Neuropathic

Research Paper

Efficient conditioned pain modulation despite pain

persistence in painful diabetic neuropathy
a,d, a c b
Yelena Granovsky *, Hadas Nahman-Averbuch , Mogher Khamaisi , Michal Granot

Abstract
Introduction: Alleviation of pain, by either medical or surgical therapy, is accompanied by transition from less efficient, or
pro-nociceptive, to efficient conditioned pain modulation (CPM). Spontaneous decrease or resolution of pain with disease
progression is reported for some patients with painful diabetic neuropathy (PDN).
Objectives: To explore whether CPM changes similarly in parallel to spontaneous resolution of pain in PDN patients.
Methods: In this cross-sectional study, thirty-three patients with PDN underwent psychophysical assessment of pain
modulation on the forearm, remote from the clinical pain.
Results: Pain duration was not correlated with neuropathic pain intensity, yet, it correlated with CPM efficiency; patients
with longer pain duration had same pain level, but more efficient CPM than those with short-pain duration (r 5 20.417; P 5
0.025, Spearman correlation). Patients with pain more than 2 years (median split) expressed efficient CPM that was not
different from that of healthy controls. These patients also had lower temporal summation of pain than the short-pain
duration patients group (P , 0.05). The 2 patient groups did not differ in clinical pain characteristics or use of analgesics.
Conclusion: Pro-nociception, expressed by less efficient CPM and high temporal summation that usually accompanies clinical
painful conditions, seems to “normalize” with chronicity of the pain syndrome. This is despite continuing pain, suggesting that pro-
nociceptivity in pain syndromes is multifactorial. Because the pain modulation profile affects success of therapy, this suggests that
different drugs might express different efficacy pending on duration of the pain in patients with PDN.
Keywords: Painful diabetic neuropathy, Pain modulation, Pain duration, Pain intensity, Conditioned pain modulation, Temporal
summation

1. Introduction An interesting feature of the neuropathic pain of diabetic origin

relates to the reports of spontaneous pain improvement or
Neuropathic pain is a common complication of diabetes 3,4,40
characterized by burning pain on the affected limbs. Similar to resolution with disease progression. Studies on surgical and
other pain states, patients with painful diabetic neuropathy (PDN) medication treatments for pain have shown that the pain
demonstrate enhanced activity of facilitatory pain pathways
26 alleviation is accompanied by improvement of pro-nociceptivity,
25 with transition from less efficient to efficient conditioned pain
and deficient activity of pain inhibitory pathways, pointing thus 6,13,18,25,49
to a pro-nociceptive pattern of central pain modulation. modulation (CPM). However, it is unknown whether
the spontaneous alleviation of neuropathic pain along with PDN
duration is similarly associated with reversal of pro-nociceptivity.
Sponsorships or competing interests that may be relevant to content are The aim of this study was to evaluate pain modulation in PDN
disclosed at the end of this article. of different pain duration and its association with pain levels. We
a The Laboratory of Clinical Neurophysiology, Technion Faculty of Medicine, Haifa, hypothesized that longer neuropathic pain duration would be
b c
Israel, Faculty of Health and Welfare Sciences, University of Haifa, Haifa, Israel, associated with lower clinical pain levels and reversal of the pro-
Internal Medicine D and Institute of Endocrinology, Diabetes and Metabolism,
Rambam Health Care Campus, Technion Faculty of Medicine, Haifa, Israel, nociceptive characteristics of pain modulatory system.
d
Department of Neurology, Rambam Health Care Campus, Haifa, Israel
*Corresponding author. Address: Department of Neurology, Rambam Health Care
Campus, Haifa, Israel. Tel.: 197248542065; fax: 197248542755. E-mail address: 2. Methods
[email protected] (Y. Granovsky).
2.1. Subjects
Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on
behalf of The International Association for the Study of Pain. This is an open Thirty-five patients with PDN participated in this study. The
access article distributed under the Creative Commons Attribution License 4.0 diagnosis of PDN was confirmed by a qualified neurologist or
(CCBY), which permits unrestricted use, distribution, and reproduction in any endocrinologist with special training in diabetes care. This article
medium, provided the original work is properly cited.
reports the results of baseline (pretreatment) assessment of the
PR9 2 (2017) e592 patients cohort who then underwent treatment with duloxetine
http://dx.doi.org/10.1097/PR9.0000000000000592 49
(reported at Yarnitsky et al, 2012 ). Patients met the following

