ORIGINAL ARTICLE
A Delphi consensus statement of the Neuropathic Pain Special
Interest Group of the Italian Neurological Society
on pharmacoresistant neuropathic pain
P. Ciaramitaro
1
& G. Cruccu
2
& M. de Tommaso
3
& G. Devigili
4
& D. Fornasari
5
& P. Geppetti
6
& M. Lacerenza
7
&
G. Lauria
8,9
& S. Mameli
10
& P. Marchettini
6,11,12
& M. Nolano
13,14
& E. Polati
15
& V. Provitera
16
& M. Romano
17
&
C. Solaro
18
& S. Tamburin
19
& V. Tugnoli
20
& M. Valeriani
21,22
& Andrea Truini
2
& On behalf of the Neuropathic Pain
Special Interest Group of the Italian Neurological Society
Received: 18 December 2018 /Accepted: 23 March 2019 /Published online: 2 April 2019
#
Fondazione Società Italiana di Neurologia 2019
Abstract
To improve patient care and help clinical research, the Neuropathic Pain Special Interest Group of the Italian Neurological
Society appointed a task force to elaborate a consensus statement on pharmacoresistant neuropathic pain. The task force included
19 experts in neuropathic pain. These experts participated in a Delphi survey consisting of three consecutive rounds of questions
and a face-to-face meeting, designed to achieve a consensus definition of pharmacoresistant neuropathic pain. In the three rounds
of questions, the participants identified and described the main distinguishing features of pharmacoresistance. In the face-to-face
meeting the participants discussed the clinical features determining pharmacoresistance. They finally agreed that neuropathic
pain is pharmacoresistant when “the patient does not reach the 50% reduction of pain or an improvement of at least 2 points in the
Patient Global Impression of Change, having used all drug classes indicated as first, second, or third line in the most recent and
widely agreed international guidelines, for at least 1 month after titration to the highest tolerable dose.” Our consensus statement
might be useful for identifying eligible patients for invasive treatments, and selecting patients in pharmacological trials, thus
improving patient care and helping clinical research.
Keywords Neuropathic pain
.
Pharmacoresistance
.
Painful neuropathy
.
Refractory pain
Introduction
Neuropathic pain—i.e., pain arising directly from a lesion or
disease that affects the somatosensory nervous system—is a
common clinical problem (http://www.iasp-pain.org/
Education/Content.aspx?ItemNumber=1698). Neuropathic
pain, having a prevalence in the range of 7–10%, represents
a significant economic and social bur den [1]. Despite the
advances, a considerable number of patients do not get
adequate pain relief or improvement in quality of life from
available drugs. Meta-analysis [2]aswellasexperiencedcli-
nicians claim that neuropathic pain is pharmacoresistant in
about 50% of patients [3].
Although pharmacoresistant neuropathic pain is a
common clinical problem [3, 4], a precise definition of
pharmacoresistance remains elusive. In most studies assessing
the efficacy of invasive procedures for patients with
pharmacoresistant neuropathic pain, the definition criterion
of pharmacoresistance is frequently ambiguous [5, 6]. The
current guideline on the clinical development of medicinal
products intended for the treatment of pain, issued by the
European Medicine Agency (http s://www.ema.europ a.eu/
documents/scientific-guideline/draft-guideline-clinical-
development-me di cinal- prod ucts- inte nde d-t reatme nt -pain _
en-0.pdf), does not provide any indication on how to define
pharmacoresistan ce to neuropathic pain. A pragmatic
definition of pharmacoresistant neuropathic pain may be
therefore useful for identifying patients for invasive
treatments in clinical practice, or for selecting patients in
pharmacological trials.
The aim of this study, issued by the Neuropathic Pain
Special Interest Group (NeuPSIG) of the Italian
* Andrea Truini
andrea.truini@uniroma1.it
Extended author information available on the last page of the article
Neurological Sciences (2019) 40:1425–1431
https://doi.org/10.1007/s10072-019-03870-y
Neurological Society, was to reach a consensus statement on
how to define pharmacoresistant neuropathic pain. To do so a
task force participated into three consecutive rounds of ques-
tionnaires, and in a face-to-face meeting designed to achieve a
final consensus statement on pharmacoresistance definition.
