Psychosomatics 49:191-198, May-June
doi: 10.1176/appi.psy.49.3.191
© 2008 Academy of Psychosomatic Medicine
Treating the Physical Symptoms of Depression With Second-Generation Antidepressants: A Systematic Review and Metaanalysis
Erin E. Krebs, M.D., M.P.H.,
Bradley N. Gaynes, M.D., M.P.H.,
Gerald Gartlehner, M.D., M.P.H.,
Richard A. Hansen, Ph.D., R.Ph.,
Patricia Thieda, M.A.,
Laura C. Morgan, M.A., M.P.H.,
Angela DeVeaugh-Geiss, M.S., and
Kathleen N. Lohr, Ph.D.
Received October 18, 2007; accepted November 28, 2007. From The Center on Implementing Evidence-Based Practice, Roudebush Veterans Affairs Medical Center; Center for Health Services and Outcomes Research, Regenstrief Institute, Inc.; the Dept. of Medicine, Indiana University School of Medicine, Indianapolis, IN; Dept. of Psychiatry, University of North Carolina, Chapel Hill, NC; Ludwig Boltzmann Institute for Health Technology Assessments, Vienna, Austria; Division of Pharmaceutical Outcomes and Policy, School of Pharmacy, University of North Carolina, Chapel Hill, NC; Cecil G. Sheps Center for Health Services Research, University of North Carolina, Chapel Hill, NC; Dept. of Epidemiology, School of Public Health, University of North Carolina, Chapel Hill, NC; and RTI International, Research Triangle Park, NC. Send correspondence and reprint requests to Erin E. Krebs, M.D., M.P.H., HSRD (11H), Roudebush VAMC, 1481 W. 10th St., Indianapolis, IN 46202. e-mail: krebse{at}iupui.edu
© 2008 The Academy of Psychosomatic Medicine
|
ABSTRACT
|
BACKGROUND: Approximately two-thirds of patients with depression experience physical pain symptoms. Coexisting pain complicates the treatment of depression and is associated with worse depression outcomes. OBJECTIVE: The authors reviewed the effect of newer antidepressants on pain in patients with depression. METHOD: The authors searched systematically for trials of second-generation antidepressants that enrolled depression patients and reported pain outcomes, pooling changes on the pain visual-analog scale (VAS), using random-effects models. RESULTS: Eight trials were eligible. Pooled analysis of head-to-head trials showed no difference in VAS between duloxetine and paroxetine. Both drugs were superior to placebo. CONCLUSION: The authors found insufficient evidence to support the choice of one second-generation antidepressant over another in patients with pain accompanying depression.
Key Words: Antidepressants • SSRIs • SNRIs • Physical Symptoms
|
INTRODUCTION
|
Approximately two-thirds of patients with depression experience physical pain symptoms.1 Coexisting pain complicates the treatment of depression and is associated with worse depression outcomes.2–4 Therefore, treatments that address both depression and pain are highly desirable.
Biologic pathways shared by pain and depression have inspired the use of antidepressant medications for pain syndromes with and without coexisting depression.5,6 Efficacy for treating pain syndromes, such as fibromyalgia, diabetic neuropathy, migraine, and chronic back pain, is best established for tricyclic antidepressants (TCAs).7–9 Second-generation antidepressants, such as selective serotonin reuptake inhibitors (SSRIs), are often favored in practice because of their tolerability and safety.10–12 However, the evidence for their efficacy in pain is relatively weak.1,13
The most recent systematic review of antidepressants for treatment of coexisting depression and pain, published in 2003, found only one relevant randomized, controlled trial (RCT) of a second-generation antidepressant available in the United States.1 Since then, interest in antidepressant treatment of coexisting pain and depression has increased, especially with the U.S. Food and Drug Administration (FDA) approval of duloxetine.
