Opioids in Chronic Musculoskeletal Conditions

Abstract

The use of opioids for benign disorders has been limited by concerns about these compounds' potential adverse events and their possible misuse. However, during the last few years an increased use in nonmalignant disorders, including rheumatologic diseases, has been observed. Herein, we review the scientific evidence for opioid therapy in three common scenarios in clinical rheumatology. Low back pain is a very frequent reason for consultation. Overall, the large majority of studies show a positive, yet rather moderate, effect of opioids in pain control, as well as in other outcomes including mood, work disability and anxiety. Similarly, opioids seem to have a role in the management of hip and knee osteoarthritis; indeed, they have been included in all international guidelines for the treatment of these conditions. However, clinical studies addressing opioid use in clinical situations are plagued by methodological limitations; furthermore, the large majority of these studies only provide short-term information about opiod utilization in these patients. Finally, opioids are currently being used as complementary therapy in inflammatory joint conditions whereby they may significantly improve the quality of life of some of these patients. Regarding their safety, severe adverse events, including abnormal drug-seeking behaviour, are very rare, but mild adverse events are frequent leading to drug discontinuation in a significant number of cases.

Introduction

Pain is a common symptom in a large number of rheumatologic conditions leading to a substantial impact on the quality of life of patients suffering from such diseases. The mainstay for pain control in the rheumatologic arena has classically been based on the use of common analgesics such as acetaminophen and/or nonsteroidal anti-inflammatory drugs (NSAIDs). However, optimal pain control is not usually achieved with these drugs in a large number of patients; furthermore, an important proportion of them cannot use some of these treatments, particularly NSAIDs, owing to their side effects or the presence of comorbid conditions.

The use of opioids (minor and major) represents a new strategy for the management of patients with refractory pain. These compounds were originally used for the treatment of pain in onco-logic patients, but during the last few decades there has been an increase in their utilization in benign conditions as well. Thus, minor opioids such as codeine, dextropropoxyphene or more recently tramadol have been used frequently for the management of musculoskeletal problems.

Moreover, the use of major opioids has also been extended to the treatment of rheumatologic patients with refractory pain. Nevertheless, the use of opioids for these types of patients and conditions raises different issues and concerns, which likely account for their still limited use in these disorders.

The aim of the present review is to examine the value of opioid therapy and its role in rheumatology according to the published scientific literature. In order to address this objective properly, we have focused our attention in the use of opioids in three common clinical scenarios: low back pain (LBP), osteoarthritis (OA) and inflammatory joint conditions. In these three scenarios we have reviewed the published randomized and open-labelled clinical trials, review articles, meta-analyses and expert consensus reports about the use of opioids.

Low back pain

More than 90% of the general population will suffer at least one episode of LBP during their lifetime; 10% of these patients will have a chronic course resulting in high levels of disability with important economic consequences due to absence from work and early retirement. With the exception of patients with a problem that can be resolved with a surgical intervention or other interventional procedures, the large majority of these patients are managed using common analgesics and/or NSAIDs. Nonetheless, in a significant percentage of cases, these treatments are either ineffective or cannot be used due to the occurrence of side effects or the presence of contraindications. In these situations opioids might be a reasonable alternative.

In the past, only a few controlled studies have addressed the efficacy of opioids for the management of patients with LBP. More recently several clinical trials with newer opioids and new formulations have been published. By and large, these studies have shown that opioids have statistically and clinically positive effects on pain control assessed with different scales [Vorsanger et al. 2008; Hale et al. 2007b, 2005, 1999; Katz et al. 2007; Wallace et al. 2007; Rauck et al. 2007, 2006a, 2006b; Allan et al. 2005; Kosinski et al. 2005; Peloso et al. 2004; Gammaitoni et al. 2003; Ruoff et al. 2003; Schnitzer et al. 2000; Schofferman, 1999; Jamison et al. 1998; Simpson et al. 1997]. Nevertheless, it is important to note that in the majority of these trials only mild-to-moderate levels of analgesic effects have been obtained from these compounds. However, in some of these studies, opioids have been shown to exert some beneficial effects in other dimensions such as mood, functional and work capacity as well as anxiety, which are common occurrences in patients with chronic pain disorders.

