What's new?

Our real‑life study of patients with psoriatic arthritis (PsA) showed that treatment with glucocorticosteroids and nonsteroidal anti‑inflammatory drugs was associated with an increased prevalence of cardiovascular comorbidities. A specific clinical phenotype of PsA, oligoarthritis, as well as the current use of glucocorticosteroids, were predictors of cardiovascular comorbidities. These findings suggest the need to limit the use of glucocorticosteroids and nonsteroidal anti‑inflammatory drugs in clinical practice, and to pay special attention when monitoring patients with the oligoarthritic form of PsA. Our results indicate that in a real‑world setting, hyperlipidemia remains an underdiagnosed condition in patients with PsA, once again highlighting the necessity of improving the screening and treatment of classical cardiovascular risk factors.

Introduction

Cardiovascular diseases (CVDs) remain a major global health concern, with a substantial impact on morbidity and mortality rates worldwide. Psoriatic arthritis (PsA), a chronic inflammatory condition characterized by heterogeneous manifestations affecting the skin, peripheral and axial joints, and entheses, has garnered increasing attention due to its association with elevated cardiovascular risk. This heightened risk includes a greater likelihood of experiencing major adverse cardiac events (MACEs) and a higher prevalence of classic cardiovascular risk factors and cardiovascular‑related comorbidities, such as arterial hypertension, obesity, diabetes, dyslipidemia, metabolic dysfunction–associated steatotic liver disease, and metabolic syndrome, as compared with the general population.¹ Although cardiovascular risk in PsA has been found to be elevated and contributes to excessive mortality,² the prevalence of cardiovascular risk factors and comorbidities in PsA as a distinct clinical entity has not been as thoroughly assessed as in rheumatoid arthritis or psoriasis, especially in the real‑world setting. To date, numerous studies conducted in the past decade focused on patients with psoriasis and did not account for the clinical phenotypes of PsA and the nuances of PsA treatment.^3-5 A recent meta‑analysis confirmed the increased incidence of CVDs in PsA patients, as compared with the general population over the last years.⁶ Despite these findings, the actual burden of cardiovascular comorbidities in real‑life populations of patients with PsA is still underestimated. It is also reflected by the fact that unlike in rheumatoid arthritis, there is no validated cardiovascular risk modifier or calculator. However, traditional cardiovascular risk factors emerge as crucial in predicting cardiovascular risk in PsA, considering they are sufficient to prognosticate it.⁷ A significant knowledge gap persists regarding the clinical characteristics of PsA patients, including disease phenotypes, such as polyarthritis, oligoarthritis, and axial disease, as well as the current usage of various treatment modalities, including nonsteroidal anti‑inflammatory drugs (NSAIDs), glucocorticosteroids (GCs), conventional synthetic disease‑modifying antirheumatic drugs (csDMARDs), and biological or targeted DMARDs (b/tsDMARDs). Understanding the prevalence of cardiovascular risk factors and comorbidities among individuals with PsA, along with their interaction with the disease phenotypes and treatment modalities, is crucial for effective clinical management and risk stratification. This paper presents real‑world data from a single center, in order to elucidate the prevalence of cardiovascular risk factors and comorbidities in PsA in relation to the clinical phenotype and therapeutic interventions.

