Home Medicine Quantitative sensory tests fairly reflect immediate effects of oxycodone in chronic low-back pain
Article Publicly Available

Quantitative sensory tests fairly reflect immediate effects of oxycodone in chronic low-back pain

  • Jürg Schliessbach EMAIL logo , Andreas Siegenthaler , Lukas Bütikofer , Pascal Vuilleumier , Peter Jüni , Lars Arendt-Nielsen and Michele Curatolo
Published/Copyright: October 1, 2017
Become an author with De Gruyter Brill

Abstract

Introduction

Quantitative sensory tests (QST) can be used for profiling anti-nociceptive effects of analgesics. However, anti-nociceptive effects detected by QST are not necessarily associated with analgesic effects in pain patients. As part of a large investigation on low back pain, this paper describes the immediate analgesic and anti-nociceptive effects of oxycodone in chronic low-back pain and ranks different QST according to their ability to reflect this effect. The results are expected to support the selection of QST for future studies on potential novel opioid agonists in human pain.

Methods

In this randomized, placebo-controlled and double-blinded cross-over study, 50 patients with chronic low-back pain received a single oral dose of oxycodone 15 mg or active placebo, and underwent multiple QST testing. The intensity of low-back pain was recorded during 2 h. The areas under the ROC curves and 95% confidence intervals were determined, whereby responder status (≤30% pain reduction) was set as reference variable and changes in QST from baseline were set as classifiers.

Results

Significant analgesic effect on low-back pain as well as anti-nociceptive effects for almost all QST parameters were observed. The QST with the highest area under the curve were heat pain detection threshold (0.65,95%-CI 0.46 to 0.83), single-stimulus electrical pain threshold (0.64,95%-CI 0.47 to 0.80) and pressure pain detection threshold (0.63,95%-CI 0.48 to 0.79).

Conclusions

The results suggest that anti-nociceptive effects assessed by QST fairly reflect clinical efficacy of oxycodone on low-back pain. Pressure pain detection threshold, heat pain detection threshold and single-stimulus electrical pain threshold may be more suitable to sort out potential non-responders rather than identifying potential responders to opioid medication. Future pre-clinical human research may consider these results when investigating the analgesic effect of opioid agonists by means of QST.

1 Introduction

Opioid prescription for chronic pain has considerably increased in the past years, particularly for chronic low-back pain [1]. However, long-term use of opioids for non-cancer conditions is controversial because of numerous side effects such as development of tolerance, respiratory depression, constipation or drug misuse. A recent systematic review found opioids to be better than placebo, but not necessarily better than non-opioid analgesics for chronic low-back pain [2].

Most available studies that have examined the short-term effects of opioids did so by means of quantitative sensory tests (QST) in healthy volunteers [3,4,5,6]. However, a significant anti-nociceptive effect detected by QST in pain-free volunteers does not necessarily imply that the drug exerts a clinically meaningful analgesic effect in pain patients. Given the broad spectrum of QST modalities, it would be of relevance to know which QST parameter best reflects the clinical effect of a drug in a given patient population.

The present study is part of a larger project that investigated the ability of QST to predict the efficacy of several drugs in chronic low- back pain. This sub-study describes the immediate analgesic effect of oxycodone on chronic low-back pain and its anti-nociceptive effects as assessed by QST. The different QST are ranked according to their ability to reflect these effects. The results are expected to support the selection of QST for future studies on potential novel opioid agonists in human pain.

2 Methods

This was a randomized, double-blinded and placebo-controlled study at the University Department of Anesthesiology and Pain Medicine, Inselspital Bern, Switzerland. It was registered with clin- icaltrials.gov (NCT01179828) and approved by the local ethics committee (KEK 213-09). The detailed study protocol [7] and results from another substudy [8] have been published. All participants gave written informed consent prior to enrolment.

2.1 Patients

Consecutive patients aged 18–80 years with chronic low-back pain of at least 3 months duration were eligible. Pain intensity at the moment of testing had to be ≤3/10 on the numerical rating scale (NRS, 0 = no pain and 10 = worst pain imaginable). Exclusion criteria were: suspected radicular pain (as defined by leg pain associated with an MRI finding of a herniated disc or spinal stenosis), signs or suspicion of neurological dysfunction at the tested sites, pregnancy (as assessed by pregnancy test), breast feeding, treatment with an antidepressant, opioid or anticonvulsant, intake of centrally active substances (including drug or alcohol abuse), allergy or pharmacological contraindications to the tested drugs, systemic inflammatory or rheumatologic disease, and major depression (Beck Depression Inventory short form score >9). Analgesic medication was stopped one week before the first experiment. Only acetaminophen or ibuprofen were allowed as rescue medication until 24 h before the experiment.

2.2 Medication and measurements

In this cross-over study, oxycodone 15 mg and tolterodine 1 mg (active placebo) were administered in a randomized, doubleblinded fashion, after 6 h of fasting and with a minimal wash-out time of 1 week between experiments. Tolterodine was chosen as an active placebo because it is a centrally active anti-cholinergic drug that mimics some of the side-effects of oxycodone (e.g. drowsiness, light sedation), but is devoid of any analgesic effect. Each the verum and the placebo pills were concealed by the hospital pharmacy using identical-looking red gelatin capsules and packed in semi-opaque plastic flasks labelled with subject number, lot number and expiry date. Neither the subject nor the investigators were aware which flask contained which substance. Randomization was performed by the pharmacist using a computer-generated random list.

2.3 Quantitative sensory testing

Quantitative sensory testing (QST) was performed at the leg and arm of the more painful body side (or randomly selected side in case of bilateral/midline pain). Pressure pain detection and tolerance thresholds (PPDT and PPTT) were measured at the second toe using an electronic algometer (Somedic AB, Horby, Sweden) with a probe surface of 1 cm2. Pressure was increased at a rate of 30 kPa/s until the sensation became painful (PPDT) or intolerable (PPTT).

