ophthalmic Neuralgia

1Department of Rehabilitation Medicine, Shanghai Tenth People's Hospital, Affi liated Tenth People's Hospital of Tongji University, Shanghai, China 2Department of Rehabilitation Medicine, the First People's Hospital of Taicang City, Taicang Affi liated Hospital of Soochow University, Suzhou, China *Address for Correspondence: Gang Xu, Department of Rehabilitation Medicine, Shanghai Tenth People's Hospital, Affi liated Tenth People's Hospital of Tongji University, Room 4305, Outpatient Building, 301 Middle Yanchang Road, Shanghai, 200072, China. Tel: +86-21-66306496; Fax: +86-21-64940427; E-mail:


INTRODUCTION
Herpes Zoster (HZ) is a common, painful and debilitating condition caused by a reactivation of the Varicella-Zoster Virus (VZV) from a latent infection of sensory ganglia [1,2]. Among the cranial nerves aff ected by VZV, Herpes Zoster Ophthalmicus (HZO) which originated from the ophthalmic division (V1) of the trigeminal nerve (V) [3][4][5], is the most common and accounts for 10-25% of all HZ infection [6,7], followed by thoracic dermatomes. Herpetic Ophthalmic Neuralgia (HON), a persistent painful condition that can become chronic, is one of the most common complications of HZO [5,8]. HON is likely to be due to nociceptive response to local infl ammation and tissue damage stimulation, which are manifested neurologically as pain [9]. Symptoms are non-specifi c and range from itching to an intense burning sensation [10]. Consequently, V1 is one of the most common sites of Postherpetic Neuralgia (PHN) in the elderly [3,11], and this condition can severely impact the Quality Of Life (QoL) in these patients [12].
Although the development of PHN would most likely be neural injury [13], no previous study has provided clinical evidence of the severity of neural damage. Th e persistence of local subcutaneous pain aft er the onset of vesicles implied that the peripheral mechanisms of cutaneous innervations, such as the sub epidermal nerve plexus, epidermal nerve fi bers and terminals, played an important role in the pathogenesis of HON [7,14], but little is known about how the phenotypic and functional changes of the aff ected pain area occur.
A sensitive, reliable, and quantitative test of V1 function in HON may be extremely essential to broaden our knowledge of the underlying pain-generating mechanisms, diagnosing HON and monitoring of therapeutic outcomes. Nerve conduction techniques allow direct measurement of the sensory nerve action potential and sensory conduction velocity, thus providing a quantitative measure of the functioning large fi bers in the nerve. V1 is a purely sensory nerve that has a large concentration of unmyelinated fi bers and consequently, a lower A/C fi ber ratio when compared with V2 and V3 [15]. Th is ratio of fi bers makes it diffi cult to evaluate the ophthalmic neurosensory functions by sensory conduction velocity. Von Frey's hairs only have moderate sensitivity and specifi city for diagnosis of neurophysiological and clinical neuropathy [16]. Quantitative Sensory Testing (QST) relies on analysis of an individual's response to external stimuli, refl ecting the integrity of the peripheral nervous system and the sensory pathway [17]. Th e sensory pathway cannot be assessed in isolation from the aff ective and cognitive characteristics of patients or testers. Many variables potentially aff ect the reliability and reproducibility of QST, which aft er all, is designed for the testing of individuals by other individuals. Relative comparisons across body regions do not off er advantages over absolute reference values. Th ere is little report to detect V1 area neurological function. Measuring the Current Perception Th reshold (CPT) of sensory nerves performed by the neurometer device could obtain quantifi ed sensory nerve function [18,19]. Th is test reportedly achieves diff erential neuroexcitatory eff ects by using diff erent frequencies of an electrical sine wave stimulus. CPT testing has been validated and well-documented in peripheral neuropathy resulting from several diff erent etiologies [20][21][22]. However, studies utilizing CPT testing for evaluation of cutaneous innervations in HON patients are limited [23]. Th e aims of the present study were to measure the aff ected sub epidermal nerve sensation with a combination of CPT and Semmes-Weinstein Monofi lament (SWM) testing to detect sensory function changes in the cutaneous innervations under pain areas with HON.

METHODS
Th e study was conducted in accordance with the Helsinki Declaration and was approved by the Research Ethics Board of Shanghai Tenth People's Hospital (2014-49). Th is research was performed from March 2014 to December 2016 in the Department of Rehabilitation Medicine. Using a statistical sampling formula with α = 0.05, β =0.2, p 1 = 0.86 and p 2 = 0.96, the number of samples was 150 people who were suff ering from HON.

