nature reviews rheumatology https://doi.org/10.

1038/s41584-024-01169-7

Review article Check for updates

Management strategies in
rheumatoid arthritis
Victoria Konzett & Daniel Aletaha
Abstract Sections

Management of rheumatoid arthritis (RA) has evolved from simply the Introduction

direct translation of drug efficacy results from clinical trials to patient Why is strategy important?
care, to a more complex longitudinal process that considers not only The precision gap in
drug efficacy but also the safety gestalt of a treatment and patient profiles rheumatoid arthritis
and preferences, as well as health-economic factors. With numerous The therapeutic matchmaking
DMARDs available to treat RA, knowledge about trial efficacy becomes approach leading to precision
tuning
less important than data that inform an appropriate clinical strategy for
Treating to the right target
their optimal selection and use. Overly ambitious approaches targeting
the ‘maximum’ level of success could, for example, be prone to failure Smart-to-target: finding
a sweet spot of effort and
and create frustration, and lead to a large number of patients then success in each patient
being considered as ‘difficult to treat’. Safety profiles might be more
Bringing it all together
informative than efficacy profiles for precision medicine approaches.
Conclusions
Contemporary RA management strategies might therefore take a more
holistic approach, beyond merely efficacy, to the setting of targets that
lead to improved compliance rather than aspirational successes, with
consideration of each patient’s multimorbidity profile and preferences,
as well as the safety profile of each treatment. Ultimately, the goal remains
unchanged: maximizing health-related quality of life; however, with a
focus on optimal balance rather than superlatives.

Department of Medicine III, Division of Rheumatology, Medical University of Vienna, Vienna, Austria.
e-mail: daniel.aletaha@meduniwien.ac.at

Nature Reviews Rheumatology

Review article

Key points are available: new questions become relevant, such as those related to
the use of a step-up or a step-down approach, monotherapy or combi-
nation therapy, predicting preferential response to one or the other
• The variety of options available for the treatment of rheumatoid compound, and many more. Answers to these strategic questions need
arthritis (RA) necessitates strategies that guide the optimal choice a solid evidence base, which has in the past decade led to the design of
for each individual patient, the objective of precision medicine. ‘strategic’ clinical trials, an approach that is clearly different from the
traditional design of ‘efficacy’ trials that were used to support regula-
• Precision medicine strategies based on predicting differences in tory approval claims14,15. Rather than asking whether a drug works bet-
drug efficacy have not sufficiently enabled the guidance of treatment ter than placebo, strategic trials answer questions relevant to clinical
choice in clinical practice, a phenomenon referred to as the ‘precision practice16–19, and might help in addressing the questions of ‘when?’,
gap’ in RA. ‘in whom?’ and ‘how?’, among others20,21. The need for strategic trials
increases with the number of compounds in the therapeutic armamen-
• To overcome the precision gap, contemporary RA management tarium (Fig. 1a), and in their absence treatment guidance rests largely
might need to evolve towards a more holistic approach that goes on empirical beliefs or a trial-and-error approach22.
beyond efficacy, to include consideration of individual patient One exception in this regard might be the use of methotrexate,
preferences and multimorbidity and safety aspects of drugs, in a which undoubtedly continues to be the first-choice DMARD in RA23,24.
‘therapeutic matchmaking’ approach. However, after treatment initiation, patients will start to show dif-
ferent patterns of response to methotrexate, and patients’ journeys
• The defined treatment target needs to be ambitious but realistic and through the course of their disease will start to diverge. Ultimately,
take into account the potentially disproportionate increase in effort the combination of disease and treatment will unfold over time and
required to achieve overly ambitious targets, which we refer to as a create highly individualistic scenarios (Fig. 1b); however, these specific
‘smart-to-target’ approach. patterns are currently not based on decision-making informed by
evidence, but instead depend on a multitude of external factors. The
scarcity of evidence that would enable decisions about the best time
Introduction for treatment in an individual patient is a relevant clinical problem.
Management of rheumatoid arthritis (RA) has come a long way. For A comprehensive management strategy for RA must therefore also
many years, treatment choices were limited, and most available thera- overcome this evident ‘precision gap’.
pies were of only low grade or questionable effectiveness, such as gold
salts and d-penicillamine1–3. Although the introduction of glucocor- The precision gap in rheumatoid arthritis
ticoids was a milestone in the treatment of RA, their cumulative use Precision medicine in RA remains aspirational. Many attempts have
soon raised concerns about their benefit-to-risk ratio, and the role been made to identify predictive biomarkers to guide treatment selec-
of glucocorticoids in the management of RA remains controversial tion, and numerous positive results have been reported25–27. However,
to this day4,5. The advent of methotrexate in the 1980s then dramati- the studies for which these analyses were reported often lacked the
cally changed the therapeutic field of RA in an unprecedented way6. breadth of treatment options necessary to appropriately mirror real-life
At that time and during the subsequent decades, methotrexate showed clinical choices. For example, many studies of these markers involved
impressive effects, which were measured by improvements in signs secondary analyses of data obtained from clinical trials, which rarely
and symptoms of arthritis, including the number of active joints, pain, compared more than two compounds. Others were circular by design
stiffness and physical function7,8. Nevertheless, the safety of metho- (for example, the analysis of markers in data from clinical cohorts):
trexate and other chemical compounds used at the time remained although this setting naturally includes the most relevant drug choices
a concern9,10, and it was also safety that guided RA management as a in clinical practice, it suffers from bias by indication; that is, in clinical
principle (the ‘first do no harm’ concept and the ‘pyramid approach’)11,12. practice drug choice is not random but based on prior, sometimes
The unexpected expansion of treatment options in subsequent years unclear and unmeasurable assumptions that conflate the interpretabil-
then led to a plethora of drugs for RA13, and transformed the needs of RA ity of the results. Certainly, the precision gap is currently being tackled
management, as the lack of strategies for the use of all these treatment from many sides, and other markers are being tested in a randomized
options became evident. setting in dedicated precision medicine trials. However, the usability
In this Review, we do not discuss mechanisms of disease or diag- of these biomarkers for clinical practice is still not advanced28,29.
nostic aspects. Instead, we elaborate on the components of a novel Precision medicine is an important unmet need in RA, specifically
approach to RA management, considering the fascinating landscape because patterns of response to biological therapies are strikingly simi-
of therapeutic options, combined with discussion of how to define lar at the group level in clinical trials30. The pattern is easily recognizable
the right treatment target for the right patient, and individualized when the major outcome metrics for RA, the ACR response criteria
approaches to drug selection that go far beyond the consideration of ACR20, ACR50 and ACR70, are considered31: all biologic DMARDs
efficacy and include patient multimorbidity, drug safety and context achieve response rates of roughly 60%, 40% and 20%, respectively,
in general. when investigated in groups of patients with an insufficient response
to methotrexate32. This clearly supports the concept of a uniform
Why is strategy important? response paradigm33,34. As good predictors of drug effectiveness in
Strategic management of a chronic disease would be of little relevance individual patients are lacking to date, new strategies will need to go
if only a single therapeutic compound were available: in the absence of beyond effectiveness and also consider a broader context to inform
contraindications, that drug would be the preferred choice. This situ- treatment choices. For example, they might include a strategic risk
ation already changes considerably if two or more treatment options assessment, consider the toxicity risk and profile of treatments, the

Nature Reviews Rheumatology

Review article

a b
100

SpA

PsA

RA
Relevance to disease
management (%)

Clinical
effectiveness Clinical
50
Informed by strategy
efficacy trials Informed
by
strategic
trials
0
Number of available DMARDs Time (months)

