What are some examples of musculoskeletal disorders?

RA is a chronic autoimmune disease that causes pain, aching, stiffness, and swelling in multiple synovial joints. It typically affects the small joints of the hands and the feet and usually both sides equally and symmetrically, although any synovial joint can be affected. It is a systemic disease and so can affect other organ systems, including the heart, lungs, and eyes (NICE, 2018), and it can cause other systemic symptoms, including fatigue, fever, and weight loss (Wasserman, 2011). Because the musculoskeletal impairments associated with RA are typically the most disabling and the major source of functional limitations for individuals with this condition, this section primarily focuses on those impairments, although the committee acknowledges that RA’s impacts on other organ systems may also influence global functioning (Filipovic et al., 2011). The common pathophysiology underlying musculoskeletal impairments in RA is inflammation of the synovium (Scott et al., 2010). During disease flares, inflammation results in a short-term worsening of joint pain and swelling; in patients with longstanding and severe disease, persistent inflammation will over time result in the erosion of cartilage and bone, leading to joint destruction and deformities that in turn cause chronic pain and functional limitations (Sokka and Pincus, 2001).

The diagnosis of RA is based on a patient’s clinical history, physical examination, and laboratory findings. The 2010 American College of Rheumatology (ACR)/European League Against Rheumatism (EULAR) classification criteria for RA form the generally accepted diagnostic criteria for the condition, although, notably, these criteria were developed for research studies to allow for the identification of individuals with earlier-stage RA and were not primarily intended for clinical practice. The criteria are outlined in Table 5-6 and are intended to be applied to individuals for whom there is clinical suspicion of RA based on definite synovitis in at least one joint, as determined by physical exam, that is not better explained by a different condition. Patients with a score of at least 6 out of 10 are considered to have “definite RA.” The 2010 ACR/EULAR criteria were designed to identify patients with recent-onset and active RA; adults with longstanding or inactive disease may be diagnosed with RA if there is a documented prior history of findings or laboratory testing fulfilling those criteria. Adults with seronegative RA who lack rheumatoid factor and anti-citrullinated protein antibody on laboratory testing might not satisfy the 2010 ACR/EULAR criteria (Humphreys and Symmons, 2013), but may still be diagnosed with RA if their clinical findings are otherwise characteristic of the disease and if alternative diagnoses are excluded. In those cases radiographic findings of bone erosions, which are characteristic of RA, may help support the diagnosis, although radiography is generally not required to establish a diagnosis (Scott et al., 2010).

The lifetime risk of RA is two to three times higher among women than men (Crowson et al., 2011). The onset of RA peaks between the ages of 30 and 50 years, although it may occur at any age (Tehlirian and Bathon, 2008). The risk factors include older age, a family history of RA, and current or prior cigarette smoking (CDC, 2019a; Costenbader et al., 2006).

The principal measures used to assess response to treatment and remission for RA are composite, multidimensional outcome measures that incorporate clinical data (i.e., the physical examination, laboratory markers such as erythrocyte sedimentation rate [ESR] and C-reactive protein [CRP], physician’s assessment), functional assessment, patient-reported symptoms, and patient-reported global assessment (Felson and LaValley, 2014). For RA, it has long been recognized that because of the heterogeneity of its manifestations, and its impacts on multiple organ systems, improvement cannot be accurately determined based on a single domain (e.g., laboratory markers); accordingly, the use of composite outcome measures reflecting multiple disease domains has become the norm (Aletaha et al., 2008). Notably and of key importance to the current study, the routine assessment of physical functioning is strongly recommended as part of any treatment strategy for RA and is more widespread than for many other disabling medical conditions (Singh et al., 2016). We review the major measures used to assess treatment response below, noting that while these measures are widely used in research and clinical trials, their application in routine clinical practice by U.S. rheumatologists is highly variable (Anderson et al., 2012).