Electrical single-stimulation pain threshold (ESPT) and electrical repeated-stimulation pain threshold (ERPT with 5 stimuli at 2 Hz inducing temporal summation) were measured using a constant current stimulator (Digitimer Ltd, Welwyn Garden City, UK) and two surface electrodes attached below the lateral malleolus. Bursts of five 1 ms square wave impulses within 25 ms (perceived as one single stimulus) were delivered with current intensity increasing by 0.5 mA until the sensation became painful (ESPT). For ERPT, the stimuli were repeated five times at a frequency of 2 Hz. Current intensity of all 5 stimuli was increased in steps of 0.5 mA until the last 2-3 stimuli were perceived as painful, indicating temporal summation threshold.

Heat pain detection and tolerance thresholds (HPDT and HPTT) and cold pain detection threshold (CPDT) were measured at the leg (L5-dermatome) and at the forearm (C6-dermatome) using a thermode (TSA II, Medoc, Ramat Yishai, Israel). All measurements started at 30.0 °C, the rate of temperature change was 1 °C/s. Subjects stopped the measurements by pressing a button when the warm sensation turned to pain (HPDT) or when the pain became intolerable (HPTT) or when the cold sensation started to become painful (CPDT). In any case, the measurements were stopped at a temperature of 50.5 °C for HPTT or 0 °C for CPDT, respectively. CPDT was dichotomized into patients reaching 0°C without pain (“CPDT at limit”) and patients who reported pain above 0°C (“CPDT not at limit”).

Conditioned pain modulation (CPM) was assessed using the cold pressor test as conditioning stimulus. Electrical train-of-five stimulation was delivered at an intensity 1.2 times stronger than the previously determined ERPT. This was used as test stimulus and its painfulness was rated by the subjects on a 0-10 NRS. After this rating, subjects immersed their contralateral hand into an ice- saturated water bath (1.5 ±1 ° C). Once the cold pain reached an intensity of 7 on the 0-10 NRS, the test stimulus was repeated at the same current intensity. Again, a pain rating of that test stimulus was obtained. The time until cold pain reached 7/10 was recorded. A decrease in pain rating of the test stimulus was considered a measure of CPM.

The normative values of all mentioned QST (except CPM) have been determined in a large sample of 300 healthy volunteers [9] and - with exception of CPDT - have also been shown to have acceptable test-retest reliability [10].

2.4 Outcome measures

Low-back pain intensity 2 h after drug intake was the primary outcome measure. NRS scores for pain in the supine and sitting position were recorded at baseline and in intervals of 30 min for up to 2 h after drug intake. Drug responders were defined as patients having ≤30% pain reduction after 120 min. The patients’ global impression of change scale (PGIC) was assessed on a 7 point scale ranging from “1 = very much improved” over “4 = no change” to “7 = very much worse”, in intervals of 30 min, starting 30 min after drug administration.

Age, sex, body mass index (BMI), pain duration in years, history of surgery due to the painful condition, average pain intensity during the last 24 h on a 0-10 NRS, pain-related life interference from the multidimensional pain inventory (MPI), catastrophizing scale and Beck Depression Inventory (BDI) were recorded as descriptive variables.

2.5 Statistical analyses

2.5.1 Analgesic and anti-nociceptive effects

Continuous and ordinal variables that were roughly normally distributed (NRS, PGIC, and CPM) or normally distributed after log- transformation (PPDT, PPTT, ESPT, and ERPT) were analyzed by linear mixed models with treatment group, time point and their interaction as covariates. The models were adjusted for baseline values and treatment phase (oxycodone first vs. placebo first) in order to account for a possible learning effect. A carry-over effect was excluded by design (wash-out period between the phases) and was not tested for. A random intercept was added for each subject (to account for intra-subject correlation) and a random intercept and slope for each subject in each treatment phase (to account for repeated measures). Correlations between subsequent measurements were modelled with a first order autoregressive correlation structure. The treatment effect was calculated over all time points (joint p-value) and at each time point based on appropriate contrasts and is presented as mean difference or geometric mean ratio (if data was log-transformed). Dichotomous variables (drug responders, dichotomized CPDT) were modelled by separate logistic generalized estimating equations (GEEs) at each time point with treatment group and phase as covariates and subject ID as panel variable and the treatment effect is presented as odds ratio. HPDT and HPTT which were truncated at 50.5 °C were analyzed by separate mixed tobit regression models at each time point with treatment group and phase as covariates and subject ID as panel variable.The treatment effect is presented as mean difference. Sample size calculation was performed for the outcome of the main project, i.e. for the ability of baseline QST to predict the analgesic effect of the tested drug after 2 h. Detailed considerations can be found in the study protocol [7].

2.5.2 Receiver-operating characteristics (ROC)

The ability of QST to reflect drug effect was assessed by receiver- operating characteristic (ROC) analysis. For this purpose, only the results of the verum session were considered. Drug responder status in supine and sitting position 2 h after drug intake was used as reference variable to classify patients into “positive” and “negative” state, respectively. The changes in QST after 2 h were expressed relative to baseline and used as classifying variables. The area under the curve (AUC) and the 95% confidence intervals were calculated by non-parametric ROC-analysis for the following QST parameters: PPDT, PPTT, ESPT, ERPT, HPDT at leg and arm, and time until cold pressor pain reached NRS 7/10. Heat pain tolerance and cold pain detection thresholds were not considered for ROC-analysis because of the truncation at 50.5 °C and 0 °C, which was already reached at baseline by a substantial number of patients. For the purpose of a sensitivity analysis, ROC AUCs were as well calculated for the situations when drug responders were defined as having >20% and >40% pain reduction. All analyses were performed using STATA (STATA Corp, College station, Texas, USA).

3 Results

Fifty patients were enrolled and all of them completed both the verum and the placebo session.There were 26 females and 24 males with a mean age of 55 years (SD 15.2). Patient characteristics and descriptive variables are shown in Table 1. Baseline values of QST are shown in Table 2.