Fig. 1 | The complexity of rheumatoid arthritis management. a, The arthritis (PsA) and axial spondyloarthritis (SpA) are conceptually placed on
importance of strategic considerations in rheumatoid arthritis (RA) a scale to indicate that RA is currently the entity with the widest spectrum
management. As the number of DMARD options increases, information on of treatment options and thus the greatest need for strategic trials. b, High
drug efficacy solely from placebo-controlled trials is discounted in relation to variability of individual RA patient journeys. In RA, most patients are initially
the importance of strategic trials that address specific and clinically relevant treated with methotrexate, but subsequent treatment segments evolve to
questions. In this illustration, three major arthritic conditions, RA, psoriatic create high variability in therapeutic journeys among individuals with RA.

susceptibility of individual patients, who are typically multimorbid, be defined for a given compound, the more informative it will become
and the risk of the underlying disease, when not treated adequately, and in a therapeutic matchmaking approach: the respective safety profile
in general will benefit from a sound evaluation and documentation of of a treatment can be used as a biomarker that can be matched with a
all relevant patient and disease characteristics and how they change patient’s multimorbidity profile and thus support precision medicine
over time in routine care35. In fact, in clinical practice it is an intuitive by, for example, excluding a specific drug from consideration in a
principle to incorporate these aspects into decision-making. However, respective at-risk patient.
this is mostly done implicitly and potentially not comprehensively, The causal sequence of potential aggravation of a pre-existing dis-
or systematically. ease is well established for some organ systems; for example, in patients
with elevated levels of liver enzymes and the risk of worsening upon
The therapeutic matchmaking approach leading treatment with methotrexate or IL-6 inhibitors51, the presence of con-
to precision tuning gestive heart failure and its potential exacerbation following treatment
Patients with RA typically have an impressive multimorbidity with TNF inhibitors52,53 or impaired insulin resistance and its worsening
pattern36–38, meaning that they are not only at risk of RA-related con- through treatment with glucocorticoids54. For some other conditions,
sequences, but that other medical conditions also need to be consid- the evidence that their presence is a risk factor for an organ-specific
ered in the therapeutic decision-making process39–41. Comorbidities adverse event is not as strong. For example, sulfasalazine can cause
can become especially relevant in advanced disease stages, as they agranulocytosis, but the presence of leukopenia before treatment is not
not only increase disease burden but also reduce available treatment necessarily a risk factor for a future treatment-related event, which in
options owing to drug–drug interactions and adverse reactions35,42,43. this case has an idiosyncratic nature55. Likewise, in patients with a history
However, the worldwide shortage of rheumatologists, which has been of malignancy, treatment with drugs suspected of increasing the risk of
highlighted by various assessments of the rheumatology work force malignancy (for example, abatacept or Janus kinase ( JAK) inhibitors56,57)
that also show a continuously increasing demand44,45, makes it impos- might not specifically increase the risk of recurrence or new manifesta-
sible in most settings for rheumatologists to proactively manage their tion of malignancies. Similarly, the presence of demyelinating disease
patients’ comorbidities40. Rheumatologists therefore often need to rely might not increase the risk of multiple sclerosis in patients treated with
on other specialties, for example to treat diabetes mellitus or hyper- TNF inhibitors (although demyelinating disease has been reported
tension, while the relevance — and consideration — of multimorbidity as a potential adverse effect of TNF inhibition)58. Regardless of these
status is key to improving the overall health-related quality of life of examples, avoidance of certain compounds is a common strategy and
their patients with RA46–48. is intuitive from a clinical perspective of risk minimization: a lack of
Similarly, rheumatologists are aware of the broad and diverse evidence of increased risk related to pre-existing disease is not evidence
spectrum of potential adverse events associated with the therapeu- for an absence of such risk. This concept is also reflected in certain
tics they use: although many safety concerns, such as overall risk of recommendations such as the 2021 ACR guidelines for the treatment
infection or injection site reactions49,50, are shared among the vari- of RA, which conditionally recommend the use of a non-TNF inhibitor
ous treatments for RA, most drug classes are associated with certain biologic DMARD or JAK inhibitor in patients with moderate-to-severe
adverse effects that are not commonly observed with others. These heart failure despite a lack of strong evidence for a risk of aggravation
more specific adverse effects are typically detected or determined in of heart failure in patients with RA treated with a TNF inhibitor59.
clinical trials, and can relate to one or more specific organ systems. The concept of therapeutic matchmaking considers both the
In their most obvious form, they will culminate in contraindications and patient’s multimorbidity profile as well as the drug’s safety profile, and can
their mention in the summary of product characteristic and package therefore effectively guide preferential treatment selection in RA even if
inserts of a drug. In totality, however, these ‘special’ adverse events efficacy outcomes alone do not enable this. This approach therefore leads
can contribute to a safety signature that is potentially unique to each to an improved precision in selecting treatments (‘precision tuning’).
compound. Practically, the more distinctively this safety signature can Figure 2 illustrates how various contraindications, potential adverse

Nature Reviews Rheumatology

Review article

4
3 5

2 Diverticulitis?
Malignancy? Autoantibodies, 6
DMD?
ASCVD,
VTE, HZ? HG,
Y
1 Y Y vaccination?
T I R 7
Y T I R A J T I R
Heart A J Y
failure? A J
T I R RA-ILD?
A J TI R TIR T I R
I AR AI R
J
A AJ A AJ A J
Y T R A A J
TJ A R AJ T I T I R I Y
IR AJ I
A I AJ I
T I R RJ A TR AJ
AJ I
A J R A I T I R
A A T J A J I I
A I R T I R I I A J
J RJ I
TI A A AJ AJ I
R A TR A
R J AJ I I A
T T I
I R J T I R A A
I R A I A
I T T R A J A
R I A
R I T I A J A
R I A
I R A A
I T I R I
I I A
T T I A J A
R R A A A
R T R I IA
R A
T A
T I I
T T T I R I IA
R T I I
A A
A
R A J A
R T R A I
I
R R A IA
R A I A
A I
T I R T I R A I A IA
TI T I
AR AR A J A J J
T I
RA AR T I
TIA T I AR 1
A A T I
I
IR I R I R A R T I 2 A I
AR A A A R J
A I
I A TR
A TR 3 I
A TR A I
TR A A T I A I I I
A R T R R J I
A T J I I
A
A R A T I I
R 4 A I
A R A T I I
A R T A I J I
A R R J A I
A I I
A T I 5 J J I
A TI R R T A I
T A I A I A J
R I R R J A J A
R T I J J
A = Abatacept I I R 6
J A J A A
I T AI A A
I = IL-6 inhibitor R R R T AI AI AI J J A
J = JAK inhibitor I T R J J J J A A
I R R A Condition not fulfilled
R = Rituximab 7 AI A
T T R AI A AI A AI A J
T = TNF inhibitor R J Condition fulfilled
J J

Fig. 2 | The therapeutic matchmaking approach leading to ‘precision tuning’. be favoured in patients with current or recent malignancies57,112–114. (4) IL-6
In the absence of effective guidance for treatment selection, an implicit clinical inhibitors can increase the risk of gastrointestinal perforations, especially in
matching of comorbidities and treatment safety profiles could take place in patients with a history of or current diverticulitis115,116; this potential risk has
clinical practice. Individual comorbidity scenarios (1 to 7) commonly narrow also been observed for JAK inhibitors57,117,118. (5) Formation of autoantibodies
the choice of therapeutic options, and their combination in a multimorbidity (for example, antinuclear or anti-double-stranded DNA antibodies)119,120 and
setting could lead to few or even a single remaining option. (1) TNF inhibitors demyelinating disease (DMD) such as multiple sclerosis or Guillain–Barré
are contraindicated in the setting of moderate-to-severe heart failure (New syndrome have been observed in patients on TNF inhibitor therapy121–123.
York Heart Association (NYHA) class III or IV)59,109, and rituximab in severe (6) Hypogammaglobulinaemia (HG) could disfavour rituximab110, and response
heart failure (NYHA class IV)110. (2) Janus kinase ( JAK) inhibitors should be to any planned vaccination might be limited after rituximab treatment has
prescribed only after careful consideration of benefit and risk in patients been started124,125. (7) Conversely, specific extra-articular efficacy can positively
with a history of or current atherosclerotic cardiovascular disease (ASCVD), select DMARDs: in RA-associated interstitial lung disease (RA-ILD), rituximab
venous thromboembolism (VTE) or herpes zoster (HZ)57,108,111,112. (3) Abatacept or abatacept might be favoured over TNF inhibitors, IL-6 inhibitors and JAK
and JAK inhibitors might increase the risk of malignancies, and might not inhibitors60,61.