2 (2017) e592 www.painreportsonline.com 1

2 Y. Granovsky et al. ·
2 (2017) e592 PAIN Reports®

inclusion criteria: (1) lower limb distal neuropathic pain for more because genuine assessment of the central effect depends
than 3 months and (2) mean pain severity during the last month on intactness of the peripheral conduction of neural signals.
of $40 on a 0 (“no pain at all”) to 100 (“most painful imaginable Mechanical temporal summation of pain (TS)—the stimuli
sensation”) numerical pain scale (NPS). The exclusion criteria were delivered with an 180gr (# 6.45) von Frey filament applied
were as follows: (1) intake of a monoamine oxidase inhibitor to the volar forearm. Subjects were exposed to a series of 10
(MAO-I) within the last 14 days, (2) narrow-angle glaucoma, (3) stimuli with an interstimulus interval of 1 sec, applied within an
inability to undergo psychophysical testing, (4) pain complaints at area of 1 cm in diameter. Subjects were asked to rate the level of
the upper limbs, and (5) chemotherapy for cancer and known pinprick pain intensity using verbal NPS for the first and the tenth
vitamin B12 deficiency. Patients who were on regular use of stimuli. The difference between the tenth and the first pain
serotonin–noradrenaline reuptake inhibitors (SNRIs) were tested scores served as a score of the TS.
after 1 month drug washout. Conditioned pain modulation was assessed using the parallel
Data from healthy subjects (N 5 29, 17 women) collected paradigm in which 2 identical test stimuli were given; once before
under same psychophysical protocol as for patients with PDN and then simultaneously with a noxious conditioning stimulus. The
served as the control group. The inclusion criteria for the test stimulus was a tonic noxious contact heat stimulus applied to the
control group were as follows: (1) absence of chronic pain volar aspect of the dominant forearm using TSA. The intensity of the
history; (2) no use of analgesic or psychiatric medication on a test stimulus was predetermined individually based on the
regular basis; (3) age between 40 and 80. All participants psychophysical parameter of Pain60 temperature. This method was
were able to communicate and understand the instructions of based on delivery of several triplets of 7-sec long stimuli of various
the study and were asked to refrain from any pain relief intensities; the closest temperature that induced pain at a level of 60
10
medications 24 hours before the experimental trial. The study on a 0-100 NPS was considered as the Pain60 temperature. The
was approved by the local ethics committee, and informed baseline temperature was 32˚C, which in-creased at a rate of 2˚C/s
consent was obtained from all participants. to the destination temperature, which was held for 30 seconds, and
decreased back to baseline at the same rate. Subjects were asked to
rate the intensity of the test stimulus along the stimulus at the 10th,
2.2. Experimental protocol 20th, and 30th second. The mean pain score served as the pain level
All experiments were conducted in the same setting by the of “test stimulus.” After a 15-minute break, subjects were exposed to
same experimenter (H.N.-A.) at the Laboratory of Clinical the conditioning stimulus by immersion of the nondominant hand into
Neurophys-iology in Rambam Health Care Campus, Haifa, a hot water bath at 46.5˚C (Heto CBN 8-30 Lab equipment, Heto-
Israel. Subjects sat comfortably in a quiet room with an Holten A/S; Allerod, Denmark) up to the wrist for 1 minute. Water-
ambient room temperature of 22˚C to 23˚C. induced pain was rated every 10 seconds during the first 30 seconds
of immersion; the mean score of the 3 ratings of the immersed hand
served as the pain level of the “conditioning stimulus.” During the last
2.2.1. Clinical assessment 30 seconds of the conditioning stimulation, an identical test stimulus
was repeated, and pain of the test stimulus was rated again every 10
Patients were asked to report duration of (1) diabetes, (2)
seconds. The CPM effect was calculated as the difference between 2
sensory neuropathic symptoms such as, (3) mean and
test-stimuli applications: the conditioned test stimulus minus the
maximal neuro-pathic pain intensity during the last week, and
stand-alone test stimulus, for each single individual pair of stimuli,
pain duration. In addition, the information regarding current or
past pain-relieving drugs consumption was collected. obtained at 10, 20, and 30 seconds of test stimuli (CPM 10, CPM20,
As part of a clinical examination, sensory detection thresholds and CPM30, respectively), and as CPMmean for averaged 3 CPM
were measured at the dorsum of the foot to assess neuropathy values. More negative values indicated more
severity. Warm detection thresholds (WDT) and cold detection
thresholds (CDT) were measured 3 times each using the thermal efficient CPM. This paradigm of conditioned pain stimulation
sensory analyzer (TSA-II) (Medoc, Ramat Yishai, Israel) with a is routinely used in our laboratory.10,23,24,43
2
Peltier 30 330 mm contact thermode according to the method of
45
limits. The 32˚C baseline temperature decreased or in-creased 2.2.3. Questionnaires
at a rate of 1.0˚C/s and was stopped by pressing the response
button at the moment that warm or cold sensation was At the beginning of the experimental session, all participants
perceived. Warm detection threshold and CDT were calculated completed several pain-related psychometric questionnaires for the
as the mean values of the 3 measurements. The mechanical assessment of pain catastrophizing (assessed with Pain
35
detection thresholds (MDT) were measured using von Frey Catastrophizing Scale [PCS]) and anxiety levels (assessed with
filaments (North Coast Medical, San Jose, CA). The weight was 34
Spielberger State-Trait Anxiety Inventory, STAI). The patients with
increased using consecutive filaments until the patient reported PDN also reported their self-perceived health by filling out a Hebrew
that sensation was perceived. Each filament was applied 3 times 20
version of the short-form 36 (SF-36) Health Survey. SF-36 contains
until the filament pressure was detected in at least 2 of the 3 the following subscales: vitality, physical functioning, bodily pain,
trials. The first filament to be detected at $2/3 times was general health perceptions, physical role functioning, emotional role
determined as the MDT. functioning, and social role functioning. Higher scores on each
summary scale agree with better general health.
2.2.2. Psychophysical assessment
All participants, the patients with PDN and the control subjects, 2.3. Statistical analysis
underwent CPM and temporal summation (TS) assessments to assess The primary interest of this study is to characterize the relationship
pain modulation. All tests were delivered to the volar aspect of the between pain duration and pain psychophysics. Preliminary
forearm. This site is less commonly affected by the neuropathy, allowing examination of data indicated that skewed pain duration data
normal or near-normal peripheral sensory function. This is distributions, with some extreme values, would result in
2 (2017) e592 www.painreportsonline.com 3