Like venlafaxine, duloxetine enhances the activity of both serotonin and norepinephrine. The mechanism of these serotonin and norepinephrine reuptake-inhibitors (SNRIs) has been hypothesized to confer analgesic effects independent of antidepressant action.13–15 Indeed, duloxetine has received FDA approval for major depressive disorder and for diabetic neuropathic pain, but it has not received a distinct indication for their comorbidity. In light of the recent attention to antidepressant treatment of coexisting depression and pain, we conducted a systematic review to answer the following question: What is the effect of second-generation antidepressants on pain in patients with depression?
|
METHOD
|
This systematic review is based on work done for a comprehensive, comparative-effectiveness review of second-generation antidepressants for depression treatment in adults, conducted for the Agency for Healthcare Research and Quality (AHRQ).16 Key questions for the comprehensive review were developed through a process involving the public, the Scientific Resource Center for AHRQ’s Effective Health Care program, and stakeholder groups (http://effectivehealthcare.ahrq.gov/aboutUs/index.cfm). The question addressed in this article was developed prospectively by the authors.
Literature Search and Study Selection
We searched MEDLINE, Embase, the Cochrane Library, and the International Pharmaceutical Abstracts from 1980 to April 2007. The search strategy combined keywords and Medical Subject Heading terms for depression, drug interactions, and adverse events with a list of the 12 second-generation antidepressants approved for use in the United States (Table 1). Searches were limited to "human" and "English language."
We also hand-searched reference lists of pertinent review articles and letters to the editor, and searched the FDA’s Center for Drug Evaluation and Research (CDER) database to identify unpublished research submitted to the FDA. The Scientific Resource Center contacted pharmaceutical manufacturers and invited them to submit dossiers, including citations. We received dossiers from three pharmaceutical companies (Eli Lilly and Company, GlaxoSmithKline, and Wyeth). We also contacted authors to clarify reports and identify additional unpublished data.
Two persons (including at least one senior investigator) independently reviewed each abstract and full-text article for eligibility. Any disagreements were resolved by discussion and consensus or, when necessary, by consultation of a third party (another senior investigator). Reviewers used a structured data-abstraction form to ensure consistency in appraisal.
Studies were considered for exclusion if they did not meet preestablished eligibility criteria for intervention design, duration, patient population, or sample size (Table 1). For the question addressed in this article, we included studies that reported any pain-specific outcome,17 such as pain severity and pain-related functioning.
Methods and results of the comprehensive comparative effectiveness review are publicly available.16 Eight studies that included pain outcomes met eligibility criteria for this review.
Quality Assessment
Two reviewers independently assigned a quality rating for each study (Good, Fair, or Poor), using predefined criteria based on those developed by the U.S. Preventive Services Task Force and the National Health Service Centre for Reviews and Dissemination.18,19 The elements of quality we assessed included randomization, allocation concealment, similarity of groups at baseline, presence of intention-to-treat analysis, and acceptable rates of loss to follow-up. Studies that met all criteria were rated as being Good quality. Those that met some criteria and had no serious flaws were rated Fair quality. Those with serious flaws (inadequate randomization, lack of intent-to-treat analysis,  40% overall attrition, 15% differential attrition) were rated Poor.
Data Analysis
If data were sufficient, we conducted a metaanalysis of the weighted mean difference (WMD) of changes on the visual-analog scale (VAS) for pain, using random-effects models. For studies with more than one arm per drug, values for the arms were combined. Relative-risk metaanalysis was not possible because no studies reported pain-response rates. We assessed heterogeneity with the I2 statistic.20 However, given the small number of studies in our metaanalysis, this test has low power and should be interpreted cautiously. Stata 9.1 (College Station, TX) and StatsDirect, Ltd., Version 2.4.5 (Altrincham, UK), were used for statistical analyses.
|
RESULTS
|
Study Characteristics
We identified seven published trials21–27 and one unpublished trial (Eli Lilly and Co.: Clinical Study Summary: Study F1J-MC-HMAT, Study Group A: Eli Lilly and Co., 2004; available at http://www.lillytrials.com/results_files/cymbalta/cymbalta_summary_4091a.pdf) that met eligibility criteria. Four trials, including the Eli Lilly trial, compared duloxetine with paroxetine and placebo;24,26,27 three compared duloxetine with placebo;21–23 and one compared paroxetine with placebo.25 All included trials were sponsored by a pharmaceutical manufacturer—seven by Eli Lilly and Company, the manufacturer of Cymbalta (duloxetine),21–24,26,27 and one by SmithKline Beecham Pharmaceuticals, the maker of Paxil (paroxetine).25 Details of included trials are summarized in Table 2.