Participants
All subjects were at least 50 years of age, and the sample consisted of healthy volunteers and patients with HON who were recruited from the Shanghai Tenth People's Hospital. Th e reference standard for the diagnosis of HON was Dworkin`s diagnostic criteria [12]. Th e subjects were suff ering cutaneous and/or subcutaneous pain in the unilateral V1 region temporally and spatially associated with their rash; had a worst pain score of four or higher on an 11-point pain intensity Numerical Rating Scale (NRS) in the past 24 hours; and had pain that persisted more than 15 days aft er rash onset. Topical agents or/and oral medications, including antidepressants, antiepileptics, and antiarrhythmics were discontinued 10 days before the initiation of the study. Exclusion criteria were the subjects (including the volunteers) having a history of neurological or dermatological disease other than HON, cognitive impairment, or systemic disease, such as diabetes mellitus or renal failure. Th e subjects with prodromal pain preceding the appearance of a rash were also excluded.

Study protocol overview
Aft er obtaining Institutional Review Board Approval, one hundred fi ft y-fi ve subjects with pain and 30 aged-matched healthy volunteers were enrolled in the study. All subjects who fulfi lled the inclusion criteria, provided written informed consent. Th e subjects with HON were divided into fi ve groups by pain duration. Th e HON-30 group had pain duration more than 15 days and less than 30 days, the HON-60 group had pain more than 30 days and less than 60 days, the HON-90 group had pain no more than 90 days, the HON-120 group had pain no more than 120 days, and the HON > 120 group had pain more than 120 days aft er rash onset. All subjects underwent a structured interview to assess the clinical and phenomenological characteristics of their pain: descriptive characteristics and pain intensity on the NRS.
Th e study was conducted in a quiet room to encourage the subject to focus on the test. Th e temperature of the testing room was maintained at 20℃-22℃, and the humidity was approximately 60%. Light and heat were blocked and indoor airfl ow was maintained constantly. Th e procedure was explained to the patient before the test. All testing was performed with subjects in stationary positions (supine) to reduce the possibility of position interference while being tested. One day before examination, patients were asked to refrain from physical therapy, thermotherapy, and electro-physiologic study.

Clinical examination
All subjects underwent a general and neurological examination. Sensory disturbances and pain were carefully assessed. Before testing, all patients were interviewed and asked to indicate the severity of their current pain, to rate the pain intensity on an 11-point NRS ranging from 0 (no pain) to 10 (worst possible pain) and to describe their predominant pain in one of the following categories: constant pain, deep pain, burning pain, paroxysmal pain, lancinating pain, allodynia, paresthesia, dysesthesia, hyperalgesia, and itching [12,24].

SWM Testing
Th e SWM test using Touch Test Sensory Evaluators (Anaesthesio® Dr. med. Martin Bloch, Düsseldorf, Germany) examines the cutaneous pressure threshold by a reproducible buckling stress fi lament that exerts a constant force onto the supraorbital area for 3-5 seconds. Monofi lament testing was performed by the third author, who had received the necessary training on the use of monofi laments. Th e monofi lament was pressed on the supraorbital area as far as the monofi lament could be bent until the subject responded to the stimulus, while the subject's eyes were closed. At each point, the test was repeated three times. Th e threshold was defi ned as the lightest fi lament that the subject responded to correctly in at least two of the three trials. Th e fi laments are labeled with a numerical marking, which is a log base ten of the force in tenths of milligrams [25].

Neurophysiological testing procedures
Th e neurometer (Neurotron, Baltimore, USA) electrodiagnostic testing device uses a standardized, automated procedure [26.27]. CPT testing was performed by the second author, who had received the necessary training on the use of neurometer. Aft er skin preparation with an alcohol swab, the paired electrode was attached to the test site with a 5-mm distance between the outer rims of each electrode. Stimulations of a constant sinusoidal electrical current with 3 diff erent frequencies (5 Hz, 250 Hz and 2,000Hz) were delivered to the test site. At each frequency, the current was slowly increased from 0.01 mA until the subject reported a sensation and pressed a button. Th e test sites of volunteers were at the bilateral supraorbital area. As a Mirro r Site (MS), the fi rst site was placed contralateral to and at the level of the supraorbital area to determine a reference value for each individual. Th e Painful Site (PS) at HZ-aff ected supraorbital area was marked for quantitative tests of sensory function, avoiding open lesions. To initially familiarize the procedure, the test was performed on the MS to the PS in the subj e cts with HON. At each point, the test was repeated two times.

