Association between systemic antipsoriatic drugs and cardiovascular risk in patients with psoriasis with or without psoriatic arthritis: A nationwide cohort study

Authors


  • This study is based in part on data from the National Health Insurance Research Database, provided by the Bureau of National Health Insurance, Department of Health and managed by the National Health Research Institute, Taiwan. The interpretations and conclusions contained herein do not represent those of the Bureau of National Health Insurance, Department of Health, or the National Health Research Institute.

Abstract

Objective

Psoriasis is associated with ischemic heart disease (IHD). Results of prior studies have suggested that methotrexate (MTX) may improve vascular disease in patients with psoriasis and rheumatoid arthritis. The aim of this study was to assess the risk of new-onset IHDs in psoriasis patients, comparing those taking MTX with those taking other nonbiologic antipsoriatic drugs.

Methods

In this retrospective cohort study, we utilized claims data from the National Health Insurance Research Database of Taiwan to identify 179,200 subjects who had received a diagnosis of psoriasis, with or without psoriatic arthritis, on at least 3 visits. The risk of hospitalization for new-onset IHD was compared between psoriasis patients taking MTX monotherapy (n = 6,578; MTX case cohort) and psoriasis patients taking other nonbiologic antipsoriatic drugs (n = 5,471; reference control cohort) between January 1996 and December 2008. Additional adjustments were made for cardiovascular risk factors, number of hospital visits, Charlson comorbidity score, and use of other antiinflammatory drugs.

Results

The incidence rates of IHD were 666 cases per 100,000 person-years in the MTX case cohort and 830 cases per 100,000 person-years in the reference control cohort (unadjusted P = 0.027). Increasing age, male sex, presence of hypertension, presence of diabetes, Charlson comorbidity score, and use of phototherapies were independent risk factors for hospitalization related to a new IHD in the study cohorts. However, the multivariate-adjusted hazard ratio for IHD-related hospitalizations among patients receiving MTX, in comparison with those receiving other nonbiologic antipsoriatic drugs, was 0.97 (95% confidence interval 0.79–1.19), after adjustment for age, sex, comorbidity score, number of hospital visits, and other antiinflammatory treatments for psoriasis.

Conclusion

In patients with psoriasis with or without coexisting psoriatic arthritis, the adjusted risk of hospitalization for an IHD among individuals receiving MTX was comparable with that among individuals receiving other nonbiologic antipsoriatic drugs.

Psoriatic disease, comprising a diagnosis of psoriasis that can include psoriatic arthritis, is a common chronic inflammatory disease affecting ∼0.2–3% of the general population (1–3). Many, but not all, of the relevant epidemiologic studies have consistently shown that psoriasis patients are at increased risk of developing multiple cardiovascular risk factors, metabolic syndromes, and ischemic cardiovascular diseases (1, 4–13). There is a 50% increase in deaths among psoriasis patients, mainly attributable to cardiovascular events, especially in those with severe psoriatic arthritis (13–15) and those with a younger age at disease onset (16). Although the exact mechanism of the relationship between psoriasis and cardiovascular comorbidities is complex, inflammation appears to be a central component in the initiation and propagation of atherosclerosis and atherothrombotic events (17, 18) that result in ischemic cardiovascular diseases.

Current treatment strategies for moderate-to-severe psoriasis include combination therapy with several systemic immune modulators, an approach that has been suggested to improve efficacy and decrease individual drug toxicity (19). Methotrexate (MTX) is a folate antagonist that inhibits dihydrofolate reductase activity and is an antiinflammatory drug that is widely used for the long-term management of moderate-to-severe psoriasis (19). MTX has 2 properties that have opposing effects on vascular diseases, each of which may have significant clinical implications after long-term use.

Long-term MTX therapy may promote hyperhomocysteinemia (20), and when MTX is administered in patients with a preexisting elevation in the serum levels of homocysteine (21), the risk of ischemic cardiovascular diseases may be increased. However, MTX also exerts antiinflammatory properties, which may decrease the risk of ischemic cardiovascular diseases. Some large observational studies have found that the use of MTX is associated with a reduced risk of vascular diseases in patients with psoriasis and rheumatoid arthritis (22, 23). The individual effects of MTX on the risk of a cardiovascular event in patients with psoriasis, in comparison with the effects of other systemic antipsoriatic immune modulators, have not been fully investigated.