problematic Pearson correlations, with extreme points having segments; CPM10 (r 5 0.199, P 5 0.302), and CPM 20 (r 5 0.292,
unacceptably high leverage. Therefore, nonparametric Spearman P 5 0.132) or CPM30 (r 5 0.130, P 5 0.517). In line, no correlation
(rank-order) correlations were performed to test overall mono-tonicity was found between pain intensity and TS (r 5 0.051, P 5 0.783).
of relationships between neuropathic pain level and duration and However, looking at the relationships between pain duration and
psychophysical measures. To further examine the effect of pain pain modulation capabilities, we found that patients with longer
duration on pain modulation profile (PMP) character-istics, additional pain duration had (1) increased efficiency of CPM mean (r 5
analyses were conducted by dividing the patients into 2 groups 20.417; P 5 0.025, Fig. 1A); when analyzed by segments along
based on the median score of pain duration. The results of 2 the CPM test, the association between the pain duration and
subgroups of patients were compared with the control subjects with CPM was significant for CPM10 (r 5 20.625, P , 0.001) and
the addition of age as a covariate, in a regression model. The post CPM20 (r 5 20.457, P 5 0.015) but not for CPM30 (r 5 20.120, P 5
hoc analysis was performed to confirm the group differences. The 0.551), (2) a trend toward lower TS magnitude (r 5 20.316, P 5
level of significance was set at P , 0.05. The data are presented as 0.079).
mean (SE) if not indicated otherwise. Statistical analyses were
performed using Microsoft Excel (Microsoft Corporation, Redmond,
WA) and JMP (SAS Institute, Cary, NC). 3.3. Pain modulation in patients grouped by pain duration vs
control subjects
Patients were divided into 2 groups based on the median split of
3. Results the pain duration. Patients with pain duration of 2 years or less
Two patients did not provide full information about their (N 5 20, mean age 60.6 6 11.7, mean 6 SD) were considered as
neuropathic pain duration, so the final sample of patients with short-pain duration group, whereas patients with pain duration
PDN consisted of N 5 33 (7 females). Demographic and longer than 2 years (N 5 13, mean age 60.2 6 6.2, mean 6 SD)
clinical parameters as well as the values of PCS and STAI of were considered as long-pain duration group. The results of 3
the patients with PDN as well as the results of sensory group comparisons for psychophysical variables are presented in
Table 2. Patients with longer pain duration demonstrated lower
quantitative sensory assessment are presented in Table 1.
TS magnitude in comparison with the short-pain duration group.
Patients were significantly older than controls (60.4 6 9.7 vs
Moreover, patients with longer pain duration demonstrated more
53.9 6 6.7 years, mean 6 SD, P 5 0.004); therefore, subjects’
age was added as a covariate into the regression models. efficient CPMmean (Fig. 1B) and the CPM at 10 (CPM10) and 20
seconds (CPM20) of the CPM assessment, as compared with
short-pain duration patients, and were not different of those from
3.1. Pain duration and neuropathy severity the healthy controls.
Patients with longer pain duration had (1) higher duration of
neuropathy (r 5 0.826; P , 0.001); (2) higher severity of the 3.4. Pain-related psychometric variables
neuropathy as measured by higher CDT (r 5 20.366; P 5
0.036) and warm detection threshold (r 5 0.350; P 5 0.046), No correlation was observed between the pain duration and
all tested on feet. Because no consistent responses were any of the pain-related psychophysical parameters; anxiety
obtained to the assessment of mechanical detection trait (r 5 0.055; P 5 0.784), anxiety state (r 5 0.276; P 5
threshold, these data are not reported. It is important that the 0.174), or PCS (r 5 0.126; P 5 0.531). Comparing the long-
pain duration did not correlate with neuropathic pain intensity and short-pain duration PDN patients with the healthy
(r 5 20.154; P 5 0.393) or age (r 5 0.116; P 5 0.522). subjects, no significant group differences were observed for
the scores for anxiety state or trait; the PCS scores were
significantly higher in the patients with PDN, with no
3.2. Pain modulation and neuropathic pain intensity differences between the 2 patient groups (Table 2).
and duration
No significant correlations were found between the neuropathic 3.5. Clinical characteristics of the short- and long-pain
pain intensity and CPMmean (r 5 0.206; P 5 0.285) or its temporal duration painful diabetic neuropathy patient groups
No significant group differences were observed for the clinical
Table 1 pain intensity (63.5 6 17.4, NPS, in short-pain duration groups vs
Demographic and clinical characteristics of patients with painful 61.3 6 20.9 in long-pain duration group, mean 6 SD, P 5 0.754),
diabetic neuropathy (mean 1/2 SD). or for the neuropathy severity in terms of foot warm (42.3 6 4.7
Measure Value vs 45.2 6 4.9, mean 6 SD, P 5 0.112) or cold sensation threshold
Age, y 60.4 6 9.7 (25.8 6 6.5 vs 19.9 6 11.9, mean 6 SD, P 5 0.081) (although a
Male/female 26/7
trend of increase in severity with time can be seen). In addition,
the short- and long-pain duration patients were not different for
Diabetes duration, y 15.1 6 10.0
any component of the SF-36 Health Survey (Table 3).
Neuropathy duration, y 3.5 6 4.1 Among the 33 patients, 17 (52%) were currently using pain
Pain duration, y 4.2 6 6.4 medications (nonsteroidal anti-inflammatory drugs, sSNRIs, or
Clinical pain intensity, NPS 62.1 6 19.3 gabapentinoids). The proportion of patients on pain medications
WDT, ˚C, foot 43.6 6 4.9 was not different between the short (10 patients with and 10
patients without medications) and long (6 and 7 patients,
CDT, ˚C, foot 23.5 6 9.3
respectively) pain duration groups (2-tailed Fisher exact test P 5
MDT, g, foot 20.2 6 72.2 1.00). In addition, the number of used drug families was similar
CDT, cold detection threshold; MDT, mechanical detection threshold; PDN, painful between short- and long-pain duration patients with PDN (1.9 6
diabetic neuropathy; WDT, warm detection threshold.
0.9 vs 1.5 6 0.5, mean 6 SD, P 5 0.317).
4 Y. Granovsky et al. · 2 (2017) e592 PAIN Reports®

