Reviews/Commentaries/ADA Statements

M E T A - A N A L Y S I S

Efficacy, Safety, and Tolerability of

Pregabalin Treatment for Painful Diabetic
Peripheral Neuropathy
Findings from seven randomized, controlled trials across a range of doses
ROY FREEMAN, MD1 There are seven double-blind, ran-
EDITH DURSO-DECRUZ, PHD2 domized, placebo-controlled trials in
BIROL EMIR, PHD2 painful DPN with pregabalin (6 –12), five
of which are published in full (7–11). The
effective dosing range for treatment of
OBJECTIVE — To evaluate the efficacy, safety, and tolerability of pregabalin across the ef- neuropathic pain syndromes is 150 to
fective dosing range, to determine differences in the efficacy of three times daily (TID) versus 600 mg/day, administered either three
twice daily (BID) dosage schedules, and to use time-to-event analysis to determine the time to times daily (TID) or twice daily (BID).
onset of a sustained therapeutic effect using data from seven trials of pregabalin in painful Among the seven trials, dosages of 150,
diabetic peripheral neuropathy (DPN). 300, and 600 mg/day were used, but only
one trial included all three of these dos-
RESEARCH DESIGN AND METHODS — Data were pooled across seven double-
blind, randomized, placebo-controlled trials using pregabalin to treat painful DPN with dosages ages. Thus, individually, the seven trials
of 150, 300, and 600 mg/day administered TID or BID. Only one trial included all three of these present an incomplete picture of the ef-
dosages, and TID dosing was used in four. All studies shared fundamental selection criteria, and fective dosing range. In addition, TID
treatment durations ranged from 5 to 13 weeks. dosing was used in the first four trials,
whereas the three most recent trials of
RESULTS — Pooled analysis showed that pregabalin significantly reduced pain and pain- pregabalin in painful DPN used BID
related sleep interference associated with DPN (150, 300, and 600 mg/day administered TID vs. dosing.
placebo, all P ⱕ 0.007). Only the 600 mg/day dosage showed efficacy when administered BID The objective of the current report is
(P ⱕ 0.001). Pain and sleep interference reductions associated with pregabalin appear to be to use the pooled data from these seven
positively correlated with dosage; the greatest effect was observed in patients treated with 600
trials to evaluate the efficacy, safety, and
mg/day. Kaplan-Meier analysis revealed that the median time to onset of a sustained (ⱖ30% at
end point) 1-point improvement was 4 days in patients treated with pregabalin at 600 mg/day, tolerability of pregabalin across the effec-
5 days in patients treated with pregabalin at 300 mg/day, 13 days in patients treated with tive dosing range. We also use these data
pregabalin at 150 mg/day, and 60 days in patients receiving placebo. The most common treat- to determine differences in the efficacy of
ment-emergent adverse events were dizziness, somnolence, and peripheral edema. TID and BID dosing schedules. Finally,
we use a time-to-event analysis of the
CONCLUSIONS — Treatment with pregabalin across its effective dosing range is associated pooled data to determine the time to onset
with significant, dose-related improvement in pain in patients with DPN. of a sustained therapeutic effect across the
range of doses.
Diabetes Care 31:1448–1454, 2008

RESEARCH DESIGN AND

T
he prevalence of diabetic neuropa- classes are used to treat painful DPN with METHODS — Study treatment dura-
thy is as high as 50% in patients who varying degrees of efficacy, safety, and tol- tions ranged from 5 to 13 weeks (Fig. 1A).
have had diabetes for 25 years (1), erability. The antiepileptic agents gaba- Four trials used TID dosing, three used
and painful diabetic peripheral neuropa- pentin and pregabalin have attained BID dosing, and all but one trial (pregaba-
thy (DPN) occurs in up to 26% of all peo- widespread use in the treatment of painful lin 300 mg/day vs. placebo) used escala-
ple with diabetes (2). Symptoms range DPN. These agents bind to the auxiliary tion to assigned fixed dosing over a period
from mild dysesthesias to severe unremit- ␣2-␦ subunit of the voltage-sensitive cal- of 1 to 2 weeks. Patients were randomized
ting pain that can profoundly impact pa- cium channel, thereby decreasing Ca2⫹ to placebo or fixed-dosage pregabalin at
tients’ lives (3,4). influx at nerve terminals and modulating
150, 300, or 600 mg/day. One trial in-
Medications of several different neurotransmitter release (5).
cluded a 75-mg/day dosage arm (8); re-
● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● sults for this dosage are not presented
From the 1Department of Neurology, Beth Israel Deaconess Medical Center, Harvard Medical School, here, as 75 mg/day is considered to be a
Boston, Massachusetts; and 2Pfizer Global Pharmaceuticals, New York, New York. nontherapeutic dosage of pregabalin in
Corresponding author: Roy Freeman, [email protected].
Received 1 November 2007 and accepted 18 March 2008.
painful DPN. One trial (10) studied both
DOI: 10.2337/dc07-2105 DPN and postherpetic neuralgia patients
R.F. has served as a consultant to Pfizer. administered flexible-dosage pregabalin
© 2008 by the American Diabetes Association. Readers may use this article as long as the work is properly (150 – 600 mg/day), fixed-dosage pre-
cited, the use is educational and not for profit, and the work is not altered. See http://creativecommons. gabalin (600 mg/day), or placebo. To en-
org/licenses/by-nc-nd/3.0/ for details.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby sure consistency for the purposes of this
marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. analysis, only the DPN patients who re-

