Greater Conditioned Pain Modulation Is Associated With Enhanced Morphine Analgesia in Healthy Individuals and Patients with Chronic Low Back Pain

Abstract

Objectives:

Conditioned pain modulation (CPM) protocols index magnitude of descending pain inhibition. This study evaluated whether degree of CPM, controlling for CPM expectancy confounds, was associated with analgesic and subjective responses to morphine, and whether chronic pain status or sex moderated these effects.

Methods:

Participants included 92 individuals with chronic low back pain (CLBP) and 99 healthy controls, none using daily opioid analgesics. In a crossover design, participants attended two identical laboratory sessions during which they received either intravenous morphine (0.08 mg/kg) or saline placebo before undergoing evoked pain assessment. In each session, participants engaged in ischemic forearm and heat pain tasks, and a CPM protocol combining ischemic pain (conditioning stimulus) and heat pain (test stimulus). Placebo-controlled morphine outcomes were derived as differences in pain and subjective effects across drug conditions.

Results:

In hierarchical regressions controlling for CPM expectancies, greater placebo condition CPM was associated with less subjective morphine unpleasantness (p=.001) and greater morphine analgesia (p’s<.05) on both the ischemic pain task (VAS Pain Intensity and Unpleasantness) and heat pain task (VAS Pain Intensity, McGill Pain Questionnaire-Sensory and Present Pain Intensity [PPI] subscales). There was no moderation by sex or CLBP status, except for the ischemic PPI outcome for which a significant 2-way interaction (p<.05) was noted, with men showing a stronger positive relationship between CPM and morphine analgesia than women.

Discussion:

Result suggest that CPM might predict analgesic and subjective responses to opioid administration. Further evaluation of CPM as an element of precision pain medicine algorithms may be warranted.

Introduction

Conditioned pain modulation (CPM) is a phenomenon in which exposure to a noxious conditioning stimulus reduces the experience of pain from a second test stimulus applied simultaneously to an anatomically remote body region¹. CPM is believed to reflect activation of descending pain inhibitory mechanisms^2,3. Protocols to evaluate CPM, although varying in the specific combinations of conditioning and test stimuli used, appear to have at least good reliability⁴, and are increasingly included in published studies to quantify altered functioning in central pain modulatory pathways. A meta-analysis of this growing literature suggests that patients with chronic pain have reduced endogenous ability to inhibit pain, as reflected in responses to CPM protocols⁵.

A number of studies have addressed the question of whether use of opioid analgesics alters CPM. Although results of some studies suggest that opioid analgesic use is associated with reduced CPM^6–9, others report no effect^10,11, and still others suggest that opioid analgesics may even increase CPM^12,13. In contrast, the inverse question of whether baseline degree of CPM is associated with responsiveness to opioid analgesics has received little attention, despite the fact that this is a highly relevant issue for precision pain medicine which seeks to target treatments to those who are most likely to benefit¹⁴. To our knowledge, only two prior studies have specifically addressed this issue. This work, conducted in relatively small samples of healthy participants¹⁵ and individuals with chronic low back pain¹⁶, suggested that magnitude of baseline CPM was not associated with subsequent level of oxycodone analgesia as indexed by laboratory evoked pain tasks. Although the few available opioid studies are negative, other research indicates that CPM predicts efficacy of non-opioid medications for pain management, such as duloxetine¹⁷ and nonsteroidal anti-inflammatories¹⁸, suggesting that CPM may have potential as a pharmacological response predictor.

The current study further explored the potential association between CPM and opioid analgesic (intravenous morphine) responses using data from a larger project that has previously been described¹⁹. This sample included males and females both with and without chronic low back pain, allowing us to directly test the potential moderating effects of both sex and chronic pain status on hypothesized associations between CPM and morphine responses. Because prior work suggested that often overlooked expectancy effects may significantly influence the extent of CPM observed^20–22, we also controlled for CPM expectancies to evaluate associations between CPM and morphine responses independent of potential expectancy bias. Finally, given previous work suggesting that analgesic effects of morphine may parallel to some degree its subjective non-pain effects (e.g., drug liking; desire to take the drug again)^23,24, we also examined whether CPM was associated with subjective morphine responses. Based on our earlier work suggesting that individuals with reduced endogenous pain modulatory capacity exhibit greater analgesic and subjective reinforcing responses to opioids²⁴, we hypothesized that lower CPM would be associated with greater morphine responsiveness.