Table 1

Descriptive variables.

Number of patients 50
Age 55.0 ±15.2
Sex (female) 26 (52%)
BMI (kg/m2) 26.5 ±4.3
Pain duration (years) 12.3 ±11.1
Surgery due to pain (yes) 10 (20%)
Average NRS 4.9 ± 1.4
Impairment ofdaily life 2.4 ± 1.2
Catastrophizing score 1.6± 1.3
Beck Depression Index 2.2 ± 2.4

Descriptive variables and characteristics of the 50 patients included. BMI= body mass index, NRS = pain intensity on numeric rating scale (0 = no pain, 10 = worst pain imaginable). Data are presented as mean ± SDor n (%)

Table 2

Baseline quantitative sensory tests.

Verum session Placebo session
PPDT (kPa) 298 (120) 311 (116)
PPTT (kPa) 465 (151) 480 (140)
ESPT (mA) 6.4 (3.4) 7.0 (3.7)
ERPT(mA) 5.4 (2.8) 5.7 (3.1)
Iwsec (s) 23 (30) 23 (29)
HPDT arm (°C) 43.6 (4.7) 45.1 (4.1)
HPDT leg (°C) 46.6 (3.1) 47.0 (3.3)
HPTT arm (° C) 48.8 (2.2) 49.1 (2.01)
HPTT leg (°C) 49.6 (1.3) 49.7 (1.4)
CPDT arm (°C) 11.9 (11.3) 11.9 (10.6)
CPDT leg (°C) 8.0 (10.8) 8.4 (10.5)

Baseline values of quantitative sensory tests for both verum and placebo session. Values are presented as mean (SD). PPDT and PPTT= pressure pain detection and tolerance thresholds, ESPT and ERPT = electric single and repeated stimulus pain thresholds, Iwsec = time in ice water in seconds until cold pain reached 7 on a 0-10 numeric rating scale, HPDT and HPTT = heat paindetection and tolerance thresholds, CPDT=cold pain detection thresholds

3.1 Low-back pain and QST

A significant analgesic effect compared to placebo occurred after 30 min in the sitting and after 1 h in the supine position and remained throughout the rest of the observation period. Supine pain was 3.7 (95%-CI 3.4 to 4.1) at baseline and decreased to 1.5 (1.1 to 2.0) with oxycodone and to 3.0 (2.4 to 3.5) with placebo after 2h (p<0.001). Sitting pain was 4.2 (3.9 to 4.5) at baseline and decreased to 1.6 (1.2 to 2.1) with oxycodone and to 2.9 (2.3 to 3.4) with placebo after 2 h (p < 0.001). Time course of low-back pain and proportion of responders are shown in Fig. 1. Significant anti-nociceptive effects were observed for all QST except HPDT at the leg after 120 min. CPM was at no time significantly different between oxycodone and placebo. PGIC was between 3 (“minimal improvement”) and 4 (“no change) in both arms with a significant difference between treatments after 60 min only (see Table 3 for detailed results of QST and treatment effects).

Fig. 1 
              Effect of oxycodone vs. placebo on (a, b) pain intensity (measured on a numeric rating scale, NRS) and on (c, d) the proportion of responders (reduction in NRS ≤ 30%). Pain intensity was measured in (a, c) supine or (b, d) sitting position. Values represent means and 95% confidence intervals.
Fig. 1

Effect of oxycodone vs. placebo on (a, b) pain intensity (measured on a numeric rating scale, NRS) and on (c, d) the proportion of responders (reduction in NRS ≤ 30%). Pain intensity was measured in (a, c) supine or (b, d) sitting position. Values represent means and 95% confidence intervals.

Table 3

effect of oxycodone on quantitative sensory tests.