Statistical analysis
An unpaired (2-tailed) t test was used to evaluate diff erences of the mean for SWM and CPT values between the female group and male group. Paired t tests were used to test the diff erence between the mean for SWM and CPT values obtained for diff erent sides (leftright). Receiver Operating Characteristic (ROC) analysis was used to evaluate the predictive value of each biomarker separately for the disease progression of HON. Th e ROC curve was used to illustrate the sensitivity versus 1-specifi city of SWM and CPT values along a sequence of cut-points with both healthy subjects and subjects with HON. Th e highest Area Under Th e Curve (AUC) and Youden index were used to select the cut-off value of the biomarker's measurement. Th e median and Inter-Quartile Ranges (IQR) were calculated for each item, the Kruskal-Wallis H test was performed to compare the PS/ MS ratio among groups and the Dunn-Bonferroni`s tests were used for post-hoc comparisons. P-values less than 0.05 were considered statistically signifi cant. All tests were performed with R soft ware.

Variability of the SWM and CPTs
Th e demographic characteristics of the study population are shown in table 1. For each volunteer, the SWM and CPTs were tested from both the left and right sides. Th e means and 95% Confi dence Intervals (CI) of the SWM and CPTs are shown in table 2. Th e fi rst step was to investigate whether there were diff erences in the SWM and CPTs related to gender and the diff erent sides in which the distributions exhibited Gaussian distributions. Th e descriptive values and results are shown in table 2. Th ere were no signifi cant diff erences when comparing the SWM and CPTs in men and women. Th e diff erences for the diff erent (left -right) sides were not statistically signifi cant in the SWM and CPT readings at each frequency according to paired t tests.
Although the SWM and CPTs values had high standard deviation in the healthy controls, no signifi cant diff erences were found between the diff erent gender or left -right side variables. Th e variables of the SWM and CPTs accompanied by changes in pain duration, intensity and features of patients with HON were investigated.

The predicting values for HON
Based on the clinical observations and pain scores, 155 subjects with HON (79 men and 76 women; mean age: 64.4 ± 6.5 years; range: 50-75 years; 93 subjects aff ected on the right side; 153 subjects were primary school or higher educational level) were selected to measure their SWM and CPTs at pain sites. Th e results of ROC analysis for each of the SWM and CPTs are summarized in fi gure 1 and table 3.
Th e SWM and CPTs indicated signifi cant predictive values for healthy controls compared to subjects with HON (AUC > 0.75). Among the 4 ROCs, CPT 2000Hz was identifi ed for predicating healthy controls to HON, and the AUC of the CPT 2000Hz was 0.93 with both high sensitivity (81.9%) and specifi city (96.7%), and the corresponding cut-off value was 199. In contrast, the SWM had lowest sensitivity (80.6%) and specifi city (73.3%), and the corresponding cut-off value was 0.135.
Based on a normal distribution of the SWM and CPTs of healthy controls, the average values and the upper and lower normal reference limits (defi ned as +/-standard deviations from the mean) for each testing were calculated [28]. It showed that these cut-off values for the CPTs and SWM were approximately equal with the mean + standard deviation, which were defi ned as the upper reference limits, except for CPT 2000Hz . Combined lower reference limits with the cut-off values were then used as a screening hyperesthesia or hypoesthesia threshold for local sensory malfunction of HON. Table 3 showed the hyperesthesia, normal or hypoesthesia distributions of 155 subjects with HON. More than eighty percent of the HON subjects had either SWM or CPT hypoesthesia at any frequency in the pain area. In contrast, only 0.64-5.16% of subjects with HON had either SWM or CPT hyperesthesia.

Variations in the SWM and CPTs with pain in HON
One hundred fi ft y-fi ve patients had a clinically manifested sensory impairment in the aff ected areas involving every sensory modality. Most of patients also had abnormal SWM and CPTs. Th e Shapiro-Wilk test for the SWM and CPTs of patients showed positively skewed distributions, and the median (IQR) of the SWM and CPTs were calculated. We used the PS/MS ratio of the SWM and CPTs to avoid the potential impact of individual diff erences on intraindividual comparisons.
According to the length of time pain lasted days aft er rash onset, subjects with HON were labeled HON-30, HON-60, HON-90, HON-120 and HON>120 days. Th e PS/MS ratio of the CPTs and SWM were compared among the diff erent pain duration groups. A Kruskal-Wallis H test showed no signifi cant diff erences in the SWM and CPTs among diff erent pain duration groups.
Th e worst pain scores reported by subjects with HON ranged from 4 to 10 on the NRS (mean 7.3). Th e PS/MS ratio of the CPTs and SWM were compared among the diff erent pain intensity groups. A Kruskal-Wallis H test showed no signifi cant diff erences in the SWM and CPTs among the diff erent pain intensity groups.
Th e PS/MS ratio of the CPTs and SWM were compared among the diff erent pain characteristic groups, and signifi cant diff erences in CPT values for 250Hz and 2,000Hz frequencies among the diff erent groups were observed (shown in fi gure 2, p < 0.05). Post hoc tests showed signifi cant diff erence in CPT 250Hz between itching and lancinating pain (p= 0.006), itching and allodynia (p = 0.002),   and burning pain and allodynia (p = 0.042). Similarly, the CPT 2000Hz between deep pain and allodynia (p = 0.001) and itching and allodyni a (p = 0.014) were signifi cantly diff erent.