Materials and Methods

Design

The parent study on which the current findings are based was a double-blind, placebo-controlled crossover design with intravenous administration of an opioid analgesic (morphine), an opioid antagonist (naloxone), and saline placebo across three separate sessions¹⁹. Order of drug administration was randomized and counterbalanced. Because the focus of this report is on associations between CPM in the intact state (placebo) and morphine responses, data regarding naloxone condition pain responses are not detailed in the interests of space (see Bruehl et al.¹⁹). This study was a multisite study, with identical data collection procedures and equipment employed in a coordinated fashion at two sites (Vanderbilt University Medical Center and Rush University Medical Center).

Participants

The total sample of 191 individuals included 92 with chronic low back pain (CLBP) and 99 without chronic pain (controls). All participants were recruited either through on-line advertisements on the Vanderbilt e-mail recruitment system, the Rush Pain Clinic, advertisements in local print media, or posted flyers. General criteria for participation included age between 18–55; no self-reported history of cardiovascular disease, hypertension, liver or kidney disorders, posttraumatic stress disorder, bipolar disorder, psychotic disorder, diabetes, seizure disorder, or alcohol or drug dependence; no use of anti-hypertensive medications; and no daily use of opioid analgesics (with absence of recent use confirmed via urine opiate screen before each laboratory study session). As in our past studies^25–27, additional inclusion criteria for the CLBP group were chronic daily low back pain of at least 3 months duration with an average past month severity of at least 3 on a 0–10 verbal numeric pain intensity scale. Individuals with chronic pain related to malignancy, autoimmune disorders, or fibromyalgia were excluded. Potential participants who were pregnant (determined by urine pregnancy screens) were also excluded to avoid unknown effects of naloxone on the fetus. In the CLBP group, 54.3% displayed pain in a possible radicular distribution. Use of as-needed opioids (but no use within 3 days of laboratory assessment) was reported by 14 CLBP group participants (15.2%). No control participants were using as-needed opioids, and antidepressant use in both groups was low (controls = 1.0%, CLBP = 4.3%).

Study Drug

The opioid analgesic used in this study was morphine sulfate, the prototypic mu opioid receptor agonist. As in similar laboratory acute pain studies with morphine²⁸, the current study employed a dosage of 0.08 mg/kg (in 20ml normal saline). This dosage (approximately 7mg for an average sized male) was selected because it was judged to be sufficient to produce analgesia, but low enough to avoid ceiling effects that might obscure key individual differences in morphine responding. Peak morphine activity is achieved within approximately 15 minutes²⁹. In the placebo condition, participants received 20mL normal saline. The investigational pharmacy at each institution prepared and provided the study drug in blinded fashion to the study nurses.

Evoked Pain Tasks

After receiving the assigned study drug in each session, participants engaged in two laboratory evoked pain tasks in fixed sequence, followed by a conditioned pain modulation (CPM) protocol. First, participants underwent an ischemic pain task based on procedures described by Maurset et al.³⁰, similar to our past studies^25–27. Participants engaged in two minutes of dominant forearm muscle exercise using a hand dynamometer at 50% of their maximal grip strength as determined at the beginning of the laboratory procedures. Then they were asked to raise their dominant forearm over their head for 15 sec. A manual BP cuff was then inflated on the participant’s dominant biceps to 200 mmHg SBP, the arm was lowered, and the cuff remained inflated until tolerance was reached, up to a maximum of 8 min. Pain ratings were then immediately obtained.

