Home Page - About the Arthritis Center -Hopkins Rheumatology - Myositis Center - Vasculitis Center - Scleroderma Center
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
  |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Cardiovascular Risk | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
These three abstracts each use large databases to investigate the known risk of accelerated cardiovascular disease encountered in persons with RA. These abstracts specifically deal with one aspect of cardiovascular disease in RA, acute myocardial infarction (AMI). OP0037 ACUTE MYOCARDIAL INFARCTIONS IN RHEUMATOID ARTHRITIS AND DIABETES: A TALE OF TWO DISEASES AND A CALL FOR ACTION
An increased awareness of the impact of diabetes on atherosclerosis in the last decade has led to improvements in outcomes for diabetics who present with acute myocardial infarctions (AMIs). Here, by utilizing a large database of inpatient hospitalizations, trends in case fatality rates over the last decade for persons with RA presenting with AMI are compared with those of diabetics. Methods: Cohorts of persons with RA and diabetes hospitalized for AMI from 1991-2001 were constructed using the Nationwide Inpatient Sample, a random coding sample of 85% of U.S. hospitalizations without regard to payor. Details about mortality and concomitant diagnoses were collected and compared between groups. Results: In both RA and diabetes, a linear increase in the annual number of hospitalizations for AMI was noted. However, the proportion of AMI hospitalizations to all hospitalizations increased in RA patients over the decade of study (increasing from 1.5% of hospitalizations in 1991 to 2.2% of hospitalizations in 2001) while the same proportion remained relatively constant in diabetics (around 4% of admissions). While trends in case fatality rates for AMI hospitalizations involving diabetics have steadily declined over the decade of study (11% in 1991 to 7.5% in 2001), case fatality rates for AMI hospitalizations involving persons with RA have remained relatively constant (9% in 1991 7.5% in 2001). Of note, in 2001 the case fatality rate for AMI hospitalizations in persons with RA equaled that for diabetics. In comparing the two groups, RA patients had a mean age at admission 1 year older than the mean age for diabetics. In addition, diabetics tended to have more concomitant hypertension, congestive heart failure, and anterior wall MIs than subjects in the RA cohort. RA patients tended to have more arrhythmias and a higher proportion were women. Conclusions: The steady declines in case fatality rates that have accompanied hospital admissions for AMI in diabetics over the last decade have not been accompanied by similar trends in persons with RA. Persons with RA admitted with AMI should be treated as high risk with the same aggressive treatment goals as diabetics. Editorial Comments: Though thought provoking and insightful, these results are more illustrative than truly comparative. Importantly, many unaccounted confounders other than RA may be partially, if not wholly, responsible for the effects seen. For example, it is well established that case fatality rates for women presenting with AMI are higher than in men. Since the proportion of women to men in RA is 3:1, and since the authors have not provided the gender breakdown for each group, it is difficult to say with any assurance that the differences between the groups observed are not due to gender differences alone. Also, it is possible that hospital coding practices for RA have changed over the last decade toward a greater awareness of comorbid illness and need for more complete coding for reimbursement. Lastly, these data cannot address if any differences in sudden death or longitudinal associated mortality from AMI exist between the groups. Regardless, the authors are to be commended for attempting to raise awareness of the consequences of RA disease activity on the heart and in attempting to relate the burden of atherosclerosis in RA to a comparable and better-understood risk group. OP0038 DMARD USE AND THE RISK OF ACUTE MYOCARDIAL INFARCTION IN RHEUMATOID ARTHRITIS
COX-2 inhibitors, glucocorticoids, and some DMARDs, particularly methotrexate, all have potential, via their individual mechanisms of action, to promote coronary atherosclerosis and/or thrombosis. However, their ability to modulate inflammation may compensate for or even surpass these mechanisms and result in a net anti-atherogenic effect in RA. Here, the authors examine DMARD use and risk of AMI in RA by utilizing two large insurance claims databases. METHODS: Using two U.S. insurance claims databases (PharMetrics and Protocare: n=26 million enrollees), subjects were identified with a diagnosis of RA who filled at least one prescription for a DMARD between Sept 1, 1998 and Dec 31, 2001. During this period, subjects were followed for inpatient hospitalization claims for AMI. Each identified AMI case was then compared to 10 randomly matched non-AMI controls from the cohort based on age, and date of enrollment into the cohort. Results: Of 41,885 DMARD treated RA subjects, 267 were identified as being hospitalized with AMI during the study period (case rate of 5.2 cases per 1000 patients per year). 2670 non-AMI controls were matched based on age and cohort entry date. The percentage of women in the case group (55%) was lower than the percentage of women in the control group (74%). Current DMARD users were at a significantly reduced risk of developing AMI than non-current users while current use of COX selective antagonists were significantly associated with an increased risk of AMI compared to non-current users of these agents:
