 |
The current American College of Rheumatology (ACR) guidelines for the treatment of RA begin with establishment of diagnosis,
determination of disease activity, and development of a prognosis. Therapy is initiated with patient education and physical
therapy, disease-modifying antirheumatic drug(s) (DMARD), and perhaps corticosteroids and/or a nonsteroidal anti-inflammatory
drug (NSAID).
If the response to treatment is not adequate after 3 months, therapy should be altered by changing or adding DMARDs,
either as monotherapy or in combination. Biologic DMARDs should also be considered either alone or as add-on therapy.
Patients who still do not respond adequately despite being on several DMARD regimens are considered multiple DMARD failures.
These patients will continue to have signs and symptoms and progression of structural joint damage.
The relationship between the development of radiographic joint destruction in RA and its long-term consequences for the
patient is not well understood, but one view of the disease process is that inflammatory joint symptoms are the main determinant
of disability early in the disease, while joint destruction dominates in late disease.
Presentations and discussions at OMERACT IV (the 4th International Consensus Conference on Outcome Measures in Rheumatoid
Arthritis Clinical Trials) are consistent with this view, suggesting a high correlation between inflammation and disability
in early RA, but that the strength of this relationship declines over time. Discussions at OMERACT IV also indicated that
fluctuations in disability will become less pronounced in later-stage RA and become more closely correlated with radiographic
evidence of joint damage.
Progressive joint damage occurs in unchecked RA. In the early stages of the disease, cartilage destruction begins, and
neutrophils, T cells, and B cells are recruited into the synovial cavity. The synovial membrane exhibits hyperplasia, and
hypertrophic synoviocytes and capillaries are starting to form. In a joint with established RA,the greatly thickened, inflamed
synovium (pannus) invades and erodes adjacent bone and cartilage. Activated macrophages, lymphocytes, and fibroblasts and
their products stimulate extensive angiogenesis, a central feature in synovial inflammation and pannus formation. Synovial
villi become evident.
TNF-alpha is one of the most potent osteoclastogenic cytokines produced in inflammation, and it is pivotal in the pathogenesis
of RA. Production of TNF-alpha and other proinflammatory cytokines in RA is largely T-cell dependent. Activated synovial T
cells express both membrane-bound and soluble forms of RANKL. In the RA synovium, fibroblasts also provide an abundant source
of RANKL and macrophage colony-stimulating factor (M-CSF).
TNF-a and IL-1 target stromal-osteoblastic cells to increase expression of RANKL. In the presence of permissive levels
of RANKL, TNF-alpha acts directly to stimulate osteoclast differentiation of macrophages and myeloid progenitor cells.
In addition, TNF-alpha induces IL-1 release by synovial fibroblasts and macrophages, and IL-1, together with RANKL, is a major
survival and activation signal for nascent osteoclasts. Thus, TNF-alpha and IL-1, acting in concert with RANKL, can powerfully
promote osteoclast recruitment, activation, and osteolysis in RA.
Cytokines exert their damaging effects by binding to specific receptors, and there are several potential approaches that
can be employed to block these effects. Cytokines can be neutralized through the use of antibodies or soluble receptors. With
this approach, the cytokine never reaches the receptor on the cell of interest. This avenue for the treatment of RA has been
taken with soluble TNF-alpha receptor fusion proteins, soluble IL receptors, monoclonal antibodies against TNF-alpha, and
monoclonal antibodies against IL-6.
Receptor antagonists or antibodies can bind to cytokine receptors on cells and prevent cytokines from binding. This blocks
their actions on the cell in question. This approach to the treatment of RA has been taken with recombinant IL-1Ra and an
antibody against the IL-6 receptor. Administration of anti-inflammatory cytokines can inhibit expression of inflammatory cytokines.
This approach has been taken with IL-4 and IL-10.
Cytokines have a very wide range of functions relevant to the pathophysiology of RA. Functions relevant to RA; Regulation
of the inflammatory respomse; a)Proimflammatory: TNF-alpha,IL-1,IL-6.IL-7,chemokines (Il-8,MIP1 alpha/beta,Rantes) b) Anti-inflammatory:
IL-10,TGF-beta,IL-4,IL-1Ra,sTNFR I/II,IL-1R II
Modulation of the immune response; I) Modify Th1/Th2 bias ;a) Th1:IL-12,IL-18,IFN-y ;b) Th2:IL-4,Il-13,IL-10,chemokines
II) Mediate activation/apoptosis: IL-2,IL-15,IFN-y,IL-3,IL-5,IL-7,
Remodel tissue; I) Bone and cartilage destruction: IL-1,TNF-alpha,OPG/RANKL (TRANCE) ;II) Angiogenesis/growth factors:
TNF-alpha,TGF-beta,VEGF
These molecules are key regulators of inflammatory responses, having actions that are either proinflammatory (TNF-alpha,
IL -1, IL-6, IL-7, chemokines, IL-8, MIP-1alpha/beta, and regulated upon activation, normal T cell expressed and secreted
[RANTES]) or anti-inflammatory (IL-10, transforming growth factor [TGF]-beta, IL-4; IL-1 receptor antagonist [Ra], soluble
TNF receptor [sTNFR] type I/II, and type II IL-1 receptor [IL-1R II]).