The second laboratory evoked pain task was a heat pain task using a Medoc TSAII NeuroSensory Analyzer (Medoc US., Minneapolis, MN). This equipment was used to assess heat pain responses using an ascending method of limits protocol. First, four heat pain threshold trials were conducted (for use in determining stimulus intensity in the CPM protocol detailed below). In these trials, the probe started at an adaptation temperature of 32°C and increased at a rate of 0.5°C/sec until the participant pressed a button on a computer mouse to indicate that the stimulus had begun to feel “painful”. Four heat pain tolerance trials were then conducted, with each trial conducted sequentially at one of four different non-overlapping sites on the non-dominant ventral forearm. An interval of 30 sec between successive stimuli was employed. For each pain tolerance trial, the probe started at an adaptation temperature of 40°C and increased at a rate of 0.5°C/sec until the participant indicated maximum tolerance had been reached. Pain ratings were obtained immediately following the final tolerance trial. CPM was unrelated to heat pain tolerance morphine outcomes (results not detailed). The maximum stimulus temperature possible was 51°C due to an automatic hardware cutoff in the TSAII device to ensure participant safety.

CPM Protocol

The CPM protocol combined the heat pain (test stimulus) and ischemic (conditioning stimulus) tasks described above. As illustrated in Figure 1, participants experienced repeated exposure to the heat pain test stimulus (non-dominant forearm) before and then during exposure to the ischemic pain conditioning stimulus (dominant bicep). The CPM protocol began with a pre-conditioning phase, involving 5 minutes of repeated, brief (3 sec) heat pain stimuli delivered at 30 sec intervals at an intensity of 1.2 times the previously-determined heat pain threshold. The participant rated the intensity of each stimulus using a 0 (No Pain) to 100 (Worst Possible Pain) verbal numeric rating scale. Then, participants engaged in 2 min of dominant forearm muscle exercise, exsanguination, and inflation of a blood pressure cuff to 200mmHg on the dominant upper arm as described in the evoked pain task description above. Finally, the conditioning phase began, in which participants provided ratings of heat pain intensity in response to the same brief heat pain stimuli as in the pre-conditioning phase, with stimuli delivered every 30 seconds for 5 minutes while the blood pressure cuff remained inflated. Prior to completing the first CPM assessment, participants received an oral description of the procedure that was complemented by a visual aid. As described in France et al.²¹, after confirming that they understood the procedure, participants were asked “What do you think that having the blood pressure cuff inflated on one arm will do to the heat pain that you feel on your other arm?” Possible responses included a) it will not change the pain, b) it will decrease the pain, or c) it will increase the pain. If participants chose either b or c, then they were asked to indicate what percentage change they expected on a scale of 0–100%. These data were then used to compute CPM expectancies ranging from −100% to +100%.

Figure 1.

Conditioned pain modulation (CPM) protocol.

Measures

Pain Outcomes

Immediately after completing pain tolerance assessment for each evoked pain task, participants completed the McGill Pain Questionnaire-Short Form (MPQ)³¹ to describe the pain experienced during the task. The MPQ is a well-validated measure that allows separate assessment of the sensory (MPQ-S) and affective (MPQ-A) qualities of pain³¹. The MPQ also includes a Present Pain Intensity (MPQ-PPI) numeric scale of overall pain intensity, on which categorical ratings are provided using a 0 (no pain) to 5 (excruciating) scale³¹. In addition, the MPQ includes a 100mm visual analog pain intensity scale (VAS) anchored with No Pain and Worst Possible Pain. Participants were similarly asked to describe the degree of pain unpleasantness associated with each task using a 100mm VAS unpleasantness scale, anchored with Not Unpleasant at All and The Most Unpleasant Possible.

Subjective Opioid Effects

Non-analgesic subjective effects of acute opioid administration were assessed as in several prior opioid studies by Zacny and colleagues^32–34 using a 26-item VAS opioid effects rating scale. This assessment approach has been used in our prior work as well^24,35,36. Items on the VAS opioid effects scale (anchored with not at all and extremely) tap into both the positive and negative subjective (e.g., cognitive, emotional) effects of opioids, including effects relevant to abuse potential (feeling good, elated, comfortable, etc.). To reduce the number of measures examined and mitigate against elevated Type I error rate, VAS opioid effect scale responses were used to create 3 opioid response factors identified based on principal components analysis in our prior work³⁶. Each factor score in the current work reflected the sum of observed values for the top 3 items loading on each factor as in Bruehl et al.²⁴. The three factor scores (and items contained within them) were labelled: Sedation (dreamy, coasting, floating), Unpleasantness (down, anxious, feeling bad), and Euphoria (stimulated, elated, having pleasant thoughts).