reactions and comorbidities could help rule out various treatment effect on arthritis, certain specific efficacy aspects might positively
options, leaving the treating physician with a reduced list of alternatives select some compounds over others. For example, individuals with
to choose from. In a sense, this method favours the alignment of complete interstitial lung disease could benefit preferentially from abatacept or
patient profiles with drug profiles over a merely efficacy-driven approach. rituximab treatment, considering their additional effects on a specific
Conversely, it may be that, in addition to the classic anti-inflammatory extra-articular disease manifestation beyond those on arthritis60,61.

Nature Reviews Rheumatology

Review article

Yet, therapeutic matchmaking goes beyond all of the above- to whether residual disease activity, as in low disease activity, might
mentioned efficacy, multimorbidity and safety aspects of drugs, and be acceptable as a target in some patients if structural and functional
further extends to patient preferences, which can help to exclude or integrity is preserved over time.
select certain compounds that remain as options after others have been Regarding the first question, several studies have investigated
ruled out. These preferences include subjective beliefs about certain subclinical imaging targets in addition to clinical targets. Two of these
drugs, which could be related to a preferred route of administration, trials used ultrasonography and one used MRI to define the subclinical
personal beliefs and treatment adherence considerations in general. target77–79. All three studies concluded that the predefined primary
Thorough evaluation and documentation of patient and disease fea- outcomes of strategies targeting subclinical and clinical outcomes
tures is also relevant for effective therapeutic matchmaking. For exam- were not different. Targeting subclinical outcomes rather than clini-
ple, injectables might be disfavoured by frequent travellers (owing to cal ones therefore failed to yield incremental benefit for patients, and
the need for cooling) or by those who are unable to self-administer a will in addition require higher levels of therapeutic effort. ‘Effort’ in
compound (owing to fear of injections)62. Finally, the overall healthcare that context is a broad and general term, but conceptually extends to
setting, including drug availability, treatment costs and reimbursement the important concept of increased risk of adverse events, increased
landscapes, is another angle of ‘context’ that can strongly influence cost, or similar.
treatment choices in RA management63. Regarding the second question, of whether an alternative tar-
All these angles can and will shorten the list of therapeutic options get that is less stringent than remission would be acceptable, several
in clinical reality, and their consideration effectively becomes an act factors can be considered. Importantly, this question needs to be
of precision medicine. Therapeutic matchmaking could be further approached in an individualized way and will only apply to a certain
expanded by comprehensively accessing data on drug safety and elec- number of patients. One scenario in which low disease activity would
tronic medical health records of patients and utilizing artificial intel- be an alternative to remission is, as mentioned in the original publica-
ligence and/or deep learning models to provide a holistic approach tion of the T2T principle, in patients with long-standing established
(taking into consideration efficacy, safety, comorbidity, preference and disease, given the expected lower likelihood of response to treatment
cost). Certainly, the matchmaking will be of better quality in patients for in individuals who have already been exposed to multiple therapeutic
whom data on health and disease have been more thoroughly collected. courses80.
A second scenario relates to patients with isolated or centralized
Treating to the right target pain syndromes, most prominently secondary fibromyalgia. Pain has a
Regardless of how well therapeutic matchmaking can be implemented, major influence on disease activity self-assessed by patients and might
every treatment puts patients at potential risk of harm. Therefore, or might not be related to the active inflammatory disease processes
understanding the most appropriate treatment target for each indi- in RA81. Although both pain and inflammation are relevant from the
vidual patient is also crucial, since a less ambitious target will in general patient’s perspective, only the inflammatory component is amena-
require less aggressive treatment. Since 2010, RA management has ble to DMARD treatment. Patients with isolated pain syndromes or
been governed by the treat-to-target (T2T) approach21. The T2T con- centralized pain might therefore be unable to attain the ambitious
cept implies that a treatment target is clearly defined on a quantitative target of remission because of secondary problems. In these patients,
disease activity scale, and that the achievement of this target (or failure composite outcomes fail to be accurate. Pain is a determinant of many
to achieve this target) should be evaluated repeatedly within the first patient-reported outcomes, such as the patient global assessment or the
6 months of therapy. If the target is not reached during this period, tender joint count82. If pain is unrelated to inflammatory disease activ-
the treatment needs to be modified. As simple as this concept might ity, DMARD treatments will not be able to further improve a patient’s
sound, challenges lie in the details of defining the target. state (for example, from a state of low disease activity to remission).
An important aspect is the metric used to measure disease activity. A state of low disease activity could therefore be an acceptable alterna-
Several continuous scales exist, including the most commonly used tive to remission in the presence of secondary pain syndromes. Of note,
clinical disease activity index (CDAI), simplified disease activity index this assertion obviously does not mean that the pain syndrome should
(SDAI), and 28-joint disease activity score (DAS28)64–67. Although the not be managed as well. The ACR–EULAR criteria for remission have
comparability of these scales might be limited, particularly between been adapted to allow for a higher patient global assessment threshold
CDAI or SDAI and the DAS28, the most important objective of T2T is in appreciation of this issue83,84. In other words, although remission was
already achieved once any target is clearly defined (regardless of the kept as a target, its definition was adapted to control for the discrepant
scale used to assess the target) and is prioritized over using clinical patient-reported outcomes and inflammatory markers that are often
impression only68. observable in patients with RA85.
Another important aspect concerns what level of disease activity Finally, a biological reason for accepting low disease activity as a
should be the target. There is a large body of literature about control target (instead of remission) is based on the type of treatment that is
of disease activity and prevention of poor long-term outcomes, such being used. Structural progression is inhibited if remission is achieved,
as structural damage and functional disability69–74. This association regardless of the type of compound used. If ‘only’ low disease activ-
between disease activity and damage is typically based on a ‘the less, the ity is achieved, available data imply that structural progression will
better’ principle. Therefore, defining an optimal cut-off point for use as continue if treatment with methotrexate (or another conventional
target on one of the continuous disease activity scales is challenging for synthetic DMARD) is in use. However, if a state of low disease activity
several reasons. First, is clinical remission a stringent enough target? is reached with the use of a biologic DMARD or a JAK inhibitor, then
In other words, can we do even better by adopting subclinical targets structural progression is still abrogated (as it is in remission). This
(such as imaging remission, laboratory remission or molecular remis- paradigm of non-progression of structural damage despite ongoing
sion) as opposed to (only) clinical ones75,76. Second, and conversely, do (low) disease activity has been referred to as the ‘dissociation’ concept
we really always need to aim for clinical remission? This question relates of RA86. The explanation for this dissociation could be that rapid control

Nature Reviews Rheumatology

 Review article

a Not present Highly present

b Inflammation Erosion WT
1.25 0.2 Human-TNF transgene

inflammation (mm2)

bone erosion (mm2)

1.00
NS

Area of

Area of
0.75
P < 0.01 0.1
TNF level

0.50
P < 0.05
0.25 P < 0.01 P < 0.05 P < 0.05
Threshold
levels for 0 0
structural

g
en

en
W

W
1m

1m

1m

1m

m
damage

lv

lv
10

10
0.