Our aim in the present study was to investigate the risk of newly developed ischemic heart diseases (IHDs) in patients with psoriasis, with or without a diagnosis of psoriatic arthritis, who were taking MTX monotherapy (MTX case cohort), in comparison with psoriasis patients taking other nonbiologic antipsoriatic drugs (reference control cohort), using claims data from the National Health Insurance Research Database (NHIRD) of Taiwan. The effects of age at diagnosis, sex, coexisting cardiovascular risk factors, and other antiinflammatory treatments for psoriasis on the development of IHD were also evaluated.

PATIENTS AND METHODS

Data sources.

This retrospective nationwide cohort study was based on data from the NHIRD, released by the Taiwan National Health Research Institute. Taiwan began its national health insurance program in 1995 to finance health care for all of its residents. There are currently more than 25 million enrollees in the program, representing ∼99% of the entire population of Taiwan.

The NHIRD contains registration files and original claims data on care reimbursements for all enrollees, making it one of the largest and most complete sets of data on nationwide population-based health care services in the world (1, 24–30). In this database, the diagnostic codes are in the format of the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM). Patients were diagnosed by board-certified physicians in the corresponding specialties. The accuracy of diagnosis of major diseases in the NHIRD, such as diabetes mellitus, hypertension, and stroke, has been validated (30–32), and this database has been used extensively in many epidemiologic studies in Taiwan (1, 24–30). Personal information, including body weight, height, family history, laboratory findings, lifestyle, and habits such as smoking and alcohol use, was not available from the NHIRD.

Study sample.

We identified a total of 179,200 patients who had received a diagnosis of psoriasis with or without coexisting psoriatic arthritis (ICD-9-CM codes 696.0, 696.1, and 696.8) at least 3 times over the period of observation, by a board-certified physician in a dermatology or rheumatology department in Taiwan. Among these patients, 18,437 were treated with systemic antipsoriatic drugs, including MTX, oral retinoids, cyclosporine, azathioprine, and mycophenolate mofetil. Since biologic agents were not approved for treatment of psoriasis and psoriatic arthritis in Taiwan until late 2009, patients receiving biologic therapy were not included in this study. We also excluded patients who had been diagnosed with an IHD, including acute myocardial infarction (AMI) (ICD-9-CM codes 410 and 411), angina, coronary atherosclerosis, and other ischemic cardiovascular diseases (ICD-9-CM codes 413 and 414), before their first entry into one of the study cohorts. Patients were entered into a study cohort at the first prescription of MTX (for the MTX case cohort) or first prescription of other nonbiologic antipsoriatic drugs (for the reference control cohort), after the presence of psoriasis (with or without psoriatic arthritis) was established.

The MTX case cohort comprised 6,578 psoriasis patients who were started on a systemic antipsoriatic drug regimen of MTX; none of these patients had received any other systemic antipsoriatic drugs. Patients enrolled in the MTX case cohort could not be simultaneously receiving other systemic antipsoriatic drugs. The reference control cohort consisted of 5,471 psoriasis patients who were receiving other nonbiologic antipsoriatic drugs, including oral retinoids, cyclosporine, azathioprine, and mycophenolate mofetil; none of these patients had ever been treated with MTX during the period of observation. There was no history of IHDs in any of the patients in either study cohort. All sampled individuals were followed up from January 1, 1996 to December 31, 2008.

Consultation with the Institutional Review Board of Taichung Veterans General Hospital confirmed that this study was exempt from full review by the Institutional Review Board, as the NHIRD consists of de-identified secondary data released to the public for research purposes.

Main outcome measurements.

To study the risk of new development of an IHD in psoriasis, patients were classified as having a new-onset IHD, defined as AMI (ICD-9-CM codes 410 and 411), angina pectoris, coronary atherosclerosis, or other ischemic cardiovascular diseases (ICD-9-CM codes 413 and 414), if they had been hospitalized for one of these conditions during the observation period after the date of enrollment.

Covariate assessment.

Risk factors for IHDs were identified in all of the patients. These included hypertension (ICD-9-CM codes 401–405), diabetes mellitus (ICD-9-CM code 250), dyslipidemia (ICD-9-CM codes 272.0, 272.1, and 272.2), and ischemic stroke (ICD-9-CM codes 433 and 434). These diagnoses must have been made in the patient at least 3 times before and during the followup period.