Procedure

All procedures were conducted at the Vanderbilt General Clinical Research Center or a dedicated research room at the Rush University Pain Center. All procedures were approved by the Institutional Review Boards at the respective institutions. After providing informed consent, participants completed a packet of questionnaires, including information regarding demographics and chronic pain. Individuals then participated in identical experimental sessions across drug conditions, with sessions scheduled one week apart and at the same time of day to control for variance due to circadian rhythms.

Participants remained seated upright in a comfortable chair throughout all laboratory procedures. During each session, participants initially completed a 10-min seated rest period, after which an indwelling venous cannula was inserted into the dominant arm by a trained research nurse under physician supervision. After a 30-min resting adaptation period, participants received (via the cannula) either saline placebo or morphine, with order of drug administration across sessions randomly determined and counterbalanced.

After a 15-min rest period to allow peak drug activity to be achieved, participants completed the VAS subjective opioid effects measure. They then engaged in the ischemic task using the procedures described above, after which the MPQ and VAS pain measures were immediately completed to describe responses to this evoked pain stimulus. Then, participants engaged in the heat pain tolerance task, with the MPQ and VAS pain measures completed immediately afterward to describe the pain just experienced during the heat pain stimuli. Finally, the CPM procedure described above was carried out. All participants remained in the lab under observation for 2 hours after peak drug activity had been achieved to allow drug effects to remit, after which they were released to a responsible adult.

Statistical Analyses

Analyses were conducted using IBM SPSS for Windows Version 24 (SPSS Inc., Chicago, IL). Prior to conducting analyses, CPM effects were derived as mean heat pain ratings during the pre-conditioning phase minus mean ratings observed during the conditioning phase. Thus, greater positive CPM values reflected more descending pain inhibition. Given the study hypothesis that CPM effects in the intact state would be associated with extent of morphine analgesia, analyses used CPM effects in the placebo condition as the independent variable. To examine the effects of morphine while controlling for placebo effects, morphine effect variables reflecting the difference between placebo and morphine condition evoked pain responses were derived separately for each targeted outcome (MPQ-S, MPQ-A, MPQ-PPI, VAS intensity, VAS unpleasantness), such that greater positive morphine effect values indicated greater morphine analgesia. A similar approach was used to derive measures quantifying the placebo-controlled effects of morphine on subjective state, with larger positive values for the three VAS opioid effects factor scores indicating greater sedation, unpleasantness, and euphoria following morphine administration.

Group differences in demographic characteristics were examined using t-tests for continuous measures and Chi-Square analyses for categorical variables. Preliminary analyses used Pearson correlations (r) to examine zero-order correlations between extent of placebo condition CPM observed and placebo-controlled morphine effect outcomes. Although outside the scope of this study, we note that naloxone administration did not have a significant overall effect on magnitude of CPM within either controls or the CLBP group (p’s>.15). Primary analyses were a series of hierarchical multiple linear regressions. Each regression focused on a different morphine effect outcome as the dependent measure. To isolate CPM effects, all hierarchical regressions entered main effects of participant type (control vs. CLBP) and sex in step one, as well as drug administration order (a potential confound) and CPM expectancy (ranging from −100% to +100%). The latter was included to control for significant expectancy effects on CPM noted in prior work by our group and others^20–22. Next, the main effect of placebo condition CPM was entered in step two. Two-way interactions (Type X CPM, Sex X CPM, Sex X Type) were entered in step three, with the three-way Type X Sex X CPM interaction entered in step four. The source of significant interactions was dissected using simple effects analyses, with these effects displayed graphically by solving regressions computed by sex for hypothetical low and high CPM values (−1 SD and + 1 SD from the mean CPM value) as described by Aiken and West³⁷. Main effects were only interpreted in the absence of significant interactions in the model. To further provide a conservative overall test of study hypotheses, follow-up analyses paralleling primary analyses were conducted using the mean standardized (z-score) morphine effect value across all evoked pain measures examined for each evoked pain task (i.e., mean ischemic morphine effect and mean thermal morphine effect). A two-tailed p<.05 criterion for significance was used in all analyses.