All included trials enrolled a narrowly-selected population of patients, excluding those with psychiatric disorders other than depression, serious medical illness, substance abuse, and treatment-resistant depression. Only two studies, one of duloxetine versus placebo and one of paroxetine versus placebo, required the presence of pain at baseline for enrollment.21,25 All trials used a VAS for pain outcome assessment (0–100-mm scale, with higher scores indicating worse pain). One trial also used the Brief Pain Inventory (BPI) severity scale,21 and another also used the Oswestry Disability Questionnaire.25
Attrition rates were high overall, despite relatively short trial durations of 7 to 9 weeks (Table 2). Six trials lost more than 30% of participants to follow-up.
We rated seven studies, plus the Eli Lilly trial, as Fair quality because of inadequate description of randomization and allocation-concealment.21–25 One study was rated Poor quality because 41% of randomized participants were lost to follow-up.26
Study Outcomes
Outcomes of included trials are summarized in Table 3. Neither of the two trials that required participants to have both pain and depression at baseline found a difference in pain outcomes between active drug and placebo.21,25 One of these trials compared duloxetine 60 mg/day with placebo.21 To be included, patients were required to have baseline pain of at least mild severity (BPI average pain severity score 2 on a 0–10 scale, with higher scores indicating worse pain); however, those with "a primary pain complaint with a diagnosis such as arthritis, fibromyalgia, migraine headache, or acute injury" were excluded. Participants’ mean baseline pain severity was moderate (VAS: 48 mm). No statistically significant differences between duloxetine and placebo for depression or pain outcomes were present at 7 weeks.
The other trial that required the presence of pain for inclusion compared paroxetine 20 mg/day with placebo.25 Patients with low back pain for at least 6 months, significant disability (Oswestry Disability Index >30%), and depressive symptoms (Montgomery-Asberg Depression Rating Scale >16) were recruited from a rheumatology clinic. Baseline pain severity was moderate (mean VAS: 56 mm). Pain and disability scores did not differ between groups at 8 weeks.
We included four trials, including the Eli Lilly trial, that compared duloxetine with paroxetine for treatment of depression, with pain as a secondary outcome.24,26,27 Inclusion criteria were similar: participants were adult outpatients who met Diagnostic and Statistical Manual of Mental Disorders, 4th Edition (DSM–IV), criteria for major depressive disorder (MDD) and had a minimum score of 15 on the 17-item Hamilton Rating Scale for Depression (Ham–D). The presence of pain was not required for enrollment in the studies. Mean baseline pain severity was mild in each trial.
Detke et al.24 compared two high-dose formulations of duloxetine (80 mg/day and 120 mg/day) with low-dose paroxetine (20 mg/day) and placebo. Mean VAS scores improved in all treatment groups; no statistically significant differences were found between paroxetine and duloxetine; however, only paroxetine was statistically superior to placebo (–16.0 mm versus –5.3 mm; p=0.035).
Goldstein et al.26 compared two different duloxetine formulations: low-dose (40 mg/day) and high-dose (80 mg/day), with low-dose paroxetine (20 mg/day) and placebo. Overall pain improved similarly in all three drug-treatment arms (p values for active drug comparisons not reported); however, only high-dose duloxetine was statistically better than placebo (–10.3 mm versus –3.2 mm; p=0.048).
Perahia et al.27 compared two high doses of duloxetine (80 mg/day and 120 mg/day) with low-dose paroxetine (20 mg/day) and placebo. Pain outcomes did not differ between active drugs at 8 weeks. Only duloxetine 80 mg/day was statistically superior to placebo (–15.5 mm versus –7.6 mm; p=0.014).
A fourth trial, which is publicly available but unpublished (Eli Lilly study, cited above) compared low-dose (40 mg/day) and high-dose (80 mg/day) duloxetine with low-dose paroxetine (20 mg/day) and placebo. No statistically significant differences between treatment arms were present at endpoint.
We also included two placebo-controlled trials of duloxetine for treatment of depression that also evaluated pain as a secondary outcome.22,23 Both showed a statistically significant difference in pain scores favoring duloxetine 60 mg over placebo.
