Treat-to-target strategies have been widely adopted as the standard of care for the management of patients with rheumatoid arthritis. The concept of ‘tight control’ is prevalent in other disciplines, particularly in diabetes and cardiovascular risk management. In these disciplines, evidence has accumulated that the utility of tight control strategies must be carefully weighed against the disutility that may arise from multiple interventions, particularly in patients at low risk. There is a lively debate in rheumatology circles about whether treatment should be targeted at achieving low disease activity, clinical remission or imaging remission. As rheumatologists we should learn the lessons from other disciplines, and ensure that we expand the evidence base to ensure our recommendations are securely underpinned by robust evidence.
- Rheumatoid Arthritis
- Disease Activity
Statistics from Altmetric.com
A great deal can be learned by rheumatologists from studying other disciplines. Parallels have been drawn between the treatment of type 1 diabetes mellitus (DM) and rheumatoid arthritis (RA). The seminal Diabetes Control and Complications Trial showed that tight glycaemic control, delivered through multiple daily insulin injections, resulted in substantial improvements in microvascular complications such as diabetic retinopathy.1 Similarly, ‘tight control’ of disease activity in RA yields significant increases in the rate of remission, improved physical function and reduced radiographic progression.2 As the evidence base has developed, treat-to-target (T2T) strategies for the treatment of RA have been endorsed by national guidelines and international consensus recommendations.3
As T2T strategies of care have been more widely adopted, debate has emerged about which target should be pursued, and remission is often stated as the appropriate target. The original studies demonstrating the advantages of T2T strategies largely employed low disease activity (LDAS) as the target,2 ,4 ,5 although some studies have used remission as the treatment target.6 The aspiration to treat to remission obscures the fact that there is no universally accepted definition of remission—is disease activity score (DAS)/DAS28 remission sufficient? Should the stricter Boolean American College of Rheumatology/European League against Rheumatism (ACR/EULAR) remission criteria be pursued? Or, perhaps, should imaging remission be the modern goal?7 There are grounds for considering the pursuit of imaging remission. It has been demonstrated that subclinical synovitis can be identified using musculoskeletal ultrasound (MSUS) or MRI in a significant proportion of patients in clinical remission.8 Moreover, the presence of MSUS synovitis is associated with ongoing radiographic progression and is associated with an increased risk of subsequent clinical flare.9 Equally, there are grounds for caution: will the incremental benefits of pursuing imaging remission in patients in sustained clinical remission outweigh any disadvantages?
It is at this point that we might learn from other disciplines (although exercising caution in extrapolating our knowledge between different diseases), in which the concept of disutility associated with the pursuit of ultratight control of risk factors for long-term disease has been studied—the utility of interventions that improve one dimension of outcome must be balanced against the disutility associated with those interventions. So, for instance, a meta-analysis of studies comparing intensive and moderate dose statin therapy explored the relationship between cholesterol reduction and important outcomes: the use of intensive statins was associated with a reduced risk of major cardiovascular (CV) events, but also with an increased risk of incident DM.10 While further reduction in the rate of CV events is clearly more important than a modest rise in diabetes in this scenario, the disutility of developing DM must nevertheless be taken into account. Is there a useful parallel between CV disease and RA? In RA, the routine use of oral corticosteroids has been advocated because of an increased rate of remission11 and reduced radiographic progression.12 However, there is epidemiological evidence that oral corticosteroid use in RA is associated with an increased risk of incident DM13 and myocardial infarction.14 The benefits of corticosteroid therapy in RA are likely to outweigh the risks in some clinical situations—for instance, short-term use of corticosteroids in early disease may harness most of the benefits of treatment while minimising any harm. Assessing the disutility associated with adverse effects is not straightforward, but the potential harm of therapy does need to be taken into account.15
Observations about the utility of tight and ultratight control of diabetes continue to be illuminating too—the evidence that intensive glycaemic control in DM results in better outcomes has become more nuanced. First, the benefits of tight control in type 1 DM cannot be extrapolated to type 2 DM, in which tight control does not lead to improved outcomes in many patients. Rather intensive glucose control in type 2 DM is associated with an increased risk of hypoglycaemia, significant weight gain and no vascular benefit or even harm in specific subgroups.16 Second, even in type 1 DM, a subgroup of patients (possibly with those with an underlying predisposition to the development of type 2 DM, so-called ‘double diabetes’) treated intensively develop significant weight gain, which is associated with adverse CV risk factors that may offset some of the benefits of tight glycaemic control.17
Importantly, trials supporting the adoption of T2T strategies aiming for LDAS have incorporated some assessments of disutility—so, for instance, the Tight Control of Rheumatoid Arthritis (TICORA) study found that that those assigned to intensive management: reported fewer adverse events; attracted similar costs from the perspective of the National Health Service (NHS) and had improved generic health-related quality of life, not just better disease-specific outcomes.2 But caution should be exercised before extrapolating these results to more stringent targets. Modelling the utility/disutility of the pursuit of tight control of risk factors for CV disease has demonstrated some important factors18:
The incremental benefit of additional therapy to control blood pressure and hyperlipidaemia reduces the more drugs that are already prescribed—the biggest impact is achieved though the first intervention(s). In terms of the management of RA, these might equate to the substantial benefits of using methotrexate, and stepping up to triple therapy in patients not achieving LDAS. The incremental benefit of adding additional disease modifying anti-rheumatic drug (DMARDs), corticosteroids or biologic drugs is likely to be smaller in patients already in LDAS or clinical remission with methotrexate (MTX) monotherapy or triple therapy.