Results

Participant Characteristics

As shown in Table 1, comparison of participant characteristics across the two groups revealed that controls and CLBP participants differed significantly only on age, with the CLBP group being slightly older. Although not significant, the CLBP group had a higher proportion of female participants than the control group, with similar racial and ethnic distributions across the groups. The CPM procedure elicited a small mean decrease in perceived pain, with wide variability [Mean (SD) = 2.7 (9.22)]. Mean (SD) responses to both evoked pain tasks during the placebo and morphine conditions are summarized by sex and participant type in Table 2. Stability of pain responses for the two evoked pain tasks between the placebo and morphine conditions was, unsurprisingly, only moderate (intraclass correlations ranged from 0.49 – 0.68).

Table 1.

Characteristics of participants in the control and chronic low back pain (CLBP) groups.

Measure	Controls (n=99)	CLBP (n=92)
Sex (% female)	50.5	62.0
Race:
Caucasian	60.6	57.6
African-American	33.3	34.8
Ethnicity:
Non-Hispanic	97.0	95.6
Age (years)*	33.0±9.14	36.5±11.57
VAS Chronic Pain Intensity (0–100)		56.9±22.97
Pain Duration (median, in months)		87.6

^*p<.05

Note: Summary statistics are presented as percentages or means (± SD). VAS Chronic Pain Intensity was a retrospective measure of overall past month chronic pain intensity.

Table 2.

Mean (±SD) placebo and morphine condition pain rating values in male and female participants by participant type.

	Controls				CLBP
	Females		Males		Females		Males
Measures	Placebo	Morphine	Placebo	Morphine	Placebo	Morphine	Placebo	Morphine
ISC MPQ-S	5.1 ± 4.07	4.0 ± 2.85	6.5 ± 5.20	5.7 ± 5.07	9.8 ± 7.50	7.8 ± 7.02	9.4 ± 7.13	9.1 ± 7.80
ISC MPQ-A	0.5 ± 1.20	0.4 ± 0.78	0.6 ± 1.39	0.8 ± 1.47	1.7 ± 2.60	1.0 ± 1.88	1.7 ± 2.09	1.6 ± 2.21
ISC MPQ-PPI	1.8 ± 0.95	1.4 ± 0.84	1.9 ± 0.97	1.7 ± 1.05	2.54 ± 1.10	1.9 ± 1.25	2.3 ± 0.93	2.1 ± 1.28
ISC VAS Intensity	36.4 ± 23.55	31.6 ± 22.24	37.1 ± 24.3	33.8 ± 24.76	55.5 ± 26.11	43.6 ± 29.45	55.3 ± 19.70	46.0 ± 25.79
ISC VAS Unpleasant	43.7 ± 26.46	33.4 ± 22.99	42.7 ± 25.28	36.9 ± 26.14	58.9 ± 26.35	46.7 ± 30.76	62.4 ± 20.87	51.5 ± 27.19
Therm MPQ-S	5.0 ± 3.60	4.0 ± 2.99	6.1 ± 5.35	5.0 ± 4.49	7.6 ± 6.07	6.4 ± 5.79	8.2 ± 6.21	8.0 ± 7.10
Therm MPQ-A	0.4 ± 1.15	0.4 ± 1.07	0.5 ± 1.75	0.4 ± 1.35	0.9 ± 1.93	0.8 ± 1.49	1.2 ± 1.87	0.9 ± 1.53
Therm MPQ-PPI	2.1 ± 1.09	2.0 ± 0.68	2.4 ± 1.14	2.3 ± 1.06	2.5 ± 1.05	2.2 ± 1.23	2.6 ± 0.98	2.5 ± 1.12
Therm VAS Intensity	47.8 ± 25.29	44.0 ± 23.87	52.0 ± 24.94	49.4 ± 27.07	58.3 ± 25.98	48.8 ± 28.71	63.6 ± 20.01	59.4 ± 25.41
Therm VAS Unpleasant	48.5 ± 27.02	42.8 ± 26.63	52.5 ± 24.33	42.8 ± 49.26	57.7 ± 26.51	51.3 ± 30.14	66.6 ± 20.16	62.2 ± 25.84