Metaanalysis
We conducted metaanalyses assessing the pooled difference in VAS overall pain for all three possible comparisons. We pooled VAS outcomes from all four head-to-head trials of duloxetine versus paroxetine (Figure 1). Results did not show a significant difference between drugs (WMD: –0.8 mm; 95% confidence interval [CI]: –3.8 to 2.3; negative values favor paroxetine). The I2 statistic, indicating the percentage of variation across studies due to heterogeneity, was 0% (95% CI: 0 to 67.9).
View larger version (47K):
[in this window]
[in a new window]
|
FIGURE 1. Effect-Size Metaanalysis of Visual-Analog Scale for Overall Pain: Paroxetine vs. Duloxetine
Pooled WMD (weighted mean difference): –0.8 mm; 95% confidence interval: –3.8 to 2.3.
|
We also pooled VAS outcomes from six trials that compared duloxetine with placebo and four trials that compared paroxetine with placebo (Figure 2). One placebo-controlled duloxetine trial21 and one placebo-controlled paroxetine trial25 were not included in the metaanalysis because of insufficient data. Results from metaanalyses of the placebo-controlled trials indicate that both active drugs are statistically superior to placebo, with similar effect sizes (WMD for duloxetine versus placebo: 5.2 mm; 95% CI: 2.7–7.7; WMD for paroxetine versus placebo: 5.8 mm; 95% CI: 2.2–9.4).
View larger version (34K):
[in this window]
[in a new window]
|
FIGURE 2. Effect-Size Metaanalysis of Visual-Analog Scale for Overall Pain: Duloxetine vs. Placebo and Paroxetine vs. Placebo
1) Pooled WMD (weighted mean difference): 5.2 mm; 95% confidence interval: 2.7 to 7.7.
2) Pooled WMD: 5.8 mm; 95% confidence interval: 2.2 to 9.4.
|
|
DISCUSSION
|
We did not find evidence of differential efficacy between duloxetine and paroxetine in treatment of pain accompanying depression. No studies were identified for the 10 other second-generation antidepressants included in our search strategy.
Our pooled analysis found no significant difference in pain outcomes between duloxetine—an SNRI, and paroxetine—an SSRI. SNRIs, namely, duloxetine and venlafaxine, inhibit neuronal uptake and therefore enhance activity of neurotransmitters, especially serotonin and norepinephrine. Because of this dual uptake-inhibition, it has been hypothesized that SNRIs may have special analgesic effects in addition to their antidepressant effects.13–15 Our metaanalysis results do not support this hypothesis. Unfortunately, we found no relevant RCTs of venlafaxine; uncontrolled and preclinical studies are summarized elsewhere.13,14 Clinical trials are still needed to test the hypothesized "dual effect" of SNRIs on pain and depression.
We found that pooled pain-score outcomes for both duloxetine and paroxetine were superior to placebo, but the clinical significance of this finding is uncertain. Pooled differences were statistically significant, but small in magnitude. Estimates of the minimum clinically significant difference on the 100-mm VAS vary, but are consistently higher than the values reported in this review.28–32 Small improvements in pain severity are likely nonspecific effects related to improvement in depression level.
The strengths of our review are its rigorous methods and comprehensive search of the literature for both published and unpublished trials. However, our review is limited by the small number of relevant trials and the lack of available trials for the majority of second-generation antidepressants. The eight studies we included had important weaknesses. Only two were specifically designed to evaluate pain outcomes in patients with both depression and pain. The others were depression trials with pain scores as secondary outcomes. Mean pain scores in these trials were mild at baseline, so floor effects could make improvement difficult to detect. Results from trials that were not designed to enroll patients with clinically relevant comorbid pain at baseline should be viewed cautiously. This review is also limited by included trials’ inadequate outcome-reporting. For example, we were unable to calculate relative risk of improvement or number needed to treat because authors did not report the percentage of patients who experienced clinically important improvement in pain scores.
In summary, we did not find evidence of differential efficacy between duloxetine and paroxetine on pain in patients with depression. However, few trials relevant to our study question were available, and eligible trials were limited by shortcomings in design and reporting. We believe that insufficient evidence is currently available to support the choice of one second-generation antidepressant over another in patients with physical pain accompanying depression. Trials designed to compare the effects of newer antidepressants on clinically relevant pain outcomes in patients with both pain and depression are still needed.
|
ACKNOWLEDGMENTS
|
Disclaimer: The authors of this report are responsible for its content. Statements in the report should not be construed as endorsement by the Agency for Healthcare Research and Quality or the U.S. Department of Health and Human Services of a particular drug, device, test, treatment, or other clinical service.