Methods
The NeuPSIG of the Italian Neurological Society organized
this Consensus, based on the Delphi technique, between April
and June 2018. The task force included 15 neurologists, two
pharmacologists, and two pain physicians, selected because of
their specific expertise in neuropathic pain. The participants
were selected due to their scientific activity in the neuropathic
pain field, the geographical representation of the Italian terri-
tory, and the clinical and scientific activity associated with the
Italian Neurological Society. For instance, several members of
this team participated in relatively recent epidemiological
studies dealing with neuropathic pain [7–9]. These experts
were contacted by email with an invitation to participate in
the Delphi method survey .
We organized three rounds of open questions, sent by email
to the expert panel. Two facilitators (AT and GC) developed
the questions. After each round, the two facilitators analyzed
the responses and elaborated the successive round questions.
Each participant was unaware of the responses of the other
participants.
Of the 19 invited experts, 17 completed the three rounds
and 16 participated to the face-to-face meeting. In this consen-
sus procedure, we set a priori a level of agreement at 80%.
The response rate for the three rounds of questions ranged
between 82 and 100%.
Rounds of questions
In the first round, we sought the general opinion of the partic-
ipants about the problem of treatment-resistant neuropathic
pain. We asked, “In your opinion when does neuropathic pain
become “pharmacoresistant?”
For the second round, a summary of the results of round
one was fed back to participants by email, along with ques-
tions of round 2. This round included items directly raised by
the answers to the round 1 questions. Accordingly, the ques-
tions of the second round were focused to reach a precise
definition of clinically meaningful pain relief, global impres-
sion of change, and how to identify effective drugs and their
dosage and minimal duration of treatment, before defining a
patient pharmacoresistant.
A summary of the results of round two was then fed back to
participants by email. The third round aimed at completing
and clarifying some of the issues raised by the replies to the
second round. We therefore included questions on whether
drugs of the same class (e.g., different tricyclic antidepres-
sants) shall be tried, how comorbidities influence the response
to treatment, and whether other, non-invasive, therapies (e.g.,
nutritional supplements, relaxation) shall be tried before de-
fining a patient pharmacoresistant
Face-to-face meeting
After the three rounds of questions, we fed back a list of all the
answers in anonymized form to participants. This procedure
step was meant to prepare participants to the face-to-face
meeting. In this meeting, the participants were asked to dis-
cuss the different items raised by the rounds of questions and
indicate their agreement with each of the components to be
incorporated into a working definition of pharmacoresistant
neuropathic pain.
In the face-to-face meeting the participants also discussed
about the possible differences between the terms
pharmacoresistance and refractoriness, and sought an agree-
ment on the distinguishing features of these two terms.
Results
Rounds of questions
The answers to the first question showed that all participants,
though using different wording, indicated that neuropathic
pain becomes pharmacoresistant when drugs do not provide
sufficient pain relief or a sufficient global impression of
change (Table 1).
In the second round, the participants indicated the def-
inition of clinically meaningful pain relief as a reduction
of pain of 30 or 50%, as assessed with a validated pain
scale (i.e., the Visual Analogue Scale and the Likert
scale). Most participants indicated as a clinically mean-
ingful global impression of change, a 2-point improve-
ment (corresponding to “very much improved” or “much im-
proved”) in the Patient Global Impression of Change scale
(PGIC), a 7-point scale commonly used in the management
of patients with neuropathic pain (Fig. 1). Most participants
considered as effective drugs those having at least a weak
recommendation in international guidelines or recommenda-
tions. Virtually all the participants considered the adequate
drug dosage as the best balance between effectiveness and
tolerability; conversely, they provided widely varying defini-
tions for adequate d uration of drug trial, indicating a trial
duration between 2 weeks and 6 months.