Figure 1. (A) Neuropathic pain duration significantly correlated with CPM efficiency, Spearman r 5 20.417; P 5 0.025. The solid line is a spline fit to show the
general monotonicity of the relationship, lambda 5 100. (B) PDN patients with longer pain duration had more efficient CPM than the short-pain duration patients (P ,
0.05), and were not different from the responses from healthy controls. CPM, conditioned pain modulation; PDN, painful diabetic neuropathy.

4. Discussion osteoarthritis.2,32 It is important that subsequent to the pro-

nociception reported during pain, evidence is available to
The main finding of this study is that the association of clinical
show that a shift back toward normal PMP occurs when pain
pain and pronociceptive pain modulation is not maintained in
is alleviated after surgical13,18 or pharmacological
long-standing PDN. Although we did not find longer pain duration
to be associated with lower pain intensities, it was, however, treatment.25,49 However, efficient engagement of descending
associ-ated with more efficient CPM response, as well as less inhibition and decreased activity in pain facilitatory pathways
enhanced TS. In other words, although clinical pain levels seem provide protection against the development of experimental
to persist, CPM and TS seem to “return to normal” along the chronic neuropathic pain models in animals,8,29 and predict
neuropathic pain syndrome. The longer pain-efficient CPM lower incidence and intensity of chronic postoperative pain in
association was unrelated to general functioning, physical or humans.44,47 Thus, a relationship between clinical pain and
emotional health or to analgesics intake. pro-nociceptive pain modulation was established; however,
We have advocated the concept of PMP, which reflects the the “chicken-egg” ques-tion remains open.
state of individual pain modulation in various clinical situations, It should be emphasized that our study is cross sectional, so we
showing the shifts in modulation in accordance with changes in cannot definitively comment on “normalization” of the modulation
46,48 along time in our patients without a preplanned long-length
clinical pain. A central point is that the presence of clinical
pain is usually accompanied by a shift toward pro-nociception, prospective study, but can extend this concept from our findings that
expressed, among other parameters, by a less efficient CPM and CPM in long-pain duration patients was not different from the CPM
enhanced TS. This has been shown for various pain syndromes, responses in healthy subjects. We hypothesize, therefore, that better
21,39,46 CPM in patients with long-pain duration may represent a recovery of
including neuropathic pain. In line, the association
between longer pain duration and pro-nociceptive PMP was the modulation in a “back to normal” fashion. There seems to be an
reported for some nonneuropathic pain states such as effort of the central nervous system pain-controlling
2 (2017) e592 www.painreportsonline.com 5