Additional covariate factors considered were a diagnosis of psoriatic arthritis, use of other antiinflammatory treatments, number of hospital visits, and the comorbidity score based on the Charlson index (33). Nonsteroidal antiinflammatory drugs (NSAIDs), selective cyclooxygenase 2 (COX-2) inhibitors, and corticosteroids are commonly used to treat psoriasis and psoriatic arthritis, and all have been reported to increase the risk of ischemic cardiovascular diseases (34–37). In addition to these antiinflammatory treatments, phototherapies used to treat the extensive skin lesions in psoriasis, including ultraviolet B (UVB) and psoralen plus ultraviolet A (PUVA), were also considered as covariate factors.

Statistical analysis.

Demographic data were compared between the study cohorts. Followup for each subject began at the date of first prescription of MTX in the case cohort or the date of first prescription of other nonbiologic immune modulators in the reference control cohort, and ended at the date of censorship, i.e., the date of diagnosis of outcome, death, or transfer out, or the end of the followup period, and was measured in numbers of years. We used time-to-event analysis (Kaplan-Meier method) to estimate the cumulative incidences of the outcomes of interest after the patients had received their treatment. Observations of subjects without a particular outcome were censored at the time of their last recorded followup. We tested differences in the full time-to-event distributions between the 2 treatment groups using the log rank test.

To determine the independent risk factors for IHDs, multivariate analyses and stratified analyses using hazard ratios (HRs) were carried out with Cox proportional hazards models. The assumption of proportional hazards was confirmed by plotting the graph of the survival function versus the survival time and the graph of the log, log(survival), versus the log of survival time. Variables included age, sex, coexisting cardiovascular risk factors (hypertension, diabetes mellitus, hyperlipidemia, and ischemic stroke), number of hospital visits, comorbidity score using the Charlson Index, and adjunct treatments, including NSAIDs, selective COX-2 inhibitors, and systemic corticosteroids, as well as phototherapies, including UVB and PUVA. Assessment of goodness-of-fit of the models with the step-down method was carried out to analyze the independent risk factors.

Sensitivity analyses included 1) restricting the outcome of interest to AMI only; 2) restricting the outcome of interest to IHDs other than AMI; 3) focusing on individuals without a history of cardiovascular risk factors; and 4) focusing on individuals without a history of other autoimmune diseases, such as rheumatoid arthritis, systemic lupus erythematosus, Sjögren's syndrome, or ulcerative colitis. These sensitivity analyses were conducted with the purpose of examining whether the main findings were robust to different assumptions.

All analyses were performed using SAS software (version 9.1; SAS Institute).

RESULTS

Study population, IHD event rate, and survival analysis.

The distributions of age and sex in the MTX case cohort and reference control cohort are presented in Table 1. There were more female patients and patients were slightly younger in the MTX case cohort compared to the reference control cohort. The median followup time was 3.4 years (mean ± SD 4.0 ± 2.7 years) in the MTX case cohort and 4.9 years (mean ± SD 5.0 ± 2.9 years) in the reference control cohort. Among the 5,471 patients in the reference control cohort, 5,028 patients (91.9%) had received systemic oral retinoids, while 362 patients (6.62%) had received cyclosporine, 277 (5.06%) had received azathioprine, and 36 (0.66%) had received mycophenolate mofetil.

Table 1. Characteristics of the patients with psoriasis, with or without psoriatic arthritis, in the methotrexate (MTX)–treated case cohort compared to the reference control cohort*
CharacteristicReference control cohort (n = 5,471)MTX case cohort (n = 6,578)P
  • *

    Except where indicated otherwise, P values were derived using chi-square tests and Fisher's exact tests. IHD = ischemic heart disease; NSAIDs = nonsteroidal antiinflammatory drugs; COX-2 = cyclooxygenase 2.

  • Derived using t-tests.

  • Derived using log rank t-tests, in unadjusted analyses.