1448 DIABETES CARE, VOLUME 31, NUMBER 7, JULY 2008

 Freeman, Durso-DeCruz, and Emir

Figure 1—A: Dosage arms of seven trials contributing to this pooled analysis, ITT populations. *Trial used a modified ITT population: 11 patients
were withdrawn by Ministry of Health/European Committee during a partial clinical hold. AEs, adverse events; PBO, placebo; PGB, pregablin. †Trial
included 338 patients total, 96 of whom had painful DPN and were assigned to a fixed dosage of 600 mg/day pregabalin. Postherpetic neuralgia
patients from this trial were not included in the present analysis nor were DPN patients assigned to flexible-dosage pregabalin. ‡No dose escalation.
PBO, placebo. B: Pooled studies patient disposition with baseline demographics and characteristics.

DIABETES CARE, VOLUME 31, NUMBER 7, JULY 2008 1449

Pregabalin in diabetic peripheral neuropathy

Figure 2—A: Change from baseline to end point in least-squares mean pain score based on last observation carried forward analysis. Patient
population comprised of patients who had both baseline and end point assessments (numbers in some groups are therefore smaller than in the ITT
population). Significant reductions in end point least-squares mean pain score were observed for all three dosages investigated: ⫺2.05, ⫺2.36, and
⫺2.75 points for patients receiving pregabalin 150, 300, and 600 mg/day vs. ⫺1.49 for patients receiving placebo (*P ⫽ 0.007 for 150 mg/day and
†P ⬍ 0.0001 for 300 and 600 mg/day vs. placebo). B: Change from baseline to week 5 in least-squares mean pain score. Reductions were observed
for all three dosages investigated: ⫺1.98, ⫺2.44, and ⫺2.75 points for patients receiving pregabalin 150, 300, and 600 mg/day vs. ⫺1.47 for patients
receiving placebo (*P ⬍ 0.0001 vs. placebo; †P ⬍ 0.01 vs. placebo). C: Proportion of patients meeting ⱖ50% improvement and ⱖ30% improvement

