Rheumatoid Arthritis (RA) is a chronic inflammatory disease of unknown etiology that preferentially affects joints in a symmetric manner. The course of the disease is variable because it leads to functional compromise from the onset, progressing over time along with joint destruction and deformity, which may lead to severe disability in a large percentage of affected personas, work loss and even shortened survival.

Prognostic factors (PF) are sociodemographic, clinical, analytical and/or radiological data present at the beginning of the disease that provide prospective information of the patients’ progress. This information is useful in order to guide therapeutic decision. The importance of PF is settled mainly in three aspects:

• –

  Classification: allows the stratification of patients into homogeneous groups.

• –

  Therapeutic: facilitate therapeutic choices for each patient, as well as the comparison of these options between each group of patients with different prognostic characteristics.

• –

  Prevention: the knowledge of PF allows us to initiate specific preventive actions.

We can classify PF into two groups: those which are modifiable (erythrocyte sedimentation rate [ESR], C reactive protein [CRP], DAS28, HAQ, and treatment), and non modifiable (gender, age, rheumatoid factor [RF], anti-CCP, and shared epitope).

One can talk about PF in relation to different aspects:

• –

  Functional prognosis.

• –

  Radiologic progression.

• –

  Disease remission.

• –

  Mortality.

Functional prognosis of disease refers to the degree of disability developed by a patient in the long term. The possibility that a patient develops severe disability reaches 33% in studies performed before the availability of anti-TNF,1 and said disability is reflected on the patients capacity to work, which may be reduced in 50% at 10 years since the onset of disease.2

PF associated with a greater disability and identified in several studies are age (OR=1058 [1017–1101]),3 positive RF (OR=3772 [1204–11813]),3 elevated baseline DAS28 (OR=2)4 and baseline HAQ>1 (OR=4023 [1373–11783]).3

Radiographic remission is defined as the lack of progression of structural damage. Irreversible structural lesions appear from the onset of the diseases. Many of the patients attain clinical remission according to current remission criteria, but in spite of a strict control of the disease or the minimum of joint clinical manifestations and normalization of the acute phase reactants, a proportion of patients present progression of the structural damage, joint deformity and reduction in quality of life. This radiologic progression may be explained by maintained subclinical inflammation of the bone and cartilage during the course of disease.5 In a review of the validity of remissions predictive value6 there was a relationship between remission of the disease and structural damage and long-term disability, concluding that the patients who reach clinical remission according to current criteria have a tendency to show less functional impairment and slower radiographic progression.

PF predicting radiologic progression as identified in different studies are: female gender (OR=3.3 [1.3–7.6])7 (OR=5.5 [1.1–28.2])8; baseline ESR (OR=3.2 [1.2–7.6])7; baseline CRP;9 positive RF (OR=3.1 [1.2–7.6])7; baseline anti-CCP titers (OR=4.03 [1.65–9.82])5 (low OR=2.6 [0.9–7.2], elevated OR=9.9 [2.7–36.7])7 (OR=3.6 [0.9–14.5])8; bone edema seen on magnetic resonance (MR) (OR=1.44 [0.95–2.20])5; sharp score (OR=1.12 [1.03–1.21]); shared epitope (OR=2.0 [1.8–2.2])10 (OR=3.1 [1.1–9]).8

Disease remission is generally a synonym of minimal clinical affection, absence of synovitis and normal acute phase reactants. If disease remission is achieved, it will be more likely that the degree of long-term disability of the patient will be minimized.

A recent systematic review11 outlining the variables that act as predictive factors of disease remission has been published. The magnitude of association of each of them is variable in relation to the design and number of included patients in the analyzed studies, as well as the variables used to adjust each model. Factors identified can be grouped into three areas: sociodemographic, disease associated and treatment associated. The factors most commonly associated with disease remission are rheumatoid factor, disease activity as quantified by DAS28, functional status (HAQ) and early onset of treatment.

• 1.

  Sociodemographic factors:

  • –

    Gender: among the studies evaluating the effect of gender on disease remission, 5 of 11 studies, among them TEMPO12 and ReAct,13 conclude that male gender is an independent predictive factor of disease remission in a maintained manner. The rest of the evaluated studies did not show gender as a remission-predicting factor.

  • –

    Age and age at onset of disease: it has been observed that age acts as a significant predictor of disease remission in an inverse manner, in 2 cohorts of patients treated with anti-TNF.

    • 1.

      GISEA14 trial: patients treated with anti-TNF over 53 years of age have less probability of achieving remission after adjusting for gender, RF and baseline disease activity (OR=0.64 [0.4–0.9]).

    • 2.

      ReAct trial13: patients treated with adalimumab under 40 years of age have a higher tendency to achieve remission after 3 years of follow up versus those older than 40 (HR=0.61–0.87).

    • 3.

      FIN-RACo15 trial: did not confirm age at onset of disease as an independent prediction factor for remission.

  The study by Pease et al.16 concludes that onset of disease in persons over 65 acts as an independent remission factor in patients treated with DMARD (OR=2.99 [1.8–5]).

  After gender and age, one may deduce that female gender and advanced age have less chance to achieve remission. These data must be used relatively because of the limited the parameters used to measure the degree of disease activity and remission criteria have in these populations.