Conclusions: The current use of any DMARD appears to have a protective effect against AMI compared to non-current DMARD use. Current COX-2 inhibitor use increases the risk of AMI compared to non-current use. The effect of current use of non-selective NSAIDS or glucocorticoids could not be determined from this data. Editorial Commnents: The advantage that large numbers of patients bring to analyses such as this abstract and abstract OP0037, in which large 3rd party databases are retrospectively analyzed, can be partially, if not entirely, counteracted by uncertainties and potential biases within vital information that has not been collected. Here, for instance, there are no collected data about baseline characteristics, disease activity/severity, or comorbid illness. It is unknown whether confounding by indication may be a significant bias. In this case, the question would be whether the patients most likely to develop AMI were less likely to be prescribed traditional DMARDs or biologics. In addition, since most RA patients currently using DMARDS frequently also are prescribed COX-2 inhibitors, how then is risk to be sorted out? Also, since all subjects included here had, at one time, been prescribed a DMARD what circumstances made them non-current users for the purpose of this analysis? All in all, these data support other similar investigations suggesting that DMARD use does not necessarily accelerate atherosclerosis in RA patients. Definitive proof requires a prospective, controlled study with an adequate number of patients to detect differences between groups. OP0039 PREDNISONE BUT NOT BIOLOGICS OR DMARDS IS ASSOCIATED WITH INCREASED RISK OF MYOCARDIAL INFARCTION IN PERSONS WITH RA
Methods: Patients with RA enrolled in the National Databank for Rheumatic Disease (n=12,142) were queried at 6-month intervals about occurrence of AMI in the previous six months. Detailed information about medications, disease activity/severity, and comorbid conditions is available on all enrollees. The association of RA treatments with traditional DMARDS, biologics, and prednisone was determined within the construct of two multivariate models adjusted for age, sex, ethnicity, smoking status, concomitant hypertension, concomitant diabetes, and history of previous MI. Model 1 tested whether the use of specific biologics or prednisone predicted AMI in the following 6 months. Model 2 tested whether AMI was predicted based on RA therapeutic modalities (no biologic or DMARD, DMARD only, biologic only, DMARD + biologic) received in the preceding 6 months. Results:
Model 1: Predictors of AMI in following 6 months
Model 2: Predictors of AMI in following 6 months treatment strategy compared to treatment with no biologic or DMARD
| DMARD only | 1.3 | 0.6 2.5 | 0.524 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Biologic + DMARD | 1.2 | 0.6 2.6 | 0.618 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Prednisone use | 1.7 | 1.1 2.6 | 0.021 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
The addition of low dose aspirin use to the models did not change the above relationships. The above relationships were obtained using Cox proportions regression modeling. The same results were obtained using GEE (general estimating equations).
Conclusions: Prednisone use is predictive of AMI in the six months after use. The use of individual biologics and methotrexate, or treatment strategies involving biologics or DMARDS in general were not predictive of AMI in the six months following use.
Editorial Comments: As opposed to the methodology of abstract OP0038 above, the authors here utilize a large prospective database in which the primary focus is the patient with RA. An important potential confounder here is whether the use of prednisone in this patient population is a uniform marker for flares of disease or overall increased disease activity (although this point could also apply in similar way to unaccountable patterns in the prescribing of biologics). Another important point, acknowledged by the authors, is that the effects of biologics or DMARDs may be slow acting and not necessarily manifest in the immediate six months after use. Because treatments and disease characteristics change over time, however, it is difficult to accurately determine the effects of interventions further in the past using this type of collection instrument and analysis technique. Nonetheless, this study is solidly constructed and compliments other investigations that, taken in the aggregate, support the notion that biologics and DMARDs used to dampen the inflammatory response in RA are beneficial in reducing the overall burden of cardiovascular disease, emphasizing the underlying principal that systemic inflammation is the mechanism common to both atherosclerosis and RA.