Adverse events, which have been mild for the most part, have been reported to occur with relative frequency in the patients studied; these side effects led to the withdrawal of treatment in up to 21% of patients enrolled in these trials. Severe adverse events have been observed very rarely; problems of special concern such as drug abuse, abnormal drug behaviour or cases of addiction and psychological dependence have only been reported in exceptional cases.

Several methodological limitations of these studies should be borne in mind in order to understand them better. For example, the majority of the clinical trials reviewed are open labelled with only a few being placebo controlled; furthermore, in some studies two opioids were compared, but they were not compared with either placebo or another type of analgesic. In only one study was a comparison with a NSAID conducted. According to these data, therefore, it is very difficult to establish firm conclusions about the exact role and hierarchy of these compounds in the management of patients with LBP. Moreover, these studies are quite heterogeneous in their design, the type of opioid studied (tramadol alone or in combination, oxycodone, oxymor-phone, morphine, transdermal fentanyl), the endpoints and outcomes as well as the assessment tools used and the period of follow up, precluding their comparison and preventing definitive conclusions about them begin reached. Furthermore, only a few of the studies used standardized outcome measurements such as standardized mean differences (SMD), which may have provided more clear information about the real effect size of these compounds allowing their comparison. Finally, with the exception of only one study that provides one-year follow-up information, all of the other studies presented only short-term results, with follow-up time ranging from 12 to 32 weeks, have been reported. This fact has special relevance when dealing with patients suffering from a chronic condition such as LBP who will probably need treatment over long periods of time.

In 2007 Martell and colleagues published an interesting meta-analysis about the use of opioid therapy in LBP estimating the prevalence of its use, its efficacy and the frequency of abnormal drug behaviours [Martell et al. 2007]. According to their analyses, opioid use in this setting varied from 3% to 66%. By reanalysing the primary data from the studies included in this meta-analysis, they calculated the SMDs for pain scores and performed pooled analyses. They found no statistically significant benefits from opioid use either in comparison with placebo (placebo-controlled studies) or with baseline data (studies comparing opioids). Nonetheless, it is necessary to note that this review only included 15 clinical trials given its stringent inclusion criteria, and almost one third of them were older studies in which codeine or dextropropoxyphene were used, opioids scarcely used for pain control nowadays. A significant number of recent clinical trials with newer opioids could not be included because they had been published subsequently in relation to this meta-analysis. These issues need to be considered when trying to define the real efficacy of opioids in LBP. As to abnormal drug behaviour, a lifetime prevalence of substance abuse disorders ranging from 36% to 56% and a current substance abuse rate as high as 43% were found; the frequency of aberrant medication-taking behaviours ranged from 5% to 24%. Notwithstanding these data, the authors recognize that firm conclusions about these particular aspects could not be established given the poor methodological quality of the studies.

In a second study also published in 2007, the authors examined the utility of diverse types of treatments for LBP [Chou and Huffman, 2007]. Regarding opioids, they confirmed our conclusions about their positive overall effect coupled with the frequent occurrence of mild side effects in this setting. They also underscored that the degree of pain reduction is only moderate and that no differences in efficacy have been found between sustained and immediate-release formulations which may be relevant for their clinical use in chronic conditions such as LBP [Chou and Huffman, 2007].

Osteoarthritis

OA is the most prevalent form of arthritis within the general population and a frequent cause of pain and disability. In this review, we focus our attention on hip and knee OA not only because they are the most frequent forms of OA studied but because they produce significant morbidity as well as economic burden for the healthcare systems. Moreover, they have good, straightforward assessment instruments for their clinical evaluation.