  • –

    Genetic markers: their use is restricted to clinical trials. It has been shown that the presence of shared epitope, both specific predisposing alleles HLA-DQB1/HLA-DQA,1 and the protective HLA-DRB1 allele are not associated with remission in RA when adjusted for RF and the use of DMARD.17

  • –

    Smoking: The results obtained in two studies are contradictory and the effect of tobacco on disease activity must be confirmed by future research.13,18

  • –

    Comorbidity:

    • 1.

      ReAct trial13: the presence of more than one comorbidity is related to a lessened probability of achieving clinical remission (HR=0.85 [0.78–0.93]).

    • 2.

      The study by Hyrich et al. did not show a significant association between the presence of comorbidity and disease remission in patients treated with ETN and IFX.

• 2.

  Disease dependent factors:

  • –

    Disease activity: most of the studies showed that the degree of disease activity quantified by DAS28 is inversely related to disease remission.13,19–21

  • –

    Functional status (HAQ): numerous studies on cohorts of patients treated with DMARD or anti-TNF have shown that the functional status as quantified by the baseline HAQ behaves as an independent predictor of disease remission in all models in an inverse manner.13,21 This association has not been documented in early onset RA.19

  Occasionally, disease activity measures or remission criteria may not truly reflect the degree of disease because they take into account the patient's perception of pain or the global disease activity evaluation. For example, it has been shown that women with RA have a tendency to evaluate in a more severe way than the disease with respect to men and these data may reflect less precision on the evaluation of disease activity by this specific population.

  • –

    Duration of disease: patients with longer diseases have less chances of achieving persistent clinical remission (OR=0.87–0.91; P≤.004).21 In other cohort studies using anti-TNF it has been impossible to determine if the time since the onset of disease is a predictor of remission.13,22

  • –

    Rheumatoid factor: most of the studies have shown that RF is inversely related to disease remission. However, the predictive value of baseline RF disappears when adjusted for anti-CCP titers, the treatment strategy employed (combination DMARD or anti-TNF) and the presence of shared epitope.21

  • –

    Anti-CCP: Baseline anti-CCP titers have been inversely related with the probability of remission at 24 months since onset (OR=0.6 [0.5–0.9]),21 adjusted for DAS,28 duration of disease, HAQ and male gender.

  • –

    Acute phase reactant plasma levels: patients with a baseline CRP plasma determination equal or over 20mg/l have a reduced probability of achieving disease remission (HR=0.8 [0.8–0.9]).13

  • –

    Radiologic affection: a Sharp score under 4 behaves as an independent remission factor when adjusting for other variables (DAS, morning stiffness, HAQ<1.25, Ritchie score) (OR=1.99 [0.98–4.0]).19

• 3.

  Treatment dependent factors: numerous published studies show that patients receiving early treatment with DMARD, anti-TNF or combinations of DMARD and anti-TNF have a greater probability of achieving disease remission.13,15,23 On the other hand, the number of DMARDs employed prior to anti-TNF is inversely related to the probability of disease remission.23,24

Lastly, patients in which the start of treatment is delayed for more than 4 months since the onset of disease have a reduced chance of attaining clinical remission.25

While the general population mortality rate has substantially decreased during the past 4–5 decades, this improvement in survival has not been shown to occur in patients with RA, with survival remaining constant.26 One of the main causes of mortality in patients with RA is that of cardiovascular origin, but classic cardiovascular risk factors by themselves do not justify this increase in RA patient mortality with respect to the general population. However, disease inflammatory activity does play an important role in it.

In studies prior to the use of anti-TNF and in current disease incidence cohorts it has been demonstrated that there is a difference between life expectancy between the general population and patients with RA which has increased in recent decades.26,27

Studies evaluating the influence of RF on survival of patients with RA observed an inversely proportional relationship in patients with positive RF, while those negative to RF had a mortality on par with the general population. The increase in the mortality rate between patients with RA and the general population is confirmed for patients with RA and positive RF.26 The other mortality associated PF identified in RA are26:

• –

  Age (HR=1.1 [1.09–1.12]).

• –

  Male gender (OR=1.90 [1.43–2.52]).

• –

  Elevated HAQ scores maintained throughout the progression of the disease (OR=1.46 [1.19–1.79]).

• –

  Comorbidities (OR=1.83 [1.38–2.42]).

• –

  Low schooling levels: lack of secondary schooling is associated with a reduction of over 50% in the functional status or 9 year mortality rates (OR=7.5).28

• –

  Depression: patients with depression have higher mortality (HR=2.2 [1.2–3.9]).29

We have identified predictors of disease among which are age, rheumatoid factor, the degree of disease activity (DAS28), functional status (HAQ) and early treatment. These prognostic factors present at the onset of the disease help us to identify patients most likely to present a more aggressive course of RA. In these patients combination therapy with DMARDs and anti-TNF at the onset of disease30 may be indicated to achieve as low an inflammatory activity as possible, maintaining it during activity and minimizing morbidity and mortality attributable to RA. However, more studies are needed to establish long-term benefits of aggressive treatment strategies.