ACR20 For patients with RA, the ACR has developed several definitions of a response to therapy, including the ACR20, the ACR50, and the ACR70, which indicate an improvement of at least 20 percent, 50 percent, or 70 percent, respectively, on a set of core outcome measures (Felson and LaValley, 2014). The core measures include the swollen joint count, the tender joint count, and three out of the following five measures: pain visual analog scale (patient-reported pain symptom scale), patient global assessment, physician global assessment, inflammatory marker levels (either ESR or CRP), and a measure of physical functioning (commonly the Health Assessment Questionnaire Disability Index [HAQ], described below). Of these, the ACR20 is the most widely used, and it has been recommended by the U.S. Food and Drug Administration as a preferred outcome measure in studies of new drugs for RA; accordingly it is commonly used as the primary outcome in clinical trials of RA therapies (Aletaha et al., 2008; Felson and LaValley, 2014). It is not recommended for monitoring treatment response in clinical practice—other disease activity scales, described below, are considered more feasible to implement in clinical settings (Greenberg et al., 2009).

Disease activity scales ACR-endorsed instruments to measure RA disease activity and to define remission include the Patient Activity Scale (PAS), the PASII, the Routine Assessment of Patient Index Data 3, the Clinical Disease Activity Index, the Disease Activity Score (DAS), and the Simplified Disease Activity Index (Anderson et al., 2011, 2012; Fransen et al., 2003; Pincus et al., 2008; Singh et al., 2011; Wolfe et al., 2005). All scales are multidimensional, composite measures drawing on data from several different domains (e.g., physical exam, laboratory markers, functional measures, pain symptoms, physician- and patient-reported global assessments) and are sensitive in discriminating between different levels of disease activity (Anderson et al., 2011). These measures are commonly reported as secondary outcomes in clinical trials of drugs for RA, and they are recommended for routine assessments in clinical practice (Anderson et al., 2012; Greenberg et al., 2009).

Disease activity scores correlate closely with the degree of functional impairment related to RA, and, indeed, several of the aforementioned scores are based in part on functional assessments (Carvalho et al., 2019). However, because RA causes progressive joint damage and deformity, functional impairment is possible among individuals whose disease is quiescent if it was previously active (Ishida et al., 2018; Norton et al., 2014).

Functional assessment The most widely used measure of functional capacity in RA is the HAQ, which was originally developed in 1978 and assesses a patient’s ability to have carried out activities of daily living (dressing/grooming, arising, eating, walking, personal hygiene, reaching, gripping, and errands) over the previous week (Maska et al., 2011). The HAQ can be self-administered by patients or administered by a clinician, and it is commonly reported as a secondary outcome in clinical trials of new RA drugs. While it does not explicitly ask patients about work activities, multiple studies have demonstrated that the HAQ is a strong predictor of work disability (de Croon et al., 2004; McWilliams et al., 2014; Wolfe and Hawley, 1998; Young et al., 2000, 2002).

Several other instruments that aim to more specifically measure work-related functioning have been validated for the inflammatory arthropathies, including the Work Productivity and Activity Impairment Questionnaire (Tucker et al., 2019; Zhang et al., 2010), the Work Instability Scale (Revicki et al., 2015), and the Work Productivity Survey (Osterhaus and Purcaru, 2014). At present, such instruments are not widely used in either research or clinical practice, although they may hold promise.