Age, median (mean ± SD) years45.6 (46.1 ± 17.2)43.8 (44.6 ± 15.7)<0.001
Male, no. (%)3,960 (72.4)4,152 (63.1)<0.001
Followup   
 Median (mean ± SD) years4.9 (5.0 ± 2.9)3.4 (4.0 ± 2.7)<0.001
 Person-years27,23226,113 
IHD incidence rate, no. of events (no. of events per 100,000 person-years)226 (830)174 (666)0.03
Psoriatic arthritis, no. (%)180 (3.3)1,003 (15.2)<0.001
No. of hospital visits, mean ± SD245.1 ± 173.9243.5 ± 171.60.63
Charlson comorbidity score, no. (%)  
 02,388 (43.6)2,901 (44.1) 
 1–103,040 (55.6)3,633 (55.2)0.66
 ≥1143 (0.8)44 (0.7)0.43
Antiinflammatory drugs, no. (%)  
 NSAIDs4,969 (90.8)6,094 (92.6)<0.001
 COX-2 inhibitors630 (11.5)1,442 (21.9)<0.001
 Systemic corticosteroids4,056 (74.1)5,144 (78.2)<0.001
Associated diseases, no. (%)  
 Hypertension1,648 (30.1)1,704 (25.9)<0.001
 Diabetes1,002 (18.3)980 (14.9)<0.001
 Dyslipidemia545 (10.0)653 (9.9)0.96
 Ischemic stroke298 (5.4)227 (3.5)<0.001
Reasons for end of study, no. (%)  
 Onset of outcomes226 (4.1)174 (2.6)<0.01
 End of observation without outcomes4,721 (86.3)6,023 (91.6)<0.001
 Death or transfer out524 (9.6)381 (5.8)<0.001

In the MTX case cohort, a total of 174 IHDs occurred, resulting in an incidence rate of 666 IHDs per 100,000 person-years. In the reference control cohort, 226 IHD events occurred, corresponding to an incidence rate of 830 IHDs per 100,000 person-years. Thus, the IHD incidence rate was slightly lower in the MTX case cohort than in the reference control cohort (P = 0.027, in unadjusted analyses) (Table 1).

These 2 study cohorts were similar with respect to observation time, comorbidity score, and number of hospital visits. However, they differed in several ways. Among patients receiving MTX, there were more cases of psoriatic arthritis and more patients receiving COX-2 inhibitors, when compared to the reference control cohort. Patients in the MTX case cohort had fewer coexisting cardiovascular risk factors, including hypertension, diabetes, or ischemic stroke, when compared to patients in the reference control cohort (Table 1).

Independent risk factors for ischemic cardiovascular diseases in psoriasis.

We next conducted a multivariate analysis to determine the independent risk factors for IHD-related hospitalizations in both groups. In adjusted Cox proportional hazards models including each demographic factor and multiple covariate factors, we found that the demographic factors of increasing age per decade and male sex, the presence of hypertension or diabetes mellitus, and a higher Charlson comorbidity score were each significantly associated with an increased risk of IHDs in the study cohorts. The risk of an IHD-related hospitalization was similar between the MTX case cohort and reference control cohort in every analysis (Table 2).

Table 2. Adjusted Cox proportional hazards models for the risk of ischemic cardiovascular disease in the methotrexate (MTX)–treated case cohort, in comparison with reference control patients treated with other nonbiologic antipsoriatic drugs*
CovariateAdjusted HR (95% CI)
  • *

    Other nonbiologic antipsoriatic drugs included acitretin, cyclosporine, azathioprine, and mycophenolate mofetil. Model 1 was adjusted for the use of MTX or other drugs, age, and sex. Model 2 was adjusted for the use of MTX or other drugs, age, and sex plus traditional cardiovascular risk factors (hypertension, diabetes, and dyslipidemia) and ischemic stroke. Model 3 was adjusted for the use of MTX or other drugs, age, and sex plus the Charlson comorbidity score (excluding the score for myocardial infarction), number of hospital visits, coexisting psoriatic arthritis, and prior use of other antiinflammatory drugs, including nonsteroidal antiinflammatory drugs, selective cyclooxygenase 2 inhibitors, and systemic corticosteroids. HR = hazard ratio; 95% CI = 95% confidence interval.

  • P < 0.001.

  • P < 0.05.