1450 DIABETES CARE, VOLUME 31, NUMBER 7, JULY 2008

 Freeman, Durso-DeCruz, and Emir

ceived fixed-dosage pregabalin (28% of icant differences versus placebo be- RESULTS — A total of 1,510 patients
the total cohort) were included in this tween BID and TID dosing regimens. represented the ITT population in the
analysis. Finally, time to onset of sustained and seven studies: 557 received placebo, and
Each of the studies shared fundamen- clinically meaningful pain relief was in- 953 received pregabalin. Ninety percent
tal inclusion criteria, including ⱖ18 years vestigated across the seven studies. This of patients were white, and 58% were
of age, an average pain score ⱖ4 (on an was defined as the first day on which pa- male. Mean age was 59 years, mean
11-point, Likert-like numeric rating scale tients demonstrated a ⱖ1-point reduc- weight was 93 kg, and mean baseline pain
[NRS]: 0 ⫽ “no pain” to 10 ⫽ “worst pos- tion in mean pain score in patients with a score was 6.5 (Fig. 1B).
sible pain”) over a 7-day baseline period, ⱖ30 and ⱖ50% reduction in mean pain
and a score ⱖ40 mm on the 0- to 100-mm score at end point. These two criteria were Efficacy
visual analog scale of the Short-Form imposed to ensure clinically meaningful Significant reductions in end point least-
McGill Pain Questionnaire at screening and durable pain relief based on evidence squares mean pain scores were observed
and randomization (baseline and ran- that in studies with a similar design, a for all three dosages investigated (P ⫽
domization in one study). All patients in ⱖ30% improvement from baseline corre- 0.007 for 150 mg/day and P ⬍ 0.0001 for
each trial were required to have A1C lev- sponds to a patient global impression of 300 and 600 mg/day vs. placebo) (Figs.
els ⱕ11%. Prior therapeutic failure of change of “much improved” or “very 2A and B). Pain reductions associated
gabapentin was an exclusion criterion much improved” at study end point (13). with pregabalin appear to be positively
in three studies (6 – 8). All patients pro- Baseline pain scores in a typical clinical correlated with dosage, with the greatest
vided informed consent before partici- trial with painful DPN patients are be- effect observed in patients treated with
pation, and all studies were conducted tween 6 and 6.5 on the 0- to 10-point 600 mg/day. The proportions of patients
in compliance with the ethics principles scale; therefore, a 30% improvement is experiencing ⱖ50 or ⱖ30% reductions in
originating in or derived from the Dec- pain levels (responders) were signifi-
⬃2 points. We defined the “event” of in-
laration of Helsinki, internal review cantly greater in the pregabalin groups
terest in this time-to-event analysis as the
board requirements, or good clinical than in the placebo group (Fig. 2C) and
time to the first ⱖ1-point reduction in the
practices guidelines. were dose related.
daily pain score, recognizing that any
The primary efficacy measure in each The number needed to treat for these
such criterion could be considered arbi-
study was end point mean pain score (on data are as follows: pregabalin 600 mg/
the 11-point NRS) derived from entries in trary. Time to onset of sustained pain re- day 4.04 (95% CI 3.3–5.3), pregabalin
patients’ daily pain diaries. A supplemen- lief was evaluated by applying the Kaplan- 300 mg/day 5.99 (4.2–10.4), and pre-
tal responder analysis using two defini- Meier procedure, and comparisons with gabalin 150 mg/day 19.06 (CI for the ab-
tions of response—patients with ⱖ50% placebo were made using the log rank solute risk reduction contains 0,
and with ⱖ30% reductions in mean pain test. rendering the CI for the number needed
scores from baseline—was also per- Safety measures included incidence to treat difficult to interpret).
formed. The studies included several sec- of adverse events, physical and neuro- More patients treated with pregabalin
ondary efficacy measures. End point logic examinations, 12-lead electrocar- reported global health status improve-
mean sleep-interference score was de- diogram, vital signs, weight change, and ments than patients treated with placebo,
rived from daily sleep diaries in which pa- clinical laboratory testing including A1C . as measured by the Patient Global Im-
tients rated daily how much their pain For the pooled analysis, all statistical pression of Change. Eighty percent of pre-
had interfered with their sleep (also done testing of efficacy measures was per- gabalin 600 mg/day patients, 74% of 300
using an 11-point NRS, with 0 ⫽ “pain formed on the intent-to-treat (ITT) pop- mg/day patients, and 65% of 150 mg/day
does not interfere with sleep” to 10 ⫽ ulation. End point mean pain scores patients were improved, compared with
“pain completely interferes with sleep”). using last observation carried forward 54% of placebo patients (300 and 600
Each study included the Patient Global and sleep interference were analyzed mg/day, P ⬍ 0.0001).
Impression of Change, in which pa- with ANCOVA (with a term for baseline In each of the above analyses, both
tients rate their improvement on a values and a term for treatment), whereas BID and TID regimens of 600 mg/day pre-
7-point scale ranging from “very much other secondary efficacy measures (re- gabalin were significantly superior to pla-
worse” to “very much improved” (6 – sponders) were analyzed using a logistic cebo (P ⬍ 0.0001 for all comparisons of
12). These measures were also analyzed regression model (with a term for baseline BID dosing to placebo and TID dosing to
to determine whether there were signif- values and a term for treatment). placebo). For 300 mg/day, only the TID