0.
So

So
Adalimumab Adalimumab

Inflammation Structural damage

Solvent Solvent

0.1 mg 0.1 mg

10 mg 10 mg

Fig. 3 | Threshold hypothesis and ‘dissociation’ concept. a, Conceptual b, Experimental confirmation. In a TNF-transgenic murine model, adalimumab
background. Rapid and substantial suppression of TNF levels to below the therapy significantly reduced the progression of erosion even at low doses,
threshold necessary for osteoclast activation might halt radiographic progression whereas its effect on inflammatory disease activity was dose-dependent. NS, not
and progression of structural damage, despite ongoing (low-grade) inflammation. significant; WT, wildt ype. Part b adapted with permission from ref. 92, Wiley.

of TNF through the use of modern therapies leads to a fall in TNF levels targeted therapies could halt radiographic progression despite ongo-
to below the threshold that is necessary for osteoclast activation, ing low disease activity86. Regardless of the ultimate target used in
whereas clinical control of disease activity correlates more linearly the T2T approach, the utmost importance lies not in reaching, but
with quantitative TNF levels (Fig. 3a). In this way, low disease activity in maintaining, the target state.
can be a situation in which osteoclast activation is (fully) controlled,
despite clinical manifestations persisting. This theory was derived from Smart-to-target: finding a sweet spot of effort and
clinical observations and secondary data analysis, and holds true for success in each patient
most targeted therapies, including TNF inhibitors87,88, IL-6 inhibitors89, Success in management of RA and the efforts required to achieve
rituximab90 and JAK inhibitors91; it has not been tested for abatacept. such success does not necessarily follow a linear relationship. As in
The theory was further underpinned by an experimental approach that most processes in biology and even in social sciences, a large amount
demonstrated an on–off switch for structural progression with TNF of success can be achieved with a small amount of effort, whereas
inhibition, which was in clear contrast to the more gradual effects of approaching full-scale success requires substantially more effort98
TNF inhibition on inflammation92 (Fig. 3b). (Fig. 5). Importantly, ‘effort’ can also represent adverse effects, costs
One important final note about T2T is that this strategy is not and/or any other dimension related to increased levels of therapeutic
only about reaching a target, but also about sustained achievement of effort. This Pareto principle98 can be translated to the definition of the
a target over time. Maintenance of a good state of disease activity right target for the right patient, which we term ‘smart-to-target’.
over long periods of time is still an unmet need in clinical reality93–96. By way of illustration, in the setting of untreated early RA, most of
Moreover, the sustainability of a target state again links into the previ- the potential success can be achieved with very little effort; for exam-
ous discussions regarding the definition of the target in general; for ple, with the standard use of first-line oral methotrexate, which leads
example, maintaining a target of clinical remission over time could to a good response in terms of disease activity in most treatment-naive
well be expected to also improve markers of subclinical remission, patients99–102 (Fig. 5, scenario 1). A further increase of therapeutic effort
without those markers being the direct target (Fig. 4). The higher the to move from remission to subclinical remission would produce small
sensitivity of a subclinical marker, the more time is required for it to incremental benefits while requiring much higher levels of therapeutic
normalize after onset of clinical remission. A clinical analysis that linked effort (which includes higher risks and costs). If the patient’s disease is
time with absence of joint swelling to ultrasonographic findings at the difficult to treat or refractory, even stronger treatment escalations and
level of individual joints supports this hypothesis97. even higher levels of effort would often be necessary to actually trans-
In summary, the true value of reaching ‘maximum’ treatment fer them into the ‘target’ zone of disease activity (Fig. 5, scenario 2).
effects is not clear. Data currently do not support the use of sub- The effort-to-success curve can also deviate in patients with comorbidi-
clinical targets over merely clinical targets77–79, and even support ties, as such patients carry higher risks of treatment-related adverse
targeting low disease activity over remission in some situations, reactions103–105 (Fig. 5, scenario 3); again, increasing risks and costs
as in long-standing disease80 or in the presence of comorbidities have to be accepted to achieve a certain level of treatment success in
such as centralized pain syndromes83,84. Also, treatment with novel this scenario.

Nature Reviews Rheumatology

Review article

Onset of of other elements, for which the evidence might not be clear or suf-
clinical
remission ficient. In clinical practice, this implicit integration is necessary to
bridge the precision medicine gap, which is still very evident in RA.
This bridge could be pragmatically established with a consequent
Specific imaging signals
trial-and-error-based T2T approach, which however will also expose
Sensitive imaging signals
Imaging and sensitive patients to medicines that they would not need. A therapeutic match-
Subclinical activity

biochemical signals making process can be utilized in clinical practice, whereby the phy-
sician’s offer of therapy is based also on safety considerations rather
than indiscriminatory expectations of effectiveness. Several relevant
comorbidities can be sequentially evaluated to more formally narrow
the therapeutic options in a given patient. Although this is a conceptual
approach (with differing levels of evidence for individual comorbidities
or the safety of different drugs), it will advance clinical decision-making
Time in sustained clinical remission
from trial-and-error to a more informed process. Importantly, this
Fig. 4 | The importance of maintaining treatment targets. The ultimate approach will probably also incorporate patient preferences, which
goal of the treat-to-target approach is not just reaching the defined target but could further reduce the therapeutic options.
maintaining the target state over time. Sustainment of clinical remission over In this way, the therapeutic matchmaking approach considers
time will result in decreases in remaining subclinical markers of inflammation,
context and also provides a living practice of shared decision-making,
such as imaging signals or laboratory or molecular parameters.
a concept that has been promoted for many years as an anchor for the
management of many chronic immune-mediated diseases21,108. Context
obviously also incorporates patient demographics and disease stage;
Being treatment naive, experienced or refractory, however, is for example, a drug might safely be used in a young, healthy individual
not only related to an expectation of the effectiveness of a new treat- at the beginning of the disease, but can have a serious negative impact
ment approach, but also to defining disease stage more generally: on the quality of life in an older, multimorbid patient with established
in the course of a patient’s therapeutic journey, their preferences, RA. In summary, there are highly variable elements on both sides of the
expectations, adherence and tolerability change, and their comor- benefit–risk scale, and it is their integration that leads to directions for
bidities as well as their willingness to take risk can evolve. All these clinical decision-making.
aspects become defining elements of disease management. Treating Success can also be viewed as achieving an effect in relation to a
to target therefore becomes a highly individualized process, with an prior expectation. As such, it is also important to set the targets for
individual effort-to-success curve for each patient that is dependent disease control in a feasible, realistic manner, and the ultimate claim
on treatment targets and personal risks and efforts. It is important
to understand, or at least estimate, the slope of each individual’s
effort-to-success curve to be able to treat them in a smart-to-target 1 'Subclinical'
remission
manner. Whereas traditional T2T-defined targets with clear thresh-
Remission
olds on quantitative scales64,65 clearly outline that moderate or high
Low disease
disease activity should under no circumstances be acceptable in 2 activity
Success (%)

patients with RA, the smart-to-target approach would also consider Moderate
a balance of success and effort in approaching the target, which could disease
3 activity
additionally affect management decisions, especially when consider-
ing ambitious targets such as near or complete clinical and subclinical High
disease
remission. For example, in some patients it might be reasonable to activity
aim for very ambitious targets, as they give greater consideration
to the burden of their disease than to potential adverse reactions, Good tolerability Severe AE
and perceive ‘treatment success’ to be achieved only when reaching Effort (%)