Note: ISC = Ischemic pain task; MPQ-S, MPQ-A, and MPQ-PPI = McGill Pain Questionnaire-Short Form Sensory, Affective, and Present Pain Intensity subscales respectively; VAS = Visual Analog Scale; Therm = Thermal pain task.

Zero-Order Correlations

Table 3 displays the zero-order correlations between placebo condition CPM and placebo-controlled morphine effects derived for evoked pain response and subjective morphine effect outcomes. Results indicated significant associations between greater CPM and greater analgesic effects for morphine, with the largest effect noted for the MPQ-PPI measure on the thermal evoked pain task. Significant associations in the same direction were also noted for VAS pain intensity across both evoked pain tasks. In terms of subjective morphine effects, greater CPM was found to be associated with significantly lower reports of subjective unpleasantness in response to morphine administration, but no associations were observed with either sedating or euphoric effects of morphine.

Table 3.

Zero-order correlations between placebo-condition CPM and placebo-controlled morphine responses.

Morphine Effect Measure	Correlation (r)
Ischemic MPQ-S	0.09
Ischemic MPQ-A	−0.03
Ischemic MPQ-PPI	0.25***
Ischemic VAS Intensity	0.18*
Ischemic VAS Unpleasantness	0.17*
Thermal MPQ-S	0.18*
Thermal MPQ-A	−0.09
Thermal MPQ-PPI	0.31***
Thermal VAS Intensity	0.22**
Thermal VAS Unpleasantness	0.07
Unpleasantness Factor	−0.24***
Sedation Factor	0.08
Euphoria Factor	0.06

^*p<.05

^**p<.01

^***p<.001

Note: MPQ-S, MPQ-A, and MPQ-PPI = McGill Pain Questionnaire-Short Form Sensory, Affective, and Present Pain Intensity subscales respectively. The CPM variable was derived so that greater negative values indicate greater descending inhibition. Morphine analgesic effects were derived so that more negative values indicate greater morphine analgesia. Factor scores indexing morphine subjective effects were derived such that greater positive scores indicated greater subjective effects with morphine.

Ischemic Task Analgesic Outcomes

For the ischemic pain task, hierarchical regressions revealed that CPM was associated with morphine analgesia beyond any influence of sex, chronic pain status, drug order, and CPM expectancies as indexed by both VAS intensity (beta = 0.160, p=.031) and VAS unpleasantness (beta = 0.162, p=.030). In both cases, greater placebo-condition CPM was associated with greater morphine analgesia. For the ischemic task MPQ-PPI outcome, a two-way interaction was observed between sex and CPM (beta = 0.209, p=.0122). Simple effects analyses revealed that the association between greater CPM and larger MPQ-PPI morphine analgesic effects in men (beta = 0.401, p<.001) was significantly larger than the positive association noted in women (beta = 0.187, p=.058). Figure 2 graphically displays the source of this interaction. All other CPM main and interaction effects were nonsignificant (all p’s>.10). The potential confounding influence of CPM expectancies on morphine effect outcomes appeared to be nontrivial (e.g., MPQ-A: p=.028). A conservative analysis using the mean standardized morphine effect across all ischemic task outcomes indicated a significant overall unique association between CPM (beta = 0.152, p=.039) and morphine analgesic responses, with no significant interactions (p’s>.10).

Figure 2.

Source of 2-way interaction between sex and conditioned pain modulation (CPM) on morphine effects for the ischemic task McGill Pain Questionnaire- Present Pain Intensity (MPQ-PPI) measure. Higher positive morphine effect values indicate greater analgesia.