Model 1 
 MTX vs. reference controls0.91 (0.75–1.11)
 Age per decade1.88 (1.76–2.00)
 Male sex1.61 (1.27–2.05)
Model 2 
 MTX vs. reference controls0.94 (0.77–1.14)
 Age per decade1.69 (1.57–1.81)
 Male sex1.68 (1.32–2.13)
 Hypertension1.83 (1.45–2.31)
 Diabetes1.79 (1.45–2.22)
 Dyslipidemia0.82 (0.62–1.07)
Model 3 
 MTX vs. reference controls0.97 (0.79–1.19)
 Age per decade1.93 (1.80–2.07)
 Male sex1.35 (1.06–1.73)
 Charlson comorbidity score1.12 (1.08–1.16)
 Psoriatic arthritis1.00 (0.68–1.48)
 Use of phototherapy1.28 (1.04–1.56)

The multivariate analysis also showed that patients who had ever received phototherapy (UVB and PUVA) for extensive skin involvement in psoriasis had an increased risk of developing an IHD (adjusted HR 1.28, 95% confidence interval 1.04–1.56). However, coexisting psoriatic arthritis in patients with psoriasis was not associated with an increased risk of IHD (Table 2).

Stratified analysis.

We conducted stratified analyses to determine the effects of MTX on the risk of an ischemic cardiovascular event in different subsets of psoriasis patients. Initiation of treatment with MTX was not associated with a statistically significant elevation or reduction in the risk of IHD when compared to that in patients initiating other nonbiologic antipsoriatic drugs, in any of these stratified analyses. The risk of IHD in patients who had received long-term MTX therapy or whose use of MTX was increased in frequency in every observation year was not different than that in patients who had received other nonbiologic antipsoriatic drugs (Table 3).

Table 3. Stratified analyses of cardiovascular risk among the patients in the methotrexate-treated cohort, in comparison with the reference control cohort*
 No. of patientsAdjusted HR (95% CI)
  • *

    Hazard ratios (HRs) with 95% confidence intervals (95% CIs) were calculated with adjustment for multiple variables, including age, sex, coexisting psoriatic arthritis, prior use of different antiinflammatory drugs, prior use of phototherapies, Charlson comorbidity score, and number of hospital visits. None of the HRs were significant. PUVA = psoralen plus ultraviolet A therapy; UVB = ultraviolet B therapy; NSAIDs = nonsteroidal antiinflammatory drugs; COX-2 = cyclooxygenase 2.

Age  
 0–39 years4,2681.34 (0.52–3.44)
 40–59 years4,4650.87 (0.62–1.20)
 ≥60 years2,1080.98 (0.75–1.29)
Sex  
 Female3,5260.92 (0.60–1.43)
 Male7,3150.98 (0.78–1.23)
Psoriatic arthritis  
 Yes1,1031.14 (0.46–2.85)
 No9,7380.96 (0.78–1.18)
Phototherapy (PUVA, UVB)  
 Yes5,1290.95 (0.73–1.25)
 No5,7121.00 (0.73–1.36)
NSAIDs  
 Yes9,9821.00 (0.81–1.24)
 No8590.59 (0.25–1.37)
COX-2 inhibitors  
 Yes8,9441.18 (0.75–1.85)
 No1,8970.91 (0.73–1.15)
Systemic corticosteroids  
 Yes8,3481.04 (0.83–1.29)
 No2,4930.64 (0.36–1.14)
Antipsoriatic drugs
 Duration of use  
  <90 days5,4380.95 (0.73–1.24)
  90–180 days1,6970.78 (0.45–1.33)
  >180 days3,7061.01 (0.68–1.51)
 Frequency of use per observation year  
  <1 month6,2271.09 (0.82–1.45)
  1–3 months2,5890.71 (0.42–1.21)
  >3 months3,2330.87 (0.60–1.25)

Sensitivity analyses.

Results of the sensitivity analyses suggested that the primary findings were robust (Table 4). In all of these analyses based on different assumptions (as shown in Table 4), the use of MTX was not associated with a statistically significant reduction in the risk of new-onset IHD when compared to that in patients receiving other nonbiologic antipsoriatic drugs.