Cytokines can also modulate immune responses. IL-12, IL-18, and interferon (IFN)-y promote a Th1 bias and IL-4, IL-13,
IL-10, and chemokines promote a Th2 bias. Activation and/or apoptosis are mediated by IL-2, IL-15, IFN-y, IL-3, IL-5, and
IL-7. Cytokines are also involved in regulation of tissue remodeling. IL-1, TNF-alpha, and OPG (osteoprotegerin) /RANKL (receptor
activator of nuclear factor-kappa b ligand) (TRANCE, TNF-related activator-induced cytokine) have been implicated in bone
and cartilage destruction, and TNF-alpha, TNF-beta, and vascular endothelial growth factor (VEGF) have been shown to have
trophic/angiogenic effects.
CD4+ T cells might initiate the disease process in RA. Activated by antigens, these cells stimulate monocytes, synovial
fibroblasts, and macrophages to produce the key proinflammatory cytokines TNF-alpha, IL-1, and IL-6 as well as to release
MMPs, enzymes that degrade connective tissue matrix.1 TNF-alpha and IL-1 also inhibit synovial fibroblasts from producing
tissue inhibitors of MMPs. These two actions by TNF-alpha and IL-1, among others, are believed to result in the joint damage
that occurs in RA. Activated CD4+ T cells also contribute to joint damage by stimulating the development of osteoclasts by
expressing osteoprotegrin ligands (OPGLs) and by stimulating B cells to produce immunoglobulins such as rheumatoid factor.
Activated macrophages, lymphocytes, and fibroblasts and their products can stimulate angiogenesis, a fact that may account
for the increased vascularity of the synovium in rheumatoid joints. Other cytokines involved in the complex cellular interactions
that occur as part of the inflammatory process include IL-4, IL-10, IL-12, and IFN-g. In addition, CD11 and CD69 cells are
involved in the cell-surface signaling that leads to the production of cytokines.
Research investigators analyzed the direct medical care costs for RA over 1- and 11-year periods, and assessed the impact
of poor function and functional decline on these expenses. Their analysis used data from the University of California, San
Francisco, RA Panel Study in which 1,156 patients were followed for up to 15 years.
Medical care costs for RA for 1995 to 1996 averaged $5,919 per patient. Hospital admissions contributed 51.7% of this
total and drugs contributed 26.1%. The respective contributions for physician visits, allied health professional visits, testing,
and other expenses were 8.6%, 1.4%, 4.7%, and 7.5%.
In an era of increased scrutiny of both costs and quality, health outcomes research has become a focus of attention in
many areas of medicine. In an effort to improve healthcare outcomes, it is first necessary to define and measure "quality."
Although both the definition and measurement of quality remain a source of continued debate, the Institute of Medicine defines
quality as the "degree to which health services for individuals and populations increase the likelihood of desired health
outcomes and are consistent with current professional knowledge." Lack of quality can be due to overuse, underuse, or misuse
of resources. Although this definition is a useful starting point for discussion, it does not incorporate the additional important
dimension of patient values or preferences.
In 1966, Donabedian first proposed a conceptual framework that has remained useful in evaluating the quality of healthcare.
Using this theoretical construct, factors that contribute to quality are separated into 3 groups: those affecting structure,
process, or outcome. Structural measures pertain to the capacities, technologies, and infrastructures that make up the organization
of a healthcare system. Process is defined as what is done for or to the patient and may be administrative, clinical, or financial
(eg, periodic laboratory monitoring for patients receiving disease-modifying antirheumatic drug [DMARD] therapy). Optimal
process measures can often be associated with positive outcomes by examining results from clinical trials.
Patient outcomes are the end point of greatest interest and include mortality, morbidity, and functional status. Outcomes
may be relatively easy to measure in carefully controlled clinical trials focused on "efficacy." However, it is much more
difficult to measure outcomes in real-world "effectiveness" studies involving populations or health systems or within a physician's
office.
Moreover, due to varying demographics and burdens of disease among patient populations, directly measuring outcomes may
be costly, time-consuming, and not sensitive enough in detecting differences in the quality of healthcare provided.
In contrast to outcomes, process of care is more easily evaluated, requires shorter end points, offers higher sensitivity
in detecting differences among treatments provided, and usually requires no risk adjustment.