The greatest benefits of CV risk reduction accrue in patients who are at highest risk of subsequent CV events (ie, those with existing cardiovascular disease), so pursuing very intensive targets for cholesterol and blood pressure control in patients at low risk of CV events may not necessarily be beneficial. In RA, there are proponents for the argument that personalised medicine should be abandoned (at least for the time being) in favour of rigorous control of disease activity,19 but this may not be in the best interests of patients who have the lowest risk of disease progression. For example, there are patients who have no persistent synovitis who nonetheless have elevation of DAS/DAS28 because of comorbidity (eg, fibromyalgia and high patient global assessment of disease, or dysproteinaemia and high erythrocyte sedimentation rate (ESR)) in whom pursuit of ‘clinical remission’ (as currently measured) may not be appropriate.
Disutility may accrue from factors such as polypharmacy or drug-related adverse effects; these can sometimes be more difficult to quantify, but are real nonetheless given the association between polypharmacy, poor adherence and poorer outcomes. It has recently been reported that very good outcomes can be achieved using triple therapy, oral corticosteroids and infliximab in patients with newly diagnosed RA.20 The question should be asked: “Is this very intensive regimen appropriate for all patients with newly diagnosed RA irrespective of their personal risk of disease progression?” There may be some patients at lowest risk, for whom the pursuit of more costly, very stringent treatment targets results in no net gain or even net harm (for instance, from rare but serious adverse events such as opportunistic infection). However, we also need to recognise that our ability to stratify patients according to prognosis is currently very limited, and allocation of less intensive treatment according to perceived good prognosis runs the risk of undertreating some patients.21
In another field, there has been concern that sometimes the pursuit of commendable goals (eg, the reduction of mortality) may become coloured by conflicts of interest—for example, recommendations to raise the age at which women in the USA should receive routine mammography have been opposed by breast radiologists, who (it has been argued) have a vested interest in the breast screening programme.22 Rheumatologists (potentially) and the pharmaceutical industry (undoubtedly) have vested interests in the pursuit of stringent targets of remission in RA, so it is incumbent upon us to ensure that our recommendations are truly in the best interests of all our patients (and healthcare systems) and supported by high-quality evidence.
What then, is the best way forward? We should learn from other disciplines such as diabetology, in which a simple message—‘Tight glycaemic control results in superior outcomes’—has become more nuanced over time. The critical issue is the development of the evidence base—before we endorse strategies of care that pursue remission, for example, we must:
be careful to define what we mean by remission;
test the incremental benefit of pursuing remission rather than LDAS, or imaging remission rather than LDAS/clinical remission;
capture the impact of therapeutic strategies on generic health-related quality of life and comorbidities as well as disease-specific outcomes;
explore whether any advantages accruing are restricted to subpopulations of patients at high risk, although more accurate predictors of prognosis and harm are needed;
ensure that trials capture disutility (adverse effects, harms and additional costs), as well as utility using a range of patient-centred outcomes.
Is this achievable? We believe so—for example, there are studies underway that are testing the benefit of integrating MSUS into our paradigms of care in the pursuit of MSUS remission (eg, Targeting synovitis in Early RA study, NCT00920478). Until this evidence emerges, we should be cautious about adopting strategies of care in routine practice that go beyond the available evidence base.
Handling editor Tore K Kvien
Contributors All authors were involved in manuscript preparation and approval of the final version.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.