In the round 3, according to several participants although
no specific evidence is available, different drugs of the same
classes should be tried, mainly due to possible differences in
pharmacokinetic profiles. Concerning the influence of co-
morbidities, the participants indicated t hree main issues:
1426 Neurol Sci (2019) 40:1425–1431
the patient compliance and adherence to therapy, psychi-
atric comorbidities, and patient’s genotype influencing
the drug effectiveness. Replies about additional treat-
ments varied considerably among the participants.
Some participants indicated among possible useful treat-
ments cognitive and behavioral therapy, physiotherapy,
acupuncture, relaxation, and nutritional supplements.
However, many participants considered that the current
evidence supporting the clinical use of these treatments
is still insufficient (Fig. 1).
Face-to-face meeting
In the face-to-face meeting there was unanimous consensus
that among the different issues raised during the three rounds
of questions, pharmacoresistance definition should include the
following items: satisfactory pain relief or PGIC, number of
efficacious drugs, dosage of the drugs, and treatment duration.
All participants agreed that satisfactory pain relief corresponds
to the 50% reduction of pain and a clinically meaningful global
impression of change corresponds to an improvement of at least
2 points in the PGIC. All the drugs indicated as first, second, or
third line in the most recent international guidelines should be
tried, for at least 1 month after titration, using the highest dose
according to the EU approved Summary of Product
Characteristics (SPC), without clinically important side effects.
All the participants agreed that pharmacoresistant and re-
fractory neuropathic pain are distinct conditions. The defini-
tion of pharmacoresistance should be used only regarding
pharmacological treatment; conversely, refractoriness is more
relevant to the lack of response to any treatment, including
neurostimulation and surgical procedures.
Box 1 Definition of pharmacoresistant neuropathic pain
Discussion
In this study, using the Delphi method, a task force appointed
by the NeuPSIG of the Italian Neurological Society, agreed
that neuropathic pain is pharmacoresistant when “the patient
does not reach the 50% reduction of pain or an improvement
of at least 2 points in the PGIC, having used all drug classes
indicated as first, second, or third line in the most recent and
widely agreed international guidelines, for at least 1 month
after titration to the highest tolerable dose.” We believe that
this consensus statement on pharmacoresistant neuropathic
pain may serve as a pragmatic and applicable definition for
the everyday clinical management and pharmacological trials.
The primary target users of our consensus definition of
pharmacoresistant neuropathic pain are medical practitioners
at all health care levels involved in the clinical management of
patients with neuropathic pain. Our definition may help phy-
sicians, health care administrators, and insurers in identifying
eligible patients for invasive therapeutic procedures. Other
target users are clinical researchers because adopting a con-
sensus definition might help for selecting patients in pharma-
cological trials.
Neuropathic pain is pharmacoresistant when “the patient does not reach
the 50% reduction of pain or an improvement of at least 2 points in the
Patient Global Impression of Change, having used all drug classes
indicated as first, second, or third line in the most recent and widely
agreed international guidelines, for at least 1 month after titration to the
highest tolerable dose.”
Table 1 Round questions and exemplary responses
First round
In your opinion when does neuropathic pain become
“pharmacoresistant”?
- When treatment with at least one drug for each class (TCA,
gabapentinoids, sodium channel blockers, opioids), used at full dose
strength for at least 6 weeks, does not reduce the intensity of pain by at
least 50% on a validated rating scale (e.g., VAS, NRS) and/or does not
change the patient global impression of change.
- Failure of more than three medications known to be effective, used
for more than 3 months and at the maximum tolerated dosage, where
pain failure is defined as a reduction between T0 and T1 < 50% or no
change in the Patient Global Impression of Change
Second round
Define sufficient pain relief
- 30% of pain (NRS/VAS) reduction
- 50% decrease at VAS/11-point NRS
Define sufficient global impression of change
- 2 points
- A noticeable change that affects significantly the quality of life
List effective drugs, including polypharmacy
- Amitriptyline, duloxetine, pregabalin, gabapentin, carbamazepine,
oxcarbazepine, tramadol, oxycodone, amytriptiline + gabapentinoids;
duloxetine + gabapentinoids; tramadol/oxycodone + gabapentinoids
Define adequate drug dosage
- Dosage in the therapeutic range sufficient to cause pain relief,
without adverse events
- It is the maximum tolerated dose providing the higher pain relief and
improvement of the Patient Global Impression of Change
Define the adequate duration of a drug trial
- 8 weeks at the maximum tolerated dosage
- Two weeks to 1 month after titration
Third round
Different drugs of the same class should be tried?