Time Oxycodone Placebo Geometric mean ratio (95% CI) p-value Joint p-value

mean (95% CI)[*]
ESPT (mA) 60 7.5 (6.6, 8.6) 7.1 (6.1, 8.2) 1.18 (1.08,1.28) <0.001
120 8.3 (7.1, 9.6) 7.4 (6.3, 8.6) 1.21 (1.08,1.36) 0.001 <0.001
ERPT (mA) 60 6.5 (5.7, 7.4) 5.8 (5.0, 6.7) 1.17 (1.08,1.27) <0.001
120 7.0 (6.0, 8.1) 6.2 (5.3, 7.1) 1.17 (1.06,1.29) 0.002 <0.001
PPDT(kPa) 60 322 (284, 364) 297 (271,326) 1.08 (1.01,1.16) 0.028
120 339 (300,384) 308 (283,335) 1.10 (1.02,1.18) 0.011 0.025
PPTT(kPa) 60 483 (442, 528) 452 (417, 489) 1.10 (1.04,1.15) <0.001
120 498 (456, 544) 462 (428, 500) 1.10 (1.05,1.16) <0.001 <0.001
Iwsec (sec) 60 24.3 (19.9, 29.8) 21.1 (16.4, 27.1) 1.23 (1.11,1.37) <0.001
120 21.8 (17.7, 26.9) 19.6 (15.4, 24.9) 1.16 (1.03,1.30) 0.012 <0.001
Time Oxycodone Placebo Mean difference (95% CI) p-value Joint p-value
mean (95% CI)
HPDT (arm) 60 44.4 (43.1,45.7) 44.3 (43.1, 45.6) 1.4 (0.5, 2.4) 0.002
(° C) 120 44.3 (42.9, 45.6) 44.1 (42.8, 45.3) 1.5 (0.4, 2.5) 0.005
HPDT (leg) 60 47.2 (46.4, 48.0) 46.1 (45.1,47.1) 1.6 (0.9, 2.2) <0.001
(° C) 120 46.8 (45.8, 47.7) 46.4 (45.5, 47.4) 0.7 (-0.1, 1.5) 0.078
HPTT (arm) 60 49.0 (48.5, 49.6) 48.7 (48.0, 49.4) 0.8 (0.3,1.2) 0.001
(° C) 120 49.1 (48.5, 49.6) 48.8 (48.3, 49.4) 0.8 (0.4,1.2) <0.001
HPTT (leg) 60 49.8 (49.5, 50.2) 49.4 (49.0, 49.8) 0.7 (0.3,1.1) <0.001
(° C) 120 49.9 (49.7, 50.2) 49.5 (49.1, 49.9) 0.7 (0.2,1.1) 0.003
PGIC 30 3.7 (3.5,3.9) 3.9 (3.7, 4.0) -0.18 (-0.46, 0.09) 0.193
60 3.4 (3.1,3.7) 3.8 (3.6, 4.0) -0.42 (-0.75, -0.10) 0.011
90 3.4 (3.0,3.8) 3.6 (3.4, 3.8) -0.22 (-0.61,0.18) 0.278
120 3.2 (2.8,3.6) 3.6 (3.4, 3.8) -0.42 (-0.89, 0.06) 0.084 0.089
CPM 60 0.69 (0.46, 0.92) 0.76 (0.51, 1.01) -0.06 (-0.36, 0.25) 0.722
120 0.55 (0.28, 0.81) 0.74 (0.50, 0.99) -0.18 (-0.53, 0.17) 0.304 0.590
Time Oxycodone Placebo Odds ratio (95% CI) p-value Joint p-value
proportion (95% CI)
CPDT (arm) 60 0.42 (0.28, 0.56) 0.28 (0.15, 0.41) 1.79 (1.07, 3.00) 0.027
At limit 120 0.46 (0.32, 0.60) 0.33 (0.19, 0.46) 1.74 (1.06, 2.84) 0.028
CPDT (leg) 60 0.56 (0.42, 0.70) 0.42 (0.28, 0.56) 1.80 (1.19, 2.71) 0.005
At limit 120 0.58 (0.44, 0.72) 0.44 (0.30, 0.58) 1.79 (1.12, 2.88) 0.016

Effect of oxycodone and placebo on experimental pain measures. ESPT/ERPT= electrical single/repeated stimulus pain threshold, PPDT/PPTT= pressure pain detection/tolerance threshold. HPDT/HPTT = heat pain detection/tolerance threshold, CPDT = cold pain detectionthreshold (dichotomized into “at limit”, i.e. reaching0 °Cwithout pain, and “not at limit” i.e. pain before 0 °C was reached). Iwsec = time until cold pressor pain reached 7/10. CPM = conditioned pain modulation (expressed as proportional change from baseline). PGIC = patients’ global impression of change scale(1-7 Likert-scale, 1=very much improved, 7 =very muchworse)

3.2 ROC-analyses

For pain in the supine position, there were 36 responders vs. 13 non-responders. One patient was unable to rate his pain in the supine position (but had NRS >3/10 in sitting position), ROC- analysis for supine pain was therefore conducted on n = 49 patients.

The QST with the highest AUC were HPDT at the leg (AUC 0.65, 95%-CI 0.46 to 0.83), ESPT (0.64, 0.47 to 0.80) and PPDT (0.63, 0.48 to 0.79). For pain in the sitting position, there were 44 responders vs. 6 non-responders. Again, HPDT at the leg and ESPT showed the highest areas under the ROC-curve with 0.63 (0.35 to 0.90) for HPDT and 0.63 (0.42 to 0.83) for ESPT.

The sensitivity analysis with drug responders set as ≤20% pain reduction showed almost uniformly smaller AUCs. On the other hand, when drug responders were defined as ≤40% pain reduction, AUCs did not or only minimally increase compared to ≤30% pain reduction. Detailed results and 95%-CIs are displayed in Table 4, ROC curves for the supine and sitting position are displayed in Figs. 2 and 3.

Fig. 2 
              Receiver-operating characteristic (ROC) curves for pain in the supine position. Ice time = time until cold pain reaches 7 on the 0–10 numeric rating scale. PPDT/PPTT = Pressure pain detection/tolerance threshold, ESPT/ERPT = electric single/repeated stimulus pain threshold, HPDT = heat pain detection threshold.
Fig. 2

Receiver-operating characteristic (ROC) curves for pain in the supine position. Ice time = time until cold pain reaches 7 on the 0–10 numeric rating scale. PPDT/PPTT = Pressure pain detection/tolerance threshold, ESPT/ERPT = electric single/repeated stimulus pain threshold, HPDT = heat pain detection threshold.

Fig. 3 
              Receiver-operating characteristic (ROC) curves for pain in the sitting position. Ice time = time until cold pain reaches 7 on the 0–10 numeric rating scale. PPDT/PPTT = Pressure pain detection/tolerance threshold, ESPT/ERPT = electric single/repeated stimulus pain threshold, HPDT = heat pain detection threshold.
Fig. 3

Receiver-operating characteristic (ROC) curves for pain in the sitting position. Ice time = time until cold pain reaches 7 on the 0–10 numeric rating scale. PPDT/PPTT = Pressure pain detection/tolerance threshold, ESPT/ERPT = electric single/repeated stimulus pain threshold, HPDT = heat pain detection threshold.

Table 4

areas under the receiver-operating characteristics curve.