Implicit in the evaluation of the structure of healthcare is that patients must have good access to care. Lacaille and
colleagues studied 29,297 individuals with rheumatoid arthritis (RA) in British Columbia, Canada, from 1996 to 2000. RA was
defined by at least 2 physician visits at least 2 months apart and no other connective tissue disease diagnosis. These investigators
found that only 32% of patients had seen a rheumatologist in this 5-year period (range, 9% to 38%, depending on province),
whereas the majority (53%) were followed up only by a general practitioner. Of those seen by a rheumatologist, 80% had received
DMARDs compared with only 14% of patients seen by a family physician.
Presenting the results of a recent North American RA study, Sokka and Pincus reported on a cohort of 296 RA patients,
pointing out that most of them (> 60%) had not been seen by a rheumatologist until at least 4 months after the onset of
their symptoms. Those with longer disease duration before evaluation and treatment were more likely to be seropositive for
rheumatoid factor and to have erosive disease as well as worse pain and health status.
In a third investigation, Kwoh and colleagues studied 944 RA or inflammatory polyarthritis patients in 2001 and found
that only half of them had ever seen a rheumatologist, and, consistent with the British Columbia study, DMARD use varied greatly
between those having seen a rheumatologist (64%) and those not having seen a rheumatologist (27%). Patients treated by rheumatologists
were 50% more likely to be prescribed selective cyclooxygenase 2 inhibitors (coxibs), and 25% were less likely to be treated
with narcotics.
In addition to barriers to seeing arthritis specialists in a timely fashion, barriers to receiving newer, more effective,
or safer drugs may adversely affect health outcomes. One of the most exciting developments in rheumatology is the addition
of biologic agents, including tumor necrosis factor (TNF) inhibitors, to the armamentarium of agents used to treat active
inflammatory disease.
Another newer class of drugs, coxibs, has made nonsteroidal anti-inflammatory drug (NSAID) therapy safer and, thus, more
rational for high-risk patients by reducing gastrointestinal morbidity. However, the expense of drugs in both these classes
is much greater than their older counterparts, and administrative barriers to receiving these agents imposed by health plans
may negate their expected health benefit in high-risk patients.
Yelin analyzed data obtained from 493 patients and found that patients in health maintenance organizations (HMOs) were
only half as likely to receive TNF-blocking agents and 40% less likely to receive coxibs compared with those in fee-for-service
plans. Since prior work has shown a negligible difference between healthcare utilization rates of other resources (eg, outpatient
visits) among different types of health plans, it appears that controlling costs by restricting access to newer medications
is a cost-saving strategy that may be used by some HMOs.
Briesacher and colleagues studied 38,072 RA and osteoarthritis patients in 2000 and found that coxib use declined dramatically
as copayments increased. Of patients in a 1-tier plan (average copayment, $8), 63% received coxibs compared with only 9% of
patients in 3-tier plans (average copayment, $72).
Another important health service research question is whether the use of both traditional and biologic DMARDs is improving
long-term health outcomes. Large, population-based studies and registries provide important effectiveness data to help answer
this question.
Bombardier and colleagues reported on the Study of New Onset Rheumatoid Arthritis (SONORA) cohort of 1000 North American
and Canadian patients with early RA and found that 79% of RA patients enrolled > 3 months after diagnosis had received
DMARDs. Those patients who had not received DMARD therapy (35% in the overall cohort) had more active disease (as measured
by the Health Assessment Questionnaire, global pain, and physician- and patient-determined disease assessments) and higher
tender and swollen joint counts. These findings are timely given the emerging evidence that early evaluation by a rheumatologist
and institution of early and aggressive DMARD therapy, more commonly implemented by rheumatologists than generalists, may
retard radiographic disease progression and delay adverse outcomes in some patients.
Despite the increasing clinical trial evidence about the efficacy of TNF-blocking agents, long-term health outcome effectiveness
data have been lacking. Kvien and colleagues studied almost 1000 RA patients in Oslo, Norway, served by 1 hospital and 1 outpatient
facility from 1994-2001. The use and type of resources provided to this population did not significantly differ in any way
over time except for increased access to new medications, including coxibs and TNF-blocking agents. Physical health status,
pain, and global scores were improved from 1996-2001 and were attributed to the access to these new agents. Incorporating
data on the incremental costs associated with use of these drugs during this period, these investigators found that cost per
quality-adjusted life-year gained was US $29,000, a value comparable to that of other commonly advocated medical therapies.
Although identifying and ultimately ameliorating disparities in access to specialist care and to new therapies is a continuing
and important issue in outcomes research, another area addresses inefficiencies owed to individual physicians. Given the national
shortage of rheumatologists, methods to minimize waiting periods for patient access to arthritis specialists are particularly
warranted.