We thank our colleagues from the University of North Carolina at Chapel Hill: Tim Carey, Janelle Shumate, and Stacey Williams. We also express our appreciation to Linda Lux and Loraine Monroe of RTI International.
This research was funded through a contract from the Agency for Healthcare Research and Quality to the RTI International University of North Carolina Evidence-Based Practice Center (contract #290–02-0016).
Dr. Krebs was supported by the Robert Wood Johnson Clinical Scholars Program. Dr. Gaynes was supported in part by an NIMH K23 Career Development Award (MH01951-03), and Dr. Hansen was supported by NIH grant K12 RR023248.
|
REFERENCES
|
- Bair MJ, Robinson RL, Katon W, et al: Depression and pain comorbidity: a literature review. Arch Intern Med. 2003; 163:2433-2445
- Bair MJ, Robinson RL, Eckert GJ, et al: Impact of pain on depression treatment response in primary care. Psychosom Med 2004; 66:17–22[Abstract/Free Full Text]
- Karp JF, Scott J, Houck P, et al: Pain predicts longer time to remission during treatment of recurrent depression. J Clin Psychiatry 2005; 66:591–597[Medline]
- Karp JF, Weiner D, Seligman K, et al: Body pain and treatment response in late-life depression. Am J Geriatr Psychiatry 2005; 13:188–194[CrossRef][Medline]
- Blier P, Abbott Laboratories FV: Putative mechanisms of action of antidepressant drugs in affective and anxiety disorders and pain. J Psychiatry Neurosci 2001; 26:37–43[Medline]
- Campbell LC, Clauw DJ, Keefe FJ: Persistent pain and depression: a biopsychosocial perspective. Biol Psychiatry 2003; 54:399–409[CrossRef][Medline]
- O’Malley PG, Jackson JL, Santoro J, et al: Antidepressant therapy for unexplained symptoms and symptom syndromes. J Fam Pract 1999; 48:980–990[Medline]
- Saarto T, Wiffen PJ: Antidepressants for neuropathic pain (systematic review). Cochrane Database of Systematic Reviews, 2006
- Staiger TO, Gaster B, Sullivan MD, et al: Systematic review of antidepressants in the treatment of chronic low back pain. Spine 2003; 28:2540–2545[CrossRef][Medline]
- Hansen RA, Gartlehner G, Lohr KN, et al: Efficacy and safety of second-generation antidepressants in the treatment of major depressive disorder. Ann Intern Med 2005; 143:415–426[Abstract/Free Full Text]
- Williams JW Jr, Mulrow CD, Chiquette E, et al: A systematic review of newer pharmacotherapies for depression in adults: evidence report summary. Ann Intern Med 2000; 132:743–756[Abstract/Free Full Text]
- Barbui C, Hotopf M, Freemantle N, et al: Treatment discontinuation with selective serotonin reuptake inhibitors (SSRIs) versus tricyclic antidepressants (TCAs) (systematic review). Cochrane Database of Systematic Reviews, 2007
- Lynch ME: Antidepressants as analgesics: a review of randomized, controlled trials. J Psychiatry Neurosci 2001; 26:30–36[Medline]
- Ansari A: The efficacy of newer antidepressants in the treatment of chronic pain: a review of current literature. Harv Rev Psychiatry 2000; 7:257–277[CrossRef][Medline]
- Mallinckrodt CH, Goldstein DJ, Detke MJ, et al: Duloxetine: a new treatment for the emotional and physical symptoms of depression. Primary Care Companion: J Clin Psychiatry 2003; 5:19–28[Medline]
- Gartlehner G, Hansen RA, Thieda P, et al: Comparative Effectiveness of Second-Generation Antidepressants in the Pharmacologic Treatment of Adult Depression: Comparative Effectiveness. Review No. 7. (Prepared by RTI International-University of North Carolina Evidence-Based Practice Center, under Contract No. 290-02-0016, Agency for Healthcare Research and Quality. Rockville, MD, 2007; available at http://effectivehealthcare.ahrq.gov/reports/topic.cfm?topic=8&sid=39&rType=3&selection=6; accessed April 19, 2007