from baseline in mean pain score at end point based on last observation carried forward analysis. Patient population is comprised of patients who had
both baseline and end point assessments (numbers in some groups are therefore smaller than in the ITT population). Among patients receiving 150,
300, and 600 mg/day pregabalin, 27, 39, and 47%, respectively, reported pain reductions ⱖ50% from baseline to end point, while 22% of placebo
patients reported comparable reductions (pregabalin 300 and 600 mg/day, †P ⬍ 0.0001 vs. placebo). Using the ⱖ30% improvement criterion, a level
of improvement deemed clinically meaningful (31), 43, 55, and 62% of patients treated with 150, 300, and 600 mg/day pregabalin, respectively, were
responders vs. 37% of patients who received placebo (*P ⫽ 0.0455 for 150 mg/day, †P ⫽ 0.0001 for 300, and ‡P ⬍ 0.0001 for 600 mg/day vs.
placebo). D: Survival curve analysis of the time to onset of meaningful pain relief, defined as the first day on which patients demonstrated a sustained
ⱖ1 point in mean pain score where sustained is defined as a ⱖ30% reduction in mean pain score at end point. The median time to onset of a sustained
(ⱖ30% at end point) 1-point improvement was 4 days in patients treated with pregabalin at 600 mg/day, 5 days in patients treated with pregabalin
at 300 mg/day, 13 days in patients treated with pregabalin at 150 mg/day, and 60 days in patients receiving placebo. Hazard ratios were 1.44 for
pregabalin at 150 mg/day (P ⫽ 0.013), 1.84 for pregabalin at 300 mg/day (P ⬍ 0.0001), and 2.26 for pregabalin at 600 mg/day (P ⬍ 0.0001). PGB,
pregablin.

DIABETES CARE, VOLUME 31, NUMBER 7, JULY 2008 1451

Pregabalin in diabetic peripheral neuropathy

Table 1—Common TEAEs and discontinuations occurring in >5% of any treatment group (ordered by greatest percentage of adverse events
in the pregabalin 600 mg/day group)

Pregabalin
Placebo 150 mg/day 300 mg/day 600 mg/day
n (%) Discontinuation n (%) Discontinuation n (%) Discontinuation n (%) Discontinuation
n 557 176 266 511
Adverse event
Dizziness 26 (4.7) 0.7 12 (6.8) 1.1 62 (23.3) 3.4 142 (27.8) 6.8
Peripheral edema* 40 (7.2) 0.5 10 (5.7) 1.1 26 (9.8) 1.5 82 (16.0) 2.7
Somnolence 16 (2.9) 0.4 9 (5.1) 0.6 38 (14.3) 3.0 68 (13.3) 4.3
Weight gain 5 (0.9) 0 8 (4.5) 0 10 (3.8) 0.4 45 (8.8) 1.0
Asthenia 12 (2.2) 0.2 4 (2.3) 0.6 13 (4.9) 2.3 44 (8.6) 2.0
Headache 38 (6.8) 1.1 12 (6.8) 0.6 16 (6.0) 0.8 35 (6.8) 2.5
Dry mouth 7 (1.3) 0 3 (1.7) 0.6 13 (4.9) 0.8 30 (5.9) 1.8
Accidental injury 16 (2.9) 0 4 (2.3) 0 7 (2.6) 0.4 26 (5.1) 0.6
Vertigo 5 (0.9) 0.4 3 (1.7) 0 8 (3.0) 1.5 25 (4.9) 1.4
Nausea 29 (5.2) 0.9 4 (2.3) 0.6 8 (3.0) 1.1 23 (4.5) 2.0
Pain 18 (3.2) 0.4 9 (5.1) 0.6 8 (3.0) 0.4 20 (3.9) 0
Infection 35 (6.3) 0 14 (8.0) 0 23 (8.6) 0.8 17 (3.3) 0.4
Edema 0 0 4 (2.3) 0 13 (4.9) 0.4 10 (2.0) 0
Data are % unless otherwise indicated. *One patient in the 600-mg group had both edema and peripheral edema.