Although not a focus of the study, we note that step one of the analysis for the ischemic MPQ-A outcome revealed sex differences, with women reporting greater morphine analgesia than men (beta = −0.158, p=.028).

Thermal Task Analgesic Outcomes

For the thermal pain task, hierarchical regressions revealed significant effects of CPM on morphine analgesic responses independent of the influence of sex, chronic pain status, drug order, and CPM expectancies on the MPQ-S (beta = 0.196, p=.008), MPQ-PPI (beta = 0.321, p<.001), and VAS intensity measures (beta = 0.218, p=.003). In each case, greater placebo-condition CPM was associated with greater morphine analgesia. No other main or interaction effects were noted (all p’s>.10). CPM expectancies significantly influenced morphine effects for the MPQ-A outcome (p=.023). A conservative analysis using the mean standardized morphine effect across all thermal task outcomes indicated a significant unique association between CPM and degree of morphine analgesia (beta = 0.159, p=.033), with no significant interactions (p’s>.10).

Subjective Morphine Effects

Hierarchical regressions were also used to examine associations between placebo condition CPM and the subjective effects of morphine, independent of the influence of sex, chronic pain status, drug order, and CPM expectancies. Consistent with the zero-order correlation analyses, results revealed an independent effect of CPM on ratings of unpleasant feelings after administration of morphine (beta = −0.253, p=.001), indicating that greater CPM was associated with less subjective unpleasantness. No main effects or interactions were noted for associations between CPM and ratings of sedation or euphoria (all p’s >.10).

Discussion

Based on our recent findings suggesting that individuals with lower endogenous pain inhibitory capacity experience greater analgesic effects with morphine²⁴, we hypothesized that individuals displaying a smaller CPM response, a purported marker of impaired descending pain inhibition, would also report greater morphine analgesia. Contrary to hypotheses, individuals exhibiting a larger CPM effect reported the greatest analgesia with intravenous morphine across two different laboratory evoked pain tasks. This association was independent of CPM expectancies, which have been shown to exert a significant influence on magnitude of CPM effects^20–22. Moreover, with one exception, the positive association between CPM and magnitude of morphine analgesia was similar in males and females and individuals both with and without chronic pain. For the ischemic task MPQ-PPI measure only, a two-way interaction indicated that associations between greater CPM and enhanced morphine analgesia were significantly stronger in males than in females. Although this effect was isolated, its plausibility is indirectly supported by a prior meta-analysis that indicated sex differences in morphine responding³⁸. Associations between CPM and magnitude of morphine analgesia were paralleled to some degree by associations between CPM and subjective morphine responses. Greater CPM was associated with fewer subjective unpleasant effects of morphine, potentially implying that CPM might identify a clinically-relevant subgroup of individuals who obtain both better analgesia and lower side effects with opioid analgesics.

In the context of the existing literature, the current findings may be informative regarding mechanisms underlying associations between CPM and opioid analgesic responses. Our past work in two separate samples indicated that lower endogenous opioid inhibitory activity is predictive of greater morphine analgesia^19,24. Our most recent work further indicated that it was the patients with both low endogenous opioid activity and low endocannabinoid levels that displayed the greatest morphine analgesia²⁴. To the extent that smaller CPM effects are an indicator of impaired endogenous inhibitory capacity, which could in part be related to low endogenous opioid activity^39–41, our prior findings suggested that low CPM might be expected to be associated with greater morphine analgesia. The fact that results indicated the opposite pattern of association suggest that endogenous opioid CPM mechanisms did not contribute to the observed associations. While determining the mechanisms of CPM was not the focus of this study, we did note that observed CPM effects in our study were unchanged by administration of naloxone, a finding further arguing against CPM-related analgesia in the current work being related to release of endogenous opioids. Although this absence of endogenous opioid mechanisms of CPM could be specific to the combination of stimuli used in this study⁴², this finding is consistent with other literature suggesting that CPM can occur via non-opioid mechanisms^42–44. Such mechanisms might include, for example, activation of blood pressure-linked analgesic systems^43,45,46 or serotonin pathways^47–49. The current work indicates that it may be the non-opioid elements of CPM that account for associations with analgesic responses to opioid medications.