Table 4. Sensitivity analyses of the risk of ischemic heart diseases (IHDs) in the methotrexate (MTX)–treated cohort compared with the reference control cohort*
CovariateReference control cohortMTX case cohort 
No. of eventsTotal no. of patientsNo. of eventsTotal no. of patientsAdjusted HR (95% CI)
  • *

    Hazard ratios (HRs) with 95% confidence intervals (95% CIs) were calculated with adjustment for age, sex, Charlson comorbidity score, number of hospital visits, coexisting psoriatic arthritis, use of phototherapies, use of nonsteroidal antiinflammatory drugs, and use of selective cyclooxygenase 2 inhibitors and systemic corticosteroids. AMI = acute myocardial infarction; RA = rheumatoid arthritis; SLE = systemic lupus erythematosus; SS = Sjögren's syndrome; UC = ulcerative colitis.

  • History of cardiovascular risk factors (occurring before the start of antipsoriatic drug therapy) included hypertension, diabetes, dyslipidemia, and ischemic stroke.

Model 32265,4711746,5780.97 (0.79–1.19)
Model 3 restricted to outcome of AMI695,471616,5781.16 (0.82–1.66)
Model 3 restricted to outcome of other IHDs excluding AMI1895,4711446,5780.95 (0.76–1.19)
Model 3 excluding history of cardiovascular risk factors865,1141044,2791.10 (0.82–1.47)
Model 3 excluding history of RA, SLE, SS, or UC2255,4091626,3470.94 (0.77–1.16)

DISCUSSION

This is the first large-scale study to use nationwide–based data to examine the relationship between MTX and the risk of IHDs among psoriasis patients in an Asian population. Both study cohorts included psoriasis patients who were treated with systemic antipsoriatic drugs, which implied that these patients had moderate-to-severe psoriasis (with or without psoriatic arthritis).

Several independent risk factors for IHDs were identified in our study subjects, such as increasing age, male sex, traditional cardiovascular risk factors, and use of phototherapy (as a proxy for extent of skin involvement). Initiation of MTX did not reduce the risk of IHD-related hospitalization in patients with psoriasis with or without coexisting psoriatic arthritis, when compared to that in patients initiating other nonbiologic antipsoriatic drugs. Results from the adjusted stratified analyses and the sensitivity analyses, in which several different assumptions were tested, further supported these findings. Different measures of MTX exposures, such as duration of exposure and cumulative dose or frequency of exposure, did not affect the association between MTX and development of IHDs.

These results appear to be inconsistent with those obtained in prior studies regarding the effect of MTX on cardiovascular risk. The cardioprotective effect of MTX in rheumatoid arthritis has been widely explored. Long-term use of MTX in rheumatoid arthritis patients with preexisting cardiovascular diseases has been reported to be associated with a 3.4 times higher risk of death during followup (23). A nested case–control study found that use of oral corticosteroids and cytotoxic immunosuppressants, except for biologic drugs, was associated with an elevated cardiovascular risk when compared with use of MTX monotherapy in rheumatoid arthritis patients (38). The protective properties of MTX and other disease-modifying antirheumatic drugs (DMARDs) in rheumatoid arthritis patients were mostly mediated through reductions in the levels of the acute-phase response marker C-reactive protein and in the extent of insulin resistance (39).

Fewer investigations regarding the cardiovascular effect of MTX on psoriasis have been conducted. Patients with moderate-to-severe psoriasis usually benefit from the antiinflammatory properties of these systemic therapies. However, these antipsoriatic drugs have diverse side effects, which may lead to toxic effects on the organs and neutralization of the antiinflammatory benefits of these treatments. One case–control study showed a protective effect of MTX against vascular diseases in psoriasis (22). A population-based cohort study from Taiwan recently demonstrated that the risk of IHD among psoriasis patients was comparable between those taking systemic treatments and those who did not receive such treatments, after adjustment for demographic and cardiovascular factors (40). However, most of these studies failed to separate the patients according to those taking MTX alone and those simultaneously taking other DMARDs for psoriasis, which may result in mixed effects due to the potentially differing outcomes from combination therapy (22, 23, 40).

Our results suggested that patients taking MTX monotherapy had no more risk of developing ischemic cardiovascular diseases than did patients taking other immune modulators. More than 91% of the patients in the reference control cohort had received systemic retinoids. As a main therapy for psoriasis, treatment with oral retinoids normalizes the proliferation, differentiation, and cornification of epidermal cells, by interfering with the expression of epidermal growth factor genes (41, 42). There is also evidence that oral retinoid drugs have immunomodulatory properties, in that they can inhibit dermal microvascular endothelial cells and neutrophil migration (43–45). One recent case–control study found no significant association between the use of isotretinoin and risk of stroke, AMI, and thrombophlebitis (46).