- Dworkin RH, Turk DC, Farrar JT, et al: Core outcome measures for chronic pain clinical trials: IMMPACT recommendations. Pain 2005; 113:9–19[CrossRef][Medline]
- Harris RP, Helfand M, Woolf SH, et al: Current methods of the U.S. Preventive Services Task Force: a review of the process. Am J Prev Med 2001; 20:21–35[Medline]
- National Health Service Centre for Reviews and Dissemination: Undertaking Systematic Reviews of Research on Effectiveness: CRD’s Guidance for Those Carrying Out or Commissioning Reviews. Report Number 4 (2nd Edition), 2001
- Higgins JPT, Thompson SG, Deeks JJ, et al: Measuring inconsistency in metaanalyses. BMJ 2003; 327:557–560[Free Full Text]
- Brannan SK, Mallinckrodt CH, Brown EB, et al: Duloxetine 60 mg once daily in the treatment of painful physical symptoms in patients with major depressive disorder. J Psychiatr Res 2005; 39:43–53[CrossRef][Medline]
- Detke MJ, Lu Y, Goldstein DJ, et al: Duloxetine, 60 mg once daily, for major depressive disorder: a randomized, double-blind, placebo-controlled trial. J Clin Psychiatry 2002; 63:308–315[Medline]
- Detke MJ, Lu Y, Goldstein DJ, et al: Duloxetine 60 mg once-daily dosing versus placebo in the acute treatment of major depression. J Psychiatr Res 2002; 36:383–390[CrossRef][Medline]
- Detke MJ, Wiltse CG, Mallinckrodt CH, et al: Duloxetine in the acute and long-term treatment of major depressive disorder: a placebo- and paroxetine-controlled trial. Eur Neuropsychopharmacol 2004; 14:457–470[CrossRef][Medline]
- Dickens C, Jayson M, Sutton C, et al: The relationship between pain and depression in a trial using paroxetine in sufferers of chronic low back pain. Psychosomatics 2000; 41:490–499[Abstract/Free Full Text]
- Goldstein DJ, Lu Y, Detke MJ, et al: Duloxetine in the treatment of depression: a double-blind, placebo-controlled comparison with paroxetine. J Clin Psychopharmacol 2004; 24:389–399[CrossRef][Medline]
- Perahia DGS, Wang F, Mallinckrodt CH, et al: Duloxetine in the treatment of major depressive disorder: a placebo- and paroxetine-controlled trial. Eur Psychiatry 2006; 21:367–378[CrossRef][Medline]
- Gallagher EJ, Liebman M, Bijur PE: Prospective validation of clinically important changes in pain severity measured on a visual analog scale. Ann Emerg Med 2001; 38:633–638[CrossRef][Medline]
- Hagg O, Fritzell P, Nordwall A: The clinical importance of changes in outcome scores after treatment for chronic low back pain. Eur Spine J 2003; 12:12–20[Medline]
- Kelly AM: Does the clinically significant difference in visual analogue scale pain scores vary with gender, age, or cause of pain? Acad Emerg Med 1998; 5:1086–1090[Medline]
- Ten Klooster PM, Drossaers-Bakker KW, Taal E, et al: Patient-perceived satisfactory improvement (PPSI): interpreting meaningful change in pain from the patient’s perspective. Pain 2006; 121:151–157[CrossRef][Medline]
- Todd KH, Funk KG, Funk JP, et al: Clinical significance of reported changes in pain severity. Ann Emerg Med 1996; 27:485–489[CrossRef][Medline]
This article has been cited by other articles:
|
|
|
|
 
K. Kroenke, M. J. Bair, T. M. Damush, J. Wu, S. Hoke, J. Sutherland, and W. Tu
Optimized Antidepressant Therapy and Pain Self-management in Primary Care Patients With Depression and Musculoskeletal Pain: A Randomized Controlled Trial
JAMA,
May 27, 2009;
301(20):
2099 - 2110.
[Abstract]
[Full Text]
[PDF]
|
|
|
Get information about faster international access.
Privacy Policy
Copyright © 2008
Academy of Psychosomatic Medicine.
All rights reserved.
Home
| Search
| Current Issue
| Past Issues
| Subscribe
| All APPI Journals
| Help
| Contact Us
|
TOP
ABSTRACT
INTRODUCTION