Newman and colleagues reported on their efforts to "do today's work today" and convert a typical office-based rheumatologist
practice with excessive waiting times for a new consultation to an "open access model," where between 6 and 8 hours of unscheduled
clinic visits were available each day for either same-day new patient or follow-up visits. New patients were able to see a
rheumatologist within days of referral. Follow-up patients were given a return appointment date only if they needed to be
seen within 3 months, but if their expected follow-up was > 3 months, they were told which week to make their appointment;
they would call the day or week of the desired appointment and would be seen in the next 72 hours. Internet-based algorithms
were created and made accessible to referring primary care physicians for common problems (eg, knee osteoarthritis)
that enabled the primary care physicians to manage routine complaints more effectively and to guide the diagnostic studies
obtained before the referral.
These algorithms included management strategies, Web-based links, and a series of predefined order sets that could be
customized for an individual patient, allowing the physician to choose patient education resources, assisting devices, physical
therapy, medications, and radiographs. Newman's group was able to reduce the third available appointment from 60 days to fewer
than 2 days. Cancellations decreased from 40% to less than 15%, and both physician and patient satisfaction rose. Net revenues
increased, and both total referrals and referrals of patients with active inflammatory disease (largely RA) increased by approximately
50% compared with a similar period historically.
The structure of healthcare provided and an adequate access to that structure are integral in achieving improved equity
and efficiency. The above research provides valuable information at a population, health plan, and office level about overcoming
the barriers to early evaluation and treatment of patients with active inflammatory disease. Other research data presented
at ACR this year showed that in the real world aggressive arthritis treatment can lead to improved outcomes at a cost that
is acceptable.
It is difficult to define treatment algorithms and clinical practice guidelines for disorders of unknown pathogenesis
and, often, unclear diagnostic criteria. Nonetheless, different expert groups periodically assemble consensus panels to try
and define appropriate care criteria. As an example, the ACR issued guidelines for RA in 2002. These guidelines recommend
initial therapy with a traditional DMARD for at least 3 months at maximal dosing (methotrexate, 25 mg/wk) before prescribing
an anti-TNF agent.
In an effort to define appropriate and inappropriate use of anti-TNF therapy, Berger and colleagues reviewed the results
obtained in 4124 RA patients receiving TNF-blocking therapy (infliximab or etanercept) covered by Caremark during a recent
1-year period to determine whether the patients had been given DMARD therapy in the period preceding administration of anti-TNF
agents. They found that only 22% of patients had received a traditional DMARD before biologic therapy, and only 17% had been
treated according to the ACR recommendations for the use of biologic agents, as outlined above.
Although no set of recommendations can or should be interpreted as the standard of care for any individual patient,
the use of biologics as first-line drugs in approximately 80% of patients may represent overzealous use, given available data
and their expense.
For the first time, the ACR devoted an entire oral symposium to measuring the quality of care in rheumatic disease. A
major theme of this session was the development of quality indicators (QIs). QIs are process-of-care measures that identify
a discrete population and define a health service that should be performed for that population. They differ from practice
guidelines in that they are very specific and usually constitute a minimal standard of care for all patients in a well-defined
population.
In contrast, practice guidelines tend to be broader, often take the form of "should be considered," and are not necessarily
applicable to an individual patient. Moreover, QIs can be developed for patients who might be traditionally prohibited from
participation in clinical trials due to strict exclusion criteria, such as age or comorbidities, making QIs useful in a real-world
practice.
Using a process outlined previously, Mikuls and colleagues defined minimal standards of care for gout management. These
researchers developed QIs using a systemic literature review and refined them further by convening 2 sequential panels of
content experts. Their ongoing work is characterizing factors associated with adherence to these QIs.
Pencharz and associates used previously defined osteoarthritis-specific QIs developed by the Assessing Care for Vulnerable
Elders (ACOVE) project to study 323 patients with OA. Although pharmacologic process-of-care measures were higher than nonpharmacologic
ones, quality scores (proportion of eligible patients receiving the process of care) were, overall, suboptimal. First-line
osteoarthritis therapy was acetaminophen for 60.4% of patients, and 79.5% of acetaminophen-treated patients were taking the
maximal dose before switching to NSAIDs. However, only about a quarter of new and established patients with osteoarthritis
received patient education, and exercise therapy for new and established osteoarthritis was prescribed to only 29% and 40.1%
of patients, respectively.
The development of QIs and assessment of adherence are necessary first steps in improving the care provided to patients.
In an era where standardized performance measures, such as the Health Employer Data Set, are used to compare health plans
and allow purchasers to make informed decisions about where to spend their healthcare dollars, it is likely that QIs of process
of care will be increasingly used to assess both health plan and physician performance and to determine whether they are conforming
to appropriate standards of care. Some health plans have already begun establishing incentive programs that could be linked
to reimbursement for physicians who consistently demonstrate excellent quality of care.
|