The goals of RA treatment include reducing symptoms of joint pain and swelling, preventing deformity, maintaining quality of life, and limiting extra-articular disease manifestations (Wasserman, 2011). Pharmacologic treatments, specifically disease-modifying antirheumatic drugs (DMARDs), are the mainstay of therapy (Singh et al., 2016). They limit progressive joint damage and improve function through different mechanisms (Scott et al., 2010) (see Table 5-7). DMARDs are typically prescribed under the supervision of a rheumatologist. Care by a rheumatologist is associated with an earlier initiation of DMARD therapy (Rat et al., 2004; Widdifield et al., 2011) and improved treatment response (Criswell et al., 1997), resulting in less joint destruction (van der Linden et al., 2010), lower functional impairment (Ward et al., 1993), and a lower likelihood of requiring orthopedic surgery (Feldman et al., 2013). Traditional (non-biologic) DMARDs include methotrexate, leflunomide, hydroxychloroquine, and sulfasalazine; biologic DMARDs include anti-tumor necrosis factor (TNF) agents (adalimumab, certolizumab pegol, etanercept, golimumab, infliximab), and non-TNF biologics (abatacept, rituximab, tocilizumab). A final class of DMARDs includes Janus kinase (JAK)-inhibitors, of which tofacitinib is the primary agent used in RA. Traditional DMARDs and tofacitinib are orally administered medications4 that may be taken at home; anti-TNF biologics are generally available in prefilled syringes that can be injected subcutaneously by patients in their homes (with the exception of infliximab, which must be administered via intravenous infusion in an infusion center); non-TNF biologics are generally administered via intravenous infusion in an infusion center (with the exception of abatacept, which is also available as a prefilled syringe). Medications used for short-term symptom relief include NSAIDs and steroids; the latter may be administered orally, intramuscularly, or intra-articularly. Non-pharmacologic treatments include physical and occupational therapy, exercise, patient education, and psychosocial interventions (Rindfleisch and Muller, 2005). Pain is among the most prominent and distressing symptoms among patients with RA (ten Klooster et al., 2007). It is managed using therapies that target the underlying disease, such as DMARDs, as well as through adjunctive therapies targeting pain symptoms. The latter are discussed in more detail above in the Musculoskeletal Conditions and Pain section of this chapter. Surgery is indicated for intractable pain, severe loss of motion, or functional impairment that exists despite medical management (Rindfleisch and Muller, 2005).

Evidence-based treatment guidelines for the pharmacologic management of established RA (defined as a disease duration of at least 6 months) include the 2015 ACR guidelines and the EULAR guidelines (Singh et al., 2016); the latter were originally developed in 2010 and most recently updated in 2017 (Smolen et al., 2017a). Both the ACR and EULAR guidelines primarily address the use of DMARDs for RA treatment. Patients who are not in clinical remission and who have any degree of disease activity as measured using validated scales (see section on Measurement of Outcomes for Rheumatoid Arthritis for more detail) are considered candidates for therapy; indeed, it is recommended that therapy for RA be initiated as soon as possible after the diagnosis is established, as there is evidence that earlier DMARD therapy is associated with better outcomes. The specific agents recommended are determined by the degree of disease activity, prior treatments used, treatment response and toxicities, and the patients’ comorbidities (Anderson et al., 2000; Nell et al., 2004; Smolen et al., 2017). The goal of therapy is sustained clinical remission or low disease activity (Ramiro et al., 2014).

Under the 2015 ACR guidelines, monotherapy with a traditional DMARD is recommended as the first-line initial treatment for RA regardless of the level of disease activity, with methotrexate being the preferred agent. For patients who do not improve sufficiently with traditional DMARD monotherapy (i.e., RA disease activity remains moderate to high), the recommended approach is either a combination of traditional DMARDs, a biologic DMARD (with or without methotrexate), or tofacitinib (with or without methotrexate).5 For patients on anti-TNF therapy alone who continue to have moderate to high disease activity, the addition of one or two traditional DMARDs is recommended (methotrexate is again the preferred agent) owing to evidence of superior efficacy compared with monotherapy with a biologic. If treatment targets are not achieved with a given biologic DMARD, it is recommended that different biologic DMARDs be tried. Short-term, low-dose glucocorticoid treatment may be added for patients on traditional or biologic DMARDs whose disease activity remains moderate or high, or for RA flares. Once low disease activity is achieved on a specific DMARD regimen, it is recommended that the regimen be continued, given that clinical experience suggests a high risk of relapse and the need for resuming therapy in the absence of DMARD treatment. If remission is achieved, tapering DMARD therapy can be considered, though the guidelines recommend against discontinuing all therapy because of the high risk of relapse.6