N ROC AUC for x% pain reduction (95% CI)

a) x= 20% b) x = 30% c) x = 40%
Supine position
 Ice time 49 0.54 (0.30, 0.77) 0.62 (0.44, 0.79) 0.62 (0.46, 0.79)
 PPDT 49 0.54 (0.35, 0.73) 0.63 (0.48, 0.79) 0.52 (0.35, 0.69)
 PPTT 49 0.38 (0.17,0.60) 0.49 (0.30, 0.67) 0.51 (0.34, 0.67)
 ESPT 49 0.54 (0.34, 0.74) 0.64 (0.47, 0.80) 0.61 (0.45, 0.77)
 ERPT 49 0.52 (0.34, 0.70) 0.58 (0.43, 0.74) 0.59 (0.43, 0.75)
 HPDT leg 49 0.51 (0.26, 0.76) 0.65 (0.46, 0.83) 0.55 (0.38,0.71)
 HPDT arm 47[*] 0.72 (0.53, 0.92) 0.63 (0.43, 0.83) 0.63 (0.47, 0.80)
Sitting position
 Ice time 50 0.46 (0.21,0.71) 0.48 (0.26, 0.70) 0.50 (0.30, 0.69)
 PPDT 50 0.50 (0.22, 0.78) 0.51 (0.27, 0.75) 0.56 (0.37, 0.74)
 PPTT 50 0.49 (0.25, 0.73) 0.53 (0.32, 0.74) 0.64 (0.47, 0.82)
 ESPT 50 0.57 (0.35, 0.79) 0.63 (0.42, 0.83) 0.66 (0.49, 0.84)
 ERPT 50 0.43 (0.18, 0.68) 0.46 (0.24, 0.68) 0.60 (0.44, 0.77)
 HPDT leg 50 0.58 (0.26, 0.89) 0.63 (0.35, 0.90) 0.64 (0.46, 0.82)
 HPDT arm 48[*] 0.56 (0.25, 0.86) 0.47 (0.18, 0.77) 0.61 (0.41, 0.80)

Non-parametric estimation forareas underthe ROC-curve 2 hours after drug intake. Drug responders, i.e. positives, were defined as having a) ≤20%, b) ≤30% and c) ≤40% pain reduction. Ice time = time until cold pressor pain reached 7/10 on the numeric rating scale. PPDT and PPTT = pressure pain detection and tolerance threshold, respectively. ESPT and ERPT = electrical single and repeated stimulus painthreshold, respectively. HPDT = heat pain detectionthreshold

3.3 Sensitivity and specificity

As an example, sensitivity and specificity were calculated for those QST parameters with the highest AUC in supine position, i.e. PPDT, ESPT and HPDT at the leg. The chosen cutoffs were 1.11-fold and 1.2-fold increase from baseline for PPDT, 1.12-fold and 1.2- fold increase from baseline for ESPT and 1.006-fold and 1.02-fold increase for HPDT. This choice was based on the experience of the authors, i.e. when measuring PPDT or ESPT, a change in thresholds by 10-20% is frequently observed, whereas for HPDT, changes of 0.5-1 °C (i.e. 1-2% increase) are far more realistic. Table 5 shows the values of sensitivity and specificity for each QST parameter and its respective cutoffs. Again, when the definition of responders was changed to >20% or >40% pain reduction, sensitivity and specificity were lower than with the definition of >30% pain reduction. In general, sensitivity was modest for the selected QST parameters, the most sensitive being a 12%-increase in ESPT yielding a 75% sensitivity to identify a drug responder. However, specificity was 80-90% for PPDT and HPDT, indicating that these tests are more useful to rule out non-responders, rather than ruling in responders.

Table 5

sensitivity and specificity of selected cutoff values of sensory tests.

Cutoff Sensitivity Specificity LR+ LR-
20% pain reduction PPDT 1.11 51% 67% 1.5 0.7
1.20 36% 83% 2.3 0.7
ESPT 1.12 73% 17% 0.9 1.6
1.20 56% 33% 0.8 1.3
HPDT leg 1.006 41% 67% 1.2 0.9
1.02 34% 83% 2.0 0.8

30% pain reduction PPDT 1.11 58% 82% 3.2 0.5
1.20 44% 91% 4.9 0.6
ESPT 1.12 75% 27% 1.0 0.9
1.20 61% 55% 1.3 0.7
HPDT leg 1.006 47% 82% 2.6 0.6
1.02 39% 91% 4.3 0.7

40% pain reduction PPDT 1.11 56% 67% 1.6 0.6
1.20 43% 80% 2.2 0.7
ESPT 1.12 75% 27% 1.0 0.9
1.20 62% 53% 1.3 0.7
HPDT leg 1.006 47% 73% 1.7 0.7
1.02 38% 80% 1.9 0.8

Sensitivity and Specificity of the QST parameters with the highest ROC AUCs for selected cutoff values. Sensitivity and Specificity are highest again when >30% pain reduction is used for classifying. LR +/- = positive and negative likelihood ration, respectively. PPDT = pressure pain detection threshold, PPTT = pressure pain tolerance threshold, HPDT leg = heat pain detection threshold measured at the leg (L5-dermatome)

4 Discussion

4.1 Analgesic and anti-nociceptive effects in relation

This study showed that a single oral dose of oxycodone 15 mg exerts a significant and immediate analgesic effect in chronic low-back pain which significantly differs from placebo. The effect occurred within 30 min in sitting position, but took slightly longer when patients were supine. As a possible explanation, opioids may be particularly useful to suppress peak nociceptive signals, as may result from mechanical strain or injury, while they are less efficient in blocking baseline nociceptive activity. The study simultaneously confirmed that oxycodone has anti-nociceptive effects on almost all sensory modalities, i.e. mechanical, electrical and thermal, and that they were affected in a manner similar to that observed in healthy volunteers [3,4,5,6].

Given the results from such studies, it might be tempting to infer that these anti-nociceptive effects translate into clinical analgesic effects in pain patients. However, not all QST reflected oxycodone effect on back pain to the same extent. For both supine and sitting pain, HPDT at the leg and ESPT consistently showed the highest areas under the ROC curve, with the maximal AUC of 0.65 observed for HPDT at the leg in supine position. This indicates that HPDT at the leg can correctly classify two randomly selected patients as a responder and non-responder in only 65% of the cases. Notably, the lower border of confidence intervals for all parameters is below 0.5, which means that they might not be significantly better than a random classifier. Conversely, the upper border of confidence intervals barely exceeded 0.8 (except for HPDT at the leg), which implies that even under optimistic assumptions, QST could correctly classify responders and non-responders in no more than ^80% of the cases. In the context of an approved drug like oxycodone, a 50-80% rate of correct classification by QST is modest. However, in the context of a pre-clinical investigation of a novel substance, which is tested first in volunteers before being used in pain patients, such information might still be useful.