states of complete resolution of symptoms such as pain or fatigue. Fig. 5 | ‘Smart-to-target’ with consideration of the balance of effort and
Conversely, in other patients careful dose reductions because of success. Effort-to-success relationships for different clinical scenarios.
adverse reactions might be preferred over complete freedom from Importantly, ‘effort’ could also represent adverse effects (AE), costs or any
symptoms, as long-term structural and functional deterioration are other dimension related to levels of therapeutic effort. (1) In early disease,
still abrogated106,107. effort-to-success relationships are typically non-linear (Pareto principle)98;
Despite all efforts, a proportion of patients still fail to evade a most success can be achieved with acceptable levels of effort, but very high
levels of success (for example, achieving subclinical remission) probably requires
difficult-to-treat phenotype, and continuous investigations into the
disproportionately higher levels of effort. (2) In some patients, such as those
characterization of these phenotypes, as well as into drugs with novel
with long-standing disease, irreversible structural damage or secondary pain
modes of action and precision medicine approaches, are warranted.
syndromes, treatment targets are generally harder to achieve, but the Pareto
principle commonly still applies. (3) Patients with unfavourable safety profiles
Bringing it all together (for example, those with multiple comorbidities) can have a skewed effort-to-
In summary, RA management does not lack therapies; instead, the success (risk–benefit) curve: more risk (adverse events, ‘effort’) needs to be
filtering and utilization of existing therapies needs to be improved. accepted to achieve more success, and the curve might emulate a more linear
More generally, the management of RA in clinical practice has become relationship (with every increase in efficacy leading to an increased risk of
a combination of evidence-based decisions and the implicit integration adverse effects).

Nature Reviews Rheumatology

Review article

for complete absence of any disease activity markers (clinical, sub- 22. Smolen, J. S. & Aletaha, D. Forget personalised medicine and focus on abating disease
activity. Ann. Rheum. Dis. 72, 3–6 (2013).
clinical, molecular, or similar) might not be appropriate in many 23. Pincus, T., Gibson, K. A. & Castrejón, I. Update on methotrexate as the anchor drug for
patients in clinical practice. We label this feasibility-centred strategy rheumatoid arthritis. Bull. Hosp. Jt Dis. 71, S9–S19 (2013).
as a smart-to-target approach, in which the position of the sweet spot 24. Di Matteo, A., Bathon, J. M. & Emery, P. Rheumatoid arthritis. Lancet 402, 2019–2033
(2023).
for success and effort (or risk) needs to be understood to maximize 25. Takeuchi, T. Biomarkers as a treatment guide in rheumatoid arthritis. Clin. Immunol. 186,
health-related quality of life. The most important part, however, is 59–62 (2018).
that beyond the mere definition and reaching of a treatment target, 26. Aletaha, D. Precision medicine and management of rheumatoid arthritis. J. Autoimmun.
110, 102405 (2020).
it should be maintained over time. 27. Wientjes, M. H. M., den Broeder, A. A., Welsing, P. M. J., Verhoef, L. M. & van den Bemt, B. J. F.
Prediction of response to anti-TNF treatment using laboratory biomarkers in patients with
Conclusions 28.
rheumatoid arthritis: a systematic review. RMD Open 8, e002570 (2022).
Humby, F. et al. Rituximab versus tocilizumab in anti-TNF inadequate responder patients
In conclusion, the combination of information about efficacy with the with rheumatoid arthritis (R4RA): 16-week outcomes of a stratified, biopsy-driven,
high-level safety profile of each drug and the individual multimorbid- multicentre, open-label, phase 4 randomised controlled trial. Lancet 397, 305–317 (2021).
29. Rivellese, F. et al. Rituximab versus tocilizumab in rheumatoid arthritis: synovial
ity profile of each patient will help to better identify the right drug for
biopsy-based biomarker analysis of the phase 4 R4RA randomized trial. Nat. Med. 28,
the right patient. If the treatment target also takes into consideration 1256–1268 (2022).
context and the preferences of the patient, and is particularly set in a 30. Burmester, G. R. & Pope, J. E. Novel treatment strategies in rheumatoid arthritis. Lancet
389, 2338–2348 (2017).
realistic way, then patients and their carers will benefit from achiev-
31. Felson, D. T. et al. American College of Rheumatology. Preliminary definition of
ing and, most importantly, maintaining treatment success over time improvement in rheumatoid arthritis. Arthritis Rheum. 38, 727–735 (1995).
in this chronic disease, and directly as well as indirectly maximize 32. Konzett, V., Kerschbaumer, A., Smolen, J. S. & Aletaha, D. Determination of the
most appropriate ACR response definition for contemporary drug approval trials in
health-related quality of life.
rheumatoid arthritis. Ann. Rheum. Dis. 83, 58–64 (2024).
33. Smolen, J. S. & Aletaha, D. Rheumatoid arthritis therapy reappraisal: strategies,
Published online: xx xx xxxx opportunities and challenges. Nat. Rev. Rheumatol. 11, 276–289 (2015).
34. Lin, C. M. A., Cooles, F. A. H. & Isaacs, J. D. Precision medicine: the precision gap in
References rheumatic disease. Nat. Rev. Rheumatol. 18, 725–733 (2022).
1. Bluhm, G. B. The treatment of rheumatoid arthritis with gold. Semin. Arthritis Rheum. 5, 35. Dey, M., Nagy, G. & Nikiphorou, E. Comorbidities and extra-articular manifestations in
147–166 (1975). difficult-to-treat rheumatoid arthritis: different sides of the same coin? Rheumatology 62,
2. Bunch, T. W. & O’Duffy, J. D. Disease-modifying drugs for progressive rheumatoid arthritis. 1773–1779 (2023).
Mayo Clin. Proc. 55, 161–179 (1980). 36. Finckh, A. et al. Global epidemiology of rheumatoid arthritis. Nat. Rev. Rheumatol. 18,
3. Lee, D. M. & Weinblatt, M. E. Rheumatoid arthritis. Lancet 358, 903–911 (2001). 591–602 (2022).
4. Hoes, J. N., Jacobs, J. W. G., Buttgereit, F. & Bijlsma, J. W. J. Current view of glucocorticoid 37. Smolen, J. S., Aletaha, D. & McInnes, I. B. Rheumatoid arthritis. Lancet 388, 2023–2038 (2016).