Table 2
Psychophysical pain parameters and the scores to psychometric questionnaires in the control subjects in the patients with painful diabetic
neuropathy of long- and short-pain duration.
Pain parameters Model P Age P Control group Long-pain duration group Short-pain duration group Group P
Pain60 temperature 0.246 0.589 46.3 6 0.7 45.3 6 1.1 44.4 6 0.9 0.252
Test stimulus, NPS 0.754 0.686 53.7 6 3.4 52.6 6 4.9 49.3 6 4.8 0.704
Conditioning stimulus, NPS 0.166 0.053 68.6 6 5.0 64.9 6 7.7 65.1 6 6.4 0.883
CPM
mean 0.007 0.628 28.6 6 3.4 26.7 6 5.1 10.7 6 4.2*† 0.002
CPM10 0.002 0.600 23.2 6 3.4 21.2 6 5.1 18.8 6 4.3*† ,0.001
CPM20 0.003 0.071 210.8 6 4.1 26.2 6 6.3 14.3 6 5.4*† 0.002
CPM30 0.317 0.501 212.0 6 4.6 212.3 6 7.0 20.1 6 6.1 0.254
Mechanical TS 0.056 0.650 15.6 6 3.7 2.8 6 5.4 21.8 6 4.6† 0.027
Anxiety state 0.139 0.440 29.2 6 2.2 36.8 6 3.4 32.2 6 2.8 0.139
Anxiety trait 0.704 0.463 35.5 6 1.6 37.4 6 2.7 35.7 6 2.2 0.831
PCS ,0.001 0.063 11.9 6 2.2 28.0 6 3.3‡ 28.7 6 2.7* ,0.001
* Refer to pairs of group means (short-pain duration group vs controls) which are significantly different at P , 0.05 (Tukey post hoc test).
† Refer to pairs of group means (short-pain duration group vs long-pain duration group) which are significantly different at P , 0.05 (Tukey post hoc test).
‡ Refer to pairs of group means (long-pain duration group vs controls) which are significantly different at P , 0.05 (Tukey
post hoc test). CPM, conditioned pain modulation; PCS, Pain Catastrophizing Scale; TS, temporal summation.

subsystems to return to equilibrium, even when the peripheral changes, suggesting an adaptation process of the
perturbing factor, pain, is still prevalent. A similar situation is pain processing system.
proposed regarding reduced TS magnitude in patients with Our findings on antinociceptive shift of PMP with no pain
longer duration of neuropathic pain. resolution are challenging. We suggest that central neuroplastic
Neuropathic pain is a result of excessive primary afferent changes that took place during the years of diabetic neuropathy
activity that generates neuronal hyperexcitability in second-order might have contributed to the efficiency of pain inhibitory
15 systems. At the initial stage, the net output of descending pain
spinal neurons and modulates their pattern of function-ing.
Alterations of pain modulatory pathways, with a shift toward a modulation is shifted to overactive pain facilitatory mechanisms,
pro-nociception, were demonstrated in psychophys-ical studies promoting the development and maintenance of persistent
in patients with neuropathic pain. Enhanced TS and neuropathic pain.30,31 As was documented by animal neuro-
less-efficient CPM were reported for various neuropathic pain pathic pain models, inactivation of supraspinal structures
1,24,37
conditions. These studies, however, did not explore the involved in descending facilitation prevents or diminishes the
relationship between the neuropathic pain duration and pain 17,42
evoked secondary hyperalgesia. We may, therefore, suggest
modulation capabilities. It is worth noting that spontaneous that pro-nociceptive features of the PMP in PDN patients with
improvement or even partial or complete resolution of sponta- shorter pain duration relate to the superior activity of pain
neous pain is reported for several studies on patients with facilitatory pathways. In line, we may propose that neuroplastic
3,4,22,40
PDN. The likely cause is loss of nociceptive neurons that alterations occurring at various levels of pain neuromatrix
generate pain along the process of ongoing metabolic and modulate the balance between activity of descend-ing pain
16,27 facilitatory and inhibitory pathways in favor of the latter. These
microvascular changes. Our results did not reveal associ-
ation between longer pain duration and low pain intensity, plastic changes, theoretically, may strengthen the anatomical
perhaps because we only recruited patients reporting pain, so connectivity within the periaqueductal gray–rostral ventromedial
our group was biased. Therefore, despite significant correlation medulla–spinal circuit, and with the subnucleus reticulates
between the duration of pain symptoms and duration of the dorsalis, a part of the spinal-bulbo-spinal loop implicated in
19,41
neuropathy, the efficient pain inhibitory capacity in long-pain diffuse noxious inhibitory control. This possibly explains why
duration patients seems to be determined by central rather than patients with longer pain duration