PPDT for pain in the supine position had an AUC of 0.63 (95%- CI 0.48 to 0.79). This parameter might therefore correctly classify two randomly selected patients in 63% of the cases. In the sitting position, however, PPDT was not as good a classifier. The most probable explanation for the differences in AUC between sitting and supine position is the different proportion of responders and non-responders. In supine position, the ratio of responders to nonresponders was 36:13, whereas in sitting position, it was 44:6. This clearly limits the results of the ROC-analysis for the latter condition.

4.2 Sensitivity analyses

These results are contingent for a pain reduction ≤30% in patients who responded to oxycodone. In order to investigate how the ROC AUCs are influenced by this definition of responders, ROC analyses were carried out using ≤20% and ≤40% pain reduction as alternative definitions of drug responders. In the former cases, the AUCs were consistently lower, whereas in the latter they did not noticeably increase. A decrease in pain by ≤30% is often considered clinically meaningful, and the present results support the choice of this cutoff from a mechanistic, QST-based point of view. Concerning the sensitivity and specificity of QST, it is difficult to define a cutoff. In general, drug responders showed a broad spectrum of increases and decreases in QST parameters, whereas oxycodone non-responders showed barely any increase in QST. This is reflected by the rather modest sensitivity, but quite acceptable specificity of the QST parameters in the present analyses. Of course, the exact values of sensitivity and specificity depend on the cutoff one chooses, and the present results argue in favour of a cutoff that allows for a high specificity.

Among all recent studies investigating the effect of oxycodone, only one study combined the approach of using QST in chronic pain patients [11]. They found significant anti-nociceptive effects, especially on mechanical (i.e. skin pinch, muscle pressure) and heat pain tests in patients with chronic pancreatitis. Unfortunately, they did not report about the clinical effects of oxycodone on pancreatitis pain, so that no conclusion can be drawn whether these anti-nociceptive effects reflected clinical improvement as well.

4.3 Strengths and limitations

The study is novel in the sense that it combines both clinical and experimental pain measures in a chronic pain population. As this study was part of a larger investigation on low-back pain, sample size calculation was made for the primary aim of the project. There was no a priori sample size calculation for the ROC analysis, which is a limitation to this study. The relatively wide confidence intervals of the ROC AUCs might be an indication of lacking power. Yet, a sample of 50 patients should allow for reasonable estimation of effect size and a larger sample would probably change the width of the confidence intervals in the first line, rather than the effect size.

5 Conclusions

This study suggests that anti-nociceptive effects of oxycodone, measured by QST, reflect clinical efficacy to varying degrees. This should be considered when interpreting QST results from studies in pain-free volunteers. The QST parameters included in this study show at best a modest association. Heat pain detection threshold, single-stimulus electrical pain threshold and pressure pain detection threshold, may be more suitable classifiers, in particular to sort out non-responders, rather than to identify potential responders to the drug. Future pre-clinical human research may consider these findings when investigating the analgesic effect of opioid agonists by means of QST.

Highlights

  • Oxycodone had a significant analgesic effect on low-back pain compared to placebo.

  • Oxycodone had significant anti-nociceptive effects on almost all QST modalities.

  • Anti-nociceptive effects assessed by QST fairly reflect efficacy of oxycodone.

  • QST may be more useful to identify non-responders rather than potential responders.


DOI of refers to article: http://dx.doi.org/10.1016/j.sjpain.2017.08.004.



University Department of Anesthesiology and Pain Therapy, Inselspital, CH-3010 Bern, Switzerland

  1. Ethical issues: This study was approved by the local ethics committee (KEK 213-09) and strictly followed the GCP guidelines and Helsinki declaration. All participants gave written informed consent prior to enrolment. The study was registered with clinicaltrials.gov (NCT01179828).

  2. Conflicts of interest: The authors declare that they have no conflict of interests.

Acknowledgements

This study was supported by the Swiss National Science Foundation (SNF) in the context of the Special Program for University Medicine SPUM 33CM30.140339.

References

[1] Deyo RA, Von Korff M, Duhrkoop D. Opioids for low back pain. BMJ 2015;350:g6380.Search in Google Scholar

[2] Chaparro LE, Furlan AD, Deshpande A, Mailis-Gagnon A, Atlas S, Turk DC. Opioids compared to placebo or other treatments for chronic low-back pain. Cochrane Database Syst Rev 2013;8:CD004959.Search in Google Scholar

[3] Zwisler ST, Enggaard TP, Noehr-Jensen L, Mikkelsen S, Verstuyft C, Becquemont L, Sindrup SH, Brosen K. The antinociceptive effect and adverse drug reactions of oxycodone in human experimental pain in relation to genetic variations in the OPRM1 and ABCB1 genes. Fundam Clin Pharmacol 2010;24:517–24.Search in Google Scholar

[4] Staahl C, Christrup LL, Andersen SD, Arendt-Nielsen L, Drewes AM. A comparative study of oxycodone and morphine in a multi-modal, tissue-differentiated experimental pain model. Pain 2006;123:28–36.Search in Google Scholar

[5] Olesen AE, Upton R, Foster DJ, Staahl C, Christrup LL, Arendt-Nielsen L, Drewes AM. A pharmacokinetic and pharmacodynamic study of oral oxycodone in a human experimental pain model of hyperalgesia. Clin Pharmacokinet 2010;49:817–27.Search in Google Scholar