The impact of cyclosporine on cardiovascular risk has also been investigated. A negative correlation with cumulative doses of cyclosporine and carotid intima media thickness among patients with systemic lupus erythematosus has been reported (47). These results suggest that the use of cyclosporine in treating lupus nephritis may have protective effects against premature atherosclerosis. Taken together, the results indicate that immunosuppression may underlie the potential beneficial role of all of these systemic treatments on cardiovascular diseases.

Since patients with moderate-to-severe psoriasis always require the use of systemic antiinflammatory treatments, it is challenging to select a cohort of unexposed patients as a referent, i.e., containing patients who have severe disease and yet have never received systemic antipsoriatic drugs. Further studies are needed to address these associations with the use of the proper person-time approach, allowing us to compare the exposed cohort (MTX users) to the nonexposed cohort (nonusers).

The strengths of the present study include the utilization of the nationwide NHIRD, which contains detailed pharmacy claims from each study subject and is widely accepted for epidemiologic studies in Taiwan (1, 24–29, 40). Furthermore, we utilized an incident user design in this study, in which exposure time began with the start of new systemic antipsoriatic drugs in both cohorts. This design also allows for a more valid comparison of treatments than would a mix of ongoing users with new users in the same analysis (48). Moreover, an active comparator cohort, as was used in the present study, allowed the analyses to be restricted to patients with active disease of similar severity. In addition, the choices of treatments in both cohorts were not identified hierarchically (i.e., there was no preference or guideline for one specific drug over another drug in patients with similar severity of disease). Finally, the period of exposure to systemic antipsoriatic medications was mutually exclusive in both cohorts; this would minimize the immortal person-time bias prior to treatment exposure, which may result in a downward trend toward underestimation of the risk rate ratios, as has been seen in many observational studies (49, 50).

There are several limitations to our present study. First, it is difficult to infer causation between drugs of interest (in this case, MTX) and the risk of IHD-related hospitalization on the basis of an observational study, without a random assignment of treatments. Confounding by indication may exist. In addition, patients in the MTX cohort may be different from the reference cohort in several, unmeasured ways. We did not have personal information on each patient, such as lifestyle, body mass index, laboratory parameters, smoking habits, alcohol use, and diet, all of which may contribute to the risk of IHD. To minimize the influence of these confounding factors, we used an active comparator group of patients as the reference cohort; however, this may not adequately adjust for unmeasured confounding. Patients in the reference cohort were older, there were more males, and more had cardiovascular risk factors, when compared to patients in the MTX cohort, which may have resulted in a higher risk of IHD in the reference cohort, before adjustment for these confounding factors (Table 1).

Disease severity was also not recorded in the database. For this, we took the use of phototherapy as a proxy for extensive skin involvement. Our results suggested that more extensive skin involvement was associated with an elevated IHD risk, independent of the presence of traditional cardiovascular risk factors, a diagnosis of psoriatic arthritis, or the use of systemic treatments.

Finally, coding errors in the disease classification are a possible limitation in the use of administrative databases. To minimize this bias, we enrolled only patients who had been given a diagnosis of psoriasis at least 3 times over the observation period, with validation by a dermatologist or rheumatologist (40). The numbers of patients with psoriasis with or without coexisting psoriatic arthritis, may therefore be underestimated.

To our knowledge, this is the first large-scale nationwide cohort study investigating the individual effect of MTX on the risk of IHD among psoriasis patients in an Asian population. We found that the use of MTX in patients with psoriasis with or without psoriatic arthritis was associated with a risk of IHD comparable to that in patients taking other nonbiologic antipsoriatic drugs. Further studies will be required to elucidate the pathogenetic relationship between MTX therapy and cardiovascular diseases.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Wu had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Y. -J. Chen, Chang, Shen, T. -T. Chen, Wang, C.-M. Chen, Wu.

Acquisition of data. Y. -J. Chen, Chang, Shen, T. -T. Chen, Wang, C.-M. Chen, Wu.

Analysis and interpretation of data. Y. -J. Chen, Chang, Shen, T.-T. Chen, Wang, Wu.

Ancillary