The 2016 EULAR recommendations for RA treatment are largely similar to the 2015 ACR guidelines: notably, traditional DMARDs (and specifically methotrexate) are recommended as the initial therapy for RA, the addition of biologic DMARDs or tofacitinib is recommended if improvement is not achieved, and if patients do not respond to a biologic DMARD, the guidelines recommend switching to a different biologic DMARD or tofacitinib (Aletaha et al., 2008). There are, however, several distinctions between the ACR and EULAR guidelines worth noting. First, the EULAR guidelines recommend that short-term glucocorticoid therapy be considered when initiating or changing DMARDs, whereas the ACR guidelines reserve glucocorticoid use for patients with moderate or high disease activity despite DMARD therapy. Second, for patients who do not respond to initial monotherapy with a traditional DMARD, the EULAR guidelines recommend that the choice of the subsequent agent be based on prognostic factors. Specifically, for patients with “unfavorable” prognostic indicators (i.e., the presence of autoantibodies especially at high levels, high disease activity, early erosions, or no response to two traditional DMARDs), a biologic DMARD or JAK-inhibitor (tofacitinib or baricitinib) is recommended. For patients in whom such findings are absent, the guidelines recommend adding or changing to a different traditional DMARD. In contrast, the ACR guidelines do not discuss the role of prognostic factors in treatment selection.

The ACR and EULAR guidelines are based on comprehensive and systematic reviews of the evidence on RA treatment; however, they have several limitations in the context of this study. First, the guidelines do not discuss non-pharmacologic treatments for RA or the optimal combination of pharmacologic and non-pharmacologic therapies. RCTs support the use of physical exercise as a strategy to improve muscle strength and quality of life (Baillet et al., 2009; Brodin et al., 2008), whereas complementary therapies such as acupuncture and dietary changes have not been found to provide benefit (Hagen et al., 2009; Kelley, 2009; Smedslund et al., 2010; Wang et al., 2008). Second, since the publication of the guidelines, several additional therapies have been approved for RA or are currently under investigation. Sarilumab is a non-TNF biologic DMARD that was approved for the treatment of moderate-to-severe RA in 2017; it has improved efficacy relative to adalimumab, a commonly used anti-TNF biologic, with a similar safety profile (Burmester et al., 2017). Baricitinib is a JAK-inhibitor7 that was approved for the treatment of RA in 2018 and is therefore not discussed in the 2015 ACR guidelines; the 2016 EULAR guidelines note that there is some evidence for its superior efficacy relative to adalimumab,8 but because long-term safety data are limited, as with tofacitinib, it is recommended that biologic DMARDs be tried first (FDA, 2018; Taylor et al., 2017). Third, neither guideline explicitly discussed the impact of DMARDs on work-related functional capacity (Nam et al., 2014), which is the outcome of principal interest to the committee as it is especially relevant to the SSA population. Many of the individual studies on which the guidelines are based do assess the impact of DMARDs on measures of physical functioning, but there are limitations in extrapolating from those scales to estimate impacts on actual work capacity.

Beyond the ACR and EULAR guidelines, an important limitation of the RA treatment literature more broadly is the limited evidence that is available to guide the management of patients with refractory RA (Singh et al., 2016). While there is no universally accepted definition of refractory RA, the term is often used to refer to patients who have not responded to at least two different biologic DMARDs or to two different biologic DMARDs with different mechanisms of action (Buch, 2018; de Hair et al., 2018; Kearsley-Fleet et al., 2018; Roodenrijs et al., 2018). The prevalence of refractory RA is not well established; the only published national registry study to date is from the United Kingdom, and it estimated that at least 6 percent of patients with RA have been exposed to at least three DMARDs, which is suggestive of a difficult-to-treat disease (Kearsley-Fleet et al., 2018). It is not known what share of SSA beneficiaries with RA satisfy this definition of refractory disease, but because those patients have a lower chance of clinical remission, it is likely that they are disproportionately represented in the SSA population. At present, there is limited evidence to inform the appropriate treatment strategy for patients with refractory RA. Baracitinib was efficacious in a study population in whom the majority of patients had refractory disease (i.e., had previously tried at least two different biologic DMARDs), so it may provide an alternative for those patients (Genovese et al., 2016). Other novel therapies are currently under investigation (Aletaha and Smolen, 2018; Cheung and McInnes, 2017).