Patients and methods

Results

In Table 1, we present detailed characteristics of the study population, including the disease activity variables, PROMs, laboratory data, and comorbidity information. In the entire group of patients, 25.09% had arterial hypertension, 29.37% were obese, 15.36% had diagnosed dyslipidemia, and 7.89% had diabetes mellitus. However, TC was above 5 mmol/l in 86 patients (33.46%), and LDL‑C over 3 mmol/l was noted in 94 patients (36.58%). We did not find differences in the prevalence of cardiovascular risk factors and the frequency of cardiovascular and metabolic comorbidities between the patients divided according to their clinical phenotype, that is, pattern of joint involvement (Table 2). Furthermore, the prevalence of cardiovascular risk factors (Table 3) and cardiovascular comorbidities (Table 4) did not differ across the groups divided by the mode of therapy. The patients with any CVD had a longer median (IQR) disease duration (13 [7–21.25] vs 8 [4.25–14] years; P = 0.003), were older (56 [45.25–66.75] vs 45 [38–54] years; P <⁠0.001), had higher BMI (29.5 [26.05–33.35] vs 26.25 [23.36–29] kg/m²; P = 0.004), and serum CRP levels (3 [1–8] vs 2 [1–5] mg/l; P = 0.004) than the patients without cardiovascular comorbidities. They also had higher disease activity, as expressed by DAS28 (3.5 [2.4–4.9] vs 2.6 [1.75–3.8]; P <⁠0.001), DAPSA (14.65 [6.25–28.15] vs 11.25 [4.45–21.43]; P = 0.04), and a higher SJC (1 [0–4] vs 0 [0–2.75]; P = 0.04) than the patients with no cardiovascular comorbidities. There were no differences in the measurements of axial disease activity (Table 5). Similar differences in these parameters were observed in the patients with hypertension vs those without hypertension. Additionally, among the patients with hypertension, the oligoarthritic phenotype of PsA was more frequent (38.81% vs 22%; P = 0.01; Table 5).

The patients on GC therapy were older than the rest of the group (P = 0.045; Table 5). Moreover, the individuals with diagnosed CVD were more often currently treated with GCs (17.95% vs 4.76%; P = 0.001) and had been exposed to them in the past (22.64% vs 13.23%; P = 0.02). The same was true for the patients with hypertension: 17.91% of those with ongoing GC treatment were diagnosed with hypertension, as compared with 5.5% of those not currently on GC therapy (P = 0.004; Table 5).

Among the patients on ongoing NSAID therapy, 83.33% had HF, as compared with 35.25% of participants without HF who were currently taking NSAIDs (P = 0.03; Table 5). Furthermore, 5.31% of the patients currently taking NSAIDs and 0.58% of those not taking these drugs had heart disease (P = 0.01). The odds ratio (OR) of heart disease in the patients with PsA who were taking NSAIDs was 9.72 (95% CI, 2.24–32.92). Additionally, 23.81% of the individuals with diabetes were on current GC therapy, as compared with 7.32% of those without diabetes who were on ongoing steroid treatment (P = 0.03; Table 5). A total of 60.87% of the patients on GC therapy and 26.22% of those not treated with GCs had at least 1 cardiovascular condition (OR, 4.375; 95% CI, 1.81–10.6; P <⁠0.001). Additionally, 91.67% of the patients on current GC treatment and 22.54% of participants not treated with GSs had hypertension (OR, 3.26; 95% CI, 1.42–11.9; P = 0.02). While 1.64% of the patients not treated with GCs had cardiovascular heart disease, this condition was present in 8.69% of the participants on ongoing GC therapy (OR, 5.71; 95% CI, 2.68–12.21; P = 0.03). A total of 6.56% of patients with PsA not on GCs and 27.78% of those treated with GCs had diabetes (OR, 3.96; 95% CI, 1.3–12.02; P = 0.04). We did not detect an additive effect of taking both NSAIDs and GCs on the likelihood of having any cardiovascular condition (OR, 3.07; 95% CI, 0.91–10.36; P = 0.12). The patients who ever smoked had twice the odds of having CVD than those who never smoked (OR, 2.073; 95% CI, 1.27–3.37; P = 0.02).

A multivariable logistic regression model (Table 6) involving data on clinical phenotype and treatment found that the oligoarthritic phenotype and current GC use are predictors of having any CVD, with an accuracy of 72.7% and Nagelkerke R² value of 0.048.