Table 3
SF-36 Health Survey component in patients with painful diabetic neuropathy of long- and short-pain duration (mean 6 SD).
Component Long-pain duration group Short-pain duration group P
Physical functioning 50.8 6 30.3 53.9 6 36.5 0.747
Role physical functioning 36.5 6 45.2 34.7 6 36.5 0.902
Bodily pain 40.4 6 23.0 42.5 6 24.3 0.811
General health 48.1 6 25.2 44.3 6 17.3 0.624
Vitality 44.3 6 26.8 40.7 6 19.5 0.672
Social functioning 70.2 6 24.8 72.2 6 22.9 0.816
Role emotional functioning 59.0 6 43.4 66.7 6 39.1 0.614
Emotional well-being 45.0 6 16.4 43.5 6 14.6 0.783
PDN, painful diabetic neuropathy; SF-36, short-form 36.
6 Y. Granovsky et al. ·
2 (2017) e592 PAIN Reports®

expressed more efficient CPM. In line, functionally active data analysis and the manuscript preparation. All the authors
descending inhibition on the second-order neurons may also approved the final draft of the manuscript.
contribute to lower TS magnitude and higher pain thresholds.
The functional changes in the pain-modulatory system at the
brainstem level may involve descending inputs from higher brain Acknowledgments
5
structures. Reports on neuropathic pain, including diabetes, The authors thank Dr Elliot Sprecher for his help with the
indicate enhanced activation in sensory and limbic pain–related data analysis.
28,36 7,9
brain structures and changes in cortical thickness. It is
important that neuropathic pain duration was positively Article history:
correlated with increased BOLD signal in some of these Received 26 December 2016
36
structures, and with enhanced cortical thickness in anterior Received in revised form 10 February 2017
7
insula along with thinness in prefrontal cortex. We, therefore, Accepted 9 March 2017
hypothesize that the proposed anatomical and functional
changes in the pain-related brain structures occurring through References
pain duration might be a part of a general long-term
[1] Aasvang EK, Brandsborg B, Christensen B, Jensen TS, Kehlet H.
reorganization of the pain-modulatory system. We further
Neurophysiological characterization of postherniotomy pain. Pain
suggest that this reorganization is more relevant to sensory 2008; 137:173–81.
rather than to affective components of pain perception because [2] Arendt-Nielsen L, Egsgaard LL, Petersen KK, Eskehave TN, Graven-
pain duration was not associated with magnitude of pain-related Nielsen T, Hoeck HC, Simonsen O. A mechanism-based pain sensitivity
psychometrics such as anxiety or pain catastrophizing. index to characterize knee osteoarthritis patients with different disease
stages and pain levels. Eur J Pain 2015;19:1406–17.
Our results shed light on 2 additional aspects of potential
[3] Benbow SJ, Chan AW, Bowsher D, MacFarlane IA, Williams G. A
clinical implication of PMP; first, the significant relationship prospective study of painful symptoms, small-fibre function and
between pain duration and CPM was restricted to first 20 peripheral vascular disease in chronic painful diabetic neuropathy.
seconds of dual stimulation (CPM 10 and CPM20), confirming a Diabet Med 1994;11:17–21.
11 [4] Boulton AJ, Armstrong WD, Scarpello JH, Ward JD. The natural
methodological advantage of shorter CPM assessment. history of painful diabetic neuropathy—a 4-year study. Postgrad Med
Second, less-efficient CPM at early stages of painful neuropathy J 1983;59: 556–9.
may suggest higher efficacy of mechanism-based analgesic [5] Calejesan AA, Kim SJ, Zhuo M. Descending facilitatory modulation
choice (such as serotonin-norepinephrine reuptake inhibitors) of a behavioral nociceptive response by stimulation in the adult rat
during the first years of the neuropathic pain, whereas other anterior cingulate cortex. Eur J Pain 2000;4:83–96.
treatment choices may be preferable for the long-time pain [6] Campbell CM, Buenaver LF, Raja SN, Kiley KB, Swedberg LJ,
Wacnik PW, Cohen SP, Erdek MA, Williams KA, Christo PJ. Dynamic
sufferers. This approach, however, should be confirmed from a pain phenotypes are associated with spinal cord stimulation-induced
study on a large group of patients with PDN. reduction in pain: a repeated measures observational pilot study.