While pharmacologic treatments for RA can substantially improve symptoms, they also have associated toxicities that are important to consider (Aletaha and Smolen, 2018; Graham, 2006; Harirforoosh et al., 2013; Huscher et al., 2009; Kamata and Tada, 2018; Nash et al., 2017; Rindfleisch and Muller, 2005; Saag et al., 1994; Sostres et al., 2010). Serious infections are among the most concerning potential adverse effects of biologic DMARDs and glucocorticoids because of their immunosuppressive properties. The toxicities of medications may limit their use in specific patients depending on comorbidities (particularly patients with liver, renal, or cardiovascular disease) and may prompt patients to discontinue or switch medications (Choquette et al., 2019).

Few studies have directly and rigorously assessed the impact of RA treatments on work outcomes. The committee identified a Swedish study comparing traditional DMARDs to combination therapy with infliximab and methotrexate for RA; it found no differences between the treatment arms in the number of work-days lost (Eriksson et al., 2016).

In the absence of direct evidence on the impact of specific RA treatments on work outcomes, the committee reviewed evidence of the impact of RA treatments on measures of physical functioning, specifically the HAQ. HAQ scores are predictive of work disability, and the HAQ is commonly used as a secondary outcome measure in clinical trials testing RA therapies. Among pharmacologic agents, a range of medications including traditional DMARDs (e.g., methotrexate, leflunomide) (Scott et al., 2001) and biologic DMARDs (e.g., golimumab, tocilizumab, baricitinib, certolizumab, filgotinib, sarilumab, tofacitinib, sirukumab, adalimumab, rituximab) have all been demonstrated to improve functional status in RA as measured using the HAQ (Bingham et al., 2014; Burmester et al., 2016; Dougados et al., 2017; Emery et al., 2017; Genovese et al., 2015, 2018; Keystone et al., 2017; Rigby et al., 2011; Strand et al., 2015a,b; Takeuchi et al., 2017; Taylor et al., 2017). Comparative effectiveness analyses and active comparator trials have generally not identified significant differences between biologic DMARDs in their impact on HAQ scores in RA (Jansen et al., 2014; Strand et al., 2016), with the exception of two recent trials that found sarilumab to be superior to adalimumab in its impact on physical functioning as measured using the HAQ (Strand et al., 2018). Among nonpharmacologic treatment strategies, resistance exercises have been found to improve physical functioning as measured using the HAQ (Baillet et al., 2012).

A key limitation of those data is that most studies do not focus specifically on patients with severe or refractory RA, who might be more likely to participate in SSA programs (Kilcher et al., 2018), so it is unclear whether the aforementioned therapies would meaningfully improve work-related functional capacity within the population of interest to SSA. Of the evidence the committee reviewed, the studies that most closely reflected the population of interest (i.e., adults with severe RA resulting in functional limitations that significantly restrict work) were those evaluating the impacts of specific treatments in patients who had not responded to at least one biologic DMARD. In RA, sarilumab (Fleischmann et al., 2017), filgotinib (Genovese et al., 2018), baricitinib (Genovese et al., 2016; Smolen et al., 2017b), and tofacitinib (Strand et al., 2015b) have all been demonstrated to improve HAQ scores in patients with an inadequate response to at least one anti-TNF DMARD. Conversely, secukinumab (Blanco et al., 2017; Dokoupilova et al., 2018) was not found to improve physical functioning as measured using the HAQ in this population.

NSAIDs and low-dose glucocorticoids can provide symptom relief within days. With DMARDs, clinical improvement is typically expected within 3 months of starting therapy, although a substantial number of patients might not respond until months 3–6 (Kavanaugh et al., 2008, 2010). Accordingly, many clinical trials of RA therapeutics now assess treatment response at both 3 and 6 months, and the EULAR treatment guidelines for RA recommend changing therapy if no improvement is seen after 3–6 months (Ramiro et al., 2014).