co-therapy with DMARDs in rheumatoid arthritis. Nat. Rev. Rheumatol. 6, 693–702 38. Radner, H. et al. The impact of multimorbidity status on treatment response in
(2010). rheumatoid arthritis patients initiating disease-modifying anti-rheumatic drugs.
5. Doumen, M., Pazmino, S., Bertrand, D., Westhovens, R. & Verschueren, P. Glucocorticoids Rheumatology 54, 2076–2084 (2015).
in rheumatoid arthritis: balancing benefits and harm by leveraging the therapeutic 39. Dey, M., Nagy, G. & Nikiphorou, E. Comorbidities or extra-articular manifestations: time to
window of opportunity. Joint Bone Spine 90, 105491 (2023). reconsider the terminology? Rheumatology 61, 3881–3883 (2022).
6. Weinstein, G. D. Methotrexate. Ann. Intern. Med. 86, 199–204 (1977). 40. Baillet, A. et al. Points to consider for reporting, screening for and preventing selected
7. Tugwell, P., Bennett, K. & Gent, M. Methotrexate in rheumatoid arthritis. Indications, comorbidities in chronic inflammatory rheumatic diseases in daily practice: a EULAR
contraindications, efficacy, safety. Ann. Intern. Med. 107, 358–366 (1987). initiative. Ann. Rheum. Dis. 75, 965–973 (2016).
8. Weinblatt, M. E. & Kremer, J. M. Methotrexate in rheumatoid arthritis. J. Am. Acad. Dermatol. 41. Agca, R. et al. EULAR recommendations for cardiovascular disease risk management
19, 126–128 (1988). in patients with rheumatoid arthritis and other forms of inflammatory joint disorders:
9. Fries, J. F., Williams, C. A., Ramey, D. & Bloch, D. A. The relative toxicity of 2015/2016 update. Ann. Rheum. Dis. 76, 17–28 (2017).
disease-modifying antirheumatic drugs. Arthritis Rheum. 36, 297–306 (1993). 42. Sandberg, M. E. C. et al. Overweight decreases the chance of achieving good response and
10. Shergy, W. J. et al. Methotrexate-associated hepatotoxicity: retrospective analysis of low disease activity in early rheumatoid arthritis. Ann. Rheum. Dis. 73, 2029–2033 (2014).
210 patients with rheumatoid arthritis. Am. J. Med. 85, 771–774 (1988). 43. Roodenrijs, N. M. T. et al. Non-adherence in difficult-to-treat rheumatoid arthritis from the
11. Smyth, C. J. Therapy of rheumatoid arthritis. A pyramidal plan. Postgrad. Med. 51, 31–39 perspectives of patients and rheumatologists: a concept mapping study. Rheumatology
(1972). 60, 5105–5116 (2021).
12. Fries, J. F. Safety issues related to DMARD therapy. J. Rheumatol. Suppl. 25, 14–17 (1990). 44. Battafarano, D. F. et al. 2015 American College of Rheumatology Workforce Study: supply
13. Aletaha, D. & Smolen, J. S. Diagnosis and management of rheumatoid arthritis: a review. and demand projections of adult rheumatology workforce, 2015-2030. Arthritis Care Res.
JAMA 320, 1360–1372 (2018). 70, 617–626 (2018).
14. Aletaha, D. et al. Reporting disease activity in clinical trials of patients with rheumatoid 45. Kilian, A., Upton, L. A., Battafarano, D. F. & Monrad, S. U. Workforce trends in
arthritis: EULAR/ACR collaborative recommendations. Ann. Rheum. Dis. 67, 1360–1364 rheumatology. Rheum. Dis. Clin. North. Am. 45, 13–26 (2019).
(2008). 46. Tournadre, A., Pereira, B., Gossec, L., Soubrier, M. & Dougados, M. Impact of
15. Felson, D. T. et al. American College of Rheumatology/European League against comorbidities on fatigue in rheumatoid arthritis patients: results from a nurse-led
Rheumatism provisional definition of remission in rheumatoid arthritis for clinical trials. program for comorbidities management (COMEDRA). Joint Bone Spine 86, 55–60 (2019).
Ann. Rheum. Dis. 70, 404–413 (2011). 47. Katz, J. & Bartels, C. M. Multimorbidity in rheumatoid arthritis: literature review and future
16. Østergaard, M. et al. Certolizumab pegol, abatacept, tocilizumab or active conventional directions. Curr. Rheumatol. Rep. 26, 24–35 (2024).
treatment in early rheumatoid arthritis: 48-week clinical and radiographic results of 48. Gossec, L. et al. Screening for and management of comorbidities after a nurse-led
the investigator-initiated randomised controlled NORD-STAR trial. Ann. Rheum. Dis. 82, program: results of a 3-year longitudinal study in 769 established rheumatoid arthritis
1286–1295 (2023). patients. RMD Open 5, e000914 (2019).
17. Verschueren, P. et al. Effectiveness of methotrexate with step-down glucocorticoid 49. Singh, J. A. et al. Risk of serious infection in biological treatment of patients with rheumatoid
remission induction (COBRA Slim) versus other intensive treatment strategies for early arthritis: a systematic review and meta-analysis. Lancet 386, 258–265 (2015).
rheumatoid arthritis in a treat-to-target approach: 1-year results of CareRA, a randomised 50. Sepriano, A. et al. Safety of synthetic and biological DMARDs: a systematic literature
pragmatic open-label superiority trial. Ann. Rheum. Dis. 76, 511–520 (2017). review informing the 2022 update of the EULAR recommendations for the management
18. Gottenberg, J.-E. et al. Non-TNF-targeted biologic vs a second anti-TNF drug to treat of rheumatoid arthritis. Ann. Rheum. Dis. 82, 107–118 (2023).
rheumatoid arthritis in patients with insufficient response to a first anti-TNF drug: 51. Di Martino, V. et al. Busting the myth of methotrexate chronic hepatotoxicity. Nat. Rev.
a randomized clinical trial. JAMA 316, 1172–1180 (2016). Rheumatol. 19, 96–110 (2023).
19. Porter, D. et al. Tumour necrosis factor inhibition versus rituximab for patients with 52. Chung, E. S. et al. Randomized, double-blind, placebo-controlled, pilot trial of
rheumatoid arthritis who require biological treatment (ORBIT): an open-label, infliximab, a chimeric monoclonal antibody to tumor necrosis factor-α, in patients with
randomised controlled, non-inferiority, trial. Lancet 388, 239–247 (2016). moderate-to-severe heart failure: results of the anti-TNF Therapy Against Congestive
20. Sokka, T. & Pincus, T. Rheumatoid arthritis: strategy more important than agent. Lancet Heart Failure (ATTACH) trial. Circulation 107, 3133–3140 (2003).
374, 430–432 (2009). 53. Mann, D. L. et al. Targeted anticytokine therapy in patients with chronic heart failure:
21. Smolen, J. S. et al. Treating rheumatoid arthritis to target: 2014 update of the results of the Randomized Etanercept Worldwide Evaluation (RENEWAL). Circulation 109,
recommendations of an international task force. Ann. Rheum. Dis. 75, 3–15 (2016). 1594–1502 (2004).