Our main study limitation is the quite small sample, which is a Pain Med 2015;16:1349–60.
general problem in clinical research. In addition, the middle-aged [7] DaSilva AF, Becerra L, Pendse G, Chizh B, Tully S, Borsook D.
Colocalized structural and functional changes in the cortex of
control subjects were significantly younger than the patients with patients with trigeminal neuropathic pain. PLoS One 2008;3:e3396.
PDN. Although age was included as a covariate in all group [8] De Felice M, Sanoja R, Wang R, Vera-Portocarrero L, Oyarzo J,
comparisons, the younger age of control subjects can be King T, Ossipov MH, Vanderah TW, Lai J, Dussor GO, Fields HL,
a potential bias for the results on CPM because of several reports on Price TJ, Porreca F. Engagement of descending inhibition from the
12,14 rostral ventromedial medulla protects against chronic neuropathic
more-efficient CPM at younger ages. However, this issue is far
33,38 pain. Pain 2011;152:2701–9.
from being conclusive because of contrary findings. The [9] DeSouza DD, Hodaie M, Davis KD. Abnormal trigeminal nerve
other limitation is that this is a cross-sectional study; the ideal microstructure and brain white matter in idiopathic trigeminal
design for a study that relates to durations is a longitudinal neuralgia. Pain 2014;155:37–44.
[10] Granot M, Weissman-Fogel I, Crispel Y, Pud D, Granovsky Y, Sprecher E,
study. Because no prospective follow-up was designed, we
Yarnitsky D. Determinants of endogenous analgesia magnitude in a
rely on patients’ memory as far the durations of both diffuse noxious inhibitory control (DNIC) paradigm: do conditioning
neuropathy and pain, rather than on documented data. stimulus painfulness, gender and personality variables matter? Pain
To conclude, this is a first report on the association 2008;136:142–9.
between duration of neuropathic pain and pain modulation. [11] Granovsky Y, Miller-Barmak A, Goldstein O, Sprecher E, Yarnitsky D.
CPM test-retest reliability: “standard” vs “single test-stimulus”
Pain modu-lation seems to return to normal for longer protocols. Pain Med 2016;17:521–29.
duration of pain, whereas clinical pain levels seem to persist [12] Grashorn W, Sprenger C, Forkmann K, Wrobel N, Bingel U. Age-
for a long time, at least in our cohort of patients with PDN. dependent decline of endogenous pain control: exploring the effect
Our findings on the difference of antinociceptive PMP in PDN of expectation and depression. PLoS One 2013;8:e75629.
patients with longer pain duration highlight the significant role [13] Graven-Nielsen T, Wodehouse T, Langford RM, Arendt-Nielsen L,
Kidd BL. Normalization of widespread hyperesthesia and facilitated
of the “time window” in which the assessment of PMP has spatial summation of deep-tissue pain in knee osteoarthritis patients
clinical meaning and should be considered as a factor in after knee replacement. Arthritis Rheum 2012;64:2907–16.
treatment choice and evaluation of treatment efficacy. [14] Hermans L, Van Oosterwijck J, Goubert D, Goudman L, Crombez G,
Calders P, Meeus M. Inventory of personal factors influencing
conditioned pain modulation in healthy people: a systematic
Disclosures literature review. Pain Pract 2016;16:758–69.
[15] Jensen TS, Baron R, Haanp a¨ a¨ M, Kalso E, Loeser JD, Rice AS,
The authors have no conflicts of interest to declare. Treede RD. A new definition of neuropathic pain. Pain 2011;152:2204–5.
This study was sponsored by an IIT grant from Eli Lilly Inc. [16] Johnson MS, Ryals JM, Wright DE. Diabetes-induced chemogenic
hypoalgesia is paralleled by attenuated stimulus-induced fos
Y. Granovsky contributed to logistics of this study, data expression in the spinal cord of diabetic mice. J Pain 2007;8:637–49.
analysis, and the manuscript preparation. H. Nahman-Averbuch [17] Khasabov, SG, Simone DA. Loss of neurons in rostral ventromedial
contributed to data collection. M. Khamaisi contributed to the medulla that express neurokinin-1 receptors decreases the
patients recruitment and consenting. M. Granot contributed to development of hyperalgesia. Neuroscience 2013;250:151–65.
2 (2017) e592 www.painreportsonline.com 7