dosing regimen was significantly superior all three pregabalin groups, with 150, nature and were not considered associ-
to placebo (P ⬍ 0.0001 for all compari- 300, and 600 mg/day showing reductions ated with treatment.
sons); however, the 300 mg/day BID dos- of ⫺1.92, ⫺2.32, and ⫺2.62, respec- Over the course of 5 to 13 weeks of
age group was included in only one of the tively, compared with ⫺1.32 for placebo treatment, the incidence of clinically
seven studies. (P ⫽ 0.003 for 150 mg/day and P ⬍ meaningful weight gain (defined using a
Kaplan-Meier analysis revealed that 0.0001 for 300 and 600 mg/day vs. pla- Food and Drug Administration– guided
the median time to onset of sustained cebo). As with change in mean pain score, criterion of ⱖ7% weight increase from
(ⱖ30% improvement from baseline) pain improvement in sleep interference ap- baseline to end point) for pregabalin ver-
relief was 4 days in patients treated with peared to be positively correlated with sus placebo was dose related: 2.01% for
pregabalin at 600 mg/day, 5 days in pa- dosage. pregabalin at 150 mg/day (P ⫽ 0.14 [95%
tients treated with pregabalin at 300 mg/ CI ⫺0.47 to 3.03%]), 2.12% for pregaba-
day, 13 days in patients treated with Safety and tolerability lin at 300 mg/day (P ⫽ 0.04 [⫺0.09 to
pregabalin at 150 mg/day, and 60 days in Incidence of treatment-emergent adverse 2.86%]), and 3.88% for pregabalin at 600
patients receiving placebo (Fig. 2D). events (TEAEs) appeared to be related to mg/day (P ⬍ 0.0001 [1.76 – 4.54%])
Comparison of the pregabalin treatment dosage, with most of the common TEAEs groups, compared with 0.73% for the pla-
groups with placebo by log rank test con- having the greatest incidence among pa- cebo group. The odds of weight gain com-
firmed that time to onset of clinically tients receiving 600 mg/day (Table 1). pared with placebo are 2.3-fold for
meaningful pain relief was statistically There was no consistent pattern in TEAE pregabalin at 150 mg/day (P ⫽ 0.14
significantly more rapid than with pla- incidence rates by dosing regimen, with [0.77– 6.60%]), 2.8-fold for pregabalin at
cebo (P ⬍ 0.0001 for pregabalin doses of some TEAEs, such as peripheral edema 300 mg/day (P ⫽ 0.04 [1.06 –7.47%]),
300 and 600 mg/day and P ⫽ 0.01 for 150 and weight gain, having greater incidence and 6.2-fold for pregabalin at 600 mg/day
mg/day). The median time to onset of sus- in the BID dosing groups relative to the (P ⬍ 0.0001 [2.82–13.67%]). Mean
tained (ⱖ50% improvement from base- TID dosing groups, while other TEAEs, changes in weight from baseline to end
line) pain relief was 6 days in patients such as dizziness and somnolence, oc- point for pregabalin-treated patients ver-
treated with pregabalin at 600 mg/day curred with greater frequency in the TID sus placebo control subjects were 0.76 kg
and 12 days in patients treated with pre- than the BID groups. For all treatment for pregabalin at 150 mg/day (P ⫽ 0.02
gabalin at 300 mg/day. Comparison of the groups, TEAEs were generally mild to [0.08 –1.11 kg]), 1.86 kg for pregabalin at
pregabalin treatment groups with placebo moderate. The discontinuation rate due 300 mg/day (P ⬍ 0.0001 [1.26 –2.14
by log rank test confirmed that time to to adverse events was greatest in the 600 kg]), and 2.04 kg for pregabalin at 600
onset of clinically meaningful pain relief mg/day group (Fig. 1B). Serious TEAEs mg/day (P ⬍ 0.0001 [1.54 –2.22 kg]); the
was statistically significantly more rapid occurred in 3.4, 2.3, and 4.9% of patients mean change was 0.16 kg for placebo.
than with placebo (P ⬍ 0.0001 for pre- receiving 150, 300, and 600 mg/day pre- The incidence of ⱖ7% weight gain by
gabalin doses of 300 and 600 mg/day and gabalin and in 3.4% of patients receiving study duration across all pregabalin doses
P ⫽ NS for 150 mg/day). placebo. The most common serious is as follows: 5 weeks, 2.8%; 8 weeks,
Mean sleep interference scores at end TEAEs reported by both pregabalin and 6.8%; and 12–13 weeks, 7.9%.
point were also significantly improved in placebo patients were of cardiovascular There was a dose-related increase in