Nature Reviews Rheumatology

Review article

54. Do, T. T. H. et al. Glucocorticoid-induced insulin resistance is related to macrophage 86. Smolen, J. S. et al. Radiographic changes in rheumatoid arthritis patients attaining
visceral adipose tissue infiltration. J. Steroid Biochem. Mol. Biol. 185, 150–162 (2019). different disease activity states with methotrexate monotherapy and infliximab
55. Curtis, B. R. Non-chemotherapy drug-induced neutropenia: key points to manage the plus methotrexate: the impacts of remission and tumour necrosis factor blockade.
challenges. Hematology Am. Soc. Hematol. Educ. Program 2017, 187–193 (2017). Ann. Rheum. Dis. 68, 823–827 (2009).
56. de Germay, S., Bagheri, H., Despas, F., Rousseau, V. & Montastruc, F. Abatacept in 87. Smolen, J. S. et al. Evidence of radiographic benefit of treatment with infliximab plus
rheumatoid arthritis and the risk of cancer: a world observational post-marketing study. methotrexate in rheumatoid arthritis patients who had no clinical improvement:
Rheumatology 59, 2360–2367 (2020). a detailed subanalysis of data from the anti-tumor necrosis factor trial in rheumatoid
57. Ytterberg, S. R. et al. Cardiovascular and cancer risk with tofacitinib in rheumatoid arthritis with concomitant therapy study. Arthritis Rheum. 52, 1020–1030 (2005).
arthritis. N. Engl. J. Med. 386, 316–326 (2022). 88. Landewé, R., van der Heijde, D., Klareskog, L., van Vollenhoven, R. & Fatenejad, S.
58. Seror, R. et al. Pattern of demyelination occurring during anti-TNF-α therapy: a French Disconnect between inflammation and joint destruction after treatment with etanercept
national survey. Rheumatology 52, 868–874 (2013). plus methotrexate: results from the trial of etanercept and methotrexate with
59. Fraenel, L. et al. 2021 American College of Rheumatology guideline for the treatment of radiographic and patient outcomes. Arthritis Rheum. 54, 3119–3125 (2006).
rheumatoid arthritis. Arthritis Rheumatol. 73, 1108–1123 (2021). 89. Smolen, J. S., Avila, J. C. M. & Aletaha, D. Tocilizumab inhibits progression of joint damage
60. Narváez, J. et al. SER-SEPAR recommendations for the management of rheumatoid in rheumatoid arthritis irrespective of its anti-inflammatory effects: disassociation of the
arthritis-related interstitial lung disease. Part 2: treatment. Reumatol. Clin. 18, 501–512 (2022). link between inflammation and destruction. Ann. Rheum. Dis. 71, 687–693 (2012).
61. Yu, K.-H. et al. Consensus recommendations on managing the selected comorbidities 90. Aletaha, D., Alasti, F. & Smolen, J. S. Rituximab dissociates the tight link between disease
including cardiovascular disease, osteoporosis, and interstitial lung disease in activity and joint damage in rheumatoid arthritis patients. Ann. Rheum. Dis. 72, 7–12
rheumatoid arthritis. Medicine 101, e28501 (2022). (2013).
62. Salaffi, F., Di Carlo, M., Farah, S. & Carotti, M. Adherence to subcutaneous anti-TNFα 91. Lopez-Romero, P., de la Torre, I., Haladyj, E., Aletaha, D. & Smolen, J. S. Baricitinib further
agents in patients with rheumatoid arthritis is largely influenced by pain and skin enhances disease-modifying effects by uncoupling the link between disease activity and
sensations at the injection site. Int. J. Rheum. Dis. 23, 480–487 (2020). joint structural progression in patients with rheumatoid arthritis. Ann. Rheum. Dis. 81,
63. Hsieh, P.-H. et al. Economic burden of rheumatoid arthritis: a systematic review of 622–631 (2022).
literature in biologic era. Ann. Rheum. Dis. 79, 771–777 (2020). 92. Binder, N. B. et al. Tumor necrosis factor-inhibiting therapy preferentially targets bone
64. Smolen, J. S. et al. A simplified disease activity index for rheumatoid arthritis for use in destruction but not synovial inflammation in a tumor necrosis factor-driven model of
clinical practice. Rheumatology 42, 244–257 (2003). rheumatoid arthritis. Arthritis Rheum. 65, 608–617 (2013).
65. Aletaha, D. et al. Acute phase reactants add little to composite disease activity indices 93. Listing, J. et al. Clinical and functional remission: even though biologics are superior to
for rheumatoid arthritis: validation of a clinical activity score. Arthritis Res. Ther. 7, conventional DMARDs overall success rates remain low-results from RABBIT, the German
R796–R806 (2005). biologics register. Arthritis Res. Ther. 8, R66 (2006).
66. van der Heijde, D. M., Van ’t Hof, M., van Riel, P. L. & van de Putte, L. B. Development 94. Gaujoux-Viala, C. et al. Suboptimal management of rheumatoid arthritis in France:
of a disease activity score based on judgment in clinical practice by rheumatologists. a real-world study based on data from the French National Health Data System.
J. Rheumatol. 20, 579–581 (1993). RMD Open 9, e003075 (2023).
67. Prevoo, M. L. et al. Modified disease activity scores that include twenty-eight-joint 95. Olsen, I. C. et al. Assessments of the unmet need in the management of patients with
counts. Development and validation in a prospective longitudinal study of patients with rheumatoid arthritis: analyses from the NOR-DMARD registry. Rheumatology 58, 481–491
rheumatoid arthritis. Arthritis Rheum. 38, 44–48 (1995). (2019).
68. Grigor, C. et al. Effect of a treatment strategy of tight control for rheumatoid arthritis 96. Brinkmann, G. H. et al. Treat to target strategy in early rheumatoid arthritis versus
(the TICORA study): a single-blind randomised controlled trial. Lancet 364, 263–269 (2004). routine care – a comparative clinical practice study. Semin. Arthritis Rheum. 48, 808–814
69. Kavanaugh, A. et al. Clinical, functional and radiographic consequences of achieving (2019).
stable low disease activity and remission with adalimumab plus methotrexate or 97. Gärtner, M. et al. Persistence of subclinical sonographic joint activity in rheumatoid
methotrexate alone in early rheumatoid arthritis: 26-week results from the randomised, arthritis in sustained clinical remission. Ann. Rheum. Dis. 74, 2050–2053 (2015).
controlled OPTIMA study. Ann. Rheum. Dis. 72, 64–71 (2013). 98. Pareto, V. Cours d’Économie Politique Nouvelle édition par Bousquet, G.-H. et Busino, G.
70. Keystone, E. C., Kavanaugh, A., Weinblatt, M. E., Patra, K. & Pangan, A. L. Clinical 313–315 (Droz, 1964).
consequences of delayed addition of adalimumab to methotrexate therapy over 5 years 99. Goekoop-Ruiterman, Y. P. M. et al. Clinical and radiographic outcomes of four different
in patients with rheumatoid arthritis. J. Rheumatol. 38, 855–862 (2011). treatment strategies in patients with early rheumatoid arthritis (the BeSt study):
71. Klarenbeek, N. B. et al. Association with joint damage and physical functioning of nine a randomized, controlled trial. Arthritis Rheum. 58, S126–S135 (2008).
composite indices and the 2011 ACR/EULAR remission criteria in rheumatoid arthritis. 100. Heimans, L. et al. A two-step treatment strategy trial in patients with early arthritis aimed
Ann. Rheum. Dis. 70, 1815–1821 (2011). at achieving remission: the IMPROVED study. Ann. Rheum. Dis. 73, 1356–1361 (2014).
72. Schipper, L. G. et al. Meta-analysis of tight control strategies in rheumatoid arthritis: 101. Bathon, J. M. & Genovese, M. C. The Early Rheumatoid Arthritis (ERA) trial comparing the
protocolized treatment has additional value with respect to the clinical outcome. efficacy and safety of etanercept and methotrexate. Clin. Exp. Rheumatol. 21, S195–S197
Rheumatology 49, 2154–2164 (2010). (2003).
73. Aletaha, D. et al. Rheumatoid arthritis joint progression in sustained remission is 102. Lee, E. B. et al. Tofacitinib versus methotrexate in rheumatoid arthritis. N. Engl. J. Med.
determined by disease activity levels preceding the period of radiographic assessment. 370, 2377–2386 (2014).
Arthritis Rheum. 60, 1242–1249 (2009). 103. Roodenrijs, N. M. T. et al. Pharmacological and non-pharmacological therapeutic
74. van Leeuwen, M. A. et al. Interrelationship of outcome measures and process variables in strategies in difficult-to-treat rheumatoid arthritis: a systematic literature review
early rheumatoid arthritis. A comparison of radiologic damage, physical disability, joint informing the EULAR recommendations for the management of difficult-to-treat
counts, and acute phase reactants. J. Rheumatol. 21, 425–429 (1994). rheumatoid arthritis. RMD Open 7, e001512 (2021).
75. Gul, H. L., Ferreira, J. F. & Emery, P. Remission in rheumatoid arthritis: is it all the same? 104. Busby, A. D. et al. The role of comorbidities alongside patient and disease characteristics
Expert. Rev. Clin. Pharmacol. 8, 575–586 (2015). in long-term disease activity in RA using UK inception cohort data. Rheumatology 61,
76. Mankia, K., Gul, H. & Emery, P. Treating rheumatoid arthritis to an imaging target 4297–4304 (2022).
produces better outcomes, or does it? Rheumatology 60, 3–4 (2021). 105. Bertsias, A. et al. Patterns of comorbidities differentially affect long-term functional
77. Haavardsholm, E. A. et al. Ultrasound in management of rheumatoid arthritis: ARCTIC evolution and disease activity in patients with “difficult to treat” rheumatoid arthritis.
randomised controlled strategy trial. BMJ 354, i4205 (2016). RMD Open 10, e003808 (2024).
78. Dale, J. et al. Targeting ultrasound remission in early rheumatoid arthritis: the results of 106. Kjørholt, K. E. et al. Effects of tapering conventional synthetic disease-modifying
the TaSER study, a randomised clinical trial. Ann. Rheum. Dis. 75, 1043–1050 (2016). antirheumatic drugs to drug-free remission versus stable treatment in rheumatoid
79. Møller-Bisgaard, S. et al. Effect of magnetic resonance imaging vs conventional arthritis (ARCTIC REWIND): 3-year results from an open-label, randomised controlled,
treat-to-target strategies on disease activity remission and radiographic progression in non-inferiority trial. Lancet Rheumatol. 6, e268–e278 (2024).
rheumatoid arthritis: the IMAGINE-RA randomized clinical trial. JAMA 321, 461–472 (2019). 107. Tascilar, K. et al. Treatment tapering and stopping in patients with rheumatoid arthritis
80. Smolen, J. S. et al. Treating rheumatoid arthritis to target: recommendations of an in stable remission (RETRO): a multicentre, randomised, controlled, open-label, phase 3
international task force. Ann. Rheum. Dis. 69, 631–637 (2010). trial. Lancet Rheumatol. 3, e767–e777 (2021).
81. Buch, M. H. Defining refractory rheumatoid arthritis. Ann. Rheum. Dis. 77, 966–969 (2018). 108. Smolen, J. S. et al. EULAR recommendations for the management of rheumatoid arthritis
82. Gossec, L., Dougados, M. & Dixon, W. Patient-reported outcomes as end points in clinical with synthetic and biological disease-modifying antirheumatic drugs: 2022 update.
trials in rheumatoid arthritis. RMD Open 1, e000019 (2015). Ann. Rheum. Dis. 82, 3–18 (2023).
83. Studenic, P. et al. Testing different thresholds for patient global assessment in defining 109. Gerriets, V., Goyal, A. & Khaddour, K. Tumor necrosis factor inhibitors (StatPearls, 2024).
remission for rheumatoid arthritis: are the current ACR/EULAR Boolean criteria optimal? 110. Buch, M. H. et al. Updated consensus statement on the use of rituximab in patients with
Ann. Rheum. Dis. 79, 445–452 (2020). rheumatoid arthritis. Ann. Rheum. Dis. 70, 909–920 (2011).
84. Studenic, P. et al. American College of Rheumatology/EULAR remission criteria for 111. Burmester, G. R. et al. Safety profile of upadacitinib over 15 000 patient-years across
rheumatoid arthritis: 2022 revision. Ann. Rheum. Dis. 82, 74–80 (2023). rheumatoid arthritis, psoriatic arthritis, ankylosing spondylitis and atopic dermatitis.
85. Studenic, P., Radner, H., Smolen, J. S. & Aletaha, D. Discrepancies between patients and RMD Open 9, e002735 (2023).
physicians in their perceptions of rheumatoid arthritis disease activity. Arthritis Rheum. 112. Winthrop, K. L. The emerging safety profile of JAK inhibitors in rheumatic disease.
64, 2814–2823 (2012). Nat. Rev. Rheumatol. 13, 234–243 (2017).