Discussion

In our PsA cohort, arterial hypertension was the most prevalent comorbidity, affecting 25.09% of the patients, closely followed by obesity at 23.97%. Notably, only 14.61% of the cohort were current smokers, which is lower than the average prevalence of tobacco use in both Poland and Europe.¹² Although only 15.36% of the patients were clinically diagnosed with dyslipidemia, elevated levels of TC and LDL‑C were detected in 33.46% and 36.58% of the patients, respectively. Of note, data from the lipid profile were missing in 33% of the patients, suggesting that at least one‑third had not undergone lipid measurement within a 3‑year period. This observation does not necessarily contradict the 2016 guidelines for cardiovascular risk management, which recommend lipid assessment every 5 years.¹³ However, more consistent monitoring and diagnosis could potentially reveal dyslipidemia as a more prevalent condition. It was exemplified in a study¹⁴ of the feasibility of cardiovascular risk assessment in patients with rheumatoid arthritis, in which missing data on cholesterolemia affected 14% of patients, and modifiable but insufficiently managed cardiovascular risk factors were identified in 30%. This is supported by a real‑life study from Greece, where dyslipidemia was identified in 42% of PsA patients, making it the second most common comorbidity after hypertension.¹⁵ Our findings are corroborated by an observational cohort study on patients with early PsA, which also reported high rates of dyslipidemia, underscoring it as a frequent comorbidity in this patient population.¹⁶

No significant differences were observed in the prevalence of cardiovascular risk factors or the frequency of cardiovascular and metabolic comorbidities among patients categorized by clinical PsA phenotype. Previous research has established that patients with PsA, as compared with those with psoriasis alone, exhibit a higher incidence of comorbid conditions, particularly hypertension and liver diseases.¹⁷ Although comparisons between patients with oligoarthritis and polyarthritis showed similar clinical characteristics, the extent of cardiovascular risk factors and comorbidities has not been extensively explored.¹⁸ To our knowledge, the association between the pattern of joint distribution in PsA and its comorbidities has not been thoroughly investigated. However, findings from a German study suggest that patients with the oligoarthritic subtype experience a higher comorbidity burden and greater disease severity than those with polyarthritis.¹⁹ Our data indicate that oligoarthritis is more prevalent among patients with hypertension and that it is a weak predictor of cardiovascular comorbidity. This may be explained by the earlier initiation of DMARDs in polyarticular PsA, while in the oligoarticular type, NSAIDs may be used more frequently and for a longer duration. Thus, current evidence does not yet support the concept that a greater number of affected joints leads to an increased burden of the disease and comorbidities.

Furthermore, in this study, patients with any CVD exhibited a longer PsA duration, were older, and had higher BMI and serum CRP levels than those without cardiovascular comorbidities. These patients also demonstrated increased disease activity, as expressed by DAS28 and DAPSA, alongside a higher SJC. These observations align with our understanding of PsA, where systemic inflammation is as critical as metabolic disturbances, particularly obesity and its related complications, in the pathogenesis of CVDs. Additionally, the patients in this cohort diagnosed with any cardiovascular condition and those diagnosed with hypertension were more frequently treated with GCs, either currently or in the past. These findings are consistent with the concept that both systemic inflammation and classic cardiovascular risk factors are pivotal contributors to the increased cardiovascular morbidity in PsA. A recent study of patients with rheumatoid arthritis and PsA indicated that systemic inflammation and GC usage independently elevate the risk of MACEs in both types of inflammatory arthritis.²⁰ Given the observation that patients treated with GCs frequently exhibited hypertension and diabetes, concerns arise regarding the use of GCs in PsA. The exacerbation of traditional cardiovascular risk factors by such treatments contradicts the therapeutic objectives. This is especially significant in light of current data suggesting that these risk factors alone are sufficient predictors of cardiovascular risk in PsA.⁷ The negative impact of systemic GCs is acknowledged in PsA management guidelines, which not only strongly advise against their use²¹ but also point to prompt escalation of therapy to b/tsDMARDs if csDMARDs fail in patients with peripheral arthritis. In the case of other manifestations (axial, enthesitis, dactylitis, etc.) first‑line therapy with bDMARDs is recommended. Rapid development of new molecules has expanded the spectrum of bDMARDs from tumor necrosis factor inhibitors to interleukin (IL)-17 inhibitors (with recent advances in dual inhibition of IL‑17, such as bimekizumab) and IL‑23 inhibitors. The exclusion of systemic GCs, even as bridge therapy (which is permitted in rheumatoid arthritis), and limiting their use only to local injections has become a cornerstone of modern PsA treatment.²² Numerous factors contributed to the use of GCs in the group of our patients. Since this was a real‑life study of patients treated in an outpatient clinic, therapeutic decisions made by practitioners were not influenced in any way. Fast symptom relief, therapeutic inertia, systemic limitations in the availability of novel therapies, and outdated clinical knowledge may all have played a role in the overrepresentation of GC use. The study group also included patients who may have had difficulty discontinuing GCs due to long‑standing PsA and limited access to novel therapies in the past, or treatment‑resistant cases where multiple lines of therapy had failed, with patients reporting improvement only with GC treatment.