1452 DIABETES CARE, VOLUME 31, NUMBER 7, JULY 2008

 Freeman, Durso-DeCruz, and Emir

edema and peripheral edema (Table 1). 1A). Pooling data from all treatment arms 600 mg/day (24). The basis for this differ-
The presence of edema across doses and in this analysis adds to our knowledge of ence in efficacy between disease states is
by severity is as follows: pregabalin at 150 the efficacy, safety, and tolerability of pre- not known, although it may be related, in
mg/day, mild 64.3%, moderate 28.6%, gabalin for treatment of painful DPN. In part, to the permitted use of concomitant
and severe 7.1%; pregabalin at 300 mg/ terms of efficacy, there was an evident pain medication in all postherpetic neu-
day, mild 56.4%, moderate 41%, and se- dose response, with the greatest efficacy ralgia pregabalin clinical trials (24 –26).
vere 2.6%; pregabalin at 600 mg/day, observed among patients treated with 600 However, the 150- and 300-mg/day doses
mild 63.4%, moderate 33%, and severe mg/day. Equally evident from this pooled were used BID in only one study in this
1.1%; and placebo, mild 81%, moderate analysis— but not from examination of pooled analysis (15), and further studies
19%, and severe 0%. the trials individually, as 150 mg/day was are required to definitively address this
In all pregabalin-treated groups, not significantly efficacious in any indi- question.
15.2% had edema or peripheral edema, vidual study—was the observation that The dose-related increase in efficacy
6.0% had a ⱖ7% weight increase, and patients treated with pregabalin at its low- was accompanied by a dose-related in-
2.3% had both weight increase and est effective dosage for chronic neuro- crease in incidence of most adverse
edema. In comparison, in the placebo- pathic pain, 150 mg/day, experienced
events. Similarly, the rate of discontinua-
administered group, 7.3% had edema or statistically significant improvements in
tion due to an adverse event was dose
peripheral edema, 1.5% had a ⱖ7% their pain and pain-related sleep interfer-
weight increase, and 0.2% had both. ence and responded to pregabalin (ⱖ30% related. Dizziness, somnolence, and pe-
There were no clinically meaningful improvement) in proportions signifi- ripheral edema were the most common
changes in laboratory values from base- cantly greater than placebo. adverse events. While there was a consis-
line to end point reported in the studies. Time to onset analysis of the pooled tent increase in the incidence of dizziness
There were neither statistically significant data revealed dose-related, rapid onset of across doses, the incidence of somnolence
nor clinically meaningful changes from durable pain relief. By day 4, 50% of the was similar in the 300- and 600-mg/day
baseline to end point in A1C values (% of subjects taking 600 mg/day had a sus- doses. Examination of adverse events by
total Hb) in pregabalin-treated patients tained (ⱖ30% at end point) 1-point im- dosing regimen, i.e., BID versus TID, for
and in control subjects over 5 to 13 weeks provement in pain score, while a response each pregabalin daily dosage did not re-
of treatment: pregabalin at 150 mg/day, of this magnitude was achieved by day 5 veal any consistent patterns favoring one
0.07% (95% CI ⫺0.07 to 0.24); pregaba- in the 300 mg/day group. All seven stud- regimen over the other. There was a dose-
lin at 300 mg/day, 0.01% (⫺0.01 to ies showed statistically significant differ- related increase in peripheral edema (Ta-
0.26); pregabalin at 600 mg/day, 0.08% ences between pregabalin and placebo ble 1). Edema was not an exclusion
(⫺0.09 to 0.13); and placebo, 0.03% by week 1 (6 –10,12) or week 2 (11); criterion in any study; however, clinical
(⫺0.05 to 0.11). however, these analyses do not necessar- judgment is warranted when pregabalin is
ily imply that the response is clinically used in patients with preexisting edema.
CONCLUSIONS — In this pooled meaningful. This approach also does not Pregabalin was not associated with cardio-
analysis of patients with painful DPN provide insight into the durability of the vascular complications, rarely led to dis-
from seven randomized, controlled trials response in individual patients. Although continuations, and was not associated
spanning the effective dose range, pre- not frequently used in pain therapeutic with laboratory changes suggestive of re-
gabalin was shown to significantly reduce trials (23), the time to onset analysis used nal or hepatic failure. The incidence of
pain associated with DPN. The pooled here provides numerical and graphic (Fig. reported weight gain was not only dose
analysis, in contrast to individual reports, 2D) information that is clinically relevant related but also dependent on duration of
revealed efficacy of the 150-mg dose; for patient and clinician, specifically the exposure. The underlying cause of the
however, efficacy with BID dosing was likelihood of a predetermined clinical re- weight gain is not known and does not
only present with the 600-mg dose. In sponse (the hazard ratio) and the time to appear to be related to the presence of
addition, time to event analysis revealed this response. The analysis in this report peripheral edema. There was no evidence
that pregabalin was associated with a was complicated by the different dose es-
that the weight increase compromised
dose-related, rapid onset of sustained calation schedules of the individual stud-
glycemic control; the pooled analysis
pain relief. ies. Since time to onset was determined
showed no clinically meaningful changes
Several anticonvulsants with diverse from the start of dose escalation and not
mechanisms of action have been sub- the time point when an effective thera- in A1C values in any dose cohort in stud-
jected to large-scale randomized con- peutic dose was attained, the analysis may ies with treatment durations of 5 to 13
trolled trials assessing therapeutic efficacy have overestimated the time to sustained weeks. Long-term studies are warranted
in the treatment of painful DPN. These efficacy. Future studies should incorpo- to address this question.
agents, which include topiramate (14 – rate this analytic technique prospectively. In conclusion, the current pooled
16), lamotrigine (17), oxcarbazepine The present analyses revealed that analysis of seven randomized, controlled
(18,19), and gabapentin (20 –22) have dosing schedule (BID vs. TID) apparently clinical trials in patients with painful DPN
shown varying efficacy in clinical trials. made no meaningful difference for pa- showed that over the effective dose range,
In contrast, pregabalin has shown ef- tients treated with 600 mg/day, as both pregabalin not only significantly reduced
ficacy in six of seven clinical trials. Among regimens were highly statistically signifi- pain associated with DPN but was also
the seven trials, one included 150-, 300-, cant versus placebo. This finding is in associated with rapid onset of sustained
and 600-mg/day treatment arms, two in- contrast to studies of postherpetic neural- pain relief. The improvement in pain re-
cluded two of these dosages, and four in- gia in which BID efficacy was demon- lief was accompanied by a dose-related
cluded only one of these dosages (Fig. strated across a range of doses from 150 to incidence of adverse events.