- Yes, due to different pharmacokinetic profile
- No, possible differences between drugs of the same class are usually
negligible
What might compromise the response to the treatment?
- Patient compliance, depression, cognitive disturbances
-Patient’s genotype influencing drug effectiveness
Other non-drug, non-invasive, therapies should be tried?
- Hypnosis, mindfulness, or other techniques able to help
self-consciousness
- Most of the non-pharmacological approaches produce only a
placebo effect
Neurol Sci (2019) 40:1425–1431 1427
The definition of satisfactory pain relief and PGIC
How to define satisfactory pain relief remains a controversial
problem. Many pharmacological trials assessing the efficacy
of drugs in patients with neuropathic pain indicate as primary
endpoint the 30% reduct ion of pain [2]. However, several
studies showed that clinically important improvement of qual-
ity of life and health status are associated with pain relief
higher than the 30% reduction of pain [10, 11]. Furthermore,
the percent change that is clinically important to patient in-
creases as the baseline pain severity increases. Hence, for pa-
tients with severe pain, the percentage of pain relief should be
higher than 30%. For these reasons, we decided to indicate the
50% reduction of pain as the appropriate targ et for pharmaco-
logical treatment, given that a pain reduction of ≥ 50% probably
represents the substantial change in pain severity for most pa-
tients [12].
In the definition of pharmacoresistant pain we decided to
include also the PGIC. The percent change of pain and the
PGIC are closely related; this correlation, however, loses con-
sistency in patients with high pain scores. Furthermore many
drugs used for neuropathic pain treatment (e.g., antidepres-
sants and gabape ntinoids) may i mprove sl eep, mood, and
therefore quality of life, regardless of the degree of pain re-
duction [13]. We therefore believe that the PGIC is a useful,
alternative item to pain reduction. It follows that we may con-
sider drug responders those patients with a pain relief lower
than 50%, but with an adequate improvement of PGIC. We
identified as an adequate improvement of the PGCI a change
of 2 points corresponding to “very much improved” or “much
improved,” given that this improvement notoriously corre-
sponds to a clinically important change in the general health
status [12].
The definition of effective pharmacological treatment
Many drugs are supposed to modulate somatosensory nervous
system activity, thus possibly reducing pain. Only some drugs,
however, have be en tested in randomized controlled trials
proving some efficacy in reducing neuropathic pain.
Accordingly, in the definition of pharmacoresistance we re-
ferred to all the drug classes indicated as first, second, or third
Fig. 1 Frequency of the different responses to round questions 2 and 3.
Given that rounds 2 and 3 consisted of open questions, for some questions
the two facilitators performed a qualitative analysis and grouped together
similar items. For instance in the question D “Define the adequate
duration of a drug trial,” the answers reporting a time span between
8 weeks and 6 months have been included in the item “2 months or
more.” The sum of percentage of answers to question G is above 100%
given that some participants provided more than one answers
1428 Neurol Sci (2019) 40:1425–1431
line in the most recent, widely agreed international guidelines
based on meta-analysis of published randomized controlled
trials. We propose that the most recent and appropriate guide-
lines are used. Currently, the guideline for neuropathic pain
treatment issued by the Neuropathic Pain Special Interest
Group of the International Association for the Study of Pain
[2] should be considered as the reference guideline unless an
updated guideline is published.
In the definition of pharmacoresistance we indicated that
the minimal duration of treatment necessary to consider the
drug insufficient should be of at least 1 month after titration.
This choice reflects two main considerations. Drugs used for
treating neuropathic pain are usually carefully titrated for
minimizing adverse events, and the adequate dosage is
reached after several days or weeks. While the analgesic
effects of antidepressants on pain manifest in a few days
to 1 week, the effects on depression and quality of life
take approximately 2 to 4 weeks [14 ], thus possibly hav-
ing late influence on the PGIC.