Significant associations observed between CPM and opioid analgesic responses in the present study are inconsistent with the two prior negative studies addressing this topic^15,16. These discrepancies could be due to several methodological differences. Both of the earlier studies reporting no association between CPM and subsequent opioid analgesic responses were conducted in much smaller samples, and both used oxycodone as the study drug. Moreover, there are a variety of combinations of test and conditioning stimuli used in CPM protocols. The current work used a combination of heat pain (test stimulus) and ischemic pain (conditioning stimulus) whereas related prior work assessed CPM using combinations of heat pain-cold pressor pain¹⁵ and electrical pain-cold pressor pain¹⁶. Therefore, it is possible that the positive current results may be due not only to a larger sample, but also to different CPM stimuli and the use of a different opioid analgesic (morphine rather than oxycodone). Indeed, different opioid analgesics are known to activate distinct signaling pathways⁵⁰, hence differential associations would not be implausible. In addition, the CPM associations observed in the current study may have been influenced to some extent by the statistical control of CPM expectancies, the influence of which in some cases was nontrivial. The potential confounding impact of CPM expectancies on magnitude of CPM effects has been reported in the literature^20–22, but is not yet commonly addressed in published CPM studies.

Potential precision medicine applications of CPM may merit additional exploration. A prior qualitative review of the literature⁵¹ concluded that quantitative sensory testing measures, including CPM, were not robust enough predictors of analgesic drug responses to justify their clinical use. However, few studies using CPM were available at that time, and those that were available reflected small sample sizes (n<60 participants). The current findings, in a large sample (n=191) of both pain-free individuals and those with chronic pain, suggest that exploring the value of CPM as a potential predictor of clinical analgesic responses may be worthwhile. A possible clinical limitation is that the magnitude of observed associations between CPM and opioid analgesic responses was relatively modest, and therefore the value of including CPM in precision medicine algorithms may hinge on whether it predicts variance not accounted for by other predictors in such algorithms (e.g., negative affect)^52–54. Whether opioid analgesic use itself enhances CPM^12,13 or reduces it^6–9 might also impact on the clinical utility of CPM for predicting longer-term opioid responses. Additional work is required before any precision pain medicine applications of CPM can be determined conclusively.

Several other limitations of this study are also noted. Because of the design of the parent study, CPM was assessed only in the context of saline placebo administration, and it is possible that placebo effects could have significantly influenced the observed pattern of associations between CPM and morphine responses. Generalizability to situations in which CPM is assessed in the absence of placebo administration is unknown. Another limitation is that drug administration order was randomized, so the placebo condition in which CPM was assessed in some cases was conducted before and in other cases after the morphine administration session. Thus, based on the current results, CPM cannot be said to be truly predictive of subsequent morphine responses (i.e., temporal precedence), but rather, these findings simply reflect associations suggesting that future predictive work is warranted. It is noted however that the current findings reflect associations observed when drug administration order was statistically-controlled. Another limitation is that there was no formal adjustment for multiple comparisons in analyses in order to avoid Type II error, although we note that results were similar when collapsing across all measures (using combined Z scores) for each pain stimulus, and that results for the thermal evoked pain task would be significant even when applying a highly conservative bonferroni adjustment. Finally, the mean magnitude of the observed CPM effect was small, and some evidence suggests that larger CPM effects may be observed with other stimulus combinations⁵⁵. Whether similar results would be obtained using different CPM protocols is unknown.

In summary, the current study indicates that greater magnitude of CPM is associated with greater analgesic effects and lower subjective unpleasantness in response to administration of a weight-adjusted dose of morphine. These findings were independent of CPM expectancies, and appeared to be relatively consistent across males and females and individuals with and without chronic pain. Endogenous opioid mechanisms appear unlikely to contribute to these effects. The potential value of including CPM as a predictor in precision pain medicine algorithms may warrant further exploration.