DIABETES CARE, VOLUME 31, NUMBER 7, JULY 2008 1453

Pregabalin in diabetic peripheral neuropathy

ropathy with pregabalin: a randomized, M: A randomized, placebo-controlled

Acknowledgments — This research was placebo-controlled trial. J Pain 6:253– study of oxcarbazepine in painful diabetic
sponsored by Pfizer. 260, 2005 neuropathy. Acta Neurol Scand 114:177–
Administrative and copyediting support 10. Freynhagen R, Strojek K, Griesing T, 180, 2006
was provided by Adelphi Inc. and also by Whalen E, Balkenohl M: Efficacy of pre-
Pfizer. 19. Beydoun A, Shaibani A, Hopwood M,
gabalin in neuropathic pain evaluated Wan Y: Oxcarbazepine in painful diabetic
in a 12-week, randomised, double-blind, neuropathy: results of a dose-ranging
multicentre, placebo-controlled trial of study. Acta Neurol Scand 113:395– 404,
References flexible- and fixed-dose regimens. Pain 2006
1. Pirart J: Diabetes mellitus and its degen- 115:254 –263, 2005 20. Backonja M, Beydoun A, Edwards KR,
erative complications: a prospective study 11. Tölle T, Freynhagen R, Versavel M, Schwartz SL, Fonseca V, Hes M, La-
of 4,400 patients observed between 1947 Trostmann U, Young JP Jr: Pregabalin for Moreaux L, Garofalo E: Gabapentin for
and 1973. Diabete Metab 3:245–256, relief of neuropathic pain associated with
the symptomatic treatment of painful
1977 diabetic neuropathy: a randomized, dou-
neuropathy in patients with diabetes mel-
2. Davies M, Brophy S, Williams R, Taylor A: ble-blind study. Eur J Pain 12:203–213,
The prevalence, severity, and impact of 2008 litus: a randomized controlled trial. JAMA
painful diabetic peripheral neuropathy in 12. Rosenstock J, Arezzo J, Pauer L, LaMor- 280:1831–1836, 1998
type 2. Diabetes Care 29:1518 –1522, eaux L, Barrett J, Durso-De Cruz E, Pfizer 21. Backonja M, Glanzman RL: Gabapentin
2006 Global R&D: Pregabalin as treatment of dosing for neuropathic pain: evidence
3. Galer BS, Gianas A, Jensen MP: Painful painful diabetic peripheral neuropathy from randomized, placebo-controlled
diabetic polyneuropathy: epidemiology, (DPN): nerve conduction and analgesic clinical trials. Clin Ther 25:81–104, 2003
pain description, and quality of life. Dia- effect in a 13-week double-blind, place- 22. Gorson KC, Schott C, Herman R, Ropper
betes Res Clin Pract 47:123–128, 2000 bo-controlled trial (Poster). Pain 8 (Suppl. AH, Rand WM: Gabapentin in the treat-
4. Boulton AJ, Vinik AI, Arezzo JC, Bril V, 1):S27, 2007 ment of painful diabetic neuropathy: a
Feldman EL, Freeman R, Malik RA, Maser 13. Farrar JT, Young JP Jr, LaMoreaux L, placebo controlled, double blind, cross-
RE, Sosenko JM, Ziegler D: Diabetic neu- Werth JL, Poole RM: Clinical importance over trial. J Neurol Neurosurg Psychiatry
ropathies. Diabetes Care 28:956 –962, of changes in chronic pain intensity mea- 66:251–252, 1999
2005 sured on 11-point numerical pain rating 23. Freeman R, Raskin P, Hewitt DJ, Vor-