The magnitude of pain relief usually correlates with the
drug dosage: the higher the dosage is, the larger the pain relief
is [15, 16]. Admittedly, this relationship has not been
homogenously reported for all drugs. For instance whereas
the efficacy of pregabalin linearly increases with the dosage,
theefficacyof60and120mgofduloxetinedoesnotdiffer
[17, 18]. Regardless of the dose-response relationship,
however, the frequency and sever ity of adverse events
increases with the drug dosage [19]. We have therefore
indicated that the drug dosage should aim at the highest
dosage, according to the EU approved Summary of
Product Characteristics, without clinically important side
effects (i.e., the highest tolerable dose).
Pharmacoresistance and refractoriness
Although most clinical trials assessing drug efficacy have
used pharmacoresistanc e and refractoriness as synonyms
[20–22], some studies suggested that the two definitions might
be perceived as different entities [3, 4]. We suggest that while
the term pharmacoresistance explicitly refers to drugs, the
term refractoriness should be intended for neuropathic pain
resistant also to non-pharmacological treatment (e.g.,
neurostimulation, physical therapies, cognitive-behavioral
therapy). We believe that a clear-cut distinction between the
two terms might be clinically useful, given that the
pharmacoresistance is a requisite for patients’ eligibility to
invasive procedures.
Limitations of the study
Our expert panel mainly includes neurologists, given that the
study is issued by the NeuPSIG of the Italian Neurological
Society. This unbalanced team might be regarded as a
limitation, because different medical practitioners may be in-
volved in the clinical management of patients with neuropath-
ic pain. However, neuropathic pain is due to neurological dis-
eases. We therefore believe that the special role for neurolo-
gists in this field is largely expected and justified.
In this study using the Delphi procedure we provide an
expert opinion consensus, unsupported by evidence-based da-
ta. However, we believe that this limitation is intrinsic to the
problem of pharmacoresistant neuropathic pain, given that no
study has systematically addressed the pharmacoresistance
and most trials including patients with pharmacoresistant neu-
ropathic pain did not explicitly define the criteria for identify-
ingpatientswithpoororabsentresponsetotreatment.
Although our definition of pharmacoresistance has some
resembling points with two earlier expert opinion-based stud-
ies [3, 4], important differences still exist. In our study we
explicitly provide previously unreported details on the pain
relief, PGIC, and drugs, useful to yield a working definition
of pharmacoresistance, essential in clinical practice.
Particularly we identify the 50% reduction of pain rather than
the 30%, reported in a previous study [4], and precisely indi-
cate the level of satisfactory PGIC. We also clearly indicate
that the effective drugs should be identified based on standard
reference guidelines. Differently from the two previously pub-
lished expert opinion-based studies [3, 4], we also addressed
the possible differences between pharmacoresistant and re-
fractory neuropathic pain.
Although many participants considered that different drugs
of the same class should be tried, we did not include this point
in the pharmacoresistance definition. Admittedly, drugs of the
same class might have a different effect for minor differences
in pharmacodynamic profile or different pharmacokinetic
properties. For instance among the tricyclic antidepressants,
amitriptyline and imipramine have a stronger activity on
voltage-gated sodium channels than nortriptyline and desipra-
mine [23, 24]. Gabapentin and pregabalin have minor phar-
macodynamic differences (pregabalin has greater binding af-
finity for the alpha-2/delta-1 subunit than gabapentin), and
major differences in terms of pharmacokinetics, especially
absorption (the intestinal absorption of pregabalin is not satu-
rable) [25, 26]. We considered, however, that the possible
differences between drugs of the same class are usually neg-
ligible; furthermore, screening different drugs of the same
class may require a long time, thus making a definition of
pharmacoresistance unsuitable to clinical practice.
Conclusions
In this study, using a Delphi procedure to reach a consensus
statement, we provide a working definition for
pharmacoresistant neuropathic pain, possibly useful for iden-
tifying patients for invasive treatments in the everyday clinical
Neurol Sci (2019) 40:1425–1431 1429