In our study, also NSAIDs were associated with cardiovascular comorbidities, specifically HF. Despite the relatively small number of patients with HF within this subgroup, there was a significant bidirectional association: a higher incidence of HF among patients treated with NSAIDs and a greater use of NSAIDs among patients diagnosed with HF. Previous research documented that NSAIDs may increase the risk of HF.²³ A study in 4 European countries showed the negative impact of NSAIDs on hospital admissions due to HF, noting that this effect varies with the type and dosage of NSAID administered.²⁴ The cardiovascular adverse effects of NSAIDs in patients with PsA have not been extensively explored to date. However, robust data from a large Korean real‑world cohort of patients with ankylosing spondylitis suggest that higher doses of NSAIDs are linked to an increased risk of congestive HF.²⁵ Given the cardiometabolic profile of patients with PsA, it would not be surprising if a similar association was observed in PsA on a larger scale. Such findings underscore the need for cautious use of NSAIDs in this patient population, considering their potential cardiovascular risk.

The primary limitation of this study is its cross‑sectional design, which constrains our ability to infer casual relationships. Although this limitation was partially mitigated by extracting data from a greater number of visits, it remains challenging to determine whether cardiovascular comorbidities existed before the PsA diagnosis or the introduction of therapies such as NSAIDs and GCs. Furthermore, the size of the cohort restricted a more detailed analysis of clinical phenotypes. Consequently, we confined our categorization to 4 subtypes despite recent research identifying up to 7 clinical clusters of patients with PsA.²⁶ Another limitation is the lack of reliable data on the severity of skin psoriasis, which is important as psoriasis itself, even without joint involvement, increases cardiovascular risk.²⁷ This limitation arises from the design of the study, where real‑world data were collected during time‑restricted routine rheumatology outpatient clinic appointments, which did not include a detailed assessment of skin disease severity in the manner that is usually provided by dermatologists. Future studies should investigate the presence and severity of skin lesions in patients with PsA in relation to cardiovascular risk and comorbidities, and aim to determine whether skin and joint involvement in PsA have a cumulative effect.

The strength of our study lies in its real‑life setting, which enhances generalizability of the findings to the broader population of patients with PsA. This approach effectively minimizes selection bias, providing insights into a wide clinical spectrum of patients. Moreover, by evaluating the impact of various treatment regimens on cardiovascular outcomes, this research contributes valuable information to clinical decision‑making. Particularly, it addresses the gaps in knowledge concerning cardiovascular risks and comorbidities in a subset of patients who are typically excluded from randomized controlled trials, thereby enriching the evidence base for managing this complex patient population.

By leveraging the real‑world data, including patient demographics, PsA phenotype characterization, cardiovascular risk factor profiles, and treatment regimens, this study improves understanding of the prevalence and patterns of CVDs in outpatient real‑world cohort of PsA. Given the significant influence of traditional cardiovascular risk factors on the development of cardiovascular comorbidities in patients with PsA, it is evident that the effective systemic treatment with novel therapies such as b/tsDMARDs alone is insufficient to reduce cardiovascular morbidity. Therefore, active screening for and intensive management of cardiovascular risk factors and comorbidities are imperative. Moreover, despite the unequivocally negative long‑term impact of GCs on patients with PsA, there remains a high demand for more innovative therapies, as GCs continue to be widely utilized in real‑life clinical practice.