5. Dooley DJ, Taylor CP, Donevan S, Feltner scale. Pain 94:149 –158, 2001 sanger GJ, Jordan DM, Xiang J, Rosenthal
D: Ca2⫹ channel alpha2delta ligands: 14. Thienel U, Neto W, Schwabe SK, Vija- NR: Randomized study of tramadol/acet-
novel modulators of neurotransmission. purkar U: Topiramate in painful diabetic aminophen versus placebo in painful di-
Trends Pharmacol Sci 28:75– 82, 2007 polyneuropathy: findings from three dou- abetic peripheral neuropathy. Curr Med
6. Pfizer Protocol No. 1008 – 040: A place- ble-blind placebo-controlled trials. Acta Res Opin 23:147–161, 2007
bo-controlled trial of pregabalin and Neurol Scand 110:221–231, 2004 24. van Seventer R, Feister HA, Young JP, Jr
amitriptyline for treatment of painful 15. Raskin P, Donofrio PD, Rosenthal NR, Stoker M, Versavel M, Rigaudy L: Efficacy
diabetic peripheral neuropathy [article Hewitt DJ, Jordan DM, Xiang J, Vinik AI; and tolerability of twice-daily pregabalin
online], 2007. Available from http:// CAPSS-141 Study Group: Topiramate vs for treating pain and related sleep inter-
www.clinicalstudyresults.org/drugdetails/ placebo in painful diabetic neuropathy: ference in postherpetic neuralgia: a 13-
?drug_name_ id⫽203&sort⫽c.company_ analgesic and metabolic effects. Neurology
week, randomized trial. Curr Med Res
name&page⫽2&drug_id⫽1952. Accessed 63:865– 873, 2004
Opin 22:375–384, 2006
28 June 2007 16. Freeman R, McIntosh KA, Vijapurkar U,
25. Dworkin RH, Corbin AE, Young JP, Jr,
7. Rosenstock J, Tuchman M, La Moreaux L, Thienel U: Topiramate and physiologic
Sharma U: Pregabalin for the treatment of measures of nerve function in polyneu- Sharma U, LaMoreaux L, Bockbrader H,
painful diabetic peripheral neuropathy: a ropathy. Acta Neurol Scand 115:222–231, Garofalo EA, Poole RM: Pregabalin for the
double-blind, placebo-controlled trial. 2007 treatment of postherpetic neuralgia: a ran-
Pain 110:628 – 638, 2004 17. Vinik AI, Tuchman M, Safirstein B, domized, placebo-controlled trial. Neu-
8. Lesser H, Sharma U, La Moreaux L, Poole Corder C, Kirby L, Wilks K, Quessy S, rology 60:1274 –1283, 2003
RM: Pregabalin relieves symptoms of Blum D, Grainger J, White J, Silver M: 26. Sabatowski R, Galvez R, Cherry DA, Jac-
painful diabetic neuropathy: a random- Lamotrigine for treatment of pain associ- quot F, Vincent E, Maisonobe P, Versavel
ized controlled trial. Neurology 63:2104 – ated with diabetic neuropathy: results of M: Pregabalin reduces pain and improves
2110, 2004 two randomized, double-blind, placebo- sleep and mood disturbances in patients
9. Richter RW, Portenoy R, Sharma U, controlled studies. Pain 128:169 –179, with post-herpetic neuralgia: results of a
LaMoreaux L, Bockbrader H, Knapp LE: 2007 randomised, placebo-controlled clinical
Relief of painful diabetic peripheral neu- 18. Grosskopf J, Mazzola J, Wan Y, Hopwood trial. Pain 109:26 –35, 2004

1454 DIABETES CARE, VOLUME 31, NUMBER 7, JULY 2008