DISCUSSION
Th ere was a strong correlation between the incidence of HZ and increasing age, with a marked rise in incidence at the age of 50-60 years and older [29]. We thus selected the age 50 years as a point for comparison against the HZ and HON populations. Previous reports using a variety of stimulation techniques suggested diff erences in sensory thresholds between genders were related to diff erences in skin fold thickness and epidermal nerve fi ber denvsity [30,31]. Our results suggested there were no signifi cant diff erences when comparing the SWM and CPTs in the supraorbital region in men and women or the right and left sides.
HON in the V1 distribution has a complex pathophysiology that  begins with viral damage and increased sensitization of peripheral sensory fi bers [4,32]. Th e persistence of local pain aft er the onset of vesicles implies that the damaged V1 branch play an important role in the pathogenesis of herpetic neuralgia [33]. However, accurately ascertaining these conditions is diffi cult, and quantifying the severity of pain and sensory disturbance has generally been considered impossible. Th e development of new methods for assessing neuropathic pain has provided an improved foundation for the diagnosis and assessment of herpetic neuralgia. During the past three decades, the CPTs have been reported to provide reliable measures of peripheral sensory nerve function from the large and small myelinated and unmyelinated nerve fi bers using a neurometer [34,35]. As the worst pain associated with HZ are the partial lesion regions innervated by distal nerve branches instead of the entire nerve distribution area [12], the local SWM and CPT changes may be refl ected in the functions of supraorbital nerve fi bers with HON.
Although it is diffi cult to prove the CPTs could selectively measure the functional integrity of the large myelinated, small myelinated and unmyelinated C-fi bers, CPT evaluation has the unique ability to detect both hyperesthesia (abnormally low electrical excitation thresholds) as well as hypoesthesia (abnormally elevated thresholds) [36]. In the present study, we evaluated the SWM and CPT values in 155 patients with HON. Using the means and standard deviations from normative data and the cut-off values from ROCs, the SWM and CPTs provided high sensitivity and specifi city for detecting the local neural function of HON. More than eighty percent of subjects with HON had either SWM or CPT hypoesthesia, and 0.64-5.16% subjects with HON had hyperesthesia at any frequency in the pain area. Assessment of the functions of these nerve fi bers involved in acute and chronic pain, is clinically most challenging for patients with HON, thus contributing greatly to improved diagnosis and treatment. However, the results of SWM and CPT testing did not correlate with pain duration or pain intensity, which implied these lower reference limits and cut-off values had little relation to underlying mechanisms.
Most patients with HON describe multiple types of pain [12]. Th e characteristics of the pain and intensity were obviously due to diff erent pathophysiological mechanisms. Th us, the diagnosis and evaluation of pain on the basis of local sensory abnormalities might be more appropriate. Th e present study showed that the PS/MS ratio of the CPTs and SWM, when comparing diff erent pain characteristic groups, were signifi cantly diff erent among the groups in the CPT values when using 250Hz and 2,000Hz frequencies (p < 0.05). Post hoc tests showed signifi cant diff erences in CPT 250Hz between itching and lancinating pain (p = 0.006), itching and allodynia (p = 0.002), and burning pain and allodynia (p = 0.042). Additionally, there were signifi cant diff erences in CPT 2000Hz between deep pain and allodynia (p = 0.001) and itching and allodynia (p = 0.014). Th ese results implied that CPT 250Hz or CPT 2000Hz of allodynia were diff erent from those of itching, burning pain or deep pain, which indicated that the neurometer can be used as an auxiliary diagnostic technique for allodynia. Our results are supported by the fi ndings of Ohsawa and Raj, et al. [37,38] Th e results of the SWM had relatively low sensitivity to identify the pain characteristics.
Th e present study showed that combined SWM and CPT testing could be used to assess local sensory function with HON, which is a less invasive test that can be performed anywhere on the body in a simple manner. Th e test can be performed without any limitations in location, and it is a useful test for evaluating the local sensory fi ber functional changes with HON.

FUNDING SOURCES
Th is work was supported by the Shanghai Science and Technology Committee (16401934900) and Shanghai Municipal Commission of Health and Family Planning (20134320). Th e sponsorship was limited to supplies and expenses. Th e fi nancial sponsor of this work had no role in the design and conduct of the study or the collection, management, analysis, interpretation of the data and preparation or review of the manuscript or the decision to submit. Th e remaining authors have no confl icts of interest.