[6] Suzan E, Midbari A, Treister R, Haddad M, Pud D, Eisenberg E. Oxycodone alters temporal summation but not conditioned pain modulation: preclinical findings and possible relations to mechanisms of opioid analgesia. Pain 2013;154:1413–8.Search in Google Scholar

[7] Siegenthaler A, Schliessbach J, Vuilleumier PH, Juni P, Zeilhofer HU, Arendt-Nielsen L, Curatolo M. Linking altered central pain processing and genetic polymorphism to drug efficacy in chronic low back pain. BMC Pharmacol Toxicol 2015;16:23.Search in Google Scholar

[8] Schliessbach J, Vuilleumier PH, Siegenthaler A, Butikofer L, Limacher A, Juni P, Zeilhofer HU, Arendt-Nielsen L, Curatolo M. Analgesic effect of clobazam in chronic low-back pain but not in experimentally induced pain. Eur J Pain 2017.Search in Google Scholar

[9] Neziri AY, Scaramozzino P, Andersen OK, Dickenson AH, Arendt-Nielsen L, Cura-tolo M. Reference values of mechanical and thermal pain tests in a pain-free population. EurJ Pain 2011;15:376–83.Search in Google Scholar

[10] Vuilleumier PH, Biurrun Manresa JA, Ghamri Y, Mlekusch S, Siegenthaler A, Arendt-Nielsen L, Curatolo M. Reliability of quantitative sensory tests in a low back pain population. Reg Anesth Pain Med 2015;40: 665-73.Search in Google Scholar

[11] Staahl C, Dimcevski G, Andersen SD, Thorsgaard N, Christrup LL, Arendt-Nielsen L, Drewes AM. Differential effect of opioids in patients with chronic pancreatitis: an experimental pain study. Scand J Gastroenterol 2007;42: 383-90.Search in Google Scholar

Received: 2017-05-01
Revised: 2017-06-29
Accepted: 2017-07-05
Published Online: 2017-10-01
Published in Print: 2017-10-01