Although the OA published literature is quite extensive, the number of randomized clinical trials, systematic reviews or meta-analyses is somewhat more limited. However, several international guidelines about the management of patients afflicted with these conditions have been published; there is consensus about the role of opioids for the treatment of these two forms of OA. Like with LBP, clinical trials of opioids in OA report a global, albeit moderate, beneficial effect on pain control with up to 88% of the patients reporting improvement and up to 75% rating the qualify of their treatment as good or excellent [Beaulieu et al. 2008; Choquette et al. 2008; Florete et al. 2008; Burch et al. 2007; Fishman et al. 2007; Hale et al. 2007a; Vorsanger et al. 2007; Gana et al. 2006; Kivitz et al. 2006; Langford et al. 2006; Le Loet et al. 2005; Malonne et al. 2004; Pavelka et al. 2004]. The majority of these clinical trials do not provide effect sizes; however, when this was calculated for one of these trials, it was found to be of a moderate level (≈0.4) [Langford et al. 2006]. Higher effect sizes (0.78 and 0.79, respectively) were estimated in two reviews in which data from published clinical trials were pooled [Avouac et al. 2007; Zhang et al. 2005]. It is important to keep in mind, that results from this type of pooled analysis are heavily dependent on which clinical trials are included as per previously established inclusion criteria, which may yield discrepant results. Head-to-head comparisons of NSAIDs and tramadol as the opioid have been performed, demonstrating similar beneficial effects on pain control [Beaulieu et al. 2008]. Positive effects of the use of opioids in hip and knee OA other than pain control have been described in different studies including the beneficial effect on sleep quality and functional capacity, reduction in the dose of NSAID, which can be quite important for some patients, and improvements in the quality of life of those patients waiting for an arthroplasty to be performed [Choquette et al. 2008; Florete et al. 2008; Zhang et al. 2008; Gana et al. 2006; Kivitz et al. 2006; Langford et al. 2006; Peloso, 2001]. Regarding their safety, similarly to clinical trials on LBP, side effects, mostly mild, have been reported frequently with approximately a quarter of the patients having to discontinue the opioid. Again, severe side effects have been observed very rarely. Among the adverse events, nausea (up to ^39%) [Kivitz et al. 2006], constipation (up ≈30%), vomiting (up to 26%) [Le Loet et al. 2005], somnolence (up to ≈25%) [Hale et al. 2007a] and dizziness (up to ≈23%) [Vorsanger et al. 2007] have been reported most commonly. Owing to the high incidence of side effects reported, a systematic review conducted by a Cochrane collaboration suggested that the small-to-moderate beneficial effects observed with nontramadol opioids may be outweighed by the risk of adverse events [Nuesch et al. 2009]; however, this conclusion is debatable.

Similar methodological limitations to those observed in the opioid LBP clinical trials can be considered here. Thus, heterogeneity of study design, the opioid studied, the assessment instruments used as well as the differences between clinical trials and ‘real life’ clinical practice should be taken into account before assuming that these conclusions are definitive.

Finally, three international organizations have published consensus guidelines for the optimal management of patients with hip and knee OA. The first was published in 2000 by the American College of Rheumatology (ACR) and an update is expected soon. According to these guidelines, tramadol is reserved for patients who have failed treatment with NSAID and acetaminophen, also providing the opportunity to use formulations combining weak opioids and acetaminophen while reserving major opioids for specific refractory patients [American College of Rheumatology, 2000]. In 2007, the European League Against Rheumatism (EULAR) published its own guidelines which were developed using a Delphi method by an expert panel based on a systematic review of the literature and in agreement with evidence-based medicine rules [Zhang et al. 2005]. These guidelines include 10 statements about the management of OA and indicate that opioids are useful in patients where NSAIDs are not effective, either because of side effects or the fact that they are contraindicated (level of evidence Ib; strength of the evidence A). Similarly, in the same year the Osteoarthritis Research Society International (OARSI) published its consensus guidelines for OA treatment based on the review of all previously published guidelines pointing out also that opioids should be indicated in nonresponders to common analgesics or NSAIDs [Zhang et al. 2008].