Nature Reviews Rheumatology

Review article

113. Montastruc, F. et al. Abatacept initiation in rheumatoid arthritis and the risk of cancer: Acknowledgements
a population-based comparative cohort study. Rheumatology 58, 683–691 (2019). The authors thank A. Mazzucato-Puchner and S. Hayer for help with preparation of the slides
114. Simon, T. A. et al. Comparative risk of malignancies and infections in patients with in Fig. 3, and also acknowledge J. Smolen and P. Mandl for their insightful considerations and
rheumatoid arthritis initiating abatacept versus other biologics: a multi-database comments on this work.
real-world study. Arthritis Res. Ther. 21, 228 (2019).
115. Kastrati, K. et al. A systematic literature review informing the consensus statement on efficacy Author contributions
and safety of pharmacological treatment with interleukin-6 pathway inhibition with biological Both authors researched the data for the article, provided substantial contributions to
DMARDs in immune-mediated inflammatory diseases. RMD Open 8, e002359 (2022). discussions of its content, wrote the article and reviewed and/or edited the manuscript before
116. Rempenault, C. et al. Risk of diverticulitis and gastrointestinal perforation in rheumatoid submission.
arthritis treated with tocilizumab compared to rituximab or abatacept. Rheumatology 61,
953–962 (2022). Competing interests
117. Smolen, J. S. et al. Safety profile of baricitinib in patients with active rheumatoid arthritis V.K. declares that she has received speaker’s honoraria and travel support from Astra Zeneca,
with over 2 years median time in treatment. J. Rheumatol. 46, 7–18 (2019). Astro Pharma, Eli Lilly, Galapagos, Novartis and Sandoz. D.A. declares that he has received
118. Hoisnard, L. et al. Adverse events associated with JAK inhibitors in 126,815 reports from grants from AbbVie, Amgen, Eli Lilly, Galapagos and Sanofi, and honoraria for speaker’s
the WHO pharmacovigilance database. Sci. Rep. 12, 7140 (2022). bureau and/or consultancy from AbbVie, Amgen, Galapagos, Eli Lilly, Janssen, Merck,
119. Aringer, M. & Smolen, J. S. TNF inhibition in SLE: where do we stand? Lupus 18, 5–8 (2009). Novartis, Pfizer and Sandoz.
120. Charles, P. J., Smeenk, R. J., De Jong, J., Feldmann, M. & Maini, R. N. Assessment of antibodies
to double-stranded DNA induced in rheumatoid arthritis patients following treatment with Additional information
infliximab, a monoclonal antibody to tumor necrosis factor α: findings in open-label and Peer review information Nature Reviews Rheumatology thanks Elena Nikiphorou and the
randomized placebo-controlled trials. Arthritis Rheum. 43, 2383–2390 (2000). other, anonymous, reviewer(s) for their contribution to the peer review of this work.
121. Kaltsonoudis, E., Voulgari, P. V., Konitsiotis, S. & Drosos, A. A. Demyelination and other
neurological adverse events after anti-TNF therapy. Autoimmun. Rev. 13, 54–58 (2014). Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
122. Winkelmann, A., Patejdl, R., Wagner, S., Benecke, R. & Zettl, U. K. Cerebral MRI lesions published maps and institutional affiliations.
and anti-tumor necrosis factor-alpha therapy. J. Neurol. 255, 109–114 (2008).
123. Buch, M. H., Johnsen, A. & Schiff, M. Can switching to abatacept therapy in patients Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this
with rheumatoid arthritis on background methotrexate reverse TNF-inhibitor-induced article under a publishing agreement with the author(s) or other rightsholder(s); author
antinuclear autoantibody/double-stranded DNA autoantibody conversion? An analysis self-archiving of the accepted manuscript version of this article is solely governed by the
of the AMPLE and ATTEST trials. Clin. Exp. Rheumatol. 37, 127–132 (2019). terms of such publishing agreement and applicable law.
124. Bonelli, M. M., Mrak, D., Perkmann, T., Haslacher, H. & Aletaha, D. SARS-CoV-2 vaccination
in rituximab-treated patients: evidence for impaired humoral but inducible cellular © Springer Nature Limited 2024
immune response. Ann. Rheum. Dis. 80, 1355–1356 (2021).
125. Mrak, D. et al. SARS-CoV-2 vaccination in rituximab-treated patients: B cells promote
humoral immune responses in the presence of T-cell-mediated immunity. Ann. Rheum.
Dis. 80, 1345–1350 (2021).

Nature Reviews Rheumatology