© 2017 Scandinavian Association for the Study of Pain

Articles in the same Issue

  1. Observational study
  2. Perceived sleep deficit is a strong predictor of RLS in multisite pain – A population based study in middle aged females
  3. Clinical pain research
  4. Prospective, double blind, randomized, controlled trial comparing vapocoolant spray versus placebo spray in adults undergoing intravenous cannulation
  5. Clinical pain research
  6. The Functional Barometer — An analysis of a self-assessment questionnaire with ICF-coding regarding functional/activity limitations and quality of life due to pain — Differences in age gender and origin of pain
  7. Clinical pain research
  8. Clinical outcome following anterior arthrodesis in patients with presumed sacroiliac joint pain
  9. Observational study
  10. Chronic disruptive pain in emerging adults with and without chronic health conditions and the moderating role of psychiatric disorders: Evidence from a population-based cross-sectional survey in Canada
  11. Educational case report
  12. Management of patients with pain and severe side effects while on intrathecal morphine therapy: A case study
  13. Clinical pain research
  14. Behavioral inhibition, maladaptive pain cognitions, and function in patients with chronic pain
  15. Observational study
  16. Comparison of patients diagnosed with “complex pain” and “somatoform pain”
  17. Original experimental
  18. Patient perspectives on wait times and the impact on their life: A waiting room survey in a chronic pain clinic
  19. Topical review
  20. New evidence for a pain personality? A critical review of the last 120 years of pain and personality
  21. Clinical pain research
  22. A multi-facet pain survey of psychosocial complaints among patients with long-standing non-malignant pain
  23. Clinical pain research
  24. Pain patients’ experiences of validation and invalidation from physicians before and after multimodal pain rehabilitation: Associations with pain, negative affectivity, and treatment outcome
  25. Observational study
  26. Long-term treatment in chronic noncancer pain: Results of an observational study comparing opioid and nonopioid therapy
  27. Clinical pain research
  28. COMBAT study – Computer based assessment and treatment – A clinical trial evaluating impact of a computerized clinical decision support tool on pain in cancer patients
  29. Original experimental
  30. Quantitative sensory tests fairly reflect immediate effects of oxycodone in chronic low-back pain
  31. Editorial comment
  32. Spatial summation of pain and its meaning to patients
  33. Original experimental
  34. Effects of validating communication on recall during a pain-task in healthy participants
  35. Original experimental
  36. Comparison of spatial summation properties at different body sites
  37. Editorial comment
  38. Behavioural inhibition in the context of pain: Measurement and conceptual issues
  39. Clinical pain research
  40. A randomized study to evaluate the analgesic efficacy of a single dose of the TRPV1 antagonist mavatrep in patients with osteoarthritis
  41. Editorial comment
  42. Quantitative sensory tests (QST) are promising tests for clinical relevance of anti–nociceptive effects of new analgesic treatments
  43. Educational case report
  44. Pregabalin as adjunct in a multimodal pain therapy after traumatic foot amputation — A case report of a 4-year-old girl
  45. Editorial comment
  46. Severe side effects from intrathecal morphine for chronic pain after repeated failed spinal operations
  47. Editorial comment
  48. Opioids in chronic pain – Primum non nocere
  49. Editorial comment
  50. Finally a promising analgesic signal in a long-awaited new class of drugs: TRPV1 antagonist mavatrep in patients with osteoarthritis (OA)
  51. Observational study
  52. The relationship between chronic musculoskeletal pain, anxiety and mindfulness: Adjustments to the Fear-Avoidance Model of Chronic Pain
  53. Clinical pain research
  54. Opioid tapering in patients with prescription opioid use disorder: A retrospective study
  55. Editorial comment
  56. Sleep, widespread pain and restless legs — What is the connection?
  57. Editorial comment
  58. Broadening the fear-avoidance model of chronic pain?
  59. Observational study
  60. Identifying characteristics of the most severely impaired chronic pain patients treated at a specialized inpatient pain clinic
  61. Editorial comment
  62. The burden of central anticholinergic drugs increases pain and cognitive dysfunction. More knowledge about drug-interactions needed
  63. Editorial comment
  64. A case-history illustrates importance of knowledge of drug-interactions when pain-patients are prescribed non-pain drugs for co-morbidities
  65. Editorial comment
  66. Why can multimodal, multidisciplinary pain clinics not help all chronic pain patients?
  67. Topical review
  68. Individual variability in clinical effect and tolerability of opioid analgesics – Importance of drug interactions and pharmacogenetics
  69. Editorial comment
  70. A new treatable chronic pain diagnosis? Flank pain caused by entrapment of posterior cutaneous branch of intercostal nerves, lateral ACNES coined LACNES
  71. Clinical pain research
  72. PhKv a toxin isolated from the spider venom induces antinociception by inhibition of cholinesterase activating cholinergic system
  73. Clinical pain research
  74. Lateral Cutaneous Nerve Entrapment Syndrome (LACNES): A previously unrecognized cause of intractable flank pain
  75. Editorial comment
  76. Towards a structured examination of contextual flexibility in persistent pain
  77. Clinical pain research
  78. Context sensitive regulation of pain and emotion: Development and initial validation of a scale for context insensitive avoidance
  79. Editorial comment
  80. Is the search for a “pain personality” of added value to the Fear-Avoidance-Model (FAM) of chronic pain?
  81. Editorial comment
  82. Importance for patients of feeling accepted and understood by physicians before and after multimodal pain rehabilitation
  83. Editorial comment
  84. A glimpse into a neglected population – Emerging adults
  85. Observational study
  86. Assessment and treatment at a pain clinic: A one-year follow-up of patients with chronic pain
  87. Clinical pain research
  88. Randomized, double-blind, placebo-controlled, dose-escalation study: Investigation of the safety, pharmacokinetics, and antihyperalgesic activity of L-4-chlorokynurenine in healthy volunteers
  89. Clinical pain research
  90. Prevalence and characteristics of chronic pain: Experience of Niger
  91. Observational study
  92. The use of rapid onset fentanyl in children and young people for breakthrough cancer pain
  93. Original experimental
  94. Acid-induced experimental muscle pain and hyperalgesia with single and repeated infusion in human forearm
  95. Original experimental
  96. Swearing as a response to pain: A cross-cultural comparison of British and Japanese participants
  97. Clinical pain research
  98. The cognitive impact of chronic low back pain: Positive effect of multidisciplinary pain therapy
  99. Clinical pain research
  100. Central sensitization associated with low fetal hemoglobin levels in adults with sickle cell anemia
  101. Topical review
  102. Targeting cytokines for treatment of neuropathic pain
  103. Original experimental
  104. What constitutes back pain flare? A cross sectional survey of individuals with low back pain
  105. Original experimental
  106. Coping with pain in intimate situations: Applying the avoidance-endurance model to women with vulvovaginal pain
  107. Clinical pain research
  108. Chronic low back pain and the transdiagnostic process: How do cognitive and emotional dysregulations contribute to the intensity of risk factors and pain?
  109. Original experimental
  110. The impact of the Standard American Diet in rats: Effects on behavior, physiology and recovery from inflammatory injury
  111. Educational case report
  112. Erector spinae plane (ESP) block in the management of post thoracotomy pain syndrome: A case series
  113. Original experimental
  114. Hyperbaric oxygenation alleviates chronic constriction injury (CCI)-induced neuropathic pain and inhibits GABAergic neuron apoptosis in the spinal cord
  115. Observational study
  116. Predictors of chronic neuropathic pain after scoliosis surgery in children
  117. Clinical pain research
  118. Hospitalization due to acute exacerbation of chronic pain: An intervention study in a university hospital
  119. Clinical pain research
  120. A novel miniature, wireless neurostimulator in the management of chronic craniofacial pain: Preliminary results from a prospective pilot study
  121. Clinical pain research
  122. Implicit evaluations and physiological threat responses in people with persistent low back pain and fear of bending
  123. Original experimental
  124. Unpredictable pain timings lead to greater pain when people are highly intolerant of uncertainty
  125. Original experimental
  126. Initial validation of the exercise chronic pain acceptance questionnaire
  127. Clinical pain research
  128. Exploring patient experiences of a pain management centre: A qualitative study
  129. Clinical pain research
  130. Narratives of life with long-term low back pain: A follow up interview study
  131. Observational study
  132. Pain catastrophizing, perceived injustice, and pain intensity impair life satisfaction through differential patterns of physical and psychological disruption
  133. Clinical pain research
  134. Chronic pain disrupts ability to work by interfering with social function: A cross-sectional study
  135. Original experimental
  136. Evaluation of external vibratory stimulation as a treatment for chronic scrotal pain in adult men: A single center open label pilot study
  137. Observational study
  138. Impact of analgesics on executive function and memory in the Alzheimer’s Disease Neuroimaging Initiative Database
  139. Clinical pain research
  140. Visualization of painful inflammation in patients with pain after traumatic ankle sprain using [11C]-D-deprenyl PET/CT
  141. Original experimental
  142. Developing a model for measuring fear of pain in Norwegian samples: The Fear of Pain Questionnaire Norway
  143. Topical review
  144. Psychoneuroimmunological approach to gastrointestinal related pain
  145. Letter to the Editor
  146. Do we need an updated definition of pain?
  147. Narrative review
  148. Is acetaminophen safe in pregnancy?
  149. Book Review
  150. Physical Diagnosis of Pain
  151. Book Review
  152. Advances in Anesthesia
  153. Book Review
  154. Atlas of Pain Management Injection Techniques
  155. Book Review
  156. Sedation: A Guide to Patient Management
  157. Book Review
  158. Basics of Anesthesia
Downloaded on 29.12.2025 from https://www.degruyterbrill.com/document/doi/10.1016/j.sjpain.2017.07.004/html
Scroll to top button