Inflammatory joint conditions

Patients with diverse types of inflammatory joint disorders may require, in some cases, in addition to specific therapies to control the inflammatory process and stop the progression of their disease, the coadjuvant use of drugs aimed at better pain control. In this context, NSAIDs have been used classically. Nevertheless, a significant number of these patients do not achieve a good response with these compounds or present adverse events or have contraindications for their use. Moreover, the increasing evidence of a high risk for cardiovascular problems in patients with these disorders, in addition to the fact that some NSAIDs can increase this risk further, make the use of these compounds not advisable in a significant percentage of cases. Under these circumstances, opioids may be worthy of consideration as complementary therapy for the adequate management of these patients.

A few studies have specifically addressed the use and role of opioids in inflammatory rheumatologic conditions, mainly rheumatoid arthritis, but no randomized clinical trials about their efficacy have been carried out in this setting. However, some data can be derived from the literature in order to examine the use of these compounds in patients with these disorders in clinical practice.

Regarding this matter, the use of opioids in rheumatoid arthritis has been examined in a relatively recent publication in which the authors highlight that the use of opioids in rheumatoid arthritis is, in fact, relatively common although the large majority of patients use ‘weak’ opioids. According with the published data, the prevalent use of opioids in these patients was only 4%, but their ‘ever’ use reached almost 25%. In fact, these figures are higher than those commonly seen in other pathologies including LBP and OA, which are usually deemed as being more suitable for opioid treatment [Solomon et al. 2006]. Furthermore, it has been stressed that the use of these compounds in rheumatoid arthritis and related pathologies has increased during the last few decades regardless of the age of the patients [Grijalva et al. 2008]. Despite this increasing use of opioids in inflammatory conditions, definitive evidence of their clinical usefulness is largely lacking. Recently, a study about the use of trans-dermal fentanyl patches in rheumatoid arthritis patients has been published [Berliner et al. 2007], demonstrating that fentanyl had a positive effect on the quality of life of these patients improving the patient pain intensity, functional capacity, sleep quality and general well being with good tolerability and overall high levels of satisfaction. However, this study was not placebo controlled; furthermore, no information was provided about the specific aetiology of pain in the different patients included in the analysis. Despite these limitations, this study provides the scientific basis for incorporating this therapeutic option into the therapeutic armamentarium of these patients.

As to safety, there is not much information about the tolerability of opioids in this specific context. Similar considerations to those mentioned for the other two clinical scenarios should be applied here. Nevertheless, lower frequencies of common side effects, such as nausea (≈10%) or vomiting (≈7%), to those observed in the OA trials have been reported [Berliner et al. 2007].

Furthermore, dextropropoxyphene has been reported to be relatively safe in rheumatoid arthritis patients even in combination with amitriptyline [Saarialho-Kere et al. 1988].

Finally, the potential anti-inflammatory effects of opioids should be considered. It has been shown for example that the levels of endogenous opioids are increased in inflamed synovium and these opioids are able to inhibit inflammatory mediators such as interleukin-1β (IL-1β), IL-6, IL-8, tumour necrosis factor alpha (TNFα) or metallo-proteases with the consequent amelioration of the inflammatory response in rats with adjuvant-induced polyarthritis [Straub et al. 2008; Takeba et al. 2001]. Further studies in this field will be needed to clarify whether these anti-inflammatory actions will have any relevance for the use of opioids in inflammatory joint disorders.

Conclusions

Although some of the issues about the use of opioids in rheumatologic disorders are still open to debate, overall our review of the present literature points to the fact that these compounds represent a valuable option in the management of these patients. Methodological limitations and the heterogeneity of the different studies probably explain some discrepancies among them, but, by and large, a positive effect in pain control and other associated problems such as anxiety, sleep disorders, disability, and quality of life are commonly recognized, although the level of improvement is only modest in a significant proportion of patients. The safety of these compounds remains an important shortcoming given the frequency of adverse events and the high rate of discontinuations due to them. Nonetheless, given that the large majority of these events are not serious, the utilization of suitable dose titration and tapering protocols, as well as the simultaneous use (at least at the beginning of opioid therapy) of preventive therapy for common side effects, such as nausea or dizziness, most probably will help to improve their tolerability and survival in clinical practice.