[18] Kosek E, Ordeberg G. Lack of pressure pain modulation by heterotopic sensitivity in the adult Danish general population: the DanFunD study.
noxious conditioning stimulation in patients with painful osteoarthritis J Pain 2017;18:274–84.
before, but not following, surgical pain relief. Pain 2000;88:69–78. [34] Spielberger CD, O’Neil HF Jr, Hansen DN. Anxiety, drive theory, and
[19] Le Bars D. The whole body receptive field of dorsal horn ultireceptive computer-assisted learning. Prog Exp Pers Res 1972;6:109–48.
neurones. Brain Res Brain Res Rev 2002;40:29–44. Review. [35] Sullivan M, Bishop S, Pivic J. The pain catastrophizing scale:
[20] Lewin-Epstein N, Sagiv-Schifter T, Shabtai EL, Shmueli A. Validation development and validation. Psychol Assess 1995;7:524–32.
of the 36-item Short-Form Health Survey (Hebrew version) in the [36] Tseng MT, Chiang MC, Yazhuo K, Chao CC, Tseng WY, Hsieh ST.
adult population of Israel. Med Care 1998;36:1361–70. Effect of aging on the cerebral processing of thermal pain in the
[21] Lewis GN, Rice DA, McNair PJ. Conditioned pain modulation in human brain. Pain 2013;154:2120–9.
populations with chronic pain: a systematic review and meta- [37] Tuveson B, Leffler AS, Hansson P. Heterotopic noxious conditioning
analysis. J Pain 2012;13:936–44. stimulation (HNCS) reduced the intensity of spontaneous pain, but not of
[22] Marchettini P, Teloni L, Formaglio F, Lacerenza M. Pain in diabetic allodynia in painful peripheral neuropathy. Eur J Pain 2007;11:452–62.
neuropathy case study: whole patient management. Eur J Neurol [38] Vaegter HB, Handberg G, Graven-Nielsen T. Similarities between
2004; 11(suppl 1):12–21. exercise-induced hypoalgesia and conditioned pain modulation in
[23] Nahman-Averbuch H, Granovsky Y, Sprecher E, Steiner M, Tzuk-Shina T, Pud humans. Pain 2014;155:158–67.
D, Yarnitsky D. Associations between autonomic dysfunction and pain in [39] van Wijk G, Veldhuijzen DS. Perspective on diffuse noxious inhibitory
chemotherapy-induced polyneuropathy. Eur J Pain 2014;18:47–55. controls as a model of endogenous pain modulation in clinical pain
[24] Nahman-Averbuch H, Yarnitsky D, Granovsky Y, Sprecher E, Steiner syndromes. J Pain 2010;11:408–19.
M, Tzuk-Shina T, Pud D. Pronociceptive pain modulation in patients [40] Veves A, Backonja M, Malik RA. Painful diabetic neuropathy:
with painful chemotherapy-induced polyneuropathy. J Pain Symptom epidemiology, natural history, early diagnosis, and treatment options.
Manage 2011;42:229–38. Pain Med 2008;9:660–74.
[25] Niesters M, Proto PL, Aarts L, Sarton EY, Drewes AM, Dahan A. [41] Villanueva L, Le Bars D. The activation of bulbo-spinal controls by
Tapentadol potentiates descending pain inhibition in chronic pain patients peripheral nociceptive inputs: diffuse noxious inhibitory controls. Biol
with diabetic polyneuropathy. Br J Anaesth 2014;113:148–56. Res 1995;28:113–25.
[26] Otto M, Bak S, Bach FW, Jensen TS, Sindrup SH. Pain phenomena [42] Wang R, King T, De Felice M, Guo W, Ossipov MH, Porreca F.
and possible mechanisms in patients with painful polyneuropathy. Descending facilitation maintains long-term spontaneous neuropathic
Pain 2003; 101:187–92. pain. J Pain 2013;14:845–53.
[27] Parkhouse N, Le Quesne PM. Impaired neurogenic vascular [43] Weissman-Fogel I, Granovsky Y, Crispel Y, Ben-Nun A, Best LA,
response in patients with diabetes and neuropathic foot lesions. N Yarnitsky D, Granot M. Enhanced presurgical pain temporal
Engl J Med 1988; 318:1306–9. summation response predicts post-thoracotomy pain intensity during
[28] Paulson PE, Minoshima S, Morrow TJ, Casey KL. Gender the acute postoperative phase. J Pain 2009;10:628–36.
differences in pain perception and patterns of cerebral activation [44] Wilder-Smith OH, Schreyer T, Scheffer GJ, Arendt-Nielsen L. Patients with
during noxious heat stimulation in humans. Pain 1998;76:223–9. chronic pain after abdominal surgery show less preoperative
[29] Peters CM, Hayashida K, Suto T, Houle TT, Aschenbrenner CA, endogenous pain inhibition and more postoperative hyperalgesia:
Martin TJ, Eisenach JC. Individual differences in acute pain-induced a pilot study. J Pain Palliat Care Pharmacother 2010;24:119–28.
endogenous analgesia predict time to resolution of postoperative [45] Yarnitsky D. Quantitative sensory testing. Muscle Nerve 1997;20:198–204.
pain in the rat. Anesthesiology 2015;122,895–907. [46] Yarnitsky D. Role of endogenous pain modulation in chronic pain
[30] Porreca F, Ossipov MH, Gebhart GF. Chronic pain and medullary mechanisms and treatment. Pain 2015;156:S24–31.
descending facilitation. Trends Neurosci 2002;25:319–25. [47] Yarnitsky D, Crispel Y, Eisenberg E, Granovsky Y, Ben-Nun A, Sprecher E,
[31] Ren K, Dubner R. Descending modulation in persistent pain: an Best LA, Granot M. Prediction of chronic post-operative pain: pre-
update. Pain 2002;100:1–6. operative DNIC testing identifies patients at risk. Pain 2008;138:22–8.
[32] Skou ST, Graven-Nielsen T, Rasmussen S, Simonsen OH, Laursen MB, [48] Yarnitsky D, Granot M, Granovsky Y. Pain modulation profile and pain
Arendt-Nielsen L. Widespread sensitization in patients with chronic pain therapy: between pro- and antinociception. Pain 2014;155:663–5.
after revision total knee arthroplasty. Pain 2013;154:1588–94. [49] Yarnitsky D, Granot M, Nahman-Averbuch H, Khamaisi M,
[33] Skovbjerg S, Jørgensen T, Arendt-Nielsen L, Ebstrup JF, Carstensen Granovsky Y. Conditioned pain modulation predicts duloxetine
T, Graven-Nielsen T. Conditioned pain modulation and pressure pain efficacy in painful diabetic neuropathy. Pain 2012;153:1193–8.